key: cord-006880-9dgmdtj8
authors: nan
title: Neurocritical Care Society 10th Annual Meeting: October 4 - 7, 2012 Sheraton Denver Downtown Hotel Denver, Colorado
date: 2012-09-19
journal: Neurocrit Care
DOI: 10.1007/s12028-012-9775-0
sha: 
doc_id: 6880
cord_uid: 9dgmdtj8

nan

AHRQ guidelines for venous thromboembolism (VTE) prophylaxis recommend risk stratification of patients and tailoring prophylaxis to that risk. While anticoagulation is a mainstay of optimal VTE prophylaxis after trauma, little data exists to determine when TBI patients warranting neurosurgical intervention become candidates for such treatment. Our group sought to determine the natural evolution of intracranial hemorrhage in these high risk patients and identify factors contributing to early radiographic stabilization.

All TBI patients undergoing craniotomy and/or intracranial monitoring and surviving at least 24 hours were followed prospectively from Feb 2010 to Nov 2011. Radiographic stabilization was defined as the time between injury and the final CT scan that showed no worsening during the hospital stay. Kaplan Meier (KM) curves were used to compare time to stabilization by type of intervention. Binary logistic regression was used to identify covariates contributing to stabilization within 72 hours of injury.

For the overall cohort (n=127), KM curves showed no difference in time to radiographic stabilization by type of neurosurgical intervention. Significant associations were found between stabilization at 72 hours and higher presenting GCS (OR:22.8, 95%CI 2.9-180.4), younger age (OR:0.96, 95%CI 0.92-0.99), and male gender (OR:13.2, 95%CI 2.1-84.3). Subjects with a presenting GCS of >11 (n=32) had an 81% PPV for radiographic stabilization by 72 hours after injury. The AUC for the logistic regression model was 0.929.

Sentinel headache refers to discrete thunderclap headache in the weeks preceding hospital admission for SAH. A large proportion of these events are thought to represent aneurysmal bleeding events. Repeat hemorrhages have been found to increase the extent of vasospasm in experimental models, and are often assumed to increase the risk of delayed cerebral ischemia (DCI) in humans (the "double bleed effect").

Cerebral Performance Category (CPC) is a standard outcome measure after cardiac arrest, but has limited ability to discriminate between mild and moderate brain injury. We hypothesized that many cardiac arrest survivors with good CPC scores would have significant deficits on blinded neurocognitive testing.

Patients initially comatose after cardiac arrest treated who awoke after therapeutic hypothermia (TH) were evaluated by a neuropsychologist prior to hospital discharge with the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS), a well-validated tool that assesses function in multiple domains compared to standardized normal values.

18 patients admitted between Nov 2011 and May 2012 awoke after TH, 9 completed the RBANS evaluation after leaving the ICU and ready for discharge. Median age was 59 yrs (range 47-76), 100% male, 7 had initial rhythm VT/VF, median time to ROSC was 10 minutes (range 1-88). Seven patients had a CPC of 1, 1 patient had a CPC of 2, and 1 patient had a CPC of 3. Seven patients were discharged home and 2 to acute rehab. Attention and delayed memory were severely abnormal half of the patients (below 5 th percentile), language and visuospatial domains were affected less often in 22% of the patients (below 5 th percentile). On cumulative scores of all domains, all patients scored below the 16 th percentile compared to age and education adjusted scores, regardless of CPC score.

Cardiac arrest survivors with CPC scores considered 'good' frequently had severely abnormal neurocognitive function just prior to hospital discharge. The cognitive domains most frequently affected were attention and delayed memory. More sophisticated testing with tools such as RBANS may better identify components of cognitive dysfunction after cardiac arrest which may be targets for additional therapeutic intervention and be a more meaningful tool for long-term follow-up studies.

Introduction Quantitative brain diffusion-weighted imaging (DWI) MRI may help predicting the degree of functional recovery in patients ain volume with an apparent diffusion coefficient (ADC) <450x10 -6 mm 2 /sec differentiated between cardiac arrest survivors who regained an independent lifestyle and those with impaired functional outcome. We aimed to validate this threshold in an external dataset.

DWI MRIs of comatose post-cardiac arrest patients were obtained between 25-120 hours post-arrest. Survivors who regained consciousness by day 14 were assigned to one of two recovery groups: good recovery (discharged to home) and impaired recovery (discharged to a skilled nursing facility, rehabilitation facility or another hospital). The quantitative DWI data were obtained blinded to patient outcomes. The brain masks were semi-automatically created on the b0 images using Medical Image Processing, Analysis and Visualization program (MIPAV). The ADC values of each voxel within the brain were determined.

Data of 111 patients from five US centers (Columbia, MGH, Mayo Clinic Jacksonville, Northwestern, and Stanford) with adequate MRIs were analyzed. Of these, 30 (27%) patients regained consciousness and survived to discharge: mean age 55±18 years, 47% female, arrest duration 18±13 minutes, 73% of patients received therapeutic hypothermia, MRI obtained at 69±24 hours post-arrest. The median (IQR) percentage of brain volume with ADC<450x10 -6 mm 2 /sec was 0.02% (0.01-0.2) in patients with good recovery (N=9) and 0.2% (0.1-0.4) in patients with impaired recovery (p=0.03). An ADC<450x10 -6 mm 2 -99) sensitive and 24% (95%CI 9-48) specific for good recovery.

The results of this validation study support earlier findings that quantitative DWI MRI in comatose post-cardiac arrest patients is a sensitive prognostic test to predict the degree of functional recovery in post-cardiac arrest survivors.

According to the Universal Determination of Death Act, death in the United States is determined in accordance with accepted medical standards, which can be national, regional, or local. As a result, significant variability in brain death (BD) determination has been reported among the best hospitals across the country. We tested the hypothesis that similar variability exists in individual States, such as Michigan.

Michigan Health and Hospital Association and Gift of Life of Michigan (the local Organ Procurement Organization) databases were reviewed for hospital BD policies. Only hospitals with > 50 beds and an Intensive Care Unit were included. Several BD determination process variables were extracted and analyzed with descriptive statistics.

Results 61/66 hospitals had BD policies, 1 did not and in 4 it was unclear. Ten different combinations of physicians allowed to perform the exam were included. In 8.5% there were no prerequisites to initiate BD and in 23.7% no established cause mentioned. Ten different temperatures to initiate BD exam were required. Five different arterial blood carbon dioxide levels to establish positive apnea test were cited. A single BD exam was requested in 23.3% of policies, a dual in 63.3% and a single or dual in 13.3%. Confirmatory tests were optional (51%), recommended (8.5%) or mandatory (20%). Electroencephalogram was the most common confirmatory test (90%) and CT angiogram the least common (7%).

We report significant variability in the BD hospital policies in Michigan despite published guidelines from the American Academy of Neurology. If one accounts for additional variability in the strict implementation of these policies at the bedside level, the urgency for a uniform State-wide BD policy becomes even more obvious. 

Intrathoracic Pressure Regulation (IPR) therapy is a novel therapy that non-invasively modulates pleural pressures to take advantage of the physiological benefits that occur by creating pressure differentials in the thorax. After each positive pressure breath IPR lowers intrathoracic pressure to subatmospheric levels relative to the rest of the body. This intervention enhances cardiac preload and output and decreases intracranial pressure (ICP). We hypothesized that IPR therapy which has been previously shown to increase calculated cerebral perfusion pressures would also increase cerebral blood flow (CBF) in a porcine model of elevated ICP.

In this pilot study, four isofluorane anesthetized pigs (29.8 ± 1.2 kg) were subjected to a focal brain injury by epidural insertion of an 8 French Foley catheter into the left hemisphere which was slowly filled with saline to simulate a traumatic brain injury with elevated ICP. In the right hemisphere, a thermal diffusion probe was used to measure CBF (Hemedex, Inc., Cambridge, Massachusetts) while a Millar catheter was used to measure ICP. Once a stable elevated ICP was confirmed, IPR therapy was applied at a level of -12 cmH 2 O for 30 minutes. End tidal CO 2 was held constant at 40 mmHg by adjusting the respiratory rate during IPR use.

TBI is a major risk factor for the development of Alzheimer's disease (AD). In previous animal and human studies, an increase in the expression of amyloid precursor protein (APP) after TBI was found to correlate with the disruption of neuronal activity, beta-amyloid plaque formation, cognitive decline, and even death.To date, no interventions used at decreasing amyloid plaque load after TBI have been identified.

In this study, using the controlled cortical impact device we produced a severe head injury in 4 month old 5XFAD mice. At 30 minutes and 12 hours after injury, the 5XFAD mice were treated intraperitoneally with either placebo or resveratrol (anti-oxidant; 100 mg/kg). At 1 month after injury, the animals were intracardially perfused with 0.9% saline followed by 10% phosphate-buffered formalin. The whole brain was removed, sliced, and stained for beta-amyloid levels using immunohistochemistry. In addition, TUNEL+ cells were measured at the indicated time-points to determine the level of neural injury.

In this study we found that treatment with resveratrol at 30 minutes and 12 hours post-injury resulted in a significant reduction in beta-amyloid plaque load near the injury zone (parietal cortex) (p<0.01) and hippocampus (p<0.05). Also, the mice treated with resveratrol had reduced (p<0.002) TUNEL+ staining.

While a multitude of etiologies may lead to coma, treatments for coma remain elusive. The hypothalamic Orexin pathway, critical in sleep/wake cycles, can stimulate multiple areas of the brain and provides a potential pharmacologic target towards improving arousal after coma. We used a post-cardiac arrest (CA) rodent coma model to assess whether postresuscitative Orexin-A intracerebroventricular (icv) infusion after bolus injection would provide immediate and long term arousal after CA.

Seventeen adult Wistar Rats (Male, 300-350gms) were implanted with a icv cannula attached to an osmotic pump. One week later, rats underwent baseline EEG followed by 7-minute asphyxial CA. Forty-five minutes after resuscitation, rats were randomized to either Orexin-A (n=8) or saline (n=9) icv bolus and infusion. EEG was monitored continuously for 4 hours after CA, and for 30 minutes at 24hrs, 48hrs, 72hrs, and 12 days post-CA. Behavioral testing (neurologic deficit scale; NDS) was also conducted at these times. EEG was quantitatively analyzed using Information Quantity (IQ), an entropy based nonlinear previously established by us.

Rats receiving Orexin-A almost immediately exhibited higher IQ when compared to saline (0.731±0.028 vs. 0.613±0.021; p<0.01). This acute improvement in IQ appeared with slowest sub-bands (e.g.ð) improving first followed progressively by faster sub--CA. Moreover, Orexin--band at 72hrs (1.160±0.039 -band at 72 hrs (0.911±0.032 vs 1.023±0.041; p<0.05). Behaviorally, Orexin-A allowed rats to perform significantly better on the NDS at 4hrs (38.3±3.2 vs. 30.1±1.9; p<0.05); 24hrs (67.1±4.1 vs. 49.8±3.8; p<0.01); 48hrs (73.4±4.3 vs. 61.0±3.6; p<0.05), and 72hrs (75.0±3.3 vs. 65.2±2.9; p<0.05).

Heart rate variability (HRV) characteristics have been associated with outcome after traumatic brain injury. We sought to determine if HRV characteristics in the first 72 hours after subarachnoid hemorrhage (SAH) are associated with hospital morbidity and mortality.

Continuous EKGs recorded (240Hz sampling) during the first 72 hours post-SAH was analyzed in 323 of 449 consecutively admitted patients between 2006 and 2011. Admission clinical scores, radiographic, surgical, ventilation and

The Pan-Tompkins algorithm was applied to identify the QRS complex. FFT calculations were generated for the following -0.4Hz), low frequency (LF: 0.04-0.15Hz), very low frequency (VLF: 0.003-nerated Sample entropy and 1/f --minute (FFT<0.04Hz), or 60-minute

Individual multivariable logistic regression analyses of hospital morbidity and mortality controlling for admission Hunt and Hess grade, Apache II Physiological sub-score, age, and mechanical ventilation status were conducted. 

Dialysis disequilibrium syndrome (DDS) is characterized by varying central nervous system manifestations secondary to cerebral edema that most often occurs after the first round of hemodialysis (HD). Literature suggests that underlying brain injury may predispose patients to the development of DDS. However, the pathophysiology has yet to be elucidated. Herniation from HD is thought to be exceedingly rare with current dialysis methods and has not been reported in the era of modern neurointensive care.

We present a case series of three patients with acute neurological injury undergoing HD in the intensive care unit that rapidly developed fatal brain edema, secondary to DDS, even after several previous uneventful rounds of HD.

Three patients, ages 17, 45 and 51 years, with traumatic brain injury, hypertensive intracerebral hemorrhage, and ischemic stroke underwent HD in the intensive care unit. The number of dialysis sessions prior to the development of DDS was 1, 3 and 5. All three patients developed clinical signs of herniation within minutes to hours of HD. CT scans showed global cerebral edema with both transtentorial and tonsillar herniation. Aggressive osmotherapy with mannitol and supersalt were ineffective in reversing the massive edema and all three patients died. Two of the patients had a significant reduction of the BUN (65% and 45%) while the third had only a modest reduction.

Our case series illustrates the potential dangers of HD in patients with acute neurological injury who have a high potential for worsening cerebral edema. It also reaffirms that DDS with fatal cerebral edema can occur even after several rounds of HD and with current HD techniques. Utilization of continuous veno-venous hemofiltration instead of HD may prevent the rapid shifts of osmoles and prove safer in neurologically injured patients.

Traumatic coma is believed to be caused by disruption of the ascending reticular activating system (ARAS), a complex network of arousal pathways projecting from the brainstem to the hypothalamus, thalamus, and basal forebrain. There is a critical lack of diagnostic tools for detecting which components of the ARAS network are disrupted in traumatic coma. We aimed to determine whether an advanced MRI technique, high angular resolution diffusion imaging (HARDI), can detect disruptions in the brainstem arousal network that are implicated in the pathogenesis of traumatic coma.

We used HARDI tractography to analyze neural network connectivity in two postmortem brains: one from a 62-year-old woman who died three days after traumatic coma, and one from a 53-year-old woman who died of non-neurological causes. Both specimens were scanned as dissected blocks of the brainstem, hypothalamus, thalamus, and basal forebrain on a small-bore, high field (4.7 Tesla) MRI scanner. HARDI tractography analyses were performed to compare the structural integrity of each component pathway of the ARAS network in the traumatic coma and control specimens. Upon completion of imaging, both specimens were sectioned and stained for correlative histopathological analysis.

HARDI tractography revealed that specific components of the ARAS network, including known cholinergic, glutamatergic and noradrenergic projections connecting the brainstem to the thalamus and basal forebrain, were severely disrupted in the traumatic coma specimen, as compared to the normal specimen. These disruptions were consistent with histopathological tissue tears and axonal swellings. By contrast, connectivity between the brainstem and hypothalamus, and within the thalamus itself, was partially preserved in the traumatic coma specimen.

HARDI tractography can detect disruptions in specific components of the ARAS network that are implicated in the pathogenesis of traumatic coma. This advanced imaging technique may be used to elucidate the neuroanatomic basis of coma in individual patients. 

Refractory intracranial hypertension (RICH) is associated with high mortality rates and is the final pathway of many neurocritical entities, such as severe traumatic brain injury (sTBI). Objective: To determine modifications in intracranial pressure (ICP) and cerebral perfusion pressure (CPP) following indomethacin (INDO) infusion after RICH secondary to sTBI.

INDO was administered in a loading dose (0.8 mg/kg/15 minutes), followed by continuous infusion (0.5 mg/kg/h) in patients with ICP>20 mmHg for more than 20 minutes who did not respond to first line therapies. Changes in ICP and CPP were observed. Clinical outcome was assessed at 30-day according to Glasgow Outcome Scale (GOS). Analysis of INDO safety profile was also conducted. Differences in ICP and CPP values were assessed using repeated-measures ANOVA with an a-level of P<0.05

Twenty-nine consecutive sTBI patients (26 men and 3 women) with a mean age ±SD 36±17 years wereincluded. Median posresucitation GCS score at admission was 6 (IQR: 4-7) with a predominance of grade IV in Marshall CT classification. 

