key: cord-006182-kck5e1ry
authors: nan
title: 17th Annual Meeting, Neurocritical Care Society, October 15–18, 2019, Vancouver, Canada
date: 2019-10-01
journal: Neurocrit Care
DOI: 10.1007/s12028-019-00857-7
sha: 
doc_id: 6182
cord_uid: kck5e1ry

nan

Aging is associated with greater stroke risk and diminished stroke recovery. While the effect of aging on stroke recovery is well defined, the influence of aging on neuronal network activity and its correlation with stroke recovery is poorly understood. To study this, we performed serial whole-cortex imaging of spontaneous and evoked neuronal activity before and after stroke in young and aged mice and correlated those findings to behavioral outcomes.

Young (2-3m, n=13) and aged (18m, n=13) Thy1-GCaMP6 mice, underwent behavioral assessment and imaging before and 1, 7, 12, and 17 weeks after infarct. Infarcts were induced via exposure of somatosensory cortex to a laser (25mW,10m) after injection of the photosensitive dye rose bengal. Imaging followed placement of Plexiglas windows and consisted of awake (15m) and anesthetized sessions (10m) with GCaMP excitation via flashing of a 454nm LED and acquisition via an EMCCD camera (16.8hz framerate). Somatosensory activation was via forepaw shock (1mA at 3Hz for 5s x 10 blocks). Behavioral response was assessed by quantifying forepaw use during cylinder exploration (5m).

Aged and young mice exhibited similar baseline contralateral forepaw use (Aged 31.25±2.88%; Young: 26.26±2.41%) and evoked somatosensory cort 2.3±0.16 x 10-3). Whole-brain GCaMP6 flourescence power in delta (1-4Hz) and infraslow (0.02-0.1Hz) ranges was significantly (p<0.05) lower at baseline in aged mice. After stroke, aged mice developed greater long-term dependence on the unaffected limb (Wk17: Aged 57.12±5.77%; Young: 37.99±4.73%) -3). -stroke decrement in whole-brain GCaMP6 fluorescence power was observed in aged animals.

Stroke in aged mice is associated with a greater decrement in local network activation, global mechanisms underlying age-related differences.

Hypernatremia and hyperchloremia is common after moderate-severe traumatic brain injury (msTBI) from saline resuscitation, osmotherapy administration, fever with insensible losses, limited free water administration, and diabetes insipidus. In isolation, hypernatremia and hyperchloremia are independent predictors of mortality in critically-ill patients; but this association has not been studied in msTBI, or in combination as physiologically occurring in patients. We examined whether hypernatremia and hyperchloremia in combination are independent predictors of in-hospital mortality in msTBI patients.

We retrospectively analyzed prospectively collected data of consecutive msTBI patients enrolled in the OPTIMISM-study over a 7-year period. A semi-automated process provided all sodium and chloride values from the index hospitalization. Time-weighted-average(TWA)-sodium and chloride representing their "burden" over the entire hospitalization were calculated using a published formula. Univariate and multivariable logistic regression were applied adjusting for IMPACT-model-variables as validated predictors of msTBI mortality, osmotherapy, ICU length-of-stay and ventilatory days.

Of 458 patients analyzed, 202 (44%) died. Unadjusted mortality rates had a dose-response relationship with increasing sodium and chloride (34% for TWA-sodium 140-150mmol/L and 100% for TWA-sodium >160mmol/L; 42% for TWA-chloride 110-115mmol/L and 94% for TWA-chloride >125mmol/L; all p<0.0001). Separately, TWA-sodium (per 10mmol/L increase adjusted-OR 4.0;95%CI 2.1-7.5) and TWAchloride (per 10mmol/L increase adjusted-OR 3.9;95% CI 2.2-7.1) independently predicted mortality. In combination, however, TWA-chloride remained an independent predictor of in-hospital mortality (per 10mmol/L increase OR 2.9, 95% CI 1.1-7.8), while TWA-sodium did not (C-statistic 0.96; Hosmer-Lemeshow p<0.0001).

To our knowledge, this is the first study to show that when concomitantly adjusting for hyperchloremia and hypernatremia burden, only hyperchloremia is independently associated with early mortality in msTBI. While not proving cause-and-effect, this suggests that hyperchloremia, and not hypernatremia as previously reported, deserves further attention in msTBI. If validated, this may have treatment implications for msTBI patients in the acute care phase.

Hyperosmolar therapy, with hypertonic sodium chloride (NaCl) solution is often used in the treatment of cerebral edema and elevated intracranial pressure. Recent reports have demonstrated that in patients with subarachnoid hemorrhage (SAH) treated with hypertonic NaCl, hyperchloremia is associated with the development of acute kidney injury (AKI). We report a trial which compared two hypertonic solutions with different chloride content on the resultant serum chloride concentrations in SAH patients.

A low ChloridE hyperTonic solution for brain Edema (ACETatE), is a single center, double-blinded, double-dummy, pilot clinical trial comparing bolus dosing of 23.4% NaCl versus 16.4% NaCl/Na-Acetate for the treatment of cerebral edema in patients with SAH. Randomization occurred once patients who received hypertonic treatment for cerebral edema and/or elevated intracranial pressure (ICP) developed hyperchloremia (serum Cl-

group, and 17 to the Hypertonic NaCl/Na-Acetate one. The groups were well balanced in terms of severity of the SAH, age, gender and risk factors. Differerences between the serum chloride and sodium measurements, assessed from randomization to maximum during the ICU course, were comparable between the NaCl and NaCl/Acetate groups (Cl:3.3±6.1 vs. 1.6±3.2, p=0.36; Na:6.5±4.4 vs. 9.1±6.6, p=0.18, respectively). NaCl/Acetate had a more prominent effect on immediate post dose sodium (increase of 1.4±2.6 vs. 2.2±2.8,p<0.01). The rate of AKI was lower in the Na-Acetate group (53.3% in the NaCl group vs. 11.8% in the Na-Acetate group, p=0.01). Hyperchloremia preceded AKI in 61.5% of the cases; However, the time interval between hyperchloremia and AKI was only a median of 0.75 days ).

Intention to treatment analysis demonstrated that treatment with hypertonic NaCl/Na-Acetate hypertonic versus standard hypertonic NaCl solution for patients with mild hyperchloremia, resulted in less events of A -center trials are needed to corroborate these results.

Up to 5.3 million people in the United States are living with physical, cognitive, and psychological sequelae after TBI. Patients that sustain a moderate to severe TBI (msTBI) are heavily reliant on caregivers during their inpatient stay and for post-discharge care. There are limited data on how best to support caregivers in their role. The purpose of this study was to develop a checklist based on qualitative data that can be utilized by caregivers and clinicians to re-examine the particular needs of the caregiver at different periods in the acute, subacute, and chronic timeframe.

Patients with msTBI and their caregivers were recruited from two intensive care units (ICUs) in one institution to participate in semi-structured interviews at 72 hours, one month, three months, and six months post-injury. Transcripts of each interview were analyzed by two investigators who independently coded responses using a predetermined code list adapted from previously identified needs and concerns of other similar populations. Based on the particular coded segments, a checklist and a list of strategies were derived to address the needs and concerns of caregivers.

A total of 18 patient-caregiver dyads were enrolled from X-Y, with 61 interviews completed; 53 interviews with caregivers and 8 with patients. Caregiver interviews resulted in 41 unique codes that corresponded to varying caregivers' needs and concerns which were developed into a checklist and list of strategies.

The needs and concerns of msTBI caregivers should be assessed over time to provide the support necessary to assist in the care of msTBI survivors. Implementation of a checklist, as well as a list of strategies, can allow for tailored interventions that improve the transitions of msTBI survivors from the ICU to subacute/chronic care environments.

Malignant cerebral edema (MCE) develops in a subset of those with hemispheric strokes, precipitating neurological deterioration and death if decompressive hemicraniectomy (DHC) is not performed in a timely manner. However, prediction of which patients will develop MCE is imprecise based on baseline clinical and radiographic features imaging quantifies development of cerebral edema. We employ a recurrent neural network that learns from serial clinical and imaging data to enhance early prediction of MCE.

We identified 105 patients with hemispheric stroke who had NIHSS and CT scans performed at baseline automated algorithm; midline shift (MLS) was measured at the level of the septum pellucidum. We trained a recurrent neural network that incorporates sequential data and compared its performance to those of traditional models. We tried to maximize sensitivity for predicting MCE (DHC or death) while optimizing prediction of those not requiring DHC (negative predictive value, NPV).

Nine patients required DHC or died from MCE. A linear classifier incorporating age, baseline NIHSS, and serum glucose had high NPV (92%) but only 11% sensitivity for MCE. A probabilistic Gaussian mixed model (GMM) improved sensitivity to 44%. Incorporating 24-hour NIHSS into GMM improved prediction (sensitivity 76%, NPV 97%). The neural network was able to predict all cases of surgery and all of those not requiring surgery with 100% accuracy prediction.

Recurrent neural networks incorporating sequential clinical and imaging data from the first 24-hours after stroke may enhance our ability to predict which patients will need DHC. Our promising pilot evaluation of this approach study requires validation in larger external stroke cohorts.

Aneurysmal subarachnoid hemorrhage(SAH) survivors live with long term residual physical and cognitive disability. We studied whether neuromuscular electrical stimulation(NMES) and high protein supplementation(HPRO) in the first two weeks after SAH could preserve neuromotor and cognitive function as compared to standard of care(SOC) for nutrition and mobilization.

SAH subjects with a Hunt Hess(HH) grade>1, assigned to SOC or NMES+HPRO. NMES was delivered to bilateral quadriceps and gastrocnemius muscles daily during two 30 minute sessions along with HPRO(goal:1.8 g/kg/day) between post bleed day(PBD) 0 and 14. Primary endpoint was atrophy in the quadriceps muscle as measured by the percentage difference in the cross sectional area from baseline to PBD14 on CT scan. All subjects underwent serial assessments of physical(short performance physical battery,SPPB) cognitive(Montreal Cognitive Assessment Scale,MoCA) and global functional recovery(modified Rankin Scale,mRS) at PBD 14, 42, and 90.

Twenty-five subjects(SOC=13,NMES+HPRO=12) were enrolled between December 2017 and January 2019 with no between group differences in baseline characteristics(58 years old, 68% women, 50% HH>3). Median duration of interventions was 12 days(range 9 -14) completing 98% of NMES sessions and 83% of goal protein intake. No difference in caloric intake between groups, but HPRO+NMES group received more protein(1.5 +/-0.5 g/kg/d v 0.9+/-0.4 g/kg/d,P<0.01). Muscle atrophy at PBD 14 was less in NMES+HPRO group(2.8+/-2.9% vs 6.7+/-3.2% ,P=0.01). On univariate analysis, higher atrophy was correlated with lower daily protein intake (r=-0.59,P=0.002); and worse 3 month MoCA (r=-0.39, P=0.06),SPPB(r=-0.31,P=0.1) and mRS(r=0.46,P=0.03). NMES+HPRO subjects performed better on SPPB(P=0.03), were observed to have a lower mRS(P=0.01) and obtained a higher MoCA(P=0.1) than SOC at PBD 90.

NMES+HPRO may reduce acute muscle wasting in lower extremities with a lasting benefit on recovery after SAH. To better understand whether NMES and/or HPRO are responsible for observed benefits, a larger, multicenter study is underway.

Increasing authorization rates for organ donation is the best way to grow the number of organs available for life-saving transplants. In order to improve our authorization rates and thereby provide more organs for life-saving transplants, our Organ Procurement Organization (OPO) partnered with donor hospitals to -led donation conversations and intensified the focus on a collaborative donation process.

ned in the the OPO during the authorization process by providing a timely notification of a potential donor and by work together on the timing of the donation discussion.

The overall authorization rate has improved from 49% in 2012 to 70% currently. During this time frame, --led conversations has been compelling and a significant factor in improving authorization rates. Equally impactful to improved authorization rates has been a 20% increase in cases involving a collaborative donation process (measured by timely referral and collaborative mention of donation).

Developing a strong partnership between an OPO and a donor hospital is paramount to a successful donation process. Critical factors such as timely referral notification and collaboration regarding the timing of the donation discussion can positively influence authorization outcomes. Moreover, we -led donation conversations will lead to further increases in authorization which results in more lives saved through donation.

Quantitative EEG analysis is one part of multimodal monitoring in the intensive care unit due to high temporal resolution and ease of deployment. Previously we have shown that dynamical properties of EEG signals can be used to differentiate focal vs. diffuse causes of coma (Kafashan et al., 2017) , and that the intrinsic reactivity of EEG signals -a measure of responsiveness of the EEG to endogenously rare events -correlates with GCS score (INRI, Khanmohammadi et al., 2018) . Here, we explore the possibility of localizing brain lesions using these dynamical features of EEG signals.

We collected retrospective data from 20 comatose patients (GCS<8) defined to have a focal injury. The patients underwent EEG recordings and imaging for routine purposes at Barnes-Jewish Hospital NNICU.

Index (INRI) , which consists of identifying intrinsic events, obtaining brain-state trajectories, and quantifying brain-state trajectories. We then used a neural network-based classifier to map the INRI to lesion location using supervised learning paradigms with cross-validation.

We used imaging to identify anatomical location of lesions and project them to a two-dimensional headmap. We trained a neural-network classifier to predict 2D lesion location from the INRI dynamics of each EEG channel. We then assessed the correlation between predicted location and actual location using a cross-validation protocol. Predicted locations significantly correlated with injury location (r>0.6) when compared to correlations with randomly selected patients (r~0). The results point to a systematic change in underlying neuronal-dynamics induced by brain lesions, that was captured through EEG dynamics and the concept of intrinsic reactivity.

Here we developed and evaluated a framework to localize brain injury through novel analysis of EEG dynamics. The results here, together with our previous work, suggest brain injuries can be detected and localized using EEG recordings.

To examine whether changes in Intracranial Pressure (ICP) waveform morphologies can be used as a biomarker for early detection of ventriculitis.

Of 1653 consecutive patients enrolled prospectively in a hemorrhage outcomes study from 2006 to 2018, 435 (26%) patients required external ventricular drainage (EVD). Only the culture-positive ventriculitis seen in 19 (4% of all EVDs) patients were included in current analysis. Based on our es per week, and additionally if infection was suspected. EVDs were left open for drainage, with ICP monitored hourly by clamping. Using wavelet analysis, we extracted uninterrupted segments of ICP waveforms. We extracted dominant-pulses from continuous high-resolution data using Morphological Clustering Analysis of ICP Pulse (MOCAIP). Then we applied hierarchical k-means clustering using dynamic time warping distance to obtain morphologically similar groupings. We applied a top-down approach to split the clusters further, which stops when the mean distance of the waveforms to the centroid is less than a pre-clusters and further-split clusters (when equipoise existed) were categorized for broad comparison by clinician consensus.

We extracted 275,911 dominant pulses from 459.9 hours of EVD data. 112,898 pulses (40.9%) occurred before positive culture, 41,300 pulses (15.0%) were during culture positivity, and 121,713 pulses (44.1%) occurred after clinical diagnosis was made. K-means identified 20 clusters, which were further grouped into meta-clusters: tri/biphasic (green), single-peak (yellow) and artifactual (red) waveforms. 61.8% of dominant pulses were tri/biphasic before ventriculitis, which reduced to 22.6% during and 28.4% after (p<0.0001). One day before the first positive cultures were collected, the distribution of meta-clusters changed to include more single-peak and artifactual ICP waveforms (p<0.0001).

The distribution of ICP waveform morphology changes significantly prior to the clinical diagnosis of ventriculitis, and may be a potential biomarker.

Inducing normothermia with temperature modulating devices (TMDs) is often associated with significant shivering. We tested the ability of a novel transnasal TMD to induce and maintain normothermia with minimal shivering in endotracheally intubated (ET) cerebrovascular patients.

Single center study utilizing CoolStat transnasal cooling device to achieve core temperature reduction by inducing an evaporative cooling energy exchange in the turbinates and upper airway thru a high flow of dehumidified air into the nasal cavity and out the mouth. Primary goal was the ability to induce normothermia(T<=37.5C) within 4 hours in ET patients with fever(T>=38.3C) refractory to acetaminophen. Continuous temperature measurements were obtained from tympanic and core(esophageal or bladder) temperature sensors. Safety assessments included continuous monitoring for hypertension, tachycardia, and raised ICP(when monitored). ENT evaluations monitored for any device related nasal mucosal injury. Shivering was assessed every 30 minutes using the Bedside Shivering Assessment Scale(BSAS). Duration of device use was limited to 8 hours, as regulated by the e care for temperature management.

Ten subjects(median age:54 years, BMI:32.5kg/m2,60%men) were enrolled with normothermia achieved in 90% of subjects. One subject did not achieve normothermia and was later refractory to other TMDs. Median baseline temperature was 38.5+/-0.1 C, with a reduction noted by 4 hours(38.5 +/-0.1 v. 37.3 +/-0.8, P<0.001) and sustained at 8 hours(38.5+/-0.1 v 37.2+/-0.7, P=0.001). Time to normothermia was 3.1+/-1.8 hours. The median BSAS was 0(range:0 -1) with only 4 episodes necessitating meperidine across 80 hours of study monitoring. No treatment was discontinued due to safety concerns. ENT evaluations noted no device related adverse findings.

Inducing normothermia with a novel transnasal TMD appears to be safe, feasible and not associated with significant shivering. A multicenter trial testing the ability to maintain normothermia for 24 hours is currently underway.

Traumatic coma is thought to be caused by disruption of the subcortical ascending arousal network (AAN). This hypothesis has not yet been tested because tools to map AAN connectivity in living humans have only recently become available. We implemented high angular resolution diffusion imaging (HARDI) on an MRI scanner in the intensive care unit to determine whether patients presenting with traumatic coma have disrupted AAN connectivity.

We performed high angular resolution diffusion imaging (HARDI) in 16 patients with acute severe traumatic brain injury who were comatose on admission and in 16 matched controls. We used probabilistic tractography to measure the connectivity probability (CP) of AAN axonal pathways linking the brainstem tegmentum to the hypothalamus, thalamus and basal forebrain. To assess the spatial specificity of CP differences between patients and controls, we also measured CP within four subcortical pathways outside the AAN.

Compared to controls, patients showed a reduction in AAN pathways connecting the brainstem tegmentum to a region of interest encompassing the hypothalamus, thalamus, and basal forebrain (Patients: median 0.08, IQR [0.04, 0.1] Controls: 0.1 [0.09, 0.12], P = 0.006). Examining each pathway individually, brainstem-hypothalamus and brainstem-thalamus CPs (Pc < .05), but not brainstemforebrain CP (Pc = 0.37), were significantly reduced in patients. Only one subcortical pathway outside the AAN showed reduced CP in patients.

We provide initial evidence for the reduced integrity of axonal pathways linking the brainstem tegmentum to the hypothalamus and thalamus in patients presenting with traumatic coma. Our findings support current conceptual models of coma as being caused by subcortical AAN injury. AAN connectivity mapping provides an opportunity to advance the study of human coma and consciousness.

Limited knowledge about the physiology underlying coma recovery has decreased clinicians' ability to identify patients likely to benefit from continued intensive therapy. Machine learning using quantitative EEG (QEEG) has shown potential to improve outcome prediction in cardiac arrest, but the relationship between QEEG trends and coma recovery had limited evaluation in large multicenter studies.

Seven hospitals contributed clinical and EEG data from comatose adult subjects with cardiac arrest who underwent continuous EEG and targeted temperature management. QEEG features evaluated included background frequency, burst-suppression ratio(>50%), epileptiform discharges, and entropy. We utilized random forests to predict good (CPC1-2) vs. poor (CPC3-5) outcome at 6-months. Model performance was evaluated using the AUC at 12h intervals up to 72h.

We analyzed 58,125 hours of EEG (+5TB) for 1,089 subjects (388 good outcomes). Unfavorable EEG features were common in subjects with good or poor outcomes (epileptiform discharges: 27%, 38% and burst-suppression: 25%, 56%, respectively). Epileptiform discharge frequency peaked after rewarming in subjects with good outcome (3 spikes/min at 48h), but continued increasing during cooling and rewarming for those with poor outcome (4-8 spikes/min from 24h-48h). Shannon entropy was always higher in subjects with good outcome. Burst-suppression strongly predicted outcome for all centers but during different times, while epileptiform discharges predicted outcomes in five centers, entropy in three, and alpha-background in only one. Outcome prediction was best with QEEG during cooling rather than after rewarming (AUC 0.83 vs. 0.77 at 24h and 48h, p<0.05). Maximal AUC at 24h for individual centers ranged from 0.54-0.86.

Early QEEG trends carry useful information for coma recovery prediction, but marked heterogeneity in QEEG trends across centers can limit performance and reproducibility of machine learning prognostication algorithms. Coexistence of favorable and unfavorable QEEG features in the same patient is common, suggesting that generalizable models for coma recovery prediction must leverage temporal trends.

Human consciousness depends on ascending projections from the brainstem. Brainstem lesion mapping studies have identified a coma-specific sub-region of the dorsolateral pontine tegmentum. However, loss of consciousness (LOC) can also occur following injury to cortical regions remote from the brainstem, a phenomenon that commonly occurs after penetrating head trauma but remains poorly understood.

Andexanet alfa has been shown to reduce anti-factor Xa activity however outcome studies are lacking. We compare the efficacy of four-facto -PCC) vs andexanet in patients with factor Xa inhibitor related bleeding.

Retrospective study was performed January 2017 to March 2019, including patients with factor Xa inhibitor related bleeding of whom 71 wer -PCC vs 50 treated with andexanet. Outcome was analyzed using Glasgow Outcome Scale (GOS) at discharge, presented as good (score 4-5) and poor (1-3); length of stay (LOS) and invariables, and T-test for continuous variables.

-PCC or andexanet were included in the study. Bleeding source --PCC; vs andexanet 17 cases, 34% of total -PCC n=20, andexanet n=11) and trauma (45-PCC -PCC group was 7.8d vs 6.7d in the andexanet group; ICU stay corresponded to 4.34 vs 4.29 days, respectively. Outcomes evaluated through GOS did not differ -PCC group vs 50% in andexanet group, -PCC group vs 29.4% on andexanet group, p=1). Unexpectedly, in-hospital mortality was higher on andexanet group -PCC group (9.9%); with a similar trend observed in the CNS subgroup.

-PCC as a factor Xa inhibitor related bleeding reversal agent was as effective as andexanet based on outcome scale, constituting an essential option for hemostatic control as cost differences can limit the use of andexanet.

The mechanism by which early administration of tranexamic acid (TXA) reduces mortality in traumatic brain injury (TBI) is poorly understood. In-vitro models suggest the glycocalyx is preserved with early TXA administration, indicating that TXA may inhibit glycocalyx breakdown. We hypothesized that early TXA administration would result in vascular endothelial preservation as evidenced by lower levels of thrombomodulin, syndecan-1, ICAM, and VCAM.

We analyzed a subset of subjects from the Prehospital TXA for TBI Trial, which examined the efficacy and safety of prehospital administration of TXA compared to placebo in patients with moderate or severe TBI who were not in shock. Blood samples were collected upon admission and at 6hrs. Glycocalyx breakdown markers were quantified using a Luminex analyte platform. Clinical variables were compared using Wilcoxon Rank-Sum tests for non-parametric continuous data and Chitests for categorical data. Differences in median marker levels were evaluated using t-tests performed on log-transformed variables. Significance was set at 0.05.

Data from 437 patients [placebo (n=201), TXA (n=236)] were analyzed. Groups were well-matched for age, sex, injury mechanism, admission injury severity score, head abbreviated injury score, and presence of intracranial hemorrhage (ICH) on admission CT. No differences were observed in any median marker levels on admission or at 6 hours. However, admission levels of syndecan-1 in patients with ICH (n=287) who received TXA were lower than those in the placebo group (254.6 pg/mL [200.7-322.0] v. 272.4 pg/mL [219.7-373 .1], P=0.05). No differences in thrombomodulin, ICAM, or VCAM levels were detected at either timepoint in the ICH subgroup.

Administration of TXA early after injury may attenuate endothelial release of syndecan-1 in patients with moderate or severe TBI and ICH, potentially suggesting a selective role for TXA in endothelial gl

Despite a rapid increase in the use of the oral factor Xa inhibitors rivaroxaban and apixaban over recent years, there remains no standard management for associated life-threatening hemorrhage. Andexanet -approved reversal agent available but its place in therapy remains controversial due to its high cost and a lack of head-to-head trials comparing it to four-factor prothrombin complex -PCC).

We conducted a retrospective review of adult patients admitted with ICH associated with rivaroxaban or apixaban and -PCC for anticoagulation reversal between May 2018 and April 2019. The primary outcome was hemostatic efficacy using the ANNEXA-4 study rating system (excellent, good, or poor) based on initial and repeat non-contrast CT head imaging within 24 hours. Secondary outcomes included the occurrence of thromboembolic events and 30-day all-cause mortality. We excluded patients whose hematoma was surgically evacuated before the 24-hour CT or who received multiple reversal products.

22 ICH patients met the inclusion criteria: 8 andexanet patients (6 spontaneous and 2 traumatic) and 14 -PCC patients (1 spontaneous and 13 traumatic). 8 (100%) andexanet patients achieved excellent -PCC patients ( -PCC patients, 2 (14%) achieved good (p=0.26) and 2 (14%) achieved poor (p=0.26) hemostasis. Thromboembolic events following -PCC patients (p=0.05). Thirty-day all-cause mortality occurred in 1 (13%) andexanet patient and 1 (7%) -PCC patient (p=0.67).

-PCC for reversing ICH associated with rivaroxaban and apixaban. Limitations include our small sample size and -PCC in this population now that andexanet alfa is widely available.

A quality improvement project was undertaken to understand the risks of central venous catheter associated venous thromboembolism (VTE) in the NeuroICU setting.

All patients who were admitted to the NeuroICU and required a central venous catheter from 7/1/2018 to 12/31/2018 were included in the study. All catheters were placed under ultrasound guidance using the Seldenger technique. The site of catheter insertion, duration of dwell time and subtype were recorded for each catheter that was placed. Catheters were categorized as cooling catheters, large bore and dialysis catheters, or standard multi-lumen infusion catheters. Clinical suspicion for VTE such as extremity edema or unexplained hypoxemia triggered the standard of care use of ultrasound and/or lung CT angiography for diagnosis. VTEs with an appropriate chronology and in the same vascular distribution as the suspected catheter were categorized as catheter associated.

174 catheters in 104 patients were included in the analysis representing 1022 catheter*days. A total of 8 catheter related VTEs were observed in our cohort. In a mixed NeuroICU cohort the overall VTE rate was 7.8 per 1000 patient days which is in line with prior published rates. Multi-lumen infusion catheters had the highest rate of VTE (8.1 ± 3.3) and cooling catheters had the lowest rate (6.3 ± 6.3). Surprisingly, the highest rate of VTEs was observed in catheters placed in the subclavian vein across catheter types (13.7 ± 7.8).

We observed that multi-lumen infusion catheters had a higher rate of VTE compared with cooling and large bore catheters. This finding may be related to longer dwell times for multi-lumen catheters (6.4 ± 4.6 vs [cooling] 4.9 ± 2.4 and [large bore] 4.8 ± 4.4). The subclavian vein was the site with the highest rate of VTE which may be related to more lateral approach taken with ultrasound guided subclavian catheter placement.

Patients on direct acting oral anticoagulants (DOACs) have high mortality after intracranial hemorrhage (ICH). Prothrombin Complex Concentrate (PCC) has been used off-label to treat ICH while on DOACs. PCCs effect on laboratory markers of anticoagulation have varied. Whether or not a change in laboratory markers of anticoagulation impact outcome is unknown.

Retrospective, single center design assessing patients on DOACs that presented with ICH and received PCC. The primary outcome is to describe changes in anti-Thrombelastography (TEG) parameters before and after receiving PCCs. Hemostatic efficacy (defined by International Society on Thrombosis and Haemostasis criteria), and thrombosis rate are also reported.

Thirty five patients were included. Patients were 73.6+/-9.6 years old and 54% were male. 48.5% had traumatic brain injury related hemorrhage, 37% had primary intracerebral hemorrhage, 11.4% had subdural hemorrhage, and 2.8% had subarachnoid hemorrhage. Median Glasgow Coma Score at was 0.38 units/mL +/-0.50 units/mL. On average TEG R time decreased 14 +/-34 seconds and TEG ACT time decreased 34 +/-84 seconds. Hemostatic efficacy was excellent or good in 70% of patients and poor in 30%. Thrombosis rate was 8.6%. Overall mortality was 49%.

There was a modest response in laboratory parameters after giving PCC to patients with DOAC associated ICH. The mortality in this cohort was high. Whether a laboratory response in coagulation dosing, laboratory response, hemostatic efficacy and patient outcomes.

In critically ill patients with TBI, agitated behaviors may often be threatening for patients safety and for clinical teams. Antipsychotics are commonly used for the acute management of these agitated behaviors. However, animal TBI models suggest that repeated use of antipsychotic agents reduce cognitive and functional recovery. It remains unknown if the use of these agents negatively impact the functional recovery of TBI patients. Our objective was to describe the use of antipsychotic agents and agitation/delirium monitoring practices in critically ill TBI patients.

We conducted a retrospective observational study of 9 adult ICUs in Canada that manage TBI patients. Consecutive adult patients with moderate/severe TBI admitted to ICU between January 2015 and December 2016 were included. Data were collected using standardized forms for up to a maximum of 21 days in ICU or until transfer out of ICU. The primary outcome was incidence of antipsychotic use.

We included 333 patients (3037 patient-days) with a moderate (44%) or severe (56%) TBI. The majority TBI included falls (48%), MVA (34%) and assaults (8%). Antipsychotics were used in 44% of patients for a total 27% of patient-days. Quetiapine, haloperidol, olanzapine, and risperidone were used in a 14%, 11%, 1%, 1% of patient-days, mostly for agitation, an unclear reason or delirium (16%, 8% and 3% of total patient-days, respectively). A delirium monitoring tool was used 16% of patient-days whereas the RASS and SAS were used in 67% and 12% of patients-days, respectively.

Despite uncertainties regarding their efficacy and safety, antipsychotics are frequently used in critically ill moderate/severe TBI patients in Canada, mostly for the management of agitation. Sedation/agitation tools are mostly used for the monitoring whereas delirium tools are more rarely used.

Traumatic venous sinus thrombosis (tVST) is increasingly detected on neuroimaging in acute head trauma, and may be an important contributor to elevated ICP refractory to standard medical/surgical treatment, and in turn, higher morbidity/mortality and more complex ICU course. We sought to identify clinical and neuroimaging features predictive of refractory ICP issues in tVST patients treated in an urban level I Trauma center.

Retrospective query of electronic radiology database from 2015 to 2018 using the phrase "venous sinus thrombosis". Cases were reviewed and scored by a fellowship-trained neuroradiologist to define degree of occlusion (partial vs complete) and cause of sinus occlusion (extrinsic compression vs intrinsic thrombus vs both). Additional patient characteristics included demographics, mechanism of trauma, cerebral venous sinus involvement, laterality, skull fracture, extra-axial hemorrhage and invasive neuromonitoring. Refractory ICP was defined as at least one spontaneous ICP elevation >=5 minutes during ICU stay despite use of first tier therapies for ICP control. Odds ratios were computed and adjusted by multivariate logistic regression for patient age, gender and initial GCS to determine association with refractory ICP.

Among 74 patients with radiologic diagnosis of tVST, 18 developed refractory ICP (18/74=24.3%). Statistically significant variables associated with refractory ICP included involvement of internal jugular vein (aOR=6.11, 95% CI 1.65-22.62), involvement of transverse sinus (aOR=6.49, 95% CI 1. 22-34.60 ) and presence of temporal bone fracture (aOR=5.27, 95% CI 1. 48-18.72) . Potentially protective factors included sinus pathology secondary to extrinsic compression (aOR=0.23, 95% CI 0.50-0.87) and coexisting epidural hemorrhage (aOR=0.33, 95% CI 0.10-1.17).

Involvement of the internal jugular vein or transverse sinus and temporal bone fracture may represent sensitive features of tVST predisposing to refractory ICP issues, while extrinsic compression of a sinus alone was found to be protective.

Monitoring cerebral autoregulation in traumatic brain injury (TBI) patients can indicate an individual cerebral perfusion pressure (CPP) target for which autoregulation is best preserved (CPPopt): this offers a precision medicine approach with hypothetical advantage over the current 'one size fits all' strategy. Large retrospective data suggest that managing CPP close to CPPopt has a benefit in outcome. A prospective evaluation of CPPopt guided therapy is needed, but before performing an outcome study it is necessary to assess the feasibility and safety of such a protocol. The primary objective of COGiTATE (CppOpt GuIded Therapy Assessment of Target Effectiveness) is to demonstrate feasibility of individualising CPP at CPPopt in TBI patients, expressed as the percentage of monitoring time for which CPP is within 5 mmHg of regularly updated CPPopt targets during the first 5 days of Intensive Care Unit (ICU) admission. Secondary objectives are to investigate the safety (increases of the Treatment Intensity Level) and physiological effects of this strategy (changes in autoregulation indexes, organ function parameters).

COGiTATE is a phase II non-blinded, randomised controlled trial currently ongoing in the ICU of Cambridge, Leuven, Nijmegen and Maastricht. Severe TBI patients requiring intracranial pressure directed therapy, are enrolled in the first 24 hours after ICU admission and allocated into two groups. In the intervention group the CPP target (CPPopt) is calculated using a (modified) algorithm previously described by Liu X et al. 2017 and clinically reviewed 4-hourly. The control group uses a fixed CPP target (60-70mmHg).

Patient re have been recruited so far.

Randomising between a fixed and variable CPP is feasible. After completion of recruitment and follow up in terms of assessment of safety and physiological parameters, we will consider progressing to a phase III study.

Selective reduction of non-classical monocytes has been associated with reduced neutrophil activation in murine traumatic brain injury (TBI) models. Similarly, CD4-/CD8-T cells or double negative T-cells (DNT) may exacerbate ischemic brain injury. This study sought to assess the expression of peripherally isolated T-cells and monocytes after acute TBI.

All patients admitted with primary TBI to the Neurotrauma ICU between November 2017 and November 2018 were eligible for study. Consent was obtained and blood samples were obtained within 24 hours of injury. Samples were compared to healthy age-and sex-matched controls. Conventional flow cytometry techniques gating on All patients admitted with TBI to the Neurotrauma ICU between November 2017 and November 2018 were eligible for study. Consent and blood samples were obtained within 24 hours of injury. Samples were compared to healthy age-and sex-matched controls. Conventional flow cytometry techniques, gating on CD3+ and CD14+ were employed to identify T-cell and monocyte populations, respectively. Data were analyzed using cytometric fingerprint binning and t-SNE embedding, which captures the set of multivariate probability distribution functions and generates maps that facilitate quantitative comparisons.

8 patients were compared to 10 controls. After computational analysis, 27 distinct T-cell phenotypes were identified, of which 11 were statistically significantly different between patients and controls expressed as a fraction of CD3+ cells. Three of these eleven subsets had a CD4-/CD8-(double negative) phenotype that were depressed among patients: CD4-/CD8-/CD127+ 0.00058% versus 0.0060%, p=0.008; CD4-/CD8-/161+ 0.0009% versus 0.0080%; p=0.013; CD4-/CD8-/127+/161+ 0.008% versus 0.046%; p=0.003. There was a three-fold decrease in the fraction of Type 3, non-classical monocytes in patients than in controls [0.068 (IQR 0.037-0.103) versus 0.02 (IQR 0.0175-0.028); p=0.04]. Similar patterns in monocyte expression were observed for the 3 patients who had repeat analysis at 72 hours.

In this preliminary study, there were notable reductions in DNT populations and non-classical monocytes in patients with acute TBI, which may suggest recruitment to the CNS. Prior studies suggest that DNT play a critical role in the perpetuation of cerebral ischemia after acute stroke and that type 3 monocytes modulate neutrophilwarranted.

Much of the secondary injury that occurs after traumatic brain injury (TBI) results from coagulation derangements related to disseminated intravascular coagulation (DIC). Extracellular vesicles (EVs) are small (0.1transduction. EVs are released from all cell types, including platelets, endothelium, and granulocytes which are responsible for DIC. We hypothesized that specialized flow cytometry techniques could identify a unique EV signature of DIC in acute TBI.

3 EV fluorescence panels were created assessing for endothelial cells (CD 144+, CD 105+), platelets (CD 41, CD41a+, CD 42b+), erythrocyte markers (CD 235+) as well as brain specific biomarkers (S100b). Using a modified flow cytometry instrument for detection of small particles, side scatter signal is used to estimate EV size. Samples were treated with triton, which disrupts vesicular membranes, abolishing EVs. Samples were prepared in Trucount tubes with a known number of lyophilized beads, which enabled the determination of the plasma volume. All combinations of positive/negative expression were counted.

There was no significant difference in the total number of EVs in the panels between the 11 patients and 7 controls. Of 84 combinatorial analyses in the first EV panel, the following were significantly elevated after Bonferroni correction: CD31+/CD144+ 6.89 EVs/uL plasma v 0 controls (Wilcoxon rank sum p=0.009); CD 105+/CD144+ 7.64 EVs/uL plasma v 0.41 controls (p=0.04); CD235+/CD41a+ 3433.6 EVs/uL plasma v 501.1 controls (p=0.009). Brain biomarkers were also elevated: S100b 32581.7 EVs/uL plasma v. 2688.6 controls (p=0.004).

evaluate whether this expression correlates with secondary microvascular brain injury. S100B EVs (membrane bound, not free soluble protein) are significantly elevated in TBI patients; if reproducible, the significance of this remains to be elucidated.

Diabetes insipidus (DI) following transsphenoidal craniotomy may lead to significant metabolic derangements. Serum sodium imbalances are frequent and important; both hypo-and hypernatremia can be devastating neurologically. A project aimed at improving DI management through predictive assessments and DDAVP protocols could potentially improve patient outcomes. However, few predictors for the postoperative development of DI have been reported.

After institutional IRB exemption, the records of 121 patients undergoing endonasal transsphenoidal craniotomy between July 2015 and December 2018 were retrospectively reviewed. Demographics, preoperative medical or radiologic diagnoses, medications, and laboratory values as well as intraoperative blood loss, urine output, and DDAVP administration were assessed for correlation with the incidence of postoperative development of DI using logistic regression.

Development of postoperative DI was defined as postoperative DDAVP administration and/or DDAVP use upon or after discharge from hospital. 17 of the 121 patients developed postoperative DI. Patients 0.037, and 0.047, respectively). Similarly, patients with increased intraoperative blood loss, increased intraoperative volume administration, nd intraoperative DDAVP or vasopressin administration were also more likely to develop postoperative DI (pwith logistic regression modeling adjusted for associations between outcome and potential risk factors, patients having a documented or clinical suspicion for a preoperative endocrinopathy had seven times higher odds of developing postoperative DI compared to their peers (p-value 0.111, 95% CI 1. 571-33.151 ).

In administration, and DDAVP were independently associated with an increased risk of postoperative DI; the odds of postoperative DI were seven times higher in patients with a documented or clinical suspicion findings.

The 5 seminal mechanical thrombectomy (MT) trials had a median age of 68 years. Though some of these trials included nonagenarians, there is little data on their outcomes. We aimed to compare the procedural, discharge outcomes and complications, of MT for acute ischemic stroke (AIS) in

Patients with AIS admitted to two comprehensive stroke centers were enrolled prospectively in a MT, procedural outcomes, complications, and discharge disposition were compared in propensity scorematched groups (matched for NIHSS, pre-stroke mRS, IVdefined as a discharge to home/acute rehabilitation.

Of the 3010 AIS patients, 46/297 (16%) nonagenarians underwent MT compared to 159/1337 (12%) 69) were propensity score-matched with a median admission NIHSS of 22 and 19, and median ASPECTS (84% vs 50%, P=0.035), whereas ICA (10% vs 13%, p=0.76), and M2 (21% vs 43%, p=0.19) occlusions were similar between the two groups. Time to groin puncture (100±65 vs 76±34; p=0.124), revascularization time (134±72 vs 110±54; p=0.145), complication rates (0 vs 5.1%; p=0.494) and inhospital deaths (11% vs 24%; p=0.155) were similar among the two groups. 44% of nonagenarians had

We present one of the largest series of MT among nonagenarians with 89% successful recanalization rates. In propensity score analysis almost half of nonagenarians (44%) were discharged to home/rehab, which is comparable to a younger cohort (51%). Aggressive management is warranted in the oldest of the old.

Early neurologic deterioration (END) occurs in up to one third of stroke patients and is associated with poor outcome. No consistent definition of END exists regarding degree of NIHSS decline and timeframe. We evaluated the definition of END, predictive factors, and 90 day outcomes in a cohort of critically ill stroke patients.

This study is a retrospective review of consecutive ischemic stroke patients with NIH stroke scale (NIHSS) intervention factors were obtained. END was defined as a delta NIHSS > 2 at 24 hours from admission. Reperfusion was defined as a Thrombolysis in Cerebral Infarction (TICI) score of >2b, cerebral edema treatment as any ICP-lowering therapy, and poor outcome as mRS >3 at 90 days. Multivariable logistic regression analyses were performed to assess factors associated with END and poor outcome.

183 patients (median age 67 years, 50% women, median NIHSS 20) met study criteria. 23% experienced END. Admission NIHSS, administration of tPA, receipt of intraarterial therapy, and successful reperfusion were not associated with END. END was independently associated with older age (P=0.036), sex (p=0.043), and treatment of cerebral edema (P=0.039) after adjusting for cerebral herniation and tracheostomy. Poor outcome was associated with older age (P=0.001), higher delta NIHSS (P<0.001), not receiving tPA (P=0.015), and placement of percutaneous endoscopic gastrostomy tube (P=0.006). END patients had a higher median 90 day mRS (P<0.001).

END as defined by a delta NIHSS > 2 at 24 hours predicts poorer outcome, but was not associated with tPA or intraarterial therapy, which contrasts with prior literature. This variance could be attributed to the END timeframe defined as 24 hours rather than the typical 48samples sizes and comparison of END timeframes could clarify observed findings.

ANNEXA-4 was a single-arm, prospective, open-label study of andexanet in patients presenting with major bleeding within patients with spontaneous intracranial hemorrhage (ICH). Brain imaging was performed at baseline, and at 1 and 12 hours post andexanet treatment. Subdural hemorrhage (SDH) thickness and ICH volumetric analysis was performed using Quantomo software. Co-primary efficacy outcomes were change in anti-

Of 352 patients enrolled in ANNEXA-4, nontraumatic ICH was present in 128 patients, including intracerebral +/-intraventricular in 99 patients, subarachnoid in 5 patients and subdural in 14 patients. In this cohort, mean age was 78 years (SD 9.2) administration was 11.9 hours (IQR 7.9-15.3); median time from symptoms to CT was 2.6 hours (IQR 1.2-2.7); and median time from CT to andexanet administration was 1.8 hours ). Median intraparenchymal volume was 9.5 mL (IQR 3.9-21.3). Among 99 efficacy evaluable patients (baseline anti-treatment overall. In patients treated <2 hours after baseline imaging, hemostatic efficacy was 82.5%; 2-3 hours after baseline imaging, 70.8%; >3 hours, 75.0%. Within 30 days, death occurred in 24 patients (18.2%).

Andexanet reduced anti--or apixaban-associated nontraumatic intracranial bleeding and with a high rate of hemostatic efficacy up to 12 hours after treatment.

Spontaneous intracerebral hemorrhage (ICH) is associated with high rates of mortality. Multiple scoring systems exist however the original ICH score remains most commonly used. We hypothesize that patients undergoing SCUBA, compared to medically managed patients, would have lower 30-day mortality than predicted.

We performed a retrospective observational cohort study of consecutive nontraumatic spontaneous ICH patients treated at a single, tertiary care, academic center from December 2015 to June 2018. Patients for each patient based on the admission ICH score.

A total of 100 ICH patients were included. The median age was 60 (Q1=53, Q3=73), GCS 10 (7, 13), and NIHSS 17 (13, 22) . Sixty-three were deep hemorrhages and 43 had intraventricular hemorrhage. Median pre-operative volume was 40.9ml (27.8, 64 .3). The expected 30-day mortality was 34.7% while the observed mortality was 9%. On 180-day follow up, a mRS of 0-3 was seen in 46% of patients.

Patients undergoing SCUBA have an absolute risk reduction of 25.7% in mortality than predicted by the ICH score. Good outcome to moderate disability, defined as mRS 0-3, was achievable in almost half the Introduction Andexanet (coagulation factor Xa [recombinant] inactivated-zhzo), a specific reversal agent for factor Xa 9% of patients with major bleeding in the ANNEXA-4 trial. However, little is known about the clinical factors associated with a hemostatic response in patients with intracranial hemorrhage (ICH) receiving andexanet.

ANNEXA-4 was a prospective, single-arm, open-label study of andexanet in patients with acute major treatment was rated by an independent adjudication committee as excellent, good, or poor/none based on pre-specified criteria. All ICH patients with evaluable HE were included in the analysis. Univariate and indication for anticoagulation, baseline antianti-platelet use, time from last dose to andexanet (and other time intervals), neurologic function, and hematoma characteristics were performed to identify factors predictive of HE.

Of 227 ICH patients enrolled, 219 (96.5%) had evaluable HE. In patients with ICH, baseline antitime from symptoms to andexanet were all significantly associated with HE. In multivariate analysis, time from last dose (12.3h for excellent/good; 8.4h for poor/none), time from symptoms to andexanet (7.0h for excellent/good; 4.4h for poor/none), and time from symptoms to scan (3.8h for excellent/good; 1.6h for poor/none) were independently associated with HE.

In ICH patients treated with andexanet in the ANNEXA-4 study, various time intervals were predictive of hemostatic efficacy. These findings suggest that shorter time intervals are associated with lower HE and are consistent with the known relationship between time from symptoms and the risk of hematoma expansion.

Alterations in functional connectivity are associated with persistent cognitive deficits in survivors of aneurysmal subarachnoid hemorrhage (SAH), but causation remains unknown. Therefore, we sought to and behavior could be assessed.

We used functional optical intrinsic signal imaging to measure spontaneous hemodynamic fluctuations -operated (n=10) and SAH (n=11) mice. We tested behavior using the Morris water maze, open field test, Y-maze, and rotarod. Timepoints were from 4 days to 3 months. We used the anterior prechiasmatic injection model of SAH.

0.13), and visual cortex (0.24 vs. 0.15) at day 4 following sham procedure or SAH, as measured by the proportion of brain surface with a correlation coefficient >0.5 (sham vs. SAH, respectively, p<0.05).

-independent ng SAH. A global connectivity index remained decreased until 1 month following SAH (0.42 vs. 0.37, p<0.01 ). An interhemispheric connectivity index was also he hidden platform test on the Morris water maze (p=0.045) and open field test (34 vs. 14 m, p=0.034) at approximately 2 weeks. There were persistent deficits on the Y-maze for at least 3 months (59% vs. 48% alternation, p=0.028 with repeated measures at 1 and 3 months). There was no significant effect of SAH on rotarod performance.

We studied whether high-protein supplementation (HPRO) and neuromuscular electrical stimulation (NMES) after subarachnoid hemorrhage (SAH) could be a safe and feasible approach to reduce muscle wasting and improve long term recovery.

assigned to standard of care (SOC) or NMES + HPRO. NMES was delivered to bilateral quadriceps and gastrocnemius muscles during two 30-minute sessions daily along with HPRO (goal 1.8 g/kg/day) between post bleed day (PBD) 0 and 14. Tolerability was measured during each NMES session by assessing for agitation or discomfort. Safety measurements included increased heart rate, blood pressure, or intracranial pressure (when monitored) during NMES. Stimulation sites were assessed after each NMES for muscle injury or skin trauma. HPRO tolerability was assessed by monitoring for gastric retention or emesis. Safety measures included aspiration and evidence of acute kidney injury. NMES and HPRO were discontinued if subjects refused. The goals were to administer at least 80% of NMES and HPRO. Muscle wasting was assessed with serial CT scans of the thighs.

Twenty-five subjects (SOC=13, NMES + HPRO=12) participated with no differences in baseline characteristics (58 years old, 68% women, 50% HH>3). Median intervention days were 12 (range: 9-14), with 98% of NMES sessions completed. Two subjects had transient muscle soreness but no other adverse events. No adverse events were associated with HPRO. The HPRO group received 83% of the goal and more protein than SOC (mean difference: 0.63+/-0.17g/kg/d, P<0.01). Muscle atrophy at PBD 14 was greater in SOC group (6.7+/-3.2% vs 2.8+/-2.9%, P=0.01).

NMES and HPRO are safe and feasible after SAH. A larger pilot study is underway to understand whether NMES and/or HPRO may beneficially impact neuromotor recovery after SAH.

Lipocalin-2 (NGAL) is released by activated neutrophils and astrocytes and mediates neuro-inflammation and iron regulation in hemorrhagic stroke models. Blood NGAL is an early biomarker in human ischemic L and neurofunctional outcome in SAH patients.

Magnetic Luminex Assay, R&D Systems) and assessed modified Rankin Scale (mRS) every 3 months. Patients with renal or severe liver dysfunction, active malignancy or intracranial infections were excluded. Poor outcome is defined as mRS>2. Vasospasm was defined as >50% reduction any vessel caliber on cerebral angiogram. Continuous variables were compared with Student's t or Wilcoxon rank sum test depending on data distribution. One-way ANOVA was used for multi-group comparison.

SAH cohort has mean age of 55.1 years, 69% women, 48% with poor 3-month outcome and 55% developed vasospasm. Higher plasma NGAL on post--SAH days 1-3 (p=0.05) and 5 (p=0.017) are associated with poor 3-month outcome. Higher plasma NGAL on postand NGAL on post-SAH days 1--0.0063)

Early elevation of plasma NGAL on post-SAH day 1 is associated with vasospasm and poor 3-month -SAH 1-5 are associated with poor 3-month outcome. Larger population studies are needed to validate plasma NGAL as a potential SAH biomarker.

Patients with aneurysmal subarachnoid hemorrhage (aSAH) are at risk of rebleeding prior to aneurysm obliteration. While placebo-controlled studies have shown that administration of either -aminocaproic acid (EACA) or tranexamic acid (TXA) can decrease rebleeding, there has not been a comparison of the two in this patient population. Because of a national shortage of EACA in 2017, our hospital changed to TXA. The purpose of this study is to describe the outcomes of aSAH patients treated with either EACA or TXA.

This is a retrospective chart review of patients who presented with an aSAH between 4/1/16 and 3/31/19 and were treated with either EACA or TXA to prevent aneurysm rerupture. Descriptive statistics were used.

There were 43 patients with aSAH who received EACA and 45 who received TXA. The groups were EACA group and 86.7% in the TXA group. The average time from admission to drug initiation was 6.0 ± 3.8 hours in the EACA group and 4.8 ± 4.0 hours in the TXA group. No patient in either group experienced aneurysm rerupture after receiving the drug. Similar numbers of patients in both groups had cerebral ischemia (EACA: 23% vs. TXA: 24%) and extracranial thrombosis (EACA: 7% vs. TXA: 9%). Although TXA is known to lower the seizure threshold, we found no increased incidence of seizures (EACA: 28% vs. TXA 13%). There was a modest cost difference in favor of TXA vs. EACA.

There does not appear to be any major differences in outcomes in patients with aSAH treated with either EACA or TXA for the prevention of aneurysm rerupture and a slight cost savings favoring TXA. A larger prospective study is required to confirm these results.

Outcome prediction after aneurysmal subarachnoid hemorrhage (aSAH) is based on scores, which are determined once at admission. However, the occurrence of delayed ischemic neurological deficits (DIND) depends on multiple concomitant and continuously changing factors. The goal of the study was to establish an automated analysis pipeline to predict DIND from multimodal data.

Multimodal data (patients' history, imaging and laboratory values among others) from 105 patients with aSAH were analyzed. DIND was defined as new ischemia or perfusion deficits in native or contrastenhanced CT/MRI and/or cerebral vasospasm in conventional, CT-/MR-angiography. A ranking of the features was performed by univariate regression analysis. Only cases with <10% of missing values were included in the model. Among the 412 tested features, the top 26, with a false discovery rate <0.01, were selected. Missing values were imput random forest machine learning algorithm was applied. The performance of the prediction was estimated on the fly by predicting the observations that were not used for building the tree ("out-ofbag") across all trees.

The final data matrix contained 464 events described by 26 features from 67 patients. In the final model, the out-of-bag estimate of error rate was 23%, which reflected a 77% accuracy. The importance plot for different features revealed the importance of some parameters known in the context of inflammatory response, which is linked to the pathophysiological cascade leading to DIND. These included counts of leucocytes, monocytes, neutrophils, and lymphocytes. However, other laboratory parameters, such as zinc and selenium, appeared to be of high importance in the model, which was somewhat unexpected.

Machine learning algorithms may be helpful to filter out predictive features from a large number. These features might be subsequently investigated regarding their predictive value on the occurrence of DIND after aSAH.

Innate inflammation is a recognized mediator of DCI after SAH. We have shown that neutrophils and the neutrophil-derived enzyme, myeloperoxidase (MPO), mediate memory deficits in DCI. How MPO affects memory is unclear. There is evidence that MPO, and its substrate H2O2, may act through astrocytes or directly on neurons. Here we test MPOs action on astrocytes and neurons.

Primary neuronal and astrocyte cultures were developed from WT and C57BL/6Thy1-GCaMP3 mice. To test if MPO or H2O2 are toxic to neurons or astrocytes, cells were incubated with MPO (0.8u/ml), H2O2 (0.0012%), and MPO/ H2O2 and evaluated with Live/Dead cell viability assay (Thermofisher). To test if neuronal firing is affected by MPO, the same experimental conditions were examined in C57BL/6Thy1-GCaMP3 using video microscopy. Neuron activation was stimulated with KCl (final concentration 10mM).

Addition of H2O2 led to death in neurons and astrocytes. MPO did not affect cell death in either group. Interestingly, MPO/ H2O2 showed less cell death than H2O2 alone suggesting a neuroprotective benefit of MPO. In neurons, KCl administered to untreated neurons led to continuous firing as evidenced by intense calcium signal. MPO addition did not change the firing rate when compared to baseline. After 2.5 hours of MPO pretreatment, activation with KCl showed a suppressed firing rate suggesting neuronal depression. The addition of MPO/ H2O2 showed the same firing rate suppression as MPO alone.

This study suggests that MPO acts directly on neurons to decrease function. In our model of neutrophilinduced development of DCI, MPO is released in the meninges, diffuses to the brain parenchyma and acts directly on neurons to affect memory. This needs to be tested more thoroughly in an in vivo model of SAH. How MPO specifically affects memory in neurons is an area of interest in our laboratory.

Delayed cerebral ischemia (DCI) is a feared complication of subarachnoid hemorrhage (SAH), leading to worse outcomes. Electroencephalography (EEG) provides a useful, continuous monitoring tool for DCI risk (Claassen 2004; Kim 2017; Rosenthal 2018) and late-onset epileptiform discharges (ED) have high predictive value for DCI (Kim 2017; Rosenthal 2018) . However, optimal parameters to assess ED contribution to longitudinal DCI risk are unknown. We hypothesize that the evolution of ED frequency after SAH can provide early identification of those at high DCI risk.

We analyzed continuous EEGs from 124 patients with moderate to severe aneurysmal SAH. ED were identified using a commercial detection algorithm (Scheuer 2017). We calculated ED frequency (per hour) after SAH and compared mean ED frequencies between DCI and control patients. We also evaluation, we performed group based trajectory analysis (GBTM) and calculated hourly receiver operating curves (ROC).

ED rates were higher in both DCI and control groups during the clinical DCI "risk period" of day 3-14. Overall mean ED frequency were significantly higher in DCI patients (t-test, p=0.03), including only pre-DCI ED assessment (t-test, p<0.01). Hourly mean ED rates remain higher in DCI patients from days 2-13. Using GBTM, we identified three distinct trajectories associated with DCI (79%, 52%, 37%, p=0.01), with group number selection optimized based on Bayesian Information Criteria. Hourly area under the ROC (AUC) calculations of ED frequency yielded a maximum performance of 0.78.

Natural history of ED frequency in all SAH patients coincides with the "high risk" time-period of DCI. Patients with DCI have higher mean frequencies that remain elevated throughout this DCI risk period. GBTM and AUC calculations suggest longitudinal analysis of discharge frequency can differentiate DCI risk, but integration of other waveform characteristics are needed to optimize prediction.

Aneurysmal subarachnoid hemorrhage (SAH) has high morbidity and mortality. Time to aneurysm repair, whether earlier or later in the course of the disease, may impact outcomes. However, optimal timing remains controversial. Our goal was to describe the association between time to aneurysm repair and mortality and functional outcome.

This study was conducted in two reference centers -one in Rio de Janeiro and one in Porto Alegre July 2015 to March 2019, every adult patient admitted to the ICU with aneurysmal SAH was enrolled in the study. Data were collected prospectively during the hospital stay. Patients were divided into four groups according to the moment of aneurysm repair after bleeding: <3 days, 4 to 8 days, >8 days and not repaired. The primary outcome was in-hospital mortality. Dichotomous variables were analyzed using twomortality as the reference group (4 to 8 days).

A total of 437 patients were included. Median age was 55 years, mostly female (73%). In the univariate analysis hydrocephalus, rebleeding, postoperative neurological deterioration (up to 48 hours after procedure), delayed cerebral ischemia, as well as mortality and poor outcome, were associated with the different timing of aneurysm repair. In the multivariate model for mortality, poor grade SAH, hydrocephalus, post-procedure neurological worsening and DCI were independently associated with higher mortality. Additionally, late repair was associated with lower mortality (OR 0.4) as compared with occlusion between 4 to 8 days.

Our study shows higher mortality in patients submitted to aneurysm occlusion procedure between days 4 and 8 after ictus, when compared to late repair. More studies are needed to define the best timing of aneurism repair in patients that are not submitted to early occlusion.

The biological mechanisms that influence abnormal cortical neurophysiology after aneurysmal subarachnoid hemorrhage (SAH) are uncertain. We hypothesized that soluble ST2 (sST2), a plasma marker of the innate immune response, is associated with events of electroencephalography (EEG) deterioration including new epileptiform abnormalities (EAs) or new EEG background deterioration.

-approved biospecimen repository, we evaluated patients with at least 3 days of EEG monitoring and an early sST2 measurement (collected <5 days following SAH). EAs were defined as sporadic epileptiform discharges, lateralized rhythmic delta activity (LRDA), lateralized periodic discharges (LPD), or generalized periodic discharges (GPD). Background deterioration was defined as decreasing Alpha Delta Ratio (ADR), Relative Alpha Variability (RAV) or worsening focal slowing. The association between sST2 level and EEG-identified EAs or new background deterioration was compared using the Wilcoxon Rank sum test.

46 patients met inclusion criteria. Early sST2 was collected at mean 3.5±0.9 days after SAH; 23 patients had a subsequent sST2 measurement at 10±2.5 days. 17 (37%) patients developed new EAs during EEG monitoring, 21 (46%) developed new background deterioration, and 8 (17%) developed neither. Median sST2 in patients developing new EAs was higher (114.8 ng/ml ]) than in patients who did not develop new EAs (74.7 ng/ml ], p=0.024). This association between elevated sST2 and new EAs was not present for sST2 samples collected at later time points. There was no difference in sST2 levels between patients who developed new background deterioration (75.7 ng/ml ) compared with those who did not (80.1 ng/ml ], p=0.56).

Among patients admitted with aneurysmal SAH, elevated sST2 in the first 5 days is associated with the development of new EAs on EEG monitoring. This association was not present at later time points, suggesting that the early inflammatory response may be linked to abnormal cortical neurophysiology.

Glial-mediated inflammation occurring early after status epilepticus (SE) in rodent models has been implicated in the subsequent development of spontaneous recurrent seizures (SRS). While this suggests anti-inflammatory strategies may be a target for therapeutic intervention, the appropriate timing for such an intervention is unclear. The aim of this work is to define the timing of early inflammatory changes using pro-inflammatory miR-155 and anti-inflammatory miR-146a as biomarkers in a kainic acid mouse model of SE.

SE was induced in 7-8 week old male C57BL/6J mice (n=5 per timepoint) using intraperitoneal injections of 30 mg/kg kainic acid. The onset of SE was defined as the first Class 3 seizure using a modified Racine Scale. The intensity of the SE episode was estimated by the total number of discrete class V seizures observed. After 2 hours, the SE was aborted with diazepam, and hippocampal tissue was harvested at 3 hr, 18 hr, 24 hr, 48 hr and 72 hr. RNA was isolated using TRIzole (Life Technologies) followed by qRT-PCR analysis to define the steady-state expression levels of miR-155 and miR-146a and their targets, SOCS1

We observed a >4 fold increase in expression levels of miR-155, reaching peak levels at 18 hours. Expression levels of miR-155 directly associated with the intensity of SE. The level of SOCS1 mRNA expression decreases after the peak expression of miR-155. As the levels of miR-146a were only

Conclusions miRNA-155 expression shows an early increase within 3 hours of SE, reaching a peak at 18 hours. miR-146a shows a non-miR-155 initiated after SE to determine if this can prevent the development of SRS.

Nurses routinely screen for changes in neurologic status with serial clinical assessments. The objective of this study was to employ mixed methods to determine inter-rater reliability (IRR), protocol adherence, and acceptability of a new tool we developed called Serial Neurologic Assessment in Pediatrics (SNAP) compared to the Glasgow Coma Scale (GCS).

SNAP assesses mental status, cranial nerves, communication, and four-extremity motor function/strength, with scales for children <6-months, 6-months to 2-ye -years-old. SNAP was designed for use in a diverse population, including patients who are intubated, sedated, and/or have developmental disabilities. IRR of independent SNAP assessments by pairs of trained nurses was assessed with multilevel Cohen's kappa and linear weighted kappa, calculated through clustered bootstrap method to account for multiple assessments. We assessed protocol adherence with standardized observations. We conducted semi-structured interviews to assess acceptability and feas We thematically analyzed interviews in accordance with modified grounded theory framework.

45 critical care nurses performed 387 paired SNAP assessments on 299 patients (148 <6-months; 91 6months to 2--years). There was substantial agreement between nurses (average kappa=0.7 <6-months; 0.8 6-months to 2--years), and IRR was unchanged for children who were intubated, sedated, and/or had developmental disabilities. IRR was unchanged based on degree of experience using SNAP and for day vs. night-shift nurses. Nurses had 100% protocol adherence. SNAP was easier to use and more precise at describing neurologic status of patients who were intubated, sedated, and/or had developmental disabilities than GCS. 88% of nurses preferred to use SNAP over GCS.