Our findings support the effective and feasibility of INDO in reducing ICP and improving CCP in RICH patients. Future studies to evaluate different doses, lengths of infusion and longer-term effects together with effects on outcome are needed.

Hematoma expansion after acute intracerebral hemorrhage (ICH) occurs most frequently in patients presenting within 3 hours of symptom onset. Therefore, most investigational therapies have been tested only in patients presenting ultra-early in their disease course. However, the majority of ICH patients present outside this time window or with an unknown time of onset. We investigated the prevalence of hematoma expansion in these patients with delayed presentation and assessed the accuracy of the CT angiography (CTA) spot sign for identifying risk of hematoma expansion.

We performed a prospective cohort study. 391 consecutive ICH patients undergoing CTA and follow-up head CT were enrolled over ten years. CTA spot sign readings were performed by two experienced readers and hematoma expansion was assessed using semi-automated software. Expansion was defined as an increase in volume of >6 mL or an increase of >33% from baseline ICH volume.

Hematoma expansion occurred in 18% of patients. When stratified by time from symptom onset to initial CT, hematoma expansion rates were: 39% within 3 hours; 11% between 3-6 hours, 11% beyond 6 hours (but with known onset), and 20% in patients with an unknown symptom onset time. Of patients who developed hematoma expansion, only 38% presented within 3 hours. The accuracy of the spot sign in predicting hematoma expansion was 0.67 for patients presenting within 3 hours, 0.83 between 3 to 6 hours, 0.88 after 6 hours and 0.76 for patients presenting with an unknown onset time.

A substantial number of patients destined to suffer from hematoma expansion present either late or with an unknown time of symptom onset. The CTA spot sign accurately identifies patients destined to expand regardless of time from symptom onset, and may therefore open a path to offer clinical trials and novel therapies to the many patients who do not present acutely.

Intraventricular fibrinolysis has been shown effective in clearing intraventricular hemorrhage (IVH) in small series of patients. We present our experience with using fibrinolytics over 12 years.

Retrospective analysis of prospectively collected data of patients with IVH admitted to two Neuro-ICUs and treated with rt-PA instillation (one patient with tenecteplase) via intraventricular catheter (IVC) until the 3 rd and 4th ventricles were cleared of blood. All patients were treated by the same neurointensivist with the same instillation protocol but different doses of drug based on individual patient characteristics. The Graeb and LeRoux semi-quantitative scales were used to measure the amount of IVH before and after the last dose.

55 patients (mean age 58.6 years, 58.2% male) were admitted with a median GCS 8. Thirty-one had intracerebral and 17 aneurysmal subarachnoid hemorrhage, 4 brain tumors, 1 head trauma, 1 arteriovenous malformation.and 1 primary IVH. t-PA was administered at a total dose of 5.1±4.7mg (individual doses ranging between 0.5 to 4 mg), with 1 st dose 57.1±55.7 hours from admission and for a duration of 3.2±2 days. The pre-fibrinolysis Graeb and LeRoux scores were 8.1±1.8 and 11.3±2.5 and decreased post fibrinolysis to 4.6±1.8 and 5.3±2.6 (p<0.001). A significant correlation between total fibrinolytic dose and difference in pre-post amount of IVH was found for the LeRoux scale (Pearson 0.413, p <0.002). Three patients had small tract bleeds (<3cc, with one bleeding profusely at the incision site requiring transfusion) and one had extension of ICH in the upper midbrain. No patient developed ventriculitis. The total dose of t-PA was lower in 11 patients who received shunt, compared to 44 who did not (3.3±2.5 vs 5.6±5, p<0.039). Eighteen (32.7%) patients had -up of 27.2±27.3 days.

In our large series of patients, intraventricular fibrinolysis significantly decreased IVH with minimal complications.

Distinguished Poster 10 ___________________________________________________________________________________ 

White matter lesions significantly impact on outcome after aneurysmal subarachnoid hemorrhage (aSAH). Brain extracellular tau is indicative for axonal injury, associated with poor neurological outcome after severe traumatic brain injury, however has not been elucidated so far in patients with aSAH.

Twenty-five consecutive aSAH patients monitored with cerebral microdialysis (CMD) and brain tissue oxygen tension (P b tO 2 ) were included. CMD total tau, phospho-tau-181 and beta-amyloid 1--42) levels were analyzed at a 12hours interval until d3 and 24-hours interval until d9 using an ELISA-technique (Innogenetics). Statistical analysis was performed with non-parametric tests and a mixed effects model as appropriate.

Median age was 56y (47-67y) and admission Hunt&Hess grade ranged from 3 to 5. CMD-tau, phospho-tau--42 were detectable in all patients. Probe location in perilesional tissue revealed a higher overall CMD-tau level (P<0.05) -42 and phospho-tau. CMD-tau positively correlated with CMD-lactate (r=0.59, P<0.001). Brain hypoxic (P b tO 2 were associated with increased CMD-tau levels (P<0.05). No correlation was found between other variables besides a higher phospho-tau level and CMD samples categorized as brain hypoxic hyperlactatemia. Patients with poor outcome -tau level during hospital-course (P<0.05) but no difference in phospho-tau--42 (adjusted for disease severity).

Cerebral tau is elevated after aSAH and associated with perilesional probe location and poor 3-months functional outcome. Association with brain-morphological abnormalities and neuropsychological deficits need further investigations.

Neurocrit Care (2012) 17:S1-S337 

To date only two studies have evaluated anemia status in acute intracerebral hemorrhage [ICH] . On admission anemia [OAA] was associated with larger hematoma volume and lower hemoglobin levels during hospital stay were related to poorer outcome. It remains unknown whether anemia impacts outcome primarily through its effects on ICH volume or itself has independent effects.

This retrospective analysis included 174 consecutive patients with spontaneous supratentorial intracerebral hemorrhage. Clinical data including the pre-admission-status, neuroradiological, initial presentation, treatment, and outcome were evaluated through institutional databases, patient's medical charts and by mailed questionnaires. Multivariate logistic and graphical regression analyses were calculated to evaluate associations of OAA with functional outcome and to determine independent effects of OAA.

OAA was associated with larger ICH volume (29.9cm³ versus 13.6cm³, p=0.001), greater extent of intraventricular hemorrhage [IVH] (p=0.044) and poorer neurological status on admission (p<0.001). Further, OAA showed a true positive and accurate association with larger hemorrhage volumes (ROC: p=0.001,AUC>0.7). Multivariately, for all patients despite age, only OAA could be elucidated as independent predictor of unfavorable functional outcome (mRS > 3) at 90 days (OR=3.179;p=0.0435). Comparison of separate multivariate models revealed: for OAA-patients no independent predictor could be identified, whereas in non-OAA patients ICH volume demonstrated known independent effects on functional outcome (OR 1.05;p=0.031).

Within this study OAA was shown to be a significant predictor of an unfavorable functional outcome and has independent effects beyond its accurate association with larger hemorrhage volumes. OAA appears to be a very relevant and previously unrecognised predictor of functional outcome at 90 days. The recognition of anemia and its treatment could possibly open up new therapeutic avenues to decrease the rate of functionally dependent patients after ICH. This strongly supports the need of prospective interventional studies to evaluate the influence of anemia in patients with intracerebral hemorrhage.

In patients with suspected subarachnoid hemorrhage and negative brain imaging, lumbar puncture is recommended. This test is frequently complicated by false-positive results due to a traumatic tap. We hypothesized that blood precipitating in the thecal sac following non-traumatic subarachnoid hemorrhage would be visible on MRI.

A prospective database for subarachnoid hemorrhage was searched for patients who received MR lumbosacral spine imaging during admission for subarachnoid hemorrhage. Electronic chart review was completed. All MR studies were read and interpreted by a neuroradiologist.

Patients (n=10) with subarachnoid hemorrhage underwent delayed MRI imaging of the lumbosacral spine an average of 10 days (range 1-21 days) after the onset of symptoms. The median Hunt-Hess grade for this cohort was 3 (range 2-5). The median Fisher Grade was 3. Blood precipitating in the thecal sac was visible in 8 out of 10 patients (80%). The density of blood compared to CSF was hyper-intense on T1 (bright) and hypo-intense on T2 (dark). The blood was most evident at L5 and S1 levels and layered in a dependent fashion. Delayed CT head non-contrast obtained at the time of the MRI LS spine demonstrated resolution of subarachnoid hemorrhage in 9/10 patients and a small amount of isodense intraventricular hemorrhage layering in the occipital horns was detected in 4/10 patients.

Delayed neuroimaging with CT head after subarachnoid hemorrhage has a high false negative rate. MR imaging of the lumbosacral spine detected persistent blood products settling in the thecal sac despite clearance of subarachnoid blood on CT head imaging. MR lumbosacral spine imaging could serve as a 'virtual lumbar puncture' in patients with suspected subarachnoid hemorrhage.

Stroke patients receiving IV tPA can be admitted to an ICU or a stroke unit (SU) but SU admission may be more costefficient. We compared ICU admission vs SU admission in tPA-treated patients.

During the initial 3 years of this retrospective study, patients were admitted to the ICU as we lacked a SU. In the following 2 years, patients were admitted to a new SU. Demographics, medical history, NIHSS, treatment interventions, neurologic and medical complications, and mortality were collected to determine if ICU admission resulted in better outcome and less complications. Categorical variables were analyzed with Fishers exact test and continuous ones with proportion of the means test (t-test).

We compared 71 ICU admissions and 104 SU admissions. ICU admission included 54% males and SU admission included 59% males (NS). Median age for ICU and SU admission was 69 and 70 respectively (NS). Admission NIHSS was 11 for ICU patients and 9 for SU patients (NS). The median length of stay in the ICU was 1 day (as per protocol) and the median SU length of stay was 4 days. Intravenous anti-hypertensives (bolus) were used in 30% of ICU patients and in 18% of SU patients (p= 0.09) and continuous infusions in 11% of ICU patients and 6% of SU patients (p=0.04). Initial NIHSS scale of > 15 predicted need for mechanical ventilation (p=0.002). Intracranial hemorrhage occurred in 3% of ICU and 3% of SU patients (NS). Complications (pneumonia, venous thromboembolism, sepsis, or death ) did not differ. There was no difference in the proportion of patients with mRS of 1 or less in the two groups (38% vs 36%). Admission to the SU resulted in savings of $ 70,000 per 100 patients/day.

Patients receiving tPA can be safely admitted to a SU resulting in significant cost savings. Patients with NIHSS > 15 are likely to need ICU admission for mechanical ventilation.

Stroke patients with dysphagia have a high incidence of aspiration, which may lead to pneumonia. Evidence suggests that ACE inhibitor use may decrease the risk of pneumonia via their inhibitory effects on substance P degradation. The objective of this study was to investigate the association between ACE inhibitor use and the development of pneumonia in hospitalized stroke patients.

A retrospective case-control analysis was performed. Eligible patients (N=1927) were individuals admitted to Saint Louis University Hospital with a diagnosis of acute ischemic stroke, spontaneous intracerebral hemorrhage, or non-traumatic subarachnoid hemorrhage between March 1 st , 2009 and November 30 th , 2011. Patients greater than 89 years of age, who died or were discharged to hospice within 5 days of admission, or who had a baseli excluded. Cases were patients with an ICD-9 code for pneumonia or antibiotic treatment course for at least 7 days with a positive respiratory culture. Controls were patients without pneumonia matched using primary diagnosis, baseline demographics, history of prior stroke, diabetes, hypertension, heart failure, and initial NIHSS scores. ACE inhibitor use, length of stay, discharge disposition, and other pertinent data were collected and analyzed using descriptive statistics, chisquare, and logistic regression.

There is growing evidence supporting the role of inflammation in aneurysmal subarachnoid hemorrhage (aSAH) pathophysiology and it is of great interest to elucidate which immune mechanisms are involved.

Methods 9 aSAH patients (SAHP) and 28 healthy control subjects (CS) were enrolled prospectively. The protocol was authorized by the Ethics Committee of our hospital and all subjects (or patient next of kin) signed an informed consent. The median age of SAH patients was 48 years (34-67) and of control subject was 45 years (25-76). We assessed leukocytes subpopulations and their activation status by multiparametric flow cytometry in cerebrospinal fluid (CSF) and peripheral blood (PB) of SAHP at the same time and in PB of CS.

We found an increase in CD16+-monocytes percentage (p=0.035) in CSF compared with PB in SAHP and a decrease in PB of SAHP compared with CS (p=0.016). SAHP also showed a marked increase in the expression of CD69 (activation antigen) in PB CD4+T cells compared with CS (p=0.002). Additionally, CSF CD4+T cells showed a decreased expression of CD4 (p=0.0001) and CD3 (p=0.008) (activation markers) compared to PB CD4+T cells in SAHP. Similarly, PB CD8+T cells in SAHP showed an increased expression of CD69 compared with CD8+T cells of CS (p=0.002). CSF CD8+T cells showed a decreased expression of CD3 (p=0.008) and an increased expression of CD69 compared with PB CD8+T cells (p=0.002). B and NK cells were decreased in SAHP compared with CS (p=0.0001 and p=0.0001 respectively).

As far as we know this is the first report that analyzes leukocytes subsets in CSF and PB in patients with aSAH. Our data suggests not only CSF leukocytes recruitment (from the blood) but also an increase status of activation at this level. Overall, these results indicate that aSAH probably stimulates both the innate and adaptive immune responses.

Subdural hematoma (SDH) is a common diagnosis in neurosurgical and neurocritical practice. Comprehensive outcome data and management guidelines are lacking for non-traumatic SDH. Thus, we aimed to determine factors associated with in-hospital mortality in a large sample of patients with non-traumatic SDH.

Using the Nationwide Inpatient Sample, we included adults with a primary diagnosis of acute non-traumatic SDH (ICD-9 code, 432.1) hospitalized in the United States between 2007 and 2009. Demographics, comorbidities, craniotomy treatment and discharge outcomes were identified. Univariable and multivariable analyses were performed to identify predictors of in-hospital mortality.

Of 14093 patients with non-traumatic SDH, the mean age was 71.4 (SD 14.8) with 62% male, and 22.2% admitted during the weekend. Surgical evacuation was performed in 51.4% of patients; 7.4% (16.9% of patients requiring surgical evacuation) required a second craniotomy. Death during hospitalization occurred in 11.8% of patients. Factors significantly associated with higher in-hospital mortality included increasing age, female sex, comorbidities (congestive heart failure [CHF] , coagulopathy, renal failure, liver disease), mechanical ventilation during the first 4 days (MV), premorbid warfarin use, repeated SDH evacuation, and admission during the weekend. Craniotomy was associated with decreased in-hospital mortality. In multivariable analysis, age (OR 1.02, 95% CI 1.02-1.03), female sex (OR 1.13, 95% CI 1.01-1.28), CHF (OR 1.42, 95% CI 1.22-1.68), warfarin use (OR 1.50, 95% CI 1.28-1.77), mechanical ventilation (OR 18.21, ) and weekend admission (OR 1.19, 95% CI 1.04-1.36) were independent predictors of inhospital mortality. Surgical SDH evacuation was a strong independent predictor for decreased mortality (OR 0.46, 95% CI 0.40-0.52).

One in nine patients with non-traumatic SDH dies during hospitalization. Of several predictors of mortality, the weekend effect and the decision for or against surgical evacuation are potentially modifiable factors. Further investigation may lead to improvement of management practice and better outcomes.

To determine the burden of structural damage of the central nervous system (CNS) in patients who died in the setting of non-neurological critical illness.

Critically ill patients who died in the medical, surgical or cardiac ICUs over a 10 year period and underwent autopsy were included. Patients with known CNS lesions, cardiac arrest, and those from neurological ICUs were excluded. Brain specimens were reviewed by a neuropathologist and classified according to location and lesion type (infarct, hemorrhage, inflammation).

Acute brainlesions were found in 60 of 97 patients studied. Mean GCS at admission was lower in patients with neuropathological findings (11.3 vs. 12.8; p=0.0329). The most common sites of injury were cortex (35.1%) and hippocampus (28.9%). Infarcts (56.7%), hemorrhages (13.4%), and signs of inflammation (9.3%) were the most frequent findings. Patients with septic shock and ALI/ARDS had more lesions than patients without these critical illnesses, albeit these differences were not statistically significant.

Ischemic brain injury is prevalent in patients dying from non-neurological critical illness and may occur secondary to CNS hypoperfusion. Efforts to optimize brain oxygen delivery during critical illness may be neuroprotective. 

After CA, microcirculatory reperfusion disorders develop despite adequate cerebral perfusion pressure. Increased blood viscosity strongly hampers the microcirculation resulting in plugging of the capillary bed, arteriovenous shunting and diminished tissue perfusion. The rheologic properties of blood depend on hematocrit and plasma constituents, mainly acute phase proteins. The aim of the present study was to assess blood viscosity in relation to cerebral blood flow in patients after a cardiac arrest.

We performed an observational study in 10 comatose patients after cardiac arrest. Patients were treated with mild therapeutic hypothermia for 24 hours and passively rewarmed to normothermia. Blood viscosity was measured ex-vivo at 0, 6, 12, 24, 36, 48 and 72 hours after admission using a Contraves LS300 viscometer. Mean flow velocity in the middle cerebral artery (MFV MCA ) was measured by Transcranial Doppler (TCD) at the same time points.

The median viscosity on admission was 9.12(8. 19-11.19 )mPa.s, remained stable at 9.13(7.57-10.51)mPa s and 9. 70(8.50-11.42 )mPa s at 3 and 6 hrs respectively (p=0.47). From 6 hrs after admission viscosity decreased significantly to 3.66(3.12-4.04)mPa s (p<0.001). Median MFV MCA was low (27.0(23.8-30 .5)cm/s) on admission, and significantly increased to 63.0(51.0-80.0) cm/s at 72 hrs (p <0.001). There was a significant association between the viscosity and the MFV MCA (p=0.0019). Median hematocrite was 0.41 (0.36-0.44)l/l on admission and subsequently significantly decreased to 0.32 (0.27-0.35) l/l at 72 hrs (p <0.001) In contrast, acute phase proteins such as CRP and fibrinogen increased during admission (from 2.5(2.5-6.5)mg/l to 101(65-113.3)mg/l and 2795 (2503-3565)mg/l to 6195(5843-7368)mg/l respectively (p <0.001).

Viscosity decreases in the first 3 days after cardiac arrest and is strongly associated (correlated) with an increase in cerebral blood flow. Since viscosity is a major determinant of cerebral blood flow, repeated measurements may guide therapy to restore cerebral oxygenation after cardiac arrest.

Initial hemorrhage burden is an independent predictor for delayed cerebral ischemia (DCI) in patients with aneurysmal subarachnoid hemorrhage (SAH). Among the different definitions of blood burden, cisternal plus intraventricular hemorrhage volume (CIHV) has been regarded as the most sensitive blood volume definition in predicting DCI. However, it is not clear whether clot clearance is associated with DCI.

Quantitative analysis of hemorrhage volume and clot clearance was made in consecutive 116 patients who were scanned within 24 hours from onset. Cistenal plus intraventricular hemorrhage volume (CIHV) was calculated for clot burden analysis. Serial CIHV was measured up to 7 days after SAH onset. Clot clearance was calculated up to 7 days as a percentage of residual clots compared to the initial scan. Initial clot burden and clot clearance were compared in patients with and without DCI.

Included patients were 55.5 ±15.2 years old with female preponderance (65.5%, (76/116)). DCI was developed in 34 patients (29.3%). Conventional risk factors were not different between patients with and without DCI including age, sex, HT, DM, smoking, Admission H&H scale and APACHE score. Patients with DCI had higher CIHV (22.1ml, ) compared with those without DCI (14.5 ml, IQR (8.58had higher odds for DCI (OR 4.3, 95% CI (1.3 -14.0, P = 0.015). However, clot clearance rate was not different between patients with and without DCI (Day 3: 48.2% vs. 57.8%, P = 0.12, Day 5: 77.2% vs. 75.6%, P =0.75, Day 7: 80.5% vs. 82.1 %, P= 0.43).

Quantitative clot clearance rate using CIHV is not associated with the development of DCI while initial CIHV is an independent predictor for DCI.

The majority of patients who die from subarachnoid hemorrhage have withdrawal or limitation of care and a focus on comfort at the end of life. Ethnic disparities at the end of life has been examined in general critical care settings but not specifically in brain injured patients.

Patients with aneurysmal subarachnoid hemorrhage were prospectively followed in an observational database from August 1996 to January 2011. Demographic information including ethnicity was collected from medical records and self reported by patients or their family. Significant in-hospital events including care withheld or withdrawn (comfort measures only, CMO) and mortality was recorded prospectively. Included were patients of White, Black or Hispanic race.

1255 patients were included in our analysis: 650 whites, 218 blacks and 387 hispanics. Age was the only baseline characteristic that was different between groups. Whites (56±15 years) were older than Blacks (52±14 years) and Hispanics (53±15 years). No difference in morality was seen: 18% in Whites, 19% in Blacks, 18% in Hispanics. CMO was more commonly ordered for Whites (14%) than Blacks (10%) and Hispanics (9%) (p=0.04). In multivariate analysis controlling for age and initially Hunt-Hess grade Hispanics were less likely to have CMO orders than Whites (OR, 0.6; 95%CI, 0.4-0.9; p=0.02). Of the 229 patients who died 77% of Whites had CMO orders compared to 54% of Blacks and 49% of Hispanics (p<0.01). In multivariable analysis controlling for age and Hunt-Hess, Blacks (OR, 0.3; 95%CI, 0.2-0.7; p<0.01) and Hispanics (OR, 0.3; 95%CI, 0.2-0.6; p<0.01) were less likely to die with CMO orders than Whites.

Multiple assessment measures are used to evaluate post-aneurysmal subarachnoid hemorrhage (aSAH) outcomes / complications. The use of a common measure has not been established, thus choosing which measure to control for becomes difficult when conducting multivariable analysis in clinical research. We compared odds ratio (OR) and positive predictive value (PPV) to determine measures with strongest associations with post-aSAH complications / outcomes.

Subjects (n=559) with aSAH were recruited from an ongoing study with 21measures were assessed on admission: Hunt and Hess (HH), Fisher, Claassen, Glasgow coma scale (GCS), World Federation of Neurological Surgeons (WFNS), and NIH Stroke Scale (NIHSS). Dependent variables were measured as follows: delayed cerebral ischemia (DCI) was defined as clinical deterioration due to cerebral ischemia, moderate/severe vasospasm was diagnosed using sonography/angiogram, infarction was diagnosed via head CT scan. Three and 12month outcomes were assessed by Barthel index and modified Rankin scale (MRS). Logistic regression and Spearman correlation were used.

When predicting vasospasm and DCI (controlling for age, gender, clipping/coiling), Fisher scale had the largest ORs (1.83 and 2.1), with a PPV of 61.8% and 60.9% (p<.01), respectively. When predicting infarction, HH had the largest OR (1.36) with a PPV of (51.6%); p=.02. All scales were significantly associated with poor MRS (4-6); p<.0001. For 3 and 12-month poor MRS, Fisher scale had the largest OR (3.2 and 3.1) with a PPV of 51.5% and 36.4%, respectively. Admission NIHSS had the largest correlation coefficient (-.36) with 3-month Barthel index while WFNS had the largest correlation coefficient (-.35) with 12-month Barthel index (p<.05).

Fisher scale has the strongest association with vasospasm, DCI and MRS, while HH has the strongest association with infarction. We recommend clinical studies control for Fisher when investigating vasospasm, DCI, and MRS and for HH when investigating infarction to determine independent risk factors.

To date, there has been a shortage of evidence-based quality improvement initiatives that have shown positive outcomes in the neurosurgical patient population. A single-institution prospective intervention trial with continuous feedback was conducted to investigate the implementation of a urinary tract infection (UTI) prevention bundle to decrease the catheterassociated UTI rate.

All patients admitted to the adult neurological intensive care unit (neuro ICU) during a 30-month period were included. The study consisted of two 1-month pre-intervention observation periods (approximately 1200 catheter days) followed by a 30month intervention phase (20,394 catheter days). A comprehensive evidence-based UTI bundle encompassing avoidance of catheter insertion, maintenance of sterility, product standardization, and early catheter removal was enacted.

The urinary catheter utilization rate dropped from 100% to 73.3% during the intervention phase (p <0.0001) without any increase in the rate of sacral decubitus ulcers or other skin breakdown. The rate of catheter-associated UTI was also significantly reduced from 13.3 to 4.0 infections per 1000 catheter days (p <0.001). There was a linear relationship between the decreased quarterly catheter utilization rate and the decreased catheter-associated UTI rate (r2 = 0.79, p <0.0001).

This single-center prospective study demonstrated that a comprehensive UTI prevention bundle along with a continuous quality improvement program can significantly reduce the duration of urinary catheterization and rate of catheterassociated UTI in a neuro ICU. Continued efforts to reduce CA-UTI beyond the study resulted in sustained reductions when all components of the bundle were in place and daily foley rounds were maintained as a nursing intervention. 

Matrix metalloproteinases (MMPs) are extracellular proteolyic enzymes that may modulate the neuroinflammatory response to brain injury. We sought to determine the effect of MMPs on pro-inflammatory cytokine production following severe traumatic brain injury (sTBI).

As part of a prospective cohort study, 8 adults with sTBI underwent multimodal monitoring with high cutoff, cerebral microdialysis and arterial and jugular venous bulb catheters. The concentration of MMPs and pro-inflammatory cytokines were measured in microdialysate and blood over 6-days.

Interleukin-1-alpha (IL-1), IL-1-beta, IL-6, IL-10, and tumor necrosis factor-alpha (TNF-alpha) concentrations were initially high in microdialysate and then declined to low levels. The microdialysate concentration of IL-8 also declined after first being high, but then increased between 84-and 90-hours. With the exception of IL-5, IL-10, and TNF-alpha, the cytokine blood concentration was low to undetectable. Using generalized estimating equations, we observed a positive change in the microdialysate concentration of IL-8 [(2.59 pg/mL)/(pg/mL); 95% CI, 1.12 to 4.05] with an increase in the MMP-9 microdialysate concentration. In contrast, a significant increase in the microdialysate concentration of MMP-9 was seen with an increase in IL-1-alpha [(34.67 pg/mL)/(pg/mL); 95% CI, 23.43 to 45.91] and IL-1-beta [(4.41 pg/mL)/(pg/mL); 95% CI, 3.67 to 5.15]. In blood, a significant change in MMP-9 occurred during an increase in the levels of IL-1-beta [(16304.15 pg/mL)/(pg/mL); 95% CI, 824.73 to 31783.57] and IL-8 [(2934.10 pg/mL)/(pg/mL); 95% CI, 1335.75 to 4532.25]. Although IL-10 levels were higher in cerebrospinal fluid (CSF), no major difference in MMP or cytokine concentration was observed between arterial and jugular venous blood or, for the three patients who were also fitted with CSF drainage catheters, between cerebral microdialysate and CSF.

sTBI is associated with a substantial central cytokine or neuroinflammatory response, which may influence or be influenced by production of MMPs.

Severity classification of traumatic brain injury (TBI) has traditionally been based on the Glasgow coma scale (GCS), with mild TBI being defined as 13-15. However, there is often a subset of "mild" TBI that requires surgical intervention. The current study examines this subgroup to decipher any symptomatology that may be helpful in identifying who these patients may be.

This observational cohort study included consecutive adult patients presenting with a TBI. Independent variables included vomiting, seizure, loss of consciousness (LOC), alteration of consciousness (AOC), and post-traumatic amnesia (PTA); these were tested for correlation with surgical intervention, the dependent variable. Data were entered into RedCap, a clinical data capture system housed in our Center for Translational Science Institute. The z-test for proportions was used to determine significance of symptomatology. Statistical analyses were performed in JMP 9.0 for the mac.

Of the total mild TBI cohort (n=1,088), 57% were male. The median age was 39 (IQR: 24-59, R: 18-96). Thirty seven patients required surgical intervention. Symptoms significantly associated with surgical intervention on univariate regression included vomiting (P=0.0320), and AOC (P0.0448). Multiple regression analysis revealed that time (length of) LOC (P=0.0118) and PTA (P<0.0022) were also significantly correlated with surgical intervention. Age was also a statistically significant predictor of surgical intervention (P<0.0001).

These pilot data suggest that older patients, as well as patients who present with vomiting, LOC, or PTA, have a significant likelihood of requiring surgical intervention. This calls attention to proactively seeking these data and ensuring adequate neuroimaging for all patients with TBI, regardless of GCS score at presentation.

The prevalence of chronic subdural hematoma (SDH) is expected to increase with an aging population and increased use of anticoagulants. We aimed to develop a tool to predict mortality after SDH.

A prospective study was conducted between 2008-2011 of patients with chronic subdural hematoma (N=102) admitted to a tertiary neuro-ICU. Three-month mortality data was collected. After testing admission demographic, radiographic and -8, 9-14, 15 ; P=0.001) and herniation (P=0.022) were found to be independent predictors of death in multivariate logistic regression analysis. A score was composed (0-4) with each variable weighted based on its independent strength of association with mortality (B value) as -14=1, GCS 3-8=2, herniation=1 point.

Overall, 24% of patients died and 3-month mortality increased with each point of the SDH score (0=0%, 1=8%, 2=41%, 3=60%, 4=100%). The SDH score predicted death (OR 5.3, 95% CI 2.5-11.5, P<0.0001) with an area under the curve of 0.821, sensitivity 87.5%, specificity 69.2%, PPV 47% and NPV 95%. The Hosmer and Lemeshow and Nagelkercke R 2 for this model were 0.380 and 0.362, respectively, indicating a strong model. SDH evacuation reduced the odds of death by 80% when added to a multivariate model including age, GCS and herniation (adjusted OR 0.2, 95% CI 0.1-0.8, P=0.025).

The SDH score allows for a reliable prediction of mortality for patients with chronic SDH. This score may help risk stratify patients for surgical treatment. 

We developed a novel method capable of determining the degree of conformance of observed morphological changes of intracranial pulses with their expected patterns associated with global vasodilatation and vasoconstriction, respectively. These patterns were formed as a template consisting of pulse morphological changes during CO 2 tests that were consistent for multiple subjects. We used this novel pulse morphological template matching (PMTM) algorithm to study 1) the incidence of cerebral vasoconstriction/vasodilatation associated with LPR increase episodes; 2) how likely cerebral vasoconstriction/vasodilatation could lead to or lag behind LPR increase. We studied 4318 Microdialysis data samples collected in an average interval of 1.3 hours from 30 severe TBI patients. The LPR increase episodes were automatically identified using a moving time-window of 4 hours. The PMTM algorithm was applied to the continuous intracranial pressure (ICP) signal time-synched to the identified LPR episodes.

Across all subjects, more than half of the LPR increase episodes are not associated with any detectable cerebral vasoconstriction or vasodilatation (p = 1e-5). Comparing LPR episodes with either vasoconstriction or vasodilatation, it was more likely that vasoconstriction rather than vasodilatation occurred during an LPR increase episode (p = 0.008). Also for 3 out of 30 subjects with dominant number of vasoconstrictive LPR episodes, a causality relationship between vasoconstriction and LPR increase were observed, i.e., vasoconstriction occurred in one hour before LPR increase started.

Across the 30 TBI subjects studied, the incidence of either vasoconstriction or vasodilatation associated with LPR increase was low. However, about 10 percent of subjects had a dominant number of LPR increase episodes associated with cerebral vasoconstriction. Furthermore, cerebral vasoconstriction occurred within one hour preceding LPR increase.

Placement of an intracranial pressure (ICP) monitorto guide the management of severe traumatic brain injury (TBI) patients has been historically performedby neurosurgeons. Trials have suggested decreased morbidity and mortality with timely resuscitationand rigorous treatment of intracranial hypertension. We hypothesize that ICP monitors can be placed by non-surgeon neurointensivists, with placement success and complication rates comparable to neurosurgeons.

We retrospectively reviewed the medical records of TBI patients who required insertion of parenchymal ICP monitors from May 2010 to December 2011 in a large level I trauma center. Monitor placement was performed by 4 neurointensivists (board certified by the ABIM in Critical Care Medicine and by the UCNS in Neurocritical care). Patient data recorded are age, gender, CT findings, ICP monitor placement location and length of placement, complications related to the ICP monitor, and patient outcomes.