When utilized by nurses, SNAP has substantial IRR, excellent protocol adherence, and is acceptable and feasible to i neurologic decline.

Several studies demonstrate significant gender disparities in professional societies for critical care and neurology, but data for neurocritical care is lacking. We examined gender representation trends within the Neurocritical Care Society (NCS), the largest international professional society for this subspecialty. We hypothesized that female representation has increased with achievement of gender equality in 2018.

A multidisciplinary writing group obtained approval from the NCS Executive Committee and endorsement by the Women in Neurocritical Care (WINCC) section. After review by the Rush University IRB, access was granted for the following rosters: general membership, Board of Directors, Officers, committees, Annual Meeting speakers, grant, fellowship and other award recipients. We differentiated between female, male and unidentified gender.

Available membership rosters from 2005 listed 524 members, with gender unknown for >75%. In 2018, of 2633 members 37.6% were females, 49.2% males, and 13.3% unidentified. As of 2018, 3/17 presidents (17.6%) and 4 -2018, female committee members increased from 15% to 47%; female committee chairs increased from 18% to 48%. To date, 1/15 (6.6%) Christine Wijman Young Investigator awardees were female with no female recipients of the Best Scientific Abstract Award (2007) (2008) (2009) (2010) (2011) (2012) (2013) (2014) (2015) (2016) (2017) (2018) . 36% of Presidential Citation awardees -80% from 2011-e representation in guidelines writing groups ranged between 5%-75% (2015-2017), and 9-45% in consensus statements writing groups (2011) (2012) (2013) (2014) (2015) (2016) .

-43% (2003awardees were women.

Within the NCS, a longitudinal increase in female representation has occurred over the last 19 years but gender equality has not been achieved. We recommend focused efforts to facilitate inclusion and gender equity within NCS.

With the push toward using large data sets in critically ill patients, the use and management of registries is becoming more relevant. Clinical registries provide insight about associations and patterns in diagnosis, disease, and treatment. The integrity of the data is of utmost importance. This poster describes the quality control and data management methods for maintaining the integrity of a multicenter trial Registry.

We employed modifications to Van Den Broeck's method of data organization to clean and manage the END-PANIC registry. The data management consisted of five phases: 1) Screening phase, 2) Data Organization , 3) Diagnostic phase, 4) Treatment phase, and 5) Missing data phase. The screening phase consisted of distinguishing missing and extraneous data elements, outliers, inconsistent patterns/distributions and unexpected analysis results. The data organization phase consisted of treating blank cells and highlighting errors with data input. The diagnostic phase was used to clarify the true nature of the data points, and make sure the data presented was biologically possible. The treatment phase consisted of correcting variables. The missing data phase consisted of determining whether the missing data was informative or noninformative.

Currently the multi-center registry houses ~3.5 million discrete data points from 3,272 patients. There was a high correlation between the Texas, Ohio and California locations, and NPi, DVL, CVL, MCVL, and Pupil Size. There was a low correlation between the Texas, Ohio and California locations, and Pupil Latency and presence/absence of cataracts. Missing data was informative for age, race and ethnicity, and distribution of missing data caused an inquiry into methods for collecting data and implementation plans for change.

This interdisciplinary method for cleaning and managing the END PANIC registry was able to identify and rectify errors. We would recommend others to use the methods to build, clean and manage clinical registries.

Objective was to describe current state of quality improvement (QI) processes implemented in neurocritical care units (NCCU).

A 27-question-survey was sent to 2000 members (physician, nurses, and pharmacists) of the Neurocritical Care Society. We describe factors affecting the presence of NCCU QI, barriers to QI, awareness of stroke (STK, CSTK), stroke get with the guidelines (GWTG), trauma quality improvement program (TQIP) and American Academy of Neurology (AAN) performance measures, and examined factors affecting satisfaction with current practices.

The response rate was 22.5%; 73.7% of respondents were from US teaching hospitals, 87.9% practiced in dedicated neurocritical care units, and 43.4% in a program with a neurocritical care fellowship. 44.6 % reported a dedicated NCCU QI program. Comprehensive stroke center (RR 1.7, 95% CI 0.92-3.13, p = 0.09), dedicated NCCU (RR 2.0, 95% CI 0.83-4.83, p = 0.07), and NCC fellowship programs (RR 1.41, 95% CI 0.99-2.02, p = 0.06) were more likely to report dedicated NCC QI staff. External ventricular drain infection was the most commonly tracked NCC QI metric (69.6%). Respondents indicated the highest level of awareness for CSTK (87.5%), STK (81.8%), and GWTG (81.8%), but indicated a relative lack of awareness for TQIP (42.7%), and AAN (46.2%) perform satisfaction with existing NCC QI were: presence of a hospital QI program (RR 8.81, 95% CI 1.30-59.7), p = 0.0004), presence of a formal NCC QI program (RR 4.33, 95% CI 1.47-12.7), p = 0.0003, and dedicated NCC QI staff (RR 2.23, 95% CI 1.51-3.32), p < 0.001). Insufficient hospital (57.7%) and departmental support (36.5%) were reported common barriers to the successful implementation of an NCCU QI program.

A dedicated staffed NCCU QI program occurs in a minority of neurocritical care units, and the lack of such programs may lead to clinician dissatisfaction. Institutional and departmental support may be critical elements of a successful and satisfactory implementation of NCCU QI.

The development and implementation of a nurse driven rounding model was instituted in the Neuro ICU of an academic medical center to increase effectiveness of team communication, practice autonomy and integration of nursing input into the interprofessional care plan.

Clinical nurses and neuro-intensivists developed a structured rounding tool to guide the nursing presentation of clinical information on rounds. The interprofessional team underwent education on expectations and processes. The rounding tool underwent a number of revisions over a 24-month period based on feedback from all team members and evolving patient care priorities. All team member roles in the rounding process were clearly defined with nursing leading patient assessment and goals. Nursing satisfaction surveys assessed nursing attitudes regarding autonomy, decision making and RN-MD communication via a 5 point Likert scale; mean values for each question domain were compared pre-and post-implementation.

In total, 106 nursing surveys were analyzed, 44 pre-implementation and 62 nurses postimplementation. Mean response values evidenced significant improvement across all domains in the post-implementation group: Autonomy (3.70 vs 4.31, p<0.01), RN Decision Making (3.16 vs 4.04, p<0.01), p<0.01) . Survey participation was good in both groups (68% pre-and 78% post-implementation).

Nursing satisfaction across multiple important domains improved following implementation of a nurse driven structured rounding model. Application of a nurse-facilitated, structured model creates a standardized reliable process that can be observed by all team members in order to deliver data driven, high quality, efficient and effective care.

Multiple models for program development and care delivery in pediatric neurocritical care (PNCC) have been proposed with varying degrees of success. Here we present a unique model for building a dedicated pediatric neuro-intensive care unit (PNICU) through creation of a community of practice (CoP). CoP represents a mechanism for collective learning and production of repertoire of best practices through knowledge sharing, development of social capital, and support for organizational change.

We utilized a Bolman and Deal 4-frame for organizational functioning (Structural, Human Resources, Political, Symbolic) to describe the development of our PNCC CoP. We evaluated our PNICU with the standards outlined by the Neurocritical Care Society (NCS) for a Level 1 neuro-intensive care unit.

Structural factors included forming PNCC leaders across specialties (neurology, critical care, neurosurgery, radiology, nursing), opening (in 2018) a state-of-the-art, unique PNICU which includes wired rooms for continuous EEG monitoring and multimodal neuro-monitoring, meeting 137/140(98%) of NCS standards. Human resource factors included creating core groups of physicians and nurses with a primary role in PNICU, providing ongoing education through workshops, lecture series, and certification including ENLS and TNCC, meeting 74/105(70%) of NCS standards. Politically, a PNCC fellowship-trained, board-certified physician serves as medical director coordinating conception of collaborative partnerships across multidisciplinary experts. Simultaneous creation of other specialty cohorts in pediatric critical care aided in departmental acceptance for the program, meeting 4/4(100%) of NCS standards. Symbolically, we set forth our shared purpose and strong commitment to foster CoP that advances knowledge and best practices for PNCC.

Using CoP principles, we have accomplished many of the NCS standards over a relatively short period of time. We plan to further develop the program with particular focus on education, certification, and expansion to include allied health professionals. Our roadmap may be applicable to any institution interested in developing a PNCC CoP/PNICU.

Intravenous (IV) anti-hypertensive infusions are often used acutely in patients with intracerebral hemorrhage (ICH). There is a lack of standardization of titration and variation in goal blood pressure, and therefore their use is associated with increased ICU length of stay (LOS) and cost. We examined the use of anti-hypertensive infusions in ICH patients in our institution and developed a quality improvement intervention to reduce duration of infusion, ICU LOS, and cost.

Patients were included if they were admitted to our ICU from September 2017-March 2019 with an ICD-10 diagnosis of non-traumatic ICH and received IV antihypertensive infusions. Interventions introduced starting in November 2018 included interdisciplinary task force formation, provider education, updated rounding checklist, and EMR order with clear blood pressure target. The primary outcome measure was duration of anti-hypertensive infusions determined by retrospective chart review, and secondary outcome measures of ICU LOS and cost data were obtained from our finance department.

Over 5 months, mean antihypertensive infusion duration reduced from 54.2 hours (n=66) to 37.9 hours (n=29). ICU LOS reduced from 3.35 to 3.0 days. Proportion of cases with discordant blood pressure goal documentation reduced from 46.0% to 24.1%, while discordance in documented goals to actual orders reduced from 57.1% to 27.6%. There were no significant increases in countermeasures (infusion restarts, ICU readmission, and AKI due to blood pressure lowering). Extrapolating from finance data, and our baseline infusion duration and ICU LOS data, IV antihypertensive infusions cost ~$36/hour. Our improvement suggested $17017 in estimated cost savings in 5 months. ICU accommodation cost was approximated at $262/hour, for an estimated $63823 additional cost savings.

A quality improvement based intervention targeting management of hypertension resulted in reduced duration of anti-hypertensive infusions, ICU LOS, and cost. The intervention was feasible and ongoing data collection is warranted to assess sustainability.

Mortality and long-term-disabilities secondary to stroke are high. Educating high-risk population with early stroke symptoms has been outstanding. However, education of post-stroke consequences (requiring resuscitation codes and goals-of-care awareness) is lagging. This study evaluates the understanding of such concepts by the admitted stroke patients (high risk population) and visitors (general population).

were asked to answer a preliminary question about their original code status then read a self-explanatory sheet followed by revealing their revised code and goals-of-care choices. We used within-group logistic-regression-analyses to determine changes of codes among original coders and types of novel codes among post-survey coders. This included proposition of new short-term resuscitation (STR-STRP [partial]) codes. We used between-group chi-square-analyses to determine differences in education between groups.

The odds of changes in no-coders were 2.429, 4.957 in patients and visitors, (p-value= 0.001, <0.0001) respectively. The odds of changes in DNR-coders=0.265, 0.151, partial-coders= 0.353, 0.277, full-coders= 0.073, 0.043 times those of the no-coders respectively (p-value< 0.00001). The odds of novel-DNR-coders= 1. 429, and 3.696, 299, 379.5, 54.625, 6 .851 times those of novel-no-coders respectively (p-value< 0.00001). STR-coders originated from other-codes> no-coders. Between-group analyses showed 42%, 52% of patients versus visitors changed their code status respectively (p-value= 0.04519). Goals-ofcare choices indicated tolerance towards temporary measures (Tracheostomy and feeding-tube placement) and hemiplegic disabilities without poor mentation among the majority (~53-77%) as a target for continuing care. Pre-event (stroke) documentation of code status was approved among the majority of participants (58%).

There is a misunderstanding of the resuscitation codes among both admitted stroke patients and general population. However, the difference between both indicates reception of some education among the stroke patients. STR-STRP are a good alternatives for many people. Pre-event documentation -stroke outcome awareness are needed.

Early integration of palliative care improves communication, decision-making and social support in patients with acute stroke in the neurocritical care unit. The primary objective of this study was to analyze how early palliative involvement impacts communication between the healthcare team and patients/families.

In this ongoing prospective study, patients with moderate to severe ischemic and hemorrhagic strokes were randomized into control and intervention arms. The control arm received routine ICU care and the intervention arm received an early palliative care consultation. Study assessments with the patient or surrogate decision maker were obtained at day 1-3, and day 5-7 of ICU care. Comparisons were made for total scores on the Questionnaire on Communication (QOC), Decisional Conflict Scale (DCS), and Hospital Anxiety and Depression Scale (HADS). We performed an interim analysis utilizing the Student's T-Test and Chi -Square Test on SPSS 25, with results below as mean + standard deviation.

Of 13 patients enrolled (7 intervention and 6 control), 50% and 44% were female (p = 0.07). The average age was 61 + 12 and 65 + 11 years (p = 0.52). The majority (83% and 57%) were ischemic strokes (p = 0.503). Admission NIHSS was 16 + 8 and 13 + 8 (p = 0.35). There was no difference in total QOC (67 + 41, 87 + 19, p = 0.26), HADS (18 + 7, 17 + 7, p = 0.79), or DCS (31 + 24, 37 + 20, p = 0.59) scores. When comparing responses to individual questions, a trend toward improvement in QOC responses was observed "Using words you can understand" (p = 0.05) and "Answering all questions about illness" (p = 0.08).

Early integration of palliative care may improve communication between healthcare providers and patients/families, specifically with regards to using appropriate language that is understandable.

Routine daily chest radiographs (CXR) in mechanically ventilated patients (MVP) are often performed in the ICU for "monitoring" purposes, despite lack of specific indications. Routine daily tests are of questionable value and may increase costs without clinical benefit. The Society for Critical Care Medicine and Choosing Wisely Campaign promote indication-based test ordering. Studies involving medical-surgical ICUs demonstrate that indication-based versus routine daily CXRs in MVPs results in cost-savings without jeopardizing outcomes. We implemented a quality improvement initiative targeting reduction of routine daily CXRs in MVPs in the NSICU.

We convened an interprofessional team of attending physicians, fellows, medical students and nurse practitioners. We conducted educational campaigns promoting evidence-based CXR utilization practices. Standardized discussion of indication for CXR was incorporated into rounds. Iterative process improvements were adopted beginning June 2018. CXR utilization rates in MVPs were measured the first 12 weeks of 2017, 2018 and 2019 and compared pre/post-intervention. Hospital length of stay (HLOS) was evaluated to monitor for complications resulting in prolonged hospitalization. 

Implementation of indication-based ordering strategies decreased CXR utilization in MVPs in the NSICU without increasing HLOS. Value-based care quality improvement initiatives can reduce costs without compromising clinical outcomes.

Patients transferred from NSICU to lower acuity units are vulnerable to readmissions and hospital acquired complications. Standardized handoffs may help reduce this risk within academic institutions where physician trainees possess varying levels of clinical experience. We sought to implement a standardized handoff (I-PASS) within inpatient Neurology, focusing on high risk patient populations. Residents and attendings were surveyed about inpatient handoff practices to inform implementation of I-PASS.

An electronic survey was administered in 2018 to residents and inpatient attendings in Neurology at University of North Carolina (UNC). Handoff practices among inpatient services (Wards, Consults, NSICU, and Epilepsy) were evaluated. Surveys assessed perceived quality of handoffs, as well as problems with handoffs leading to adverse events. Surveys were sent to 30 physicians (19 residents, 11 inpatient attendings); 21 responses (12 residents, 9 inpatient attendings) were obtained (response rate, 70.0%).

-six percent of residents and 81% of attendings reported that problematic handoffs had been the primary or contributing factor to one or more adverse events. Overall quality of handoffs involving NSICU patients transferred to lower acuity units was reported as a concern, with 50% of residents indicating the quality of these handoffs to be poor. In ranking inpatient services for prioritization of handoff interventions, 69% of residents identified NSICU handoffs as either their first or second highest priority. We also found residents exhibited a self-performance bias, with 84% reporting that they provided all pertinent information during handoffs most of the time, and only 22% reporting that they received all pertinent information during handoffs most of the time.

Inpatient handoffs are perceived as problematic by residents and attendings, with handoffs involving transfer of NSICU patients identified as high priority for targeted intervention. UNC Neurology has since implemented I-PASS protocols to improve the safety of handoffs involving NSICU patients.

Targeted temperature management (TTM) to 33-36c is the standard of care for post-cardiac arrest patients. Recent literature has demonstrated a new trend of worsening morbidity and mortality postarrest due to under-utilization of TTM. Management of post-arrest patients is a multidisciplinary health care effort, and knowledge of TTM rationale and protocol varies. Normothermia (37-37.5c) also could have neuroprotective benefit in other clinical scenarios and is another indication for TTM. We hypothesized that a focused educational intervention would improve TTM protocol compliance.

A multidisciplinary team developed a standard educational presentation and a 10 question exam given as a pre-and post-test to residents, fellows, and critical care nurses. Baseline data on TTM use was established followed by 6 month prospective data collection post-intervention. Data was extracted from Arctic Sun® machines on all TTM cases (post cardiac arrest and normothermia). The primary outcome was compliance with the TTM protocol measured by correct temperature target goals and appropriate duration, assessed by chi-square analysis. The secondary outcome measure was individual score improvement, evaluated by 1-variable students T test.

There was a total of 62 TTM cases pre-intervention, and 51 TTM cases post intervention. There was a trend toward increased TTM protocol compliance (29% to 43%), however this was not statistically from pre-test (n=54) to post-test (n=43) after the education presentation (P<0.0001, CI 2.48 to 3.52) among all health care participants. The resident, fellow, and nursing scores increased from 57% to 88%, 62% to 90%, and 70% to 86%, respectively.

Educational interventions for physicians and nurses caring for post-cardiac arrest and neurocritical care patients improved knowledge gaps and helped improve compliance with TTM protocol. Additional education and process improvement activities are warranted to further improve protocol compliance, which may improve patient outcome.

Identifying the appropriate level of care needed for a patient presenting with acute intracerebral hemorrhage (ICH) is often imprecise. The utility of prior work in triaging patients is limited by exclusion of non-primary ICH patients, which is often difficult to determine prior to admission. This study aims to identify which admission factors are associated with ICU level of care on presentation.

This is a single-center retrospective review of patients admitted to our institution with ICH in 2018, regardless of etiology. All patients were admitted to the neurocritical care unit (NCCU). ICU level of care was defined as the need for mechanical ventilation, administration of vasoactive or insulin infusions, continuous renal replacement therapy, ventriculostomy, treatment of cerebral edema, temperature management, management of status epilepticus, or neurosurgical intervention. Logistic regression was used to identify characteristics associated with ICU level of care.

183 patients (median age 62, 49% female, median admission GCS 10, median ICH volume 22 mL, 55% with IVH, 48% lobar, 19% infratentorial) were admitted with ICH. 141 (77.1%) required intensive care. The most common interventions required were mechanical ventilation (100 patients, 69.9%), antihypertensi with need for intensive care included age (62 vs. 60), admission GCS (9 vs. 15), deep location of ICH (36.4% vs. 15.6%), ICH volume (28 mL vs. 8 mL), and presence of IVH (61.0% vs. 33.3%). On multivariate analysis, age (p = 0.01), admission GCS (p < 0.001), and deep location (p = 0.05) were independently associated with the need for intensive care.

Among all patients presenting with ICH, age, admission GCS, and location of hemorrhage may help identify ICH patients who need ICU level of care.

The impact of Emergency Neurological Life Support (ENLS) course on provider knowledge and selfreported comfort in management of neurocritically ill patients in a low-middle income country such as Cambodia is unknown and explored in this study.

In-person ENLS courses with English to Khmer translated slides were conducted in 2 hospitals in Phnom Penh, Cambodia in May, 2019. Wilcoxon Signed Rank test and matched paired t-test were used to examine pre and post-course scores on translated knowledge-based multiple choice tests. A descriptive analysis was performed to evaluate provider comfort in management of neurocritically ill patients pre and post-course and amongst individual ENLS modules.

Overall, 52/57 healthcare providers participated; 27 (52.9%) physicians and 25 (47.1%) nurses. Thirtythree (73.3%) had acquired base specialty training in Cambodia, 23 (51.1%) had completed subspecialty training in critical care medicine and 31 (68.9%) previously cared for neurocritically ill patients. Pre-test sores were 50% [IQR21]; post-test scores were 53.8% [IQR 15]. Though not statistically significant, posttest scores were higher for providers who had base specialty training in Cambodia (54.8 % vs. 49.9%, p = 0.06), subspecialty training in critical care medicine (55.7% vs. 50.7%, p = 0.06) and previous experience caring for neurocritically ill patients (54.9 % vs. 50%, p = 0.06). Most (82%, n = 37) reported that ENLS training had prepared them for management of neurocritically ill patients.

ENLS courses may enhance the knowledge and comfort of healthcare providers in managing neurocritically ill patients in low-middle income countries, however this may depend on prior experience and minimizing language barriers. The impact of ENLS courses on outcomes in neurocritically ill patients in low-middle income countries warrants further study.

Neurocritical care has become increasingly subspecialized.Yet, due to limited availability of dedicated Neurocritical Care units (NCCUs), often patients may need to be admitted to ICUs other than NCCUs. This survey based study was conducted to explore self-reported knowledge in recognizing and managing some common neurological emergencies such as stroke, status epilepticus, raised intracranial pressure etc among critical care nurses at a Comprehensive Stroke Center.

In January 2019, we engaged 240 nurses from 8 ICU units in this QI project-which included medical, surgical, neurocritical care, cardiac and cardiothoracic units as well as post-anesthesia care unit (PACU) and interventional radiology units. Using institutional RedCap anonymized surveys were sent to the nurses.Information on demographic and critical care work experience was recorded. All participants answered questions with a likert type scale on their knowledge of several common neurological emergencies.

240 Nurses (199 females, 41 males) participated in the survey. 112 (58%) had been working in an ICU for 3 years or longer. Their self-reported level of knowledge in managing neurological emergencies revealed that more than half the participants did not feel comfortable managing patients with EVDs, ICH, SAH, raised intracranial pressure, TBI and traumatic spine injury patients. More than 70% of Nurses were not satisfied with their current level of training to deal with neuroemergency and supported the need for dedicated training/ study time.

ICU nurses report gaps in fundamental knowledge in recognizing and managing common neuroemergencies. This highlights the need for providing ongoing training and education about neuroemergencies to critical care nurses to help maintain competencies.

Simulation training has been increasingly adopted in critical care specialties to promote active learning and create a reproducible platform for feedback. The role of advanced simulation as a core component of training in neurocritical care remains unclear, which may be due to uncertainty about the degree of fidelity needed. Our objective was to determine if trainee knowledge and/or confidence differs when using standardized patients as compared to a multi-media simulation platform in a Neurocritical Care Concepts training course.

Methods 20 junior neurology residents engaged in 3 simulated neurologic emergencies: a right MCA stroke case, status epilepticus case, and a pontine hemorrhage/coma case. The MCA stroke and status epilepticus cases were portrayed by trained standardized patients for half of the residents (Group SP), while the other half interacted with the manikin supplemented with video clips of pertinent neurologic exam findings (Group MV). Both groups interacted with the manikin for the pontine hemorrhage/coma case. Before and after the course, residents completed a 40-question multiple-choice test on management of neurologic emergencies and a survey about their confidence in managing 15 neurologic emergencies. A detailed task checklist was used to assess decision making during the simulations.

Both resident groups had statistically significant higher knowledge and confidence scores after their training sessions (Knowledge: pre: 49% vs post: 72%, p<0.001; Confidence: average pre: 0.98 to post: 2.01, p<0.001). However, there was no statistically significant difference between the two groups in either knowledge or confidence. The task checklist demonstrated significant variations in treatment practices and provided individualized areas for teaching.

This pilot study suggests that trainees' knowledge and confidence in the management of neurocritical care concepts increases following simulated encounters, regardless of whether an actor-patient or multi-media simulation platforms is used. Use of a task checklist uncovered important variations in protocol adherence among novice physicians.

The accurate evaluation and determination of brain death has broad consequences on life-saving organ donation, closure for families, and length-of-hospital-stay. We have observed a concerning variability of brain death testing knowledge and comfort amongst neurology attendings and trainees at our institution. We aimed to create and apply a combined didactic and simulation training program to increase the knowledge and comfort in brain death evaluation, using our approved institutional brain death policy as reference. We hypothesized that participants who attended the training would show a measurable increase in their knowledge and comfort in the clinical evaluation of brain death.

An experienced neurointensivist (>15 years of clinical practice) presented a 1-hour didactic session on brain death criteria, evaluation, and pitfalls to neurology residents and attendings. A high-fidelity simulation was implemented to allow practicing the brain death examination. Knowledge and comfort levels were measured before and after learners had attended both sessions using electronic -Exact-tests were applied to examine changes in knowledge and comfort in brain death testing pre-and post-exposure to the educational sessions.

27 participants (9 residents, 18 attendings) completed pre-exposure, and 12 (3 residents, 9 attendings) have completed post-exposure questionnaires thus far. Knowledge significantly improved from pre-to post-exposure (59% correct, range 18-91% improved to 76% correct, range 64-91%; p=0.0002). Comfort levels in performing the brain death examination pre-exposure also increased from pre-to postexposure (pre: "Very Comfortable-30%","Somewhat Comfortable-37%","Neutral-7%","Somewhat or Very Uncomfortable-26%" to post: "Very Comfortable-50%", "Somewhat Comfortable-33%","Very Uncomfortable-17%" [p=0.005]).

Exposure to a single combined didactic and simulation session improved the knowledge and comfort levels immediately post--exposure questionnaire response rates, as well as measurements of knowledge retention over a 3-and 12-month period and accurate application in practice.

The safety and benefit of early mobilization in general intensive care units (ICU) has been found to improve outcome and decrease length of stay. However, there is a lack of literature on early mobilization in the Neuro ICU (NICU) specifically, due to the complexity of the patients in the NICU and their disease processes. Traditionally, patients were kept on bedrest after subarachnoid hemorrhage, intracerebral hemorrhage, ischemic stroke, and neurotrama, due to neurologic limitations such as fluctuation in mental status, requirement for sedation and paresis. Additional challenges associated with mobility in this population include the potential for positional changes to impact intracranial pressure Physician comfort level and concern for adverse neurological outcomes such as vasospasm or increased bleeding also decrease mobilization. While it is imperative to be cautious with NICU patients, prolonged bedrest and restricted mobility come with its own set of complications including muscle atrophy, decreased activity tolerance, delirium, pressure sores, nosocomial infections and deep vein thromboses. We sought to develop an early mobilization guideline that would help multidisciplinary staff identify which patients in the NICU should be mobilized early.

A NICU physical therapist and the director of the NICU identified criteria for patients who were appropriate/inappropriate for early mobilization.

All patients in the NICU should be mobilized early with the exception of the following exclusion criteria: unstable respiratory status, status epilepticus, contraindication to holding sedation, Rass -4, changing/worsening neuro exam, ICP > 20mm Hg, mean arterial pressure <65 or >110mm Hg, oxygen saturation < 88%, acute myocardial infarction, >2 vasopressors, clinical vasospasm, perfusional state,

Guidelines on early mobilization in the NICU can optimize patient mobility while minimizing complications associated with mobilization.

Introduction delivery, nurses must develop leadership skills and serve as full, collaborative partners with physicians and health professionals (2011). Registered nurse (RN) inclusion into rounds has been shown to: improve interdisciplinary collaboration, incorporate learning in the workplace, increase leadership skills and improve team members' perception of unit flow and culture.

Attending physicians, RNs, Neurocritical Care fellows, nurse practitioners, pharmacists and respiratory therapists were surveyed via SurveyMonkey to examine opinions regarding current rounding processes and potential opportunities in the Neurocritical Care Unit (NCCU). Responses were aggregated to create scores for each topic, with the priority areas being the lowest relative scores based on a 5-point Likert scale.

Survey responses were collected from 43 NCCU staff members (62% response rate). Based on survey results, priority areas to enhance rounding satisfaction included: increasing collaborative decision making, creating entire team efficiency, completing rounds in a timely manner, increasing engagement and minimizing extraneous conversations and activities. Other targeted areas for improvement included reserving time for prolonged family meetings for post-rounds, as well as focusing educational time and consistently utilizing the rounding checklist. Based on areas of opportunity, a multidisciplinary committee was developed. One item created to enhance processes was the development and implementation of an RN facilitated presentation tool. To support this, a standardized presentation script and handoff tool were created and executed. Six-month follow up survey results are pending at the time of submission.

Strategies to improve communication in multidisciplinary rounds are key to decreasing errors and improving care delivery. It is likely that a systematic data presentation by bedside RNs will improve: staff perceptions of rounds, collaboration among all multidisciplinary staff members and rounding efficiency.

The Department of Neurosurgery has a readmission rate goal of less than 9.7 for the fiscal year 2018 and less than 9.2 for the fiscal year of 2019. Over the past four fiscal quarters there has been an increase in the department's readmission rate, always exceeding the institutional goal.

All readmissions in the institution's dashboard for Q3 and Q4 for 2018 and Q1 and Q2 of 2019 were reviewed by way of chart review. These were divided into spine vs cranial, planned vs unplanned readmission, reason for readmission and consistency vs inconsistency with the institution's dashboard.

In Q3 2018 the dashboard reported a readmission rate of 10.86. The final calculated actual readmission rate was 9.32. In Q4 2018 the dashboard reported a readmission rate of 11.04. The final calculated actual readmission rate was 9.27. In Q1 2019 the dashboard reported a readmission rate of 10.76. The final calculated actual readmission rate was 8.02. In Q2 2019 the dashboard reported a readmission rate of 10.73. The final calculated actual readmission rate was 7.30. The most common reason for unplanned n reason for planned readmissions were shunt placements after lumbar drain trials. The dashboard was correct in predicting planned vs unplanned readmissions 45.5% of the time.

The coding on the backend of the institution's dashboard is missing many staged and planned readmissions and is only accurate in coding planned vs unplanned readmissions half of the time. This is resulting in falsely elevated readmission rates. Despite the initial uptrend in readmissions, the actual readmission rates of the department are down trending and always below the institutional goal. This likely translates to other departments within the hospital. There needs to be a more efficient way to improve the coding and accuracy of the institution dashboards.

The critically-ill neurological patients managed by specialized neurocritical care team is associated with improved outcome. In Korea, limited data are available on improved outcomes after initiation of neurointensivist co-management in neurocritical care units (NCU). We evaluated the impact of a newly appointed neurointensivist on the mortality of patients admitted to the NCU.

The study was conducted in 18 intensive care unit (ICU) beds of a large academic tertiary care hospital. Neurointensivist co-management was initiated in March 2016. The retrospective observational study compared the outcomes of patients before and after neurointensivist co-management.

A total of 1211 patients were included, 647 prior to and 564 after the initiation of neurointensivist comanagement. Patients admitted after neurointensivist co-management were older and had higher APACHE II scores. ICU mortality was significantly decreased in patients managed by neurointensivist (14.4% vs 10.5%, p=0.04). The length of ICU stay and duration of ventilator days were shorted in patients without co-management.

Neurocritically ill patients managed by specialized neurointensivist showed better clinical outcomes despite increased severity.

Social media has changed the way individuals communicate with each other and has altered the way society obtains information. In the past ten years, multiple articles have been published highlighting the ability to utilize social media for education of medical, nursing and pharmacy students. To our knowledge, cross discipline education utilizing these platforms has yet to be evaluated. With over 2.23 to implement a pharmacist led, social media based nursing education program and evaluate the perceived value of this education.

A curriculum consisting of basic pharmacy related issues was developed and topics were posted to the ok users weekly. A pre-and post-education survey was sent out evaluating the program's effectiveness.

email. Of those nurses who received the pre-and post-education survey, a total of 29% and 19% completed the survey respectively. Of those who completed the survey 60% received education via -education survey, there were no statistically significant differences in nursing performance on fact based questions after receiving education (p-value > 0.05 on all 17 assessment questions). Overall, 100% of the respondents reported a positive learning experience and wanted to continue this method of education delivery.

The educational content. This project demonstrates the potential of utilizing social media as a means of cross discipline education; however, the solitary utilization of this platform should be used cautiously as this did not improve performance on assessment questions.

Consequently, targeted temperature management (TTM), either to maintain normothermia or induce hypothermia, is often advocated as a therapy to improve outcomes in brain injured patients. The physiological pathways that promote fever associated brain injury, and how these pathways might be modulated by TTM, remain unclear. This study examined the effect of fever and hypothermia on cerebrovascular pressure reactivity, a validated proxy of cerebral autoregulation.

We included 72 patients treated for brain injury from a single academic center. All patients had intracranial pressure (ICP), invasive brain temperature, and arterial blood pressure (ABP) recorded patient, mean PRx over all periods of fever (>38°C), normothermia (35-38°C), and hypothermia (<35°C) were calculated. Differences in mean PRx during normothermia, fever, and hypothermia epochs were then analyzed using paired Student's t-test. The relationship between PRx differences and total time spent normothermic was analyzed using linear regression.

spent at a normal brain temperature (p = 0.0008, R2 = 0.202). In contrast, hypothermia was not associated with impaired cerebral autoregulation (p = 0.9968).

This study supports the hypothesis that impaired cerebral autoregulation may be one mechanism through which fever worsens outcome in brain-injured patients. The effect of fever on cerebral autoregulation appears to be more pronounced in patients that spend a longer amount of time in a normothermic state. Interestingly, hypothermia was not associated with reduced PRx, suggesting that the possible benefits of therapeutic hypothermia do not occur by improving the autoregulatory state.

Veno-arterial extra corporeal membrane oxygenation (VA-ECMO) provides hemodynamic support in patients with refractory cardiogenic shock. These patients have a 15% incidence risk of cerebrovascular complications according to the Extracorporeal Life Support Organization Database. Reliable neuroassessments and neuroimaging are often limited by heavy sedation and risks of transporting these patients. Transcranial Doppler (TCD) can be a useful tool for cerebral hemodynamic assessment in these patients. We present four VA-ECMO patients where TCD spectral waveforms provided key information on cerebral blood flow despite non-pulsatile flow.

Interpretable spectral waveforms were obtained in three of four patients. Extensive embolization obscured flow patterns in one patient but clear cerebral perfusion with non-pulsatile waveforms was seen in the rest. Two of the three remaining patients had high intensity transient signals (HITS), suggesting cerebral microembolization. One patient showed pulsatility in cerebral waveforms despite no gross change in cardiac output on echo that helped guide decision to initiate ECMO wean. ECMO settings included flow at 4-6 L/min, MAP 60-80 mmHg, and PaCO2 between 30-50 mmHg. MCA mean flow velocities were comparable to the systemic BP, and ranged from 60-80 cm/s in three patients and 30-50 cm/s in one patient. One patient suffered cerebral edema and two expired from withdrawal of care on sedation after multisystem organ failure without a chance neurological or neuroimaging assessment. The fourth patient retained consciousness and the ability to follow commands, but died from a massive GI bleed.

TCD spectral waveforms can be useful bedside tools for patients on VA-ECMO to assess for cerebral perfusion patterns. Presence of HITS reflecting microembolization can guide perfusionists to check for pulsatile flow, their relationship with systemic hemodynamics and VA-ECMO settings is needed.

Cranial ultrasonography has a long history of use in neonates, but inadequate windows have limited its use in adults. A hemicraniectomy provides an obvious window for point-of-care intracranial imaging, providing similar views traditionally seen on CT and MRI. We describe a standard approach and settings, presenting sample imaging demonstrating key anatomic landmarks.

The hemicraniectomy ultrasonography preset was created and optimized using a phased array transducer with a 4-2 Mhz frequency range. Imaging parameters were tested and saved for 2D grayscale mode, with an emphasis on tissue harmonic imaging, adaptive image processing, and dynamic range. Axial views are obtained from the ipsilateral temporal window, approximating the pterion, adjusting the probe to display a well-aligned view using the lateral ventricles as a landmark. By convention, the probe marker is placed anteriorly. The depth and focus are set to visualize the brain he probe craniocaudally permits visualization of the entire cranial vault. Parasagittal and coronal views are obtained by placing the probe at the vertex, off midline, ipsilateral to the hemicraniectomy.

Several structures are clearly visualized and are available as landmarks for orientation. The ventricular system can be easily identified as hypoechoic spaces similar in appearance to CT or MR imaging. The brainstem and cerebellum, with its associated folia and peduncles are also easily seen. The thalami are identified as strongly hypoechoic paramedian structures. Pathologic findings that can be easily seen include hydrocephalus, hemorrhage, and edema. Aneurysm clips are hyperechoic with streak artifact, and ventriculostomy catheters can be seen as subtle hypoechoic areas within the cortex.

Hemicraniectomy POCUS can be used to visualize the intracranial vault to facilitate evaluation of structural lesions and pathology at the bedside. The authors advocate adding hemicraniectomy POCUS to the neurocritical care imaging arsenal in patients where this view is available.

Pain assessment is a challenge in critically ill patients with impaired consciousness, either because of sedation or concomitant severe brain injury. Automated pupillometry has been used to assess the response to noxious stimulation in such patients. Skin conductance, which has been used in the operative setting, has not been tested in this setting yet. The purpose of the study was to compare the pupillary response and skin conductance to pain stimulation in critically ill unconscious patients.

Prospective ongoing study including adult (>18 years) patients admitted to the Intensive Care Unit of a University Hospital and who were unconscious (Glasgow Coma Scale < 8 with a motor response < 5) for several reasons. Automated pupillometry (Algiscan, IDpupillary reflex dilation during tetanic stimulation. The tetanic stimulation (100 Hz) was applied to the skin area innervated by the ulnar nerve and was stepwise increased from 10 to 60 mA until pupil size had increased by 13% compared to baseline. The maximum intensity value allowed the determination of a pupillary pain index score ranging from 1 (no nociception) to 9 (high nociception): a pupillary pain peak per second ( concomitantly to tetanic stimulation.

Twelve patients (median age 66 [ranges=18-77] years; male gender 9/12) were included so far; eight patients had a primary brain injury (5/8 anoxic injury) and 4 others were sedated because of shock with concomitant respiratory failure. All patients were under continuous intravenous sedation and analgesia; 9/12 were on vasopressors and 6/12 on continuous neuromuscular blockade. Median GCS at the moment of pain assessment was 3 [3-7] and median PPI was 6 [1-9]; 4 patients (33%) had adequate pain control. No changes of skin conductance variables were reported during pain stimulation.

Skin conductance was unable to detect insufficient nociception in critically ill unconscious patients.

The cerebral arterial time constant (Tau) reflects the time it takes to fill the cerebral arterial bed with blood during one cardiac cycle, and is derived from arterial blood pressure (ABP) and middle cerebral artery flow-s with/without vasospasm (VSP) and delayed cerebral ischemia (DCI) after aneurysmal subarachnoid hemorrhage (aSAH). (2017) (2018) . Angiographic VSP and DCI were adjudicated by neurointensivists. Artifact-free cerebral arterial compliance and resistance. Statistical comparisons were made using a two-tailed Mann-Whitney U-test.

Of 20 aSAH patients, 13(65%) developed VSP and 8(62% of VSP) developed DCI. 4 patients had unilateral and 9 bilateral VSP (31 & 69% of VSP). One patient with unilateral VSP was available for monitoring prior to diagnosis. This patient had increased asymmetry in Tau over time prior to diagnosis (slope: 0.08 s/day, R¬2: 0.94). TCD measures in 3 patients were available prior to angiographic diagnosis of bilateral VSP, showing initial marginal asymmetry similar to the unilateral VSP case, then slightly decreasing asymmetry over time (mean slope: -22cm/s higher in DCI (p<0.001). Tau was 0.07s greater for DCI patients, however this did not reach significance (p=0.46).

explore the relationship of Tau asymmetries with VSP and DCI after aSAH. These may provide further insights into the pathomechanisms of VSP and DCI while also having potential as a tool for earlier diagnosis of these important complications.

Pupillometry is more accurate and has higher inter-rater agreement than subjective pupil size and reactivity estimation. Limitations include using a single high-intensity flash to evaluate the direct pupillary response only. We present preliminary data on using virtual reality-based pupillometry (VRP) with graded-intensity flashes and bilateral pupillary recording to monitor patients with large hemispheric infarction (LHI).

We utilized a virtual reality headset-based system, I-PAS (Neurokinetics, Pittsburgh, USA) to perform pupillometry. A total of 13 homogeneous illumination flashes (0.395 to 65.4 cd/m², 0.300 sec on, 2.00-4.00 sec off) were presented to each eye while infrared cameras recorded pupillary area (mm2) continuously at 100 samples/sec. This permits measurement of latency, magnitude and velocity of direct and consensual pupil constriction and dilation at each light intensity.

: We performed pupillometry as described above in 4 patients admitted with LHI from middle cerebral artery strokes. 2 patients required decompressive craniectomy (DC) during the hospital course while the other 2 patients did not require DC. Bilateral graded-intensity pupillometry detected subtle changes in pupillary reactivity (peak constriction velocity in mm2/s) prior to clinical deterioration, which were very pronounced when compared to normal control performance. Singleneuropupillary index (NPI) did not detect a change in pupillary reactivity in all but the most severe deterioration.

Virtual reality-based, graded intensity pupillometry is feasible in the intensive care unit and appears ed to set cutoff values that may aid in clinical decision making.

Limited access to conventional EEG results in significant delays to important diagnostic information, especially in patients with suspected non-convulsive seizures (NCS). Recently, the Rapid Response EEG technology has proven to be clinically valuable. However, the economic aspect of this new technology has not been studied in detail.

We retrospectively reviewed the use of the Rapid Response EEG device in our small community hospital over 6 months since its launch in December 2018. We performed limited chart review and collected information regarding EEG diagnosis, length of stay, and transfer to mothership hospital. We evaluated the clinical and economic impact of the device by considering the patients' clinical outcome and the estimated cost of hospitalization (~$3000-6000/day) and transfer ($4000-20,000). Metrics are not precise and are only estimates.

The device was used in a total of 20 patients. The treating physician or the nurse applied the device with and one with post-anoxic burst suppression. In patients with status epilepticus, seizures were aborted successfully, and median length of stay was 8.5 (national average of 15 days). All 20 patients were treated locally without requiring transfer to the main University hospital. Considering the cost of Rapid Response EEG infrastructure and disposables (<$30,000) compared to conventional EEG systems (~$100,000-140,000) and EEG technologists (estimated to cost ~$420,000-562,000), and estimated range of $50,000 to $266,000 in annual savings because of shorter LOS and lesser transfers, this new technology seems economically advantageous.

Rapid Response EEG System enabled significantly faster and easier access to EEG and helped detect a relatively high number of patients with gross EEG abnormalities. Adopting the Rapid Response EEG improved emergent NCS detection and treatment in a cost-effective manner.

Patients requiring neurocritical care frequently have neurologic fluctuations of uncertain significance. We hypothesized that severe and prolonged events of neurologic deterioration (ND) have the greatest impact on discharge neurologic status and serve as intermediate indicators of poor outcome.

We extracted nurse-documented GCS scores from electronic health record (EHR) data of consecutive patients admitted to a Neurosciences Intensive Care Unit (ICU) or undergoing intracranial pressure monitoring (April 2016 -1) best initial 24-hour GCS (bestGCS-24h), 2) maximum magnitude of GCS decline (maxGCSdecline), 3) duration of the episode of maximum GCS decline (dur-max), and 4) the maximum duration of any GCS decline >= 3 points (max3pt-dur). We fit a 10-fold cross-validated logistic regression model predicting the final GCS 3-10 (vs. 11-15) and tested it in a 30% hold-out sample. We then evaluated the rates of poor outcome for combinations of these parameters.

,077 consecutive admissions (4,522 unique patients) met inclusion criteria (10% with severe bestGCS-24h (3-8), 11% with moderate bestGCS-24h (9-12), and 79% with mild bestGCS-24h (13-15)). BestGCS-24h, maxGCSdecline, dur-max, and max3pt-dur, respectively, were independently associated with poor discharge GCS (OR per standard deviation were 0.44[95%CI 0.08-0.98], 2.02[1.01-4.08], 1.35[1.01-2.32], and 1.27[1.01-with a 4-point maxGCSdecline, the rate of poor outcome was 57% for patients with a severe bestGCS-24h and >=24-hour max-dur; 35% for patients with a severe bestGCS-24h and <24-hour max-dur; 9.3% for patients with a mild bestGCS-24h and >=24-hour max-dur; and 2.5% for patients with a mild bestGCS-24h and <24-hour max-dur.

Both the magnitude and duration of ND events are independently associated with neurological status at discharge. These empiric, informatics-derived thresholds may serve as useful intermediate outcomes facilitating the testing of biological associations and therapeutic interventions aimed at promoting neurologic recovery.

unit. Deteriorati worsening. We hypothesized that nonearlier than clinical deterioration.

We prospectively collected data from patients with acute brain injury who are at a high risk of perfusion disturbance (SAH, MMD, and severe anterior circulation ischemic stroke) between May 2017 and May 2018. Non--ry 3 seconds neurological worsening were assessed using perfusion imaging and were categorized as hypoperfusion group and hyperperfusion group. Baseline compared. 

Non-monitoring should be highlighted in patients with high risk of deterioration.

Intracranial cerebral pressure (ICP) monitoring is an integral part of acute brain injury management. While invasive ICP monitor is the gold standard, there are several medical conditions that preclude its placement. Non-invasive ICP assessment tests (e.g. optic nerve sheath diameter, optic nerve disk elevation, pulsatility index, pupillary reactivity etc) have moderate accuracy when used individually. The aim of the present study is to validate a multimodal approach for intracranial hypertension detection.

In this prospective study, patients with acute brain injury who had an EVD placement for both ICP measurement and treatment were included since March 2019. We measured bilateral optic nerve sheath diameter (ONSD) by ultrasound, bilateral optic nerve disk elevation (ONDE) by ultrasound, bilateral middle cerebral artery (MCA) pulsatility index (PI) by using transcranial Doppler and assessed pupillary reactivity with or without pupillometer as part of multimodal assessment for measuring intracranial pressure. We assessed the correlation and agreement of these values with ICP measured by the EVD.

We included 11 measurements in 7 patients with acute brain injury. The presence of two or more values of mean ONSD greater than 5 mm, unilateral or bilateral presence of ONDE and mean MCA PI greater than 1.0 has 100% sensitivity (95% CI 0.29 -1.00) and 87.5% specificity ( 95%CI 0.47-0.99) for predicting ICP greater than 15 mmHg.

Non-invasive multimodal assessment can be easily done by bedside, requires minimal training and seems to correlate well with increased ICP.

Raised ICP following acute brain injury is associated with poor outcome. Monitoring with early detection is important in reducing sustained ICP crisis. Previous studies demonstrated Rheoencephalography (REG) reflects cerebrovascular reactivity and may substitute invasive monitoring techniques. We hypothesized using a correlation coefficient between slow spontaneous changes in REG and Systolic Arterial Pressure to calculate REGx.

REG measurements were obtained from ten patients with acute brain injury. Analog waveforms of REG and arm bioimpedance pulse waves were recorded with a bioimpedance amplifier. We used the ICM+ program (PRx) calcu bioimpedance pulse waves (REGx) instead of ICP and invasive arterial pressure.

visualized by previously established waveform changes on REG. A change in mean REGx greater than the previous recording's mean REGx value was clinically significant as opposed to absolute mean REGx . One patient with a right ICA infarction clinically deteriorated from moving all extremities to extensor posturing on the right and flaccid paralysis on the left with significant delta mean REGx. Another with bilateral ACA distribution ischemic infarctions worsened from flexor to extensor posturing with significant delta mean REGx. Lastly, a patient with ventriculoperitoneal shunt malfunction repair improved from GCS 7 to 14 with multiple significant delta mean REGx values between recordings.

Our series demonstrated clinical significance of patient specific delta mean REGx suggesting importance of presenting mean REGx for detection of changes in intracranial compliance. Like presenting blood pressure and relative changes in blood pressure rather than absolute changes in blood pressure or specific values, REGx was shown significant in a similar manner. REGx is a realistic means of future noninvasive neuromonitoring.

Dialysis is characterized by markedly increased rates of stroke and cerebral micro-vascular disease, though the mechanisms by which dialysis modalities impact cerebral hemodynamics have not been well studied. This case series compares intra-dialytic cerebral hemodynamics measured by transcranial Doppler (TCD) in patients receiving intermittent hemodialysis (IHD) versus peritoneal dialysis (PD).

Ten outpatient end-stage renal disease (ESRD) without stroke were identified. TCD mean flow velocity averaged. Six patients administered hemodialysis were followed over 240 minutes, with mean arterial d every 30 minutes.

there was no statistically significant difference between dialyses group and no significant change over time. To quantify volatility in patient measurements over time, we calculated the coefficient of variation -Sum test. To test if there was a difference in volatility between dialyses groups, we used a Wilcoxon Rankgroup (p <0.05).

In this small case series, though cerebral hemodynamics are not significantly different among stable measures are more stable over time for patients on the peritoneal dialyses group.

End-stage renal disease (ESRD) patients with acute neurologic injury are at risk of altered cerebral hemodynamics during dialysis. Here, we present transcranial Doppler (TCD) images revealing marked intra-dialytic increased distal vascular resistance and compromised flow velocity in an ESRD patient with acute traumatic brain injury.

The patient underwent continuous TCD monitoring during hemodialysis to monitor intra-dialytic cerebral hemodynamics.

A 73 year-old man with ESRD on chronic presented with headaches after a fall. CT head revealed 6 mm right convexity acute subdural hematoma with 2-3 mm leftward midline shift and right parietal parenchymal contusion. On arrival to the neuro-ICU, the patient was afebrile, hemodynamically stable, and fully oriented with no focal deficits. Repeat CT head six hours from initial was stable. The patient was started on his outpatient prescription of dialysis (dialysate Na 140 mEq/L, blood flow rate 350 ml/min), run without heparin. Within first 3 hours of hemodialysis patient developed progressive rightsided headache, which evolved to vomiting, decreased in level of consciousness, and left-sided weakness. He intermittently opened eyes to stimulation but required persistent painful stimulation to answer orientation questions. He had no changes in mean arterial pressure during hemodialysis. His serum BUN had decreased from 80 to 43 mg/dL, and his serum sodium remained unchanged. Emergent CT head was stable from prior. Intra-dialytic TCD waveforms revealed progressively increased distal resistance to flow, measured by pulsatility index (PI) at his bilateral middle cerebral arteries (MCA), and compromised MCA velocities. This change was dramatic on the right, the same side as his subdural hemorrhage and cerebral contusion.

ESRD patients with critical neurologic injury are at risk for altered cerebral hemodynamics during dialysis. TCD ultrasonography may be a practical bedside tool to screen for patients at particular risk, and guide medical decision-making regarding dialysis prescription for ESRD patients in the neuro-ICU.

Point of care ultrasound (POCUS) differs from diagnostic ultrasound in being often performed by clinicians and focused to acquire only relevant images to answer a specific clinical question. Most ultrasound modalities have differentiated clinical indications where POCUS is appropriate: the use of echocardiography to rule out tamponade in shock is considered POCUS while the assessment of diastolic dysfunction in heart failure deserves a diagnostic exam. Neuroultrasound has been used in various clinical indications like vasospasm, intracranial stenosis, collateralization, and emboli monitoring. These studies are mostly performed by sonographers as diagnostic studies. With emerging interest in assessing POCUS indications, we performed a systematic literature review to identify all clinical indications of neuroultrasound and used a Delphi based review by three experts to differentiate clinical indications where neuroultrasound could have point-of-care uses.

Two authors (LMH, GB) performed a systematic review to identify all reported modalities and clinical indications of neuroultrasound (TCD, Duplex, B-mode, Carotid, Ocular and Temporal) in MEDLINE, EMBASE, Cochrane, and Scopus databases. Three experts (JGD, CT, AS) were surveyed using the Delphi method to review each clinical indication and modality on whether it was focused on diagnosis or management and whether the clinical indication was a valid POCUS. Differences in opinion were settled with a final face-to-face discussion to reach a consensus.

The systematic review determined 78 total clinical indications of point of care use of neuroultrasound individualized by disease and modality. In 35 indications it was considered a diagnostic adjunct, in 5 instances it was considered an aide in management, and in 38 instances it was determined to aid in both diagnosis and management decisions.

There are many point of care indications of neuroultrasound in neurocritical care. This consensus opinion can guide clinicians to clinical indications where point of care use can aide in bedside diagnosis and management.

In a systematic review, we reported current literatures on neuromonitoring methods in Left Ventricular Assist Device(LVAD) population.

We searched five databases (PubMed, EMBASE, Cochrane Library, Web of Science, Scopus, clinicaltrials.gov) related to LVAD and neurological monitoring methods from inception through January 2018.

Of 3597 unique citations, 32 studies (2055 participants) met the inclusion criteria. The median age was 48.5(interquartile range 40.3-52.8, 67.7% male). Study designs were retrospective observational studies (n=19) and prospective observational studies (n=13). Neuromonitoring methods studies included transcranial dopplers(TCD) for emboli monitoring(n=13) or cerebral autoregulation monitoring (n=2), traditional neuroimaging (CT/MRI) (n=7), cerebral oximetry(n=2), carotid ultrasound (n=3) and plasma VAD, 4 articles studied pulsatile- 

Current evidence on neuromonitoring in LVAD is limited and there is no consensus on the indication and effectiveness on use of any neuromonitoring methods. The publications have significant heterogeneity adequate power are warranted to develop an optimal neurological monitoring protocol and prevention strategy.

Midbrain compression secondary to cerebral edema or hemorrhage results in high mortality and morbidity. Quantitative pupillometry holds promise as a bedside indicator of worsening anatomic tissue shifts. Because pupil reactivity relies on an intact neural network through the diencephalon and brainstem, compression can lead to changes in pupil size and reactivity. We studied markers of compression and pupillometry within 2 hours of head CT in 35 patients with anterior ischemic stroke (AIS) or supratentorial intraparenchymal hemorrhage (IPH) causing mass effect.

We reviewed 62 scans from 35 patients with unilateral injury from AIS (>1/3 of MCA territory) or IPH (>30 mL). We assessed midline (MLS) and pineal gland shift (PGS), as well as novel measurements of midbrain compression including interpeduncular shift (IPS) and the ipsilateral and contralateral cerebral peduncle hemi-distances to the interpeduncular cistern (ICPHD, CCPHD). Multilevel modeling was used to analyze radiographic measurements with quantitative pupil metrics including pupil reactivity (DNPi) and size (DSize) differences between eyes. 

Pupil reactivity and size differences were significantly associated with radiographic markers of midbrain noninvasive indicators of brainstem compression.

Evaluation of optic nerve sheath diameter (ONSD) has been widely examined as both a correlate of intracranial pressure (ICP), and a potential predictor of outcome after neurological injury. Recent studies have evaluated sonographic measurement of ONSD, yet clinical limitations to this approach persist. Evaluation of ONSD measurements via routine brain computed tomography (CT) imaging has been less studied, but offers potential for detection of increased ICP in the absence of invasive monitoring. Previous studies have employed a cross-sectional approach to ONSD measurements via CT scan, primarily among patients with traumatic brain injury (TBI). However, no studies have evaluated serial correlations between CT ONSD measurements and ICP to evaluate strength of correlations during hospitalization, and across diagnosis types. The purpose of this study was to investigate correlations between ONSD via serial CT imaging and ICP among adult patients with neurological injury.

Retrospective cohort study of all adult patients admitted with acute neurological injury requiring ICP monitoring and critical care admission.

N=48. Diagnosis type included TBI (33%), aneurysmal subarachnoid hemorrhage (29%), intracranial hemorrhage (18%), cranial mass (6%), and other (12%). There was a strong, positive correlation between right/left ONSD across all time points (r=0.7-9, p<0.001), suggesting a consistent bilateral response. Correlations were strongest between initial inpatient CT scan ONSD readings and ICP (r=0.5, p<0.05), but decreased over time. Patients with increased ICP across all diagnosis types experienced higher ONSD values upon presentation to the emergency department (ED) and on serial CT scans throughout hospitalization (range 5.7mm-6.4mm, p<0.05).

urements as a potential indicator of increased ICP in the absence of invasive monitoring. Serial CT brain imaging is often performed to evaluate for intracranial changes during hospitalization, and measurement of ONSD during this imaging can contribute to decisions regarding more invasive monitoring.

Monitoring of burst-suppression-pattern (BSP) in electroencephalography (EEG) is relevant to control barbiturate-induced coma. Currently, the assessment of BSP is based on continuous observation of the EEG with manual counting of bursts per minute (BPM) by experts, which is prone to inter-rater variability.

We evaluated the reliability of a new algorithm for automatic BSP-detection compared to manual assessment in two Thiopental-induced burst-suppressed patients. A bipolar 8-channel EEG-montage was recorded. The montage was bandpass filtered into typical EEG rhythms and segmented into 2 secs -Moonen metric, a distance matrix between all epochs in the first hour of data from patient 1 us to cluster this matrix into 3 clusters: Burst, Suppression and Artifact. We labelled the rest of the (test) data from patient 1 and patient 2 by training Support Vector Machine classifier from the labels produced by clustering. The EEG was scored by a neurologist to get ground truth BPM ranges (min, max for intervals of 10 minutes to 1 hour) for both patients. The algorithm provided estimated ranges of BPM for these intervals.

The pilot data shows a high correlation of automatic burst counts compared to the manual counting. We found a significant Pearson correlation (patient 1: 0.84, p<0.001, patient 2: 0.83, p<0.001) and linear regression coefficient (patient 1: 0.85, p<0.001, patient 2: 0.7, p<0.001) between estimated and ground truth BPM ranges.

The automatic detection of the bursts provides an objective and fast assessment of BSP. The algorithm showed a slightly lower sensitivity due to the missing detection of very short or low bursts. We are ation.

ventilated neurocritically ill patients is unknown and explored in this study.

A retrospective cohort study was performed on patients admitted to the neurocritical care service between 1/2/2018 and 2/26/2019, Hospital-wide O2 shut down for maintenance and a switch to olerated with lowest being 21%

owest SpO2 of > 92% and SpO2 < 92% amongst the patients in the pre and post-O2 shutdown groups. 

-tolerated. With the risk of hyperoxia and its potential negative effects on neuronal injury, a subset of neurocritically

Whole body hypothermia has been used as a treatment for patients with severe traumatic brain injury (TBI) since many years. Invasive brain temperature monitoring is the most commonly practiced for target temperature management in these patients; however, complications are common due to the invasive nature of the procedure. The objective of the current investigation was to evaluate the association between brain temperatures obtained using a non-invasive sensor (AccuCor) and an intracranial pressure/temperature (ICP) catheter during selective brain cooling in patients with TBI.

aluated during a selective brain cooling over 24 hours using both a parenchymal ICP catheter (Raumedic -PT) and the AccuCor sensor, with a catheter positioned in the nasopharynx. Mean temperature values for each participant were obtained along the cooling intervention. Outlier values derived from the AccuCor sensor were detected and removed prior to comparison. The variation in brain temperatures was calculated by mean temperature differences obtained using both measuring devices for each participant.

Mean brain temperature values were very similar between devices: 36.04°C (35.14°C-38.36°C) for the ICP catheter and 36.10°C (34.55°C-38.77°C) for the AccuCor sensor (p-value: 0.05, 95% CI: -1.72 to 1.84). The median temperature difference between the devices was 0.41ºC (minimum: -1.21°C, maximum: 1.18°C, p-value: 0.89).

Our results suggest that there were no differences between brain temperature measurements conducted using the ICP catheter and the non-invasive AccuCor sensor. This conclusion highlights the precision of non-invasive temperature monitoring, a safe alternative to the current invasive practice. monitoring procedures.

Sepsis-associated encephalopathy (SAE) is a multifactorial syndrome, characterized as diffuse brain dysfunction that occurs secondary to infection in the body without overt central nervous system infection. The prognosis for SAE is associated with the degree of cerebral damage. We investigated the relationship between the wavelet coherence of cerebral oxyhemoglobin (OxyHb) among different channels and outcomes in patients with SAE.

14 consecutive patients with SAE were included. Moreover, we included normal controls (n=26) for comparison. The cerebral OxyHb data were collected using functional near-infrared spectroscopy (NIRSIT, OBELAB Inc.). The coherence between sections of prefrontal OxyHb oscillations in five frequency intervals (I, 0.6-2 Hz; II, 0.15-0.6 Hz; III, 0.05-0.15 Hz; IV, 0.02-0.05 Hz; and V, 0.0095-0.02 Hz) were analyzed using wavelet coherence. In addition, we analyzed the coherence of electroencephalography (EEG) signal in three frequency intervals (delta, 2-4Hz; theta, 4-7Hz; and alpha, 8-13Hz). We evaluated the outcomes using Glasgow Coma Scale (GCS) cores at discharge. The patients were categorized into three groups of normal control, good outcome (GCS 13-15), and poor outcome

Among the included SAE patients (mean age, 63.1 years; and male, 57.4%), 5 patients (35.7%) had a good outcome. In the poor outcome group, phase coherence was significantly lower compared to good outcome and the normal groups, especially for the myogenic frequency interval III (0.63±0.16 vs. 0.68±0.19 vs. 0.28±0.06, P <0.001, respectively). However, the phase coherence of EEG signal was similar in two groups.

Our results demonstrated that the lower phase coherence of OxyHb in the myogenic signal, which originated from the vascular smooth muscle cells in the brain, was related to the poor outcome in SAE patients. This suggests that evaluating cerebral dysfunction using wavelet coherence of OxyHb could be a useful outcome predictor following SAE.

External ventricular drain (EVD) placement is a common procedure in the neurointensive care unit and intracranial hemorrhage (ICH) is a recognized complication. In this study we sought to determine the factors associated with ICH development after EVD placement.

Retrospective study performed at a tertiary hospital. We identified all patients in whom an EVD was placed over a 13 month period. Electronic chart review was done to obtain basic demographics, past medical history, use of antiplatelets/anticoagulants, type of catheter placed and presence of intracranial hypertension (IH). Computed tomographies were reviewed to identify EVD-associated ICH. ICHs were classified into symptomatic (GCS decline > 2 points, intubation, outcome of death, or new focal continuous variables were analyzed with a proportion of the means test.

The sample was comprised of 106 subjects, 36 had EVD-associated ICH. The median age was 59 years. There was no significant difference in race or gender between patients with ICH and those without ICH. Age, catheter type, history or inpatient use of anti-thrombotics, recent surgery, tPA use, heparin use, history of hypertension, hospital outcome, prior stroke, symptomatic hemorrhages, and ICP spikes were analyzed, but only age (61.1 hemorrhage and 53.6 non-hemorrhage, p = 0.023), history of antithrombotic use (19/36 hemorrhage and 16/70 non-hemorrhage, p = 0.004) and ICP spikes (26/36 hemorrhage and 35/70 non-hemorrhage, p = 0.038) were significantly associated with ICH occurrence.