Twenty seven (27) 

These findings were comparable to published outcomes from neurosurgeon placements. We believe that insertion of parenchymal ICP monitors should be considered a core skill for neurointensivists and should be included in neurocritical care fellowship training. Insertion of ICP monitors by neurointensivists is safe and may aid in providing prompt monitoring of patients with severe TBI. 

Use of computers at the bedside for trending primary signals like ICP or CPP brings obvious advantages in neuro-critical care unit. Software can be extended to calculate secondary indices reflecting underlying pathophysiological phenomena, like disturbance of cerebral compensatory reserve and vascular reactivity.

During 2003-2011 more than 400 severe TBI patients were monitored using ICM+ software. Various modalities were used, including ICP, ABP, PbtiO2, NIRS, TCD blood flow velocity, brain temperature, etc. From ICP and ABP waveforms secondary indices were extracted. Compensatory reserve was assessed using moving correlation index between slow changes in pulse amplitude and mean ICP (RAP). Pressure reactivity index (Prx) was calculated as moving correlation between mean ICP and ABP. 'Optimal CPP' (CPPopt) was estimated as CPP corresponding to the best cerebrovascular reactivity within the period of past 4 hours.

Trending compensatory reserve showed that usually it is good (RAP around 0) in the first few hours after admission (RAP around 0), with gradual deterioration triggered by aggravating brain edema. In most cases RAP stayed close to +1 (impaired reserve). It decreased to negative values (exhausted reserve) on top of plateau waves and in refractory intracranial hypertension, indicating critical ICP. PRx proved to be highly variable, responding to changes in ABP, ICP and ventilation. It deteriorated on top of plateau waves, and at extreme values of CPP. In cases of refractory intracranial hypertension, deterioration of reactivity seemed to preceed the elevation of ICP above 30 mmHg. CPPopt fluctuated during the monitoring period. Absolute distance between current CPP and CPPopt was strongly associated with outcome. Too low CPP (below CPPopt) correlated with greater mortality rate (p<0.0001) and too high CPP -with greater rate of severe disability (p<0.025).

Individual observations of secondary indices calculated by ICM+ software help in better interpretation of primary signals in intensive care of TBI patients.

Financial Support: The software for brain monitoring ICM+ is licensed by the University of Cambridge (Cambridge Enterprise). Authors PS and MC have a financial interest in a part of the licensing fee.

To determine the differences in hospital outcomes among adult mild traumatic brain injury (TBI) patients where the severity of TBI is defined by Glasgow Coma Scale (GCS) score.

This is a retrospective chart review of consecutive adult who came to the ED department of a tertiary care hospital in North Central Florida. The TBI severity was classified according to GCS score, with patients with GCS score of 13-15 categorized as having mild TBI. Outcome variables such as admission status, ICU admission status, in-hospital death and 3-month death among patients with different mild GCS scores of 13, 14, and 15.

We had a total of 809 mild TBI patients in the specified period of time. The majority of this cohort had a GCS of 15 (697 or 86.1%). This was followed by a GCS of 14 (84 or 10.4%) and GCS score of 13 (28 or 3.5%). There was a statistically significant difference between mild TBI with GCS 13, 14, 15 (p<0.0001, ANOVA) with the outcomes of hospital admission (92% vs. 67% vs. 30%), ICU admission (61% vs. 31%, vs. 9%), in hospital death (4% vs. 4% vs. 0.3%), and 3 month death rate (7% vs. 5%, vs 1%). There is a 30% increase in hospital admission rates for each point decrease in GCS score. The 3-month death rate nearly doubles with each incremental decrease in GCS score.

There is a significant difference in outcome within "mild" head trauma across the continuum.

To characterize the patterns of presentation of children with head trauma to the Pediatric Emergency Department.

This is an observational cohort study that sought to collect injury and outcome variables with the goal of characterizing the very early natural history of pediatric traumatic brain injury in children over the age of 3 years. Statistical analyses were performed using JMP.

Cohort (n=351) . Similar multivariate model showed that as children grew older, they were more likely to be admitted in hospital because of a TBI as a result of recreational activities (p=0.0002) and traffic accidents (p<0.0001), and less likely due to sport TBIs (p=0.0003) with AdjR 2 =14%. 42% of the children who were admitted ended up in ICU with mean ICU-LOS of 1 day with an IQR of 0-3. One percent had an in hospital death. Kids with amnesia were significantly more likely to be admitted to the ICU (p=0.0018, R 2 =11%). Children who got admitted to ICU (p=0.0012) and were older (0.0262), were significantly more likely to be readmitted to the hospital within 30 days.

These preliminary data suggest that pediatric brain injury is not without significant morbidity.

The objective of this study was to identify pre-hospital markers of in-hospital mortality in traumatic brain injury (TBI) patients due to fall.

This study was an observational cohort study performed at a Level-1 Trauma center. Study subjects included all adult arriving in emergency department with history of TBI due to fall over a period of 20months. Study variables were symptoms such as vomiting, seizure, loss of consciousness (LOC), alteration of consciousness (AOC) & post-traumatic amnesia (PTA), Glasgow coma scale (GCS) scores, vitals, pre-hospital glucose. JMP 10 for windows & Z-test of proportions were used to perform statistical analysis.

The study cohort comprised of 612 adult (median age of 49yrs and IQR of 28-68). In-hospital death (IHD) was observed in 5% (N=28) of the total cohort, with Male IHD=19 (68%) greater than Female IHD=9 (32%). PTA (25%,p<0.0001),LOC (64%, p<0.0422) & AOC (54%, p<0.0001), higher pre-hospital glucose (p=0.0048) were individually found to be much more significantly associated with IHDs versus the whole cohort. Multivariate regression analysis showed significant correlation with IHDs with: 1) Higher age (p<0.0001) when adjusted for Severe GCS. 2) Vomiting (p=0.0131) and Longer duration of LOC (p=0.0236), when adjusted for rest of the symptoms. 85% of patients presenting to ED with vomiting (p=0.0373) had GCS score=15, and 3% of that sub-group suffered IHD

Patients presenting to emergency department with higher blood glucose and symptoms such as PTA, AOC, longer duration of LOC & vomiting, were more likely to have worse outcome of in-hospital deaths compared to rest of the patients. Hence identifying these symptoms in fields might help to make key decisions for providing intensive care and improving the overall patient outcomes.

To determine which symptoms affect severity in pediatric traumatic brain injuries (TBI).

Study design-this observational cohort study was performed at a level one trauma center that has a dedicated pediatric emergency department. Consecutive patients age 4-17 were included. The age cutoff of 4 years was used because it was decided that younger children may not be ale to report their symptoms, particularly to endorse AOC (alteration of consciousness) or PTA(post-traumatic amnesia). The dependent variables were vomiting, Seizures(SZ), LOC (loss of consciousness), AOC, and PTA at the time of the head injury. The independent variable was ED TBI severity based of the Glasgow coma score, with mild being defined as 13-15, moderate as 9-12, and severe as less than 8.

The median age of the cohort was 13 with an interquartile range of 7 to 16. 59% were boys. In the univariate model, all symptoms except vomiting were statistically significant: Seizures (p=0.0437); LOC (p=0.0139); AOC (p<0.0001), PTA (p<0.0001). Multiple regression analysis of these factors revealed all of variables to retain statistical significance. The R 2 coefficient of determination was 33%, which means that almost one-third of the variance can be explained by just these five factors (symptoms), suggesting that our multivariate model is a robust one.

Symptoms at time of head injury in children including seizure, LOC, AOC and PTA were statistically significant predictors of the severity of TBI. This data allows clinicians to judge the severity of the TBI depending on the symptoms at presentation. These pilot data may be useful in designing clinical care algorithms.

ICP dynamic system of an injured brain is susceptible to various acute changes disturbing the system homeostasis that should be detected by ICP monitoring. Such a capability is particularly useful for comatose patients. Our aim was to demonstrate a novel approach to detect acute deviation from steady state of an ICP dynamic system without involving significant mean ICP changes.

Steady state of ICP dynamic systems is reflected as ICP pulses of similar mean ICP resembling each other. Therefore, a steady state indicator can be calculated by quantifying inter-pulse distances after matching their mean ICP. Besides Euclidean distance and Pearson correlation, geodesic distance was introduced as a novel metric. These different metrics were evaluated on three types of continuous ICP: 1) those between two consecutive imaging studies showing new acute ventricular enlargement for slit ventricle syndrome patients undergoing a trial of shunt externalization and clamping (SVS+); 2) those between consecutive brain imaging studies from patients under the same trial without ventricular enlargement (SVS-); 3) overnight recordings from patients with suspected normal pressure hydrocephalus (NPH). It was expected that both SVS-and NPH recordings represent steady state.

We observed that only geodesic distance correctly differentiated between SVS+ and SVS-and between SVS+ and NPH while avoiding discriminating between SVS-and NPH. It was also found that 45% SVS+ cases, none of SVS-, and 3.8% of NPH cases had a multimodal geodesic distance histogram. Pulses with a large number of distant pulses at similar mean ICP for the five multimodal-histogram SVS+ cases fell in short time windows indicating that acute ventricular changes may have occurred in these confined time windows during which no significant changes of mean ICP occurred.

Geodesic inter-pulse distance is a promising metric to quantify distance intrinsic to the underneath geometric structure of ICP signals. 

Patients with severe traumatic brain injury have multiple causes for acute respiratory decompensation. Computed tomography pulmonary angiography (CTPA) is being used extensively to evaluate acute cardiorespiratory changes. We reviewed the use of CTPA in critically ill patients with traumatic brain injury to evaluate the results and their impact on patient care.

All adult trauma patients with traumatic brain injury who were admitted to our Level 1 Trauma center intensive care units for greater than 48 hours, were identified (January 2007-December 2010). Those who underwent CTPA for acute respiratory decompensation were reviewed to determine the findings of these studies and the resulting interventions.

We identified 881 patients that met these criteria [196 admitted to Neurosurgery/Neurocritical Care(NCC) , 685 admitted to Trauma Service(TS)]; 86 of these patients underwent CTPA studies for acute physiologic changes (NCC-10, TS-76, p=0.013). TS patients were significantly younger with higher severity of injury and longer length of stay. Pertinent clinical finding were identified in 81 of the 86 (94%) studies; and included atelectasis/collapse 52 (60%), pleural effusion 18 (21%), pneumonia 9 (10%) and pulmonary embolus 14 (16%). These results prompted 153 targeted interventions, most frequently consisting of modifications of ventilator therapy (58,67%), a change or initiation of antibiotic therapy (31,36%), mini-BAL (21,24%) bronchoscopy (16,19%), vena cava filter (10,12%), and anticoagulation (5,6%). No change in patient management occurred after 10 studies. Agreement, for different findings, between chest x-ray and CTPA ranged from 46-98%.

Patients admitted to a TS are more likely to undergo a CTPA evaluation. CTPA is a useful tool in the evaluation of critically ill patients with acute physiologic decompensation beyond the diagnosis of PE. The results of these studies provide significant insight into the underlying pathology in this patient population and offer an opportunity to direct subsequent patient care. 

Somatosensory evoked potentials (SSEP) provide valuable information of the neurophysiological state of the patient throughout a surgery and the errors in the surgical procedure are easily noticed. It is hence important to analyze and monitor the SSEP during scoliosis surgery in a minimum amount of time. The study uses PCA-Walsh algorithm to analyze posterior tibial nerve SSEP and compare with the conventional signal averaging method in twelve surgical procedures.

The tibial SSEP from twelve different subjects were recorded and assessed throughout the respective surgeries using a unique PCA-Walsh algorithm by using only 10 trials at a time and compared the extracted SSEP information with conventional method. The SSEP were recorded in two bipolar channels C 3 -C 4 and C z -F z throughout the surgery and analyzed remotely using an automated software PCA-Walsh algorithm. The results are compared with the actual clinical information and presented with the merits.

In all the twelve cases, the algorithm results presented consistency throughout the surgery with an average accuracy of 91.5% when compared to the conventional method, which takes several hundred trials. The average variation in time latency was 4.4% and in amplitude was 23.7%, well within the limit of 50% following the clinical criteria.

The PCA-Walsh algorithm is capable of automated extraction of the tibial SSEP during a surgery using a minimum number of trials. The analysis using the algorithm was successful and proved conclusive to the clinical information through the different surgical procedures. The faster recording and analysis of SSEP signals provides a much better perspective for neurophysiological monitoring through the surgical procedure.

The authors appreciate the support provided by the National Science Foundation under grants CNS-0959985, HRD-0833093, CNS-1042341, and CNS-0426125. The authors are also thankful for the clinical suppo

Certain admission characteristics are known predictors of adverse outcomes in moderate-severe traumatic brain injury (msTBI) patients, but explain only 1/3 of outcome variability. Retrospective studies suggest that non-neurologic organ failure may contribute to 2/3 of all deaths after msTBI, but actual incidence rates of intensive care unit (ICU) complications and their impact on outcome are not known. We examined the incidence rates of pre-specified medical and neurological ICU complications, and their impact on in-hospital mortality and functional outcome at hospital discharge.

In a prospective observational study, 183 consecutive msTBI patients from a single Level I trauma center between 11/2009-5/2012 were analyzed. Poor outcome was defined as Glasgow Outcome Scale 1-3.Multivariable logistic regression was utilized to adjust for admission characteristics and ICU length-of-stay.

The mean age was 51 years, 71% were men, and the median Glasgow Coma Scale and Injury Severity Scores were 4 and 29, respectively. The five most common medical ICU complications were: hyperglycemia (77%), fever (62%), hypotension requiring vasopressors (40%), systemic inflammatory response syndrome (38%), anemia requiring transfusion (32%). Neurological ICU complications were: intracranial pressure crisis (ICP; [51% of n=72 with ICP monitor in place]), brain edema (38%), herniation (38%), intracranial rebleed (36%), clinical seizure (12%). Among medical complications, hyperglycemia was associated with poor outcome (OR 2.6; 95% CI 1-6.9]) while cardiac complications (e.g. cardiac arrest, arrhythmia, acute myocardial infarction) were associated with death (OR 6.1; 95% CI 1.8-20.6). When combining medical with neurological ICU complications, brain edema (OR 5.1; 95% CI 1.5-18) was associated with poor outcome, while cardiac complications and brain edema were associated with death (OR 6.9; 95% CI 1.5-31.6 and OR 11.3; 95% CI 3-43, respectively).

ICU complications are very common after msTBI. We identified specific potentially modifiable predictors of adverse outcomes after msTBI. Confirmation of our findings in a larger cohort is warranted. 

Too much oxygen may increase oxygen free radical production, possibly triggering cellular injury and apoptosis. Although laboratory investigations support the potentially detrimental effects of hyperoxia exposure after TBI, clinical data are lacking.

We retrospectively identified TBI patients admitted to our Neuro-ICU between July 2009 and February 2012. We identified a total of 169 patients with complete data including GCS, APACHE II, Age, Gender, ABG within 24 hours of injury, and outcome (Glascow Outcome Scale-GOS at discharge from the hospital). Patients were divided into 2 groups defined a priori based on PaO2 on the first ABG values obtained after injury. Hyperoxia was defined as PaO2 of 245 mm Hg or greater, and normoxia as PaO2 between 60 and 240. Poor outcome is defined as GOS of 1-3.

The patients in the normoxia group (n = 102) and the hyperoxia goup (n = 63) were matched on baseline characteristics, age ( 

Among a small number of patients admitted to the Neuro-ICU following Traumatic brain injury, patients with arterial hyperoxia had a trend towards worse outcome compared with patients with normoxia. This provides scientific rationale for large prospective clinical trials of controlled oxygenation in TBI patients. 

Elevated intracranial pressure (eICP) contributes to secondary injury in sTBI, therefore its control is paramount. Boluses of hypertonic solutions are usually used to reduce ICP but the impact of early continuous infusions has not been widely explored. We conducted this study to compare the effect and security of Hypertonic Saline 3% (HS3%) infusion vs normal saline.

All sTBI patients arriving to the emergency room within 3 hours of trauma were enrolled to receive an isovolumetric infusion of HS3% or normal saline (placebo) during 72 hours. ICU physicians and investigators were blinded to the sodium levels during the trial. Main endpoint: number of eICP episodes (>20mmHg). Secondary endpoints: neurologic outcome (GOS, mRankin), electrolyte and osmolality levels, and adverse events (AE).

Twenty non-penetrating sHBI patients were included. Median age was 33.5 years (IQ25-75:26.5 -41.5). Median ISS was 29(IQ25-75:25-34). We didn´t find significant differences for the total number of episodes of eICP at 72h between groups (6, IQ25-75:1-9 vs. 14, IQ25-75:0-23, p=0.49); however, when we analyzed patients with at least one episode of eICP we found a significant low number of eICP episodes in HS3% group (9, IQ25-75:2-9 vs. 22, IQ25-75:17,5-40,5; p=0.003). We found a sodium plateau at 42h of infusion (HS3%:151.8±5.8 vs. Control:139.2±6.2mEq/L, p=0.02) which lasted until the beginning of weaning from HS3%. The most frequent AE was hypokalemia and no patient had renal failure. The sixmonth GOS and mRankin scores had a non-significant tendency towards better outcomes in HS3% group.

An early infusion of HS3 is feasible and seems to be safe in sTBI patients. Serum sodium kinetics showed a plateau after 42h of HS3% infusion with no consequences in renal function and no rebound effects after tapering. HS3% continuous infusion could reduce eICP episodes and it could conduct to better neurologic outcomes at six months. 

Traumatic brain injury causes diffuse shearing of long fiber tracts. This can be detected by quantitative DTI imaging even in patients who have primarily localized contusions. In our population the cingulum, cotricospinal tracts and external sagittal striatum were preferentially affected compared to age and gender matched controls. These findings support the use of advanced MRI to assess the degree of injury and inform prognosis and goals for rehabilitation.

Neurocrit Care (2012) 17:S1-S337 

Most deaths following severe traumatic brain injury (TBI) are associated with a decision to withdraw life-sustaining therapies (WLST)(1). However, the incidence and the impact of WLST in clinical trials is unknown. This systematic review was performed to assess if and how WLST are dealt with in clinical trials involving patients with severe TBI.

We searched MEDLINE, EMBASE, Cochrane Central, BIOSIS and CINAHL databases and references of included studies. All randomized controlled trials (RCTs) published over a 10-year period (January 2002 (January -2012 , in one of 18 selected journals in general medicine, critical care medicine and neurology/neurosurgery were considered for eligibility if 8) and reporting data on mortality. Our primary outcome was the assessment of WLST. Secondary outcomes were the timing of evaluation, justification for WLST, proportion of WLST among deaths, factors that may have influenced the WLST and risk of bias of RCTs. Two reviewers selected RCTs and collected data independently using a standardized case report form.

From 3737 citations retrieved, 37 RCTs were included (n=15,460, ranging from 11 to 10 008 patients). 24 were single center RCTs and 13 were multicenter. The incidence of WLST was reported in 4 studies (10.8%). Three studies reported crude numbers of patients, 3 studies reported the timing of WLST and 3 studies reported the justification for the decision to WLST. 16 studies were considered at high risk of bias, 1 study at low risk of bias and 20 studies did not give enough information to conclude on the risk of bias.

WLST was rarely reported in RCTs involving patients with severe TBI over the last decade. Considering the variation of WLST in clinical practice, we suggest that WLST should be systematically reported in RCTs performed in TBI. Reference : 1. Turgeon et al. CMAJ 2011.

Previous pediatric brain injury studies have considered fevers as discrete events instead of as a "temperature dose." We sought to evaluate the population size difference captured at various fever thresholds in severely brain-injured pediatric patients, considering fever burden in terms of degree-hours; and to compare fever burden in pediatric traumatic brain injury (TBI) vs. cardiac arrest (CA).

Charts from brain-18y, admitted in 2005-8 within 24 hrs of admission were included. No temperature modulation protocols existed in the pediatric ICU during this period. 7-day core temperatures were used to generate areas-under-the-curve (AUC) above fever thresholds of 37.5, 38, 38.5, and 39 o C. These were normalized for different lengths of stay.

46 charts met inclusion criteria, with mean patient age 7.2y (range 2d -18y). Diagnoses distributed (non-exclusively) as 19 CA, 14 accidental TBI, 9 non-subarachnoid hemorrhage (SAH) intracranial bleeds, 3 SAH, 4 strokes, 4 non-accidental TBI, 4 CA after TBI, and 11 other CNS pathologies. Cohort mortality was 50%, with 30% suffering brain death. Fever burdens were measurable in 93% of patients over 37.5 o C, in 80% over 38 o C, and in 39% over 39 o C. Normalized fever burdens at these thresholds were statistically different by 1-way ANOVA (p<0.0001), with all fever burdens being statistically less than at 37.5 o C. Remarkably, a shift in threshold from 37.5 to 38 o C resulted in a 56% reduction in measured fever burden. Fever burdens fell from a peak of 6.9 ± 11.2 o C-h on day 1 to 1.7 ± 3.0 o C-h on day 3 after admission. Accidental TBI (n=14) and CA (n=19) patients did not experience different fever burdens above 37.5 o C.

Measured fever burden is markedly affected by shifting the threshold from 38 to 37.5 o C. TBI and CA appear to induce similar fever burdens. Pediatric fever burden reference values will allow more quantitative comparisons in severely braininjured children. 

Little is known about the natural history of function after traumatic brain injury. Our objective was to track the stability of DRS scores over time and to identify factors associated with worsening DRS scores.

We collected Disability Rating Scale (DRS) scores, which capture the cognitive ability to perform activities of daily living such as communication, motor response, feeding, toileting, overall functioning and employability, longitudinally on severely brain injured patients in neurosurgery clinic. Multivariable logistic regression was used to identify patient factors that were independently associated with changes in DRS score over time.

65 patients with severe brain injury had more than one DRS score collected. Of these patients, 27 had worsening scores over time. Changes in scores ranged from -8 to 9 (mean -0.3, standard deviation 3.4). This represented a change from partial/no disability to moderate disability for 3 patients and from moderate to severe disability for 3 patients. 8 patients improved from moderate to partial/no disability while only one patient improved from severe to moderate disability. Using multivariable logistic regression, there were no patient factors that were associated with worsening DRS scores including gender, age, comorbidities, race, insurance status, mechanism, injury severity score, GCS or final disposition. While half of worsening DRS scores were seen within the first 100 days after discharge, 14 were seen 3 months or more after the hospital stay, with one seen over a year after hospital discharge.

For the most part, DRS scores were stable over time. A group of patients were identified who experienced significant decline in function as far out from discharge as a year. This preliminary study highlights the need to identify those at risk for decline and to set up mechanisms for long-term follow-up for those patients in need. 

The identification of traumatic axonal injury (TAI) lesions that undergo neuronal recovery could improve prognostication in patients with traumatic brain injury (TBI) and facilitate the development of novel therapies for preventing secondary axotomy. We aimed to determine whether diffusion tensor imaging (DTI) detects neuronal recovery after TAI.

We retrospectively identified 7 TBI patients (5 severe, 1 moderate, 1 mild) who underwent at least 2 acute-to-subacute DTI scans and who had at least 1 TAI lesion in the corpus callosum (CC), as defined by hyperintensity on DWI or T2 FLAIR. The median number of days from TBI to image acquisition was 2 (range 1-7) for the first DTI scan and 26 (range 14-88) for the second scan. TAI lesions were manually outlined on the acute DWI datasets and then coregistered to the subacute datasets to measure longitudinal changes in lesional fractional anisotropy (FA) and apparent diffusion coefficient (ADC). "Neuronal recovery" within a TAI lesion was defined on the final scan by mean lesional FA within 2 standard deviations of published normal FA values for the CC. Initial FA and ADC values in lesions with and without neuronal recovery were compared (unpaired t-test).

Eleven CC TAI lesions (7 splenium, 2 body, 2 genu) were identified. FA recovered in 3 splenium lesions (0.63+/-0.03 [mean+/-SD]) and 1 genu lesion (0.53) on the final scan. Three of these 4 lesions were FLAIR hyperintense, 3 were associated with GRE microbleeds, and 3 were initially ADC hypointense. Splenium lesions with neuronal recovery did not differ significantly from lesions without recovery for FA (0.65+/-0.06 vs. 0.49+/-0.15, p=0.14) or ADC (568+/-45 vs. 772+/-222 x10^-6 mm^2/s, p=0.19) on the initial scan.

DTI may detect neuronal recovery within TAI lesions, as indicated by subacute normalization of FA. Acute DTI biomarkers of TAI reversibility were not identified in this preliminary analysis. 

Increased intracranial pressure (ICP) in patients with traumatic brain injury (TBI) is associated with higher mortality and poor outcome. Mannitol and hypertonic saline (HTS) have both been used to treat high ICP, but it is unclear which one is more effective. We compared the effect of mannitol and HTS on lowering cumulative ICP burden after severe TBI.

The Brain Trauma Foundation TBI-trac® New York State database was used for this retrospective study. A total of 26 patients with severe TBI who received only HTS were identified. 26 patients who received only mannitol were matched for age, pupillary reactivity, occurrence of hypotension on day 1. Univariate analysis was performed to compare ICP burden, cumulative hyperosmotic doses, number of ICU days (Wilcoxon signed rank test), and two-week mortality (McNemar test). ICP burden was defined as the total number of days with ICP spikes (ICP>25mmHg) expressed as a percentage of total number of days of ICP monitoring. Cumulative HTS and mannitol doses were converted to osmolar doses for comparison.

The mean age and GCS were similar in the two groups (36.26 vs. 38.79 years; 5.50 vs. 5.27; HTS vs. mannitol, respectively) . 25 patients received 3% HTS and 1 received 23.4% HTS. All patients in the mannitol group received 20% mannitol. There was no difference in number of days of ICP monitoring (p=0.49) or number of ICU days (p=0.95) in the two groups. ICP burden was significantly lower in HTS group vs. mannitol group (14.67% vs. 33.62%. p=0.02). There was no significant difference in the cumulative dose of HTS and mannitol (p=0.2), and two-week mortality in the two groups was similar (p=0.32).

HTS is more effective in lowering cumulative ICP burden after severe TBI compared to mannitol. This did not translate into reduction in two-week mortality, possibly due to the small sample size. 

Spreading depressions (SDs) have been consistently associated with hypoglycemia in animal studies. The frequency of these depolarization events, while influencing infarct size, also appears to be influenced by the plasma glucose concentration during experimental ischemia. Low cerebral dialysate glucose have also been correlated with SD events in humans. We hypothesized that low serum glucose should be associated with an increase in the frequency of SD events in human acute brain injury.

To determine the relationship between serum glucose and cortical spreading depolarizations (SDs) after traumatic brain injury (TBI), subdural electrode strips were placed on peri-contusional cortex in 103 patients from 4 centers who underwent craniotomy following TBI. Prospective electrocorticography was performed during neurointensive care with retrospective analysis of hourly serum glucose data. Patients were divided into those with SDs and those without and the distribution of glucose values among these two groups were compared using the 2-way Kolmogorov-Smirnov method.

In 61 patients (59%), 1682 SDs (spreading depressions and peri-infarct depolarizations) were observed. The probability of a depolarization occurring increased significantly as a function of rising serum glucose (p<0.000008). Median glucose values in patients with and without SDs was 7.6 and 7.0 mmol/L, respectively. Among patients with SDs, glucose values recorded within 60 minutes of the onset of an SD were higher than those occurring <75 minutes before an SD (p<3.8e-13) ( Figure 1 ).

Serum glucose does not appear to affect the generation of SDs as it does in animals but paradoxically may be elevated. This may reflect a stress response to the initial acute brain injury and critical illness or a physiologic mechanism to increase glucose supply during SD events in which cerebral glucose utilization is increased. Overall, the data suggest that plasma glucose is being managed within appropriate levels in this study group.

To determine difference in TBI severity and abnormal radiologic findings in different age groups.

This was an observational cohort study on all adult patients (>18yrs) arriving to the Emergency Department, with a history of Traumatic Brain Injury as a result of "fall" at a Level 1 Trauma center in the Southeastern United States. Data collected included ED GCS score and head CT results. Abnormal CT scans have the presence of either an intra-cranial bleed and/or cranial fracture.

There were 612 patients in the cohort with history of fall with median age of 49yrs (IQR of 28-68). We divided them into two age groups: Group A: 18-55yrs(60%) & Group-B: >55yrs(40%). Group A comprised mostly of males (66%) compared to females, meanwhile Group B was equivocal in gender composition (51% male). Out of the whole cohort of 612 patients, 92% patients had head CT performed. Out of these, 46.74% (n=265) showed an abnormal head CT. Age Group B (65%) had a significantly greater percentage of abnormal CT scans compared to Group A (34%) (p<0.0001). Among abnormal CT scans, Group A (50%) had a significantly greater percentage of skull fractures than Group B (22%) (p=0.0084). Among patients with mild and moderate TBI (GCS>9), Group B (64%) was more likely to have an abnormal CT scan than Group A (24%) (p<0.0001), however, there is no significant difference between likelihood of abnormal head CT between Group A and B for severe TBI.

Younger adults are at a higher risk of cranial fractures after a fall related TBI, probably due to severe mechanisms of injury. On the other hand elderly population with mild TBI mostly due to ground level falls had worse outcomes on CT scans.

Accumulating pre-clinical data suggests that matrix metalloproteinase (MMP) expression following cerebral trauma contributes to brain injury. We sought to characterize the temporal MMP response to severe traumatic brain injury (sTBI) in humans and its relationship with outcomes.

We conducted a prospective cohort study that included 8 adults with sTBI. High-cutoff, cerebral microdialysis and arterial and jugular venous bulb catheters were used to measure the concentration of MMPs and other markers over 6-days.

The concentration of MMP-1 was initially low in microdialysate and blood, but increased between 96-and 156-hours. MMP-2 blood levels were high and stable throughout the study while blood levels of MMP-3 were initially low and then gradually rose. In microdialysate, MMP-2 and -3 increased and then peaked between 42-and 48-hours. MMP-7 also increased in microdialysate following sTBI while its levels were low and stable in blood. MMP-8 and -9 were initially high in microdialysate and then slowly decreased over time. While the concentration of MMP-9 was also initially high in blood and then progressively declined, the MMP-8 blood level increased with time. Among the patients that also had cerebrospinal fluid (CSF) drains, marked and sometimes opposite concentration trends were observed for MMP-7 in microdialysate versus CSF. Generalized estimating equations suggested that significant changes in mean microdialysate concentrations of MMP-1, -2, -3, and -8 and MMP-1, -2, -3, and -9 occurred with increases in microdialysate glucose and the lactate pyruvate ratio, respectively. Moreover, the mean microdialysate level of MMP-8 increased with intracranial pressure (ICP) [(49.27 pg/mL)/mmHg; 95% confidence interval, 9.03 to 89.51] while that of MMP-7 decreased with cerebral perfusion pressure (CPP) [(-3.25 pg/mL)/mmHg; 95% confidence interval, -5.92 to -0.58].

Monitoring of MMPs following sTBI is feasible, and their expression may be associated with cerebral metabolism, ICP, and CPP.

To determine significance of Laboratory markers for In-hospital death after fall related adult Traumatic Brain Injury.

This was a consecutive cohort observational study done at a Level-1 trauma center serving surrounding 12 counties. Cohort consisted of all adult patients (> 18yrs) arriving to the ED with a history of fall. Study variables were lab values of the following parameters on ED admission: sodium, potassium, bicarbonate, lactate, blood glucose, INR, aPTT, WBC, RBC, platelets; along with pre-hospital glucose values in the field.