Three significant factors were associated with tract hemorrhages; age, history of anti-thrombotic use, and ICP spikes. Two of these factors have been previously supported by prior studies however, no prior study has correlated ICP spikes to EVD hemorrhages. Additional studies may further validate the association between ICP spikes and EVD-related tract hemorrhages.

Targeted temperature management(TTM) aimed at helping to improve neurological outcomes associated with ischemic stroke have been studied continuously. However, it is not well known whether the parameters in TTM initiation, induction, maintenance will affect neurologic prognosis.

We restrospectively reviewed medical records of the patients with large hemispheric infarction(LHI) who underwent TTM at SNUBH neurological intensive care unit from 2011.02.13. to 2015.12.31. Onset to TTM initiation, induction period, TTM maintenance duration were investigated and dichotomized. Neurologic prognosis was determined by the 3 month death and modified Rankin Scale(mRS).

A total of 49 patients were included in the study. Longer onset to TTM initiation(>24 hours) was associated with less 3 month death. Shorter TTM induction period(<=2 hours) was associated with less death rate, more fair outcome(mRS 0-4). TTM maintenance duration(within 5 days or more) was not statistically correlated with neurologic prognosis.

Shorter TTM induction period may reduce death in LHI through maximizing ICP control effect. The high mortality rate in patients with shorter onset to TTM initiation is likely to be related to the severity of initial symptom(mean NIHSS 20 vs 16).

Non-pulsatile continuous blood flow can cause endothelial dysfunction and small vasculature injury. The impact of non-physiologic blood flow on cerebral autoregulatory function and brain injury has not been extensively studied. We report a case of posterior reversible encephalopathy syndrome (PRES) in a patient supported by a continuous flow pump, venoarterial extracorporeal membrane oxygenation (ECMO) for acute cardiogenic shock secondary to iatrogenic ventricular septal defect (VSD).

A 74 year-old male with hypertrophic cardiomyopathy was admitted for elective septal myectomy with an ascending aorta and hemi-arch replacement. The surgery was complicated by an iatrogenic VSD requiring urgent VA-ECMO cannulation for cardiogenic shock. On day 6, CT brain achieved for poor neurological examination revealed extensive bilateral parietal, occipital and cerebellar hypodense lesions consistent with the typical imaging features of PRES. A repeat CT brain on day 7 depicted further extension of brain injury to the bilateral frontal lobes. Due to worsening neurologic status, the decision was made to place an intracranial pressure monitor and lower the ECMO flow to return to a pulsatile flow state. The patient was closely monitored for improvement with PaCO2 levels, serial CT scans, and neurologic examinations. Repeat CT scans on POD 10 and 13 depicted improvement in the bilateral cytotoxic edema with PaCo2 levels improving to 34 -38mmHg at a reduced ECMO flow rate of 2.2 -3.3L/min. His neurologic examination also improved with spontaneous movements noted in all four extremities. Although neurologically cleared for heparin loading, he remained too hemodynamically unstable for open surgical repair and his surrogate decision makers decided to withdraw life-sustaining therapy.

Our case report illustrates the limited knowledge on the consequences of ECMO's impact on cerebral dynamic cerebrovascular autoregulatory changes in real-time that occur with patients with continuous flow pumps.

Hospital-onset unresponsiveness (HOU) may occur in patients hospitalized for non-neurological conditions; while HOU tends to be a transient systemic event, it may also indicate underlying neurological problems. Quantitative pupillometry provides NPi (neurological pupillary index), a quantitative measurement of pupillary light reflexes that have been traditionally assessed via subjective visual impression. We determined the clinical usefulness of NPi in predicting the outcomes of patients who have experienced HOU.

HOU was defined as a newly developed altered mental status and cases coded as "unresponsive" in the ACDU (alert, confused, drowsy, and unresponsive) scale. We analyzed the demographics, radiological findings, etiology of HOU, NPi, in-hospital mortality, and 3-month modified Rankin Scale (mRS) scores.

A total of 345 cases in 331 patients were analyzed, out of which 214 cases (62%) had been assessed with quantitative pupillometry. Cerebral herniation syndrome (CHS) was found in 52 (15%) cases; higher NPi was associated with decreased risk for CHS (odds ratio, 0.61; 95% confidence interval [CI], 0.26-0.77; p=0.003), and no other factors were associated with the risk of CHS. A total of 102 (30%) cases showed in-hospital mortality. After controlling for clinical covariates and the presence of CHS, lower NPi was independently associated with increased risk for in-hospital mortality (odds ratio, 0.52; 95% CI, 0.29-0.94; p=0.02). At a cutoff value of 1.8, the specificity and sensitivity of NPi for predicting in-hospital mortality were 92% and 50%, respectively. Multivariate analysis showed an independent association between lower NPi and unfavorable clinical outcomes (common odds ratio, 0.64; 95% CI, 0.51-0.80; p=0.02).

NPi, a quantitative index of pupillary light reflex, was significantly associated with the risk of cerebral herniation and in-hospital mortality in non-neurological patients with HOU. Measuring pupillary light reflexes through quantitative pupillometry may be useful when responding to HOU cases.

Target Temperature Management (TTM) improves survival and neurologic outcome and is recommended for cardiac arrest (CA) survivors by international guidelines. Shivering is both an anticipated consequence and a major adverse effect of TTM. The Bedside Shivering Assessment Scale (BSAS) is a simple, validated four-point scale that enables repeated quantification of shivering at the bedside. In this study, we examine the association between time to return of spontaneous circulation (TTROSC) and shivering (defined as BSAS > 0).

Data on 30 post-cardiac arrest patients treated with TTM were collected from APACHE Outcome database and medical records. Baseline characteristics included age, APACHE III scores, TTROSC (minutes), time to target temperature (TTT, minutes), and BSAS > 0 (percentage of hours BSAS > 0/total number of hours BSAS was done). Outcome was survival to hospital discharge with good neurologic outcome. Group 1 and Group 2 included patients with TTROSC below or above the median respectively. All patients received continuous infusions of fentanyl and sedatives (propofol, midazolam, and/or dexmedetomidine) as per our institution's protocol.

Compared to Group 1 (n = 15), Group 2 (n = 15) had similar age (57 ± 21 vs 59 ± 17, p = 0.8), similar APACHE III scores (104 ± 28 vs 104 ± 33, p = 1.0), longer TTROSC (27 ± 12 vs 5 ± 2, p = 0.0001), similar TTT (485 ± 181 vs 439 ± 353, p = 0.7), more shivering (5.8 % vs 3.2 %, p = 0.03), and similar survival with good neurologic outcome (20 % vs 27 %, p = 0.7) respectively. TTROSC was strongly positively correlated with shivering (Pearson correlation coefficient, r = 0.78).

In comatose survivors of cardiac arrest who received TTM, longer TTROSC (indirect measure of brain injury) was associated with more shivering. These findings should be further investigated in prospective studies.

Pupillometry assessment of the pupillary light reflex (PLR) is gradually replacing manual PLR assessment. This new technology has led to a recent increase in clinical research and subsequent need to validate those results. McNett et al. recently investigated the association between intracranial pressure (ICP) and serial pupillometer values and found that pupillometry readings are different significantly in the setting of increased ICP. This is a replication of the McNett study in a larger multicenter cohort to explore these findings.

Data from the Establishing Normative Data for Pupillometer Assessments in Neuroscience Intensive Care (END-PANIC) Registry include over 3,000 patients with a neurological condition. 273 subjects with documented ICP readings provided 16,221 observations (daily mean ICP values) which were included in this analysis. Statistical analysis (SAS v9.4) included descriptive statistics and to examine the differences .

Subject mean age was 53 years, 52% were female and 76.6% were Caucasian. Student t-test analysis was used to explore for differences. Excepting latency and right eye NPi, lower PLR values were associated with higher ICP (compared to low or normal ICP) for all mean pupillometer/PLR variables for both left and right eyes (t range [-9 .78 to 33.67]; p-value range [<0.0001 to 0.03]).

The findings confirm and extend those of McNett. Patients with increased ICP tend to have lower pupillometer readings. Automated pupillometer is a non-invasive method that provides prediction of the ICP trends which can help neurocritical care professionals in assessing patients with neurological conditions.

Encephalopathy is a common complication in cirrhotic patients. Clinical manifestations are diverse, but few data are available on pupillary abnormalities in such patients. The aim of this study was to evaluate whether automated pupillometry could detect pupillary dysfunction in this patients' population.

Prospective ongoing study including the assessment of the pupillary changes to light stimulation using automated pupillometry (NeuroOptics, Irvine, USA) in adult cirrhotic patients after ICU admission. The degree of encephalopathy was scored by the Glasgow Coma Score (GCS). Severity of cirrhosis was assessed by the Child-Pugh and MELD scores. Severity of liver encephalopathy was assessed according to standard criteria. Different biological variables, including ammonium (NH4), was measured to pupillary assessment. The median values of pupillometry-derived variables were collected for both eyes. -Pugh and NH4 levels were found with any of the pupillometry-derived variables. No differences in pupillometry-derived variables were observed across different degree of liver encephalopathy.

Automated pupillometry did not show correlations between pupillary abnormalities and the severity of critically ill patients with liver cirrhosis.

Prognostication in comatose survivors of cardiac arrest (CA) remains challenging. The purpose of this study was to determine if early quantitative analysis of resting EEG can improve prediction of commandfollowing by post-CA day 10.

We prospectively enrolled patients admitted after CA. Clinical care was performed according to our institutional protocol, which includes continuous EEG monitoring. 20-minute resting EEG epochs were clipped daily; clips were excluded if seizures or other confounders were present. Epochs from post-CA days 1-6 were preprocessed for artifact reduction, then analyzed for three quantitative metrics: power spectral density, permutation entropy, and coherence. We created a predictive model using partial least squares regression analysis to distinguish EEG data as from patients who would or would not recover command-following by post-CA day 10. Cross-validation of results was accomplished with a 1000-times random assignment of 50% of data as training set and 50% as testing set.

368 EEG clips were analyzed from 98 patients (59.1% female, age 57.8+/-1.8 years, pre-morbid mRS 1.7+/-0.2 and CPC 1.4+/-0.1). Cardiac arrests occurred out-of-hospital in 52%, witnessed in 80.4%, and had bystander CPR in 74.5%. Mean time to ROSC was 19+/-2 minutes, 31.4% had a shockable initial EKG rhythm, and 74.5% of patients received therapeutic hypothermia. Prior to day 10, 39.2% regained consciousness and 19.4% had withdrawal of care. Using EEG data alone, predictive ability (expressed as average area under the receiver operating characteristics curve) yielded AUC 0.773+/comparison, the same model was constructed using clinical features (absence of pupil and corneal reflexes by day 5) or laboratory testing (peak NSE level). The model combining clinical, laboratory, plus EEG data yielded AUC 0.835+/-0.029, an improvement vs clinical features (AUC 0.720+/-0.002, p<0.001) or NSE levels (AUC 0.750+/-0.03, p<0.001) alone.

Quantitative EEG analysis may provide adjunctive prognostic information regarding short-term recovery of consciousness.

International Guideline recommended pupillary light reflex (PLR) and/or cortical response (N20) to short-latency somatosensory evoked potentials (SSEPs) at 72 hours after return to spontaneous circulation as the only strong predictors of unfavorable outcome in comatose patients after cardiac arrest. The aim of this study was to compare this algorithm with a multimodal approach including other prognostic tools.

Post hoc analysis of an international multicenter (n=10; n=456 patients) prognostic study on automated pupillometry in comatose post-CA patients. The primary study endpoint was the accuracy of NPI in predicting 3-month unfavorable neurological outcome (UO), defined as Cerebral Performance Category (CPC) of 3-5 (severe disability, unresponsive wakefulness or death). Patients with findings on PLR, SSEPs, NPI and EEG were included; the highest NSE was also recorded, whenever available. An NPI <2 on day 1, a discontinuous EEG background or clinical myoclonus over the first 3 days, bilaterally absence of N20 calculated as: false positive / favorable outcome.

We included 186 patients; 131 (70%) of those had UO. Using the approach of Guidelines, unfavorable outcome at day 3 was observed in 41/44 patients with absent PLR and 63/63 with absent N20; 80/131 (61%) patients with UO were identified. Using the multimodal approach, UO was identified in 15/15 patients with NPI <2, in 67/81 patients with discontinuous EEG, in 50/52 patients with myoclonus and

This study suggests that a multimodal approach, including NPI, EEG, SSEPs and NSE, could identify a

After physicians introduced the idea to declare death based on loss of brain functionality, many countries incorporated brain death into their legal criteria for death. We sought to learn about the global legal perspective on brain death declaration (BDD).

We collected legal documents about declaration of death around the world by searching national legislative databases and Google. We utilized Google Translate to convert all documents into English then searched for references to criteria for BDD. In cases where there was conflicting information, we consulted local experts.

We located legal documents on death declaration for 49 countries, 45 of which included a reference to brain death. Legally stipulated criteria for BDD were identified for 32/49 countries. With respect to prerequisites for BDD legal stipulations existed in: 28/32 countries on confounders to exclude, 16/32 countries on an observation period before BDD, 19/32 countries on the minimum temperature for BDD and 8/32 countries on the minimum blood pressure for BDD. An assessment for coma was legally required in 30/32 countries. The fact that spinal reflexes do not preclude BDD was included in the legal criteria for BDD in 12/32 countries. A broad reference to an assessment for brainstem areflexia was legally mandated in 27/32 countries. The legal criteria included specific reflexes to test in 25/32 countries (pupillary 23/32, corneal 23/32, oculocephalic 18/32, oculovestibular 24/32, gag 20/32, cough 22/32, and other 4/32). Every country legally required an assessment for the inability to breathe spontaneously, but only 17/32 described apnea testing in detail. The number of clinical exams required for legal BDD ranged from 1-5. Ancillary testing was legally required in 9/32 countries.

The legally stipulated criteria for BDD differ around the world. Standardizing the global legal perspective on BDD would help prevent 1) variability in practice and 2) false BDDs.

Up to 65% of patients monitored with pupillometry during therapeutic temperature management (TTM) after cardiac arrest will have sluggish (SL) or non-reactive (NR) pupils. The neuroimaging findings and injury patterns of these patients have not been reported.

115 adult patients treated with TTM after cardiac arrest with available pupillometry data from the NeurOptics NPi-200 were studied. Discharge outcome was classified as poor (PO) if the Cerebral Performance Category score was 3-5, and as good if 1-2. Pupil size, percent constriction, and constriction velocity were determined throughout TTM using data from the worst eye at each assessment. The Neurological Pupil index (NPi) was scored from 0 (NR) to 5 (brisk), with values <3 considered SL. Computed tomography (CT) and magnetic resonance (MR) neuroimaging was reviewed by a neuroradiologist blinded to pupillometry and outcome data.

Poor outcomes occurred in 34/35 (97%) patients with NR pupils during TTM, 22/34 (65%) patients with SL pupils, and 25/46 (54%) with normal (NL) pupil reactivity. Pupil size did not predict outcome, but pupillometry data during TTM predicted poor outcome with AUC 0.73-0.78. When nonreactive pupils were first detected, 24/35 (68%) were <5 mm. 44% of patients had CT imaging, and 30% had MR imaging a median of 102 (IQR 78-123) hours after recovery of spontaneous circulation. Cerebral edema or herniation were identified in 7/19 (37%) NR vs 1/22 (5%) SL and 0/21 NL patients (p<0.001). Midbrain injury identified by T2 sequences was identified in 15/41 (36%) NR/SL patients versus 3/21 (14%) NL patients (p=0.06). Midbrain abnormalities were identified more often in patients with NR/SL pupils than edema/herniation (36% vs 20%, p=0.003).

A minority of patients with sluggish or non-reactive pupils after cardiac arrest have evidence of cerebral edema or herniation. Midbrain injury is a more common mechanism to explain this common neurologic deficit.

Cardiac arrest (CA) survivors are often comatose and their arousal recovery is dependent on the extent of hypoxic-ischemic injury (HII). Long-term neurologic outcomes are variable, difficult to predict, and biased by withdrawal of life-sustaining therapy. Somatosensory evoked potentials (SSEP) remain the gold standard for predicting arousal potential, but is not broadly available. We hypothesized that early hi-resolution MRI may help assess arousal recovery potential as predicted by electrophysiologic outcome.

Comatose survivors of cardiac arrest admitted to an ICU between June 2015 and January 2018 who underwent SSEP and MRI were retrospectively identified. 3D-HII burden in 16 predefined regions. Semi-automated region-of-interest (ROI) tools in MIPAV were used to draw borders on DWI around the upper brainstem including the ascending reticular activating system (ARAS) to assess voxel intensity and derive HII volumes. Our outcome of interest was SSEP findings classified in two prognostic categories: indeterminate (bilaterally present N20s or unilateral presence of N20s) and poor prognosis (bilaterally absent N20s). We used paired t-tests to compare presence of signal abnormality and ROIs between patients with SSEPs predicting poor outcome or indeterminate prognosis.

42 consecutive CA survivors (mean age of 51.3, 31% female) were included. No significant differences were noted in baseline characteristics between groups though time to ROSC was noted to be 9 vs 20 mins for indeterminate and poor outcomes (p = 0. extent did not predict SSEP status. No significant difference was noted in the voxel intensities on ADC in the midbrain or pontine tegmentum.

Quantitative MRI measures of HII extent may be superior in predicting arousal potential in comatose survivors of CA compared with manual rating. A quantitative image analysis pipeline is being developed for measuring ARAS lesion burden and predicting electrophysiologic based outcomes in CA.

Despite promising preclinical results, the application of intra arrest therapeutic hypothermia (IATH) during cardiopulmonary resuscitation have produced controversial results in clinical trials. The aim of this review was to analyze the effects of such therapy on relevant outcomes in patients suffering from out-of-hospital cardiac arrest (CA).

The following databases have been searched up to 26th May 2019 for human trials: PubMed (from 1966), EMBASE (from 1974), CINAHL (from 1982), the Cochrane Library (from 1974) and Ovid/MEDLINE (from 1966). The search strategy will use the following terms: "arrest" OR "cardiac arrest" OR "heart arrest" AND "intra arrest" OR "during CPR" OR "intra CPR" AND "hypothermia" OR "therapeutic hypothermia" OR "cooling". References from identified studies and relevant review articles have also been searched for additional eligible citations. The search has been limited to English publications and has been conducted in accordance with the International Liaison Committee on Resuscitation (ILCOR) process of evidence evaluation.

A total of six human studies (n=2869; 1386 treated with IATH) including four randomized controlled trial (LOE 1), one retrospective and one prospective controlled study (LOE 3) were identified. Two studies used trans-nasal evaporative cooling and 4 others intravenous cold fluids. Overall rate of return of spontaneous circulation was similar between IATH patients and controls (542/ ) when compared to control group. No differences were found in the subgroup of shockable vs non-shockable rhythms. Different effects on outcomes were observed according to the method used to induce IATH when compared to controls.

IATH was not associated with improved outcomes when compared to standard of care. However, the method used to induce IATH may potentially influence the beneficial effects of such intervention.

Amantadine may improve functional recovery in the subacute state following brain injury. We aimed to characterize EEG signatures in patients with acute brain injury (ABI) receiving amantadine that did and those that did not recover consciousness.

We studied a consecutive series of 44 patients with acute brain injury patients who were treated with amantadine as a neurostimulant between September 2013 and December 2017. All patients were initially comatose and underwent EEG prior to and after the initiation of amantadine. The ability to follow commands was assessed daily based on prior published methodology (Claassen et al, AnnNeurol 2016). EEG features that were assessed included sleep stages, posterior dominant rhythm (PDR), and power spectral density plots. We applied a multivariate regression model using generalized estimating equations (GEE) to identify EEG features correlated with recovery of command following. EEGs were analyzed by a board certified neurophysiologists.

-free EEG clips), 30 patients (68%) recovered consciousness during hospitalization. ICH was the most common etiology in 22 (50%) patients, followed by SAH in 7 (16%) patients. On average amantadine was given for 11 +/-8 days. 14 patients (32%) had seizures, only 2 patients (14%) after starting amantadine. In our GEE model, age (p=0.01), sleep structures (p=0.02), PDR (p=0.03), and cumulative dose of amantadine (p=0.03) were all associated with recovery of command following. Spectral features corresponding to higher levels of anterior forebrain corticothalamic integrity correlated with higher levels of consciousness in 100% of recorded patients after 7 days of amantadine use. The best spectral pattern per patient was seen 1.5 days on average prior to recovery of consciousness.

EEG may provide a biomarker that indicates subsequent recovery of consciousness in unconscious patients with an acute brain injury that are treated with amantadine.

Depletion of cerebral glucose (i.e., cerebral glucopenia) occurs commonly and is associated with poor outcome in traumatic brain injury and subarachnoid hemorrhage. However, the incidence of cerebral glucopenia after diffuse hypoxic-ischemic brain injury (HIBI) is unknown. We characterized the burden of cerebral glucopenia after HIBI and its association with markers of physiological distress and outcome.

We retrospectively analyzed cerebral microdialysis data from a cohort of 22 patients with HIBI. 21 patients survived sudden cardiac arrest and 1 patient had severe hypoxia after polysubstance overdose. Hourly values of cerebral glucose, lactate, pyruvate, and glycerol as well as continuous intracranial pressure (ICP), arterial blood pressure (ABP) and interstitial brain oxygen (PbtO2) were recorded. Associations between average glucose/patient-day versus average lactate:pyruvate ratio, glycerol, ICP, PbtO2, and ABP were analyzed using linear regression. Burden of glucopenia (defined % time with glucose < 0.7 mmol/L) was analyzed by patient-day. The relationship between glucopenia burden and discharge outcome was analyzed using the Wilcoxon rank sum test.

Lower cerebral glucose was associated with higher cerebral glycerol (p=0.005), higher LPR (p=0.0002), higher ICP (p<0.0001), and lower PbtO2 (p=0.012) levels. There was no association between ABP and cerebral glucose (p = 0.12). Glucopenia burden increased progressively over time and peaked by postinjury day 8. 5/22 patients had good outcome (defined as return of consciousness prior to discharge). There was no association between outcome and cerebral glucopenia burden (p = 0.4351).

Cerebral glucopenia is common after HIBI and associates with markers of cellular distress. The burden of cerebral glucopenia progressively increases over several days and appears to peak more than 1 week after injury. Although there was no association between outcome and glucopenia burden, the number of patients in this study with good outcome was low. The utility of cerebral glucose monitoring after HIBI merits further study.

International Guideline recommends using bilaterally absence of pupillary light reflex (PLR) and/or bilaterally absence of the cortical response (N20) to short-latency somatosensory evoked potentials (SSEPs) at 72 hours after return to spontaneous circulation to predict unfavorable outcome in comatose patients after cardiac arrest. The aim of this study was to compare this algorithm with a multimodal approach including other prognostic tools.

Retrospective study of adult (>18 years) cardiac arrest patients admitted from January 2016 to March 2019 and who underwent multimodal monitoring. We collected demographic characteristics and cardiac arrest data, together with SSEPs, the presence of burst-suppression on early EEG, a neurological pupillary index on the automated pupillometry <2 at 24 after arrest and a neuron-specific enolase (NSE) 3-month unfavorable neurological outcome (UO) with Cerebral

We included 143 patients; 104 (73%) of those had UO. Using the approach of Guidelines, unfavorable outcome at day 3 was observed in 7/14 patients with absent PLR, in 16/16 with absent N20 and 14/14 with combined absent pupillary light reflex and N20; 51/104 (49%) patients with UO were identified. Using the multimodal approach, UO was identified in 30/30 patients with NPI <2 and 36/37 patients with BS on EEG. Among the others, UO was associated with absent N20 in 4/4 patients and with high NSE values in 14/18 patients. This approach identified 84/104 (81%) patients with unfavorable outcome. The Area Under Curve to predict UO for the approach of Guidelines was 0.66, which increased to 0.80 with the multimodal approach.

This study suggests the a multimodal approach, including NPI and BS on EEG, SSEPs and NSE, has a higher predictive value for UO than recommended predictive tools.

There is a high prevalence of seizures following cardiac arrest (CA), but not well studied among survivors with good neurological recovery. We describe the prevalence of clinical and electrographic seizures, anti-epileptic use, and EEG characteristics of CA survivors with good neurological outcomes.

Adults with return of spontaneous circulation (ROSC) after in-hospital or out-of-hospital CA between 9/2015-2/2019 were eligible. A consecutive sample of survivors with included. Prevalence of seizures and antiepileptic drugs (AED) use within 12-months after discharge were collected using a questionnaire administered via in-person or phone. A board-certified clinical neurophysiologist reviewed the EEG.

Of 148 patients surviving to discharge, 93 (63%) with 12-months follow-up were analyzed. Average age was 56±16 years, 32 (34%) were women, 81 (87%) patients had witnessed arrest, 44 (47%) received defibrillation, with an average ROSC duration of 12±12 minutes, and a median CPC of 2 at discharge. There were no clinical seizures reported during hospitalization. Of available 63 (68%) patients with raw EEG (median duration of 4 days), only 2 (3%) patients had electrographic seizures, 47 (74.6%) had continuous background as their best EEG pattern, 6 (10%) with discontinuous background, 5 (8%) with epileptiform discharges, and 2 (3%) patients had burst suppression pattern that recovered later to a normal EEG pattern. None of the patients had any malignant EEG patterns, 58 (92%) exhibited reactivity to a verbal or tactile stimulation and 38 (60%) had the presence of sleep structures and posterior dominant rhythms. Surprisingly, 20 (13%) patients were discharged on an AED. Clinical seizures and AED use were reported in 2/93 (2%) at 12-months follow up.

Both short and long-term seizure burden are very low among the cardiac arrest survivors with good neurological recovery. Underlying factors related to high utilization of AED before discharge warrants further investigation.

Objective: Early neuro-prognostication in the intensive care unit pediatric patients is essential to enable effective care planning, triaging level of care, and family support. In coma, the reliability of biomarkers such as electroencephalogram (EEG), anatomical neuroimaging to determine potential for consciousness and future functional capacity are less established in children. Herein we present two case studies highlighting resting state functional MRI (rs-fMRI) as a clinically new means defining real-time brain function in the pediatric critically ill population.

rs-fMRI measures spontaneous low-frequency fluctuations in the blood oxygen dependent (BOLD) signal to investigate the networks of the brain. A standardized acquisition of data on a 3 Tesla MRI under light tool MELODIC. Whole brain networks determined by independent component analysis with false discovery rate at p<0.05 to detect major brain networks.

Cases describe two critically ill children. One, with severe brain injury related to acute necrotic encephalopathy, and the other with diabetic ketoacidosis induced cerebral edema and uncal herniation. Both had slow EEG background with sleep features approximately a week after presentation and were comatose by exam on the day of rs-fMRI. Rs-fMRI detected normal brain function in the long-range fronto-parietal network, intact language-area networks, and default mode network. Atypical networks were detected in brainstem and deep grey in both children. By hospital discharge, both children were awake and communicative with spontaneous movements. Case one remain with tracheostomy with intermittent ventilation, case two had residual left hemiparesis, vision and language intact, mild cognitive deficits.

In the cases reviewed, rs-MRI may offer an objective measure of functional brain capacity and potential for meaningful recovery with preservation of language and long range connectivity networks in critically ill pediatric patients.

Provision of positive end-expiratory pressure (PEEP) through a conventional ventilator during apnea testing for brain death determination removes the need for additional equipment such as a PEEP valve, allows for use of high PEEP during apnea in patients with severe hypoxic respiratory failure and facilitates detection of respiratory effort on flow scalars. The advent of ventilators that permit deactivation of the apnea backup setting has made such testing possible. Our goal was to examine the feasibility of PEEP use with conventional mechanical ventilation during apnea testing, with a focus on premature termination and inadvertent external triggering.

performed without disconnection from the ventilator (Dräger Evita® Infinity® V500), with deactivation of the apnea backup. This was a convenience sample based on availability of appropriately trained -Support and PEEP 5-15cmH2O. Apnea was confirmed by absence of chest rise and respiratory effort on the flow scalar. Adequacy of respiratory stimulus was established by a CO2>60mmHg and 20-point CO2 rise from baseline. Endpoints included early termination of the apnea test prior to 10 minutes because of patient instability, any oxygen desaturation below 90% and inadvertent external triggering. Inadvertent external triggering required repeat of apnea testing.

Ten patients underwent apnea testing while connected to the ventilator. Apnea testing for at least 10 minutes was successful in all patients. Apnea was confirmed in all cases. No patient suffered oxygen desaturation below 90% or other instability. There was one instance of inadvertent external triggering caused by jostling of tubing, necessitating repeat testing.

Apnea testing with provision of PEEP through a conventional ventilator to improve tolerance is feasible. Inadvertent external triggering is uncommon but may occur.

Despite well-defined AAN guidelines on brain death declaration, there is marked variability in its practice nationally. This highlights the need for targeted brain death education initiatives. Communication with surrogates or families about a brain death diagnosis and its implications is integral to brain death declaration, yet this has not been studied in a simulation setting. We developed a brain death simulation curriculum at our institution addressing knowledge and surrogate communication skill development.

As part of this curriculum, multi-disciplinary critical care fellows completed a pre-curriculum multiple choice (MC) knowledge test and survey (Likert 1-10 scale) evaluating comfort and confidence. A mandatory one-hour neurocritical care attending-led didactic regarding guidelines and technical aspects of brain death examination was conducted. Subsequently, each fellow performed an observed brain death examination (SimMan3G mannequin) with feedback followed by a standardized family scenario with delivery of a brain death -simulation survey, MC questions, and provided feedback. Statistical analyses used 2-tail Wilcoxon signed rank test (p<.05).

Thirteen critical care fellows participated (neurology[1], anesthesia[5], trauma[4], pulmonary[3]). Only one fellow had previous formal brain death training with the majority [80%, (N=12)] only participating in 0-5 brain death declarations. There was significant improvement across all measures: self-rated knowledge (5.6 to 7.9, pre-simulation to post-simulation, p= 0.004), knowledge relative to peers (48% to 81%, p=0.002), confidence (4.5 to 8.1, p=0.001) and comfort (4.3 to 8, p=0.001) with performing a brain death exam, and comfort with family discussion (6.5 to 8.6, p=0.01). Test scores improved from 56% to 73% after simulation (p=0.004). All fellows found the curriculum beneficial (with all aspects wellreceived).

Critical care fellows may lack experience with brain death declaration. Didactics coupled with simulation-based education can improve objective knowledge and comfort with brain death declaration and surrogate communication.

There is a growing disparity between availability and demand for neurologic expertise, particularly in smaller community hospitals. Telemedicine has helped to bridge this disparity with respect to cerebrovascular disease and is used increasingly to deliver other types of neurologic expertise to patients. While the NIHSS is widely used in telestroke, other formalized neurologic exams have not been well studied. We seek to determine whether the components of a brain death exam can be reliably performed via telepresence.

Patients suspected of meeting brain death criteria were enrolled from July 2017 to May 2019. Standard bedside neurologic exam (BNE) performed by the attending neurointensivist in accordance with our institutional protocol was compared with the telepresence neurologic exam (TNE) performed by a study neurointensivist blinded to the findings of the BNE and a trained bedside assistant. We analyzed the agreement between examiners regarding findings of coma, corneal reflex, pupillary light reflex, oculovestibular reflex, oculocephalic reflex, cough, gag, motor response, and apnea.

We enrolled 7 patients over 22 months. Proximate causes were intracerebral hemorrhage (5/7), anoxic brain injury, (1/7), and cerebral infarction (1/7). All examination components performed in the BNE could be completed by TNE. In 3 cases, neither examiner could assess all exam components. In 2 cases spinal cord injury precluded oculocephalic testing. In 1 case refractory hypoxia precluded apnea testing. BNE and TNE agreed in 100% of testable components. In 2 cases testing pupillary light reflex was reported as difficult in the TNE but not the BNE. All telepresence examiners reported high confidence that the exam findings were consistent with brain death.

Preliminary findings from our pilot study suggest that the use telepresence for brain death examination Introduction Traumatic brain injury (TBI) is often followed by the loss of con increases each day following the injury, but the contents of consciousness, also known as qualia, do not uniformly return. While there is some information about brain regions supporting arousal, less is known about circuits encoding contents of consciousness. Some evidence supports a role for the thalamus in consciousness, but it is controversial whether it supports arousal, or has a more nuanced role in consciousness. To address this question, we combined intracranial recordings in patients recovering consciousness with neuroimaging of thalamocortical circuits.

Electrophysiology We recorded electrocorticography (ECoG) from prefrontal cortex and anterior cingulate cortex, as well as scalp electroencephalography (EEG) from a standard 10-20 montage, during singleand parietal cortex based on coherence between the evoked responses in these regions when ACC was stimulated. Radiology. Regions of structural damage were extracted from the post-TBI MRI and diffusion tensor imaging (DTI) radiographs. Tractography using DSI Studio™ was performed with seed regions placed in the bilateral mediodorsal nucleus of the thalamus.

We found that in patients with injury isolated to the cortex and/or white matter, the cortico-cortical functional connectivity across frontoparietal networks was preserved, and these patients recovered consciousness. However, a patient with thalamic injury failed to recover consciousness, despite an increased level of arousal following injury. The functional connectivity across cortical regions was drastically lower following thalamic injury, even when the cortical damage was minimal.

We propose that integration and communication of information across frontoparietal networks, which is required for contents of consciousness, is dependent on thalamic input. Thus future efforts have to be focused on restoring this input.

Brain herniation is a deadly event that requires rapid administration of hyperosmotic agents (HOAs) such as 23.4% NaCl. A recent retrospective study showed that Intraosseous (IO) cannulation provides a safe route for rapid administration of HOAs compared to central venous catheters (CVC) and peripheral intravenous catheters (PIV). Prospective study to measure the time-to-treatment for 23.4% NaCl or mannitol via IO, CVC, or PIV.

A data collection form ("brain code narrator") was created by nurses and providers to prospectively collect clinical data, hemodynamic measures, and time-to-treatment and administration route for HOAs during brain codes. In addition, demographics, diagnosis, serum sodium (Na+) and complete blood cell count, as well as immediate and delayed complications, and outcomes were collected.

Brain code narrator was used to collect data for 35 patients: 19 males with median (IQR) age 55(44-67) years. Diagnosis included intracerebral hemorrhage (n=16), subarachnoid hemorrhage (n=9), and other (n=10). All patients were intubated. Most patients were co-treated with induced hyperventilation. 23.4%NaCl (30cc) via CVC and IO route and mannitol (50 gm) via PIV were administered during 41, 18 , and 21 events with median time-to-treatments of 4(2,5), 3(1,3) and 15(14,18) minutes, respectively (p value <0.001 for all comparisons). No adverse events, such as hypotension or tissue injury were noted.

Preliminary data suggest that during brain herniation, administration of 23.4% NaCl via IO or CVC is more rapid than IV mannitol. IO cannulation for 23.4% NaCl may be an alternate route of administration of HOAs during brian code. Additional data will be provided regarding herniation reversal and long-term hematologic abnormalities.

Stress hyperglycemia is common in the critically ill and is associated with poor neurological outcomes in cardiac arrest patients. It is unknown whether glycemic dysregulation have different prevalence according to cardiac arrest etiology. We hypothesized that overdose-related cardiac arrest (ODCA) patients are more vulnerable to hypoglycemic events given the circumstances of arrest.

We retrospectively studied 243 cardiac arrest patients treated at two urban hospitals from the Multimodal Outcome CHAracterization in Comatose Cardiac Arrest (MOCHA) registry from 2011-2018. We examined glucose dysregulation (hypoglycemia blood glucose [BG]<80 mg/dL, hyperglycemia BG>180 mg/dL) within first 72h from arrest in ODCA and non-ODCA cohorts. Statistical analyses included paired/unpaired t-tests, Chi-al dysfunction was defined by scores of GOS-

Of the 243 patients, 76 (31.3%) were ODCA. There were no differences in BMI, gender, ethnicity, or therapeutic hypothermia (TH) treatment across cohorts, but ODCA patients were younger (43±15 vs 59±17 year-old; p<0.0001), had lower prevalence of diabetes (21.7 vs 40.5%; p=0.007) and lower hemoglobin A1c (5.8 vs 6.5%; p=0.02). Mean BG reduction from 0-24h to 48-72h in ODCA patients was significantly smaller (86.2±99.4 vs 148.2±106.5 mg/dL; p=0.006) despite no difference in mean peak BG. BG nadirs were lower in ODCA patients (77.8±28.8 vs 95.3±43.4 mg/dL; p=0.001). 113 patients developed glycemic dysregulation: 52(46%) ODCA vs 61(54%) non-ODCA; ODCA patients were nearly two times more likely to develop hypoglycemia (RR 1.9[1.2-3.1]; p=0.012) but had no increased risk of hyperglycemia (RR 1.0[0.8-1.2]). Among patients with glycemic dysregulation, ODCA was associated with higher risk of in-hospital death or neurological dysfunction (OR 2.9[1.0-8.4]; p=0.044).

Despite exhibiting blunted BG reductions to hyperglycemic treatment, ODCA patients were more susceptible to hypoglycemia in the first 72h postmanagement strategies should account for cardiac arrest etiology.

Sedation and neuromuscular blockade (NMB) in patients undergoing targeted temperature management (TTM) after cardiac arrest (CA) are recommended for patient discomfort and management of shivering. This study assessed the association between NMB use and neurological outcome in comatose survivors of CA who received TTM.

Data on 77 post-cardiac arrest patients treated with TTM were collected from APACHE Outcome database and medical records. TTM of 33°C or 36°C is chosen based on critical care physician's discretion. Baseline characteristics included age, APACHE III scores, time to return of spontaneous circulation (TTROSC, minutes), time to target temperature (TTT, minutes), and shockable rhythm (SR, %). Outcome was survival with good neurologic outcome.

Compared to the no NMB group (n = 22), the PRN NMB group (n = 41) and continuous NMB group (n = 14) had similar age (63 ± 15 and 59 ± 22 vs 56 ± 21, p = 0.1, 0.7),similar APACHE III scores (108 ± 33 and 106 ± 51 vs 111 ± 32, p = 0.7, 0.7), comparable TTROSC (16 ± 12 and 27 ± 24 vs 25 ± 22, p = 0.04, 0.8), longer TTT (526 ± 250 and 578 ± 300 vs 385 ± 158, p = 0.02, 0.02), comparable percentage of SR (17 % and 21 % vs 22 %, p = 0.4, 0.7), and similar proportion of patients with TT of 33 vs 36 (81 % and 87 vs 68 %, p = 0.3, 0.2) respectively. Survival with good neurologic outcome was achieved in 41 % in no NMB group vs 32 % in PRN NMB group (p = 0.5) and 36 % in continuous NMB group (p = 0.8)

In the present study, in comatose survivors of cardiac arrest who received TTM, use of NMB had no effect on neurologic outcome.

The apnea test is an essential examination for the determination of brain death. However, hypotension, hypoxemia, and other complications during the apnea test can affect the stability of brain-dead patients, as well as organ function for recipients. Therefore, it is necessary to establish standard guidelines for apnea testing.

The modified apnea test (MAT) comprises delivery of 100% oxygen through the endotracheal tube connected to manual resuscitator (Ambu® bag) with the positive end-expiratory pressure (PEEP) valve after disconnection of -nine instances of the conventional apnea test (CAT) were performed in 25 brain-dead patients; 77 instances of the MAT were performed in 39 brain-dead patients.

The mean duration of the apnea test was 3.5 ± 1.4 minutes in the CAT group and 3.0 ± 1.2 minutes in the MAT group. There were no significant changes in PaCO2, PaO2, or pH between the CAT and MAT groups (p = 0.341, 0.593, and 0.503, respectively). In overweight patients (body mass index prevented dramatic reductions in PaO2 and SaO2 (p < 0.05 for both). In the patients who had hypoxic brain injury due to hanging, differences in PaO2 and SaO2 in the MAT group were significantly smaller than in the CAT group (p < 0.05).

Although MAT, which was invented to maintain PEEP, was not efficient for all brain-dead patients, it could be helpful in selected patient groups, such as overweight patients or those who had hypoxic injury due to hanging. Clinicians should consider this reliable short-term apnea test.

Coma is a serious complication that currently has no good biological markers. The hypothalamus plays an important function in consciousness circuity. Orexin A/B, a neuropeptide produced in the hypothalamus has an excitatory effect on multiple target areas in the brain. Previous orexin studies ry (TBI), stroke and comatose states. The goals of our study: (1) the utility of orexin as a marker of coma recovery, (2) the correlation between orexin and recovery at 7 and 14 days, (3) correlation of orexin and Glasgow Coma Score/Score (GCS) over time,

A prospective, IRB approved study with a target n= 20 with a diagnosis of coma due to stroke, including hemorrhagic, and TBI, treated in the Neuro Critical Care Unit at Stony Brook University Ho collected from an external ventricular drain (EVD) and corresponding blood serum samples on Days 0, 7, and 14. There was no modification to the clinical treatment of individual patients.

dictive of whether patients recovered consciousness vs deteriorated. Logistic regression showed the relative risk of recovery vs. deterioration: , (95%CI -299.8±44.4, 72.3±10.2, respective p-values=0.4954e-39, 0.001epredictive of initial coma severity (GCS), with a correlation coefficient, R =0.338. Correlation between -0.964, -0.2881).

dictive of poor overall not appear as significant as the baseline level in predicting recovery.

There has been limited research over the past decade on how race impacts survival from cardiac arrest. It has been suggested that black patients are more likely to have unsuccessful resuscitation and lower rates of survival to discharge, however, it is unclear if this difference is secondary to hospital factors or patient specific factors. More research is needed on racial disparities in post-arrest outcomes at urban medical centers.

Multimodal outcome characterization in comatose cardiac arrest (MOCHA) is an IRB-approved multicenter observational study. This study sample consists of 191 consecutive cardiac arrest patients treated at two urban hospitals from 2011-2018. The sample includes both patients who experienced in-hospital and out-of-hospital cardiac arrest. The outcome of interest was in-hospital mortality. Associations between race and mortality were evaluated by chi-square and relative risk (RR) with 95% confidence interval.

We included 123 white (64%) and 68 black patients (36% were all found to be at no increased risk for in-hospital mortality relative to other gender and race combinations. There was no difference in location of cardiac arrest (i.e., inhospital vs. out-of-

The lack of racial differences in mortality could possibly be explained by the similar rate of out-ofhospital arrests, similar initial non-perfusing rhythms, lower socioeconomic status of all patients, and strong focus of the participating hospitals on addressing racial disparities in the healthcare system.

Hyperglycemia is associated with poor clinical outcomes in critically ill patients, such as post-cardiac arrest (CA) patients. Post-CA prognostication studies have studied clinical examinations, electrophysiology, biochemical changes, and/or neuroimaging, but studies regarding patient blood glucose levels are mostly limited to mortality outcomes. New analysis of glucose trends is needed to guide CA prognostication in order to determine favorable outcomes regarding neurologic functioning.

This study was conducted using the IRB-approved Multimodal Outcome CHAracterization in Comatose Cardiac Arrest (MOCHA) registry. The sample included 113 CA patients admitted to a university-affiliated urban hospital from 2014-2018. Case selection was determined by availability of serial glucose measurements over the first 24 hours post-CA and outcome scores at hospital discharge. Poor functional outcome was defined as Modified Rankin Scale (mRS) 3-6 or Glasgow Outcome Scale Extended (GOSE) 1-4. Statistical analysis included chi-square tests, and prognostic value was calculated by sensitivity.

There was no significant difference in outcome regarding age, sex, race, or ethnicity. The study sample consisted of 31% diabetic patients, with no significant difference in outcome. Patients with glucose levels >175 mg/dl at least once during the first 24 hours post-CA were associated with poor functional outc

There appears to be a correlation between glucose >175 mg/dl within the first 24 hours and poor functional outcome. However, it is still difficult to reliably predict poor vs. good functional outcome due glucose management are needed to better understand this relationship.

Post-cardiac arrest organ injury is associated with high mortality rate after ICU admission. Despite improvement in the post-cardiac arrest care, temporal changes in patients' severity, intensity of care and neurological outcome remain poorly defined. The aim of this study is to describe how epidemiology of cardiac arrest characteristics, therapies and outcome have changes over years.

Retrospective study including adult (>18 years) cardiac arrest patients admitted from January 2004 to March 2019 after CA to a University Hospital. We collected demographic characteristics and cardiac arrest data, together with main therapies and monitoring during ICU and hospital mortality. 

A total of 878 patients (median age 64 [52-75] years; male gender 68%) were included over the study period. Time to ROSC was significantly longer in period I and IV when compared to others (p<0.001). ICU length of stay and lactate levels on admission were also significantly higher in the period IV than others. There was a progressive and significant increase of out-of-hospital CA, non-cardiac origin of arrest and non-shockable initial rhythm from period I to period IV. Also, there was a significant increase in the number of patients developing acute kidney injury and hypoxic hepatitis over time, from period I to period IV. Despite a more frequent use of coronary angiography and multimodal neurological monitoring, hospital mortality increased (from period I, 50% to period IV, 70% -p< decreased (period I, 43% to period IV, 28% -p=0.01) over time.

In this study, severity of anoxic injury and the incidence of post-cardiac arrest organ dysfunction increased over time. This was associated with a higher proportion of patients with poor outcome.

Pressure reactivity index (PRx) based optimal cerebral perfusion pressure(CPPopt) is associated with outcome after traumatic brain injury, but is not explored after cardiac arrest. We examined post-arrest patients who underwent invasive intracranial monitoring to explore characteristics of PRx and CPP, and whether these were useful predictors of survival.

We included all comatose cardiac arrest patients without primary neurological pathology that underwent invasive intracranial monitoring between 2008-2018 at our institution. CPP, mean arterial pressure(MAP), PRx, CPPopt, and deltaCPP (CPP-CPPopt) were calculated. Systemic and brain physiologic measures were compared across the primary outcome of survival. 

In this pilot study we demonstrated the feasibility of acquiring CPP, PRx, and CPPopt for post-cardiac arrest patients. In this sample, none of the systemic and brain physiologic measures were associated with survival but the approach is limited by the bias towards poor outcomes in patients receiving monitors. Interestingly, CPPopt obtained from invasive intracranial monitoring generally ranged within physiologic norms. DeltaCPP for the single patient with good outcome was positive and small, consistent optimizing cerebral perfusion after cardiac arrest improves outcome are warranted.

Prognostication after cardiac arrest is challenge because of many confounding factors during hypothermia, Severity of the brain injury is a key determinant of whether maximal resources, such as the use of extracorporeal membrane oxygenation (ECMO), mechanical circulatory support, or even coronary artery bypass grafting, are advisable or appropriate. Therefore, early and accurate prognostication is essential for decision of therapeutic plan including maxima intensive modalities. In this study, we focused not only the prognosis estimation using MRI but also initial CT-based prognosis estimation where features captured by modern deep learning (DL) technique were commonly used.

We selected total 193 cardiac arrest patients having initial CT at ER, and brain MRI after 72 hours from cardiac arrest. Diffusion weighted image (DWI, b = 1000), and apparent diffusion coefficient (ADC) images calculated. Cerebral performance category (CPC) scores were used as the main outcomes of survivors after cardiac arrest. Both experienced neurologist and emergency medicine tried to predict the devised two cascaded deep convolutional neural networks (deep CNNs).

Even fully experienced neurologist and emergency physician could not predict the CPC score exactly with the initial CT scan only and even additional Diffusion MRI (Accuracy :16%-30% with initial CT only 60-70% with additional Diffusion MRI). By using DL technique, among 96 subjects of train set, 94 subjects had the correct prognosis score (97.9% accuracy) and among 97 subjects of test set, 96 subjects had the correct prognosis score (98.9% accuracy) with initial CT scans only. With additional diffusion MRI, 99.0% accuracy and 100% accuracy.

In visually equivocal initial CT scans, DL was more related to quantification than visual assessment. DL is superior and very useful for accurate prognostication especially with visually equivocal initial CT scan.

Cardiac arrest (CA) is associated with a high risk of dying and of neurologic impairment in survivors. Target Temperature Management (TTM) improves survival and neurologic outcome and is recommended by international guidelines. This study assessed the association between the initial Acute Physiology and Chronic Health Evaluation (APACHE) III score and neurological outcome in comatose survivors of cardiac arrest who received targeted temperature management (TTM).

Data on 77 post-cardiac arrest patients treated with TTM were collected from APACHE Outcome database and medical records. TTM of 33°C or 36°C is chosen based on critical care physician's discretion. Baseline characteristics included age, gender, APACHE III scores, time to return of spontaneous circulation (TTROSC, minutes), time to target temperature (TTT, minutes), and shockable rhythm (SR, %). Outcome included hospital mortality, and good neurologic outcome (defined as discharge to home or rehab).

Compared to the bad outcome group (n = 50), the good outcome group (n = 27) had similar age ( 

In comatose survivors of cardiac arrest who received targeted temperature management, the APACHE III score calculated in the immediate post-cardiac arrest period was a poor predictor of neurological outcome.

Brain dead patients are victims of trauma, entering the health care system through emergency department (ED).In the ED, these patients are received with injuries and de-arranged physiological conditions that depends on time sensitive treatment and have the potential for improvement with proper management. Our study tries to find out the predictors at admission that contributes to brain death (BD) so that their timely intervention can prevent BD

A retrospective analysis of the data related to severity of injuries, physiological parameters and laboratory investigation including CT scan of the head at the time of ED admission of each patients were assessed once they were diagnosed brain death. Logistic regression analysis was employed to determine the independent factor. p value of <0.05 was considered significant.

Results 28 brain dead patients records at the time of admission were analysed. On univariate analysis we found Glasgow Coma Scale (GCS) < 8, Blunt trauma chest (BTC),Skeletal injury, Intraventricular hemorrhage (IVH),Skull fracture,Subarachnoid hemorrhage (SAH),Midline shift (MLS),Mean blood pressure (MBP)< 60mmHg,Use of Ionotropes, Hemoglobin (Hb)< 8mg/dl,International normalization ratio(INR)> 1.5,Albumin< 3mg/dl,Sodium level (Na)> 150meq/dl,Urea > 50mg/dl significantly related to BD.On further multivariate analysis ,we found GCS< 8 (OR-4.2), BTC (OR-2.1), IVH (OR-3.4), MLS (OR-3.1), MBP <60 mmHg (OR-3.9), INR> 1.5(OR-3.1), albumin <3 mg/dl (OR-2.7) and Na level > 150meq/dl(OR-4.3) at the time of admission are strongly associated with BD.

Our study tried to find the predictors at the time of admission which may contribute to BD. Addressing them may prevent patient from becoming brain dead.

Biomedical technology in critical care is advancing at a rapid rate, offering the potential to substantially improve performance through improved efficiency and productivity. Recent evidence suggests that visual assessment of pupillary size and reactivity has limited interrater reliability and accuracy, hence, we examined the introduction and implementation of an automated pupillometer in an academic neurological ICU. We evaluated clinicians' perceptions about the added utility of the pupillometer to the standard visual pupillary exam.

-minute bedside education and demonstration of the pupillometer by a 'superuser', we conducted usability testing at the bedside. Participants completed the end-user testing methodology, where they completed specified tasks designed to test the pupillometer's features and later completed a questionnaire regarding their ease of use and interpretation of results, comfort and confidence using the pupillometer, and their behavioral intention to use the pupillometer if adopted into the clinical environment

To date, 46 participants have completed 50 questionnaires. Participants were allowed repeat enrollment in the study. The participant's professional designations include 41 registered nurses, 3 residents and 2 fellows and the majority have practised in the ICU for 1 to 5 years. Most of the participants are somewhat comfortable (24/46) performing the traditional visual pupillary exam and somewhat confident (29/46) with the results obtained from this exam. Twenty-one, out of 50 responses, were very comfortable in using the pupillometer, 22/50 were somewhat comfortable, and 7/50 were neutral. If this technology is introduced into ICU, the majority (31/46) will use this device to conduct pupillary exams, and 30/46 would consider changing management based on the pupillometer results.

This study outlines a strategy to evaluate usability and implementation of a newly adopted technology into the critical care environment. Improved implementation methods and evaluation of implementation processes are necessary for successful adoption of new technology in acute care settings.

Propofol infusion syndrome (PRIS) is a rare complication of propofol infusion. It is characterized by metabolic acidosis, rhabdomyolysis, acute renal failure, hyperlipidemia, and rapid cardiac failure. Risk factors for developing PRIS are: propofol infusion >48 hours, dosing >4mg/kg/hr, critical illness, malnutrition, and use of vasopressors. We present a case of PRIS that developed after propofol infusion was turned off.

A 43 year old woman with medically intractable epilepsy and developmental delay, presented with generalized tonic clonic status epilepticus. She was refractory to benzodiazepines, so she was intubated and started on a propofol infusion. At 50mcg/kg/min of propofol, she was still having generalized clonic tonic seizures. She was transferred to our neurological ICU for continuous EEG monitoring. Propofol infusion was increased to 180mcg/kg/min (10 mg/kg/hr) to control her seizures. She remained seizure free for 24 hours. Propofol was weaned over 12 hours because she became hypotensive and required norepinephrine. When the propofol was turned off, CPK was 800, lactate was 1.3, and creatinine was 0.4. She received propofol for 36 hours. Twelve hours after propofol was stopped, she developed a metabolic acidosis, lactate increased to 3.6, creatinine increased to 1.5, urine output decreased, and CPK increased to >20,000. She then developed bradycardia with wide complex QRS, which progressed to asystole. She could not be resuscitated and died.

Our patient developed PRIS after propofol infusion was off for 12 hours. She had many risk factors for developing PRIS, including high dose of propofol, critical illness, malnutrition, and use of vasopressors.

PRIS can occur after propofol infusion has been stopped, and should be monitored for after the infusion has been discontinued in patients that are at increased risk.

Subdural hemorrhage (SDH) is a common cause of morbidity. We sought to study the impact of antithrombotic drugs on nontraumatic SDH.

We retrospectively reviewed medical records of 1,211 patients admitted at Massachusetts General Hospital for SDH during 2017 to 2018 based on a research patient data registry. There were 78 patients without history of head trauma included in the analysis. Baseline demographic and clinical characteristic data were collected. The outcomes including GCS, modified Rankin Scale (mRS), SDH size, SDH expansion, surgical evacuation, mortality rates, length of stay (LOS), bleeding and thromboembolic complications were compared between two groups. Multivariate logistic regression was performed to analyze association between poor outcome (mRS 3-6) and all potential predictors (age, diabetes, conditional variable regression method was used because of relatively small sample size to avoid overfitting the model.

Among 78 patients included, 47 (60.3%) were on antithrombotic agents, either antiplatelets or anticoagulants, at presentation and 31 (39.7%) were not. Anticoagulant and antiplatelet agents constitute 25.6% and 19.2% of nontraumatic SDH, respectively. All antithrombotic agents were discontinued on admission. Nontraumatic SDH patients who were on antithrombotic agents had longer LOS (6.82 ± 2.94, p=0.023), higher rate of SDH expansion (OR 3.85; 95%CI 1.26-11.78; p=0.025), higher rate of disability at discharge (mRS 3-6) compared to no antithrombotic group (OR 12.38; 95%CI 2.5-61.05; p<0.001). On multivariate logistic regression analysis, antithombotic group had higher rate of poor outcome than no antithrombotic group (OR 12.38; 95%CI 2.5-61.05%; p=0.002).

Use of antithrombotic agents prior to admission in nontraumatic SDH patients correlates with longer LOS, higher SDH expansion and increased disability at discharge.

Maintaining goal sodium levels in the neurocritical care population can be challenging. Historically, at our institution, the supplementation of enteral sodium occurred by addition of table salt to tube feeding formulas by our dietary team. To make this therapy easier to standardize, monitor, and titrate, a new process was developed. Continuous 5% hypertonic sodium chloride solutions are now administered enterally via feeding tubes. This also allows for the charting of the medication and immediate dose titrations.

This pre-post analysis includes patients admitted six months prior to the implementation of the new enteral sodium process compared to patients admitted within one year after the new process change. Demographic variables, as well as the indication for sodium goals, initial sodium levels, sodium level for 96-hours post-addition of enteral sodium supplementations, concomitant use of intravenous hypertonic saline, and achievability of goal sodium levels were collected. Descriptive analytics were performed to compare groups.

A total of 92 patients were included in the analysis: 51 in the pre-implementation group and 41 in the post-implementation group. The most common indication for goal sodium levels in both groups was traumatic brain injury with head bleed; 43 patients (84%) in the pre-implementation group and 27 (66%) in the post-implementation group. Ability to maintain serum sodium concentrations (defined as the ability to maintain goal sodium without the need for intravenous hypertonic saline for >24h) within goal in the pre-implementation group was successful in 67% of patients (n=34) compared with 76% (n=31) in the post-implementation group.

The use of continuous enteral 5% hypertonic sodium chloride solutions to target and maintain goal sodium levels provided similar efficacy compared to the addition of table salt to tube feeding formulas and is safer and easier to monitor and titrate.

Coagulation factor Xa (recombinant), inactivated-Xa inhibitor associated life--factor prothrombin complex concentrate (PCC) was utilized off-

Retrospective, single center, cohort study including adult intracranial hemorrhage patients who received discharge between efficacy (defined by International Society on Thrombosis and Haemostasis criteria), thrombotic events, ICU and hospital length of stay, and mortality. 

Andexxa, coagulation factor Xa (recombinant), inactivated-zhzo is indicated for patients treated with rivaroxaban and apixaban, when reversal of anticoagulation is needed due to life-threatening or indication. There is no available literature supporting the use of this drug in acute neurosurgical emergencies. We present our experience of patients treated with Andexxa who required acute neurosurgical interventions as a life saving measure.

Patients were identified from May 2, 2018 to May 30, 2019 using an electronic database report identifying those who received Andexxa and subsequent chart review at a single center quaternary care academic medical facility.

factor Xa inhibitor and time of dosing. Patient 1 and 3 both had an external ventricular drain placed while in the emergency room. Patient 1 suffered from a cerebral hemorrhage with hydrocephalus while patient 3 was found to have a primary ventricular hemorrhage with hydrocephalus. Both were treated with four factor prothrombin complex concentrate (PCC) at an outside hospital. There were no bleeding complications during the procedures. Two patients had a craniotomy performed. Patient 2 was diagnosed with an acute subdural hemorrhage with worsening midline shift despite receiving PCC at the outside hospital. Patient four had an acute-chronic subdural hemorrhage with midline shift but did not receive PCC. In both craniotomy cases, there were no bleeding complications.

Andexxa was used in four patients taking apixaban or rivaroxaban undergoing lifesaving neurosurgical procedures despite no

The utilization of acute extracranial and intracranial stents for the treatment of cerebrovascular pathology is increasing. The optimal antiplatelet agent and dose in this population and the utility of platelet function testing is unclear.

All patients from January 2015 to April 2019 who were hospitalized and received ticagrelor to maintain intracranial or carotid stent patency in which platelet function testing (VerifyNow) was utilized to guide dosing were collected. Relevant demographic, clinical, platelet reactivity unit (PRU), and ticagrelor administration data was collected and qualitative assessment of PRU results was performed.

Data was collected on 104 patients and the maintenance doses utilized were 30, 45, 60 (most frequent) or 90 mg BID and loading doses of 120 mg or 180 mg. A total of 18 patients' doses were titrated in order to achieve the goal PRU range (50-150). Among patients given a dose of 120 mg 44% had a PRU in the optimal range (50-150) as compared to 11% among patients given a dose of 180 mg. Twice as many patients given a dose of 180 mg as compared to 120 mg (66% vs 31%) had a PRU between 10-50. Among the patients whose dose was titrated the average PRU prior to dose escalation was 154, the average PRU subsequent to dose escalation was 103, and the average PRU prior to dose decrease was 21 and the range in 55% of cases and was between 10-150 in 77% of cases.

The utilization of platelet function testing to guide dose titration of ticagrelor to a desired PRU range is feasible. A major limitation of this study is the lack of patient outcomes related to thrombosis or bleeding.

rivaroxaban. The efficacy and safety of andexanet alfa have been evaluated in the ANNEXA-4 study, which excluded patients receiving prothrombin complex concentrate (PCC) within the 7 days preceding enrollment. However, there have been limited reports of patients receiving both PCC and andexanet alfa for oral factor Xa inhibitor-associated major bleeding, without adverse effect. While thrombotic events were observed in 10% of ANNEXA-4 patients, potential for additive risk when combining andexanet alfa and PCC is undefined. We describe a patient who received PCC followed by andexanet alfa for an apixaban-associated intracerebral hemorrhage, who subsequently suffered devastating embolic strokes.

De-identified patient data were retrospectively collected from the electronic medical record

A 74-year-old male presented with acute left-sided hemiplegia caused by a large right-sided temporal lobe intracerebral hemorrhage. The patient had a history of atrial fibrillation, for which he was anticoagulated on apixaban. The patient initially received intravenous (IV) PCC 50 units/kg for prevention of hematoma expansion. The following day, minimally expanded hemorrhage was observed on repeat imaging concurrent with a measured apixaban level of 138 ng/mL (reference range 41-321 ng/mL). As a result, high dose andexanet alfa was administered as an 800 mg IV bolus, followed by an IV infusion of 8 mg/minute for 120 minutes. Over the next several days, the patient's neurologic exam supratentorial strokes, likely embolic in origin. Unfortunately, the patient did not survive hospitalization.

The combination of PCC and andexanet alfa may carry with it substantial thrombotic risk, and cannot be routinely recommended.

Targeted temperature management (TTM) is used for neurological protection in patients with neurological injury but shivering during TTM can reduce therapeutic effect by increasing oxygen consumption and metabolic rate. Cisatracurium used to prevent shivering has a shorter half-life than vecuronium and is not affected by liver and renal function. The objective of this study was to compare the efficacy and safety between two neuromuscular blockers in order to determine the benefit of cisatracurium.

We reviewed medical records of adult neurological intensive care unit (NCU) patients who received 1st, 2011 to May 31st, 2018. The efficacy between the two groups was confirmed by the presence of shivering and the recovery time of motor function. Safety was determined by the incidence of bradycardia and hypotension, the duration of antibiotic use and the mortality rate after discontinuation of the neuromuscular blocker in NCU. Recovery time of motor function was assessed using 'Motor Power' and 'Glasgow Coma Scale (GCS)'.

A total of 103 patients were included in the study: 53 patients in cisatracurium group and 50 patients in vecuronium group. The incidence of shivering was 18.0% and 18.9% (p = 0.910) in vecuronium and cisatracurium, respectively. The median recovery time of motor function was 7.0 [3.7-20.3], 1.7 [1.0-3.3] hours (p <0.001) based on the Motor Power score, 6.5 [3.7-23.5] hours and 1.7 [1.0-3.3] hours (p <0.001) based on the motor response score of GCS, respectively. The safety was not significantly different between the two groups.