Study cohort comprised of 612 adult subjects arriving at ED with history of fall with median age of 49yrs (IQR of 28-68). In-hospital death (IHD) was observed in 5% (N=28) of the total cohort, with Male IHD=19 (68%), and Female IHD=9 (32%). Older age groups [>55yr] (8%, n=252) had higher incidence of In-hospital deaths compared to younger age group 18-55yr (2%, n=360) with p=0.0004 (CI= 0.0267-0.0933). In a univariate regression model higher levels of: Pre-hospital Glucose (p=0.0048), ED Blood Glucose (p<0.0001), Lactate (p=0.0445), INR (p=0.0048), aPTT (p<0.0001) and WBC (p<0.0001) were significant individual predictors of In-hospital death. While lower levels of bicarbonate (p<0.0001) and RBC (p<0.0001) were significant individual predictors of In-hospital death. The following multivariate regression models showed statistical significance with higher probability of In-hospital death: 1) Higher: ED blood glucose (p=0.0014), aPTT value (p=0.0035) | Lower: bicarbonate (p=0.0482), RBC (p=0.0087) with Adj.R 2 =19% 2) Higher: aPTT value (p=0.0019) and WBC count (p=0.0035) 3) Higher: aPTT value (p=0.0112), Older age (p=0.0033) | Lower RBC (p=0.0220) and GCS scores (p<0.0001) with Adj.R 2 =29%

Lab parameters such as ED blood glucose, RBC count, WBC count, bicarbonate level & aPTT level are individually or simultaneously important predictors of In-hospital death in adult TBI patients with history of fall. 

Traumatic brain injury (TBI) is an epidemic with severe consequences. Brain tissue oxygen tension (PbtO2) monitors detect secondary injury and direct clinical therapies to mitigate damage. Blood transfusion is one therapy often used, however its effect in TBI patients is not well defined. We studied PbtO2 data in patients who received transfusion after TBI.

Sixty-nine severe TBI patients were consecutively admitted to a neurocritical care unit and received PbtO2 monitoring as part of standard clinical care for this unit. Data were collected from electronic medical records as entered by the bedside nurse. Patients were managed according to the Brain Trauma Foundation guidelines. Transfusions were identified through nursing comments. Hourly PbtO2 values were analyzed for up to six hours after starting transfusion. Other factors were also analyzed for their potential influence on PbtO2 readings.

Of 69 patients, 19 received a total of 28 transfusions in the setting of PbtO2 monitoring. Two groups were identified: transfusions that led to an increase in PbtO2 and those that did not. Six transfusions resulted in increased PbtO2, with an average increase of 11.5 mm Hg. Twenty-two transfusions did not: of these 12 were unchanged and 10 decreased. The groups did not differ in age (mean 33.5 and 33.6, respectively), maximum temperature (mean 98.7 and 97.6), minimum cerebral perfusion pressure (mean 61.2 and 60.0), or initial Glasgow Coma Scale (GCS) (mean 4.3 and 5.2).

Blood transfusion is often used in the critical care setting. The effect of transfusion on brain tissue oxygen tension is variable. Age, temperature, cerebral perfusion pressure, and initial GCS were not useful in distinguishing patients who benefited from transfusion. PbtO2 only rises in a minority of patients; therefore additional prospective studies are needed to evaluate which patients are likely to benefit from transfusion.

Mannitol use in patients with traumatic brain injury can lead to acute renal failure and may worsen outcome. The purpose of this study is to determine the rate of acute renal failure (ARF) among patients treated with Mannitol and its impact on outcome in a multicenter review.

We analyzed a one-year data (2005) (2006) from the PREMIER database, a nationally representative hospital discharge database in the United States. We compared baseline and clinical characteristics of patients with traumatic brain injury (TBI) treated with Mannitol in the first 7 days of admission who developed ARF to those who didn't. Length of stay, cost of hospitalization and discharge status were ascertained.

From a total of 2388 admissions with a diagnosis of traumatic brain injury requiring Mannitol within the first 7 days of admission, 8% (N=178) of patients had ARF. 

ARF is a common complication of TBI treated with Mannitol. It is associated with longer length of hospital stay and increased rates of in-hospital mortality. The result highlights the importance of using alternative therapy to hyperosmotic agents such as hypertonic saline in treatment of TBI patients at risk for acute renal failure.

Cervical Spine Immobilization (CSI) is a relative contraindication for percutaneous dilatational tracheostomy (PDT) because of the inability to extend the neck, making tracheal puncture at the correct level more challenging. Patients with CSI routinely undergo PDT at our institution, however, with both traditional bronchoscopic as well as with real-time ultrasound (RTU) guidance. Our objective was to review the incidence of complications related to PDT in patients with CSI versus patients without CSI.

We reviewed the records of consecutive patients who underwent PDT performed by a single operator at our neurocritical care unit between 6/2008-5/2012. All patients requiring tracheostomy are screened for eligibility for PDT by the attending neurointensivist on service. We recorded the percentage of patients who successfully underwent PDT vs requiring conversion to surgical tracheostomy, the specific guidance used (bronchoscopy, RTU) and all short-and long-term complications including placement of the tube above the first tracheal ring.

A total of 112 patients underwent PDT performed by a single neurointensivist. All patients screened by the operator underwent an attempt at PDT, and all patients successfully completed the procedure without conversion to surgical tracheostomy. Ninety-eight of 112 (87%) did not require CSI and 14 (13%) required CSI. In the CSI group, bronchoscopy alone was used in 6/14 (43%) and bronchoscopy plus RTU in 8/14 (57%). No complications occurred in the CSI group. In the no-CSI group, there were 3 (3%) complications (one tracheal granuloma and two tube dislodgments within 7 days). No other short or long term complications were recorded. All tubes were placed below the first tracheal ring.

It is feasible to safely perform PDT in patients with Cervical Spine Immobilization using bronchoscopic and real-time ultrasound guidance. 

Following traumatic brain injury (TBI), increased serum biochemical marker levels reflect the extent of neurological damage, prognosis and clinical outcomes. Effective TBI management strategies are lacking. Despite the neuroprotective effects of therapeutic hypothermia after cardiac arrest, its TBI use remains controversial. Delays in achieving target temperatures in human trials taking 6-12 hours (NABISH-I; NABISH-II) may have contributed to the lack of benefit. We hypothesized prompt, rapid induction of hypothermia, immediately following TBI would lower predictive serum biomarkers of brain injured swine.

Sixteen domestic cross-bred pigs (34-35 kg) were subjected to a 5 ATM (100 ms) fluid percussion TBI. Eight injured animals were cooled to 32°C within 90 minutes of injury and maintained for 48 hours using transpulmonary hypothermia. Eight control animals were maintained at 37°C using similar doses of inhalational and intravenous general anesthesia. Brain temperature was monitored with Camino.® Serum markers of TBI: S-100 calcium binding protein B (S-100B), Neuron-specific enolase (NSE), Glial fibrillary acidic protein (GFAP) and Phosphorylated axonal form of the neurofilament subunit NF-H (pNF-H) were measured prior to injury and seven times over 96 hours. Surviving animals were euthanized and necropsied five days post-injury.

At 3, 6, 24 and 96 hours, S-100B, NSE and pNF-H, were lower in the hypothermia group vs. controls. GFAP levels were decreased at 96 hours. After injury, peaks and troughs of the biomarkers occurred at various intervals. S-100B levels were reduced in both groups during the initial 36 hours post-injury, with control levels increasing at 48 hours.

Early initiation and rapid cooling of brain temperature to 32-33°C for 48 hours was associated with attenuated S-100B, NSE, GFAP and pNF-H levels in swine. General anesthesia was associated with early mitigated S100B levels. Prompt therapeutic hypothermia and prolonged anesthesia may offer neuroprotection after TBI. 

Mild traumatic brain injury (mTBI) from blast exposure represents a significant threat to military personnel. Until now there has been no way of knowing what the individual service member experienced during an exposure. We report the first individual measurements recorded during combat operations and how those readings were used to assist evaluation of the injured service member.

The NATO ROLE-3 hospital, Kandahar Afghanistan received the index case of a service member (SM) exposed to an improvised explosive devise (IED) blast while wearing a blast dosimetry system composed of 3 Blast Gauges placed on the back of the helmet, chest, and shoulder. The gauges include status lights that allow immediate feedback for injury risk via colored lights: green = negligible (<4 PSI peak), yellow = moderate (between 4 and 16 PSI), and red = severe (16 PSI and above). In addition, time traces of the overpressure and 3-axis acceleration are recorded and available for download through a micro-USB port. The SM's gauges were initially checked 1 hour 24 minutes after the blast, demonstrating a yellow status light.

The blast data downloaded from the gauges demonstrated a consistent exposure of 5.5 msec composed of a primary flow immediately followed by a secondary wave. The head gauge recorded a peak overpressure of 13.1 PSI and impulse pressure of 9.5 PSI-sec. There was 1 msec of sustained pressure above 4 PSI from the primary flow. All 3 gauges demonstrated similar blast profiles, including a secondary reflective wave.

These measurements are firsts in both the recording of an individual's exposure during a blast related attack and the use of that data for patient triage and medical evaluation. Blast Gauges measure environmental exposure and do not diagnose mTBI, however; they do provide clinicians with important information in the evaluation of patients subjected to blast.

To consider the definition of initial signs and symptoms to compare outcomes after "severe" Traumatic Brain Injury regard to mechanism of injury.

Design-This study included all adult patients who presented to ED at a level-1 trauma center with severe (GCS score<9) traumatic brain injury.

From the total cohort(n=126), 47% suffered TBI because of "fall" and 54% due to traffic accident(MVC). Significant proportion of each sub-group was comprised of males (72% in-MVC with median-age=39; 74% in-falls with median-age=53) · For all the patients arriving to ED after a traffic accident with severe GCS: 88% had LOC, 62% had AOC, 15% had PTA, 100% got admitted to hospital, 94% had an abnormal head CT (bleed/fracture), 91% got admitted to ICU, 49% had some neuro-surgical intervention and 29% patients died in hospital. · For all the patients arriving to ED after a history of fall with severe GCS: 87% had LOC, 70% had AOC, 23% had PTA, 97% got admitted to hospital, 93% had an abnormal head CT (bleed/fracture), 90% got admitted to ICU, 42% had some neuro-surgical intervention and 35% patients died in hospital. · Decrease in systolic blood pressure (P=0.348) and increase in diastolic blood pressure (P=0.0329) are more likely to have a fracture after a traffic accident in severe TBI. Increasing of blood pressure p=0.0256) and decreasing of pulse (p=0.0381) is significantly associated with ICU admission after a fall.

Comparing data for two most common mechanism of injury in severe TBI suggest that some vital signs and symptoms have significant impact with outcomes depends on mechanism of injury. These observations should be studied in larger cohort to find more significant association between mechanism and outcomes. 

Cerebral edema is the one of the most significant predictors of poor outcome after traumatic brain injury. It is still unclear what the pathophysiological and cellular mechanisms and predictors of post-traumatic edema are. The exponential growth in genetic information has opened an avenue for investigation in traumatic brain injury and implicated specific genes in the pathophysiology of post-traumatic injury edema. Two examples are the Aquaporin-4 and CACNA1 genes, which respectively encode water and calcium channels. The Aquaporin-4 gene on chromosome 18q11.2-12.1 encodes the Aquaporin-4 protein (AQP4) water channel. AQP4 is one of the bidirectional high capacity water channels that is primarily expressed in astrocytic foot processes in the central nervous system at the blood-brain barrier and is thought to be critical for brain water homeostasis. Experimental studies showed that AQP4 deficient mice had significantly reduced cerebral edema and better survival in a water intoxication model. The CACNA1 gene on chromosome 19p13 encodes the a1A subunit of a neuronal calcium channel. Patients with Familial Hemiplegic Migraine and delayed fatal cerebral edema and seizures from minor trauma have been found to have mutations in CACNA1, which are hypothesized to enhance development of cytotoxic edema. A missense mutation is reported to enhance risk of delayed fatal cerebral edema. Hypothesis: The CACNA1 gene missense mutation S218L and AQP4 polymorphisms will be over-represented in patients with post-traumatic cerebral edema.

To perform full exon sequence analysis of these two genes in 20 well-defined cases of excessive cerebral edema. Our long term goal is to systematically investigate genetic variants as determinants of risk of excessive cerebral edema.

Patient recruitment is currently ongoing.

It is hoped that this will further elucidate secondary mechanisms of injury specifically in the formation of post-traumatic edema and lead to targeted therapies in the future.

Microwave occurs when Improvised Explosive Devices was exploded. However, the effect for brain by Microwave has not been clarified.

Under general anesthesia, S-D rats were irradiated by head-focused microwave by Microwave fixation system (Model MMW-05/ Muromachi Kikai Co., Ltd.), which were classified in three groups (3.2Kw/ 0.1sec (I), 2.6Kw/ 0.1sec (II), 2.0Kw/ 0.1sec (III), and sham group) by intensity (n=3 in each group). Vital signs were evaluated, Arterial blood gas was examined, and we checked pathologic findings by Hematoxylin-Eosin (HE) stain immediately after Microwave irradiation, post 3 hours, 6 hours, 24 hours, 72 hours, 1 weeks, and 2 weeks in each group.

Blood pressure was elevated transiently immediately after irradiation, and recovered in short period. PaO2 was unchanged in post-irradiation phase, except in Group I. In HE stain, Neuron was degenerated and left out especially in cerebral cortex and hippocampus, microglia cells were accumulated in these regions. These pathological changes were observed frequently and earlier, when irradiation was intense.

The result was firstly reported that head-focused microwave irradiation induced brain injury in S-D rats, and this brain injury was related with intensity of microwave. Pathological change was impressive because it was occurred gradually and progressive. Further study will be required, whether this type of brain injury is similar with traumatic brain injury, or cerebral ischemia or not, and the study of behavioral effects of microwave irradiation is necessary, especially when the intensity of irradiation was not severe.

The efficacy of decompressive craniectomy (DC) in the treatment of moderate-severe traumatic brain injury (msTBI) is a topic of debate in Neurocritical care. Despite the recently published randomized DC in diffuse TBI (DECRA) trial, it is still unclear when and for which TBI patients this procedure should be considered. In order to assess the utility of DC in evidence-based clinical practice, we present a matched case-control study that compares surgical and non-surgical outcomes among patients with msTBI.

We conducted a retrospective analysis of msTBI injuries treated at a single Level I trauma center from 2009 to 2012. Twenty msTBI patients aged between 18 and 67 years, who underwent DC, were enrolled. Paired controls that underwent medical therapy only were selected according to Glasgow Coma Scale (GCS) score and age. Primary lesion type, pupil reactivity, hypotension, hypoxia and ICP crisis were secondarily considered in matching cases with controls. We focused on mortality, Glasgow Outcome Score (GOS) score upon hospital discharge and GOS score at 3 months as the primary measures of outcome.

In the DC group, we found that 35% of patients died; 15% had a favorable outcome at discharge (4 or higher on GOS); and 10% had favorable outcome at 3 months. In the control standard-care group, we found that 30% of patients died; 25% had favorable outcome at discharge; and 35% had favorable outcome at 3 months. Pupil reactivity and GCS score on admission were the variables highly correlated with mortality. Statistical analysis will be available at the meeting and presented for the first time.

In this cohort, undergoing DC did not seem to confer a mortality benefit to patients with msTBI. Good recovery after msTBI was observed in a larger percentage of the non-surgical group, which is consistent with the findings of the DECRA trial.

Each year in the United States, over 1.4 million patients present to Emergency Departments as a result of traumatic brain injury (TBI). Severity classification of TBI is based on the Glasgow Coma Score (GCS), with severe TBI being a GCS score between 3 and 8. There is always a subset of "severe" TBI that requires surgical intervention. The current study examines this subgroup to decipher any symptomatology that may be helpful in identifying who these patients are. The objective is to determine which if any factors predict the need for surgical intervention in patients with severe traumatic brain injury (TBI).

This study is a subgroup analysis of the larger cohort of consecutive adult TBI patients that presented to the ED. Our sample included only severe TBI patients (GCS<9). Besides descriptive analysis, logistic regression analysis was done to determine the significant predictors of surgical intervention in this subset of patients. Lab values (sodium, potassium, bicarbonate, lactate, blood glucose, WBC, RBC, platelets, INR, aPTT) and symptoms (such as-seizures, vomiting, loss of consciousness, alteration of consciousness, post-traumatic amnesia) were the dependent variables compared with surgical intervention (independent variable).

Of the total severe TBI cohort (n=126), 40% required surgical intervention. Presence of abnormal head CT (bleed in 93% of the total cohort) is significantly associated with surgical intervention (p=0.0131). Vomiting (p=0.0173), Lactate (P=0.0326), higher WBC (p=0.0400) and lower platelet count (P=0.0430) individually showed significant association with surgical intervention on a univariate regression model.

These data suggest that abnormal head CTs, particularly those that result from bleeding, as well as Lactate, platelet count, WBC count and vomiting are significantly associated with surgical intervention. The association of lab values with likelihood of undergoing surgical intervention is an interesting future research point.

To study the potential usefulness of initial vital parameters and laboratory evaluations to predict short term prognostic

This is an observational cohort study of all adult patients who came to the emergency department(ED) of a tertiary care hospital, in a 20 month period during 2008-2010. For the purpose of analysis, we considered initial vitals and lab values available for all patients. We individually compared vitals (pulse and mean arterial pressure-MAP) and laboratory values [Sodium(Na + ), Potassium(K + ), Bicarbonate(HCO 3 -), glucose, WBC, RBC, Platelets, INR) for the following prognostic variables: abnormal head CT finding(yes/no), hospital admission (yes/no), ICU admission, in-hospital death, hospital length of stay(HLOS), and 3-month mortality using t-tests and correlations. The significant variables were then entered into a logistic regression model (for categorical variables) and a multiple regression model (for continuous variables) simultaneously to determine significant predictors of prognostic outcomes. Significance level was set at p=0.05.

Increase in glucose(p=0.004) and WBC(p<0.001) lead to a higher likelihood of having an abnormal head CT, when controlling for MAP and HCO 3 -; Increase in glucose (p=0.001) and WBC (p=0.02), and decrease in HCO 3 -(p=0.009) and platelets(p=0.006) increases the likelihood of getting admitted, when controlling for MAP, K + , and RBC; Increased glucose(p<0.001), decreased HCO 3 -(p=0.01) and decreased platelets(p=0.03) increases chances of in-hospital death, when controlling for WBC and INR; increased glucose(p=0.004), increased WBC(p<0.001), and decrease in RBC(p=0.005) increases HLOS, while controlling for pulse, K + , and HCO 3 -.

The study indicates that some initial vitals and lab values can help to determine the prognostic outcomes in adult traumatic brain injury. Though our study is limited by a single-site patient population, the interesting findings warrant further research efforts in this specific area. 

Ultrasonic assessment of optic nerve sheath diameter (ONSD) as a non-invasive measure of intracranial pressure (ICP) has been evaluated in the literature as a potential valid technique for rapid ICP estimation in the absence of invasive intracranial monitoring. The technique can be challenging to perform and little literature exists surrounding intra-operator variability. In this study we propose an examination of ONSD utilizing a variety of novel ocular models, to both define the ability of the ultrasound linear array probe to capture different known ONSD, and to assess intra-operator variability with the technique. Here we present the model and data.

We designed ocular models composed of gelatin spheres and variable three dimensional printed cylinders, which simulate the globe of the eye and variable ONSD's respectively. These models will then be suspended in a gelatin background. Operators will then utilize the linear array ultrasound probe on these models in order to determine ONSD of 3 sizes, with 10 measurements each in order to assess intra-operator variability with the technique.

Our optic nerve sheath model offers ultrasound images comparable to in vivo, and is quick to manufacture. Analyzing the data, we removed the first two measurements from the series of ten. We defined those as "practice attempts" with the technique. For the ONSD models, the means were: 2.59mm with SD of 0.34mm (95% CI of 0.24), 5.55mm with SD of 0.33mm (95% CI of 0.23), 8.18mm with SD of 0.34mm (95% CI of 0.24).

Utilizing the standard linear array ultrasound probe for ONSD measurements in our model provided reliable results with minimal intra-operator variability across variable sheath sizes. Knowing this, we can further apply this novel model of ONSD to US teaching and training courses with confidence in its ability and the techniques ability to produce consistent results.

The objective of this study was to identify factors (signs and symptoms after injury, vital parameters, and glucose) that can be used as predictors of an intracranial bleed. This will improve identification and treatment of patients who present to the emergency department (ED) with TBIs.

This is an IRB approved observational cohort study done at a level 1 trauma center, and included all adult patients presenting to the ED following a TBI. Data for patients presented during Study variables included age, loss of consciousness (LOC), seizure (SZ), vomiting, alteration of consciousness (AOC), post-traumatic amnesia (PTA), glucose level, pulse, and blood pressure (BP). All variables were tested for association with intracranial bleeding using chi-square tests of independence, t-tests, and then significant variables were included in a regression model. Limitation of study were chart review and a single ED.

The cohort consisted of 1,241 patients, of which 92% (n=1124) had a CT scan of head. 43% of total patients had an abnormal head CT, and 80% of those had an intracranial bleed. Statistical analysis indicated that LOC (P=0.0005), AOC (P=0.0001), PTA (P<0.0001) and advanced age (P=0.0002) were significantly correlated with having a bleed. Vomiting and SZ were not statistically significant. Among patients who had head CTs, both pulse and systolic blood pressure decreased between the first and second measurements; both pulse(p=0.001) and BP(p=0.045) decreased significantly less in patients with bleeds compared to those without bleeds. Additionally, higher ED glucose level was associated with having a bleed (p<0.0001) on head CT.

These data indicate that older age, LOC, AOC, PTA, and elevated glucose levels can be used as predictors of intracranial bleeds. Sustained elevation of pulse and systolic blood pressure may also be indicative of a bleed in TBI patients. 

Patients with moderate-severe traumatic brain injury (msTBI) commonly die from withdrawal of support, likely as a consequence of an unfavorable outcome prognosis provided to the family by the treating physician. It is unknown whether prognostication may lead to self-fulfilling prophecies, and whether the presence of intensive care unit (ICU) complications may accentuate possible provider bias. In this study, we surveyed clinicians caring for patients with msTBI to examine the variability of outcome prognostication and the influence of ICU complications on these predictions.

We conducted an anonymous electronic survey of clinicians, including faculty members (Neurology, Neurosurgery, Trauma, Anesthesia/Critical Care), neurology house staff, ICU affiliate practitioners and neuroICU nurses at a single Level I trauma center. The survey included three TBI case vignettes and their respective ICU courses. Questions were designed to assess the utilization of known TBI prognostic models, relative importance of ICU complications for outcome prognostication and aggressiveness of care recommended by the survey participant.

A total of 152 surveys were distributed by email or paper, and 51 have been returned. So far, we have found that 51% of participants consider medical ICU complications as very important in TBI prognostication. Age, ICU course and head CT findings are the prognostic variables considered most important to outcomes. 68% of non-critical care neurologists are uncomfortable providing TBI prognostication. Case responses suggest that clinicians tend to recommend aggressive care (surgery), but predict unfavorable outcomes. The survey is ongoing, but complete results will be available at the meeting and presented for the first time.

We have discovered great variability in outcome predictions made by clinicians with different levels of experience in treating msTBI. Self-fulfilling prophecies may exist among msTBI outcomes. Outcome studies should focus not only on admission variables, but also on ICU complications in order to guide clinicians in providing prognostication.

The objective of this study was to identify pre-hospital factors that are associated with worse severity of head injury in order to help physicians identify when TBI treatment may be necessary.

This is an observational cohort study that included adult patients presenting to the(ED) following a motor vehicle collision. Study variables included age, gender, seatbelt use, loss of consciousness (LOC), seizure, vomiting, alteration in consciousness (AOC), and post-traumatic amnesia (PTA). Severity of TBI was classified according to the Glasgow coma scale, with mild defined as 13-15, moderate being 9-12, and severe being anything less than 9. The GCS was obtained both in the pre-hospital and ED settings.

The cohort of 451 was 57% male. The median age was 34 (IQR: 23-49, R: 18-91). The breakdown of severity in the prehospital setting (n=290) was 74% mild, 7% moderate, and 19% severe. In the ED (n=380), the breakdown was 79% mild, 4% moderate, and 17% severe. Pre-hospital factors significant for the Z-test included seatbelt (SB) use, LOC, AOC, PTA, and gender. Males, patients who did not wear seatbelts, and patients who had a positive LOC, AOC, or PTA were more likely to sustain a moderate or severe TBI. Having a seizure was also significantly associated with increased TBI severity (P=0.001). (See Table 1 ) Additionally, the data show that the likelihood of having an abnormal head CT increases with age (P<0.0001). Although vomiting was associated with greater TBI severity, the results were not statistically significant.

Early symptoms such as LOC, AOC, seizures, and PTA are early predictors of worse severity in patients who sustain a head injury during their motor vehicle collision. Age, male gender, and lack of seatbelt use also correlate with greater TBI severity.

Identifying crucial symptomatic predictors of ICU admissions, ICU length of stay and Mortality rates in Traumatic Brain Injury (TBI) patients with history of fall.

Retrospective chart analysis was performed on all adult patients arriving to emergency department with history of fall at a level one trauma center for parameters like vomiting, alteration of consciousness (AOC) & loss of consciousness (LOC) after TBI; post-traumatic amnesia (PTA) and history of seizures before or after injury, along with outcomes such as ICU admission & ICU length of stay.

From the total cohort (n=612), 26% (N=139) of patients were admitted to ICU, most of them were males(73%,p=0.0017). AOC was found to be strongly associated with ICU admission (57%, p<0.0001)[including the patients who had brief loss of consciousness of <30mins(52%)], and 3month mortality rates(p=0.0287) when adjusted for mild GCS scores. · ICU length of stay was higher in patients admitted to ICU with AOC (p=0.0482) and PTA (0.0076). ICU admissions had higher 30day readmission (p=0.0006), In-hospital death (p<0.0001) and 3month mortality rate (27%, p<0.0001). · 86% of patients were found to have intra-cranial bleed when presented to ED with AOC(p=0.010), and 75% of these patients were admitted to ICU. On a multivariate regression model analysis, patients who had abnormal head CT with mild GCS on ED presentation had higher 3month mortality rates (p=0.0066) when adjusted for age.

Patients with symptoms such as alteration of consciousness and post-traumatic amnesia after traumatic brain injury as a result of fall are more likely to be admitted to ICU with significantly longer ICU length of stay. Mild traumatic brain injuries in fall patients should not be overlooked in daily practices because of significant mortality rates.

Cardiovascular disturbances remain a leading cause of morbidity and mortality in patients with acute spinal cord injury (ASCI). ASCI patients often develop symptomatic and potentially life-threatening bradycardia. Our practice has been to use albuterol elixir prophylaxis in ASCI patients, taking advantage of its side effect profile associated with a typical dose of 4 mg tid or qid, to prevent further symptomatic bradycardia. Evidence of efficacy with this regimen is, however, lacking. We set out to determine whether treatment with oral albuterol would decrease the frequency of bradycardic episodes in patients with ASCI.

We retrospectively identified adult patients admitted to University of New Mexico Hospital between 2006-2011who sustained an ASCI and received oral albuterol therapy. The frequency of bradycardic events (HR <60 bpm) before and after initiation of albuterol was collected. We compared the number of bradycardic events before and after albuterol within each subject using the Wilcoxon Signed Rank Test. Bootstrap methods were used to further validate our findings.

We identified 11 ASCI patients who had evidence of symptomatic bradycardia before the initiation of the albuterol therapy, including hypotension and in 2 cases bradycardic cardiac arrest. The median number of bradycardic events was 17 (6.5, 42 IQR) before albuterol and was 4 (0, 11 IQR) after albuterol. We found that patient's had a significantly lower number of bradycardic events after the initiation of albuterol (p = 0.012). Ten patients experienced less bradycardic events. The median difference was 12 less bradycardic episodes. Bootstrap estimates of the median difference were consistent with our initial analysis.

Albuterol appears to be an effective means of treating bradycardia in patients with acute spinal cord injury.

Severe Traumatic Brain Injury (TBI) is frequently associated with EEG changes like epileptiform discharges, seizures; periodic lateralized epileptiform discharges PLEDs or paroxysmal delta activity. We report a case of TBI with generalized 3Hz spike and wave pattern that did not represent seizures

A 21y old girl without Epilepsy history presented after being involved in a Motor accident. Initial GCS was 10 and remained the same over the next 2 days. CT showed Contusions with small Left subarachnoid hemorrhage. Phenytoin was started for seizure prophylaxis. On day 3, she improved clinically, however, on day 5 she had fluctuating consciousness and Continuous EEG monitoring was initiated. Various antiepileptic medications were tried over the next several days including Lacosamide, valproate, topiramate, levetiracetam and ethosuximide (ETH) without significant change clinically or on EEG. She started improving clinically on day 7 but became extremely drowsy on day 8, all meds except ETH were weaned. She showed improvement and was discharged to rehab on day 14. A prolonged EEG after 2 months was normal and ETH was weaned off. She continues to do well almost one year after and is maintaining her school grades at pre-injury level.

The patient's initial EEG (day 5 post injury) showed generalized 3Hz spike and wave pattern occurring every 1-3 seconds which continued for 2 days despite treatment with various Anti epileptics as described. On day 7 EEG pattern changed to generalized Rhythmic delta activity(1-2Hz) especially during arousal. MRI during the stay showed micro hemorrhages in both frontal lobes and right temporal lobe reflective of Diffuse Axonal injury.

A 3hz spike and wave pattern mimicking Absence Seizures can be seen on EEG transiently after TBI, however its clinical significance is unclear. Whether it needs to be aggressively treated or not cannot be conclusively established but the longterm prognosis appears to be benign. 

Free radical-induced lipid peroxidation (LP) has been demonstrated to lead to the formation of isoprostanes from arachidonic acid and neuroprostanes from docosahexaenoic acid. LP is common after traumatic brain injury (TBI) and constitutes one of the key mechanisms of pathology related to secondary injury after TBI. One of the consequences of LP is the compromise of neuronal calcium (Ca ++ ) homeostasis, leading to Ca ++ overload and activation of the proteolytic -spectrin. The purpose of this project is to characterize the concentration--spectrin degradation after TBI.

This study is a prospective, single-center study of adult moderate to severe TBI patients. Inclusion criteria are age >18yo, closed head injury, within 24 hours of TBI, and Glasgow Coma Score (GCS) <12. Serial samples from urine, blood, and cerebrospinal fluid (CSF, when available) are obtained for up to 2 weeks after injury. Demographic data and pertinent clinical information are also collected. The biomarkers (5 & 15 F 2t -isoprostanes, F 2 -isofurans and F 4 -neuroprostanes) are measured via -spectrin breakdown products (SBP) by western blot analysis.

We have enrolled fifteen patients to date. Preliminary results suggest that the study population is typical of TBI (mean age 30.4 years, 73% male, median admission GCS 7). Serum and CSF 5 & 15 F 2t -isoprostane values are above published values for normal individuals, with CSF values peaking at 24 hours after TBI. SBP are also measured in elevated amounts in CSF compared to non-TBI controls (in whom they are not measurable).

Preliminary data suggests that serum and CSF isoprostane values are elevated after TBI. Continued patient accrual, further sample analysis, and comparison to control groups is needed to more precisely define the effect of TBI on the time course of LP biomarkers.

Traumatic intraventricular hemorrhage (tIVH) is generally considered to be associated with moderate to severe traumatic brain injury and a significant mortality rate. There exists, however, a rare subset of individuals who manifest with isolated traumatic intraventricular hemorrhage and have a good prognosis and outcome. We present a case of an 18-year old female who suffered polytrauma and an isolated ventricular hemorrhage following a traumatic fall while mountain climbing. Her history indicated mild transient confusion and amnesia occurring around the time of the fall. Her Glasgow Coma Score was 15, her neurologic exam was normal and she had no neurologic complaints other than positional lightheadedness and nausea. A comprehensive exam was notable for a right hip dislocation, nasal fracture, L4 vertebral body fracture, right apical pneumothroax and pulmonary contusion. Computed tomography of the head showed an acute hemorrhage in the left lateral ventricle prompting concerns for traumatic brain injury. No additional pathology was noted on a follow-up magnetic resonance imaging. Repeat CT scan showed mild interval decreases in the size of her ventricular hematoma. The patient was discharged one week after admission and had developed no neurologic complications. She was diagnosed with concussion and isolated intraventricular hemorrhage.