Recovery time of motor function was significantly shorter in the cisatracurium group than in the vecuronium group and There was no significant difference in the others. this study identified the benefits of cisatracurium in NCU under TTM.

Amantadine and modafinil are neurostimulants that may improve or accelerate cognitive and functional recovery after a stroke. This systematic review describes amantadine and modafinil administration patterns post-stroke, evaluates their impact on cognitive and functional outcomes, and identifies the incidence of adverse drug effects.

An investigator-initiated MEDLINE search identified all full-text English-language publications describing the administration of amantadine or modafinil post-stroke from inception through October 5, 2018.

-stroke); Intervention (amantadine or modafinil treatment); Comparison (not required); Outcomes (cognitive or functional recovery). Amantadine and modafinil administration practices, cognitive and functional outcomes, and incidence of adverse drug effects were collected according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols (PRISMA-P) approach. Quantitative analysis was not performed due to heterogeneity in the measures of clinical effectiveness.

Initially, 4,109 publications were identified. Eight amantadine (73 patients) and 11 modafinil (114 patients) publications were included. Only 6 (8%) amantadine patients and 20 (18%) modafinil patients received treatment during an acute hospitalization. Time from stroke to amantadine initiation was 25 (11, 103.5) days and the initial dose was 150 (100-200) mg/day. Time from stroke to modafinil initiation was 170 (17, 496) days and the initial dose was 150 (100-375) mg/day. Under-responsiveness was the most common indication for neurostimulants (n=6/19 publications; 32%). Thirty-eight unique measures of clinical effectiveness were reported. A positive response in at least one measure of clinical effectiveness was reported in 75% and 82% of amantadine and modafinil publications, respectively. Visual hallucinations (amantadine) and excitability/agitation (modafinil) were the most common adverse effects.

Amantadine and modafinil may improve or accelerate cognitive and functional recovery post-stroke, but higher quality data are needed to confirm this conclusion, especially in the acute care setting.

Levetiracetam is an antiseizure medication that is used in neurocritical care (NCC) patients to prevent or treat seizures. Behavioral adverse events (ADE) are reported to occur in approximately 10% of patients taking levetiracetam; however, the incidence of these ADEs in NCC patients are unknown and may be exacerbated due to their unique CNS pathology. The purpose of this study is to identify the incidence of levetiracetam-associated behavioral (LAB) ADEs in NCC patients.

Adult NCC patients receiving levetiracetam, admitted between November 1, 2014 and October 31, 2018, and diagnosed with TBI, SAH or ICH, or cerebral infarction were included in this study. Criteria for determination of LAB ADEs included the following: 1) diagnosis codes for delirium, agitation, irritability, hostility, violent behavior, insomnia, anxiety, or depression during this hospital admit; 2) administration of an antipsychotic; 3) positive CAM-ICU; and/or 4) physical restraints. Day of LAB ADE onset was determined by the start date of the antipsychotic or a positive CAM-ICU.

There were 965 patients included in this study; 52% males, median admit GCS was 13. The most common neurological injuries were ICH (31%) and TBI (29%). LAB ADEs were identified in 456 (47%) patients. These were identified by diagnosis codes in 35% of patients, with delirium, depression, and agitation being most common; 20% received an antipsychotic, 12% had a positive CAM-ICU, 6% had restraints ordered, and 22% had more than one determining factor. lAB ADEs were reported a median of 2 (range 0 -29) days after levetiracetam initiation. Patients with TBI had the highest reported incidence of LAB ADEs (51%).

Almost half (47%) of NCC patients that received levetiracetam experienced a behavioral ADE, which was of levetiracetam use in NCC patients.

The recommend the use of 50 units/kg of four--PCC) or rting lower dosing strategies of aPCC. In 2014, a fixed, lowimplemented at our institution. The objective of this study was to evaluate the efficacy and safety of fixed, low-dose aPCC

This single-center, retrospective chart review included adult ICH patients who received aPCC for oral tcome was achievement of ICH hemostasis. Hemostasis was defined as no progression of hematoma on head CT within 48 hours post-aPCC. Safety outcomes included in-hospital mortality and incidence of thromboembolic event (VTE) within 30 days post-aPCC administration or up to the time of discharge, whichever came first.

-four patients receiving aPCC for reversal of factor Xa inhibitor associated ICH (26 traumatic and 28 spontaneous) were included for analysis. Median age was 77 years; 63% of patients had a past medical history of atrial fibrillation and 50% were anticoagulated with apixaban. Median aPCC dose was 1941 units (1808-2028 units), with a median weight-based dose of 24 units/kg (21 -29 units/kg). Hemostasis was achieved in 69% of all patients with ICH (73% in patients with traumatic ICH, and 64% of patients with spontaneous ICH). Mortality rate was 20% and VTE incidence was 7%.

of hemostasis in the majority of patients and a low incidence of VTE.

ally ill patients, yet the optimal monitoring method is unknown. The purpose of this study was to describe the correlation between aPTT and anti-Xa levels in patients receiving prophylactic SQ-

A retrospective chart review of patients admitted 18 years were included if they received SQ--Xa level drawn within 24 hours of each other. aPTT and anti-Xa levels were then compared to determine correlation and descriptive analyses were performed. Correlation was defined as normal aPTT levels (23.9-34.7 seconds) paired with undetectable anti-Xa levels (< 0.1 IU/mL), sub-therapeutic aPTT (34.8-75.9 seconds) with sub-therapeutic anti-Xa (0.1-0.29 seconds), therapeutic aPTT (76-112 seconds) with therapeutic anti-Xa (0.3-0.7 IU/mL), and supra-therapeutic aPTT (>112 seconds) with supra-therapeutic anti-Xa (>0.7 IU/mL) levels.

A total of 43 patients and 52 paired levels were analyzed. The median time between paired aPTT and anti-Xa levels drawn was 4.8 hours, and 40.4% (21/52) of levels were drawn within 1 hour of each other. Anti-Xa levels were drawn at a median of 6.2 hours after the SQpaired levels correlated, while 42.8% (9/21) of levels drawn within 1 hour of each other correlated. A Spearman's correlation coefficient of 0.32 (p=0.154) was found between aPTT and anti-Xa levels drawn within 1 hour of each other. A sub-therapeutic aPTT with undetectable anti-Xa was demonstrated in 42.9% of levels drawn within 1 hour of each other. The SQanti-Xa levels.

There was no significant correlation between aPTT and anti-Xa levels in patients who received SQ--SQH monitoring method in the neurocritically ill population.

The utilization of acute extracranial and intracranial stents for the treatment of cerebrovascular pathology is increasing. The optimal intravenous antiplatelet agent for short-term bridging of patients who are unable to tolerate or do not respond adequately to oral antiplatelet agents is unclear. Cangrelor offers potential advantages over glycoprotein IIb/IIIa inhibitors because response can be readily measured using platelet function testing (VerifyNow) and it has superior pharmacokinetics including a rapid on-set of effect and rapid clearance.

Patients with intracranial or carotid artery stents who were administered cangrelor for bridging purposes when oral antiplatelet agents were not feasible were assessed. Relevant demographic, clinical and procedural data as well as cangrelor dosing and platelet function testing data were collected.

patients had carotid artery stents. The indications for bridging were acute GI bleeding, inability to tolerate oral medications due to severe nausea/vomiting and two patients had an inadequate response to initial oral ticagrelor dosing based on platelet function testing. The dose of cangrelor utilized for all patients was 0.75 mcg/kg/min and all patients were on a cangrelor infusion for less than 48 hours. Platelet function testing (VerifyNow) was utilized to ensure adequate platelet inhibition and all patients demonstrated adequate inhibition on the prescribed dose. No stent thrombosis or bleeding was observed.

Cangrelor is a reasonable option when patients with intracranial or carotid stents necessitate an intravenous antiplatelet for bridging when oral antiplatelet medications are not feasible.

Current guidelines for the prevention and management of pain, agitation/sedation, delirium, immobility, and sleep disruption in adult intensive care unit (ICU) patients recommend a multimodal analgesia-first strategy to minimize opioid and sedative requirements and encourage early mobilization. The purpose of this study was to evaluate the success of a stepwise multidisciplinary implementation of an analgesiafirst sedation pathway followed by introduction of an early mobility protocol in a neuroscience ICU (NSICU).

We retrospectively evaluated mechanically ventilated adult NSICU patients admitted to a single-center academic medical center. Three-month time periods were evaluated at baseline (Phase I), after implementation of the sedation pathway (Phase II), and after implementation of the early mobility protocol (Phase III).

Total of 156 patients were evaluated: Phase I (n=67), Phase II (n=40), and Phase III (n=49). We observed a progressive decrease in propofol use during each phase (I, II and III) (median 2243.7 mg/day versus 2065.6 mg/day versus 1360.8 mg/day, respectively; p=0.04 between Phase I and III) and increased dexmedetomidine utilization (3% versus 10% versus 42.9% of patients, respectively; p<0.0001). Opioidanalgesia requirements during mechanical ventilation were similar between groups. We observed a quicker time from admission to PT evaluation between Phase II and Phase III (median [IQR] of 8 days [6] [7] [8] [9] [10] [11] versus 3 days [2-4], respectively; p<0.0001). Rehabilitation therapy was provided in 59.7%, 40%, and 81.6% of patients while admitted to the ICU in Phase I, II, and III, respectively (p=0.0003) and increased number of PT sessions provided per patient (median of 1 [0-2], 0 [0-3], and 8 [1-14] sessions/patient during each phase, respectively). No adverse events related to early mobility were observed.

Interdisciplinary coordination and communication is necessary for effective unit-based practice changes as education alone is insufficient. A multidisciplinary approach to goal-directed therapy targeting pain management and light sedation increased opportunity for early mobility.

The use of opioids in the neuroscience intensive care unit offset the balance of analgesia and reliability in performing neurological exam. In lieu of the current opioid crisis, we describe our center experience about the use of ketamine as an alternative medication with opioid sparing/lowering effect.

Retrospective chart review of patients admitted to NSICU with severe brain injury between November 2018 to April 2019 were performed. Patients were separated into two groups of twenty by randomization and matching, each receiving either ketamine or propofol infusion. Data collected includes age, gender, diagnosis, comorbidities, duration of ketamine, propofol and morphine equivalent (ME) opioid dose. Statistical descriptive analysis and independent samples t-test analytical analysis were performed to determine the difference of opioid use between two groups using SPSS software.

The range of ketamine used over the mean period of 3.15 (range 1-8) days was 2-15 mcg/kg/min, while that of propofol over the mean period of 3.25 (range 1-14) days was 10-75 mcg/kg/min. 12/20 (60%) and 11/20 (55%) patients in the ketamine and propofol group required opioids respectively. The cumulative and mean Morphine Equivalent (ME) dose for the ketamine group was 1023.8mg and 51.5mg respectively, while on propofol, it was 1527.5mg and 763.8mg. Results of independent t-test analysis showed a significant p-value of 0.033, indicating significant opioid dose reduction with ketamine.

It is essential to recognize the effectiveness of ketamine as an opioid sparing/lowering agent with potential analgesic-sedative medication without significant side effects.

Introduction different indications. However, serious complications such as I -current pulmonary embolism in patients with a contraindication to unknown. This information would be needed to determine if opportunities for improvement exists.

With approval from the local investigational review board (IRB), during the period of 2013-2018 were identified from the interventional radiology department. Only identified patient data was manually extracted via chart review to determine patient characteristics and

A total of 128 patients met inclusion criteria.61.7% were male. The most common neurocritical care diagnosis were intracranial hemorrhage(48%), ischemic stroke (16%), central nervous system (CNS) neoplasm (16%) and CNS trauma (9%). 47.7% of patients had at least 1 venous thromboembolism (VTE) was the most common indication (53%) followed by VTE with contraindication for AC (27%), Primary Adjunctive Treatment (13%) Adjunctive Prophylaxis (6%) and Secondary Adjunctive Treatment (1%).

In this single center study, to anticoagulation.

Andexanet alfa was approved in May 2018 for reversal of life-threatening hemorrhages for patients on anticoagulation with apixaban and rivaroxaban. Since its approval the reversal of direct oral anticoagulant (DOAC) associated intracranial hemorrhages (ICH) has been controversial. The objective of this study was to describe real world utilization of andexanet alfa at a large academic health system.

We retrospectively reviewed patients who received andexanet alfa for an ICH. Patients were included if they received andexanet alfa from its time of approval to formulary through April 29, 2019. Baseline demographics, anticoagulation and reversal information was collected. A neurointensivist reviewed all imaging. Intracerebral hematoma expansion was defined as > 35% increase in hematoma volume. Subdural (SD) and subarachnoid hemorrhage (SAH) expansion was defined as > 35% increase in maximal hematoma diameter.

Thirteen patients received andexanet alfa for ICH. Nine patients had an intracerebral hematoma, 1 patient had an isolated intraventricular hemorrhage, 2 patients had SD, and 1 patient had a SAH. The median age was 75 (IQR 71-83) and 46% of patients were male. Six patients were receiving a DOAC for stroke prevention, and a majority of patients (77%) were taking apixaban. The median Glasgow Coma Scale was 13 (IQR 13-15), and for patients with intracerebral hematomas the median ICH score was 2 (IQR 1 -3). There was follow-up imaging available for 11 patients, and 1 patient had hematoma expansion. One patient died and another had interval surgery prior to repeat imaging. No patients had in hospital thromboembolic events up to 30 days. Of the 13 patients, 46% of patients would have met exclusion criteria from the ANEXXA-4 trial.

In this small sample of patients who received andexanet alfa for ICH it appears hemostatic efficacy was achieved in a majority of patients with no thromboembolic events; however, larger trials are needed.

Lacosamide is a monotherapy or adjunctive therapy used for treatment of partial onset seizure that enhances slow inactivation of sodium channels. Uncommonly reported adverse effects include PR interval prolongation, bradycardia, atrioventricular block, and ventricular tachyarrhythmias.

An 82 year-old male with history of atrial fibrillation, hypertension and aortic valve replacement on warfarin presented with an acute subdural hematoma after feeling lightheaded and falling. The patient reported having multiple recent syncopal episodes. He received Prothrombin Complex Concentrate and Vitamin K for warfarin reversal with an initial INR of 4.0. He was started on levetiracetam and home medications of metoprolol and diltiazem were continued. The next evening, he had focal seizures, was given lorazepam and transferred back to the ICU. He received lacosamide 300 mg IV loading dose, and within 10 minutes had a 63 second episode of asystole. His blood pressure remained stable and he did not lose a pulse. He was given atropine x 3 doses with no response therefore transcutaneous pacing was initiated. Several minutes later, he became hypotensive and was started on isoproterenol and epinephrine infusions. EKG showed complete heart block. Cardiology was consulted and placed a transvenous pacer. Vasopressors were eventually weaned off however neuro exam remained poor. About a week later, family made the decision to transition to comfort measures and the patient passed away.

Lacosamide is an anticonvulsant primarily used for partial complex seizures. Only a few cases of third degree atrioventricular block have been reported in the literature. This case of extreme atrioventricular bock with a lacosamide loading dose is not common, but a drug-drug interaction with metoprolol and diltiazem was suspected. Prescribing lacosamide with beta-blockers or concomitant medications that prolong the PR interval should be done cautiously due to increased risk of atrioventricular block.

Tissue plasminogen activator (tPA) is currently the preferred agent for treatment of acute ischemic stroke. In about 7% of cases, patients will develop life threatening intracranial hemorrhage. Currently the AHA/ASA Guidelines and NCS Guidelines recommend reversal of intravenous tPa with cryoprecipitate and platelet infusion. Both society recommendations are based off low quality evidence and are given weak recommendations.Theoretically, the mechanism of action of tranexamic acid (TXA) makes it an appealing agent for reversal of tPA ; TXA competitively inhibits activation of plasmin countering the mechanism of action of tPA. The purpose of this case report is to report and support usage of TXA for reversal of thrombolysis with tPA.

This is a patient case report in which an extensive review of the patient chart was conducted to provide an accurate history of events. Extensive literature review was compiled to reflect current therapy guidelines and the off-label use of TXA for reversal of tPA.

year-old male presented to a tertiary care medical center with signs and symptoms of ischemic stroke symptomatic cerebellar hemorrhage. The delay in obtaining cryoprecipitate and platelet transfusion led the medical team to discuss alternative agents for the reversal of tPA. Reversal with TXA was discussed based on the medication's mechanism of action. TXA 10 mg/kg (1000 mg) was prepared at bedside and administered over 20 minutes. Repeat head CT showed no further progression of hemorrhage and there was an improvement in the patient's neurologic condition was noted

Hemorrhagic transformation following thrombolysis for ischemic stroke is a life threatening emergency. TXA is an appealing option for reversal of tPA as it directly counters the mechanism of tPA and can be easily and quickly accessed. This case reports further strengthens and supports its usage.

Drug level monitoring is essential to optimize valproic acid (VPA) efficacy and minimize toxicity. Total serum VPA levels of 50-100mcg/mL are recommended, though free drug is more precisely responsible for VPA's pharmacologic effect. The interpretation of total VPA levels is complicated by the drug's complex protein binding characteristics. The use of free serum VPA levels has garnered interest, though the therapeutic range is not well defined. Little is known about the relationship between free VPA levels and toxicity. We present a novel and unambiguous case of hepatotoxicity associated with elevated free VPA levels.

De-identified patient data were retrospectively collected from the electronic medical record

A 36-year-old male with a past medical history of refractory epilepsy was hospitalized for generalized tonic-clonic seizures. His prior home antiepileptic drugs (AEDs) included carbamazepine and the VPA precursor divalproex. The patient's total and free VPA levels upon admission were 48.6mcg/mL and 4.9mcg/mL (laboratory reference range 2normal. The patient's home divalproex ER dose was increased from 1000mg twice daily to VPA suspension 1250mg twice daily for his low total VPA level. On hospital day (HD) 3, the patient had a therapeutic total VPA level of 77.3mcg/mL, but an elevated free VPA level of 23.2mcg/mL in the setting concurrent with a free VPA level of 25.9mcg/mL. The patient's VPA was then transitioned to alternative AEDs due to hepatotoxicity concerns. The patient's clinical status later improved, and he was discharged Probability Scale implicated VPA as the probable cause of hepatotoxicity in this patient.

Measurement of free VPA levels helps guide dosing decisions and may reduce drug-related toxicity.

Limited case reports of osmotic demyelination syndrome (ODS) treated with intravenous immunoglobulin (IVIG) with or without plasma exchange (PE) are published, demonstrating variable neurologic recovery. The combination of IVIG and PE led to complete neurologic recovery of our ODS patient.

Electronic chart review to collect data for this case report.

21-year-old male presented with asymptomatic serum sodium of 96 mEq/L in the setting of intractable vomiting and decreased oral intake secondary to small bowel obstruction. His sodium was overcorrected by 21 mEq/L within first 24 hours. He subsequently developed altered mental status with lethargy and became unresponsive on day 7 with flaccid quadriparesis and minimal motor response to noxious stimuli. MRI of brain revealed osmotic demyelination of central pons and bilateral basal ganglia. IVIG was initiated on the day when ODS was confirmed on MRI. His serum sodium normalized. After 4day course of IVIG 2 g/kg, he could intermittently track with eyes but did not recover motor function. Plasma exchange was initiated 4 days after IVIG. After 2 sessions of PE, he started to move his right upper extremity antigravity and was attempting to verbalize. After 6 sessions of PE, he moved all 4 extremities antigravity, could talk although he had staccato speech and was able to ambulate with assistance. After 8 sessions of PE, he was ambulating independently; his motor strength was 4+/5 throughout. He was cognitively intact. At one month follow up in the clinic, he was neurologically completely intact, except for minimal upper extremity intention tremor.

IVIG with plasma exchange led to the remarkable neurologic recovery of a patient with ODS. A randomized control trial comparing IVIG monotherapy versus PE monotherapy versus the combination of IVIG and PE is warranted to better clarify the appropriate treatment protocol in ODS patients.

Digoxin is a commonly used drug in the treatment of heart failure patients but with no intrathecal indication. We describe a rare case of accidental intrathecal administration of digoxin during an elective caesarian section that lead to severe neurological deficits.

A 34-year-old Hispanic female underwent elective Caesarian section with 2 separate attempts at regional spinal anesthesia with bupivacaine due to failure of achieving adequate anesthesia with the first injection. Risk management discovered that patient had erroneously received Digoxin as the initial injection, confirmed by therapeutic serum levels of Digoxin. Two hours after delivering a healthy child, the patient's mental status deteriorated and she became unresponsive. She had three witnessed generalized tonic-clonic seizures and was emergently intubated for airway protection and received Keppra. Twenty-four hours in, patient remained comatose, continuous Electroencephalogram revealed no seizures, Magnetic resonance imaging (MRI) brain showed diffuse, patchy hyperintensities involving bilateral frontotemporal lobes and basal ganglia. MRI spine showed extensive cervical and thoracic cord edema. Cerebrospinal fluid analysis showed 476 White blood cells and protein count of 211. She received Solumedrol 1000 milligram intravenous for 5 doses followed by 5-day course of Intravenous Immunoglobulin (IVIG). Eleven days in, she was extubated. At discharge, she had intact Upper Extremity strength, intact speech, with no sensation or motor response below T10 level. MRI showed mild thoracic cord edema. At 90 day follow up, she had intact mental status and minimal improvement in motor strength and sensation below T10.

This is an extremely sad case of severe neurological deficits resulting from a grave medical error. There are only 2 previously reported cases of intrathecal administration of Digoxin in literature but the MRI findings, duration of symptoms and neurological deficits were far more severe in our patient. Neither cases reported use of high dose steroids or IVIG either.

Neurological complications following organ transplantation can be a result of a myriad of infectious, toxic-metabolic, vascular and iatrogenic causes. Given the wide range of possibilities, accurate diagnosis can be challenging. We present a case of acute hyperammonemia complicating renal transplantation.

A 56-year-old female with a remote left MCA stroke was evaluated for progressively worsening lethargy that started approximately a week after she had undergone deceased donor renal transplantation. Her immunosuppression comprised induction with alemtuzumab plus methylprednisolone with long-term mycophenolate mofetil plus tacrolimus, and antibiotic coverage included valganciclovir, trimethoprimsulfamethoxazole and fluconazole. Progressive deterioration in the level of consciousness progressing to coma with absent cough, gag reflex, sluggish pupils and no motor response resulted in the patient being intubated. Neurological examination did not reveal any focal deficits besides her pre-existing right hemiparesis.

Pertinent investigations included an MRI brain that showed no acute changes, EEG suggestive of triphasic waves and serial lumbar punctures showing elevated pressures in the 45-55 cm H2O range. level of 889 umol/L. In addition to appropriate pharmacotherapy and dietary protein restriction, the patient underwent continuous venoher mentation to baseline. Additional investigations done to determine the etiology of the hyperammonemia showed the patient to be infected with Ureaplasma urealyticum which was treated successfully with doxycycline and moxifloxacin.

To our knowledge, this is the first report of Ureaplasma urealyticum infection resulting in hyperammonemia fo management of hyperammonemia. In the absence of hepatic impairment, alternate etiologies of hyperammonemia should be sought. Acute hyperammonemia requires prompt evaluation and treatment to reduce the mortality and morbidity associated with it.

Prevalence, characteristics, and outcomes related to ventilator associated events (VAE) in neurocritically ill patients is unknown, and explored in this study.

A retrospective study was conducted to examine prevalence, factors, and outcomes of patients with VAE admitted to the neurocritical care service at Harborview Medical Center between January 1, 2014 and December 31, 2018. Chi-square test, analysis of variance was used to compare patients by VAE status.

Amongst 855 neurocritically ill patients, 132 VAEs occurred in 129 (15.1%) patients. Most common VAE was Ventilator Associated Condition, VAC, 73(55.3%), followed by Infection related VAC (IVAC), 32(24.2%), and Possible Ventilator Associated Pneumonia (PVAP), 27(20.5%). Most common trigger for VAE was an increase in positive end-expiratory pressure (PEEP). Age (median 56[IQR 31], male sex (61%), and BMI (26.7%) were comparable across groups with and without VAEs. Patients with VAE experienced higher intracranial pressures than those without VAE(33.5 mmHg vs. 23 mmHg, p < 0.0001). Compared to patients without any VAE, patients with any VAE spent longer time on mechanical ventilation (17.4 vs. 7.9 days, p < 0.001), and in the intensive care unit (16.5 vs.10 days, p < 0.001). Mortality (34.1% vs. 31.3%), median hospital length of stay (26.5 vs 21 days) and discharge to home (12.8% vs. 16.3%) were similar across both groups.

Ventilator associated events are prevalent amongst the neurocritically ill. They are commonly triggered by changes in PEEP, and are associated with intracranial hypertension, increase length of mechanical ventilation and intensive care unit stay but may not affect mor associated with VAE in subgroups of neurocritically ill patients and their impact on clinical outcomes warrants further examination.

Synthetic cannabinoids (SC) are a heterogeneous group of compounds initially developed to study the endogenous cannabinoid system. Most SC interact with CB1 and CB2 receptors with much higher affinity -tetrahydrocannabinol. The popularity of SC is increasing in adolescents and young adults because of the ability to produce a marijuana-like high without being detected on routine drug screens. We hereby present a case of SC related status epilepticus, hypoxic respiratory failure, severe acute kidney injury (AKI) and cerebral edema with fatal outcome.

19-year-old man with suspected SC adulteration of CBD oil presented with headache and status epilepticus. Labs showed leukocytosis, triple acidosis, and tetrahydrocannabinol in urine. CT head showed diffuse cerebral edema with sulcal subarachnoid hemorrhage. Intracranial Pressure was elevated to 25-30 mmHg. Hospital course was complicated by severe and refractory metabolic acidosis into hospitalization patient suffered cardiac arrest from pulseless ventricular tachycardia secondary to severe acidosis and metabolic derangements. After multiple attempts of resuscitation, care was withdrawn, and patient passed away.

In this case, severe refractory metabolic acidosis proved to be fatal. This case highlights the many challenges in managing a critically ill patient with cerebral edema and renal failure with medically refractory metabolic acidosis. SC are undetectable on routine drug screens and exposure is difficult to establish. SC can lead to multi-organ failure and death that may result from cardiovascular events, respiratory depression, pulmonary complications, and AKI. A high clinical suspicion is warranted in atrisk patients.

Exposure to SC may lead to cardiovascular, cerebral and renal complications that respond poorly to devise appropriate therapeutic strategies in managing such patients.

Benzodiazepines are the standard medication class for treating alcohol withdrawal symptoms (AWS). In acute brain injury benzodiazepines may worsen delirium and its central nervous system (CNS) depressant effects may decrease level of consciousness and make the neurological-exam unreliable. Barbiturates have similar actions to benzodiazepines on GABA receptors and cause less CNS depression. We present our center's experience with the use of phenobarbital in patients with AWS and acute brain injury.

Retrospective chart review of twenty patients admitted in neuroscience intensive Care unit(NSICU)with acute brain injury and AWS was done. Treatment protocol consisted of 10mg/kg Ideal Body weight(IBW) of phenobarbital loading dose divided into three intramuscular doses three hours apart, followed by a tapering daily oral maintenance dose for total of seven days. Alcohol withdrawal symptoms were assessed using the CIWA score for severity. Serum phenobarbital levels were drawn five hours after the third intramuscular dose. Liver function tests were performed before loading dose and daily for -times the upper limit of normal triggered protocol discontinuation.

None of the patients developed alcohol withdrawal seizures, one patient developed severe transaminitis. Loading doses of phenobarbital did not cause hypotension. Systemic toxicity was absent and phenobarbital serum levels drawn after the loading doses ranged between 6.0-26.7mcg/ml (normal range 10-30mcg/ml). Patients decreased their CIWA score after the loading doses of phenobarbital suggesting improvement of withdrawal symptoms and there was decreased use of adjunctive medications (benzodiazepines) for management of AWS. Nine patients required adjunctive benzodiazepines and received 4mg or less of lorazepam.

Phenobarbital for management of AWS was associated with minimal adverse effects and did not lead to systemic toxicity. Phenobarbital can be used in patients with acute brain injury without exacerbating delirium and can decrease the need for adjunctive benzodiazepines.

Aneurysmal subarachnoid hemorrhage (aSAH) has a case fatality rate of up to 70% in patient that rebleed. Cerebral arterial vasospasm (VSP) after aSAH is a leading reason for death and disability. Nicardipine is used to treat hypertension and angina, and has been investigate for a potential use in the treatment of VSP after aSAH.

Intraventricular nicardine was used for treatment of severe aSAH after traditional methods failed (ie. IR, hypervolemia, permissive hypertension and intravenous inotropes). 4mg of nicardipine was mixed with preservative free saline by pharmacy to total 8ml in volume. 8ml of cerebral spinal flu drawn from the patient external ventricular device (EVD). then the nicardipine solution was instilled and the EVD was clamped for 60 minutes. Patient had transcranial dopplers (TCDs) prior to injection and 2 hours after injection and reopening of the EVD.

Patient's vasospam temporized and neuro exam returned to pre spasm baseline.

Patient survived vasospam window and was transferred to long term care facility.

In neuroscience ICU (NSICU) maintaining balance between performing reliable neurological exam with adequate analgesia without causing significant sedation is challenging. Ketamine has significant neuroprotective and anti-seizure properties. In spite of these unique neuro-friendly pharmacological profile, it's role in NSICU unit is not well defined. We describe our experience about the use of ketamine in neuro-critical care unit.

Retrospective chart review of patients admitted to NSICU in whom ketamine was used as first line agent for sedation and analgesia in intubated patients with varied brain injury from January 1 to April 30 2019 was performed. Safety parameters collected includes blood pressure changes, intracranial pressure changes, heart rate, arrhythmias, excess secretions and apneic spells. Pco2 was monitored and hypercarbia was avoided. Effectiveness was measured by requirement of additional sedation-analgesic medications while receiving ketamine.

Twenty patients with varied brain injury who were on ketamine infusion as first line agent were selected. Mean age was 62.9 years (range 22-88 years) and 12 patients were male. Admitting diagnosis was hemorrhagic stroke (40%), ischemic stroke (40%), seizures (10%), carotid stenosis (5%) and tumor mass (5%). Mean duration of ketamine infusion was 3.15 days (range 1-8 days) and dose range was 2-15 mcg/kg/min. No ICP elevation was noted among the 5 patients where the ICP was monitored. None of the patients had uncontrollable elevated blood pressures nor major fluctuation in heart rate or respiratory rate requiring discontinuation of ketamine. 13 (65%) patients had increased secretions without respiratory compromise. Opioid use decreased significantly moreover additional sedation was not required while on ketamine infusion.

Ketamine is a safe and effective sedative-analgesic in neuro-critical care patients while at the same time allow for a reliable neurological examination to perform while on sedation. More research is warranted before it could be considered as the standard of care.

Oromandibular dystonia (OMD) is a movement disorder characterized by involuntary, sustained muscle contractions of varying severity resulting in sustained spasms of craniopharyngeal muscles affecting the jaws, tongue, face, and pharynx that can lead to abnormal jaw opening or closing or tongue protrusion.1 These disorders are often treated with botulinum for improvement of symptoms. There is minimal literature related to OMD treated for Botulinum in the neurocritically ill patient population.

We conducted a retrospective electronic medical record review from 2011-2018 of all brain-injured patients admitted to our neurocritical care unit who were diagnosed with OMD and received botulinum toxin injections. Etiology and location of brain injury along with clinical characteristics including resolution of symptoms were recorded.

Over a 7-year period, we injected 9 patients with Botulinum type A injection (100 mouse units or m.u.) into bilateral masseter muscles for severe OMD causing tongue biting/maceration and difficulty with oral care, and refractory to antispasmodics and muscle relaxant medications. Among the 9 patients, 6 patients were SAH, 1 patient with ICH/IVH, 1 patient with bilateral brain injury after post pituitary neurosurgical procedure and 1 patient with diffuse bilateral ischemic stroke related to sickle cell disease. All patients tolerated the procedure with no immediate complications. All patients had gradual improvement of OMD albeit variable and only 1 out of 10 patients required a 2nd treatment.

In this small series, injection of botulinum toxin for severe OMD from brain injury causing tongue injury appears to be safe, tolerable, and efficacious in reducing enteral antispasmodics/muscle relaxants. No short-term or long-term adverse effects were noted and it helped nursing with oral care over time.

Larger randomized controlled trials should be performed to evaluate the effectiveness and safety of treatment with botulinum in the critically ill neurologic population.

The Neurosurgical Intensive Care Unit (NSICU), a level 1 trauma center in San Antonio, cares for neuro critical patients. The use of central access catheters is essential for hypertonic fluid administration, vasoactive medications, and general critical care. In this unique population the risk of developing deep vein thrombosis (DVT) is higher compared to other patients due to reasons related to neurological injuries. The objective of this research was to determine the incidence and prevalence of DVT between the use of Peripherally-inserted central catheters (PICC) versus central venous catheters (CVC) in the NSICU.

We prospectively evaluated 250 consecutive patients with a CVC or PICC in the NSICU from 2017 to 2019. Data was collected, by a team of APPs on: surveillance vs non-surveillance ultrasounds, blood stream infections (CLABSI), indwelling time, complications, and ICU length of stay.

A total of 97 PICCs were placed for 786 catheter days, 9 patients were diagnosed with a DVT related to the catheter, rate of 1.1 per 100 catheter days. A total of 195 CVCs were placed for 924 catheter days, 3 patients were diagnosed with a DVT related to the CVC, rate of 0.3 per 100 catheter days. A total of 15 DVTs were diagnosed, one symptomatic patient and remaining 14 DVTs were identified during surveillance ultrasound. Two complications were encountered during insertion of a CVC and PICC which included development of hematoma on insertion of each catheter. The average length of stay for patients with a PICC line was 11.22 days. The average length of stay for patients with CVC was 16 days.

The NSICU surveillance ultrasounds identified more DVTs with the use of PICC lines versus CVC warranted if surveillance ultrasounds should be routinely performed for NSICU patients.

Mortality with Acute Respiratory Distress Syndrome (ARDS) is as high as 60% in patients with subarachnoid hemorrhage (SAH). Many of the therapeutic modalities of ARDS carry potential deleterious effects on ICP. We are presenting a challenging case of severe ARDS and SAH.

Single case report.

63-year-old male who developed a sudden severe headache. Emergent workup revealed a large cerebellar hemorrhage, SAH with IVH and hydrocephalus secondary to a ruptured arteriovenous malformation (AVM). Emergent suboccipital decompressive craniectomy followed by external ventricular drain (EVD) placement were performed and transferred to our facility for further aggressive care. Hospital course was complicated by severe pseudomonas pneumonia with progression to severe ventilation strategies, sedation, paralysis and inhaled nitric oxide (iNO) failed to correct hypoxia. On hospital day (HD) 12 he continued to show refractory hypoxia and was placed on Roto-Prone® bed. Continuous Intracranial Pressure (ICP) monitoring was utilized with EVD open at 5cmH2O. Prone positioning was attempted for 16 hours daily. Hypercarbia during prone positioning lead to elevated ICP patient showed improvement of hypoxia, with termination of prone positioning and subsequent weaning of paralytics and sedation. He started following commands and was discharged to a long term care facility after AVM embolization, placement of a tracheostomy, feeding tube and ventriculoperitoneal shunt.

Our patient made remarkable recovery from ARDS in the settings of obstructive hydrocephalus and SAH. Strict ICP monitoring, ongoing ventilator adjustment and careful utilization of kinetic maneuvers for ARDS, including prone positioning, contributed. Proning may be a consideration in patients with SAH, obstructive hydrocephalus and ARDS with ongoing ICP monitoring and ventilator adjustment, but larger scale studies are needed to explore its potential.

Paroxysmal sympathetic hyperactivity (PSH) has been associated with worse outcomes following traumatic brain injury, possibly representing both a marker of injury severity and a source of secondary injury. Prior studies suggest that PSH is under-recognized and its treatment often delayed. The identification of admission risk factors for PSH may facilitate earlier recognition, treatment, and targeted prevention.

Adults with severe TBI admitted to a neurotrauma ICU for at least 72 hours and hospitalized for at least 14 days between January 2016 and December 2018 were retrospectively identified. Consecutive PSH-TBI patients (n=80) were identified via review of medication administration records as having been treated with propranolol and/or bromocriptine for at least 72 hours. Control-TBI patients (n=160) were matched to the PSH-TBI cohort for age (36+/-14 years) and GCS (median 7 (3,8) ). Admission Head CTs were scored using Marshall and Rotterdam criteria. Independent-samples t-tests, chi-squared, and multivariate analyses of variance were performed.

Age-matched cohorts did not differ by sex, race, BMI, trauma type, trauma mechanism, ISS, or TRISS. ICU admission vital signs differed between groups with PSH-TBI demonstrating a higher HR (p=0.04) and a trend towards higher SBP (p=0.09), but no difference in core body temperature. Neuroradiographic features associated with PSH included significantly higher Rotterdam CT Score (p=0.01), presence of IVH/SAH (p=0.02), basal cistern compression (p=0.04), and trends toward higher Marshall CT score (p=0.07), presence of epidural hematoma (p=0.1), and CT DAI (p=0.08). A multivariate analysis adjusting for admission GCS and SBP identified Rotterdam score (p=0.003), presence of CT DAI (p=0.09), and ICU admission HR (p=0.03) as independent predictors of PSH.

Admission CT findings along with HR may help predict subsequent development of PSH requiring treatment. Early identification, treatment, and prevention of PSH may mitigate its negative impact on TBI outcomes.

Hyperchloremia in patients receiving chloride-containing solutions can contribute to metabolic acidosis and acute kidney injury (AKI), and has been associated with increased inpatient mortality, length of stay and AKI in patients with spontaneous intracranial hemorrhage. Whether hyperchloremia is a risk factor for mortality in patients with traumatic brain injury (TBI) is unknown. The purpose of this study is to determine if patients that develop moderate hyperchloremia while receiving continuous hypertonic saline (HTS) have a higher risk of inpatient mortality.

This was a retrospective chart review of patients admitted between January 2016 and September 2018. Included patients were over 16 years old, admitted to the trauma service with a diagnosis of TBI, and received continuous 3% HTS for at least 12 hours for the management of cerebral edema. Exclusion criteria were baseline end stage renal disease or hemodialysis, transition to comfort measures within 48 hours or inconsistent documentation. The primary objective was inpatient mortality. Secondary objectives were AKI, hospital and intensive care unit (ICU) length of stay. 

After TBI, mortality was higher in patients who experienced hyperchloremia, while AKI and length of stay were similar.

Although randomized controlled trials (RCTs) did not prove benefits of hypothermia for severe traumatic brain injury (TBI), brain CT images have not been evaluated in detail in these studies. We aimed to explore the prognostic value of brain CT findings in BHYPO Study.

BHYPO Study was a multicenter RCT to investigate the effect of therapeutic hypothermia in patients with severe TBI. The protocol included collection of brain CT data on admission and around day 7. Using the CT database, we evaluated following findings: presence of intracranial lesion (acute subdural hematoma: ASDH, acute epidural hematoma, cerebral contusion, subarachnoid hemorrhage: SAH, or intraventricular hemorrhage: IVH), basal cistern compression, lesion laterality, Marshall CT classification, and Rotterdam CT score. Hematoma thickness and midline shift were also measured. Unfavorable outcomes were defined GOS of 1 to 3 by Glasgow Outcome Scale (GOS) assessed at 3 months.

CT data were obtained from 125 patients on admission and 110 patients around day 7. There were no differences in CT findings between hypothermia group and fever control group. In the initial CT, univariate analysis showed that odds ratio (OR) and 95% confidence interval (CI) for unfavorable outcomes were: shift > hematoma thickness (6.0, 1.6-23.5: p=0.0053), SAH (3.7, 1.7-8.3, p=0.001), SAH or IVH (3.0, 1.2-7.6, p=0.0148), absent cistern (2.9, 1.3-6.6; p=0.0077), and midline shift > 5mm (2.3, 1.1-4.9, p=0.0219). Rotterdam score was significantly higher in patients with unfavorable outcome (3.7 vs. 4.5, p<0.001). Regarding the day 7 CT, bilateral lesion (8.5, 3.4-21.5, p<0.001) and SAH or IVH (3.3, 1.5-7.2, p=0.0025) were significant. No patients with absent cistern survived.

Patients were appropriately assigned in BHYPO Study in terms of CT findings. Shift > thickness, SAH, absent cistern, and Rotterdam score were powerful prognosticator in severe TBI patients undergoing targeted temperature management.

Cerebral edema (CE) following traumatic brain injury (TBI) causes secondary injury and increased mortality. Yet, conventional measurements of CE on head computed tomography (CT) inadequately accounts for CE. Serial volumetrics may facilitate estimation of total brain volume. The objective of this study was to measure the reliability of this technique and identify a threshold for brain volume (BV) change which could be indicative of CE.

A subset of patients (n = 18) with intracranial hemorrhage on admission CT were identified from a prospectively enrolled cohort of subjects with trauma sufficient to warrant ICU admission. Using Medical Image Processing, Analysis, and Visualization (MIPAV), two independent raters calculated BV on admission and follow-up head CT scans by measuring the volume of the intracranial vault and

The absolute difference (mL^3) and percent difference between the BV values of the two scans were calculated. Intraclass correlation (ICC) and Pearson's correlations were calculated, and significance set at 0.05.

The overall reliability of BV measurements between raters was excellent (initial scan ICC 0. 

Volumetric analysis to estimate BV appears to be a reliable technique across serial head CT scans. BV changes of more than 3.1% may represent a clinically significant threshold and should be further investigated.

Beneficial effects of Therapeutic hypothermia in adults with traumatic brain injuries are controversial. We wanted to study the effect of Therapeutic Hypothermia (TH) on outcomes after severe traumatic brain Injury (TBI) in real practice using the Nationwide Inpatient Sample in the United States.

The Nationwide Inpatient Sample was used to obtain data on all adults who had been discharged from 2005 to 2014 with a primary diagnosis of TBI who required mechanical ventilation, intracranial pressure monitoring, or craniotomy/craniectomy. The patients with TH were assigned to the TH group, and the rest were assigned to the control group. The primary outcome was in-hospital mortality, and the secondary outcomes included mean the length of stay, non-routine hospital discharge, mean hospital charges.

Only 722 patients (0.15%) out of a total of 488,231 underwent TH. TH group was younger (40.2 versus 54.6 years, p <.0001),had a lower proportion of females (17.8% versus 28.5%, p= 0.05) and a higher rate of in-hopsital complication of deep venous thrombosis (8.5% versus 2.0% p = 0.0098). When controlling for age, gender, comorbidities, in-hospital complications, hospital characteristics and disease severity, TH was associated with an increased rate of in-hospital mortality (odds ratio, 1.71;95% confidence interval, 1.08-2.70), longer mean length of stay (24.0 vs. 11.8 days; P<0.001), and greater mean total hospital cost ($371,940vs. $134,823; P<0.001). There was no difference between the two groups in terms of non-routine discharge (odds ratio, 1.61; 95% confidence interval, 0.65-3.97),

Therapeutic hypothermia was associated with poorer outcomes in patients with severe TBI. Our findings disfavor therapeutic hypothermia in severe TBI in routine clinical practice. It warrants further investigation in a prospective, randomized study.

A rising incidence of subdural hematomas (SDH) has been attributed in part to increased use of anticoagulants and antiplatelets. Anticoagulants also worsen the severity and prognosis of SDHs, but the impact of antiplatelets on prognosis is unclear. We hypothesized that antiplatelets would not affect SDH severity or outcome, while anticoagulants would be associated with more severe features and a worse functional outcome.

We systematically identified and collected data on patients presenting with a new diagnosis of SDH in 2018 at a Level I trauma center. We examined common markers of SDH severity in three cohorts of patients: those not on any antithrombotics, those on antiplatelets alone, and those on anticoagulants.

Categorical data was compared with Chi-Squared tests, and continuous data was compared with Mann-Whitney U tests. Multivariable logistic regression was used to assess the impact of antiplatelet use on functional outcome at discharge, with a poor functional outcome defined as a score of 3-6 on the Modified Rankin Scale.

We identified 441 patients with a new SDH during 2018: 224 (50.8%) did not take antithrombotics, 141 (32%) took antiplatelets, and 76 (17.2%) took anticoagulants. Antiplatelets were not associated with increased SDH volume, thickness, or midline shift; anticoagulants were associated with increased volume (P<0.001), thickness (P<0.001), and a trend towards increased midline shift (P=0.08). Antiplatelets were associated with a better admission score on the Glasgow Coma Scale (P<0.001). When adjusted for age and gender, antiplatelets did not affect functional outcome (OR 0.84, P 0.53, 95% CI 0.48-1.44), while anticoagulants were associated with poorer functional outcome (OR 2.83, P 0.008, 95% CI 1.35-6.4).

Despite its known association with overall SDH incidence, premorbid antiplatelet use was not associated with SDH severity or a worse functional outcome at a level 1 trauma center.

The Common Data Elements Therapeutic Intensity Level (CDE-TIL) score, quantifies the intensity of nursing and medical care aimed at preventing intracranial hypertension for patients with severe traumatic brain injury. We validated the CDE TIL in our neurotrauma intensive care unit (NTICU) and found the CDE-TIL to be highly reflective of perceived and measured therapeutic burden but noted that the scale had a ceiling effect. Specifically when ICP was 20-25 mmHg and higher, the CDE-TIL did not capture the escalating burden. In an attempt to eliminate that ceiling effect and to incorporate current h TIL (P-TIL).

Under a quality assurance approved protocol, retrospective chart review was performed on 22 adult patients with severe TBI. The TIL score was derived using both the CDE-TIL and the P-TIL for each 12 hour nursing shift for the first 5 full days of admission. The relationship between the CDE-TIL and P-TIL and the ICP were investigated. Reliability testing of the P-TIL, including interrater reliability, and validation of the P-TIL are ongoing.

The P-TIL and the CDE-TIL are highly correlated (r=0.71) and the relationship between the scores and the maximum ICP are similar at ICP less than 20mmHg. At higher ICPs however, the slope of P-TIL increases to 0.18 compared to the CDE-TIL slope of 0.07 and illustrates a 2.5 times stronger correlation between the intensity of care level as measured by P-TIL and ICP.

The P-TIL has greater sensitivity for quantifying the intensity of therapy aimed at controlling ICPs, most significantly for patients with the highest ICPs, ICPs 20-25mmHg and above, making it an ideal scoring system for communicating current nursing and medical needs of individual TBI patients as well as potentially predicting post-intensive care or post-discharge needs.

Patients are frequently brought into neurologic intensive care units in cervical spine immobilization after sustaining ground level falls or after being "found down." Currently there is no consensus regarding cervical spine clearance in these patients as they are unable to participate in neurologic examination. After normal CT scans, MRI scans are frequently employed to evaluate for ligamentous injury and radiographic signs of cervical instability.

We conducted a retrospective chart review of patients who were admitted to the Neurologic Intensive Care Unit between 2017 and 2018 in cervical collars after ground level falls or after being found down (presumed ground level falls). Patients were included in the study if they were obtunded on admission (GCS<15) with neurologic exams consistent with their cranial pathology. All patients underwent a high definition CT Cervical spine or CTA of the Neck and were cleared if there was no radiographic evidence of fracture or instability.

Between 2017-2018, eight patients were admitted to the Neurologic Intensive Care Unit that met inclusion criteria. Average age at presentation was 75.75 years. Cranial pathology on presentation included intraparenchymal hemorrhage, ischemic stroke, and subdural hemorrhage. All patients underwent a high definition CT cervical spine or CTA neck which showed degenerative changes without fractures, subluxations or other evidence of instability such as increased atlantodental interval, or prevertebral soft tissue swelling. Average follow up was 129.75 days range (1-406). There were no cases of cleared patients that suffered secondary neurologic injury or symptoms of cervical instability during the follow up period.

Our study illustrates that obtunded patients after ground level falls can safely be cleared of cervical spine precautions after a high definition CT Cervical spine fails to demonstrate fractures, subluxations, or other evidence of cervical instability. This protocol limits the costs associated with MRI scans and the risks associated with cervical immobilization.

The elderly comprise the highest incidence of traumatic brain injury (TBI) hospitalizations and death, yet most TBI studies neglect the geriatric population. Previous studies suggest women have better outcomes after TBI but are inconclusive. We examined differences in outcomes between sexes after TBI in the geriatric population.

This is an observational study of patients 60 and older admitted with TBI to a level 1 trauma center. Clinical variables including medical history, severity of injury (GCS>12, GCS9-12, and GCS<9), mechanism of injury, and CT findings were collected. Good clinical outcomes were defined as a GOSE >4 and measured at discharge and 6 months. The Chianalysis were used where appropriate.

152 subjects were included in the analysis. 66(43%) women and 86(57%) men. Average age was 72.5(SD 8.9) with no significant differences between sexes. 117(77%) were mild, 15(10%) moderate, and 20(13%) severe. The most common etiologies were mechanical fall 64(42%), motor vehicle accident 22(15%), and syncopal fall 21(13.8%). No differences in severity of injury or mechanism of injury were found. On admission CT, men had more contusions (40%v22%;p=0.04) and skull fractures(21%v45%;p,0.01) compared to women. Older age, and history of atrial fibrillation or congestive heart failure were associated with increased incidence of death. Men were more likely to have in-hospital mortality (13%v3%; p<0.01). In multivariable logistic regression analysis controlling for other factors associated with mortality, men were significantly more likely to have in-hospital death (OR-14;p=0.01). At 6 months, men were still found to have higher mortality (OR-12.4;p<0.01). However, there were no significant differences in good outcomes between sexes at discharge (39%v43%; p=0.65) or 6 months (50%v51%;p=1).

Men have significantly higher mortality rates compared to women in the geriatric TBI population.

differences are needed.

Partial brain tissue oxygen tension (PbtO2) can be regulated by the fraction of inspired oxygen and the level of oxygen carrying capacity. We performed a systematic review of the literature using PbtO2directed treatment with red blood cell transfusion (RBCT) to analyze clinical and physiological outcomes as well as adverse events following RBCT.

We performed a systematic review following the PRISMA guidelines and pre-registered with the PROSPERO database. The following terms were used: [(brain tissue oxygen OR brain tissue hypoxia OR pbO2) AND treatment] OR [(brain tissue oxygen) OR red blood cell transfusion) OR pbO2) OR traumatic brain injury) AND red blood cell transfusion]. Inclusion criteria were studies in which PbtO2 was measured before and after RBCT. The tool used for qualitative scoring was the GRADE score. Risk of bias was assessed via RTI and ROBINS-I.

A total of 2390 articles were screened of which four articles were included in the final analysis. The intervention performed was to administer 1 to 2 units of RBC depending on the hemoglobin level and the threshold set in each study. The clinical outcome was not described in any of the studies. There was an increase in PbtO2 in all the studies, but it was primarily significant when pretransfusion PbtO2 was less than 15 mmHg. The GRADE certainty rating for the included articles was low to moderate.

Our review shows that a significant increase in PbtO2 is primarily seen when pre-transfusion PbtO2 is less than 15 mmHg. Clinical outcome and adverse events were not described in any of the included studies. In view of the known adverse effects of RBCT in critically ill patients and the limited available literature we found, transfusion should only be reserved as a later tier measure for PbtO2 correction, and possibly only when PbtO2 is less than 15 mmHg.

Withdrawal of life-sustaining therapy (WLST) is associated with 70% of deaths after severe traumatic brain injury (TBI). WLST frequently occurs within the first 3 days of hospitalization, when prognosis is most uncertain. While patient factors play a role in the decision, institutional practice patterns and physician perception of prognosis also contribute, as demonstrated in Canadian studies. We hypothesized that the rate and timing of WLST among patients with severe TBI vary across the United States.

We conducted a retrospective cohort study of patients with severe TBI admitted in 2016 to 463 US trauma centers included in the Trauma Quality Improvement Program. Severe, isolated TBI was defined by diagnosis code and Glasgow Coma Scale (GCS) score <9. Patients under 18, with severe non-head injuries, or with advanced directives were excluded. Centers were grouped by US census region (Northeast, Midwest, West, South). Multiple logistic regression for WLST was performed with region, patient demographics, GCS motor score, pupillary reactivity, and midline shift as covariates. Regression -hospital mortality. 

Variability may reflect inconsistent institutional practice patterns, regional cultural differences, and the difficulty of prognostication. More reliable and standardized prognostic assessments are needed in this population.

Introduction: Pre-injury use of antiplatelet agents may increase hemorrhage size and hematoma expansion after traumatic brain injury (TBI). However, empiric platelet transfusions may result in significant morbidity and unnecessary expense and may not be justified. We sought to determine whether a thromboelastography (TEG) platelet-mapping (PM) algorithm could safely reduce platelet transfusion without clinically relevant hematoma expansion.

Methods: A prospective standardized TEG PM-based treatment algorithm was instituted to guide reversal of antiplatelet medications in TBI patients. The algorithm established reversal thresholds for arachadonic acid inhibition (AA-inhibition >50%) and adenosine diphosphate inhibition (ADP-inhibition >60%). Consecutive TBI patients were enrolled and compared to a historical cohort. Hematoma volume was calculated by ITK-SNAP. 

Conclusions: A TEG-guided antiplatelet reversal algorithm may significantly reduce platelet transfusions without clinically significant hemorrhage expansion.

Increasing partial oxygen arterial tension is one method to increase the partial brain tissue oxygen (PbtO2). However the effects of hyperoxia on clinical outcomes and adverse effects remain elusive.

To investigate the effects of normobaric and hyperbaric hyperoxia on PbtO2 in patients with TBI, we performed a literature review following the PRISMA guidelines and pre-registered with the PROSPERO database. The following search terms were applied: [(brain tissue oxygen OR brain tissue hypoxia OR pbO2) AND treatment] OR [(brain tissue oxygen) OR brain tissue hypoxia) OR pbO2) OR traumatic brain injury) AND hyperoxia]. Prospective trials and observational cohort studies were included in this review. Two reviewers assessed the risk of bias of each study using the RTI item bank.

A total of 2390 articles were screened, of which 16 articles were included. Only one study investigated the effects of combined hyperbaric/normobaric hyperoxia and another used hyperbaric as a separate intervention; the majority of studies were of normobaric hyperoxia. Overall, an increase in pbtO2 was observed with both normobaric and hyperbaric. Clinical outcome was mostly missing; one study showed an absolute reduction in mortality and improvement in favorable outcome using Glasgow outcome score at 6 months. Adverse events were also only scarcely reported; 3 studies showed that hyperoxia did not induce cerebral toxicity by using markers of oxidative stress, and one study showed no evidence of pulmonary oxygen toxicity in either the hyperbaric or normobaric hyperoxia groups.

Normobaric and hyperbaric hyperoxia consistently induced an increase in pbtO2. Improvement in clinical outcome was reported in some studies but did not reach statistical significance except in one. Adverse events were not adequately investigated. Larger prospective studies are required to investigate the clinical outcome effects of hyperoxia, its adverse consequences, and its role in the tiered approach towards brain tissue dysoxia.

Early prognostication, either from clinical and/or radiological information, is an important aspect in the settings of neurocritical care with limited resources. We sought to determine the values of two radiological scoring systems in predicting the outcome of traumatic brain injury (TBI) patients, which are Marshall and Rotterdam CT scores in Indonesia. Therefore, a physician can make a better priority to provide high-yield care to all TBI patients.

A retrospective cohort was conducted in a national referral hospital from July to December 2017. All TBI patients admitted to the emergency department (ED) and had an initial CT scan were included in this study. Their classification of TBI and initial CT scan were reviewed and all patients were followed to see whether the patient died or alive until discharge from the hospital (in-hospital mortality). Statistical analyses were conducted to find the predictive values (sensitivity, specificity, cut-off point, relative risk) of both scoring systems.

Of 141 TBI patients admitted to ED, there were 136 patients had an initial CT scan. Most of them were categorized as mild TBI (80.9%), then moderate (14.0%) and severe TBI (5.1%). In-hospital mortality was 9.6%. With cut-off point 1 in Marshall and Rotterdam CT scores, their sensitivity (92.3% vs. 92.3%, respectively) and specificity (73.1% vs. 74.8%, respectively) were similar. Same things also found in their relative risks, which are 24.267 (95% CI 3.256 -180.835) and 25.953 (95% CI 3. 485 -193.258 ).

Both Marshall and Rotterdam CT scores have significant values in predicting the outcome of TBI patients, thus it should be implemented in daily emergency practice to assist a physician in making further clinical decisions.

Midline shift (MLS) in brain is a critical condition. If not diagnosed timely, it could lead to a devastating outcome. Computed tomography (CT) scan is the gold standard technique to diagnose MLS in neurosurgical patients. The aim of our study was to find out association between transcranial sonography [TCS] and CT scan in assessing midline shift in patients with TBI.

In this prospective ongoing study, adult patients ages 18 -65 years, of either gender, with TBI were included. Demographic details were noted. All patients underwent CT scan, followed by TCS. MLS on TCS was determined using standard technique. We noted the MLS on CT scan and time window between CT scan and TCS was also measured. Consciousness was assessed using Glasgow Coma Scale (GCS) and GCS -Pupil [GCS-P] scales. Descriptive data are given as Mean (SD) or Number. Spearman's correlation test was used to detect relationship between GCS and MLS assessed by CT scan and TCS, and also GCS-P. The value of p<0.05 was considered significant.

A total of 16 neurosurgical patients were studied. Male to female ratio was 15:1. The age was 44 [17.74] years with weight of 64.5 [10.17] kg. Ten patients had GCS<8. The mean value of MLS measured by TCS -0.1, p = 0.69). The correlation between TCS and CT scan with GCS was in significan respectively. However the value of GCS-

In patients with TBI, MLS can be successfully assessed using bedside, non-invasive and non-radioactive monitor TCS when compared to a CT scan. There is a good correlation between GCS and GCS -P.

Early post-TBI seizures are reported to occur within 24 hours and between days 1-7 following TBI in 8.9% and 1.9% patients, respectively. Early seizure prophylaxis with phenytoin in severe TBI patients is drugs with better safety profile have emerged as potential alternatives. The objective was to describe seizure prophylaxis practices in critically ill TBI patients.

We conducted a retrospective observational study of 9 adult trauma ICUs. We included consecutive adult ICU patients with moderate and severe TBI admitted between Jan 2015 and Dec 2016. Data were collected using standardized forms. Our primary outcome was the incidence of seizure prophylaxis use.

We included 379 patients with a moderate (43%) or severe (56%) TBI. The majority were men (70.5%) with mean age of 54.9 (SD 21.6) an (51%) and MVA (30%). A total of 30% required invasive ICP monitoring. A total of 259 patients (68%) received early seizure prophylaxis, 67% for moderate and 69% for severe TBI. Phenytoin, levetiracetam or their combination were used in 228 (88%), 9 (3%) and 10 (4%) of cases, respectively. Twelve patients (5%) were previously treated for pre-existing epilepsy. A total of 60 (16%) patients experienced a seizure (29 at the trauma scene, 15 in ER, 15 in ICU and 1 on the ward). Among the 118 severe TBI patients in ICU for 7 days or more, anticonvulsants were continued for the recommended 7 days in 57% of cases.

Early seizure prophylaxis is inconsistently used in severe TBI patients in Canada. Phenytoin still remains the agent most used. Despite the current recommendations, 31% with severe TBI did not receive prophylaxis and 43% for a shorter period than 7 days.

Raised ICP persistently in severe TBI patients may be detrimental. However, chest physical therapy (CPT) is equally necessary for preventing secondary factors influencing the risk in these patients. This study was intended to observe the impact of short-term rise in ICP with manual CPT in severe TBI patients on outcome along with hemodynamics.

This was a prospective, observational trial on 92 adult patients, of either sex, aged 18-60 years, with severe TBI, on mechanical ventilatory support with continuous ICP monitoring, and receiving CPT on regular basis, included in this study. The CPT was applied for 10 minutes' duration and repeated after an interval of 6 hours in between for a total 2 sessions in a day. The measurement measured intracranial pressure, cerebral perfusion pressure, heart rate, mean arterial pressure (from start of the intervention until 10 min after the intervention at 1 min interval each), and GCS after each session of CPT along with final outcome/GOS at the time of discharge and 3 months.

The rise in median intracranial pressure of 3.0 (-8.0, 8.5) and median cerebral perfusion pressure of 5.25 (-37, 29 .5) was significantly higher during intervention and after intervention phase. In contrast, a median heart rate rise of 7.5 (2.0, 28.5) and mean arterial pressure rise of 7.0(2.0, 29.0) were comparable. However, in patients with high baseline ICP (>20 mmHg), poor outcome was noted in terms of low GOSE 1(1, 3), and higher mortality (40.9%) at hospital discharge or 3 months after injury.

Significant increase in ICP in severe TBI patients post CPT for 10 minutes at a time (total 20 minutes each day) was not tolerable in this cohort. Moreover, we observed significantly low GOSE in patients with sustained intracranial hypertension. The effect of manual technique of CPT on final (long-term) neurological outcomes remain inconclusive but with favorable respiratory outcome.

Survivors of moderate and severe traumatic brain injury (msTBI) require substantial care, much of which is provided by friends and family. We sought to describe the experience and unmet needs of survivors and their informal caregivers follow msTBI, particularly related to care transitions.

This study was conducted in two intensive care units (ICUs) at a level 1 trauma center. We conducted qualitative, semi-structured interviews with both patients and informal caregivers of msTBI survivors at 72 hours, one month, three months, and six months post injury. Informal caregivers were defined as friends or family who planned to provide care for the patient. Patients were 18 years or older with an msTBI, and not expected to imminently die of their injuries.

Eighteen patient-caregiver dyads were enrolled. One patient died within 72 hours. At 72 hours, 17 caregivers were interviewed; at one-month 13 caregivers were interviewed; at three months 11 caregivers and one survivor were interviewed; and, at six months 11 caregivers and seven survivors were interviewed. Three themes were identified in the qualitative analysis of caregiver interviews: caregiver burden, caregiver health related quality of life, and caregiver need for information and support. Experiences varied depending on time since injury, discharge disposition, functional neurologic outcome, caregiver access to resources, and likely multiple other additional factors. Interviews with survivors were not insightful secondary to post-traumatic amnesia.

This study provides new information about the experience of informal caregivers during the six months after their friend or family member survived an msTBI. Caregivers reported that needs evolved over time. At three to months, few moderate to severe TBI patients were well enough to be interviewed, and information obtained by survivors was not insightful. Interventions to promote caregiving may be a substantial opportunity to improve patient and caregiver-centered outcomes following TBI.

Vasospasm following traumatic brain injury (TBI) has a high incidence and a detrimental effect on the neurological prognosis. Yet, it remains a neglected, poorly understood phenomenon and there are no guidelines for its management. Herein we present a case of severe vasospasm following TBI that caused secondary delayed cerebral ischemia (DCI). We further appraised the current literature aiming at identifying predictors of vasospasm in TBI.

A 31 y/o white woman presented to the Hospital after a mechanical fall resulting in mild TBI with associated subarachnoid hemorrhage (SAH). Glasgow Coma Scale (GCS) at presentation was 15, with no neurological deficits. A non-contrast CT head revealed diffuse bilateral fronto, parietal and temporal SAH without evidence of aneurysm or vascular malformations on CT angiogram (CTA). Toxicology screens were negative. At 72 hours from TBI patient developed acute severe headache. A repeated CTA showed right Internal Carotid Artery (ICA) and Middle Cerebral Artery (MCA) vasospasm with no ischemia identified on MRI brain. Patient was started on Nimodipine. On day-10 patient developed acute left side hemiparesis and neglect with neuroimaging evidence of a complete right MCA infarct. Hemodynamic augmentation therapy was initiated with partial improvement of deficits. Patient subsequently developed hemorrhagic conversion of the right MCA infarct. On day-25 neuroimaging revealed resolution of vasospasm. Patient had residual left side neglect and anosognosia.

In line with prior literature our patient developed vasospasm in the large intracranial vessels, at 72 hours from the TBI and earlier than in aneurysmal SAH. However, differently from previous reports, GCS at presentation was > 9, age was > 30 and despite vasospasm developing later than 24 hours it was not associated with good outcome.

eded to identify accurate predictors of vasospasm following TBI with secondary DCI that could improve detection and management of this detrimental phenomenon.

Therapeutic hypothermia and/or cooling therapy has been hypothesized to have benefits in patients with traumatic brain injury (TBI). Several systematic reviews (SR) are being performed to address this question, but their results are inconsistent. The objective of this study was to assess the methodological quality of SR that included randomized clinical trials (RCTs) that assessed the effects of therapeutic hypothermia and/or cooling therapy in patients with TBI.

A critical appraisal study was performed in order to assess any SR that fulfilled the inclusion criteria. An unrestricted search of the literature was carried out in March 2019 at four major electronic databases (MEDLINE, EMBASE, LILACS and Cochrane Library). Two independent reviewers selected the studies, extracted the data and appraised the methodological quality of the included SR using the AMSTAR-2 (A MeaSurement Tool to Assess systematic Reviews) tool. An overall assessment of the confidence in the results was performed using the checklist available in AMSTAR-2 website (https://amstar.ca/Amstar_Checklist.php). The confidence of the results may be graded as high, moderate, low or critically low. This grading is based on the adequacy of the SR to the 16 domains of the AMSTAR-2.

The search strategy retrieved 149 references. After the selection process, 16 SR were included. The SR were published between 2003-2018 and included 0 to 37 RCTs. The overall confidence in the results from included SR was graded as critically low in 53.3%, low in 13.3%, moderate in 20%, high in 13.3%.

A high number of SR addressing similar clinical questions were published in a short period of time. The methodological quality was adequate in only few SR. Clinical practice guidelines should considered this result when choosing the evidence synthesis to recommend for practice.

Neurogenic pulmonary edema (NPE) is a clinical syndrome characterized by acute onset after central nervous system injury. The aim of this study was to investigate the clinical features of NPE in patients with subarachnoid hemorrhage (SAH).

The authors retrospectively analyzed a total of 425 patients with SAH who were treated at our hospital from April 2014 to September 2017. Of these 425 patients, 350 were included in this study after the application of predefined exclusion criteria. Patient demographics, aneurysm size and location, clinical characteristics, and patient outcomes were reviewed and compared between an NPE and a non-NPE group.

Sixteen patients (4.6%) presented with NPE at admission. Among them, 10 patients (62.5%) recovered from NPE immediately, and ventilatory support was withdrawn within 2 days from onset. A univariate analysis showed that patients with NPE were of younger age (p=0.04), had a higher rate of vertebral (p=0.01), and lower systolic blood pressure on admission (p=0. revealed significant differences in the frequency of vertebral artery dissection (odds ratio (OR) 4.83, 95% CI 1.50--13.66, p=0.04) between the groups with and without NPE. No significant group differences were found in other factors, including heart rate, neurologic outcomes at discharge.

Vertebral art factors for NPE. However, neurologic outcomes at discharge did not differ between groups, suggesting that poor outcome due to NPE could be reduced by appropriate diagnosis and treatment.

Antibiotic-impregnated catheters (AIC) are recommended for the prevention of ventriculostomy-related infections (VRI). Other antibiotic prophylaxis strategies following external ventricular drain (EVD) placement vary widely by institution. The role of systemic antibiotics for this indication remains controversial.

We retrospectively reviewed the charts of all patients having an EVD placed between January 1, 2016 and December 31, 2018. After excluding patients who died or were discharged within 48 hours of EVD placement or had an EVD placed due to suspected meningitis, patients were categorized into the periprocedural (P) or no periprocedural (NP) antibiotics group. Patients were determined to have a VRI if catheter and up to 7 days after catheter removal. Mann-Whitney U test was used to analyze descriptive data and baseline demographics. Chi-squared models were used to analyze the incidence of infection.

included in the no periprocedural antibiotics group (age 60 [52-69] years; 50% male) and 120 were included in the periprocedural antibiotics group (age 51 [39-62] years; 56% male). The most frequent indications for EVD were subarachnoid hemorrhage (SAH) [NP: n=51 (54%), P: n=30 (25), p<0.001], intracranial hemorrhage (ICH) [NP: n=20 (21%), P: n=23 (19%), p=0.731), and other, which included colloid cysts and tumors [NP: n=10 (10.5%), P: n=37 (31%), p<0.001]. There were 4 infections in the no periprocedural antibiotics group compared to 8 in the periprocedural antibiotics group (p=0.436). The most common pathogen was coagulase-negative staphylococci (n=4, 33%).

The use of periprocedural systemic antibiotic prophylaxis did not significantly reduce the incidence of VRI. Periprocedural systemic antibiotics may not be necessary in the setting of antibiotic impregnated catheters to reduce the incidence of infection.

Cerebral artery vasospasm is a rare complication of craniopharyngioma resection but can have life altering consequences including delayed cerebral ischemia if not quickly recognized and managed appropriately. We present a case of craniopharyngioma resection in a 16 year old male complicated by refractory vasospasm and its management with intraventricular nicardipine.

Data regarding the operative management, time course, vasospasm and management was accessed retrospectively after patient discharge.

A 16 year old male with recurrence of a craniopharyngioma presented with left eye vision loss and was admitted to the neurosciences intensive care unit after transsphenoidal resection. Intraoperatively, the tumor was noted to be adhered to the posterior communicating artery and the left anterior cerebral artery. Dense invasion into the hypothalamus was noted. This portion was carefully resected to avoid progressive lethargy. Computed tomography angiography revealed new mild narrowing of the left anterior and middle cerebral arteries and bilateral posterior cerebral arteries consistent with vasospasm. The patient was treated with a vasospasm bundle including nimodipine, euvolemia, and blood pressure augmentation. Over the next twenty days, the patient continued to have a variable amount of vasospasm despite aggressive medical and intra-arterial management. On post-operative day 27. Nicardipine was then infused into the EVD once a day for 3 days, resulting in rapid and sustained improvement in vasospasm.

The mechanism of vasospasm following skull base tumor resection is unknown. Presence of blood in the operative bed, direct surgical injury to the blood vessels, hypothalamic dysfunction and the release of inflammatory chemicals have all been proposed. Treatment remains similar to treatment used in SAH, utilizing nimodipine, euvolemia, blood pressure augmentation and intra-arterial verapamil. This case demonstrates the effectiveness of intraventricular infusion of nicardipine on refractory vasospasm.

To present a rare case of bilateral Internal Carotid Artery (ICA) aneurysms presenting as trigeminal neuralgia (TN), with good outcome post surgical treatment.

A 70-year-old woman presented with disabling TN for 1 year, exclusively affecting the right maxillary and mandibular divisions. Symptoms did not abate with trial of adequate doses of Gabapentin, Duloxetine, Oxcarbazepine and Indomethacin. Thin-cut magnetic resonance imaging (MRI) brain with and without contrast showed rare contact with wide-necked aneurysms of bilateral petrous-cavernous ICAs producing prominent mass effect on bilateral adjacent trigeminal nerves. Carotid arteriogram redemonstrated ICA aneurysms with left measuring 13.1 mm x 7.70 mm and right measuring 10.5 mm x hours post procedure, TN had completely resolved. Patient was started on Aspirin 325 mg and Clopidogrel 75 mg daily and is being tentatively planned for intervention on left aneurysm. On her 3month follow-up appointment with neurology, she reports no recurrence of TN.

In cases of aneurysmal causes of TN, presence of bilateral aneurysms causing mass effect on the trigeminal nerve at its root is a rare occurrence and needs high clinical suspicion. Due to the high risk of rupture associated with giant and symptomatic aneurysms, treatment should be expedited and aggressive in order to not only address symptomatic TN but also to avoid the risk of aneurysm rupture in the future. Surgical clipping and endovascular coiling with or without stenting has demonstrated remarkable symptom relief in reviewed literature for other types of intracranial aneurysm.

Moyamoya disease is a chronic cerebrovascular disease characterized by spontaneous and progressive stenosis or occlusion of the internal carotid artery and its branches. Revascularization procedures have been shown to improve cerebral hemodynamics and decrease the risk of strokes, but several postoperative complications are known to occur.

We present a case with a fairly rare complication with characteristic radiological findings after surgery. A 14-year-old girl with moyamoya disease underwent left superficial temporal artery (STA)-to-middle cerebral artery (MCA) anastomosis with encephalo-duro-myo-synangiosis (EDMS), and did right STA-MCA anastomosis and EDMS one year after the initial surgery.

The procedures were uneventful and the occlusion time was 38 minutes. She recovered from the anesthesia without neurological deficit, and MRI on postoperative day (POD) 1 demonstrated no ischemic lesions and patent bypass, although swelling of the temporal muscle attached to the brain surface was noted. On postoperative day 4, she experienced a transient neurological event (left hemiparesis). Magnetic resonance imaging revealed large cortical and subcortical hyperintense lesions in the middle cerebral artery territory on diffusion-weighted imaging and apparent diffusion coefficient imaging. Subsequently, the radiographic findings improved within several days with resolution of the symptoms.

Revascularization surgery for improving a patient's hemodynamics can prevent the development of strokes, but is known to be associated with perioperative cerebral infarction and cerebral hyperperfusion causing transient neurological deterioration, delayed intracerebral hemorrhage, and vasogenic edema.This case is a reminder that hemodynamic complications can develop subacutely in patients who have undergone successful revascularization for moyamoya disease. The radiological features and mechanisms of this rare condition associated with revascularization surgery for moyamoya disease are discussed.

Vasospasm with delayed cerebral ischemia is a rare but known complication of endoscopic transsphenoidal resection of pituitary adenoma. This complication has rarely been reported in cases of -arterial treatment have been favorable in some cases.

Electronic medical record review.

The patient is a 42 year old male who underwent subtotal resection of pituitary adenoma via an open right fronto-temporal approach. Eight days post-resection he developed progressive headache and leftsided weakness which acutely worsened the following day. His NIHSS on presentation was 12, consistent with right MCA syndrome. CT Brain showed mass effect in the right frontal lobe with 3.5 mm midline shift. CTA showed sluggish flow through right M1 branch suggestive of vasospasm. He was taken to cerebral angiogram post-op day 11 and received right ICA intra-arterial verapamil and right ICA and MCA angioplasty. He was started on nimodipine following the procedure. His exam improved significantly over the course of 1-2 days. He was discharged home on verapamil 80mg q8 hours. At three month follow-up his NIHSS was 0 and his modified Rankin Scale was 1.

In the case we present, the patient received intra-arterial treatment with verapamil and angioplasty 2-3 days after onset of symptoms. Despite delayed presentation the patient ultimately achieved a favorable functional status. Vasospasm and stroke post-pituitary tumor resection are complications of which patients should be adequately informed, especially when considering the possibility of good functional outcome with intraof this potentially debilitating and life-threatening complication and attention should be paid to utilizing techniques for early detection of vasospasm.

Neuromonitoring is an essential part of the management of neurocritical patients. Many ICUs in developing countries manage their patients without monitoring ICP. Intensivists play a vital role in clinical judgments to manage their patients. Raised ICP are handled either by medical management or surgical procedures like decompressive craniotomy. The study aimed to see the outcome of patients with raised ICP and compare medical vs surgical management in these patients without monitoring ICP.

A retrospective observational study was conducted among patients admitted from January to December 2018 in the ICU of Dhaka medical college hospital, Bangladesh. Patients who had etiologies of brain code, clinical presentations and or radiological findings consistent with raised ICP were included. Patents were grouped into neurosurgical and medical management groups. Length of ICU stays and mortality were observed. Student's t-test and Chi-Square tests were used to see the statistical significance.

Total of 187 patients was selected. Mean age was 40.51±20.5 years, and 75.4% were male. Traumatic brain injury was the most common cause of raised ICP (41.2%) among selected patients. 52.9% of patients were managed medically, and neurosurgical procedures managed 47.1% of patients. Length of ICU stay was higher in neurosurgical patients compared to medical management group (13.22±9.91 vs 10.58±8.84; p=0.56, non-significant). Mortality was higher in neurosurgical patients compared to medical management group (55.7% vs 48.5%; p=0.326, non-significant). Mortality was also higher in traumatic brain injury patients who underwent neurosurgery compared to medical management (58.2% vs 50%; p=0.51, non-significant).

Neurosurgical management didn't show a better outcome in patients with raised ICP when monitoring was unavailable in a resource-limited ICU.

Chronic kidney disease (CKD) independently increases the risk of stroke and burden of ischemic small vessel disease (SVD). Effects of CKD on intracranial hemodynamics remain poorly defined. This study compared SVD and a transcranial Doppler (TCD)-based marker of intracranial vascular resistance (pulsatility index, PI) in post-stroke patients with and without CKD. 

within three months of a stroke. Anterior and posterior circulation PI (ACA, MCA, and PCA) significantly correlated with MRI lesion volume in all patients. CKD strongly correlated with higher distal resistance (median CKD ACA PI 1. 

In patients with recent stroke, MRI SVD volume is significantly associated with anterior and posterior circulation PI. Significantly higher SVD lesion burdens and anterior circulation PIs were observed in patients with CKD. CKD is an independent determinant of increased intracranial vascular resistance in both anterior and posterior cerebral circulations.

Atrial fibrillation is associated with an increased risk of stroke and systemic embolism. We investigated the prevalence of coexisting subdiaphragmatic visceral infarction (SDVI) in patients with acute ischemic stroke due to atrial fibrillation and also evaluated independent factors of acute SDVI.

We enrolled a consecutive series of 100 acute ischemic stroke subjects with atrial fibrillation between MRA or CTA were excluded. All subjects were prospectively examined using abdominal MR imaging at 1.5T and transthoracic echocardiography (TTE) within 7 days of onset. A multivariable logistic regression analysis with predefined variable (age and sex) and the potential confounders that were associated with SDVI i

The mean age was 72.9 ± 9.3 years (47% males). Onset-to-abdominal image time was 3.8 ± 2.1 days. Among 100 patients, 23 acute coexisting SDVI (20 renal and 2 splenic infarctions and 1 superior mesenteric artery occlusion) were found in 20 patients with acute ischemic stroke and atrial fibrillation. Twelve patients had a chronic SDVI; 10 renal and 2 splenic infarctions. No hepatic and bladder infarction was shown. Severe significantly associated with the coexistence of acute SDVI and acute ischemic stroke attributed to atrial fibrillation in the logistic regression model. (adjusted OR, 11.0; 95% CI, 1.2-98.0; p = 0.03).

There was a significant relationship between the presence of acute SDVI and severe left atrial remodeling in acute ischemic stroke patients attributed to atrial fibrillation. Based on these results, we suggest that abdominal MR imaging for evaluating coexisting acute SDVI should be considered in patients with acute ischemic stroke due to atrial fibrillation, especially with left atrial enlargement on TTE.

Patients with large hemispheric infarction are likely to accumulate chloride due to commonly used hypertonic saline for lowering elevated intracranial pressure. However, the effect of chloride burden on clinical outcomes in these patients is not well studied. This study aims to investigate the impact of maximum serum chloride concentration during admission on in-hospital mortality in critically ill patients with large hemispheric infarction.

We conducted a retrospective observational study of patients with large hemispheric infarction who were admitted to the neurocritical care unit, between March 2013 and June 2018. Patients were excluded if they had baseline creatinine clearance less than 15 mL/min, required neurocritical care for less than 72 hours. Multivariable logistic regression models were used to evaluate the association of maximum serum chloride concentration during admission with in-hospital mortality.

Of 106 eligible patients, 23 (21.7%) were died in hospital. Compared to patients who survive to hospital discharge, those who died in hospital had higher maximum serum chloride level during admission (137.0 ± 10.8 vs 117.8 ± 10.1, p<0.001). Each 5 mmol/L increase in maximum serum chloride concentration was associated with increased risk of in-hospital mortality with an odds ratio of 2.324 (95% CI, 1.622-3.330, p<0.001). After adjusting for confounders including Acute Physiology, Age, Chronic Health Evaluation II (APACHE II) score, baseline serum glucose, base deficit, use of mannitol, hypertonic saline, therapeutic hypothermia, and incidence of acute kidney injury, maximum serum chloride level remained an independent risk factor associated with in-hospital mortality (adjusted odds ratio for every 5 mmol/L increment, 1.966; 95% CI, 1.222-3.163, p=0.005).

Higher maximum serum chloride concentration was associated with higher in-hospital mortality in critically ill patients with large hemispheric infarction. These results suggest serum chloride level should be monitored as high chloride burden may cause poor outcomes on those populations.

Patients with acute ischemic stroke caused by large vessel occlusion may receive both CT-angiogram (CTA) and digital subtraction angiogram in the process of evaluation and management of restoring perfusion. Neither AHA/ASA stroke/imaging guidelines address indications for Transcranial Doppler (TCD) and/or Carotid Duplex Ultrasonography (CUS) in early stroke evaluation and most patients do not receive additional cerebrovascular imaging after reperfusion. We investigated the clinical utility of performing TCD/CUS after reperfusion in guiding post-acute care stroke management.

We reviewed 200 inpatient ischemic strokes admitted to a comprehensive stroke center in 2017. 41 of these had TCD/CUS done and 25 had CTA done prior to TCD. 24 of these underwent either tissue plasminogen activator or thrombectomy for reperfusion. These cases were reviewed by two experts (KH, QV), who were blinded to each other, to determine if TCD/CUS provided any added value after CTA affecting patient management. A nominal group process was performed, using a third blinded expert (AS) in case of disagreements to reach consensus. The reviewers reported 10 cases where TCD/CUS provided incremental value for management. Value added by TCD/CUS, as noted by experts, included detection of residual/recurrent mobile thrombus requiring anticoagulation, confirmation of reperfusion in a symptomatic patient, distinguishing between carotid stenosis and occlusion by showing string sign on carotid ultrasound, confirming hemodynamic significance of angiographic stenosis helping triage the need for stenting/endarterectomy, and new information on chronicity of carotid stenosis based on collateral flow patterns hence deferring further intervention.

Our experience shows a significant added value of performing TCD/CUS in more than 40% of stroke cases in our review. The incremental information provided by ultrasound-guided further evaluation and management decisions in most of these patients.

Axons of the Wallerian degeneration slow (Wlds) mutant mice survive weeks after traumatic and ischemic nerve injuries. Prior characterization of the mutant Wlds protein showed that it is a fusion gene product between the non-functional, truncated N70 amino acids of UBE4B and full functional sequence of nuclear NMNAT1, a rate-limiting enzyme in NAD+ synthesis. However, the molecular mechanisms by which the mutant Wlds protein protects axons from stroke injuries remain unclear. We sought to understand how Wlds is able to robustly protect axons from ischemic injuries, and in doing so possibly identify novel therapeutic targets to attenuate axonal loss in stroke.

We first sought to understand the temporal and spatial requirements of Wlds activity in protecting axons from ischemic injuries. To achieve this, we developed a novel tool to conditionally regulate the expression of Wlds protein by modulating its post-translational protein stability. Using this powerful technique, we asked how conditionally "turning on" or "turning off" Wlds activity affects axonal survival following ischemic insults. Moreover, as the only known function of Wlds is in catalyzing NAD+ synthesis, we designed a high-throughput pharmacological screen for NAD+ analogs to evaluate whether the NAD+ synthetic pathway mediates Wlds axon protection.

We found that conditional expression of Wlds protein within 4-6hrs after stroke injuries was necessary and sufficient to confer axonal survival, whereas turning off Wlds activity post-injury abolished axon protection. This indicates that Wlds activity is a local event in the axon, and exerts axonal protection within a critical time window even after the injury has occured. We further observed that exogenous addition of NAD+, but not its precursors or immediate metabolites, was sufficient to confer axonal protection, while attenuating NAD+ levels abolished WldS axon protection. This suggests that NAD+ is a molecular mediator of Wlds axon protection in stroke.

We showed that Wlds activity is a local axonal event, and uncovered a critical window of 4-6hrs poststroke injury in which the course of axon degeneration can be halted or even reversed in mammalian neurons. Moreover, we showed that this process is mediated by rising NAD+ levels in axonal compartments through a novel NAD+ dependent cell signaling cascade. These findings provide powerful insight into the molecular bases of Wlds activity, and uncover new therapeutic targets to delay and potentially even reverse axon degeneration in stroke.

Unruptured intracranial aneurysm (UIA) are incidentally found on the computed tomography (CT) or magnetic resonance angiography in about 2% of patients. Because of the risk of intracranial hemorrhage (ICH), the presence of UIA is contraindication to intravenous thrombolysis for acute stroke. As noncontrast CT (NCCT) is mostly used for thrombolytic therapy and UIA is difficult to diagnose using a NCCT, UIA may be found after thrombolysis.

Among the patients with acute ischemic stroke treated with intravenous thrombolysis for 5 consecutive years in one stroke center, patients diagnosed with UIA by CT angiography immediately after thrombolysis, were enrolled. Characteristics of UIA and clinical outcomes such as ICH and modified Rankin Scale (mRS) score at discharge were analyzed.

Among 172 patients treated with intravenous thrombolysis, 8 (2.3%) patients were diagnosed with UIA. ally relevant artery and 7 patients an aneurysm less than 10mm in diameter. The median value of the initial National Institutes of Health Stroke Scale score was 12 (range 5-19). The median mRS score at discharge was 3 (range 2-5). There was no patient who had ICH or aneurysm rupture during admission.

Intravenous thrombolysis could be safe and necessary to the patients with hyperacute ischemic stroke and incidental UIA.

Recent studies suggest that variations in the constitution of the gut microbiome contribute to atherosclerotic burden and cardiovascular disease. While many gastrointestinal (GI) diseases are known to cause disruption of the normal gut microbiome in humans, the clinical impact of GI diseases on subsequent vascular disease remains unknown. We conducted an exploratory analysis evaluating the relationship between GI disease and ischemic stroke or acute myocardial infarction (MI).

We performed a retrospective cohort study using claims between 2008-2015 from a nationally composite of ischemic stroke or acute MI. Stroke and MI were assessed separately as secondary outcomes. In an exploratory manner, we evaluated the association of each GI disorder in the ICD-9-CM classification with our outcomes. We then categorized individual GI disorders by anatomic location, disease chronicity, and disease mechanism. We used Cox proportional hazards models to examine associations with adjustment for demographics and established vascular risk factors. Since this was an exploratory, hypothesis-generating study, we report only notable positive associations.

Among approximately 1,730,000 beneficiaries, the following GI disorders were associated with an increased risk of subsequent ischemic stroke: gastric ulcer (HR, 1.16, 95% CI, 1.06-1.26), duodenal ulcer (1.13, 1.01-1.23), gastritis and duodenitis (1.11; 1.05-1.17), disorders of function of stomach (1.16, 1.07-1.27), other disorders of stomach and duodenum (1.14; 1.03-1.27), gastrointestinal mucositis (6.76; 2.18-20.97), unspecified noninfectious gastroenteritis and colitis (1.14; 1.07-1.22) and gastrointestinal hemorrhage (1.06; 1.01-1.13). The following categories of GI disorders were associated with an increased risk of ischemic stroke: stomach disorders (1.14; 1.07-1.22), stomach and small intestine disorders (1.05; 1.00-1.09), ulcerative disorders (1.07; 1.01-1.13) and chronic GI disorders (1.08; 1.01-1.15). GI disorders were not associated with an increased risk of MI, and some demonstrated a reduced risk.

Several GI disorders were associated with an increased risk of ischemic stroke, but none were associated with an increased risk of MI

To evaluate the relationship between serum neutrophil-to-lymphocyte ratio (NLR) levels and early neurological deterioration (END) in ischemic stroke patients with large-artery atherosclerosis (LAA).

We evaluated consecutive ischemic stroke patients due to LAA between January 2010 and December within the first 72 hours of admission. The NLR was calculated by dividing the absolute neutrophil counts by the absolute lymphocyte counts.

Among the included patients (n = 349; male, 59.9%; mean age, 69 years), 18.1% (n = 63) had END events. In multivariate analysis, serum NLR level was independently associated with END (adjusted odds ratio, 1.09; 95% confidence interval [1.02 to 1.17], P = 0.016). Visit time from symptoms onset, and insitu thrombosis and artery-to-artery embolization mechanisms were also found to be significant factors for END events. In the analyses regarding the relationship between serum NLR values and burden of vascular lesions, NLR levels were positively correlated with both the degree of stenotic lesions (P for trend = 0.006) and numbers of vessel stenosis (P for trend = 0.002) in a dose-response manner. We also compared the difference of serum NLR levels according to the stroke mechanisms from underlying vascular lesions. Then, hypoperfusion and in-situ thrombosis mechanisms showed higher levels of NLR. However, only in-situ thrombosis mechanism had higher NLR values among the END groups compared to non-END groups (P = 0.005).

Serum NLR levels were associated with END events in ischemic stroke patients with LAA mechanism. Since NLR was also closely correlated with the relevant vascular lesions, our results indicated clues for underlying mechanisms of END events.

Transcranial Doppler (TCD) can detect emboli in numerous cerebrovascular settings. Although previous studies have suggested that microembolic signals (MES) may predict recurrent stroke, the practical significance of such findings remains unclear. This uncertainty has deterred the widespread use of embolic monitoring among clinicians.

In a retrospective fashion, we investigated the real-world applicability of TCD by examining whether the presence of MES portends worsened clinical outcomes. We reviewed the charts of all ischemic stroke patients (n = 136) who underwent MES monitoring from January 2017 to December 2018.

Of the stroke subtypes reviewed, 35% were atheroembolic, 29% were cardioembolic, 4% were lacunar, 9% were dissection, 10% were hypercoagulable, 10% were cryptogenic, and 8% were due to other causes. 6 +/-9 MES were detected in 18% of patients. MES were detected at an average of 14.99 +/-4.72 dB (with a detection threshold > 9.00 dB). Recurrent stroke was seen in 10% of patients (monitored over 7.2 +/-6.6 days). Patients with MES were more likely to have recurrent stroke (46% vs. 2%, p < 0.001), undergo a revascularization procedure (25% vs. 6%, p = 0.005), have a longer length of stay (10 vs. 4 days, p = 0.001), and have a discharge mRS 3-6 (83% vs. 30%, p < 0.001) compared to those without MES. Multivariable logistic regression analysis showed that MES was an independent predictor of recurrent stroke (OR 42.1, 95% CI 6.7-265.6) and of poor discharge mRS 3-6 (OR 15.7, 95% CI 4.0-61.7) despite controlling for antithrombotic treatments and stroke subtypes.

In the largest series of patients who underwent embolic monitoring with TCD, MES predicted ischemic stroke recurrence leading to worsened disability and prolonged hospital stays. Given that MES can provide important prognostic information, TCD with embolic monitoring may be clinically useful in the workup of ischemic stroke.

Expanded patient eligibility for mechanical thrombectomy (MER) of acute ischemic stroke (AIS) has resulted in a proportional increase of patients who require emergency angioplasty and/or stenting (EAS) to achieve recanalization. Post-stenting antiplatelet medication management continues to remain a challenge due to lack of immediate effect and rapid reversibility ideal for patients at high risk of stent thrombosis and hemorrhagic complications, especially after intravenous alteplase (TPA). Cangrelor is an immediate-acting intravenous P2Y12 receptor inhibitor with rapid clearance and restoration of normal platelet within one hour of infusion termination. We describe our preliminary experience with administration of Cangrelor in AIS patients undergoing MER and requiring EAS as rescue therapy.

Ten patients with AIS who received Cangrelor after MER were identified. Median admission National TPA prior to MER. Cangrelor drip was started immediately prior to EAS. Median duration of Cangrelor drip was 22 hours. Dual antiplatelet was given a median time of 5 hours before discontinuation of Cangrelor. Seven patients had repeat imaging at 3 months confirming durable vessel patency and no restenosis. None of the patients experienced clinical deterioration, symptomatic intracranial hemorrhage, or recurrent strokes during the hospital stay. One patient underwent surgical decompression but did not develop any hemorrhagic complications. Median mRS at discharge was 3, and median NIHSS at discharge was 8.

In our case series, Cangrelor was observed to be a safe alternative to oral antiplatelet drugs in the immediate perioperative period among AIS patients who underwent MER and required EAS, , including patients who received TPA and at high risk for malignant cerebral edema or hemorrhagic transformation who may require emergency surgical decompression.

The response of the neonatal brain to hypoxic ischemic injury (HI) is developmentally specific therefore therapies for brain HI cannot be standardized across the ages. While arginases (ARG; isoforms ARG-1/ARG-2) are enzymes actively studied for their neuroprotective/neuroregenerative effects in various neurological conditions, in neonatal HI the ARG effect remains unknown.

To test the hypothesis that ARG changes with neurodevelopment and after HI we exposed mice C57BL/6 (wild-type) to hypoxia-ischemia on postnatal day 9, as follows: permanent coagulation of left common carotid artery to induce ischemia, a 1h recovery period and exposure to 10 % oxygen/balance nitrogen at 37°C for 50 min to induce hypoxia. Animals were perfused at 1h,4h,12h, 24h and day 5 with 4 % paraformaldehyde, brains were post-fixed, sectioned on a cryostat (12 um) and examined histologically with cresyl violet stain to assess the degree of damage and ARG spatiotemporal localization via immunohistochemistry. ARG expression was measured by Western Blot and ARG activity spectrophotometrically.

ARG expression and activity increase during development, however this increase is suppressed by HI. ARG-1 expression increases on day 5 after HI which corresponds to our findings of ARG-1 accumulation at the penumbra site. Cortical ARG activity remains suppressed after HI, compared to that in the hippocampus, where it increases. Spatiotemporally, ARG-1 localizes into myeloid cells in CNS. ARG-1expression increases in microglia as early as 4h after injury and remains elevated for a prolonged time. ARG-2 is localized in pyramidal neurons of the indusium griseum, fasciola cinerea, neocortex and hippocampus (CA2, CA3). ARG-2-expressing cells are damaged by HI, however they do not undergo spatial changes.

Microglial ARG-1 strongly responds to HI and may play role in neuroinflammation and neuroprotection, while ARGand therapeutic potential of the ARG-pathway in neonatal HI.

SISCO: Helping Stroke Patients with ThermaSuit Cooling Trial is a Phase 2 study in ischemic stroke with rapid induction of hypothermia to 32 within one hour. This patient had induction followed by early malignant edema requiring decompressive hemicraniectomy while 33 C. This is the first report of hemicraniectomy in a therapeutically hypothermic patient.

Results 59 y/o woman presenting with a left MCA syndrome. Initial imaging demonstrated left M1 occlusion. She received IV tissue plasminogen activator (tPA) followed by thrombectomy with TICI 3 recanalization within practice guidelines. She was enrolled in SISCO trial. She was sedated with propofol, fentanyl, and versed for induction, reaching target temperature of 34 degrees within 60 minutes. She remained on sedation for shivering and temperature was maintained at 32 degrees with the Artic Sun. Imaging 12 hours after stroke demonstrated completed infarct with edema, midline shift, and lateral ventricle effacement. Hypertonic saline was initiated, and she underwent emergent decompressive hemicraniectomy. Balancing the risk of worsening edema and coagulopathy caused by mild hypothermia, rewarming was initiated at 0.3 degrees C per hour. At the time of procedure patient was at 33.5 . A successful hemicraniectomy was performed without complications. six months demonstrated improvement with the patient returning home with modified Rankin 2, and cranioplasty performed without complications.

During SISCO, an emergency decompressive hemicraniectomy for malignant MCA syndrome was performed for a cooled patient without complication or increased bleeding. While therapeutic hypothermia has not shown an outcome benefit in previous clinical trials, these trials have had limitations rapidly reaching targeted temperature. This may have blunted the therapeutic effect. Using ThermaSuit, patients are able to reach target temperature significantly faster. Additional clinical trials are needed to determine if the therapeutic window for targeted temperature management in ischemic stroke patients improves outcome.

IV rt-PA guidelines exclude therapeutically anticoagulated or thrombocytopenic patients. These exclusion criteria may limit thrombolytic therapy to patients who might benefit. The objective of this study is to determine if IV rt-PA is safe and whether it increases neurocritical care resource utilization in this patient population.

Retrospective analysis of IV rt-PA treated patients receiving oral anticoagulation (warfarin (INR >1.7)), novel oral anticoagulant (NOAC), therapeutic heparin, low-molecular weight heparin (LMWH), or with thrombocytopenia (platelets < 100K). Patients were treated using SMART criteria (consent obtained for off label rtafter treatment. Increased neurocritical care resource utilization was defined as transfer from a primary to comprehensive stroke center solely for additional monitoring after off-label IV rt-PA use.

77 patients were identified. 35 patients received therapeutic warfarin and one had coagulopathy (unclear etiology); mean INR=2.2 (range 1.7-3). 5 received therapeutic IV heparin, 6 full dose (1 mg/kg BID) LMWH, and 18 therapeutic NOACs. 12 had thrombocytopenia (mean platelet count 77K). 7 received intra-arterial (IA) rt-PA, and 8 thrombectomy. There were 3 sICH (3.9%); for all sICHs there were mitigating factors that contributed (undiagnosed malignancy, adjunctive IA rt-PA, incorrect time of onset). Two developed hematoma at the catheter site with no clinical effect. 1 patient was transferred for the sole purpose of monitoring post off-label IV rt-PA.

These data suggest that IV rt-PA can be safely administered in therapeutically anticoagulated and thrombocytopenic patients, and sICH rates were similar to the NINDS cohort. The use of IV rt-PA in these patients may increase eligibility for acute stroke therapy, particularly where IA therapy is unavailable.

-PA in such patients does not appear to increase neurocritical care resource utilization though further study with a larger population is warranted.

Although proteinuria has been reported as a predictor of neurological deterioration, poor functional outcome and in-hospital mortality after ischemic stroke, scarce study investigated the relationship between proteinuria and the malignant middle cerebral artery infarction (MMCAI). This study aimed to determine whether proteinuria is associated with the development of MMCAI.

Patients with infarction in middle cerebral artery territory were reviewed. On admission, all patients underwent brain computed tomography (CT), the assessment of National Institutes of Health Stroke Scale (NIHSS) and Alberta stroke program early CT score (ASPECTS), and laboratory surveys, including urine analysis by using urine dipstick. Patients with known intracranial lesions or possible urinary tract infection were excluded. Patients with proteinuria were defined if urine dipstick demonstrates reading of 1+ to 4+, while others were defined as patients without proteinuria. Chronic kidney disease (CKD) was defined if either proteinuria or estimated glome identified. MMCAI was determined if a progressive conscious disturbance or signs of uncal herniation were recorded with a midline shift > 5mm on a follow-up brain CT.

We screened 1047 patients, and -five (37.2%) patients developed MMCAI, and 53 (43.8%) patients had proteinuria. Patients with MMCAI had a significant higher score of NIHSS, lower ASPECTS, less likely being dyslipidemia, and more likely having CKD and proteinuria than patients without MMCAI did. After adjustment for age, sex, dyslipidemia and ASPECTS, patients with proteinuria (OR= 2.658, 95%CI= 1.137 -6.216, p=0.0241) and CKD (OR = 2.595, 95%CI = 1.004 -6.711, p=0.0491) had a signifi 60ml/min/1.73m2 did not.

In conclusion, proteinuria is associated with the development of MMCAI. We suggest that proteinuria may be considered as a clinical predictor for the development of MMCAI.

Although tPA has been shown to improve outcome in ischemic stroke across various etiologies, tPA is contraindicated in stroke secondary to septic emboli due to a significantly higher risk of bleeding. The goal of this study is to determine the safety and short-term outcomes of acute ischemic stroke patients who underwent mechanical thrombectomy due to septic emboli from infective endocarditis (IE).

In this multi-center retrospective case series, we reached out to 48 thrombectomy centers known to our principal investigator. We have so far collected data from 5 hospitals across the US to look at outcomes after thrombectomy in patients who had an ischemic stroke from infective endocarditis. 8 centers reviewed their database and did not have eligible cases.

To date, we have collected a total of 10 cases (90% male; average age 53; 40% had a known history of IVDU). In 50% the valve implicated was bioprosthetic. 80% of the occlusions were M1, with the remaining being the carotid terminus (10%) and M2 (10%). Microbiology revealed that 30% were caused by streptococcus, 30% staphylococcus, 30% enterococcus, and 10% were polymicrobial. The average NIHSS on presentation was 19.8. 60% had received tPA prior to the thrombectomy (of those, 1/3 were known to have IE). The average best NIHSS after thrombectomy was 8.7 (averaged across 9 cases, the other case expired from new cardiomyopathy and multi-organ failure). 60% had hemorrhagic transformation (of those, 2/3 were tPA recipients).

Thrombectomy may be a safer and promising option in patients with ischemic stroke secondary to infective endocarditis. More data is required to compare the outcome of patients who received thrombectomy alone versus tPA followed by thrombectomy, and data collection is ongoing.

Therapeutic hypothermia may be an effective therapeutic measure for malignant cerebral infarction alternative to or in combination with decompressive craniectomy. The neuroimaging marker that suggests the favorable clinical course during therapeutic hypothermia is needed to predict the outcome and/or determine best and earliest timing to rewarm the patients.

We included 21 cases who received therapeutic hypothermia for malignant middle cerebral infarction in Seoul National University Bundang Hospital between July 2017 and May 2019. We measured Hounsfield unit of ischemic core in serial computed tomography scans in each patient. The nadir of Hounsfield unit of each patient was calculated. The difference of the nadir by the early clinical outcome (the survival at discharge) was analyzed.

The mean age was 71.6 ± 14.9 and the male comprised 61.9% (n=13). Three patients underwent early decompressive craniectomy plus therapeutic hypothermia and 18 patients received only therapeutic hypothermia. The mean target temperature was 34.1 ± 0.8 . A total of 14 patients (66.6%) survived at discharge. A total of 125 computed tomography scans were analyzed (about 6 scans per patient). The mean of the nadir Hounsfield unit of each patient was 19.3 ± 2.9 in the deceased patients and 16.1 ± 2.8 in the survived patients, and the difference was statistically significant (P-value = 0.034)

The nadir of Hounsfield unit in the ischemic core was lower in the survived group than the deceased group in malignant ischemic stroke patients who received therapeutic hypothermia. The change in Hounsfield unit in serial computed tomography scan may be used to estimate clinical course and optimal timing of rewarming or rescue craniectomy after therapeutic hypothermia. The volumetric analysis using semi-automated planimetry is currently being performed to elucidate this association further.

2 MHz Pulsed-wave Transcranial Doppler (TCD) increases the exposure of an intracranial thrombus to Tenecteplase (TNK-tPA) and facilitates early reperfusion. The aim of the present study is to ascertain if TCD along with TNK-tPA could improve functional outcome in patients treated with TNK-tPA after acute ischemic stroke (AIS).

This is a single center, prospective, interventional study. Patients with AIS with National Institutes of -tPA bolus) within 3 hours of symptom onset, were randomly allocated (1:1) to either 2 MHz pulsed-wave ultrasound for 120 min. (Sonothrombolysis)-intervention group or only TNK-tPA group. Ultrasound was delivered using a Mark head frame, immediately after the bolus of TNK-tPA. The primary outcome was improvement in the modified Rankin Scale score at days 30 and 90. The secondary end points were the occurrence of symptomatic intracerebral haemorrhages and death.

Between January 2016 and March 2019, 78 patients were randomly allocated to the Sonothrombolysis group and 77 patients received only TNK-tPA. At the end of 90 days, the Sonothrombolysis group achieved mRS 0-2 in 38/78 (48.70 %) compared to 32/77 (41.55 %) in the TNK-tPA group. The p-value is 0.02828. The result is significant at p < 0.05. The rate of sICH and mortality were 2.56 % in each group.

Sonothrombolysis of patients treated with TNK-tPA for AIS was feasible and safe, with some clinical benefits at 90 days. The recanalization rates and outcome are better than studies done with Alteplase. There was no increase in sICH or mortality. TNK-tPA should be the preferred drug for thrombolysis in AIS. The study should be carried out in multiple centers to see if the results of the present study can be validated.

Acute ischemic stroke is the second leading cause of death, especially if the patient did not receive the appropriate treatment geared towards a timely recanalization of the occluded vessels, including intravenous tissue plasminogen activator (IV t-PA) or endovascular thrombectomy. Little emphasis is given to the augmentation of collateral flow to offset the deleterious effect of ischemia or lessen the progression of the penumbral tissue into infarction. We present our initial experience with such vasoaugmentation strategy in patients with acute ischemic strokes.

We present o University. Our series included 12 patients with acute ischemic strokes. We excluded patients with a large vessel occlusion. All other patients were included regardless of whether they received IV-tPA or not. All patients had a CT angiogram including collateral imaging and CT perfusion study at baseline. After explaining to the patients or their next-of-kin, we started the patients on a standardized protocol of milrinone (50mcg/kg bolus followed by 0.75mcg/kg/minute). Outcome assessment was comparing the initial mRS and that of the mRS at discharge. Chi square contingency analysis was used with a set level of significance of p < 0.05.

Out of the 12 patients, 7 had good collaterals and 5 had poor collaterals. One of those poor collaterals patient had good cross flow from Pcom to the affected hemisphere, but still demonstrated poor collateral score. In our cohort, 8 (66%) achieved good neurological outcome of mRS of 4 or below with 3 patients (25%) achieving a discharge mRS of 0.

Conclusions collaterals and small infarction core. The presence of cross flow wasn't helpful.

The symptomatology of delayed cerebral ischemia (DCI) after aneurysmal subarachnoid hemorrhage (aSAH) is variable and often challenging to detect, particularly in patients with poor-grade aSAH. We report severe symmetric quadriparesis as a previously unreported symptom of DCI.

A 41-yearwas significant for intact brainstem reflexes and withdrawal in extremities. Initial treatments included aminocaproic acid and external ventricular drain insertion, followed by intra-aortic balloon pump placement for stress-induced cardiomyopathy. She subsequently underwent coiling of a ruptured left anterior choroidal artery aneurysm. On post-bleed day 5, she was noted to have new onset of decreased tone and minimal complex posturing to noxious stimulation in all extremities and severe inattention. Transcranial Doppler and digital subtraction angiography revealed moderate left greater than right middle cerebral artery and bilateral anterior cerebral artery vasospasm and she received balloon angioplasty and intra-arterial nicardipine twice. Post intervention, her quadriparesis moderately improved but she continued to have decreased tone and delayed movement initiation. On post-bleed day 11, brain MRI demonstrated infarcts in bilateral medial frontal lobes, bilateral basal ganglia & subinsular cortices. On day of discharge, she was able to spontaneously raise her left arm and legs, but only minimally moved her right arm to noxious stimulation.

This case report adds severe symmetric quadriparesis to the myriad of possible clinical symptoms of DCI after aSAH. Awareness of this uncommon clinical presentation could lead to timely detection and management of delayed cerebral ischemia after aSAH and improved clinical outcomes.

Brain MRI to determine infarct size is common in acute stroke management. Many patients cannot undergo MRI imaging due to instability, or imaging contraindications. A common CT head finding postthrombectomy is contrast extravasation, thought to be secondary to "leakage" of the blood brain barrier due to ischemia. We hypothesized that extravasation volume on post-thrombectomy CT scan correlates with final infarct size. Using CT head as a proxy for final infarct size may help guide clinical decision making when MRI scan is not possible.

We retrospectively examined a prospectively collected, IRB approved Stroke Code database from 1/1/2014 to 7/1/2018. Inclusion criteria included: anterior strokes that underwent thrombectomy, CT scan within 12 hours of thrombectomy with contrast extravasation, a MRI within 5 days of the thrombectomy. Demographics, diagnosis, imaging findings were extracted via chart review. We used the Alberta Stroke Program Early CT Score (ASPECT) score, to approximate the area of contrast extravasation (CT) and area of DWI hyper intensity (MRI). Each region of extravasation on CT head was deducted from a score of 10, resulting in the "estimated infarct size (EIS)". Each region of MRI DWI was was calculated.

We demonstrated usefulness of ASPECT scoring for comparing infarct volume between CT extravasation and final DWI infarct size. Post-thrombectomy contrast extravasation consistently underestimated final MRI infarct volume by 30%. This relationship, if validated, may be useful to approximate MRI stroke volume.

Dizziness is a vaguely-defined complaint involving the subjective experience of lightheadedness, disequilibrium, and room-spinning sensation. It is a frequently encountered problem in office visits and in the acute care setting, with approximately 2.5 million presentations to the emergency department annually. The differential diagnosis is broad, ranging from benign and of peripheral origin, to timesensitive and potentially fatal of central origin, including ischemic or hemorrhagic stroke and MS. It is estimated 10-40% of patients with dizziness receive stroke diagnoses . Despite the low percentage, diagnosis of posterior stroke is the one most feared to be missed by clinicians. We hope to establish a clinical scoring system for timely triage of presentation with dizziness.

We retrospectively reviewed charts of patients admitted at Hahnemann University Hospital between 2016 and 2018, following IRB approved protocol. Charts were chosen with primary inpatient admitting diagnosis: cerebral infarction due to thrombosis of basilar artery, dizziness and giddiness, and cerebral infarction due to embolism of post cerebral artery for a total of 176 charts. Patient charts were reviewed to identify predisposing factors, 15 data points for each patient.

Of 176 patients reviewed, 15 were found to have infarctions involving the posterior circulation, 8.5% diagnostic yield for stroke. We collected a total of 6336 data points to understand the disease process. Predisposing factors identified were chronic kidney disease, diabetes, hypertension, and hyperlipidemia. Surprisingly, previous stroke was not found to predispose to posterior fossa strokes. Common exam findings on presentation were hemiparesis and hemisensory loss. Statistical analysis is currently in process

We discuss our results in the context of previous efforts aimed at developing clinical predictors for posterior fossa stroke in patients presenting with dizziness.

High HIR (Hypoperfusion Intensity Ratio) is known to correlate with core size, infarct growth and worse clinical outcomes. Traditionally larger infarcts have been associated with higher rates of malignant cerebral edema and need for decompressive hemicraniectomy. Patients with High HIR and malignant profile (Tmax >10s greater than 50% of penumbra) are associated with increased risk of malignant cerebral edema.

As part of an ongoing study, we retrospectively identified all AIS patients with LVO who underwent CTP imaging between January 2018 to June 2018 in our healthcare system within 24 hours from symptom imaging studies (CT or MRI) were analyzed. HIR was dichotomized based on proportion of greater and less than 0.5 into malignant vs favorable profile and correlation for development of malignant cerebral edema and need for hemicraniectomy was analyzed using chi-square test of proportion for nominal variables and Wilcoxon Ranked Sum tests for the (skewed) continuous and ordinal variables.

A total of 62 patients with LVO were identified with a median age of 78 (IQR 62-87), NIHSS of 16 . Patients with high HIR suggestive of a malignant profile (n=23), regardless of reperfusion, were associated with increased risk of malignant cerebral edema compared to those with a favorable profile (n=39) (p<0.0007). 9 Patients with malignant HIR developed malignant cerebral edema compared to 2 patients with favorable HIR (RR=7.63, OR=11.89). 1 patient with malignant HIR underwent decompressive hemicraniectomy compared to none with favorable HIR.

Higher HIR and malignant profile, regardless of reperfusion, is associated with 7 times increased relative risk of development of malignant cerebral edema. These patients benefit from close monitoring and aggressive care for malignant cerebral edema including osmolar therapy and potential surgical intervention.

We present the case of a patient with basilar artery dissection with thrombus, who underwent successful mechanical thrombectomy with stenting and was ambulatory at discharge.

Patient is an 85-year-old female with past medical history of Ehler's Danlos Disease who presented with left sided weakness after being found down by her family.

NIHSS on arrival was 14 (left sided weakness 1/5), BP 173/86, glucose 184. Her CTA showed a basilar angiography was notable for a basilar dissection with reocclusion, which was treated with Enterprise stent placement with TICI3 reperfusion. MRI post intervention revealed right pontine infarct, punctate infarcts in cerebellum. Her exam at discharge was notable for improvement in her left sided strength, at 4/5. She was subsequently discharged to inpatient rehabilitation.

Mechanical thrombectomy and stenting of the basilar artery remains a largely experimental procedure, with few guidelines and little data on outcomes. We present a case of a patient with a basilar dissection who at discharge was ambulatory and near baseline.

Blood viscosity (BV) is the intrinsic resistance of blood to flow and characterizes blood stickiness. Several clinical and epidemiologic studies demonstrated an association between BV and the occurrence of major thromboembolic events. Though BV appears significantly higher in cases of lacunar or cardioembolic strokes, relationships with demographic and laboratory findings during the acute stage of ischemic stroke are unknown. We investigated the relationship between baseline characteristics and BV within 24 hours of symptom onset in patients with acute ischemic stroke.

We enrolled patients aged 40 years or older with documented histories of ischemic stroke or transient ischemic attack within 24 hours of symptom onset. A scanning capillary-tube viscometer (SCTV) (Hemovister, Pharmode Inc., Seoul, Korea) was used to assess the whole blood viscosity (WBV).

The mean age was 69.6±12.03 years and 44.4% were female. Of 189 patients, 68.3% had a history of hypertension; 27%, diabetes; 42.9%, hypercholesterolemia; 3.7%, coronary artery disease; and 18%, stroke. Additionally, 40.7% were current smokers. Sixty-one (32.3%) patients were taking antithrombotics regularly. Multiple linear regression analysis revealed that hematocrit was positively related with increased BV and prior antithrombotic use was related with decreased BV. Hematocritadjusted partial correlation demonstrated that prior antithrombotic use was significantly associated with decreased BV.

Prior antithrombotic use is significantly associated with decreased blood viscosity within 24 hours of symptom onset in patients with acute ischemic stroke. Our findings indicate that antithrombotic medications prevent stroke by inhibiting platelet function and by changing the hemorheological profile.

Ischemic stroke accounts for 80% of stroke and is the second cause of death in Brazil. The decision regarding thrombolytic treatment depends on clinical history, physical examination, and imaging. One challenge is the exclusion of situations called Stroke Mimics (SM).

A total of 132 patients admitted to the stroke unit were prospectively analyzed. They received a full clinical and laboratory evaluation for the diagnosis of stroke and aiming to rule out the SM possibility.

The study looked up for stroke etiology, demographical and epidemiological data, stroke-specific scales, 

sis, the occurrence of seizures and blood pressure lower than 150 mmHg at admission as variables of interest. The prevalence of SM and the use of thrombolytic therapy in this situation was concordant with medical literature.

The risk associated with anticoagulation in acute ischemic stroke (AIS) is uncertain. Anticoagulation is generally not indicated for early secondary stroke prevention, but may be considered in certain conditions. We assessed the use of a weight-based institution-specific heparin nomogram in AIS patients.

-New Haven Hospital who received anticoagulation with a continuous heparin infusion in the setting of AIS over a 12-month period. Anticoagulation was initiated with an initial infusion rate of 10 units/kg/hr without bolus, with subsequent increases in the infusion rate by 1 unit/kg/hr, based on aPTTs obtained every six hours until two subsequent aPTTs were within goal range. We collected indication for anticoagulation, dose at therapeutic aPTT, time to target aPTT duration of anticoagulation, transition to oral anticoagulant therapy, cerebrovascular/cardiovascular events and major and minor bleeding complications.

61 patients were included in analysis, 52% of which were male, with a mean age of 60 ± 17 years and an average weight of 82.96 ± 17.82 kg. Indications for AC were: intracardiac thrombus (31%), (sub)occlusive intra-arterial thrombus (35%), arterial dissection (18%), thromboembolic events and hypercoagulability (16%). The median time between diagnosis of stroke and initiation of anticoagulation was 53 hrs 21 mins. The time to goal aPTT was 44 ± 5.5 hours with a mean infusion rate of 14 units/kg/hr at time of goal aPTT. 60% of patients were transitioned to an oral anticoagulant and 3% of patients experienced a cerebrovascular event while on heparin infusion.

Our institution-specific heparin nomogram provides a safe anticoagulation strategy in AIS, but with a longer time to reach therapeutic goal aPTT range compared to previously published data. A more aggressive titration strategy with consideration of a higher infusion start rate may facilitate reaching the target aPTT within a shorter time frame.

Vertebral artery dissection (VAD) is one of the most common identifiable causes of ischemic stroke in young age patients forming intramural hematoma. VAD may occur spontaneously or secondarily to trauma, infection, or underlying arteriopathy. We report 2 cases of spontaneous bilateral VAD presenting with lateral medullary infarction

A 51-year-old woman transferred to the emergency room with vertigo. 5 days ago, she felt severe headache on the left temporal area.

On neurologic examination, ptosis, facial hypesthesia, dysmetria on the left side were noted, and dysarthria, dysphagia, right beating nystagmus were noted also. She had no past medical history and no familial history of stroke or cephalo-cervical trauma. Brain MRI depicted acute infarction in left lateral medulla and dissecting aneurysm of right VA and near occlusion of left VA on carotid enhanced MRA. disease was normal. She was treated with warfarin. A 43-year-old man visited to the emergency room with headache on the right occiput. On neurologic examination, ptosis, miosis, facial hypesthesia, dysmetria on the left side and hemibody hypesthesia on the right side were noted. He had no trauma history or risk factors for stroke except hypertension. Brain MRI depicted acute infarction in right lateral medulla and dissecting aneurysm in the bilateral vertebral arteries on carotid enhanced MRA. Laboratory tests showed no abnormal findings. All results were normal for young age stroke evaluation. He was treated with warfarin.

Although unilateral or bilateral VAD due to trauma or underlying medical conditions has been reported, spontaneous bilateral VAD is rare. It can present with lateral medullary syndrome or nonspecific symptoms such as headache only. Physicians should include VAD in the differential diagnosis for patients presenting with brainstem neurologic abnormality or headache, especially young patients.

Cerebrovascular complications (CVCs) occur in 20-55% of patients with infective endocarditis (IE) and manifest as ischemic stroke, meningitis or cerebritis with 75% occurring during first 2 weeks of treatment. CT or MRI brain can diagnose CVCs but are insensitive early on, precluded in critically ill patients and only demonstrate the sequelae. Transcranial Doppler (TCD) can identify high-intensity transient signals (HITS) associated with cerebral microembolization and may have a role in detecting emboli and preventing CVCs in IE.

Retrospective chart review and literature review.

We found 89 patients with strokes caused by IE at our institution from 12/2011 to 4/2018. TCDs were obtained on 26 patients, 16 abnormal for cerebrovascular abnormalities. Only 4 patients had 30 minute emboli monitoring performed of which one revealed HITS. Though MRI studies have shown microemboli in 82% of IE patients (Duval Ann Intern Med 2010), we only found 3 studies using HITS on TCDs as indicators of stroke risk in IE. In a prospective study of 30 patients with left-sided IE, CVCs occurred in 83% of patients with HITS on TCDs versus 33% of patients who did not (p=0.021) ( Lepur Scand J Infect Dis 2009). Two studies investigated 100 and 300 patients with cardiac sources of embolism and documented occurrence of HITS in 36% and 23% of subjects, respectively, with highest prevalence of HITS in patients with IE (Sliwka Stroke 1997 , Georgiadis Stroke 1997 .

Detection of HITS using TCD emboli monitoring has a potential to be an important tool for identifying cases of IE at highest risk for CVCs, especially in the early stages of antimicrobial therapy. This can aid further research into preventative interventions beyond antibiotics like earlier valvular surgery or vacuum assisted vegetation extraction.

Therapeutic hypothermia is considered as an effective therapy to reduce cerebral edema and intracranial pressure for malignant middle cerebral artery infarction, which can be used as a life-saving treatment alternative to or combined with decompressive craniectomy. However, malignant hemispheric infarction involving whole anterior, middle and posterior cerebral artery territory has been regarded as untreatable by any measures.

A 68-year-old man who had had right ventriculoperitoneal shunt for hydrocephalus since several years ago presented with global aphasia and right hemiplegia in May 2018. The brain magnetic resonance imaging showed large acute infarction involving whole left hemisphere including anterior, middle and posterior cerebral arterial territory by occlusion of distal internal carotid artery. As his family refused decompressive hemicraniectomy, therapeutic hypothermia using surface-cooling method (Arctic Sun® 5000) was initiated with a target temperature of 33.5 . The maximal midline shift on brain CT was approximately 15mm, five days after stroke onset, which led to foramen of Monro obstruction and hydrocephalus in the lateral ventricle of the opposite side. Since the hydrocephalus was controlled by draining of the cerebrospinal fluid into the ventriculoperitoneal shunt, the right hemisphere was saved and brain edema combined with midline shift gradually improved. The patient finally survived and was discharged.

This case may be the first that therapeutic hypothermia successfully treated large hemispheric infarction involving 3 cerebral arteries without decompressive craniectomy. Since the mass effect in our case was much larger than that of malignant middle cerebral infarction, we extended the duration of therapeutic hypothermia (33.5 ) to 7 days, which prevented herniation syndrome. Another interesting point is that we could manage contralateral hydrocephalus caused by extensive midline shift, heralding a fatal clinical course in malignant ischemic stroke, using the preexisting ventriculoperitoneal shunt.

Current AHA / ASA stroke guidelines list arteriovenous malformation (AVM) as a contraindication for intravenous alteplase (IV tPA) in ischemic stroke. While the associated risk of spontaneous intracerebral hemorrhage varies across the differing types of intracranial vasculature malformations, very little data or case reports exist regarding the risk of hemorrhage with intravenous thrombolytics for ischemic stroke in patients with vascular malformations.

A 65-year-old male with history of cirrhosis and known atrial fibrillation (not on anticoagulation) presented with acute onset left facial droop and left hemiplegia, NIHSS 9. Onset of symptoms were within the 4.5 hour window for IV tPA. A CT head demonstrated an ASPECTS score of 10. IV tPA was thus initiated. CTA of the head and neck revealed a right middle cerebral artery occlusion. Additionally, there was a subtle tortuosity of blood vessels within the dural surface of the right temporal lobe, suggestive of possible AVM. Given the stroke severity, tPA was continued and successful recanalization was completed by thrombectomy of the right M2 occlusion by aspiration, with confirmation of a dural based AVM. The patient did well, with no complications from tPA or thrombectomy and was discharged home with an NIHSS 2.

The decision to administer IV tPA in patients with symptoms of acute ischemic stroke is determined by last known well time and a non-contrasted CT. Vessel imaging should not delay administration of IV tPA as incidental findings may arise which may cloud the use of IV tPA in patients who otherwise may benefit from therapy. This case provides further insight that IV tPA in those with intracranial vascular malformations may be given safely with minimally increased risk.

The prevalence of Stroke Mimics (SM) can reach 30% of presumable stroke, according to some authors. Its presentation can predict the diagnosis of SM with a sensitivity and specificity of 90% and 91%, respectively. This study aimed to comparatively evaluate these data in a population hospitalized in a Stroke Unit.

The study prospectively analyzed a total of 132 patients admitted according to the suspicion of SM, the definitive diagnosis, etiology, demographic and epidemiological data, specific scales for stroke including features and its sensitivity and specificity in a specific population.

A cross-sectional analysis comprised 132 (53.1% female) patients, median age 66.5 years (56-74). The median NIHSS was 8 (4--2 in 87.1% of patients. Twenty-four patients (18.2%) presented with initial suspicion of SM, which was confirmed in 19 (14.4%). After univariate analysis on were statistically significant (p = 0.022 and p = 0.001, respectively). The multivariate logistic regression showed that the absence of facial paralysis (OR=4.57, p=0.004, 95% CI=1.60-13.03), seizure convulsion on admission (OR=11.02, p=0.001, 95% CI=2.82-43.03) and blood pressure at admission lower than 150 mmHg (OR=0.30, p=0.027, 95% CI=0.10specificity of 63.16% and 60.18% respectively, with an area under the curve of 0.682 (SE=0.0675, 95% CI=0.55-Conclusions sensitivity and specificity, probably secondary to selection bias. These data are inferior to the literature but better adapted to this study population.

Information collected from chart review and direct patient care.

A 51 year-presented with pre-syncope, abdominal pain, and malaise. He was febrile and tachycardic, and subsequently admitted for sepsis. Shortly thereafter, he experienced transient diaphoresis, expressive aphasia and right-sided weakness. MRI brain showed punctate ischemic cerebellar infarcts. There was high suspicion for embolic phenomena from sepsis or He acutely decompensated to complete non-responsiveness during the echocardiogram. CT brain showed diffuse air emboli in cerebral vasculature and subarachnoid air. He was placed in the left lateral decubitus position and managed with high concentration oxygen. Additionally, his antimicrobials were broadened to include fungal coverage. Thoracic CT revealed free air in the mediastinum between the candidate for surgical repair of his left atrium due to hemodynamic instability. Instead, he underwent urgent endoscopic esophageal stent placement. He then developed a STEMI, also thought to be due to air embolus, and went into cardiac arrest with return of spontaneous circulation achieved. The following day, he developed renal failure and coded again. Autopsy, in addition to massive cerebral edema and cardiac ischemia, demonstrated Strep Oralis bacteremia, bilateral adrenal infarcts and acute tubular necrosis.

is crucial for the ability to coordinate aggressive care. Open surgical repair of the left atrium and esophagus offers the best chance of survival, but its use may be limited by severe sepsis and hemodynamic instability.

The efficacy of mechanical thrombectomy (MT) for acute ischemic stroke (AIS) due to large vessel occlusion (LVO) is well established in the anterior circulation (AC). AIS from LVO in the posterior circulation (PC) differs from the AC in myriad ways, including presentation and resistance to hypoxia. We aim to characterize the differences in risk factors and outcomes of MT for AC vs PC stroke.

Demographic data was collected for cases of AIS undergoing MT from January 2014 to January 2018 with follow-up imaging and documented functional status at discharge. Operative reports were reviewed for procedural data including stroke onset to groin puncture time, number of passes of the stent retriever, and onset to recanalization time. Radiology reports of postprocedural non-contrast CT images of the head were assessed.

During the study period there were 156 eligible patients (131 AC and 25 PC). Atrial fibrillation (36.6% and 12.0%, p=0.017) and hyperlipidemia (42.0% and 20.0%, p=0.040) were more common in AC strokes while family history of stroke was more common in PC strokes (36.0% and 13.0%, p=0.005). Mortality erence in procedural factors or hemorrhagic complications.

AC stroke but not PC stroke.

Our data shows that PC stroke has a higher mortality rate than AC stroke after MT with no difference in procedural factors or hemorrhagic complications. The higher mortality rate in patients with PC stroke is likely inherent to severe disability from basilar artery occlusion rather than recanalization therapy. The data also support worse functional outcome in AC strokes with increasing age and number of passes.

Calcinosis is a dysregulation of vascular calcium deposition characterized by small vessel calcification and secondary fibrosis. The effect of systemic calcinosis on mineralization within the central nervous system is underreported and poorly understood.

A 65-year old man presented to ICU for possible hemorrhagic transformation of a recent left MCA stroke. His medical history was notable for atrial fibrillation, end-stage renal disease, calciphylaxis on warfarin, and parathyroidectomy. His post-stroke hospital course was notable for mildly elevated serum phosphorus. The patient started apixaban two weeks post-stroke as anticoagulation for atrial fibrillation, and underwent a routine CT head one day later. The scan showed extensive high-density signal along the cortex of the recently infarcted left MCA territory, initially misinterpreted as hemorrhagic transformation. The signal measured at 80-100 Hounsfield units, higher than expected for acute blood. A dual-energy calcium overlap map post-processing revealed the high-density material was consistent with acute mineralization, possibly potentiated by the patient's previous calciphylaxis. This case illustrates accelerated mineralization as a mimic of acute hemorrhagic transformation. Dual-energy CT is useful for differentiating hemorrhagic transformation from mineralization, and may play a special role in patients with renal disease or history of calciphylaxis.

This case illustrates accelerated mineralization as a mimic of acute hemorrhagic transformation of stroke in a patient with ESRD and history of calciphylaxis. Dual-energy CT can differentiate between intraparenchymal hemorrhage and calcification with high accuracy using material decomposition. This imaging technique may have an especial benefit in patients with renal disease or disordered mineralization. Accelerated mineralization post-stroke may worsen cerebral vessel compliance and risk of future stroke, and merits further investigation.

Systemic inflammatory response syndrome (SIRS) without infection is a surrogate of a systemic immune response and has been related with poor outcome in several vascular diseases. We investigated associations of SIRS with long-term functional outcome and contributing factors after intracerebral hemorrhage (ICH).

We analyzed consecutive spontaneous ICH-patients from our prospective cohort-study (2008) (2009) (2010) (2011) (2012) (2013) (2014) (2015) . SIRS was defined according to standard criteria: i.e. two or more of the following parameters during hospitalization: body-temperature <36°C or >38°C, respiratory-rate >20 per minute, heart-rate >90 per consisted of the modified Rankin-Scale(mRS) at three and twelve months investigated by adjusted ordinal shift-analyses. Bias and confounding were addressed by propensity score matching and multivariable regression models.

Of 780 patients with ICH 21.8% (n=170) developed SIRS during hospitalization. SIRS-patients showed more severe ICH compared to without; i.e. larger ICH-volumes (18.3cm³, IQR(4.6-47.2) versus 7.4cm³, IQR(2.4-18.6);p<0.01), increased intraventricular hemorrhage (57.6%,n=98/170 versus 24.8%,n=79/319;p<0.01), and poorer neurological admission status (NIHSS 16, IQR(7-30) versus 6, IQR(3-12);p<0.01). ICH severity-adjusted analyses revealed an independent association of SIRS with poorer functional outcome after three (OR 1.80, 95% CI(1.08-3.00);p=0.025) and twelve months (OR 1.76, 95%CI(1.04-2.96);p=0.034). Increased ICH-volumes on follow-up-imaging (OR 1.38, 95%CI(1.01-1.89);p=0.05) and prior liver dysfunction (OR 3.01, 95%CI(1.03-10.19);p=0.04) were associated with SIRS.

In ICH patients we identified SIRS to be predictive of poorer long-term functional outcome over the entire range of mRS-estimates. Clinically relevant associations with SIRS were documented for prior liver dysfunction and hematoma enlargement.

Acute major bleeding secondary to trauma is a significant complication of anticoagulated patients. In -threatening in the absence of a specific reversal agent.

ANNEXA-4 was a prospective, single-arm, open-label study evaluating the efficacy and safety of -primary efficacy endpoints were percent change from baseline in antiefficacy over the first 12 hours after treatment, as determined by an independent adjudication committee. Safety outcomes (including thrombotic events and death) were evaluated over 30 days.

Among 352 patients enrolled in the study, 113 (32.1%) had a bleed associated with trauma (99 intracranial [ICH] , 14 non-ICH). Mean age was 80.5 years. Eighty-three patients took apixaban, 25 rivaroxaban, 4 enoxaparin, and 1 edoxaban. Of the 99 ICH patients, 41 (41.4%) had bleeding in multiple compartments. The mean hematoma volume in the 13 trauma patients with single-compartment intraparenchymal bleeding was 11.3 cc. Among efficacy-evaluable ICH patients, 58 of 70 (82.9%) had excellent or good hemostatic efficacy. The percent reduction in anti-ICH patients taking apixaban and rivaroxaban, respectively. The 30-day rates of thrombotic events and mortality were 9 of 113 (7.9%) and 13 of 113 (11.5%), respectively.

Conclusions high rate of excellent or good hemostatic efficacy, with a relatively low occurrence of thrombotic events. These results are comparable to what was observed for ANNEXA-4 patients with spontaneous bleeding events, and suggest that andexanet alfa could be a safe and effective treatment in the traumatic population.

-factor prothrombin complex concentrates -related ICH.

Adult patients (18 years or older) admitted to Yale--related ICH who evaluated at approximately 6 hours after the baseline CT scan. Secondary outcomes included mortality and modified Rankin Score (mRS) at hospital discharge. Chi-square test and multivariable logistic regression analysis were used for unadjusted and adjusted analyses, respectively.

Twenty--related ICH were included in the s 13 patients received AA). Majority of the patients were anticoagulated for atrial fibrillation (n=23, 79%). group (unadju patients (77%) in AA group (unadjusted p=0.63). There was no difference in mRS at discharge, 6 patients -3 compared to 5 patients (38%) in AA group (unadjusted p= 0.96). Multivariable analyses adjusted for age, sex, race, and baseline mRS confirmed the absence of these associations (all p>0.20).

In our limited sample size, there was no significant difference in the degree of hemostasis achieved, allvalidate these results are warranted.

Symptomatic intracranial hemorrhage (sICH) following mechanical thrombectomy (MT) for acute ischemic stroke (AIS) due to large vessel occlusion (LVO) is a rare but devastating complication. However, it is difficult to differentiate sICH from contrast extravasation on early post-procedural computed tomography (CT). We aim to evaluate the rate of sICH and whether the presence of hyperdensities (HD) on post-procedural CT predicts functional outcome after MT.

Demographic data was collected for cases of AIS undergoing MT from January 2014 to January 2018 with available follow-up imaging and documented functional status at discharge. Operative reports were reviewed for procedural data and radiology reports of CT head performed immediately, and 24-36 hours post-MT were assessed.

Of the 135 patients studied, 60 (44.4%, 60/135) had HD on immediate postoperative CT and 35 (25.9%, 35/135) were contrast extravasation (CE) due to resolution of HD on CT at 24patients developed new HD on follow-up CT, resulting in a total of 39 patients (28.9%) having ICH and 8 (5.9%) having sICH. In subgroup analysis, cardiac comorbidities were more common in CE patients than ICH patients (51.4% and 23.1%, p=0.016) with no mortality or outcome differences. Diabetes mellitus (DM) was more common in sICH patients than those with CE and asymptomatic ICH (75.0% and 30.3%, p=0.015). The mortality rate of sICH patients was higher (75% vs 9.1%) and the survivors had worse discharge NIHSS than pat difference in procedural factors or preference for circulation between any groups.

Our data show that presence of HD on immediate postoperative head CT does not predict mortality and is not related to circulation or procedural factors. sICH is more common in patients with DM and associated with higher mortality rate and poor functional outcome.

Consecutive patients admitted to the health system with tSD -PCC between May 2017 and April 2019 were analyzed. Baseline demographics and outcomes were abstracted through -up CT. tSDH volume was calculated using the ABC/2 method. Descriptive statistics were used to analyze the -PCC and its association with outcomes.

-PCC for tSDH were analyzed. The median age was 87 [56 to 100], -PCC was 27.6 units/kg. Patients with HE had a median dose of 24.4 units/kg (14.9 to 52.8) versus 27.7 units/kg (12.6 to 48.4) for patients without HE. HE was seen in 48% of patients (Rivaroxaban in 2, Apixaban in 13). Among patients with HE, 27% of patients died or went to a skilled nursing facility vs 19% in those without HE. Discharge to home or acute rehab was lower in those with HE (73%) versus those without HE (81%)

DOAC use was associated with higher rates of hematoma expansion and worse outcomes in patients -PCC. Treatments for the reversal of DOAC related tSDH should be investigated

Intracerebral hemorrhage (ICH) is associated with peripheral immune dysfunction and infection. We aim to evaluate peripheral immune responses to ICH and associations with infection and ICH outcome.

100 consecutive spontaneous ICH patients admitted to a tertiary center (12/2013-9/2015) were included. Patients with secondary ICH and transition to comfort measures within 24 hours were excluded. ICH score, discharge modified Rankin Score (mRS), antibiotics use, acute clinical infections including pneumonia, bacteremia, and urinary tract infection were systematically adjudicated using modified PANTHERIS criteria. Peripheral immune dysfunction was characterized by lymphopenia and lower lymphocyte to neutrophil ratio (LNR). Continuous variables were compared using student's t or Wilcoxon test; univariate associations assessed using Pearson's or Spearman's correlation depending on distribution. Ordinal logistic regression used to evaluate independent effect of LNR on discharge mRS. (JMP Pro 14.0).

Cohort had mean age 70 years, 46% female, median ICH score 1 [IQR 1-3] and median discharge mRS of 4. Thirty-nine patients had suspected clinical infection treated with antibiotics, where only 11 met modified PANTHERIS criteria for infection. Lower %lymphocyte (p<0.003) and LNR (p<0.0001) on post-ICH days 2-3 were associated with worse discharge mRS and higher ICH scores (p's <0.01). Lower mean LNR on post-ICH days 1higher mean LNR on post-ICH days 3-5 (p<0.003). LNR on post-ICH day 2 is independently associated with mRS (p=0.0002) after adjusting for ICH score and sex.

Acute post-ICH lymphopenia and reduced LNR are associated with ICH score, infection and worse discharge outcome. LNR emerged as independent predictor of ICH outcomes in preliminary analysis.

determine how acute lymphopenia mediates ICH infection risk and outcome.

Prolonged length of stay (LOS) in the intensive care unit (ICU) is associated with significant medical complications and higher costs in patients with spontaneous intracerebral hemorrhage (ICH). Aim of this study is to assess predictors of prolonged ICU LOS in ICH.

We conducted a retrospective analysis of ICH patients admitted to our institution over a seven-year period. Demographics, clinical data, and laboratory studies at presentation were recorded. Initial CT scans were reviewed to determine location, hematoma volume, and presence of intraventricular extension. Surgical interventions, insertion of an external ventricular drain (EVD), and medical complications, including infections and deep vein thrombosis/pulmonary embolism (DVT/PE) were reviewed. LOS was calculated based on the number of midnights spent in the ICU. Patients spending less than 24-hours in the ICU were excluded.

130 ICHs were analyzed. The mean age was 61.2 ± 14.1 years and 50.7% were females. Prolonged LOS, defined by using the point of change and cumulative sum methodology analysis after normalization of the sample, was found to be > 8 days. Intubation at presentation (p<0.01), presence of IVH (p<0.05), insertion of EVD (p<0.01), surgical evacuation (p<0.01), chest infections (p<0.0.1) and DVT/PE (p<0.01) were associated with prolonged LOS, while location of the hemorrhage, hematoma volume, and ICH score at presentation were found not to be significant.

This is a preliminary analysis to identify predictors of prolonged ICU-LOS in intracranial hemorrhage. Chest infections and DVT/PE were associated with prolonged LOS. Surgical intervention, intubation at presentation, and insertion of EVD were also independent predictors. These findings suggest that early EVD weaning or shunt placement, and potentially early tracheostomy could help in decreasing the ICU-LOS in patients with ICH.

Diffusion weighted imaging (DWI) lesions are found in nearly 50% of patients with acute spontaneous intracerebral hemorrhage (sICH). However, the timing of DWI lesions after sICH ictus remains unknown. The purpose of this study is to estimate the timing of new DWI lesions after acute primary sICH. By establishing a time frame, potential pathophysiologic mechanisms for DWI lesions can be elucidated.

Between September 10, 2011 and January 15, 2017, patients were enrolled in a prospective study examining DWI lesions in acute primary sICH. Enrolled subjects received a research brain MRI after admission blinded to the clinical teams. During the same admission, select patients received a separate brain MRI as part of clinical care. Subjects with 2 scans were identified from the study cohort, and their imaging evaluated for DWI lesions. When compared to the first MRI scan, the presence of a new DWI lesion on the second MRI scan was defined as a new DWI event. A Kaplan-Meier analysis was performed to estimate the time to a new DWI event from the first MRI scan.

Among 121 enrolled subjects, 65 (53.7%) had two brain MRIs. Mean age was 56.2 years, 60% were male, and 49.2% were African American. The median ICH score was 1 (IQR 1). Median time from sICH onset to first MRI was 1.2 days (IQR 1.2). Median time from first MRI to a new DWI event was 4.9 days (95% CI, 3.0 to 7.8). Median time between the first and second MRI was 2.1 days (IQR 2.6).

Our data suggest that new DWI lesions occur 6 days after sICH ictus. Therefore, acute interventions during the first 24 hours after sICH admission may not be associated with DWI is needed to elucidate potential mechanisms associated with DWI lesions in sICH.

Intracranial hemorrhage (ICH) is a common complication in children on ventricular assist device (VAD) support, though bleed severity is highly variable. This study examined factors associated with ICH requiring neurosurgical intervention in this at-risk population.

Children aged 1 month-18 years old admitted between 2003-2018 with a diagnosis of intraparenchymal hemorrhage (IPH) or subdural hemorrhage (SDH) while on VAD support were identified retrospectively from an institutional database using ICD-9 and ICD-10 codes, after obtaining IRB approval. Patients requiring neurosurgical intervention (NS+) were compared with those who did not (NS-) using Manniables).

In total, 23 children met inclusion criteria. Of those, 4/23 (17.4%) required neurosurgical intervention bleeds occurred in 3 patients (2/4 NS+, 1/19 NS-). NS+ patients were older at bleed (mean 11.7 ± 5.6 years vs 4.8 ± 5.2 years, p = 0.04). All 4 NS+ patients were taking warfarin, versus 2/19 NS-patients (p=0.03); none of the NS+ patients had supratherapeutic INR. Number of antiplatelet agents did not differ between groups (1.4 ± 1.3 NS-vs 1.0 ± 0.8 NS+, p = 0.65). Patients received a median of 4 CT scans (IQR 3-7) with no significant difference between surgical and nonsurgical groups (p = 0.56).

Among our cohort, older children and those on warfarin were more likely to require neurosurgical neurosurgical treatment, though results should be interpreted cautiously given small numbers. Patients received multiple CT scans, though only a minority ultimately required neurosurgical intervention.

unnecessary CT scans in this population.

Elevated intracranial pressure (ICP), usually monitored by invasive ICP-measurements, is associated with mortality in intracerebral hemorrhage (ICH). The non-invasive evaluation of pupillary function using automated pupillometry is increasingly used in critical-care settings. The association of various pupillary parameters assessed by automated pupillometry with ICP is unestablished, specifically the sensitivity and specificity during ICP-elevation and the performance of sympathetic versus parasympathetic parameters.

We enrolled 8 ICH patients admitted to our neurocritical-care unit who received invasive ICPmeasurement by an external-ventricular-drain (EVD). We monitored parameters of pupillary reactivity [i.e. light-reflex latency (Lat; s), constriction and re-dilation velocities (CV, DV; mm/s), and percentage change of apertures (per-change; %)] using a portable pupilometer (NeurOptics®) as well as corresponding ICP values up to every 30 minutes for the duration of hospital stay. Receiver Operating Characteristic (ROC) analysis was performed to investigate associations between changes in pupillary reactivity and elevated ICP. Sensitivity and specificity of sympathetic and parasympathetic pupillary parameters were analyzed to evaluate associations between pupillary reactivity and ICP-elevation

In 4 patients (3 women, mean age 71.5±4.2 years), without ICP-elevation and no midline shift upon neuroimaging, 170 assessments were compared to 72 assessments in 4 patients (2 women, 52.0±17.2 years) during ICP-levels >20mmHg and corresponding midline shift. ROC-analyses revealed a significant negative association of all assessed pupillary parameters with ICP-elevation. Best discriminative thresholds for ICP-elevation were: CV<0.8mm/s, per-change<13%, Lat<0.3s, and DV<0.3mm/s. The highest sensitivity and specificity (i.e. 84.5% and 90.6%; p<0.001) for an association with concomitant ICP-levels >20 mmHg were found for a combination of the parasympathetic parameters CV<0.8mm/s and per-change<13%.

Our data suggest an association between non-invasively detected changes in pupillary reactivity and elevated ICP. Parameters of parasympathetic pupillary modulation seem most reliable to indicate ICPelevation.

Spontaneous ICH (sICH) remains a deadly complication from the use of direct oral anticoagulants -PCC for the reversal of DOAC -PCC in the prevention of hematoma expansion (HE) in DOAC associated sICH across a large health system.

Consecutive patients who were admit -PCC between May 2017 and April 2019 were analyzed. Baseline demographics and outcomes were abstracted through retrospective chart review. HE was defined as volume>33% or >12.5mL between baseline and follow-up CT. sICH volume was calculated using the ABC/2 method and IVH score. Descriptive statistics were used -PCC and its association with outcomes.

-PCC for sICH. The median age was 80 (48-95), 72% were Caucasian and --PCC dose of 24.9 units/kg (15.5-56.4) compared to 20.5 Units/kg (15.7 to 53.4) with HE. HE was seen in 21% of patients (Rivaroxaban in 4, Apixaban in 5). Among patients with HE, 22% of patients died or went to a skilled nursing facility vs 26% in those without HE. Discharge to home or acute rehab was similar in both groups

While rates of mortality and discharge disposition were similar between those with and without HE, -PCC. Treatments for the reversal of DOAC related sICH should be investigated further.

Elderly patients with mild traumatic brain injury (mTBI) are frequently admitted to an intensive care unit (ICU), which is potentially both harmful and unnecessary. However, little exists to inform early decision making to determine appropriate utilization of ICU care. Here we sought to elucidate factors available upon admission to identify geriatric patients who could safely be monitored in a non-ICU setting.

Adults 65+ years admitted with isolated mTBI, defined as positive radiologic study and Glasgow Coma Scale (GCS) 13 -15, between January 2011 -December 2016 were identified. Primary outcomes were ernight stay, no surgery, no intubation, and discharged home) and Glasgow Outcome Scale (GOS). Positive outcome was defined as GOS 4 -5 and

A total of 234 patients met criteria. Of these, 27 underwent emergent neurosurgical intervention., leaving 207 for analysis. Most presented with GCS 15 (77.8%) and were admitted to ICU (85.5%). Nearly point decrease in GCS during hospital stay. Upon discharge, 89.9% were classified GOS 4 -5. Predictors 0.007), and no home use of anticoagulant/antiplatelet medication (p = 0.007). Presence/type of a single intracranial hemorrhage (ICH) was not significantly associated with outcome, but presence of bilateral or multiple lesions independently predicted poor outcome (p = 0.040).

Overtriage of patients to an ICU is costly, resource intensive, and avoidable. Here, we suggest a conservative framework to assist the determination of which patients can be safely observed in non-ICU population who present with mTBI.

Perihematomal edema (PHE) is a known predictor of outcome after intraparenchymal hemorrhage (IPH), but factors contributing to edema formation are incompletely understood. Tissue water uptake measured using Hounsfield unit density on CT scan has emerged as a predictor of edema in ischemic stroke. The aim of this study was to examine this association in IPH, where the theoretical driver for edema volume is not anoxic cellular injury, but rather exposure of tissue to blood.

Women's Hospital were prospectively enrolled between September 2017 and March 2019. PHE and hematoma were identified on CT scans performed at admission and an average of 10.5 +/-4.3 hours later. Hematoma volume, hematoma surface area and PHE volume were measured. Net water uptake (NWU) was calculated as the percent change in PHE Hounsfield unit density compared to normal contralateral hemisphere. Associations between variables were examined with Pearson correlations and regression analyses.

Hematoma volume and surface area at admission were significantly associated with PHE volume on the admission scan (r = 0.90, p < 0.001 and r = 0.69, p < 0.001 respectively) and at the follow-up time-point (r = 0.86, p < 0.001 and r = 0.84, p < 0.001 respectively). There was no association between NWU and PHE volume at either time-point (r = 0.23, p = 0.14 and r = -0.01, p = 0.99 respectively). In multivariable analysis, hematoma volume at admission remained an independent predictor of PHE volume on the follow-

These results suggest that, unlike in ischemic stroke, PHE volume is not related to water content. Rather, hematom may suggests new avenues to predict edema formation.

The risk of hematoma expansion (HE) in patients with recent intracranial hemorrhage (ICH) receiving therapeutic anticoagulation (AC) is not known. We aim to characterize complication rates and factors associated with HE in these patients.

We performed a retrospective cohort study of adult patients at Harborview Medical Center between 2014-2019, who presented with ICH and were therapeutically anticoagulated within 4 weeks after the ICH for a venous thromboembolic event (VTE). We excluded patients with ICH due to hemorrhagic conversion of ischemic stroke, venous sinus thrombosis, or an aneurysm consequently secured. We assessed the rate of HE, defined as either radiographically proven expansion requiring cessation of AC, or death due to HE. T-tests and chi-squared tests were used to analyze factors associated with HE.

-94), 20% were female. We identified 42% SDH, 30% IPH, and 28% multicompartment ICHs, 70% due to trauma, 14% hypertensive, and 16% other etiologies. Anticoagulation was initiated an average of 10.5 +/-1.0 days after ICH. Overall, 18% developed HE, one third of whom died. Most patients (64%) experienced no complications, 10% developed minor extracranial bleeding events with AC subsequently resumed. Patients with HE were older (61 vs. 52), had higher GCS (12 vs. 9.2), lower hematoma volume (36% vs. 48% > 30 cc), larger maximal SDH diameter (13.2 vs 8.31 mm), anticoagulated earlier (8 vs. 10 days), and lower maximal PTT (110 vs. 122), although trends were not statistically significant. There was a marginally significant association between HE and the presence of hydrocephalus (p<0.07).

While AC in patients with acute ICH can be safely tolerated, there is a substantial proportion demonstrating HE. Our analysis was limited by the sample size. Larger studies are needed to identify clinical and radiographic features associated with complications.

Intracerebral hemorrhage (ICH) is a disease that is associated with high morbidity and mortality. We examined our center's experience with surgery for ICH and clinical outcomes.

We prospectively enrolled patients with spontaneous ICH from 2009 to 2018. Patients were divided into two groups based on whether they received surgical or conservative management. Surgical interventions included hemicraniectomy and/or hematoma evacuation. Multivariable regression analysis was conducted to compare the clinical outcomes after adjusting for potential confounders. Adjusted odds ratio (aOR) or adjusted mean difference (aMD) were reported.

We included 442 patients, 83 (19%) had surgery and 359 (81%) did not. Of the surgical group, 32 (39%) had hematoma evacuation, 7 (8%) had hemicraniectomy, and 44 (53%) had both. Clinical characteristics were comparable in both groups. In the surgical group, NIHSS and glucose were higher and creatinine was lower compared with the nonoperative group. Through multivariable analysis, we identified independent predictors of surgery in ICH patients including baseline hematoma volume (aOR 1.04, 95% CI 1.03-1.05; p=0.001) and enlargement with (aOR 2.81, 95% CI 1.44increase in hematoma volume, there was a 4% increase in the odds of having surgical intervention.

was less likely to have a favorable discharge disposition to home or inpatient rehabilitation (49% vs. 66%; p=0.01). Surgery was independently associated with longer ICU length of stay (aMD 2.96, 95% CI 1.38,-4.54; p=0.001) and hospital length of stay (aMD 3.63, 95% CI 0.73-6.54; p=0.01) after controlling for potential confounders.

In our patient population, baseline hematoma volume and expansion were independent predictors for surgery in ICH patients. After controlling for other variables, surgery did not impact ICH outcomes and was associated with prolonged ICU and hospital length of stay.

Moyamoya disease (MMD), an intracranial vasculopathy characterized by internal carotid artery hypoplasia, often presents with intracerebral hemorrhage (ICH) presumably due to rupture of fragile collateral vessels. Although MMD-related ICH is generally managed similarly to spontaneous ICH, we present a case in which standard management strategies may have led to an unprecedented catastrophic outcome.

Case report.

A previously healthy 37-year-old female presented to the emergency department with right-sided weakness, dysarthria, and headache. She was intubated for airway protection. A head computed tomography (CT) demonstrated a large left basal ganglia ICH. CT angiogram revealed diffuse narrowing of the entire anterior circulation with robust posterior communicating arteries. Brain magnetic resonance imaging (MRI) revealed prominent collateral vessels and sulcal hyperintensities ("ivy sign") consistent with MMD. Given these findings, systolic blood pressure was kept under 140 mmHg for the first 24 hours. The following day, the patient's mental status gradually worsened. Workup including repeat head CT, infectious and metabolic panels, as well as electroencephalogram (EEG) were unrevealing except for a decreased end-tidal carbon dioxide (CO2). Two days after presentation, the patient acutely developed fixed and dilated pupils. EEG concomitantly revealed slowing and attenuation of the background. Repeat CT head showed new diffuse cerebral edema with tonsillar herniation. Despite hyperosmolar therapy, paralytics, pentobarbital, and cerebrospinal fluid diversion, no improvement was noted. Unfortunately, brain MRI revealed multifocal brainstem infarcts with superimposed Duret hemorrhages.

Herein, we report diffuse cerebral edema as a complication of MMD-related ICH. We hypothesize that disruptions of delicate cerebral autoregulatory mechanisms led to extensive hypoxic-ischemic injury. In the setting of ICH, aggressive blood pressure management coupled with relative hypocapnia may have likely caused vasoconstriction of poorly compliant arteries leading to worsened cerebral blood flow and ischemia. Therefore, because of its complex pathophysiology, traditional blood pressure and CO2 targets should be revisited in MMD-related ICH.

It is unknown whether admission systolic blood pressure (SBP) differs among etiologies of intracerebral hemorrhage (ICH). Such differences may have implications for blood pressure -lowering strategies after ICH.

We compared admission SBP across ICH etiologies among patients in the Cornell AcutE Stroke Academic Registry (CAESAR), which has enrolled all adults with non-traumatic ICH at Cornell from 2011 through 2017. Trained analysts prospectively collected demographics, comorbidities, and admission SBP, defined as the first recorded value in the emergency department or upon transfer from another hospital. ICH etiology was adjudicated by a panel of board-certified neurologists using the SMASH-U criteria. We used ANOVA to compare mean admission SBP among ICH etiologies. After verification of model assumptions, multiple linear regression was used to adjust for age, sex, race, and Glasgow Coma Scale (GCS) score.

Among 494 ICH patients in CAESAR, admission SBP varied significantly across ICH etiologies, ranging from 138 mm Hg in those with structural vascular lesions to 167 mm Hg in those with hypertensive ICH (P <0.001). The overall difference in admission SBP across etiologies remained significant after adjustment for age, sex, race, and GCS score (P <0.001 by the Wald test). The mean admission SBP in hypertensive ICH cases was 17 mm Hg (95% CI, 11-24 mm Hg) higher than in ICH cases of all other etiologies combined. Among patients with a history of hypertension, the mean admission SBP was 12 mm Hg (95% CI, 5-19 mm Hg) higher in hypertensive ICH than in ICH cases of all other etiologies combined.

In a single-center ICH registry, admission SBP varied significantly among different ICH etiologies. Our results suggest that admission SBP is associated with ICH etiology rather than simply representing a physiological reaction to the ICH itself.

Incidence of clinical seizures after intracerebral hemorrhage (ICH) has been reported to range from 2.7 % to 17 %, with the majority occurring at or near onset. In the present study, we investigate incidence of clinical seizures in ICH subjects during hospitalization and evaluate whether clinical seizures are associated with poor clinical outcomes at discharge.

A retrospective review of consecutive patients with ICH admitted to the Baylor St. Luke's Medical center between January 2008 and December 2012 was conducted. Demographics, admission GCS, admission NIHSS, admission mRS, incidence of clinical seizures and clinical outcome at discharge were recorded. Associations between the presence or absence of seizures and clinical outcome, measured by the Glasgow Outcome scale (GOS), were investigated using a multivariate logistic regression model.

604 patients with ICH were included. Clinical seizures were identified in 15 subjects (2.5 %), presenting in a median time of 1.5 days post-admission (IQR 2). Outcome was significantly worse for subjects who experienced a seizure compared to subjects who remained seizure-free, poor outcome (GOS<4) was found on 73.3 % and 47.6 % respectively (OR 6.45, ]; p=0.035). This increased risk was significant after controlling for gender, ethnicity, admission GCS, admission NIHSS, admission mRS higher in the seizure group compared to the seizure-free group, 73.3 % vs 46.3 % respectively (OR 6.74, 95 % CI [1.16-39.2]; p=0.033) After adjusting for mortality and severe vegetative state (GOS 1, 2) there was no statistical significant difference between both groups (p=0.9632).

Our study shows a significant association between clinical seizures and poor clinical outcome at hospital discharge after controlling for admission status and other type of complications; however, the presence of clinical seizures did not influence in-hospital rates of mortality.

Despite the well-established use of the National Institutes of Health Stroke Scale (NIHSS) score as a severity scale for ischemic stroke patients, it is still unclear which score is best for intracerebral hemorrhage (ICH) patients. While some studies have looked at NIHSS and Glasgow Coma Scale (GCS) as a predictor of mortality and 3-month mRS, there is a dearth in the literature looking at how they affect longer functional outcomes. In this study, we look at and compare how initial NIHSS and GCS predict 6month functional outcomes in ICH patients.

One-hundred patients who underwent minimally invasive ICH evacuation, a standardized patient population, from December 2015 to October 2018 were retrospectively reviewed. We looked at NIHSS and GCS as a predictor of functional outcome at 6-months, defined as modified Rankin scale (mRS) 0-3. Multivariate regression models were constructed using clinical and statistical inferences to predict mRS. These variables were also correlated with 6-month mRS in multivariate analyses.

Of 100 patients, 33.0% (n=33) were female and the average age was 62.2 (SD=13.8). On admission, the median NIHSS was 17.0 (IQR 12.3-22.0) and the median GCS was 10.0 (IQR 7.0-13.0). Multivariate logistical analyses showed that higher NIHSS predicts worse 6-month mRS, however GCS does not (p=0.001 and 0.574, respectively). Correlation analysis with mRS at 6-months reveals that for every 9.17point increase in NIHSS, mRS increased by 1.

In this cohort, the admission NIHSS predicts 6-month mRS in ICH patients while controlling for significant covariates, while GCS does not appear to. Despite its simplicity and generalizability, the GCS lacks critical ICH elements that the NIHSS includes. The usefulness of the NIHSS as a predictor of ICH outcomes has been questioned, since ICH patients often have depressed consciousness on presentation, however we demonstrate its utility as a predictor of 6-month functional outcomes.

Among patients with intracerebral hemorrhage (ICH), it is unclear whether red blood cell (RBC) transfusions impact outcomes. We investigated the association between RBC transfusions and inhospital mortality in patients with ICH.

We performed a retrospective analysis using the National Inpatient Sample (NIS) database. We used standard diagnosis codes to identify non-traumatic ICH hospitalizations from 2002 through 2011. Our exposure was RBC transfusions during the ICH hospitalization and the outcome was hospital mortality. We performed multivariable logistic regression to estimate the association between RBC transfusion and outcomes after adjusting for demographics, Charlson Comorbidity Index (CCI), and hospital characteristics. However, given the absence of ICH severity and physiologic variable data within NIS, we performed additional analyses in a separate, single-center ICH cohort, adjusting for admission ICH and APACHE-II scores.

Of 619,167 non-traumatic ICH hospitalizations in the NIS, 22,904 (4%) patients received RBC transfusions. Patients receiving RBC transfusions had more comorbidities than those not receiving RBC transfusions (CCI >3: 28% vs 14%). RBC transfusion was associated with increased odds of hospital mortality (adjusted OR 1.29; 95% CI 1.22-1.38). In a separate cohort of 427 primary ICH patients, 32 (7%) patients received RBC transfusions during their hospitalization. RBC transfusion was not associated with hospital mortality after adjusting for ICH and APACHE-II scores (adjusted OR 1.96; 95% CI: 0.71-5.39).

RBC transfusion was associated with increased odds of hospital mortality after ICH. However, underlying medical comorbidities, acute physiologic derangements, and ICH severity may account for some of these ns on outcomes after ICH.

Deep venous thrombosis (DVT) is a common cause of morbidity and mortality in patients admitted to the neuro-intensive care unit (NICU). The aim of this work is to assess the incidence of DVT in patients diagnosed with intracerebral hemorrhage (ICH) and study its demographic characteristics.

A retrospective chart review of consecutive patients with ICH admitted to the Baylor St. Luke's Medical center between January 2008 and December 2012 was conducted. Subject demographics, admission status, incidence of DVT, hospital length of stay (HLOS), intensive care unit length of stay (ICU-LOS) and clinical outcome at discharge were recorded. Data was analyzed to assess the prevalence DVT in this period.

604 patients with ICH were included. DVT was identified in 53 subjects (8.8%). Median time to DVT from diagnosis was 120 h (IQR 192) after the initial symptoms of ICH. The mean age of patients with DVT was 59.5 (SD 1.3) and 25 subjects (47.1%) were female. 39.6 % subjects were of Caucasian ethnicity, 30.2 % African-American, 18.9 % Hispanic, 1.9 %Asian and 9.4 % were from other ethnicities. Median HLOS was 20 days (IQR 16.7) and ICU-LOS was 6 days (IQR 13.7). Moreover, 59.6% of patients who presented a hospital-acquired DVT had a poor clinical outcome at discharge (GOS<4).

ICH patients admitted to the NIC large prospective trials are needed to understand the baseline characteristics of patients at risk of DVT as well as the utility of surveillance and different prophylaxis methods.

Studies have demonstrated an association between high average systolic blood pressure (SBP), and increased SBP variability with worse clinical outcomes in non-traumatic intracerebral hemorrhage (ICH). Nevertheless, the optimal blood pressure target remains elusive. We aim at introducing an alternative approach to assess blood pressure in the acute phase of ICH, by using the metric of SBP dose, and showing that it provides a more robust association with clinical outcome.

We retrospectively evaluated 130 ICHs admitted to our institution over a seven-year period. Initial CT scans were analyzed to confirm the presence of intraparenchymal blood. Blood pressure was recorded at presentation and hourly for the first 24-hours. Mean SBP (mSBP) in the first 24-hours was calculated; SBP dose (dSBP) was calculated via the trapezoidal method from area under the curve (AUC), and divided in three groups: no dSBP (no time spent above 160 mmHg), moderate dSBP (AUC spent above 160 mmHg), and high dSBP (AUC above 180 mmHg). Discharge dispositions were used as surrogates of clinical outcome. Poor outcome included death, hospice, and long term acute care hospital.

-one patients (39.2%) had poor outcome. Of the patients in the no dSBP group none suffered poor outcome; 20% of the patients in the moderate dSBP group, and 52% of patients in the high dSBP group suffered poor mean SBP in predicting patient outcomes (p<0.001).

High dSBP in the first 24-hours was associated with worse clinical outcomes, and was a better predictor compared to mSBP. Blood pressure dose is a promising novel metric that deserves further study in the management of ICH.

Despite lack of ICH-specific therapies that improve outcome, current guidelines recommend treatment of ICH at tertiary care centers. As such, ICH comprises a large proportion of inter-hospital transfers (IHTs) to comprehensive stroke centers (CSCs) despite studies suggesting lack of mortality benefit and low CSC resource utilization. The subset of patients who derive the most benefit from a CSC is unclear.

Here, we create a triage model to identify ICH patients who can safely avoid transfer to a CSC.

A retrospective cohort of patients with spontaneous ICH transferred to our CSC was used to develop our triage model. Patients with early discharge from the Neuro-ICU without use of any CSC resource during hospitalization were identified as low risk, non-utilizers (LR--NU were identified and used to develop a triage model which minimized the likelihood of release of patients requiring CSC resource. This model was tested in a replication cohort for accuracy.

The development and replication cohorts comprised 358 and 99 patients respectively of whom 78 (22%) and 26 (26%) were LR-NU. Initial GCS and baseline ICH volume were associated with LR-NU in multivariate analysis. Presence of IVH and infra-tentorial location of ICH were also included. Initial GCS >13, ICH volume <15ml, absence of IVH, and supratentorial location had an AUC, specificity, sensitivity, PPV, and NPV of 0.72, 91.4%, 52.6%, 63.1%, and 87.4% respectively for identifying LR-NU. In the development cohort and 1 patient in the replication cohort had a neurosurgical intervention. However mostly these were for non-emergent AVM interventions.

Spontaneous ICH patients with initial GCS >13, ICH volume <15ml, no IVH, and supratentorial location might safely avoid IHT to a CSC. Validation in a prospective, multicenter cohort is warranted.

Metastatic cardiac myxomas have many neurologic complications, including intracerebral hemorrhage. Cardiac myxomas are rare intracardiac tumors. Though most myxomas are benign, the risk of malignant spread to the central nervous system (CNS) is well known. We describe a case of multiple recurrent intracerebral hemorrhages (ICH) occurring in the setting of a recently treated cardiac myxoma.

A 61-year old woman with a history of resected left atrial myxoma presented with a one-day history of left ear paresthesias. Computed tomography (CT) of the head was performed and demonstrated ICHs within the right frontal and parietal lobes and left cerebellar hemisphere. She had presented to an outside hospital several weeks earlier with similar symptoms with imaging demonstrating similar definitive evidence of malignancy or infection. Conventional angiography was negative for vasculitis. Brain biopsy showed no evidence of amyloidosis or glioma. At our institution, magnetic resonance imaging (MRI) of the brain with double inversion recovery also revealed no evidence of vasculitis; however, the study was concerning for multiple cavernous malformations. underwent genetic testing. No mutations associated with familial cerebral cavernous malformations syndromes were identified. Several months later, she returned to the hospital with recurrent symptoms. Head CT and MRI re-demonstrated multiple cavernous malformations with surrounding vasogenic edema, which were mildly increased compared with prior studies. Given progression of her MRI findings, concern for metastatic cardiac myxoma was raised.

Considering that 20-30% of patients with cardiac myxoma will have some form of neurologic complication, all should receive a comprehensive neurologic evaluation. Diagnosis is made with neuroimaging and brain biopsy. Primary treatment of cardiac myxoma includes surgical resection. When CNS lesions are present, chemotherapy or stereotactic radiosurgery should be considered.

An association between spontaneous hyperventilation, severity of disease at presentation, and poor clinical outcomes has been reported in patients with subarachnoid hemorrhage (SAH). We evaluated the relationship between early breathing changes and outcomes in patients with intracerebral hemorrhage (ICH).

Consecutive patients with spontaneous ICH were enrolled in an observational cohort study conducted between 2006 and 2019 at a comprehensive stroke center. Patient characteristics and functional outcome at discharge were prospectively recorded. Arterial blood gas (ABG) measurements and mechanical ventilation settings in the first 72 hours of admission were retrospectively collected, when available. Hyperventilation was defined as pCO2 <35 mmHg concurrent with pH >7.45 in spontaneously breathing patients, excluding mechanically ventilated patients not overbreathing the set rate of a control mode. We assessed for an association between early breathing changes, hemorrhage severity and hospital outcomes by univariate and adjusted analyses.

Early ABG data were available for 220 of 632 patients. Patients with ABG data had more severe hemorrhages than those without (median ICH Score 2 versus 1, p<0.001). Hyperventilation occurred in 52 (28%) of cases. There was no univariate association between hyperventilation and ICH Score, admission GCS score or initial hematoma volume. Lower initial pCO2 was associated with greater risk of in-hospital death (OR 0.94 per mmHg, 95%CI [0.89, 0.996], p=0.042) after adjustment for ICH Score, pneumonia and mechanical ventilation requirements.

Spontaneous hyperventilation is less common after ICH than SAH (55% vs 28%, respectively) and not associated with initial disease severity. The association between lower pCO2 and in-hospital mortality after ICH, independent of neurologic severity and comorbid respiratory complications, is consistent with findings of greater delayed ischemia and worse outcomes in spontaneously hyperventilating SAH patients. These associations may be mediated by a potentially modifiable underlying mechanism such as acute shifts in cerebral hemodynamics due to pCO2 changes.

ICH or SAH patients often undergo interhospital transfers to tertiary centers. Acute clinical deterioration diversion is often implemented via external ventricular drains (EVD'S). The safety and efficacy of leaving the EVDs clamped or open during inter-hospital transfer is not known. We aimed to implement a pilot during inter-hospital transport for hemorrhagic stroke patient.

Under the Neuroemergencies Management and Transfers (NEMAT) Program, Department of Neurosurgery at Mount Sinai Health System, we implemented this protocol in October, 2018. Patients with ICH or SAH requiring EVD placement prior to inter-transfer to a specialized center for ICH or SAH within our health system were enrolled. Recommendations for ICP management, for post-EVD drainage H and 10 cm or lower for ICH were included. EVD was clamped for transportation and a dose 1 g/kg of Mannitol was given just prior to transportation. ICP precautions were maintained throughout transportation.

15 (6 male,9 female_ patients who underwent inter-hospital transfers for ICH (n=3) and SAH (n=12) after placement of EVDs for raised ICP at the transferring hospital were included. All 15 patients required endotracheal intubation for transfer. 11/15 patients had an ICP less than 15 mmHg on arrival at the receiving hospital.

Conclusion: Protocolized care for ICH and SAH patients with EVDs and ICP management during interhospital transfers for patients is safe and feasible. Such a protocol could an help facilitate potentially rapid and safe life saving inter-hospital transfers for hemorrhagic stroke patients with EVDs in large urban health system to to hospitals with specialized definitive neurosurgical and neurocritical care.

Intracerebral hemorrhage (ICH) during pregnancy is abound with diagnostic and therapeutic dilemmas and contributes to pregnancy-related mortality. We present a pregnant patient with ICH due to Moyamoya disease to highlight these issues.

Case report.

A 40-year-old 32-week pregnant Asian woman presented after developing an acute onset headache followed by loss of consciousness. In the emergency department, she was comatose with bilateral pinpoint pupils and required intubation for airway protection. Initial CT head showed predominantly intraventricular hemorrhage (IVH) that emanated from the left thalamus. CT angiogram revealed highgrade stenosis of the left M1 segment with moyamoya collateralizations. Due to hydrocephalus, an external ventricular drain (EVD) was placed. The patient required admission to the neurocritical care unit for further monitoring of exam and vitals. Continuous fetal monitoring, and ultimately, successful Csection on day 3 of hospitalization was performed through collaboration with the obstetrics and gynecology (OB/GYN) team. Cerebral angiogram confirmed the diagnosis of unilateral Moyamoya disease as the cause of the patient's IVH. The patient was discharged initially to acute rehab and then home with minor cognitive deficits.

The work-up and management of ICH in pregnant patients can be challenging. Moyamoya disease is a non-atherosclerotic cerebral vasculopathy that can be included in the differential diagnosis for ICH in pregnant woman. The most common presentation of Moyamoya disease in adults is ICH, and it's mainly due to the rupture of dilated and fragile vessels in the basal ganglia, and rupture of saccular aneurysms within the Moyamoya collaterals. Pregnancy might increase the risk of ischemic or hemorrhagic stroke in women with Moyamoya, but available data is controversial. Cooperation between the neurocritical care and OB/GYN teams can assist in determining the risks and benefits of medications, imaging, and the need and timing for delivery, thus assuring optimal outcomes for the patient and infant.

Spontaneous intracerebral hemorrhage (ICH) is severely disabling, and survivors often require extensive rehabilitation to maximize recovery. Recovery for survivors discharged from index hospitalization is variable and incompletely explained by discharge functional capacity. We assessed whether discharge disposition was independently associated with long term recovery potential.

Patients with acute ICH hospitalized at a tertiary care comprehensive stroke center between 2006 and 2019 were enrolled in a prospective, observational study that recorded demographics, standard severity s was measured by the modified Rankin Scale (mRS) at discharge and three months. Discharge disposition were ordinalized by activity engagement level from highest to lowest as follows: home, 4; acute inpatient rehabilitation (AIR), 3; skilled nursing facili 2; and long-term acute care hospital (LTACH), 1. Ordinal regression was used to assess the prognostic association between discharge disposition and three month functional status by mRS, adjusting for the ICH Score and mRS at discharge.

Among 656 patients enrolled, 445 survived and had complete in-hospital data for analysis, and three outcomes at three months (mRS 4-6; 91.8% and 96.5% respectively), with most either bedbound or dead (58.8% and 63.2% respectively). Poor outcomes were less common among patients discharged to AIR (39.7%) or home (9.0%). The adjusted model found that a better discharge disposition was associated with more favorable three month mRS (odds ratio 0.54, 95% CI [0.40, 0.74], p=0.00011).

Discharge disposition captures prognostically important characteristics in patients with intracerebral hemorrhage beyond traditional case severity and functional status measures. Outcomes are poor for a large majority of patients unable to return home or qualify for acute rehabilitation. Whether the prognostic characteristics requiring nursing facility care are modifiable by increasing rehabilitation services in those care environments is not known.

2018 as a reversal agent for uncontrolled or life-threatening bleeding for patients taking apixaban and rivaroxaban. Approval was based on the results of interim analysis of the ongoing ANNEXA-4 multicenter, prospective, open-label clinical trial. Our institution began using the drug in August 2018. We report our clinical experience.

We conducted a retrospective observational study of patients admitted to Stanford Medical Center from -associated intracranial hemorrhage.

-associated ICH. The mean age was 74 (+/-10). 13 patients were male. The mean Glasgow Coma Scale score was 14. Hemorrhage types included intraparenchymal hemorrhage (9 patients), subarachnoid hemorrhage (4 patients), and subdural hemorrhage (2 patients). Hemorrhage was associated with head trauma in 4 patients (40%). Ten patients (67%) had "excellent" or "good" hemostasis defined by the ANNEXA-4 criteria. Three patients (15%) developed deep venous thrombosis. No patients developed pulmonary embolism or myocardial infarction. 30-day mortality was 20% (3 patients).

We describe a case series of patients who received andexanet alfa for intracerebral hemorrhage at a large medical center.

The incidence of intracerebral hemorrhage (ICH) is 24.6 per 100,000 person years. Nontraumatic spontaneous ICH is usually seen in setting of uncontrolled hypertension or cerebral amyloid angiopathy and commonly occurs in basal ganglia, cerebral cortex, brainstem or cerebellum. Spontaneous ICH in corpus callosum with intraventricular hemorrhage (IVH) is very rarely seen and reported. We present an unusual case of corpus callosum hemorrhage with IVH associated with a reversible cerebral vasoconstriction pattern (RCVS) on cerebral angiography.

The demographic information and clinical reports were obtained from electronic medical records retrospectively. Select neuroimaging was obtained from neuroradiology department.

year old Caucasian male with a past medical history of chronic obstructive pulmonary disease, essential hypertension, and prior ischemic stroke with residual right hemiparesis presented in unresponsive state when he was discovered on bathroom floor. Neurological examination on admission showed no verbal response, eyes open, with reactive pupils, and withdrawal to pain in left arm and leg. Blood pressure on admission was 224/92 mmHg. Computer tomography (CT) of head showed large ICH in rostrum, genu and trunk of corpus callosum with intraventricular extension and hydrocephalus. He was intubated for respiratory distress and external ventricular drain (EVD) was placed. He was also treated with intraventricular alteplase 1 mg injection for total of 9 doses, 8 hours apart. Blood pressure was controlled with nicardipine infusion initially, a up CT head showed resolution of IVH over the next several days, however, no significant clinical improvement was seen. Patient remained abulic and akinetic. Cerebral angiography performed showed right pericallosal artery beading pattern consistent with RCVS. After transition to comfort care, the patient expired on the 11th day of hospitalization.

Spontaneous non-traumatic corpus callosum ICHs are rare, and while other causes have been reported, this particular etiology is likely due to RCVS.

Intracerebral hemorrhage (ICH) is a leading cause of disability and mortality. Infections are a common complication observed in ICH and might be associated with worse outcomes. We aim to evaluate the association between infections and clinical outcomes at hospital discharge.

A retrospective chart review of consecutive patients with ICH admitted to the Baylor St. Luke's Medical center between January 2008 and December 2012 was conducted. Subject demographics, admission status, rates of infections including; pneumonia, urinary tract infection (UTI), bacteremia and clinical outcome at discharge were recorded. Associations between the presence or absence of infections and clinical outcome, measured by the Glasgow Outcome Scale (GOS), were investigated using a multivariate logistic regression model.

604 patients with ICH were included. Infections occurred in 141 subjects (23.5%). UTI was the most common infection (45.2 %) followed by pneumonia (28.8 %) and bacteremia (19.2 %). Clinical outcome was significantly worse for subjects who experienced any type of infection during hospitalization, compared to non-infected subjects, poor outcome (GOS <4) was found on 63.8 % and 42.8 % respectively (p=0.0168). This increased risk was significant after controlling for gender, ethnicity, ermore, an unfavorable discharge disposition -infected group, 62.4 % and 42.2 % respectively (p=0.004).

Our study shows a significant association between infections and poor clinical outcomes at hospital

Intracerebral hemorrhage (ICH) is a subtype of stroke associated with a high morbidity and mortality. Low serum calcium levels have been previously associated with larger hematoma volumes, hematoma expansion and worse outcomes; however, the pathophysiological mechanisms are still not well understood. A confounding effect among serum calcium and magnesium levels has been previously considered. In the present study, we investigate whether hypocalcemia is associated with poor clinical outcomes controlling for serum magnesium levels.

A retrospective chart review of consecutive patients with ICH admitted to the Baylor St. Luke's Medical center between January 2008 and December 2012 was conducted. Serum calcium and magnesium levels were measured during hospitalization, hypocalcemia and hypomagnesemia were defined as serum levels below 8.4 mg/dL and 1.5 m/dL respectively. Associations between serum calcium level and clinical outcome, measured by the Glasgow Outcome Scale (GOS), were investigated using a multivariate logistic regression model.

510 patients with ICH were included. Hypocalcemia was identified in 82 subjects (16.1%). Clinical outcome was significantly worse in the hypocalcemic group compared to the normocalcemic group, poor outcome (GOS <4) was found in 68.9% and 43.3% respectively (p=0.0035). This increased risk was significant after controlling for gender, ethnicity, serum magnesium levels, admission GCS, admission death) was also higher in the hypocalcemic group compared to the normocalcemic group, 69.6% and 41.8 % respectively (p=0.0016).

Our study shows a significant association between hypocalcemia and a poor clinical outcome after association.

Treatment of patients with intracerebral hemorrhage (ICH) typically requires advanced care at a tertiary medical center. Many patients present initially to regional or local emergency departments and require interfacility transportation to a referral center. Mission Hospital (MH) is a community-based nonacademic 763-bed tertiary care facility with Comprehensive Stroke Center certification. We serve as the referral center for 5 affiliated Mission Health hospitals and 11 regional non-affiliated hospitals across 18 counties. These hospitals are distributed throughout a mountainous and rural area with challenging terrain for transportation and limited resources for critical care transport. Here, we aim to describe the current transfer paradigm and consistency of care provided during interfacility transport of ICH patients prior to implementation of a dedicated ICH regional interfacility transfer protocol.

Retrospective review of the electronic medical record was performed to identify all patients in calendar year 2018 admitted to MH with a principal diagnosis of nontraumatic ICH who initially presented to another facility prior to transfer to MH. Data, including demographics, transport service type, and transport sequential blood pressures, were collected. Blood pressures during transport were analyzed to determine whether blood pressure exceeded our guidelines.

51 patients with ICH transferred to our referral center were identified. 10/51 (19.6%) were transported via critical care transport, and 41/51 (80.4%) were transported by local EMS using general adult transport protocols. 9/51 (17.6%) had uncontrolled hypertension as defined by 2 or more BP readings above our guidelines. Of these, 3/9 (33.3%) were transported via critical care transport and 6/9 were transported via local EMS. Transport records were incomplete in 15/51 (29.4%).

Elevated blood pressures during transport of ICH patients are common. Rural health systems are challenged by lack of critical care transport capabilities. We are currently implementing a dedicated protocol for interfacility transport care of ICH patients.

Infratentorial intracerebral hemorrhage (ICH) is associated with worse prognosis than supratentorial ICH; however, infratentorial ICH is often excluded or underrepresented in major studies of ICH. We sought to evaluate the natural history of infratentorial ICH stratified by brainstem or cerebellar location using a prospective observational study inclusive of all spontaneous ICH presenting to our institution.

Using a prospective, single center cohort of patients with spontaneous ICH between 2008-2018, we conducted a descriptive analysis of baseline demographics, severity of injury scores, and long-term functional outcomes of infratentorial ICH stratified by cerebellar or brainstem location.

Infratentorial ICH occurred in 82 (13%) of patients in our ICH cohort. Cerebellar ICH occurred in 45 (55%) and brainstem ICH occurred in 32 (45%). Compared to cerebellar ICH, brainstem ICH had significantly worse severity of injury scores, including: admission Glasgow Coma Scale (P <0.001), ICH score (P = 0.02), and National Institute of Health Stroke Scale (NIHSS) (P = 0.001). Modified Rankin Scale (mRS) scores at 3 months were significantly better in patients with cerebellar ICH compared to brainstem ICH (median 3 [1.5 -6.0] versus median 6 [5.0 -6.0], P = 0.04). Patients with cerebellar ICH were more likely to be discharged home or to acute rehabilitation (OR 4.7, 95% CI 1.6 -14.4) but there was no difference in in-hospital mortality (OR 0.39, 95% CI 0.13 -1.12) or cause of death (P = 0.5). Patients with cerebellar ICH who were alive at 3 months had smaller hemorrhages and lower severity of injury on admission.

Patients with cerebellar ICH have less severe symptoms at presentation and more favorable functional outcomes compared to patients with brainstem ICH.

It has been known that patients with intracerebral hemorrhage (ICH) have a higher rate of acute renal tality. The factors such as medications for blood pressure control, blood pressure (BP) variations and use of contrast for imaging without history of previous kidney disease.

We analyzed the records from hospitalized patients in the ICU from 2008 to 2012 in a single academic center with primary diagnosis of ICH and renal failure. A total of 604 were analyzed, 86 patients (14.23%) were reported to (47.6%) patients did not meet the criteria for renal risk, injury or failure and 23 (26.7%) did not have enough data for the study. Antihypertensive therapy used within the first 48 hours of admission was a combination of ACEI, ARBs and B-blockers.

Patients showed a wide variability in blood pressure (max-min within a day) which could not be and use of iodinated contrast, since CT without contrast was the imaging study of choice in all patients.

Our observations did not show an association in between BP variability, type of antihypertensive therapy or use of iodinated contrast within the first 48hrs of admission to acute renal failure in ICH patients either with or without history of renal disease. A larger study may be required to support this statement.

Milrinone, a phosphodiesterase inhibitor, has limited data as salvage therapy for cerebral vasospasm (CVS) secondary to aneurysmal subarachnoid hemorrhage (aSAH). To date, no study has compared patients treated with intravenous milrinone to a control group receiving standard treatment, primarily hemodynamic augmentation. We compared CVS duration in milrinone-treated patients to a control group, and evaluated additional safety and efficacy outcomes.

This was a retrospective, single center, case control study. Adult patients admitted to Spectrum Health or inclusion. The primary outcome was duration of CVS recorded on daily Transcranial Doppler exams. Secondary outcomes assessed efficacy and safety. Efficacy endpoints included, but were not limited to: incidence of ischemic stroke, interventions to treat CVS, ICU/hospital length of stay (LOS), and in-hospital mortality. Safety endpoints included vasopressor/inotrope requirements and incidence of arrhythmias.

-treated and 27 control patients. Milrinone use was associated with a longer duration of CVS (p = 0.048), increased use of intraventricular medications for CVS (p=0.003), greater vasopressor requirements (p =0.01), and longer vasopressor duration (p=0.0004). There was no difference in arrhythmias or in-hospital mortality. ICU LOS in milrinone versus control groups was 24.8 vs. 20.2 days (p=0.06) and hospital LOS was 25.7 vs. 21 days, respectively (p = 0.05). There were 6 ischemic strokes in the milrinone group versus 2 in the control group (p=0.11).

Intravenous milrinone was associated with a longer duration of CVS in aSAH patients, greater vasopressor requirements, and trended towards a higher incidence of ischemic stroke, though not statistically significant. Prospective, randomized, controlled trials are needed to further define the risks and benefits of milrinone therapy in aSAH patients.

Aneurysmal subarachnoid hemorrhage (aSAH) patients sustain several physiologic changes, including a rupture. MRI is potentially useful for prognostication in aSAH but has not been well-studied in this patient population. We present our preliminary experience with multimodal MRI in the acute period after aSAH. We hypothesized that changes in nodes of network critical to consciousness differ between patients with good and poor outcomes.

Thirty-four aSAH patients and 10 healthy volunteers underwent multimodal MRI at 3T. MRI T1 images were segmented, and ASLconsciousness (i.e., salience network, central executive network, default mode network). Wilcoxon rankto test odds of modified Rankin Scale (mRS) 0-3 at 6 months.

aSAH patients had a mean age (±SD) of 57.3±14.2 years, and controls were 26.1±5.0 years (p<0.001). prefrontal cortex -3 and 4-6).

r age-matched studies with more subjects and additional MRI sequences are needed to better determine MRI's potential utility in aSAH prognostication.

Aneurysmal subarachnoid hemorrhage (SAH) classically presents with the "worst headache of the patient's life" which can be very debilitating and persist for weeks. Headache is often refractory to standard treatment, including opiates. Pain is thought to be derived from meningeal irritation in the subarachnoid space. The sensory fibers in the anterior meninges are innervated by branches from the ophthalmic division of the trigeminal nerve, which is closely associated with the sphenopalatine ganglion (SPG). SPG blockade with local anesthetic, first described in 1908, has used as a treatment for various types of headache disorder but has not been described in SAH-associated headache. Treatment approach is either transnasal or transcutaneous injection. This case series describes five patients who received SPG blockade for intractable SAH-associated headache.

Patients with acute aneurysmal SAH in the Neurocritical Care Unit were offered adjunct SPG blockade for headache refractory to standard treatment. Patients rated pain on a 0-10 numerical scale, both before and 30 minutes after the procedure, which included either transnasal administration of ropivacaine using the TX360 device (Tian medical) or transcutaneous administration of ropivacaine with decadron.

ess score on admission 2 (range 2-3)); two (40%) received transnasal blockade and three (60%) received transcutaneous blockade. Median pre-treatment pain score was 8 (range 6their pain within 30 minutes; the fifth reported 30% reduction of pain. Transcutanous SPG blockade resulted in complete pain relief in all patients. The effects were transient, and pain typically returned within 24 hours. There were no complications associated with the procedure.

Repetitive SPG blockade is a safe and effective adjunct treatment for SAH-associated headache. A larger clinical trial is planned.

Tranexamic acid is recommended in the first 72 hours after subarachnoid haemorrhage (SAH) and before aneurysm treatment to reduce rebleeding. In Brazil, patients are frequently submitted to delayed aneurysm occlusion after SAH (>72 hours from ictus). The objective of this study was to evaluate the effects of tranexamic acid on hospital complications and outcome of patients with SAH.

All consecutive patients admitted with SAH between 2016 and 2018 at a reference center were included. Data were collected prospectively during the hospital stay. All SAH patients within 72 hours of ictus were considered eligible for tranexamic acid (TA) up to aneurysm occlusion. We analysed 3 groups: no TA, low dose TA and high dose TA. The primary endpoint was mortality at hospital discharge. Other outcomes included hospital complications such as rebleeding, delayed cerebral ischemia and adverse events such as deep venous thrombosis DVT) and pulmonary embolism (PE).

One hundred forty five patients were included in the study. Approximately half (53,1%) received TA, with 37(48%) receiving low dose and 40 (52%) high dose. At baseline, the high-dose TA group had more -dose group (38% vs 22%). Patients in the low-dose group had lower rebleeding rates (2.7%; P=0.05) than the no-TA and high-dose TA groups. Mortality was lower for the no-TA and low-dose TA groups as compared to the high-dose TA patients. Moreover, patients that did not receive TA had longer ICU and hospital lengths of stay. DVT/PE rates were very low in our cohort and not different between groups.

Our study showed that patients that received low dose of tranexamic acid had lower rates of rebleeding as compared to those that received no TA and high-dose TA. Mortality was also lower in this group when compared to patients that received high-dose TA.

Aneurysmal subarachnoid hemorrhage (aSAH) carries high mortality and morbidity. Symptomatic vasospasm is an important complication of aSAH. About thirty percent of patients with severe vasospasm do not respond to conventional management and will go on to develop delayed ischemic strokes. Medical management in these patients are limited and require endovascular therapy with intraarterial vasodilators and angioplasty. Milrinone has vasodilator properties and inotropic activity which has been used by intravenous and intraarterial routes for symptomatic vasospasm. In this study, we tested the safety and feasibility of intraventricular milrinone (IVM) in patients with severe vasospasm administered through the external ventricular drain (EVD).

A retrospective review of medical records of patients with subarachnoid hemorrhage who received IVM between 2016-2018. IVM was given at a dose of 0.87 mg in 2ml sterile saline every 8 hours through an EVD that was subsequently clamped for 2h.

28 patients received IVM for refractory vasospasm. Among those, 27 patients had ruptured aSAH and one patient had ruptured internal carotid artery pseudoaneurysm secondary to pituitary macroadenoma resection. The mean IVM doses were 13 (range 1-30 doses). Only one patient (3.5%) developed ventriculitis 26 days after IVM. There were no elevations of intracranial pressures with intraventricular administration of IVM.

In patients with refractory vasospasm from aneurysmal subarachnoid hemorrhage, intraventricular milrinone administration seemed to be relatively safe. Prospective trials are needed to further determine the safety and efficacy.

Rupture of cerebral aneurysm is the most common cause of subarachnoid hemorrhage (SAH). Hypertension is a particularly important risk factor for growing and rupture of cerebral aneurysm. In clinical practice, the non-adherence to anti-hypertensive medications is the most important cause of uncontrolled blood pressure. The aim of this study is to evaluate the effect of non-adherence to antihypertensive medications on the long-term prognosis of patients with hypertension and ruptured cerebral aneurysm based on the nationwide health claims database in Korea.

This study is retrospective cohort study using the National Health Insurance Service-National Sample Cohort (NHIS-NSC) in Korea. We included non-traumatic SAH patients (ICD-10; I60) with hypertension who underwent endovascular coil embolization or surgical clipping for ruptured aneurysm. The primary outcome is defined as composites of recurrent stroke, myocardial infarction, all-cause death. Adherence to anti-hypertensive medications is measured by calculating the proportion of days covered (PDC) based on the prescription records, which is treated as a time-dependent variable. We performed multivariate time-dependent Cox regression analysis with adjustments for sex, age, diabetes mellitus, treatment morality (coil embolization or surgical clipping), and household income.

-NSC, we found 880 patients who received coil embolization or surgical clipping for aneurysmal SAH. Among them, 297 patients with hypertension were included for analysis. During the 3.4 years of mean follow-up period, there were 116 patients who had primary outcome. In the multivariate Cox regression, poor adherence to antiindependently associated with increased risk of primary outcome (adjusted HR 1.87, 95% CI 1.17-2.97, p-value=0.009).

In this cohort study with real-world data, poor adherence to anti-hypertensive medications is a strong risk factor for worse prognosis in the hypertensive patients who underwent treatments for ruptured aneurysm. There is need for greater attention to adherence to anti-hypertensive medications in the high-risk patients.

TCD is routinely used in aneurysmal subarachnoid hemorrhage (SAH) for vasospasm surveillance. The value of TCD monitoring in non-aneurysmal SAH (NASAH) is unclear. In this study we sought to determine the clinical utility of performing TCD monitoring in a cohort of patients with NASAH.

Retrospective case series study performed at a comprehensive stroke center in a University hospital. Patients with SAH in whom an aneurysm or other vascular lesion was not identified were extracted from a prospective database covering a 5 year period. Patients with NASAH were categorized into perimesencephalic and diffuse SAH based on the CT appearance. Baseline demographics and clinical variables were obtained from the database. TCD results were obtained from a TCD database and conventional criteria were used to diagnose sonographic spasm. Categorical variables were compared

A total of 50 NASAH patients were identified; 32 perimesencephalic and 18 diffuse. Spasm was identified in 9/30 (30%) perimesencephalic NASAH patients and 6/18 (33%) diffuse NASAH patients (p=0.7542). No differences were observed between groups in age (p= 0.896 ), discharge disposition (p= 0.4707), median her score (p= 0.4119) when comparing patients with spasm to those without spasm. Similarly the median number of TCDs (p= 0.4119) did not differ among patients with and without spasm. The location of NASAH did not influence the diagnosis of spasm (p=0.7544).

Sonographic spasm occurs in 30% of NASAH patients but no specific clinical variable appears to influence its occurrence. The clinical significance of such finding needs further validation.

Complications following aneurysmal subarachnoid hemorrhage (aSAH) may be associated with early fluid status. This study aims to assess the relation of fluid balance and intravascular volume to outcomes including acute kidney injury (AKI), delayed cerebral ischemia (DCI), and vasospasm (VSP) in aSAH.

64 consecutive aSAH patients were retrospectively collected including patient demographics and admission characteristics. Intravascular volume on admission was measured by IVC ultrasound. Daily fluid balance in the first 14 days of admission were recorded along with changes in BUN and Cr. Outcomes including DCI and VSP were collected. Spaghetti plots were used to illustrate trajectory patterns. A linear mixed effect model was used the test the trajectory of slopes. An interaction term between time and patient condition was used to test the slope difference between patient conditions.

59 of 64 patients underwent IVC ultrasound assessment of intravascular volume. 21 patients were hypovolemic on admission with IVC collapsibility index >50% or distensibility index >18%. IVC slopes were found to be different by patient m balance decreased by -64.6±28mL/hr (p=0.02) while it increased 26.8±21.7mL/hr (p=0.22) in those -54.4±27.1/hr (p=0.05) while it increased 19.2±22.5/hr (p=0.39) in those without DCI (interaction p=0.04).

-87.5±27.8/hr (p<0.01) in those without VSP (interaction p<0.01).

Patient hemodynamics on admission as determined by IVC ultrasound does not correlate with development of AKI. However, fluid balance in the first 14 days of admission may be associated with outcomes in aSAH.

Early prediction of delayed cerebral ischemia (DCI) will improve management of subarachnoid hemorrhage (SAH) patients. We used mass spectroscopy (MS) to undertake an unbiased interrogation of plasma proteins associated with DCI.

This is an observational prospective single-center study of patients admitted to a tertiary care center. Serum samples from patients were obtained within 24 hours post-admission. We performed analysis in 2 cohorts separately at different times. The first cohort was a retrospective cohort of 24 matched subjects (13 no-DCI vs 14 DCI). The second cohort consisted of 45 matched subjects (24 no-DCI and 21 DCI). In both cohorts subjects were matched across DCI status for age, sex and modified fisher scale. We performed t-tests across DCI groups in both cohorts to identify proteins with a difference in concentrations between DCI groups. We selected proteins with a p-value of <0.2 for difference across DCI in both cohorts as potential candidates.

2017 and 2228 proteins were identified in cohort-1 and cohort-2 respectively. We identified 23 potential candidates in cohort-1, and 225 potential candidates in cohort-2. Six proteins were identified in both cohort-1 and cohort-2 (p-value cohort-1 and p-value cohort-2): complement factor H (p=0.02 and p=0.003); complement factor I (p=0.09 and p=0.021), antithrombin-III (p=0.19 and p=0.032), histidinerich glycoprotein (HRG) (p=0.03 and p=0.09), fetuin-B (p=0.13 and p=0.13), and hemopexin (p=0.03 and p=0.19). All plasma protein levels were lower in the DCI group.

In our unbiased approach to identifying biomarkers of DCI we identified 6 potential candidates. The compliment cascade and antithrombin-III has previously been identified as important in the pathophysiology of SAH. Of interest, we also identified hemopexin (part of the CD91-heme-hemopexin scavenging system) and HRG which is associated with cerebral vessel contraction as potential -B has not been previously reported in SAH. Confirmatory testing needs to be performed to validate our findings.

Glycemic gap (GG), determined by the difference between glucose and the HbA1c-derived average glucose (ADAG), predicts poor outcomes in various clinical settings. Our main objective was to evaluate various admission factors and outcomes in relation to GG.

We retrospectively reviewed prospectively collected data on 121 adult patients with aneurysmal subarachnoid hemorrhage. Admission glycemic gap (AGG) was defined as ADAG (28.7×HbA1c-46.7) subtracted from admission glucose (AG). Poor composite outcome was defined as death, tracheostomy, gastrostomy, and/or discharge to a nursing facility. Spearman method was used for correlation. Generalized linear model was used to test the difference in GG between patient categories. Mixed effects model was used to test the difference in trajectory slopes in GG. Area under the curve (AUC) for ROC curve was used to estimate prediction accuracy. SAS 9.4 was used for all data analyses.

The overall mean AGG was 58.9±51.8 mg/dL. AGG was significantly correlated with AG (r=0.91, p<0.01), GCS (r= -0.51, p<0.01), lactic acid (r=0.40, p<0.01), and procalcitonin (r=0.40, p<0.01) on admission, but not with HbA1c (r=0.04, p=0.65). There was a nonsignificant trend of higher AGG in those with delayed cerebral ischemia (73.1±8.7 vs. 53.2±5.5, p=0.056). Patients with poor composite outcome had both higher AG (194.4±7.7 vs. 167.8±7.5, p=0.02) and AGG (77.9±6.3 vs. 40.9±6.2, p<0.01 ), but the difference in AGG was more profound. Trajectory slope in the first 24 hours for GG did not differ in patients with poor vs. good composite outcome (-1.56±0.53/6hr vs. -1.47±0.52/6hr, p=0.9), nor did it differ for pointof-care glucose testing (-1.32±0.52/6hr vs. -1.27±0.51/6hr, p=0.95). AGG had significantly better prediction accuracy than AG in predicting poor composite outcome (AUC: 0.69±0.05 vs. 0.64±0.05, p=0.02).

Admission glycemic gap served as a better predictor of poor outcome than admission glucose. Additionally, AGG was correlated with AG, lactic acid, and procalcitonin, and inversely correlated with GCS.

The use of standardized management protocols (SMPs) has been shown to improve patient outcomes for multiple neurocritical diseases. However, whether SMPs improve outcomes after subarachnoid hemorrhage (SAH) is currently unknown. We aimed to study the effect of SMPs on 6-month mortality and neurologic outcomes following SAH.

A systematic review of randomized control trials (RCTs) and observational studies was performed by searching multiple indexing databases from their inception through January 2019. Studies were limited -traumatic SAH reporting mortality, neurologic outcomes, and delayed cerebral ischemia (DCI). Data on patient and SMP characteristics, outcomes, and methodologic quality was extracted into a data collection form. Methodologic quality of observational studies was assessed using the Newcastle Ottawa Scale (NOS).

A total of 11,260 studies were identified; 484 were assessed in full and 36 met the criteria for inclusion. Two studies were RCTs and 34 were observational. SMPs were divided into four broad domains: management of acute SAH, early brain injury, DCI, and general neurocritical care. The most common SMP design was control of DCI, with 21 studies targeting this domain. Overall, studies were of low quality; most described single-centre case series with small patient sizes. Observational studies scored between 3 and 8 on the 9-point NOS. DCI and neurologic outcomes were defined inconsistently in the literature, leading to significant challenges in their interpretation. Given the substantial hetereogeneity in reporting practices between studies, a meta-analysis could not be performed.

The effect of SMPs on SAH remains unknown due to major limitations in study design and quality. Notable deficiencies relate to heterogeneous definitions of DCI and inconsistent application of standardized neurologic assessment scales. Our study highlights the need for rigorous RCTs to determine whether the use of a protocol impacts outcomes in critically ill patients with SAH.

Elevated serum chloride has been associated with increased inflammatory markers, worsened systemic hypotension, and renal injury. Little is known regarding the effects of hyperchloremia on neurological outcomes after subarachnoid hemorrhage (SAH).

We reviewed prospectively collected data on adult patients who were admitted for spontaneous SAH from 2012 to 2016. Chloride values were examined on days 1-5. Hyperchloremia was defined as serum chloride of 110 mEq/L or greater. The primary outcome was delayed cerebral ischemia (DCI). Secondary outcomes included hospital mortality and 3 month modified Rankin scores (mRS). Chi-square test and two sample t-test were employed to assess DCI and 3 month mR analyze hospital mortality.

217 SAH patients were included in the analysis, 61 (28%) developed DCI and 156 (72%) did not. Patients with DCI had higher rates of hyperchloremia on day 3 (71% vs. 49%, p=0.005), day 4 (59% vs. 40%, p=0.01), and day 5 (51% vs. 25%, p<0.001) than patients without DCI. After controlling for age, Hunt and 2.16, p=0.03) and day 5 (OR 2.40, p=0.01) were associated with higher likelihood of experiencing DCI. Good functional outcome (mRS 0-2) was seen in 119 of 158 patients (75%) at 3 months. Rates of hyperchloremia were significantly lower in the good outcome group at all time points. After multivariate analysis, hyperchloremia on day 2 (OR 0.24, p=0.006), day 3 (OR 0.07, p<0.001), day 4 (OR 0.2, p=0.0030, and day 5 (OR 0.15, p<0.001) were independently associated with decreased odds of good functional outcome at 3 months.

Early hyperchloremia was associated with DCI and worse functional outcomes from SAH. The impact of chloride load and fluid management strategy on SAH outcomes warrants further investigation.

Headache is the most common complaint of patients presenting with aneurysmal subarachnoid an efficacious adjuvant therapy in the management of SAH-induced headache.

We performed a retrospective chart review of patients treated for SAH in the neurocritical care unit at a eceived steroids. Dexamethasone (4 mg every 6 hours) is typically administered for 1-2 days in patients with headache refractory to acetaminophen and oxycodone. Nursing documented numeric (0-10) pain scores were collected every two hours. We used paired t-tests to compare mean, maximum, and minimum daily pain scores on the day before and during steroid administration. We used multivariate analysis to assess for factors associated with steroid responsiveness, defined as an improvement of 1 or more points in mean daily pain score.

There were 51 steroid treatment periods among 40 patients (75% female, mean age 54 ± 11.8, median Hunt--two (43%) were classified as steroid responsive. Mean daily pain scores decreased by 0.9 points (p = .01) during steroid administration. Responders reported higher pre-treatment pain scores (5.6 vs 3.5, p = .009) and demonstrated greater decrease in mean pain scores (3.1 vs -.9 points, p < .001). There was no decrease in mean pain scores during the two days following therapy. In multivariate analysis, there was a weak signal that patients who underwent surgical clipping were more likely to have steroid responsive headaches (OR 16.8, . No other demographic or clinical characteristics were associated with steroid responsiveness.

A subset of patients with SAH induced headache may have a favorable, transient response to steroids. tterns and influence on opioid requirements.

Cerebral vessel vasospasm (CVV) is a feared complication following aneurysmal subarachnoid hemorrhage (aSAH). There has been an association between CVV and delayed cerebral ischemia which accounts for a great deal of morbidity and mortality following aSAH. Though the majority of patients with CVV respond to blood pressure augmentation, many patients go on to develop delayed ischemic neurologic deficits despite aggressive therapy. There is some suggestion in the literature that intraventricular milrinone (IVM) may be useful in the treatment of CVV.

Retrospective case series of 28 patients with aSAH that were treated with one or more doses of 0.87mg index (PI) and frequency of intraventricular milrinone dosing was collected. All patients were treated at Cleveland Clinic in the Neurologic Intensive Care Unit between 2016 and 2018. Paired t-test analysis was

Patients in our cohort were dosed with IVM between 1 and 30 times. There were no significant differences territory. There was also no effect of IVM on CVV over time. There were no direct complications secondary to IVM in these patients.

Based on our results, IVM was non-therapeutic for the treatment of CVV in patients with aSAH. Our data be conducted to evaluate the safety and efficacy of this treatment. Our retrospective analysis suggests that the use of intraventricular milrinone may be non-therapeutic for the treatment of CVV.

clinical and research tool for riskelement by the National Institute of Neurological Disorders and Stroke SAH Working Group. There are few data assessing the

We distributed a survey to a convenience sample of attending physicians that care for patients with questions regarding the definitions of the scale components (thin vs. thick, intraventricular blood vs no to determine the overall inter-ing.

Thirty-three respondents (100% neurocritical care fellowship trained, 79% UCNS certified in neurocritical care, 84.8% neurologists, median 4 years (IQR 2-6) in practice, treating median of 75 patients (IQR 50-100) with SAH annually from 19 institutions) completed the survey. Twenty-three (69.7%) reported r measurement of thin vs. thick blood, and 42.4% correctly identified that blood in any ventricle is scored -0.072) for the 15 CT scans, which is considered poor agreement.

Agr regarding the definitions of the score components. The National Institute of Neurological Disorders and Stroke SAH Common Data Elements may require further clarification in order to standardize research in

Cerebral vasospasm leading to delayed cerebral ischemia (DCI) is one of the most significant factors impacting functional outcome following subarachnoid hemorrhage (SAH). Although vasospasm is prevalent in this population, treatment options are limited. In recent years, several published case series have reported a positive effect of intrathecal (IT) Nicardipine for the treatment of vasospasm. We now report a single center one year retrospective cohort experience with intrathecal (IT) Nicardipine for the treatment of cerebral vasospasm following SAH.

All patients discharged in 2017 with a diagnosis of non-traumatic SAH, either aneurysmal or idiopathic, were included in the analysis. Demographics, risk factors, clinical course, radiological DCI and functional outcome were analyzed.

During 2017, 199 patients were admitted with aneurysmal (n=146) or idiopathic (n=53) SAH. The mean age was 55.5±13.9 and 64.8% were women. Low grade hemorrhage (H&H1-2) was found in 39.7%, medium (H&H3) in 37.7% and high grade (H&H4-5) in 22.6%. Cerebral vasospasm was diagnosed in 41.2% of the patients, and IT nicardipine was used in 83% of these patients (n=68). Only 8.5% of the patients required angiography to treat vasospasm. TCD data was available for 52 patients who received IT Nicardipine. Treatment reduced mean velocities in all arteries within one day (reduction of 6.6-15.6%, p<0.01). This effect remained through the treatment, until the vasospasm resolved. One patient suffered from bacterial ventriculitis. The overall rate of radiological DCI, as found in a blinded post treatment assessment of patients' imaging, was 10.6%. In this cohort, 50.8% had a favorable functional

IT Nicardipine is a safe and potentially effective treatment for cerebral vasospasm and prevention of the subsequent ischemic changes. We are currently expanding the analysis to prior years, however, future prospective controlled trials are still needed to evaluate the safety and efficacy of this treatment.

Patients remain at high-risk for vasospasm, delayed cerebral ischemia (DCI), and hydrocephalus after diversion is often necessary in ma additional benefit over standard management by facilitating intracranial blood clearance and decreasing rate of vasospasm and DCI, albeit with a possible increased risk of shunt dependency in historical studies. In this study, we assessed safety outcomes among patients who underwent this procedure.

Retrospective review of outcomes in pa cisternal drain placement at a single institution.

Between 2013drain placement. The median Hunt-Hess score was 2, but the study population skewed towards large drain dwell duration was 7.4 days. Radiographic vasospasm occurred in all but one patient (97.5%) and developed meningitis/ventriculitis, none fatal. The mean length of stay in the ICU was 16.3 days. Sixteen patients (42.1%) were discharged home, twenty-one to acute rehab (55.3%), one to subacute rehab (2.6%), and two died (5%). Among survivors, shunt-dependency occurred in 5/38 (13.2%), compared to the 5.3%-33.9% range reported in prior literature.

In the study population, cisternal drains appear to be safe as measured against historical cohorts, with comparable or lower shunt-dependency rates. This suggests the viability of further prospective studies to determine the appropriate population for and role of cisternal drainage in the management of aSAH.

Estimates of seizure onset after aneurysmal subarachnoid hemorrhage (aSAH) vary widely, reported rates range from 1% to 28%. Moreover, seizures increase mortality and disability in patients with aSAH regardless of common aSAH complications such as: rebleeding, delayed cerebral injury and vasospasm. We sought to establish the frequency of seizures in aSAH patients, along with their impact over prognosis, during hospitalization and upon discharge.

A retrospective review of consecutive patients with aSAH admitted to the Baylor St. Luke's Medical center between January 2008 and December 2012 was conducted. Subject demographics, admission GCS, admission mRS, incidence of clinical seizures and clinical outcome at discharge were recorded. Associations between the presence or absence of clinical seizures and outcome, measured by the Glasgow Outcome scale (GOS), were investigated using a multivariate logistic regression model.

226 patients with aSAH were included. Clinical seizures were identified in 7 subjects (3.1 %). Outcome was significantly worse for subjects who experienced a clinical seizure compared to subjects who remained seizure-free during hospitalization, poor outcome (GOS<4) was found on 71.4 % and 31 % respectively (OR 6.01, ]; p=0.04) This increased risk was significant after controlling hospice, death) was more common for the seizure group compared to the seizure-free group, 57.1 % vs 27.2 % respectively, however this difference did not reach significance (OR 3.89, ]; p=0.11)

Our results showed a low frequency of clinical seizures (3.1%) after aSAH, when compared to other series that have identified an increased incidence of seizures through multimodal approaches.

as indexed by GOS, along with a non-significant trend towards an unfavorable discharge disposition, among patients with seizures.

Vasospasm and delayed cerebral ischemia (DCI) account for 30% of the morbidity and mortality after aneurysmal subarachnoid hemorrhage (aSAH). Perfusion CT has been shown to be useful in identifying vasospasm, but this technique is less sensitive to microvascular perfusion changes. MR perfusion (MRP) has been increasingly used in the acute ischemic stroke population and avoids ionizing radiation. We hypothesized MRP may predict the presence of vasospasm by providing measures of impaired cerebral perfusion.

We performed a retrospective cohort study with consecutive aSAH patients between December 2012 and August 2015. Patients who underwent MRP for concern of DCI followed by digital subtraction angiography (DSA) within 24 hours were included. Quantitative volumetric analysis was performed at several thresholds of cerebral for the presence of a Tmax>6 lesion. Exact Wilcoxon Rank Sums test was used to compare perfusion volumes between patients treated endovascularly versus not treated for vasospasm.

We identified 22 patients with a total of 26 MRI studies meeting inclusion criteria (17 patients treated, 9 patients not treated). No Tmax>6s hypoperfusion lesion was identified in the untreated group, while 10/17 (59%) of the treated patients had at least some delay of Tmax>6s (p=0.009). Performance of MRP to detect vasospasm was sensitivity 0.59 (95%CI 0.33-0.81), specificity 1.0 (95%CI 0.63-1.0), PPV 1.0 (0.66-1.0), NPV 0.56 (95%CI 0.31requiring treatment for vasospasm.

Significant perfusion delay by Tmax >6s is present in patients requiring endovascular vasospasm treatment after aSAH. These results suggest that MRP may be a useful tool for patient triage for vasospasm therapy, and further studies are indicated for comparison to other screening methods for vasospasm.

Recent studies have suggested inflammation and immune dysregulation are important pathophysiology in aneurysmal subarachnoid hemorrhage (aSAH), and neutrophil to lymphocyte ratio (NLR) was considered as significant clinical predictor of unfavorable outcome including delayed cerebral ischemia (DCI). We analyzed NLR of aSAH patients during TTM, and proposed that the changes in NLR may reflect therapeutic effect of TTM in aSAH.

This retrospective single-center study included 50 aSAH patients from November 21, 2013 to May 30, 2018, among which 25 patients underwent TTM after surgical procedures, and other 25 patients didn't undergo TTM. Target temperatures were 33.5°C to 35°C and the durations of TTM were 2 to 10 days. We reviewed the changes of NRL of each patient during TTM and identified whether they had DCI, and analyzed in-hospital outcome and 3-month outcome as measured by the modified Rankin Scale (mRS).

There was no statistically significant difference of overall outcome between TTM group and non-TTM group, but TTM group showed slightly lower rate of DCI and better functional outcomes. Among the patients, patient who developed DCI had higher NLR, and the decreasing rate of NLR was higher in TTM group than non-TTM group.

Higher decreasing rate of NLR in aSAH patients while undergo TTM may show the therapeutic effect of TTM. Monitoring the trend of NLR value may be helpful in predicting the prognosis of aSAH patient and estimate the efficacy of TTM for individual patient. Eventually, NLR may play important role in deciding practice strategy while treating TTM in aSAH patients.

Hydrocephalus is generally regarded as a progressive or static process, but there are few reported cases of transient obstruction of the ventricular system. Here, the authors present a rare case of spontaneous symptomatic obstructive hydrocephalus that self-resolved. Additionally, a brief review of the literature is performed.

Records of the patient presented were reviewed in a retrospective manner for all relevant information. PubMed was then searched for all relevant articles.

Here, we discuss the case of a gentleman in his late 80's who presented with worsening confusion and lethargy in the setting of spontaneous subarachnoid and intraventricular hemorrhage. Pre-hospital medication includes a daily fish oil supplement; he takes no anticoagulation or antiplatelet agents. Approximately one year prior to admission, he experienced an episode of spontaneous left temporal intracerebral hemorrhage. This was attributed to amyloid angiopathy, as evidenced by multiple microbleeds observed on susceptibility weighted imaging at that time. The patient's neurocognitive status steadily declined admission, and he eventually became obtunded. In the process of transferring the patient to the intensive care unit for intubation and external ventricular drain placement, he suddenly became more awake and interactive. The patient's clinical symptoms completely resolved within 2-3 hours. No surgical intervention was undertaken. Repeat Head CT demonstrated that blood products seen in the third ventricle on previous imaging had now migrated into the fourth ventricle. Lateral ventricular size had decreased from the prior scan. The following morning, his family members commented that the patient was back to his baseline.

Transient episodes of obstructive hydrocephalus have rarely been reported in the literature, and are generally associated with an inciting event such as trauma or hemorrhagic stroke. It is possible that there is a higher incidence of transient hydrocephalus, but medical/surgical interventions are performed before the condition is permitted to resolve on its own.

Raised intracranial pressure (ICP) can be a dire consequence of extensive neurologic injury. Medical management of elevated ICP using intermittent doses of 23.4% hypertonic saline (HTS) and/or mannitol is relatively safe and effective for treating refractory intracranial hypertension. At our institution, prior to escalating to sedation and paralysis, HTS and mannitol are scheduled. In this study, we aim to describe our experience with scheduled 23.4% HTS.

Methods doses of 23.4% HTS during acute admission were included in our retrospective evaluation. Only patients who received scheduled 23.4% for anticipated or acute elevated ICP in the setting of high-grade subarachnoid hemorrhage (SAH) were included. The primary outcome was to characterize efficacy of sustained ICP control, by measuring frequency of ICP >25 mmHg and need for escalation of ICP management. Safety outcomes included incidence of hypernatremia (sodium > 160 mEq) and metabolic acidosis.

Seven out of 46 (15%) patients who received intermittent scheduled doses of 23.4% HTS were in the setting of highwere greater than 25 mmHg and no patients required escalation of ICP management for the duration of therapy. The median number of doses and duration of 23.4% HTS therapy were 11 doses and 7 days (IQR 5.9than 160 mEq and 2 patients (30%) developed metabolic acidosis in the setting of hyperchloremia.

Administration of scheduled intermittent 23.4% HTS in the setting of high-grade SAH is relatively safe and achieves sustained ICP control without need for escalation of ICP management. Comparative studies of scheduled intermittent 23.4% HTS vs alternative medical therapies for ICP management are warranted.

Use of contrast-enhanced computed tomography (CT) studies to evaluate neurological disorders have increased due to its non-invasiveness, fast image acquisition, easy accessibility, and minimal complications. One such procedure is CT myelogram that delineates the extent of spinal stenosis and helps in neurosurgical planning. However, it can result in intracranial migration of contrast medium leading to contrast-induced-encephalopathy (CIE).

We report 2 cases mimicking as subarachnoid hemorrhage after CT myelogram who subsequently developed CIE.

Case 1: A 68-year-old man with chronic low back pain (cLBP) was evaluated for confusion, headache due to "intracranial bleed". CT showed diffuse cerebral edema and hyperdensity in the subarachnoid space. External ventricular drain (EVD) was placed for suspected post-SAH hydrocephalus. However, CT and CT angiogram did not show any cerebrovascular malformation. The patient developed severe encephalopathy and left hemiparesis. Repeat CT head showed worsening cerebral edema and hours prior to presentation. Severe cerebral edema and left hemiparesis necessitated the use of dexamethasone with improvement in clinical symptoms and examination returning to near baseline. Case 2: A 64-year-old man with cLBP was admitted for "SAH" and associated cerebral edema with hydrocephalus. The initial presentation was confusion, double vision, and headache. CT showed diffuse cerebral edema with sulcal effacement, loss of basal cisterns and dilated lateral ventricles. CT and CT angiogram did not show a days prior to presentation. EVD placed for hydrocephalus was quickly weaned off the improvement of ventriculomegaly. The patient was discharged with complete resolution of symptoms.

CIE should be suspected in patients with encephalopathy after CT myelogram. Non-contrast CT head is to be interpreted in conjunction with clinical history to avoid unnecessary procedures that might further worsen CIE.

Seizures and ictal-interictal continuum (IIC) activity may impact recovery from acute brain injury (ABI). Empiric antiepileptic drug (AED) intensification for electrophysiologic activity of uncertain significance is challenging to evaluate given structural neurologic deficits, variable pharmacodynamics, and potential sedative effects. We analyzed the EEG and electronic medical records to identify electrographic biomarkers predicting clinical response to AED therapy.

We ascertained patients undergoing continuous electroencephalography (cEEG) during admission for ABI from a prospective big data repository of clinical data including regularly sampled Glasgow Coma Scale (GCS) scores and med -specific spectral power (alpha 8-13 Hz, theta 4-7 Hz, and delta 0.5-4 Hz) and graph theoretical metrics of EEG functional connectivity were compared at time intervals before and after AED therapy.

308 patients met inclusion criteria. 9,291 AED doses were administered (mean 1.64 +/-0.96 unique AEDs per patient). Initiating the first AED was followed by a 0.44-point average improvement in GCS (p=4.32 x10-7); initiating a second or third AED yielded no significant change, and adding a fourth, fifth, or sixth AED was followed by a 0.69-point worsening in GCS (p=0.044). Improvement in GCS 6 hours after AED administration was heralded by decline in EEG delta power and rise in network density in the hour following treatment. Decline in GCS was heralded by an early rise in delta power and decline in network density. Patients with the highest tertile of EEG improvement (greatest combination of rising EEG density and declining delta power) had a consistently improving GCS trajectory in the 48 hours following medication administration, whereas those in the lowest tertile had a consistently worsening GCS trajectory.

Empirically intensifying AED treatment for disorders of consciousness after ABI has diminishing benefit after the initial agent. Quantitative EEG biomarkers of early treatment response appears to robustly predict clinical response following AED treatment.

New-onset refractory status epilepticus (NORSE) describes patients with no seizure history who develop refractory status epilepticus (SE). The majority progress to super refractory status epilepticus (SE). We present a single-center case series of super refractory NORSE patients to highlight unique features of this group.

Retrospective chart review was performed to identify adults (age>18) admitted to the Columbia University Neurological ICU from 2/2009-2/2016 who required continuous midazolam infusions for treatment of super refractory NORSE. Outcome was defined as modified Rankin Score (mRS) at hospital discharge. Descriptive statistics were performed using Microsoft Excel.

Of the 27 cases, 96%(n=26) had a prodrome prior to seizures (infectious, psychiatric or both). Patient age was bimodally distributed with 56%(n=15) less than 30 years old and 37%(n=10) over 50. The most common comorbidity was an underlying autoimmune/rheumatologic condition (22%,n=6), though most patients had no pre-existing conditions (41%,n=11). The average STESS score was 3 (standard deviation 1.0). The majority (77%,n=21) remained cryptogenic despite extensive testing. Etiologies were identified in 22%(n=6) -4 with NMDA encephalitis and two with CNS infections. Immunomodulatory treatment included steroids in 59%(n=15, started on average 6 days from seizure onset, range 0-30), intravenous immunoglobulin in 41%(n=11, day 16, range 1-50) and plasmapheresis in 41%(n=11, day 10, range 3-13). The average ICU and hospital stays were 33 (range 2-150) and 55 (range 7-200) days, respectively. On discharge, 7%(n=2) had a good outcome (MRS 0-1), 19%(n=5) had fair outcome (MRS 2-3), 52%(n=14) had poor outcome (MRS 3-5) and 22%(n=6) died.

Compared to prior studies of all NORSE patients, our cohort with super refractory SE were younger, had more frequent prodrome, longer ICU and hospital stays and fewer identified autoimmune/paraneoplastic antibodies. The mortality rate was similar to prior studies, but among survivors, super refractory patients were less likely to have a good or fair outcome.

We aimed to assess the management of refractory status epilepticus (RSE) in developing (DING) and developed (DEV) economies, as the management of this condition is resource intense and poorly standardized.

Investigators from 4 continents collected a large cohort study of RSE patients treated between 01/2015-12/2018. Case-report-forms were finalized at the 2018 annual NCS meeting. RSE was defined as SE that failed to respond to a benzodiazepine and at least one non-anesthetic antiepileptic agent, and was managed with midazolam (MDZ) or propofol(PRO). The 2018 United Nations World-Economic-Situation-Prospect was used to identify sites as being from DEV or DING economies.

four from DEV (237 patients) economies were included. Patients from DEV economies were slightly sicker (STESS score 3.1±1.3 vs. 2.9±1.2, p<0.05). Management of patients from DEV economies more frequently involved prolonged EEG monitoring (continuous 77% vs. 23%, p<0.05) but MDZ (0.5±0.7 vs. 2.0±1.6 mg/kg/h) and PRO (37±24 vs. 75±52 mcg/kg/min, p<0.05) doses were higher in DING economies. Breakthrough seizures were more common in DING (36% vs. 53%, OR 2.1, P=0.003), but no difference in vasopressor use (64% vs. 57%; N.S.) or withdrawal seizures (22% vs. 31% N.S.) was seen. Hospital (32±31 vs. 54±83 days, P<0.001) and ICU stays (17±21 vs. 31±26 days, P<0.001) were longer for patients in DING economies. Modified Rankin Scale at discharge was associated with higher STESS scores (P=0.003) but did not differ between DING and DEV economies.

Direct comparisons between RSE patients managed in DING and DEV economies are challenging as the baseline level of illness differed but this dataset provides unique insights into differences in utilization of technology (i.e., EEG monitoring), medications (duration and dosage of anesthetics), and length of stay in different health care systems. Larger follow-up studies need to explore matched cohorts and explore differences between private-public hospital settings.

Unlike most anesthetics ketamine acts as an NMDA antagonist. We examine the efficacy of intravenous ketamine in the treatment of RSE in a large series.

Retrospective case series of 68 status epilepticus patients admitted between 2009 and 2018 who underwent treatment with ketamine, 4 patients underwent multimodality monitoring (MMM). We compared patients with complete seizure cessation after ketamine with those without using chi-square and 2 sample t-test.

Mean age was 53 +/-18 years old, 46% of patients were female. Seizure burden was decreased by 50% within 24 hours of starting ketamine in 55 patients (81%), with complete cessation in 43 (63%). Average rate of ketamine infusion was 2.2 +/-1.8 mg/kg/h, with duration of 3.6 +/-3.9 days. Average dose of midazolam was 1 +/-0.8 mg/kg/h. Ketamine was started on average 1 +/-3 day after midazolam. Patients without complete seizure control after initiation of ketamine (25/43 patients) were more commonly cardiac arrest patients 36% vs 21% (p=.17), and had lower STESS score 3 +/-1 vs 4 +/-1 (p=.03). All other characteristics were not statistically significant between the two groups including; age, gender, ketamine infusion dosages and duration, APACHE score, and midazolam infusion dosages. 8 patients (12%) were weaned off pressors after initiating ketamine infusion. When compared the MMM values 24h before and after ketamine initiation, intracranial pressure values (8 +/-4 vs 10 +/-5), cerebral perfusion pressures (73 +/-14 vs 72 +/-13), cerebral blood flow (24 +/-3 vs 21 +/-9), and Lactate/pyruvate ratio (45+/-21 vs 45 +/-12) were relatively stable. PbO2 values increased from 7 +/-1.5 to 17 +/-7.

In our cohort ketamine infusion had a meaningful decreased in seizure burden in RSE. Our preliminary data also suggests that Ketamine infusion didn't affect the intracranial pressure.

Continuous EEG (cEEG) is widely used to detect seizures (SZ) in patients with acute brain injury. However, studies examining SZ and epileptiform abnormalities (EA) using cEEG in acute ischemic stroke (AIS) are limited. Therefore, we aimed to describe the prevalence of electrographic patterns (SZ and EA) in AIS and its association with outcomes at discharge.

Retrospective chart review identified 94 patients with AIS who underwent cEEG between 1/2018 and 1/2019. Demographics, comorbidities and other relevant clinical factors including NIH stroke scale (NIHSS) and treatment interventions were abstracted. cEEG closest to admission (median 2 days) was reviewed for background, SZ and EA (lateralized and periodic discharges (LPDs and GPDs) lateralized rhythmic delta activity (LRDA) and sporadic epileptiform discharges (sEDs). Computed tomography or magnetic resonance imaging of brain closest to the time of cEEG was analyzed for midline shift, hemorrhagic transformation (HT) and cortical involvement. Outcomes measures were mortality and functional outcome in modified Rankin scale (mRS) (0-2 good and >2 poor outcome) at discharge.

Of the 94 patients, 4 had SZ and 31 had EA (18.1% LPD, 8.5% LRDA, 9.6% GPDs and 3.2% sEDs). Those with cortical involvement had higher rate of EA and Sz compared to those with subcortical stroke (85.3% vs 60.0%, p=0.011). No difference was found in SZ and EA prevalence with regards to age, sex, NIHSS, midline shift or HT. Overall mortality was 29.8%. Absence of posterior dominant rhythm (PDR) was associated with increased mortality (85.7% when PDR absent vs 56.1% when present, p=0.009). SZ and EA did not affect mortality or mRS at discharge.

Despite high frequency of EA (33%), the risk of SZ in AIS was low at 4.3 % and their presence did not impact functional outcome or mortality. However, EEG background with absence of PDR was associated with increased mortality.

Nonconvulsive seizures (NCS) are a common complication in patients admitted to Neuroscience Intensive Care Units and are associated with worse outcomes. NCS can only be diagnosed with continuous EEG (cEEG) monitoring. Intermittent conventional cEEG review by neurophysiologists typically occurs 2-3 times a day, therefore patients may be seizing for extended periods of time before the seizure is detected. Our study aims to evaluate the accuracy of a quantitative EEG (qEEG) trend, the Automated Seizure Detector (ASD) in detecting patients' first seizure, which could aid in rapid detection of NCS.

This retrospective study includes review of cEEG and qEEG data from adult patients admitted to a single institution Neuro ICU who developed NCS on cEEG monitoring. Independent conventional cEEG review without qEEG by two board-certified neurophysiologists determined the first seizure occurrence for each patient (gold standard). This was compared to the seizure detection sensitivity of the P12 ASD (Persyst, Inc., Prescott AZ), an algorithm with no user-adjustable settings.

Recordings from 45 NCS patients were used. Mean age was 61.2 years and 56% was female. Seizures had variable durations and spatial extents. The sensitivity of P12 ASD was 77.4% (95% CI 58.5-89.7) and specificity was 92.9% (95% CI 64.2-99.6). Mean false alarm rate was 0.17/hour (SD 0.46) in the time elapsed from the start of cEEG recording until first seizure occurrence. Overall, P12 ASD accurately detected the first seizure in 69% of patients, disregarding false positives. Overall, median time to clinical seizure detection was 2.0 hours (IQR 5.5 hours).

This analysis shows that the Persyst P12 ASD may have clinically useful sensitivity and specificity in critically ill patients admitted to a Neuroscience ICU. In conjunction with a low false alarm rate, incorporation of qEEG ASD may lead to a reduction in time for seizure recognition.

The incidence of early seizures (ES) in traumatic brain injury (TBI) ranges between 1-7%. However, the incidence of ES after a non-severe TBI (NSTBI) with traumatic hemorrhage (TH) is unknown. Moreover, the data about seizure prophylaxis (SP) in this population remains inconclusive. We aim to determine the incidence of ES in NSTBI and the efficacy of SP.

We respectively reviewed all adult patients with NSTBI with evidene of a TH on presentation from 2015 to 2018. Patients with history of epilepsy or receiving antiepileptic drugs (AED) were excluded. We collected demographic data, the type, severity and mechanism of injury; the need for neurosurgical intervention (NSI); ES; and SP use.

A total of 633 patients met our inclusion criteria, 94.4% had mild TBI; mean age of 70.5 years (SD 16.9); 55.3% males; and 49.1% had subdural hematoma (SDH). Same level fall was the most common 310 (7.1%) patients had an ES in the SP group (16 clinical) vs 5 of 310 (1.6%) in the non-prophylaxis group (all clinical) (P = 0.001). Levitiracetam as SP was used in 83.5%. Patients with combined SDH and traumatic subarachnoid hemorrhage or with multicompartment hemorrhage were more likely to have ES than SDH alone (p = 0.02 and 0.001, respectively). NSI was not a predictor for ES in our cohort.

The incidence of ES in NSTBI patients in our cohort falls within the previously reported ragne. however, it appears to be higher compared to reported rates for mild TBI. ES were more likely in the SP group, which might indicate a clinical selection bias. Prospective studies are required to further determine the predictors of ES and the effect of SP on outcomes in NSTBI patients.

Patients with psychogenic non-epileptic attacks (PNEA) sometimes receive aggressive treatment leading to intubation. This study aimed to identify patient characteristics that can help differentiate PNEA from true status epilepticus (SE).

We retrospectively identified patients with PNEA and SE who were intubated and underwent continuous had acute brain injury or progressive brain disease as a cause of status epilepticus were excluded. We compared clinical features, treatments and outcome between patients who were intubated for PNEA and those who were intubated for SE.

Of 1,735 patients who underwent cEEG monitoring, we identified 24 and 124 patients intubated for PNEA and SE, respectively. Compared with patients intubated for SE, intubated PNEA patients were more likely to (1) be <50 years of age (92% vs 40%, P<0.001), (2) be female (79% vs 39%, P<0.001), (3) be white (91% vs 29%, P<0.001), (4) have a history of a psychiatric disorder (88% vs 25%, P<0.001), (5) have no history of an intracranial abnormality (92% vs 38%, P<0.001), and (6) have a maximum systolic blood pressure <140 mm Hg (67% vs 16%, P<0.001). Patients with 0-2 of these risk factors had a 0% (0/88) likelihood of having PNEA, those with 3-4 had a 15% (6/39) chance of having PNEA, and those with 5-6 had an 86% (18/21) chance of having PNEA. Sensitivity for PNEA among those with 5-6 risk factors was 75% and specificity was 98%.

PNEA in patients presenting with emergent convulsive symptoms can be predicted with a high degree of certainty based on the presence of specific demographic, past medical, and physiologic risk factors. Care should be taken to avoid over-sedation and unnecessary intubation in this at-risk patient population.

A recent systematic review indicates that the mortality of status epilepticus (SE) is about 15.9% with a non significant downward trend in recent years. Mortality has not changed much despite aggressive management. This study investigates trends and predictors of in-hospital mortality due to Status Epilepticus at national level in United States.

We performed a cross-sectional analysis using the Nationwide Inpatient Sample (NIS), 2005-2014, of US adult hospitalizations with Status Epilepticus. Annual rate of in-hospital mortality was calculated using NIS weighting. We identified our Status Epilepticus patient subset from using codes (DX1 = 345.3) from the International Classification of Diseases, 9th edition. Potential factors associated with in-hospital mortality were assessed using logistic regression.

Of 147,548 hospitalized patients with Status Epilepticus, 5,271 (3.57%) died during the index hospitalization. Across 2005-2014, 3.57% of SE patients died; with a downward but not statistically significant trend in-hospital mortality from 4.28% (2005) to 3.73% (2014) (p = 0.14). SE patients with inhospital mortality were more likely to be women, older, and with a higher proportion of medical comorbidities, in-hospital complications and extreme loss of function as per All Patients Refined Diagnosis al failure, APR DRG severity, mechanical ventilation, tracheostomy, sepsis, pulmonary embolism, acute kidney injury and respiratory insufficiency.

Mortality due to SE was lower than previously reported. Mortality has had a non-significant downward trend in the years studied. Age, female gender, medical complications and poor baseline functional status are important predictors. Availability of aggressive treatment has not modified significantly mortality which requires further study.

Pregabalin (PGB) is an approved adjunctive treatment for focal epilepsy in adults. PGB lacks drug-drug interactions, has a favorable safety profile and can be rapidly titrated-attractive characteristics for its use in the neurocritically ill. However, data remain limited regarding its use in the ICU setting. We are sharing our experience with PGB in neurocritically ill patients with refractory seizures.

Charts of eight adult patients admitted received PGB were reviewed retrospectively. Demographics, antiseizure drug (ASD) regimen, and 24h of EEG data pre-and post-PGB were analyzed descriptively.

The cohort comprised eight patients (4 females) with mean age of 68.5 years. Mean ICU stay was 9.1 days. Three patients underwent a neurosurgical procedure related to their primary admission diagnosis, an ASD prior to first seizure captured on EEG. Prior to PGB, patients had failed on average 3 (2-4) other ASDs trials. PGB was dosed 300-400mg/day in 2-4 divided doses, following a load of 75-300mg. PGB lead to a significant reduction on hourly median seizure burden: 5.1 to 1 seizure/h and 5.3 to 0.97 min/h. PGB led to complete seizure cessation in 4 patients within 24h and in 6 out of 8 within 48h of administration. PGB allowed for de-escalation of ASD regimen in 5 out of 8 patients. PGB was well tolerated with the exception of mild sedation in 3 patients, which did not warrant further intervention/neurodiagnostics.

In this critically ill cohort with refractory seizures, PGB successfully aborted seizures in 75% of patients.

include prospective pregabalin treatment protocols.

To describe the first known reported case of utilization of Electroconvulsive therapy (ECT) to treat super refractory status epilepticus (SRSE) in pregnancy.

We present the case of a 21 year old Caucasian female at 8 weeks gestation with PMH focal and generalized seizures who was treated for SRSE successfully with ECT after failed pharmacological treatment. The most likely etiology of SRSE was sudden cessation of medications upon pregnancy. EEG showed 2 types of seizure activity: rhythmic theta waves over right temporal region with evolution and independent generalized seizures. Treatment included use of approximately 12 antiepileptics including , Propofol, Pentobarbital, Magnesium, Ketamine, Topiramate and Valproic acid over the course of 30 days in addition to modifying epilepticus remained super refractory with appearance of mixture of sharp waves on weaning off sedation. She underwent ECT with right unilateral electrode placement on day 31 with remarkable improvement in EEG pattern and resolution of SRSE with single session.

Patient was back to baseline level of awareness at the time of discharge. On follow up in clinic, she had significant improvement in seizure control with normal fetal development and delivery.

Treatment of status epilepticus in pregnancy is challenging given the unknown effect of prolonged sedation or hypothermia on fetal development. Alternative treatments like ECT, VNS, DBS, ketogenic diet and hypothermia are sporadically used. Use of ECT is not considered first or even second line treatment in SRSE, despite its safe profile, especially in pregnancy. This case adds to the available literature on the success of ECT for treatment of SRSE and puts emphasis on the need for a clinical trial regarding use of ECT in SRSE.

The importance of Neurocritical care (NCC) has been recognized. But no dedicated educational system for it exists in Japan. We have established version 3 of an educational NCC hands-on seminar. This study investigated its effects.

This study was a prospective, before-after study using questionnaires and examinations. It was a full-day version 3. The learning concept was to identify the various methods for maintaining cerebral oxygen balance to prevent secondary brain injury. Participants attended five skill sessions: intracranial pressure monitoring, trans-cranial color flow image, targeted temperature management, neuro examination, and EEG, and four scenario sessions: post-cardiac arrest syndrome, subarachnoid hemorrhage, traumatic brain injury, and non-traumatic acute weakness. They had examinations before and after the seminar. The primary outcome was the improvement on examination scores after the seminar. Secondary outcomes were the degrees of satisfaction with it and confidence of participants in NCC. We evaluated the improvement of the outcome using Wilcoxon signed rank test. A p-value of 0.05 or less was considered as significant.

Thirty-nine physicians and one nurse participated in the seminar. We excluded 19 (47.5%) participants because their answers were incomplete. We had 16 (76.2%) physicians who are in emergency or intensive care medicine, and 5 (23.8%) other professionals. Their median age group was in their 30s (IQR:20-30) with median intensive care medicine experience of 2 years (IQR:1.0-5.0). The percentage of correct answers, scores in the examination, improved significantly from 70 (IQR:47.0-87.0) to 80 (IQR:63.0-93.0) after the seminar (p<0.001). Eighteen (85.7%) participants were satisfied with it, and the number of professionals who could not feel NCC-confident decreased from 12 (50%) before the seminar to 0 after its completion (p<0.001).

Our seminar successfully improved the physicians' knowledge of NCC, and gave them more confidence in NCC.

Glutamic acid decarboxylase (GAD) is the rate-limiting enzyme to convert glutamate to gammaaminobutyric acid (GABA). Autoantibodies targeted against GAD65 have been implicated in a number of syndromes with neurologic manifestations including stiff-person syndrome, cerebellar ataxia, limbic encephalitis, and epilepsy. We highlight an atypical presentation of this rare disorder with several unique features to the neurological intensive care unit.

62-year-old woman with PMH of DM, remote left insular ischemic stroke, and recent right leg dystonia presented after being found down with rightward eye gaze deviation, GTC shaking, and urinary incontinence. She required midazolam, lorazepam, loading doses of levetiracetam and fosphenytoin, and propofol infusion to achieve clinical seizure control. Despite these interventions, EEG showed NCSE with left temporal seizures and anterior midline epileptiform discharges. Propofol was titrated to burst suppression. She had several other active medical problems including kidney injury, transaminitis, and myoclonus. Seizures and myoclonus were greatly improved after the addition of clonazepam; however, she remained encephalopathic. Pertinent diagnostic results included ferritin 16,130 ng/ml, LDH 2,070 units/L, IL-2R 795 U/ml, B2-micr and serum GAD65 Ab titer 332 nmol/L. MRI Brain showed prominent superior frontal lobe cortical edema. Bone marrow biopsy demonstrated good cellularity without malignancy. Skin biopsies on three random samples were positive for perivascular dermatitis with telangiectasia. She was started on high dose steroids with subsequent progressive mental status improvement.

Anti-GAD65 Ab associated vasculitis is an exceedingly rare occurrence whose diagnosis previously involved brain biopsy. This case is unique given her acute presentation with refractory status epilepticus, systemic involvement, and diagnosis on skin biopsy. While management has involved immunotherapy, specific treatment guidelines do not exist. Given her marked response to clonazepam and corticosteroids, we advocate for early initiation of GABAergic medications such as benzodiazepines and use of immunotherapy.

Epileptic seizures are a serious complication in patients with subdural hemorrhage (SDH), resulting in increased mortality rates. The incidence of new onset seizures in these patients is unclear. We examined the incidence for new onset seizures and status epilepticus (SE) in SDH patients.

We examined patients diagnosed with SDH and epilepsy between September 2015 to December 2018. We included patients with new onset seizures and extracted those who had seizures after SDH evacuation. Clinical and radiographic characteristics, and outcomes of those patients were described.

We screened 427 patients diagnosed with SDH, traumatic or non-traumatic. 248 underwent a surgical intervention and 23 (5%) patients had a seizure during their hospital stay. Among those who had a seizure, 9 patients had prior history of epilepsy, and 14 had a new onset seizure. Although SDH patients with history of epilepsy showed higher incidences of seizures than those with no history (p= 0.08), SDH patients with history of epilepsy mostly did not evolve into SE and those who had no history of epilepsy usually did. There was no significant difference in patients developing SE when compared to those without SE between the SDH thickness, midline shift, temporal lobe involvement or age of blood (acute or chronic).

Seizure occurrence in patients with SDH is commonly new onset; however, they are infrequent. In addition, SDH patients with no history of epilepsy have a higher tendency to develop SE as opposed to patients with history of epilepsy. Larger multicenter cohort studies need to be done for evaluation of these findings.

Sequoia Hospital in Redwood City, CA implemented the Ceribell Rapid Response EEG system in 2018 to expand its access to EEG for in-patient usage. Previously, the hospital had no access to after-hours EEG and the majority of their EEGs happened in the ICU. This quality improvement project was initiated to understand how access to rapid EEG impacted clinical care and financial metrics across at Sequoia Hospital.

Data was analyzed for all patients who received either conventional or Ceribell EEG from January 1, 2017

including the department where EEG was conducted, time of day of EEG was ordered, time when EEG began, and clinical diagnosis based on the EEG. Data was also captured on patient transfer due to lack of EEG.

55% of EEGs were ordered after hours after the introduction of Ceribell, compared to nearly no EEGs done after hours before Ceribell. 25% of patients with Ceribell EEGs were diagnosed with seizures. In 2018, of 66 Ceribell EEGs, 41 EEGs occurred in the in-patient unit or ED. In 50% of patients with a high suspicion of seizure, seizures were ruled out as a result of reading the Ceribell EEG.

The introduction of Ceribell EEGs has greatly expanded access to EEGs at Sequoia hospital. Before Ceribell was introduced, EEGs mostly occurred in the ICU and nearly all happened during regular hours. After Ceribell was introduced, EEG was also heavily utilized in the ED and the in-patient unit and gave Sequoia EEG access during after hours. As a result of this expanded access and earlier application of EEGs, patients have been treated more appropriately.

Tranexamic Acid (TXA) is an intravenous antifibrinolytic agent that is used routinely for elective surgery. We report a case of inadvertent intrathecal injection of TXA resulting in refractory status epilepticus.

Case report.

A 71-year-old healthy female admitted for bilateral total knee replacement was inadvertently administered 300mg of TXA intrathecally instead of bupivacaine. Soon after administration, she intubated, administered levetiracetam, started on a propofol infusion, and transferred to the neurointensive care unit (NICU). She developed persistent spontaneous and stimulus induced generalized myoclonus refractory to propofol. Midazolam infusion was added. NCHCT and CTA demonstrated pneumocephalus, but no acute arterial or venous thrombosis or stroke. vEEG revealed generalized nonconvulsive seizures occurring once per minute, not correlating with spinal myoclonus . Propofol and midazolam infusions were increased to 150 mcg/kg/min and 2.6 mg/kg/hr, respectively, to achieve burst suppression, and valproic acid was added. Over the following week, the drips were adjusted to suppress seizure activity. By hospital day 8, she was weaned off all infusions without recurrence of seizures. By hospital day 19, she was on levetiracetam monotherapy. She was discharged to rehab after a 22-day hospital course, and was discharged home 45 days after initial presentation. Residual deficits at the time of discharge included mild cognitive impairment and gait instability. She remains seizure-free since hospital day 45 on levetiracetam 500mg BID.

We report a case of refractory status epilepticus and spinal myoclonus after accidental intrathecal TXA administration. With aggressive management, the patient survived with mild residual deficits. The mechanism by which TXA causes status epilepticus and spinal myoclonus is hypothesized to be related to its inhibitory effects on GABA and glycine receptors, respectively.

Ictal bradycardia (IB) is a serious complication of temporal lobe epilepsy. If left untreated, IB can cause serious injuries related to syncope, complete heart block and death. Management of this phenomenon is controversial: should you treat the seizures or the arrhythmia? We describe the management of a patient who presented with multiple syncopal episodes and found to have symptomatic bradycardia in the setting of temporal lobe seizures.

A 22-year-old male with a recently resected brainstem cavernoma presented with episodes of 'spacing out', face tingling and transient periods of amnesia. He was started on Topamax and Lamictal. Several months later, he began having multiple syncopal events (upwards of 7 a day) that eventually brought him the hospital for evaluation. He was found to be bradycardic with a heartrate in the thirties and had sinus pauses lasting up to ten seconds requiring Atropine, an Isoproterenol infusion and transcutaneous (TC) pacing. He was also found to have another cavernoma in the right temporal lobe. EEG revealed epileptic activity within the right anterior temporal lobe with correlation to his TC pacing and IB events. Lamictal was replaced with Keppra and the seizure activity was controlled. He had a pacemaker implanted, after which he did not have any further episodes of syncope and no further seizure activity. The cavernoma was resected a few months later, and he did well postoperatively.

IB is an uncommon, but serious, complication of temporal lobe epilepsy. The temporal insula plays a role in the parasympathetic activity of the heart which can cause IB. It may be beneficial for patients who present with symptoms characteristic of temporal lobe seizures or repeated falls/drop attacks to have a full cardiac work up to rule out IB in order to determine if a pacemaker is warranted.

The cEEG has had rapid growth within neurological monitoring within the ICU, however its still disparate resource in the ICUs of Latin America. Is important to know the real situation in Colombia about the accessibility to cEEG monitoring.

An anonimus survey of 19 questions was conducted from October 2018 to April 2019. It was answered by Intensivists from Latin America, Europe, Asia and USA. (n=62) 

Considering the accessibility to the cEEG, the cEEG clinical indications and the cEEG monitoring extends (hours) in the ICU, we can conclude that Colombia is aligned with other countries in the world. In the ICUs of Colombia less than half of the intensivists make decisions in ¨real time¨ with the cEEG and have access to the QEEG modality. The most common cause for non-presciption of cEEG was Scarce resources (Equipment and human resorces support from a Neurology service).

Cefepime is a fourth-generation cephalosporin with broad-spectrum coverage used to treat infections in critically ill patients. Neurotoxic effects have been associated with cefepime, including myoclonus, reduced consciousness, and seizures. We report a case of a patient receiving cefepime who developed non-fluent aphasia and non-convulsive status epilepticus (NCSE). Two seizure drug trials (levetiracetam and fosphenytoin) failed before marked clinical and electrographic improvement with clobazam. Other than cessation of the offending agent, there is little known about the management of cephalosporin associated non-convulsive status epilepticus.

Data was collected from our institution's health record.

A 74-year-old female with a history of diabetes, chronic kidney disease, recent coronary artery bypass grafting, and mitral valve repair presented with Pseudomonas aeruginosa cellulitis of the sternotomy site. On day six of cefepime therapy she developed non-fluent aphasia. MRI brain and toxic-metabolic work-up was unrevealing. EEG was consistent with non-convulsive status epilepticus. She failed to respond to standard levetiracetam or fosphenytoin therapy. Lorazepam was given with marked improvement in her EEG. Clobazam was subsequently started resulting in marked improvement in the patient's language and sustained resolution of ictal pattern on EEG.

Epileptogenic effects of ß lactam antibiotics are thought to be due to competitive antagonism of the GABAA receptor. Beside the recommendation of withholding offending agents when safe to do so, there is no guidance in the literature regarding the appropriate antiepileptic drug choices for the treatment of cephalosporin associated NCSE. In this case, clobazam, a benzodiazepine, was an effective treatment. Given the theorized mechanism GABA antagonism of cefepime, it is possible that benzodiazepines may ch is needed regarding the optimal seizure control for various etiologies of NCSE. When treating seizures and NCSE, consideration should be given to the possible mechanism of action of the suspected offending agent.

Hashimoto encephalopathy is a rare disease. Clinical manifestations include abnormal behavior or psychosis, seizures, encephalopathy. Pathophysiology is not completely known but it has been associated with autoimmune thyroiditis. We report a case of hashimoto encephalopathy with status epilepticus which responded well to steroids and relapsed following steroid taper.

20-year-old previously healthy woman was admitted with encephalopathy, new-onset seizures, and delusional behavior for past 8-10 weeks. MRI brain was unremarkable. EEG showed status epilepticus with right fronto-central origin. She was treated with multiple antiepileptic medications including evaluation for infections, autoimmune and paraneoplastic etiologies revealed elevated thyroid peroxidase, antithyroglubulin and mildly elevated GAD 65 Antibodies. Whole body CT showed no malignancy. She was diagnosed with hashimoto encephopathy. She was treated with IV steroids and IVIG. Her clinical improvement correlated with decrease in thyroglobulin antibody levels from15.6 to 4.4 and thyroid peroxidase antibody levels from 32.40 to 17.70. She was discharged on oral steroids and admitted again in few weeks with a relapse of behavioral issues and seizures following steroid taper. She was treated with high dose IV steroids, this time followed by rituximab with significant improvement. She was discharged again on oral steroids with very slow taper and close follow up.

Our patient had hashimoto encephalopathy and had relapse following taper of steroids.

Hashimoto encephalopathy is rare condition and is often under-diagnosed. Anti-thyroglobulin and thyroid peroxidase antibodies should be checked in patients where no other etiology of new onset status epilepticus is identified. Along with seizure management, they should be treated with immunomodulators. Closer follow up is needed while tapering the steroids as relapse can occur with behavioral issues and seizures and they may benefit from steroid sparing long term immunomodulatory treatment.

Non-convulsive seizures (NCSZs) and non-convulsive status epilepticus (NCSE) are common in critically ill patients. Both are associated with neurophysiological disturbances, and even mortality if untreated in a timely manner.

[1]Continuous electroencephalogram (cEEG) monitoring has been proven to be effective in diagnosing NCSZs and NCSEs, and assessing the efficacy of treatment thus it is a vital investigation.

[2]

We conducted a national survey on the availability of cEEG monitoring within Neuro Critical Care Units (NCCU) in the UK. To ensure accuracy the consultant in charge or ST 5-7 covering the NCCU was contacted by telephone and asked a serious of questions regarding their use of cEEG and reporting.

28 hospitals were identified as having either stand alone or mixed NCCU. Responses were obtained from 27 of the 28 units contacted. Only 46% of NCCUs were able to perform cEEG monitoring from 9am-5pm

This dropped to 39% at night. In 68% of NCCUs the ITU consultant did not feel confident to analayse the cEEG and make treatment decisions based upon in.

The inability of 54% of NCCU to perform cEEG is very concerning, as a single EEG may miss episodes of status, and also makes treatment to achieve burst suppression very difficult. In addition, there appears to be a training gap in ability of ICU doctors ability to interpret cEEG. Commissioning standards may need to be modified to encourage take of this vital monitoring technique. In addition systems such as possibly setting up a central remote analysis site for all cEEG data for England might improve time to diagnosis and treatment whilst still remaining economically.

Traumatic brain injury (TBI) is the leading cause of disability in children. Neuroimaging is essential for the acute evaluation of moderate-severe TBI, although its prognostic utility is unclear. Magnetic resonance imaging (MRI) allows for detailed characterization of diffuse axonal injury (DAI), the hallmark pathology described in non-penetrating TBI. Higher DAI grade in adults correlates with worse outcome, but this association has not been rigorously tested in children. We hypothesize that acute Rotterdam Score and DAI grade predict short-term functional outcome in children with acute TBI.

Patients admitted to Stanford Children's Hospital for acute TBI were identified via retrospective chart review based on ICD9 and ICD10 codes for TBI. Inclusion criteria were age >1mo and <19yrs with blunt, closed head trauma and MRI brain obtained during hospitalization. Exclusion criteria included history of epilepsy, prior TBI, developmental delay, and penetrating or non-accidental trauma. The first head CT and brain MRI obtained during hospitalization were used for analysis of Rotterdam Score and DAI grade, respectively. Discharge destination (home versus facility) was used as a marker of short-term functional outcome.

Multiple logistic regression analysis on cohort of 44 children revealed that lower GCS and ventriculostomy were independent predictors for discharge to acute rehabilitation (OR 1.6 and 27, respectively) versus discharge home. Neuroimaging analysis revealed that more severe DAI significantly correlated with discharge to a rehabilitation facility (p=0.011), while Rotterdam CT score did not correlate with discharge destination (p=0.106).

Our study demonstrates that higher DAI grade is associated with worse short-term outcome in pediatric patients understand the short-and long-term prognostic value of acute neuroimaging in pediatric TBI.

, Niteroi, Brazil

Zika virus has been associated with several neurological complications. We aim to present three cases of Zika associated subacute encephalitis, all requiring intensive care.

All patients derived from the RIO-ZIKV-GBS study cohort. All were diagnosed with MAC-Elisa and PCR for

Case 1: 33-year-old man admitted with lower extremities weakness and urinary retention, preceded by -capsular area, extending to the corona radiata and cerebellar peduncles. He was treated with a 5-day cycle of intravenous immunoglobulin (IVIg). He was discharged one year later due to protracted weaning from mechanical ventilation. Case 2: 53-year-old man admitted with lower extremities weakness, dysphagia, and dysphonia. 11 days before he presented with and middle cerebellar peduncles, extending to pyramidal tracts. He was treated with IVIg. He was discharged after acute treatment and, one year later, presented only with ataxic gait. Case 3: 48 year-old woman admitted with disorientation and behavioral impairment. A week before she presented with 0% mononuclear) with mild protein elevation. MRI revealed hyperintense -T5 levels. She was also treated with IVIg. A year later her neurological exam returned to baseline.

All patients had similar clinical presentation, starting with atypical measles syndrome, later evolving to a subacute encephalitis. All showed similar radiological findings, resembling the ones observed with Japanese Encephalitis, another flavivirus. This new entity is likely a result of ZIKV-mediated autoimmune activation and it is a challenge for neurocritical care units worldwide.

There are two described forms of necrotizing encephalopathy: multifocal necrotizing leukoencephalopathy (MNL) and acute necrotizing encephalopathy (ANE). MNL is characterized by multiple microscopic foci of white matter necrosis and is sporadic with predilection for the pons in patients with sepsis or immunosuppression. ANE is characterized by multiple foci of grey and white matter disease and is either sporadic or familial; it is typically triggered by febrile viral illness in children without evidence of cerebral infection.

A case report with review of the clinical, laboratory, radiographic, and pathologic data.

A 39-year-old woman with post-traumatic epilepsy was admitted with acute encephalopathy and respiratory failure secondary to H1N1 and strepotococcal pneumonia. She developed refractory hypoxemia requiring proning and eventually veno-veno extra corporeal membrane oxygenation. Her neurological exam declined with no response to painful stimuli and absent corneal reflexes. Continuous restricted diffusion lesions of the cerebral white matter, splenium of the corpus callosum, brainstem, cerebellar peduncles, and deep cerebellum. She died after transition to comfort care and autopsy was pursued by family. Neuropathologic evaluation revealed microscopic acute and subacute necrotizing lesions throughout the white matter of the cerebrum, pons, and cervical spinal cord. There were similar lesions throughout the thalamus with sparing of other gray matter structures. There was no significant lymphocytic inflammation or meningoencephalitis.

This presentation is consistent with MNL, yet the thalamic involvement is more characteristic of ANE. However, ANE is rare in adults and typically affects both the grey and white matter. Our case affected mostly white matter with microscopic lesions in the grey matter of the thalamus. This case is unique in that it has features of both known necrotizing leukoencephalopathies without clear classification.

Pharmacotherapy after traumatic brain injury (TBI) aims to prevent secondary insults by optimizing brain homeostasis. To better understand the relationships between medication infusions and cerebral dynamics, we investigated their associations with cerebral compliance (CC), autoregulation (CA) and heart-rate variability (HRV).

A retrospective analysis of severe TBI patients admitted to the pediatric ICU who underwent brain multimodal monitoring was performed. CA, CC and HRV were estimated by using different parameters: CA by using the pressure reactivity index -a Pearson correlation coefficient; CC by using the RAP indexa correlation between ICP and pulse amplitude; HRV by heart-rate root mean square of successive differences. Analysis of variance was used to investigate cerebral dynamics differences during narcotic/sedation (dexmedetomidine, fentanyl, propofol), barbiturate (pentobarbital), vasoactive (epinephrine, milrinone, nicardipine, norepinephrine, phenylephrine) and paralytic (vecuronium, rocuronium) medication infusions.

46 children were identified (13 female; ages 0-17 years). CA values were significantly higher (i.e. larger positive values) in patients who received vasoactive infusions than those who did not (epinephrine (0.326±0.001), norephinephrine (0.150±0.001)). CC values were much larger (closer to 1) in patients who received barbiturate and paralytic infusions compared to those who received narcotic/sedation infusions (pentobarbital (0.517±0.001), vecuronium/rocuronium (0.506±0.001), fentanyl (0.371±0.001), dexmedetomidine (0.326±0.001), propofol (0.316±0.001)). HRV displayed significantly larger values in patients who received narcotic/sedation infusions compared to those who received barbiturate infusions (propofol (35.040±0.446), dexmedetomidine (34.732±0.139), pentobarbital (13.704±0.171)).

These results suggest vasoactive infusions (epinephrine and norepinephrine) are associated with impaired CA, narcotic/sedation infusions (dexmedetomidine and propofol) are associated with improved CC and greater HRV, and barbiturate infusions (pentobarbital) are associated with impaired CC and less HRV after severe TBI. Prospective analysis is needed to validate these associations and investigate whether these medications may be contributors or epiphenomena of altered cerebral dynamics.

Sleep wake disturbances (SWD) after pediatric traumatic brain injury (TBI) requiring critical care admission are poorly quantified, but may have important implications for patient recovery. We conducted a systematic review to quantify SWD after pediatric TBI requiring critical care, identify interventions for SWD, and determine the association between SWD and other Post-Intensive Care Syndrome (PICS) morbidities after TBI.

injury requiring neurocritical care published after 1999 and reporting a sleep or fatigue outcome. Studies focused on concussion or mild TBI without differentiation of intracranial injury requiring critical care hospitalization were excluded. Risk of bias was assessed for included studies. A meta-analysis was not performed due to heterogeneity of included studies.

Search results yielded 966 articles. Abstract review yielded 126 articles, and 21 studies were included in the final analysis (17 observational, 4 case reports). We found children with TBI had significantly more SWD when compared to controls. Studies reported over one third of TBI patients have SWD, some persisting for years after injury, but often failed to delineate phenotypes of sleep problems. Most studies used subjective measures with questionnaires or interview. Seven studies used a validated sleep questionnaire. Three studies with 16 total patients presented objective data on SWD using actigraphy (n=1), polysomnography (n=1), and electroencephalography (n=1). Outside of one case report, no studies evaluated interventions for SWD following pediatric TBI. SWD in children surviving TBI were associated with PICS morbidities including reduced quality of life, behavioral problems, and neurocognitive impairment. Heterogeneity and risk of bias among studies was high.

Research is needed to quantify SWD, including identifying phenotypes and utilizing objective measures of sleep. Evaluation of pharmacological, psychological, and behavioral interventions for SWD is warranted given associations between SWD and PICS.

Current guidelines for pediatric severe traumatic brain injury (TBI) recommend maintenance of mean intracranial pressure (ICP) under 20 mmHg. Increasing evidence has suggested that ICP waveform characteristics may be important in understanding the impact of pressure on cerebral physiology. Our study objective is to investigate strength of association of brain tissue oxygenation with ICP waveform characteristics.

Retrospective analysis was performed on pediatric patients with TBI who underwent multimodality monitoring including measurements of PbtO2 and ICP between January 1, 2014 and January 1, 2019. Data were limited to relatively normal values of PbtO2 between 15 and 50 mmHg and ICP values between 0 and 60 mmHg. Univariate linear regression was performed to assess strength of association between PbtO2 and ICP waveform characteristics including, mean ICP values, ICP pulse amplitude (AMP), and minimum and maximum values of the ICP waveforms.

17 patients were identified (4 female, ages 3-19 years [mean 8.6; interquartile range 11.8 -15.6]). PbtO2 was negatively associated with all 4 ICP characteristics following analysis. The correlation coefficient (r) was stronger with respect to the relationship of PbtO2 to AMP (r = -0.34) as compared to mean ICP (r = -0.08), maximal ICP (r = -0.14) and minimal ICP (r = -0.04). P-values were < 0.001 for all measurements.

These data provide preliminary evidence that ICP pulse amplitude is associated with PbtO2. These findings suggest that ICP waveform amplitude should receive greater scrutiny in understanding the impact that ICP has on PbtO2 after pediatric severe TBI though further research is necessary to confirm this finding.

Sarcoidosis is a systemic disease characterized by formation of noncaseating granulomas. In 5-15% of cases, sarcoid infiltrates the central nervous system causing a myriad of clinical symptoms and imaging findings. Although rare, neurosarcoidosis commonly involves the brainstem, hypothalamic-pituitary axis, leptomeninges, and spinal cord, causing symptoms such as cranial neuropathies, hypopituitarism, aseptic meningitis, and seizures. Based on the review of literature, neurogenic shock as a complication of neurosarcoidosis has not been previously reported.

A retrospective chart review was performed on the patient's medical records to obtain laboratory results, imaging studies, and treatment modalities. We demonstrate a case of neurosarcoidosis that initially presented with neurogenic shock, seizure-like activities, and anterograde amnesia.

A 28-year-old African American man with neurogenic shock and seizure-like activities was transferred to our neurointensive care unit. Initial workup revealed panhypopituitarism, including hypothyroidism and central diabetes insipidus. MRI of neuro-axis was significant for diffuse parenchymal and leptomeningeal enhancing lesions of unclear etiology, including the hypothalamic-pituitary axis, bilateral mesial temporal lobes, and cervical spinal cord. He was intubated for airway protection and treated with dopamine infusion for hypotension and bradycardia thought to be a manifestation of neurogenic shock from his extensive cervical spinal cord lesion. Despite significant cervical cord involvement, he remained with good strength throughout. He was extubated after a short course of high dose steroids and stabilization of electrolytes and endocrine function however was found to have anterograde amnesia -PET revealed hypermetabolic lymphadenopathy throughout the neck, chest, abdomen, and pelvis without cardiac involvement. He subsequently underwent lymph node biopsy which revealed noncaseating granulomas.

Neurosarcoidosis is an infiltrative disease process with varied clinical and imaging presentations. Although neurogenic shock is classically seen as a complication from spinal cord injuries above the T6 segment, neurosarcoidosis affecting the cervical spinal cord can also present with neurogenic shock.

The primary goal of traumatic brain injury (TBI) management is the prevention of secondary injury achieved by invasive intracranial pressure (ICP) monitoring. Near infrared spectroscopy (NIRS) is a continuous, noninvasive surrogate measure of cerebral blood flow and oxygenation making it a potentially useful adjunct in the management of TBI. We aimed to determine the association between regional oximetry (rSO2) and ICP in pediatric TBI.

The association between rSO2 and ICP was estimated retrospectively in 56 pediatric patients with severe TBI. Digital record using univariate dynamic structural equations modeling with a 95% credible interval (95% CI) for the standardized regression coefficients (SRC).

54 of 56 study patients had documented events. The association between rSO2 and ICP varied between patients and event type. No events triggered by changes in rSO2 occurred. A significant positive (SRC=0.035, 95% CI=0.023 -0.046; SRC=0.043, 95% CI=0.030 -0.054 respectively). A negative r this was not significant (SRC=-0.012, 95% CI=-0.026 -0.005).

During times without intracranial hypertension, changes in ICP were positively associated with changes in rSO2, which may be related to changes in cerebral blood flow. Our results also suggest that cerebral desaturation may be seen during periods of intracranial hypertension. Our data supports the utility of NIRS as an adjunct to understanding changes in ICP, however further research is needed to determine if these findings are clinically relevant.

Rapidly progressive (< 24 hours) primary angiitis of the central nervous system (PACNS) has rarely been reported in the literature. Most cases have resulted in death. Here, we describe the neurocritical care course of a patient with rapidly progressive PACNS who survives with a good outcome.

Data was collected prospectively through direct patient care and chart review.

A 20-year-old previously healthy male presented to an emergency room in acute coma. Initial head CT showed diffuse cerebral edema and a left thalamic intracerebral hemorrhage. Non-contrast brain MRI c perivascular enhancement suggestive of cerebral vasculitis. An external ventricular drain was placed for intracranial pressure monitoring and cerebrospinal fluid sampling, which showed a neutrophilic pleocytosis (WBC=42, 88% PMN). Brain biopsy on hospital day (HD) #2 was consistent with a diagnosis of necrotizing PACNS. Rheumatologic evaluation was negative for systemic inflammatory disease. Therapy included methylprednisolone, plasma exchange, and cyclophosphamide. His hospital course was complicated by ventilator-associated pneumonia, thrombocytopenia, cerebral salt-wasting, and malignant intracranial hypertension which was treated with hypertonic therapy, barbiturate coma, and hyperintensities and resolution of perivascular enhancement. He required tracheostomy and percutaneous gastrostomy and was discharged to a ventilator facility on HD #31. On discharge, he was awake and texting on his cell phone. At 2-month follow-up, his modified Rankin Score was 3.

Our case demonstrates that rapid diagnosis, early immunosuppressant therapy, and aggressive neurocritical support in collected on the optimal therapy of the patients with rapidly progressive PACNS.

, Detroit, MI, United States

Cerebral amyloid angiopathy (CAA)-related inflammation, or cerebral amyloid angiitis is an uncommon disease that presents with acute symptoms secondary to a solitary area of vasogenic edema. This series examines 6 patients presenting with acute neurological symptoms and imaging out of proportion to their exam, suggesting this is a common trend in this diagnosis.

Cases were collected through EPIC review, using slicer/dicer to select patients with both snomed diagnoses of CAA and CNS vasculitis, and snomed diagnosis of CAA concurrently treated with prednisone 2010-19.

Cases: (1) 67 year old female with prior diagnosis of CAA presents with transient worsening of right arm dexterity and word-finding difficulty.

(2) 78 year old female presented with loss of vision in the right eye lasting for 2 hours (3) 73 year old female presents with two days of word-finding difficulty and confusion, using her car remote for her television (4) 72 year old male presenting after being unable to find words and acting out for two days (5) 65 year old male with prior diagnosis of CAA presents with one day of confusion and nonsensical speech.(6) 84 year old male with history of bilateral occipital hemorrhages of cryptogenic etiology presents with two days of new onset dizziness and left hemianopsia.

In each case, patient was identified to have a focal area of vasogenic edema on MRI that was significant and alarming in comparison to the patient's presenting symptoms. SWI MRI showed numerous microbleeds elsewhere to the vasogenic edema consistent with CAA. Considered differentials included herpes encephalitis, MELAS, CADASIL, and CNS vasculitis due to lupus, however all patients exhibited a neurological exam less severe than expected of differentials mentioned prior. All patients were administered an oral steroid regimen with taper for an average of 4 weeks and their symptoms resolved on follow up.

Use of cranial ultrasound (CUS) in pediatrics has been limited to neonates or infants and transcranial doppler (TCD) for stroke risk in children with sickle cell disease. We describe a clinical case showing the utility of performing CUS/TCDs to assess for new intracranial process in a pediatric patient where head CT was difficult to obtain due to high frequ assessment of waveforms on TCD can be a useful bedside tool in assessing progression of cerebral edema in pediatric patients unable to get a head CT.

14-month child with acute respiratory distress syndrome required veno-venous ECMO and therapeutic anticoagulation complicated by intracranial hemorrhage with intraventricular extension, 15 mm leftwards midline shift, and hydrocephalus. Heparin was reversed and EVD was placed. Since heparin sedation/paralysis. Osmotic therapy was guided by elevated ICP. 3 days later, the ability to monitor ICPs became unreliable due to intermittent EVD dra repositioning was deferred because of bleeding risk and lack of clarity whether device malfunction or unsafe because of waveforms with robust arterial diastolic flow and venous flow signifying that ICP was lower than plaining unreliability of and repeat head CT showed no gross change.

CUS and TCD can be a useful tool to screen for high ICP using midline shift and spectral waveform analysis in pediatric patients where CT may be contraindicated or challenging to obtain.

The structure of intensive care has evolved as the field of medicine has created needs for specialized care. Large pediatric hospitals frequently have separated cardiac ICU from general pediatric ICUs, however further subdivision is rare, which differs from adult institutions that often have surgical and neuro ICUs. This subdivision capitalizes on concentration of expertise and collaboration across providers to improve patient outcomes. Texas Children's Hospital recently opened a new pediatric ICU tower and subdivided the PICU into six specialty units: surgical, neurology/neurosurgery, pulmonary, hematology/oncology, medical and transitional (for patients with complex needs). We sought to retrospectively review similar patients fitting predefined neuro ICU criteria both pre and post move to determine if patient outcome measures were different after cohorting patients.

We conducted a retrospective review of neuro ICU patients before and after our specialty ICU model by comparing June-August 2017 to June-August 2018. Patients were identified using local data from Virtual Pediatric Systems (VPS, LLC) and outcomes collected from the electronic medical record utilizing automated data query. Primary analysis included patient demographics and outcomes including ICU length of stay (LOS), mortality, PRISM-3 and PIM-3 risk of mortality scores. Early subgroup analysis included patients with ICP monitoring devices in both cohort groups.

172 and 273 patients were in the pre and post cohort group respectively, 10 of which had ICP monitors in each group. Median time to ICP measurement was 700 (IQR 117-1598) and 574 minutes (IQR 56-980) respectively in pre and post groups (p = 0.56). ICU LOS, mortality, PRISM-3 and PIM-3 were not statistically different.

We have developed an algorithm to capture the neuro ICU population for future study. Preliminary investigations will hopefully confirm patients benefit from this model after programmatic maturity is achieved.

West Nile Virus (WNV) is a mosquito transmitted arbovirus that is endemic in the United States. Only 25% with acute infection develop fevers, and only less than 1% develop neuroinvasive disease. Although the presentation of acute flaccid paralysis is not uncommon, it is extremely rare to visualize the destruction radiographically. Here we highlight a case of aggressive neuroinvasive disease with radiographic changes.

Results 67 y/o Caucasian male with arthritis on methotrexate and tofacitinib presented with encephalopathy and generalized weakness. Initial evaluation included MRI and lumbar puncture. Initial MRI did not demonstrate etiology of symptoms. Lumbar puncture was consistent with viral meningitis (WBC 260, RBC 350, glucose 72 and protein 62). Patient was started on broad spectrum coverage. There was no growth on bacterial or fungal cultures. PCR Biofire was negative for acute viruses. Weakness progressed, and required intubation for neuromuscular respiratory failure. Diagnostic evaluation was repeated 10 days later. Repeat MRI demonstrated changes on DWI and T2 weighted imaging, following the motor addition to continued acyclovir, plasma exchange was initiated for an attempt at treatment. The patient's mental status improved, and he refused further treatments including tracheostomy. He was extubated and comfort care was provided given his continued neuromuscular respiratory failure.

This case demonstrates severe neuroinvasive West Nile encephalitis and flaccid paralysis with radiographic findings. Being immunocompromised and age increase his risk for rare presentation of aggressive disease.

Evidence regarding adequate caloric requirements of critically ill patients with acute brain injuries is suggesting potential risk of caloric debt in neurocritically ill patients. The primary objective of this study was to determine whether guideline recommended weight-based dosing provides adequate caloric requirements compared to indirect calorimetry (IC) measurements in this population.

This was a single center, retrospective, observational case-crossover study that included adults admitted within 14 days from admission. We compared resting energy expenditure (REE) determined via IC to the lower (BMI<30 kg/m2: 25 kcal/kg and BMI 30-50 kg/m2: 11 kcal/kg) and higher (BMI<30 kg/m2: 30 kcal/kg and BMI 30-50 kg/m2: 14 kcal/kg) actual body weight-based dosing guideline recommendations. We hypothesized that guideline recommended lower-weight based nutrition will not match the caloric demand of patients with acute brain injuries.

A total of 52 metabolic studies were performed in 43 patients (36% ICH, 20% non-traumatic SAH, 20% ischemic stroke, 16% TBI, 6% status epilepticus, 2% other etiologies). The mean age was 58+18 years, mean weighed 89+27 kg with a BMI of 30+10 kg/m2, and had mean baseline GCS of 7 + 4. On average IC was obtained on day 5 of admission. Lower weight-based recommended nutrition did not provide adequate caloric needs as measured by IC adjusted for obesity (1652±485 vs 1952±611 kcal/day, p<0.001). However, higher weight-based recommendation matched the caloric demand as measured by IC (2006±563 vs 1952±611, p=0.417) .

In this preliminary analysis, higher weight-based dosing for nutrition matched the caloric demand of critically ill patients with acute brain injury. Our results need to be confirmed in future larger prospective studies.

Central venous catheter (CVC) insertion is common in neurocritically ill patents. Standard practice is to obtain a chest radiograph (CXR) to evaluate for the presence of complications, such as pneumothorax (PTX) and catheter misplacement. Point-of-care ultrasound (US) has been suggested as an alternative methodology to assess for these complications by using a flush test.

Patients admitted to our Neuro ICU between 6/5/18 -5/8/19 who required CVC placement were the subject of this quality improvement analysis. CVC's were placed in the internal jugular (IJ) or subclavian (SC) vein followed immediately by lung US to assess for PTX. Then, apical or subcostal four-chamber view of agitated saline injected through the distal port of the CVC (ie. flush test) was performed to assess for proper placement. We observed the time delay between start of agitated saline instillation and visualization of contrast in the right atrium and ventricle. This was then interpreted as appropriate (contrast present in </= 2 cardiac cycles) or not.

Twelve patients were included in this analysis. No patients had a PTX on US or CXR. One patient had a misplaced catheter detected by an abnormal flush test after subclavian CVC insertion using the landmark technique. This was further confirmed by US visualization of the CVC terminating in the ipsilateral IJ vein. After immediate removal and reinsertion, followed by an appropriate flush test, CXR then confirmed accurate placement. Ultrasound evaluation of CVC complications took < 10 minutes in all cases while it took an average of 45 minutes for the CXR to be avai a potential savings of $200.00 for each CVC inserted by avoiding the cost of a CXR.

In conclusion, we demonstrated US post CVC placement to be an accurate, fast, and potentially less expensive method of confirming CVC placement compared to CXR.

Major professional critical care societies recommend shared decision-making using decision aids (DAs) to reduce surrogate decisions incongruent with patient values and preferences and decisional conflict. No DAs currently exist for any decisions in the neurologic intensive care unit (neuroICU), including goalsof-care decisions. Using multilevel input from 88 stakeholders, we recently developed a goals-of-care DA for critically-ill traumatic brain injury (TBI) patients that meets international DA standards. We aimed to adapt this DA to intracerebral hemorrhage (ICH) and hemispheric acute ischemic stroke (AIS) to widen its applicability and fulfill the shared decision-making needs of surrogate decision-makers of stroke patients.

we adapted the existing goals-of-care DA by replacing the TBI prognostication model with diseasespecific outcome models as recommended by DA quality-standards. We selected the validated 12month ICH-model and 12-month hemicraniectomy trial outcome data for AIS because they include validated long-term outcome assessments. Next, we performed acceptability and usability testing of both DAs using validated scales in family members recruited from two academic neuroICU waiting rooms.

We successfully modified the DAs and created new icon-arrays to optimize the display of predicted outcome probabilities and uncertainties, as has been shown in risk communication studies. Twenty family members from two centers participated in acceptability and usability testing. Median usage time was 22±7 minutes. Usability was very good (mean System-Usability-Scale-score =77/100). 89% of amount, balance and comprehensibility.

We developed three goals-of-care DAs for surrogates of critically ill TBI, ICH and AIS patients that include disease specific icon-arrays to communicate outcome probabilities meeting international DA standards.

DAs is ongoing, in preparation for a future multicenter comparativeeffectiveness trial. Our rigorous development process following international DA standards may serve as a template for other neuroICU DAs.

Neurocritical care patients receiving rate-based enteral nutrition (RBEN) consistently received less than their full prescription of enteral nutrition (EN) volume due to interruptions in feeding. Rate-based EN does not allow the nurse to increase the rate after interruptions leaving patients at risk for underfeeding (Kim et al. 2010 , Yeh & Peev, 2016 . Low caloric intake over time is associated with a) immunosuppression, b) impaired wound healing, c) muscle wasting, d) increased hospital and intensive care unit (ICU) lengths of stay, e) increased costs, and f) increased mortality (Barker et al. 2011, Elke et increase the likelihood of adult neurocritical care patients receiving the predetermined goal of 85% or more of the prescribed volume of enteral nutrition over the course of the NCCU length of stay through the use of a volume-based EN (VBEN) protocol which allows the RN to increase the tube feeding rate after interruptions occur.

A retrospective pre and postimplementation chart review was conducted on patients aged 18 or greater and less than 90 years with a neurological-related injury or disease process who received EN for three or more days during their NCCU stay.

Despite no significant differences in characteristics or gastrointestinal complications between the groups, there was a significant increase in the percentage of prescribed EN volume delivered over the course of NCCU stay of 23.15 percentage points in the VBEN group (M = 95.3, SD = 4.92) as compared to the RBEN group (M = 72.15, SD = 10.55), t(27, n = 40) = 8.89, p <<0.001, two tailed, unequal variances.

VBEN can be safely implemented in the neurocritical care population and is associated with significant improvement in EN volume delivery.

pain, agitation, delirium, and machine support in the ICU. A--Intensive Care Syndrome (PICS), manifest as psychological and physical debilitation that persists beyond discharge. A-neurocritically ill patients is not known. Practices aligned with individual elements of the Apopulation were deemed beneficial for patients within this cohort. We asked the question, does the implementation of the A-

Systematic training on A--professional teams was implemented. Baseline data, total bundle compliance, and individual element compliance were measured and analyzed using an approach adopted from the published SCCM ICU liberation campaign. Mortality, ventilator hours, delirium, length of stay (LOS), and cost were compared to baseline against the third month post education to evaluate the impact of A-

Preliminary analysis suggests that increased compliance with A-whether wholly or in part, resulted in improved value of care, as evidenced by a decrease in mortality from 7% to 4%, as well as a reduction in total ICU LOS adjusted by acuity from 2.19 days to 1.36 days. This translated into a reduction in cost. As additional experience (and data) become available, further analysis is planned to evaluate the impact of A-

Early results from a systematic implementation of Aand pertinent to improving patient outcomes and value of care.

Many neurologic diseases are incurable, and patients are faced with prognostic uncertainty, progressive decline and early mortality. As a result, ACGME requires training in serious illness communication for all neurology residents; however, only 52% of programs offer formal education. To address this gap at our institution, we created a training program for neurology residents to improve their skills, attitudes, and behaviors regarding serious illness conversations

We designed and implemented a 3-hour training program utilizing the Serious Illness Conversation Guide (ariandnelabs.org). The training program is comprised of a didactic session, a simulated patient experience and debriefing. Residents were asked to have a goals of care (GOC) conversation with family members (actors) of a patient with catastrophic stroke. A faculty facilitator observed each interaction, allowing residents to self-assess, obtain feedback from the actors and peers and then practice specific components of the conversation. Residents filled out a survey both prior to and following the simulation and the faculty filled out an evaluation for each of the residents.

led out a pre-survey. The majority (8/12) felt that the residency did not adequately prepare them to have GOC discussions though more than 50% of residents reported feeling comfortable and confident performing these discussions. Despite this confidence, the observing faculty found 10/14 residents require direct supervision and are not ready to have these conversations independently. The residents struggled with conveying prognosis, maintaining silence, and dealing with strong emotions. All residents reported the training was a helpful, rewarding experience and improved their overall abilities to discuss GOC, address prognosis, and explore end-oflife issues.

This small pilot study demonstrates an innovative model to improve residents' communication skills surrounding effective GOC discussions even when they report increased confidence and independence in these difficult conversations.

Studies of patient outcomes in mechanically ventilated ICU patients have discovered associations among sedative drugs, patient characteristics (demographics and illness), burst suppression (brain inactivation), and mortality. However, association does not imply causation. It remains unclear (1) whether unintentional burst suppression has a causal effect on mortality, (2) the degree to which drugs directly impact mortality, and (3) whether drugs impact mortality indirectly by causing burst suppression. Patient characteristics are a possible source of confounding. We estimated causal effects of unintentional burst suppression, critical illness severity and the anesthetic sedation on in-hospital mortality in mechanically ventilated ICU adult patients.

This was a longitudinal study with an observational cohort of 474 adult ICU patients. EEG recordings were obtained for each patient and reviewed by 2 clinical neurophysiologists per institutional protocol to determine the presence of burst suppression in each epoch (8-12 hour segment). The causal effects were estimated using causal statistical inference (Pearl's do-calculus). To assess the robustness of our results towards unmeasured confounding, we performed sensitivity analysis using the E-value. 95% Confidence Intervals were created using bootstrapping.

On average, increased drug exposure causes a 4% increase in mortality (95% bootstrapped confidence interval -.01, .1) and a 5% increase in burst suppression incidence (.01, .09). Burst suppression mediates 15% of the impact of drug exposure on mortality and is associated with a 15% increase in mortality (.04, .24). Increased severity of illness causes a 43% increase in mortality (.13, .66), and a 13% increase in burst suppression burden, measured by the percent of epochs during which patients experienced burst suppression (-.12, .32).

Unintentional burst suppression contributes substantially to in-hospital mortality. However, unintentional burst suppression appears to be affected more by patient characteristics such as severe critical illness than by high exposure to sedative drugs.

Infectious endocarditis (IE) is a growing health concern, with increasing incidence in the context of the opioid epidemic. We described a large case series of patients with IE at an urban, academic safety-net hospital.

Clinical data on patients with diagnosis of IE were entered into an endocarditis work group single-center registry. The mission of the multidisciplinary group is to optimize clinical care, especially as it pertains to cardiac surgical management. We reviewed demographic information, history of prior cardiac valve replacement, intravenous drug use (IVDU), blood culture results, and diagnostic studies (ie echocardiogram, brain CT/MR imaging, and conventional angiography).

Between November 2017 and May 2019, 99 consecutive patients (mean 41.2 years; aged 20-82 years; 46% male) with IE were admitted to our center. Patients had a history of IVDU in 73% of cases. 37 unique microbes were isolated, most commonly Staphylococcus species (33 MSSA, 22 MRSA, 10 other). Polymicrobial infection occurred in 33 patients. On echocardiogram, vegetations were seen on tricuspid (48), mitral (49), aortic (18), and pulmonic (2) valves. 10 patients had multi-valve involvement. 63 patients underwent brain CT or MRI demonstrating-cerebral infarctions (20), subarachnoid hemorrhage (SAH) (6), intracerebral hemorrhage (ICH) (9), brain abscesses (3), and mycotic intracranial aneurysms (4). Neurologic complications occurred most frequently in the setting of mitral valve disease (15 infarcts, 5 SAH, 5 ICH) and in those with Staphylococcus (11) and polymicrobial infection (7). Tricuspid valve vegetations were associated with 4 cerebral infarcts and 1 ICH.

Despite frequent association with IVDU, left-sided cardiac valve vegetations were common. Mitral valve involvement, Staphylococcus species and polymicrobial infections were associated with increased risk for neurologic complications. Neurologic complications also occurred in patients with right-sided valvular disease, suggesting a need for increased surveillance in this population.

Acute neurological conditions such as traumatic brain injury (TBI), cerebrovascular accident (CVA), and hypoxic-anoxic injury (HAI) can significantly impact a patient's functional status, and life expectancy. Older adults may have poorer outcomes, especially those with premorbid conditions. Studies have shown a preference for less aggressive care if patients and/or families anticipate a shorter life expectancy, or poor outcome. As such, early delivery of prognostic information and discussion of goals of care (GOC) can be crucial for patients prior to transfer to a tertiary care facility.

This is a retrospective cohort study of patients aged 18 and older with CVA, HAI, or TBI who died within 72 hours of Charts were reviewed to assess documentation of pre-transfer GOC discussions, code status change after transfer, prevalence of palliative care consultation, and other utilization markers.

90 patients met inclusion criteria with 52.2% being male and 47.8% female with median age 66.5. The majority of patients were transferred from an emergency department (84.4%). Only 5.6% of patients had documented GOC discussions prior to transfer. Only 6.7% patients of patients underwent surgical intervention. After transfer, 87.8% of patients downgraded their scope of treatment to comfort measures prior to death, with this being even higher in those 65 and older (95.8%). Palliative care consultation took place in only 5.6% of patients after transfer.

While transfer to a tertiary medical center is sometimes necessary, only 5.6% of patients had documented GOC discussions and the majority (87.8%) downgraded their scope of treatment to comfort measures after transfer suggesting a potential opportunity for earlier GOC discussion and possible avoidable hospital transfers. These findings can inform the design of future interventions to improve communication and decision-making regarding goals of care in patients with high mortality neurological conditions.

A neurological emergency response team (NERT) was implemented at a dedicated neuroscience hospital in an academic medical center in anticipation of higher patient volumes with the extended treatment window for acute ischemic stroke. Goals of the NERT are to ensure safe and effective care transition into the acute care setting and a standardize approach to neurological emergency.

A dedicated clinical team examined the current structure for safe & effective patient triage, monitoring and transition from the interventional neuroradiology (INR) suite to the neurocritical care unit. Clinical nurses with over three years' experience in neurocritical care were vetted, trained & scheduled to provide 24/7 continuous coverage. After patients were triaged via telemedicine technology, the NERT nurse was alerted via automatic text message system and prepared to receive the patient via ambulance to the INR suite. During the assembly of the INR team, the NERT nurse assessed and monitored the patient and initiated diagnostic testing. Upon completion of intervention, the NERT initiated transfer of care to the neurocritical care unit. Stroke volume and door to treatment times were evaluated pre-and post-intervention.

In 2017, 289 ischemic stroke patients were admitted through the INR suite, with 74 being treated with endovascular intervention. In 2018, 560 ischemic stroke patients were admitted, with 157 patients being treated with endovascular intervention. Since the implementation of the NERT, average median arrival time to skin puncture for intervention decreased by seven minutes with a 50% increase patient volume.

A dedicated Neurological Emergency Response Team is an effective model to support the increase in patient volume for evaluation of acute ischemic stroke intervention. This approach demonstrates the potential for efficient, safe and timely triage, intervention and care transition to the Neuro ICU for stroke patients.

Arctic ground squirrels (AGS) endure harsh winter environments through hibernation. Curiously, AGS are resistant to stroke and other ischemic injuries independent of season or metabolic state. The details of physiologic and cell autonomous genetic adaptations underlying this metabolic resilience is largely unknown.

The resilient metabolic and mitochondrial phenotype of AGS neural cells was examined in vitro after exposing cells to hypoxia, hypothermia, or mitochondrial toxins and assaying for cell survival, mitochondrial morphology, and oxygen consumption rate. A cDNA library prepared from AGS cells and nucleofected into mouse neural cells exposed to metabolic stressors was used to perform a functional genetic screen. A bio-informatics pipeline identified screened candidates with key amino acid substitutions that may have evolved to optimize stress response function. Newly described dCas9 base editors capable of substituting C->T (and G->A) were used to systematically mutate and explore the role of identified proteins in mediating the AGS optimized adaptive stress response.

We found that AGS neural cells exhibit marked resistance to all metabolic stressors. This is associated with enhanced mitochondrial function and improved morphology. The functional genetic screen identified a network of evolutionarily-conserved AGS transcripts imparting cytoprotection. Use of dCas9 base editors on candidates suggested by the bio-informatics pipeline, confirmed the coordinated role of specific components of the oxidative phosphorylation (OXPHOS) and endoplasmic reticulum (ER) stress response systems in imparting mitochondrial and neuroprotection in our in vitro model.

We gained key functional insights into how specific amino acid substitutions in the machinery of the OXPHOS and ER stress responses systems alter mitochondrial function to impart cytoprotection to metabolic insults. This detailed dissection of the AGS optimized adaptive stress response pathway will serve as an template for the development of new neuroprotective treatments.

Acute ascending weakness with respiratory failure is a frequent syndrome encountered in the neurocritical care unit (NCCU), often related to demyelinating or infectious etiology. However, here we describe a case of acute ascending weakness with encephalopathy, respiratory failure and autonomic instability that was related to confirmed endocrinological etiology.

Prospectively collected data was retrospectively extracted from the electronic health record in a patient known to our NCCU team.

A 24-year-old male with medical history of childhood meningitis was transferred to the NCCU after initially presenting to an outside emergency department (ED) with a chief complaint of bilateral lower extremity weakness progressing to paraplegia over 4 hours. Six hours into his course in the ED, he developed bilateral upper extremity paresis and respiratory distress. Physical exam in this ED was additionally notable for areflexia and a sensory level at T7. He was intubated, initiated on IVIG and methylprednisolone, and airlifted to our institution. Upon arrival, telemetry showed frequent supraventricular tachycardias refractory to standard treatment. Labs (including cerebrospinal fluid) were notable only for serum potassium <1.5 mEq/L, thyroid stimulating hormone <0.01 uIU/mL, T3 5.2 uIU/mL (0.8--1.7). He was diagnosed with thyrotoxic periodic paralysis. At endocrinology's urging, the patient was given propranolol 1 mg IV every 10 minutes for 3 doses, propylthiouracil and hydrocortisone. In the hours following propranolol, his potassium improved, his paralysis and encephalopathy resolved, and he was ultimately extubated without difficulty <48hours after admission. Review of symptoms performed after improvement revealed recent symptoms consistent with hyperthyroidism.

Intensivists should remain aware of the differential diagnoses that can manifest with motor weakness and respiratory failure. In this patient, severely elevated thyroid hormone led to thyrotoxicosis and subsequent profound hypokalemia. Acquiring a thorough history and reviewing laboratory abnormalities remain paramount for timely diagnosis.

The objective of the study is to determine the prevalence of disability among ICU survivors one year after admission, and factors influencing functional outcome.

We conducted a population based cohort study in the ICUs of the Mayo Clinic, Rochester, MN. We enrolled consecutive patients from the Mayo Clinic Study of Aging (MCSA) and then admitted to medical or surgical adult ICUs at Mayo Clinic, Rochester between January 1, 2006, and December 31, 2014. Patients admitted to the Neuroscience ICU were excluded. We collected their demographic and clinical variables, length of ICU stay, functional and cognitive status (before and after ICU admission), comorbidities (components of Charlson score), and APACHE 3 were retrieved from the Electronic Medical Records using Multidisciplinary Epidemiology and Translational Research in Intensive Care (METRIC) Data Mart. One-year functional outcome was categorized using the modified Ranking Scale (mRS) with scores 0 to 2 representing good functional outcome.

831 cases were included and 569 (68.5%) patients were alive one year after ICU admission. Of them, 546 patients had one-year follow-up functional assessment and 367 (67.2%) of them had good functional outcome. On multivariable analysis, poor one-year functional outcome (death or disability) was more common among women, older patients, baseline cognitive impairment (mild cognitive impairment or dementia), higher Charlson scores, and longer ICU stay (all P<0.01). After excluding deceased patients, these associations remained unchanged. In addition, 120 (32.3%) of 372 patients who had post-ICU cognitive evaluation, experienced cognitive decline after the ICU admission.

Approximately two-thirds of survivors maintained or regained good functional status one year after ICU hospitalization. Older age, female sex, greater comorbidities, abnormal baseline cognition, and longer ICU stay were associated with poor functional recovery.

Shared decision-making using decision aids (DA) is recommended by major professional critical care societies for surrogate decision-making in the ICU to reduce decisions incongruent with patient values and preferences and decisional conflict. We converted a paper-based goals-of-care DA in critically-ill TBI patients to a digital DA. We applied eye-tracking-technology in a single-masked randomized study to understand the effects of and optimize the DA navigation design to facilitate information processing.

We created two digital DAs: (1)unmodified conversion of the paper-DA with horizontal, top-justified static navigation (control) vs. (2)vertical, left-justified navigation with page subsections and page completion checkmarks (experimental), which encourages users to view pages in order. Sixteen healthy participants were randomly assigned to the two groups (n=8/group, masked to DA assignment) and navigated through the DAs. Using t-tests, we compared user disorientation and usability using validated scales, and eye movements (fixation and saccades) recorded with eye-tracking-technology. Impact of navigation on usability was assessed with linear regression, adjusting for disorientation(System-Usability-Score= b0+ b1*Disorientation).

Disorientation was significantly less in the experimental DA (mean 6.5 vs.5.2;p=0.005;smaller values indicating increased disorientation) with no difference in usability (mean System-Usability-Scale scores 81 vs.78;p=0.64;scores>72 indicating good usability[range 0-100]). Regression analysis revealed a significant association between disorientation and usability (p=0.01), with disorientation explaining 31% of the variation in System-Usability-Scale scores (adjusted R2=0.31). Eye-tracking measurements revealed longer average fixation per page in the experimental DA (mean 0.24s vs.0.23s;p=0.03) and a higher ratio of information processing to search per page (fixation-duration over total duration of both fixations and saccades on a page; mean 0.89 vs.0.88;p=0.0003).

Eye-tracking-technology suggested that the experimental navigation design significantly improved the navigation experience resulting in less disorientation and participants spending less time searching and more time processing the information. While there was no difference in subjective usability, we found a significant association between improved navigability and higher usability.

High-fidelity simulation has become an important mode of learning in medical education. Currently, there is little data regarding the impact of simulation-based learning in neurocritical care training. In May 2019, we presented a poster at the American Academy of Neurology Annual Meeting introducing a comprehensive simulation-based curriculum for neurocritical care training at UC San Diego (UCSD). In this poster, we aim to present additional preliminary findings regarding trainee comfort levels, interest, and areas of improvement.

This is a single-group pre-post study involving current residents of the UCSD Department of Neurology. Simulation sessions consist of interactive, faculty-led, and checklist-based clinical scenarios (ischemic stroke, intracranial hemorrhage, status epilepticus, spinal cord emergencies) followed by debriefing sessions. Collected data assesses for self-perceived comfort/confidence levels, future interest, and checklist item completion.

Between January 2017 and July 2018, 22 PGY 2-4 neurology residents participated in various simulation sessions on ischemic stroke, intracranial hemorrhage, and status epilepticus. Prior to the session, 62.5% of all trainees reported no more than somewhat comfortable in treating neurological emergencies despite having received some type of neurological emergency training through didactic lectures. rtable in treating the specific simulation case in observation of each simulation session pinpointed specific areas of improvement amongst trainees on an individual basis (i.e. time to intubation after benzodiazepine administration in refractory status).

Preliminary results suggest that simulation-based learning is valuable and applicable in the neurocritical care training process, allowing trainees to feel more comfortable in managing acute neurological deterioration and faculty to directly observe trainee skill in a controlled setting. Through this project, we hope to highlight the need for simulation-based education in neurocritical care training by providing evaluative information and generalizable curricular examples.

Chimeric antigen receptor (CAR) T cell therapy for refractory/relapsed hematologic malignancy often causes severe neurologic side effects ranging from encephalopathy and aphasia to fulminant cerebral edema and death. The cause of neurotoxicity is poorly understood. We sought to develop a score based on clinical and laboratory parameters to predict which patients would develop CART-associated neurotoxicity.

All patients undergoing CART therapy at Brigham and Women's Hospital for relapsed/refractory hematologic malignancy were prospectively studied. Patients were assessed daily during their admission for cytokine release syndrome (CRS) and neurotoxicity. Vital signs, laboratory data, and medication administration records were extracted from the medical record. Logistic regression was used to determine which clinical and laboratory features were significant predictors of developing neurotoxicity.

159 patients were included. 121 experienced CRS and 73 experienced neurotoxicity. Early (within 3 days after CART infusion) fever and elevated serum C-reactive protein (CRP), timing of CRS onset, CRS grade, and treatment with tocilizumab were all significant predictors of neurotoxicity. Using ROC curves, optimal discriminators were defined and used to derive a score to predict neurotoxicity. One point was assigned for fever, serum CRP > 79.3 mg/dL, and each dose of tocilizumab administrated, zero to four points for CRS grade, and zero to three points for day of CRS onset. This score ranged from 0 to 14 for our cohort and had an AUC of 91%; a score >= 5 predicted neurotoxicity with a sensitivity of 82% and a specificity of 84%. Bootstrap analysis was used to demonstrate robustness.

We used regression analysis to develop a score that can prospectively predict which patients are most likely to suffer from neurotoxicity related to CART therapy. This score can be used for triaging and resource allocation during the care of the patients after treatment with CART therapy.

When brain herniation is impending, every minute matters; so the efficient and expedient procurement of all components required for external ventricular device (EVD) placement is vital to neurological preservation. The neurosurgical residents at the University of Rochester Medical Center often struggled to assemble the appropriate supplies for an EVD placement in a timely manner when patients were not yet admitted to the Neuro Intensive Care Unit (Neuro ICU). Additionally it was difficult to track equipment use and supply costs. In response, the Neuro ICU's Quality Improvement (QI) team designed an EVD "Go Bag" in an effort to improve delays in care, patient experience, and avoidable costs.

The multidisciplinary Neuro ICU QI team collaborated to design a portable bag that contained all equipment necessary for EVD placement. Two neurosurgery residents performed time trails, in real emergency situations, by measuring the time from decision to place an EVD in Emergency Department (ED) Critical Care Bay, to collecting the equipment from the Neuro ICU and return to the bedside in the ED. Times were compared with and without using the EVD "Go Bag".

The EVD "Go Bag" decreased the time to placement of an EVD by up to 17 minutes when compared to the traditional method of retrieving all EVD equipment from the Neuro ICU stockroom.

Time reduction was due to the speed of gathering supplied and the ability for the Neuro ICU staff to bring the EVD "Go Bag" to the patient's bedside. The EVD "Go Bag" allowed for better tracking of monetary costs and equipment, allowing for appropriate billing and stocking of supplies. A system was developed where the bag was checked and restocked daily by the critical care equipment technicians and the Neuro ICU charge nurse

Despite a growing number of prognostication models in neurologic emergencies, prognostic uncertainty remains inevitable and plays a central role during goals-of-care decision-making for incapacitated critically ill patients. We aimed to examine surrogate decision-makers' communication needs and physicians' strategies for communication of prognostic uncertainty during family meetings for critically ill traumatic brain injury (ciTBI) patients.

We qualitatively analyzed semi-structured interviews of 16 surrogates of ciTBI patients from two level-1 U.S. trauma-centers and 20 TBI expert physicians from 7 U.S. trauma-centers. Open-ended questions about prognostic uncertainty were asked. Interview transcripts were analyzed with the investigatortriangulated-inductive-framework-approach in NVIVO-software.

Prognostic uncertainty was identified as the most difficult aspect of decision-making for surrogates by physicians and surrogates alike, although most surrogates had some pre-existing expectation or understanding of it. 95% of physicians observed that uncertainty is distressing for families, with 90% employing specific measures to limit uncertainty. Over half of physicians described explaining the concept of uncertainty so surrogates understand that physicians can estimate the odds but not predict the future. Physicians typically conveyed prognosis using a range of outcomes, and conveying certainty only for prognostic extremes. Surrogates found uncertainty around prognosis was lessened when physicians explained all possible treatment options, with support from clinical data. Roughly half noted that too much certainty in providing a prognosis, without a range of possible outcomes, led to distrust in the information provided by the physician, increasing decisional conflict. The vast majority of physicians admitted statistical uncertainty in deriving prognosis, particularly for patients with TBI, and cited mistrust of prognostic models when deriving long-term prognosis. Most physicians felt that uncertainty around prognosis led to increased incidence of tracheostomy and feeding tube placement.

These results provide foundational knowledge for physician-family communication, by identifying important gaps between surrogates' communication needs and physicians' practices about prognostic uncertainty.

The rapid rise in social media utilization among both patients and healthcare providers has moved a considerable portion of conversation around health and disease to the digital space. Today, roughly nine-in-ten American adults use the internet, with 79% of internet users participating in social media. The power and reach of social media platforms makes it imperative for clinicians to be aware of the trends in the public narrative around common disease processes. In this study, we analyzed the last 2.5 years of postings ("tweets") from a popular social media platform, Twitter, to characterize themes and trends in the digital conversation around stroke, the leading cause of long term disability in the US.

Tweets under the hashtag #stroke, published from January 1st 2016 to April 26th 2019, were extracted through Symplur Signals, LLC. A total of 836,719 #stroke tweets were qualitatively coded and sentiment analysis was performed after selection for relevance among all homographs.

Accounts owned by stroke-related advocacy groups were found to be the most prolific contributors of #stroke postings, with content mostly around primary stroke prevention (risks and signs). Among the 100 most popular associated hashtags, over half of the tweets focused on comorbidities and the challenges of the stroke recovery process (top trending words included #aphasia, #lockedin, #survivor, #depression).

Our preliminary analysis describes trends in themes and stakeholder participation in the current #stroke online conversation. It also exposes important gaps in the public discourse beyond the setting of academic and research online communities, namely around existence of therapeutic treatments, availability of resources for patients and families navigating the recovery process, and possibility of successful recovery and long term outcomes. Such knowledge around the digital stroke narrative may provide valuable context to intensivists and stroke clinicians interacting with patients and families affected by stroke.

The field of autoimmune neurology, specifically the autoimmune encephalitides, has expanded since the early 2000's. Increasingly newer antibodies to various parts of the nervous system are being identified in discovered in patients with meningoencephalomyelitis, or some spectrum of these three singular entities.

Data was reviewed from electronic medical records for this case report.

A previously healthy 31 year-old male initially developed a case of aseptic meningitis, progressing to encephalitis and then extensive longitudinal myelitis leading to profound paresis and respiratory failure. An extensive workup was performed, including evaluation for rare infectious and ominant leukocytosis (491/μL and 328/μL) and elevated protein (> 200 mg/dL). He was treated empirically with antibiotics which were discontinued after negative results and cultures. After therapy with high dose IV steroids he had minimal improvement and pl had improvement in his symptoms. He was started high dose prednisone with plans to slowly taper after return with positive anti-

In review of the literature our patient had several characteristics consistent with others who were also antipsychiatric symptoms. Many reports state steroids lead to remission and improvement, however in this case our patient did not have substantial recovery until after the initiation of PLEX. At this time it is hether these antibodies instead represent a marker of other underlying disease from cytotoxic T cell damage to astrocytes.

The United Council for Neurologic Subspecialties (UCNS) accredits neurocritical care (NCC) subspecialty fellowships and certifies neurointensivists. In 2018, the American Board of Medical Specialties (ABMS) approved the application for NCC subspecialty certification by American Board of Psychiatry and Neurology (ABPN) and the Accreditation Council for Graduate Medical Education (AGME) approved NCC fellowship training in 2019. Previous studies have shown significant heterogeneity in NCC fellowship training and procedural competencies and that many programs do not have the necessary resources for a transition to ACGME accreditation.

In 2018, an online survey of ABPN neurology diplomates was utilized to estimate the number of neurologists practicing NCC, their NCC fellowship training experiences, whether their institutions required certification in NCC, their scope of practice, and their interest in pursuing ABPN certification in NCC.

300 survey respondents indicated that they practiced NCC. Based upon UCNS and other data, this is estimated to be at least 30% of all neurologists practicing NCC. 50% of UCNS-certified NCC respondents identified the primary scope of their practice as academic involving a fellowship program, and 45% of non-UCNS-certified NCC responders identified themselves as private practitioners. Nearly 90% of fellowship trained NCC respondents obtained UCNS certification. 55% of UCNS-certified NCC respondents reported that their institutions required UCNS certification, whereas 83% of non-UCNScertified NCC respondents reported no institutional requirements for certification. Over 80% of respondents thought NCC training was relevant to their current clinical practice. Most respondents indicated that they planned to take the ABPN NCC examination, and >60% of respondents reported that ABPN certification would most benefit them by improving their colleagues' perceptions about the quality of certification.

NCC training and certification is valued by most neurologists practicing NCC, and most believe that ABPN NCC certification will advance the recognition of the field of NCC.

Cerebral edema is a severe complication of acetaminophen-induced acute liver failure (Apapprimary objective was to describe the characteristics of patients with cerebral edema in the setting of Apap-

This analysis is part of a large, retrospective observational study inclusive of Apap-year period from a regional transplant center. We used standardized data collection tools and trained defined cerebral edema based on the interpretation of this CT by a blinded radiologist. We performed univariate analysis based on the presence of cerebral edema.

Of a total of 327 patients, 243 had data on CT brain imaging. The mean age was 39.2 ± 14.6 years, and 167 patients (68.7%) were female. Of patients with neuroimaging, 35 (14.4%) had evidence of cerebral edema. Patients with cerebral edema had higher average ammonia levels on day 1 of hospital admission (226, 95% CI 137 -315 vs. 90, 95% CI 75 -106 mcg/dL). Patients with cerebral edema also had significantly higher MELD scores by 48-hours (32.1, 95% CI 27.9 -36.2 vs. 24.6, 95% CI 22.9 -26.3). This significant difference persisted for subsequent hospital days. Thirteen patients (37.1%) with cerebral edema received intracranial pressure monitoring. Mortality within 28-days was 42.9% (n=15) if cerebral edema was present vs. 14.4% if absent (n=30). The odds of death within 28-days, if cerebral edema was present, was 4.5 (95% CI 2.1 -9.6). One patient with cerebral edema died awaiting transplant, and 3 received liver transplant.

In this study, cerebral edema was present in 14% of patients hospitalized for Apapwith higher mortality.

Elevated intracranial pressure and cerebral edema are leading predictors of poor outcomes and mortality in patients with head trauma, intracranial hemorrhages, or acute ischemic strokes. While hypertonic saline (HTS) is the mainstay of treatment, recent trials in critically ill populations have demonstrated a reduction in kidney related adverse events with the use of balanced crystalloid groups when compared to 0.9% sodium chloride (NaCl). The purpose of this study is to assess adverse kidney outcomes and risk of in-hospital mortality associated with HTS in a neurocritical care population.

A retrospective cohort study was conducted at a large academic medical center on 112 adult patients in the neurosciences ICU who received 3% NaCl and/or 23.4% NaCl from July 1, 2016 to July 31, 2018. The primary endpoint was Major Adverse Kidney Events (MAKE-30), defined as at least one component of the composite: in-hospital mortality, receipt of new renal-replacement therapy, or persistent renal ays. Baseline characteristics, indication for HTS, pertinent lab values including changes in serum electrolyte concentrations, total HTS volume and associated sodium and chloride milliequivalents, and patient outcomes were collected. Statistical analysis was performed using SPSS software.

In the chloride increase > 5 mmol/L group, 22 patients (23.2%) experienced the primary outcome of MAKE-30, 19 patients (20.0%) experienced in-hospital mortality and 4 patients (25.0%) experienced AKI primary outcome of MAKE-30, and 3 patients (17.6%) experienced in-hospital mortality (P= 0.62).

The primary outcome occurred more often in the chloride increase > 5 mmol/L group and in-hospital mortality accounted for the majority of the outcome in both groups. This was not statistically significant due to the sample size and unbalanced comparator groups.

Social media has been shown to be a valuable tool to improve knowledge, attitudes, and skills. It has been theorized that the success of medical education through social media can be contributed to increased learner engagement, real-time feedback, and enhanced collaboration. We hypothesize that social media is underutilized in critical care medicine in comparison to other specialty fields of medicine and surgery.

A list of medical specialties as hashtags were run through "Hashtagify" software. This software crossreferences up to 500,000 data points on Instagram and Twitter and assigns a "popularity score" for certain topics. The phrase "critical care" was cross-referenced through a database of medical news run by Doximity over a month in comparison to other topic tags.

In total, 1420 articles concerning the topic "critical care" were posted on Doximity News over 30 days. In comparison, there were 54 articles posted under "cardiology," 460 under "internal medicine," and 600 under "emergency medicine." With respect to hashtag utilization on social media, critical care was under-represented, with a popularity score of 38. This was in comparison to other specialties such as neurology (49), dermatology (49), emergency medicine (48), and ophthalmology (44). Within the critical care hashtag, the major influencers were those representing critical care nursing.

Despite the large amount of news pertaining to critical care on professionally-curated forums such as Doximity, there is significant under-representation in social media. Within the hashtag, "critical care," the major influencers represented critical care nursing suggesting that critical care physicians are even further underrepresented. This is in line with previous research suggesting the underrepresentation of medical doctors in social media. Given that social media has been shown to be a valuable tool in enhancing medical education, we believe that a greater effort should be made to engage critical care physicians on social media outlets.

There is a call for increased diversity in national and international annual meeting participation in terms of attendance, committee participation, leadership, awards and speakers. The Neurocritical Care Society Annual meeting(NCS-AM) speaker qualifications are not specified in the bylaws. The speakership patterns of the NCS-AM have not been examined. We described the speakership patterns in NCS across a 3-year time span (2016) (2017) (2018) and delineated the trends of United States-Neurocritical-Care-fellowship-

Longitudinal cohort study. The NCS-AM conference program, a readily available online document, for the years 2016-2018, were reviewed by the study authors. Speakers were identified from the conference program. Our primary outcome was the trend of speaker characteristics across the 3-year time span. Our secondary outcome was to determine speakership trends among United States-Neurocritical-Care-fellowshipinstitution of employment at the time of the meeting.

A total of 422 speakers were included in this study, of which 57% were male. Majority of the speakers were US-based(88%), mid-to late-career (60%) and were physicians (69%). The speakers were 12±9 years from fellowship. In 3-years, there was an increased trend towards international, non-physician and early-career speakers' trained from Johns Hopkins University (JHU) (47,19%), Massachusetts General Hospital (MGH) (44,17%) and Cornell/Columbia University (22,8%); while the most common sites of employment at the time of the meeting were JHU (15,6%), MGH (15,6%) and University of Pittsburgh Medical Center (14,6%).

This is the first study to evaluate speakership trends across a 3-year period of the NCS-AM. Diversity has ble institutional bias are unclear and deserves to be studied further to better define speaker selection in the NCS annual meeting. These data may also be utilized to explore opportunities for collaboration and diversity in future NCS-AMs.

Urinary tract infections (UTIs) are the fourth most common type of healthcare-associated infection, primarily caused by instrumentation of the urinary tract. There is a 3%-7% increased risk of patients acquiring a catheter-associated urinary tract infection (CAUTI) for each day an indwelling urinary catheter (IUC) remains in place. In critically ill patients, IUC placement is often required for precise urine output measurement. Subarachnoid hemorrhage (SAH) patients often require IUC's during the cerebral vasospasm period (i.e. post-bleed day, PBD 3-12) to maintain euvolemia. This places SAH patients at increased risk for developing a CAUTI. In our local neurosciences intensive care unit (NSICU), an infection control team observed higher CAUTI rates as compared to the hospital and national average necessitating changing our urinary catheter utilization policy. We report change in practice pattern with implementation of new unit policy

The Intermittent Catheterization (IC) Algorithm includes clinician review of the patient's total intake and output and current clinical status. Retrospective chart review of CAUTI incidence (rate per 1000 catheter days) and device utilization ratio (no. urinary catheter days/ no. patient days) 12 months before and after implementation of the new policy. Time periods were compared using appropriate statistical tests pre-and post-intervention

The IC algorithm was implemented to reduce IUC utilization rate with aim to reduce CAUTI rates. The time periods studied were May 2017 to April 2018 (pre-intervention period) and May 2018 to April 2019 (post-intervention period). CAUTI rates decreased from 4.5±3.4 during the former time-period to 0.42±1.45 during the latter time period (p=0.0081). Similarly, device utilization ratio decreased from 0.8±0.08 to 0.53±0.09 (p<0.00001). In addition, use of female and male external catheter devices were encouraged leading to increased utilization

Systemic team based implementation of policies can result in adoption of positive practices and reduce hospital acquired infectious complications.

Managing neurological emergencies, particularly overnight, is very challenging for Neurology trainees at the beginning of their residency. Preparation is key to ensure residents have the skills, confidence, and knowledge to manage acute scenarios.

We developed a one-week immersive bootcamp to educate new Neurology residents about neurological emergencies prior to the start of the academic year.

The bootcamp includes the fourteen Emergency Neurological Life Support (ENLS) modules designed by the Neurocritical Care Society, thirteen faculty-created didactics, nine case-based discussions, and four resident-created simulations. The bootcamp teaches residents about the management of acute ischemic stroke, acute non-traumatic weakness, anoxic brain injury, coma and brain death, intracranial hemorrhage, intracranial hypertension, meningitis, neuromuscular emergencies, status epilepticus, spinal cord emergencies, subarachnoid hemorrhage and traumatic brain injury. Residents are also taught about communication with families during and after neurologic emergencies in a didactic session on breaking bad news.

It is important for all Neurology residents to be adept at managing neurological emergencies. However, having these skills is particularly important for residents in a military program, as residents in the military may ultimately be deployed overseas or stationed at facilities with minimal support, responsible for handling all neurological emergencies, regardless of their sub-specialty. ENLS training and didactics teach residents about the fundamentals of neurological emergencies. Case-based discussions provide residents to act out the way they would utilize this knowledge in a risk-free environment that is translatable to acute clinical situations. The combination of ENLS training, didactics, case-based discussions and simulations into a one-week immersive bootcamp early in residency should, therefore, provide a solid knowledge base about management of neurological emergencies for incoming Neurology residents and allow them to consolidate that knowledge leading to safe and effective management of neurological emergencies.

Trends and Predictors of In-hospital Mortality for Status Epilepticus: National Inpatient Sample Study

Head or Heart: Ictal Bradycardia and Temporal Lobe Epilepsy Julia Bevilacqua

Higher DAI Grade Correlates with Worse Short Term Outcome in Pediatric Traumatic Brain Injury Anna Janas; Scott Hamilton; Zachary Threlkeld; Max Wintermark

Post-intensive care syndrome amongst families of ICU patients, including post-traumatic stress disorder (PTSD), is highly prevalent after patient discharge but understudied. The psychological model of "attachment theory" describes how people respond when being separated from loved ones; various "attachment styles" have been associated with the development of PTSD in other settings. Adults can be "secure" (comfortable depending on others and being alone) or "insecure." The hypothesis of this exploratory study was that insecure family members of Neuro ICU patients would be more likely to report PTSD six months after patient hospitalization compared to secure family members.

Eligible participants were 111 family members of Neuro ICU patients at a single center who already had attachment styles (secure vs. insecure) defined via a standard survey, the Relationship Questionnaire, during an earlier study in 2018. Over 2018-2019, these subjects were asked by mail to complete the Impact of Events Scale-Revised (IES-R) six months following discharge or patient death. Participants were considered to have PTSD if IES-

41/111 returned a completed IES-R (37.0%). 27 (65.9%) of these subjects reported a secure attachment style vs. 1 out of 14 (14.2%) insecure respondents (p=1.0).

This small study did not show a significant difference in rates of post-discharge PTSD amongst Neuro ICU family members with secure vs. insecure attachment styles, however was only powered to discover a large difference between groups and the rate of PTSD in our population was markedly lower than sible association in larger cohorts with an overall higher prevalence of post-discharge PTSD would be insightful.