Isolated intraventricular hemorrhage is a rare complication of traumatic head injury that can have a good prognosis and outcome. The case shows the difficulty in categorizing this particular condition within the current spectrum of traumatic brain injury and specifically highlights shortcomings with classification systems that utilize neuro-imaging abnormalities to determine severity of injury.

Traumatic intracranial aneurysms (TIAs) are distinctly uncommon, comprising fewer than 1% of all cerebral aneurysms. TIAs that develop following blunt head injuries present the clinician with both diagnostic challenges and clinical difficulties. The natural histories of giant intracranial aneurysms are generally grave owing to mass effects, severe hemorrhage, and distal thromboembolism.

Case report.

We present the case of a 68-year-old male was involved in an accident in which he suffered severe head injury from a falling heavy iron hammer. The immediately unenhanced head computerized tomography showed hemorrhagic contusions, subarachnoid hemorrhage, skull fracture and basal fracture. He had been in a deep coma ever since. The computed tomographic angiography (CTA) revealed a giant aneurysm of right internal carotid artery about one month after the blunt head injury. The aneurysm was measured 6.5cm at its maximal diameter on image. Of note, the patient failed to improve the following day and died on the fiftieth hospital day.

Giant TIAs are very rare but fatal complications of blunt head injury probably related to effects of vessel wall trauma and possibly a combination of neurological deterioration. In our case, the involved mechanism was suspected to be related to skull base fractures or resulted from stretching of the artery across the process during the impact. CTA has a high sensitivity of about 97.2% and a high specificity of about 97.9% for diagnosing cerebral aneurysms (including traumatic aneurysms). Apart from this, CTA permits 3-dimensional visualization of aneurysms and assesses surrounding intracranial structures that are not visible on DSA. Therefore, although 2-dimensional digital subtraction arteriography is currently the diagnostic gold standard in cerebral aneurysmal disease, fast and noninvasive CTA may be preferred in the acute setting of TIAs.

Julio Cabrera 1 , Corina Puppo

Major burnt patients require large volumes of fluid replacement due to a generalized increase in permeability and edema caused by cytokines. Fifty percent of the administered fluids produce edema in "preserved" tissues. Multiple organ edema follows fluid replacement. Escharotomy is frequently performed to decompress limbs and thorax, but not neck. Our objective was to describe and diagnose neck-head compartment syndrome in patients with neck circumferential burns and/or neck edema by 1) suspectng it and 2) confirming diagnosis with the help of Transcranial Doppler (TCD) Ultrasonography, searching for a high resistance pattern in cerebral blood flow velocity at basal cerebral arteries.

TCD examination was performed before and after escharectomy in two 

Both patients presented a neck-head compartment syndrome, evidenced by the cerebral hemodynamic repercussion of neck compression: hypoperfusion with an increased resistance pattern in DTC. P1: secondary compartment syndrome due to massive fluid replacement; without circumferential burn. P2: compartment syndrome in circumferential neck burn. TCD confirmed the clinical suspicion of cerebral hypoperfusion, guiding the decision to perform surgical decompression to treat it, and helped to assess the results of the decompressive surgery.

Introduction 23.4% hypertonic saline is used for the treatment of increased intracranial pressure (ICP) and in the prevention and reversal of brain herniation syndromes. The use of hypertonic saline in the management of combat related penetrating and severe traumatic brain injury is described. 23.4 % hypertonic saline effectively managed ICP with decreased risk of hypovolemia and secondary hypotension compared with Mannitol. 23.4% hypertonic saline also preserved cerebral blood flow, decreasing the risk for secondary cerebral ischemia in acute neurotrauma patients, where hyperventilation is contraindicated.

The NATO hospital, Kandahar Afghanistan treated eleven (11) patients with twenty-seven (27) doses of 23.4% saline from 3-March 2012 to 25-April 2012. Hypertonic saline was used to treat acute elevation in ICP, as well as to maintain an elevated serum sodium concentration during periods of cerebral edema. All patients were treated with initial conservative ICP management. External ventricular drains were placed and drainage of 5-10cc of CSF was performed in an attempt to maintain ICP before using hypertonic saline. Patients with life-threatening clinical signs of elevated ICP secondary to brain edema or acute neurologic deterioration were potential candidates for 23.4% hypertonic saline therapy. 30-60mL of 23.4% sodium chloride was administered via a central line infusion.

23.4% hypertonic saline was successful in acutely reducing ICP. A 30ml bolus of 23.4% saline predictably increased the serum sodium levels allowing reliable titration and maintenance of serum sodium levels and efficient management of the patient's volume status (30cc of 23.4% = 250cc of 3%).

Penetrating and severe closed head injuries have the potential to lead to neurologic emergency as a result of brain edema associated with primary TBI or following neurosurgical intervention. In a combat TBI population, 23.4% hypertonic saline demonstrates a clinical benefit over alternative treatments by decreasing the risk of secondary cerebral injury during the management of elevated ICP and was well tolerated.

Unintentional death was the ninth leading cause of death among elderly patients. Given their comorbidity profile, many of them are also on antiplatelets or anticaogulants. We sought to characterize the burden of "pro-bleeding" medications such as antiplatelets and anticoagulants in the population aged over 55 who sustain a head injury.

This observational cohort study was conducted at a level one trauma center that has a 12 county catchment area serving over 1 million. The trauma acuity is high, with over 50% of our patients haveing ISS scores over 15. The age cutoff of 55 for "elderly" is based on our trauma alert activation criteria.

Thirty-nine percent of the cohort was on at least one type of anticoagulant or antiplatelet, as follows: warfarin 8%, aspirin 21%, clopidogrel 10%, ASA+dipyridamole 1%, heparin/LMWH 3%.A third of the cohort required ICU admission. ICU length of stay ranged from 1-37days. Patient in particular, on warfarin had a significantly longer ICU length of stay (p=0.05) when adjusted for INR level. The median INR for the whole cohort was 1.0 with an IQR of 1.0 to 1.20. The median INR amongst those on warfarin was 2.3 with an IQR of 1.7 to 2.6. Patients on an antiplatelet or anticoagulant agent were significantly more likely to have an abnormal head CT (p=0.05). 11% of the patients who were on warfarin needed some sort of anti-coagulant reversal to minimize bleed. Patients on warfarin were more likely to undergo neurosurgical intervention (p<0.0001) when compared to cohort not on warfarin.

Antiplatelet and anticoagulant drugs can confer additional morbidity to persons who sustain a TBI. It may be important to recognize this early, and prepare for higher level care needs. 

Introduction Therapeutic hypothermia (TH) is know to cause immune suppression. Determining the degree of immune suppression at the bedside is often difficult or impossible. Immune Cell Function (ICF) measures the concentration of ATP from circulating CD4 cells following in vitro stimulation with phytohemagglutinin (PHA) as an indicator of immune cell function. ICF is often used in solid organ transplant programs to modulate the immunosuppressive treatment. We propose the use of IFC to determine the degree of immune depression in the patient treated with TH.

Immune Cell Function, Cylex Inc, Columbia, MD was obtained in three populations of patients: Group 1: Patients treated with TH, IFC obtained while at target temperate, 33 degrees C Group 2: Patiients that were admitted to the care of the Neurocritical care team, requiring ICU care. Group 3: Patient from Sanford Renal transplant program with stable immunosuppressive therapy.

The average ICF of group 1 were 195, of group 2, 316 and of group 3, 197.

Patient being treated with TH have a profoundly depressed ICF. The level of immunosuppression is equal to if not greater that those with solid organ transplants. According to the Cylex data a level of less than 225 represents an immune suppressed state. This does not appear to be a phenomenon of the critically brain injured patient since those without TH had a normal ICF While further studies are in process, this data has effected out practice. We now treat patients on TH as immunosuppressed patients. 

Very early prediction of neurological outcome after cardiac arrest (CA) remains challenging. Several single center studies have suggested that bispectral index (BIS) can predict outcome for patients treated with therapeutic hypothermia (TH). We evaluated the ability of BIS to predict outcome in a multicenter study.

3 medical centers prospectively enrolled comatose CA patients treated with TH. Outcome was defined as good (GO) if cerebral performance category (CPC) score was 1-2, and poor (PO) if CPC 3-5 at hospital discharge (HD) and at 6 months (6M). BIS data was assessed blind to outcome for initial value after first dose of neuromuscular blockade (NMB -BISi) and at 6 hours post-ROSC (BIS6).

256 patients were enrolled with a mean age of 60 (SD 16) years, 70% were male, 68% witnessed, initial rhythm was VT/VF in 48%, PEA in 26%, asystole in 23%, and time to ROSC was 24.7 (16) minutes. At HD, 71 (28%) had GO with similar age as PO but shorter median time to ROSC at 15 (IQR 10-24) mins vs 25 (14-34, p=0.003). GO patients also had more VT/VF as initial rhythm and witnessed CA (p<0.001), and more males (p=0.01). On ROC curve comparisons, both BISi (AUC 0.80) and BIS6 (AUC 0.80) performed better than time to ROSC (AUC 0.65) or age (AUC 0.50) -p<0.002 for all comparisons.

Among CA-TH treated patients, this is the first multicenter trial to confirm that bispectral index values after first dose of NMB and at 6 hours post-ROSC predicted outcome better than time to ROSC, rhythm, or age. BIS appears promising as a tool to predict outcome very early after CA, and may be helpful during clinical trials to stratify the severity of brain injury sustained during CA.

Hypotension negates the cerebral protective effect of therapeutic hypothermia (TH). Myocardial depression, "cold-induced diuresis," and hypokalemia can lead to refractory hypotension during the maintenance phase of TH. Intravascular volume replenishment and inotropic infusion are effective but cause wide swings in heart rate, blood pressure, cardiac output and acid-base status. We propose the use of vasopressin as a physiologically appropriate agent to correct hypothermiainduced hypotension. Hypothesis: In swine, the investigators tested the hypothesis that an infusion of vasopressin would restore blood pressure to normal levels during TH.

Six domestic cross-bred pigs (34-35 kg) were subjected to a 5 ATM fluid percussion injury to the brain followed by systemic hypothermia (32°C) for 48 hours. The animals were turned side to side and to sternal recumbency every six hours. During phase I (first 24 hours), the blood pressures were maintained in the normal range with intermittent doses of epinephrine and fluid boluses. During phase II (second 24 hours), continuous vasopressin infusion (0.04 ug/min) was added to maintain blood pressure. The number of episodes of hypotension (MAP <55 mm Hg), the volume of fluids (liters), and the total dose of epinephrine (mg) used during both phases were compared using Student's paired t-test (P>0.05).

In all animals, the infusion of vasopressin effectively mitigated the occurrence of hypothermia-induced hypotension. The episodes of hypotension (8.3±2.1 v 4.0±1.7), the total volume of fluids (15.3±2.1 v 10.6±2.3), and the total dose of epinephrine (2.6±1.1 v 0.6±0.3) administered were significantly reduced during phase II.

In order to maximize the benefits of TH, hypotension must be avoided. Animal studies show that despite hypothermia, hypotension causes cerebral cortical tissue depletion of ATP and phosphocreatine and an increase of lactate and NADH levels. The infusion of a low dose of vasopressin reverses these anomalies and effectively mitigates hypotension. 

Hypotension, hyperoxia, and hypoxia early after the return of spontaneous circulation (ROSC) are each associated with increased mortality, while early hypertension is associated with good outcome. We assessed these variables and their relationship to outcome in cardiac arrest (CA) survivors treated with therapeutic hypothermia (TH).

With IRB approval, we reviewed prospective and retrospectively collected data in a single-center database of patients undergoing TH after CA. Demographics and clinical factors were compared among patients with CPC 1-2 (good outcome) and CPC 3-5 (poor outcome) in a bivariate model. Various definitions of hypotension, hypertension, hypoxia, and hyperoxia were evaluated. We constructed logistic regression models including potential confounders and the variables of interest.

Among 265 patients, age, VT/VF rhythm, shorter time to ROSC, witnessed arrest, bystander CPR, and STEMI on initial ECG were each strongly associated with good neurological outcome, as were a lower peak neuron-specific enolase level and higher bispectral index (BIS) score after neuromuscular blockade. Hyperoxia (PaO 2 > 300mmHg) was common (present in 45.5 with good and 35.9 with poor outcomes, respectively) as were hypoxia (PaO 2 <60mmHg) and hypotension. None of these factors was a predictor of outcome. Logistic regression models intended to adjust for the potential confounding influences of age, time to ROSC, heart rhythm, witnessed arrest, and bystander CPR, also did not identify a relationship between the variables of interest and outcome.

Our data did not confirm the previously described relationship between post-resuscitation factors and outcome. This may reflect an inadequate sample size, but it is also possible that post-resuscitation hemodynamic and biochemical factors are minimally important to outcome, compared to the duration and type of the arrest. Further investigation in larger data sets is warranted.

Determining the presence of an infectious process during therapeutic hypothermia (TH) can be difficult. In addition, differentiating central vs systemic fever is difficult in the brain injured patient. Procalcitonin (PCT) was been used to guide the use of antibiotics in sepsis and pneumonia in patients that are critically ill. We propose the use of PCT to predict the presence of a systemic infection in patients during TH.

All patients treated with TH had PCT measured at the start of TH. All patients were cooled with the Medivance Arctic Sun 2000. When the water temperature was <10 degree C, PCT and two sets of blood cultures (BC) were drawn. Sputum cultures (SC) were obtained if there was a change in sputum or during bronchoscopy. Antibiotic use was determined by the Neuro-intensivist Results 14 patients were evaluated; 2 ICH, 7 TBI, 1 CVA and 4 cardiac arrest (CA). A total of 49 PCTs were obtained. One patient (7%) had positive BC, PCT of 2.48; 8 patients (57%) had positive SC. Remaining 5 patients had negative BC and SC. All 4 CA patients had increased PCT > 1.49 (normal <0.08) of which 3 (75%) had positive SC and none had positive BC. Of the remaining 9 without positive BC (64%), 5 (55%) had positive SC, all had PCT <0.39. Of the 4 (44%) patients without positive SC, all had PCT <0.25

PCT is a reliable method to exclude an infectious process in patients being treated with TH that have not had a CA. While further studies are warranted, a PCT <0.25 appear to exclude both pulmonary and blood infections, while a PCT <0.39 appears to exclude a blood stream infection. From this data, PCT is not a good marker for infection in the CA patient.

Therapeutic Hypothermia (TH) has become widely accepted practice for neuroprotection and improved mortality in comatose survivors of out of hospital V-fib cardiac arrest. Evaluation for appropriateness of TH is now part of ACLS algorithm. Its use in non-shockable rhythms such as PEA and asystolic arrest is less well established. We present our center's experience with TH after cardiac arrest and review the clinical and electrophysiological parameters that may impact prognosis.

This is retrospective review of medical charts including patients undergoing TH after cardiac arrest at a single center from 2010 through the first quarter of 2012. Demographic and clinical data were collected. Continuous EEG results were reviewed by two independent epileptologists who were blinded to the outcome of the patients. EEGs were graded based on the Synek scale for grading severity of EEGs. Patient's neurologic outcome will be assessed by grading Cerebral Performance Category (CPC) score at the time of discharge. Multivariate regression analysis will be performed on the data to identify parameters that would affect prognosis in cardiac arrest after cooling.

Fifty-eight patients were identified from our database. The overall rate of survival to discharge was 45%. The survival rate for V-fib arrest was 80% whereas the survival rates for asystolic arrest and PEA arrest were 50% and 33%, respectively. Results from the multivariate analysis will be forthcoming.

Our results affirm the predominant view that TH indeed improves outcomes after cardiac arrest. In particular with ventricular fibrillation and pulseless ventricular tachycardia arrest, we have seen very encouraging results. Patients with PEA/asystolic arrest fared worse but outcomes are still improved compared to historical control.

Since 2002, mild therapeutic hypothermia (MTH) has been the standard of care when spontaneous circulation returns after a witnessed, out-of-hospital ventricular fibrillation arrest[1]. At our institution, we have initiated MTH for approximately fifty patients since February 2010. A knowledge, attitude, and practices survey was conducted querying neurology residents and attendings, emergency medicine (EM) residents and attendings, and internal medicine (IM) residents. Our aim was to identify areas of weakness so that we could strengthen the overall awareness of the utility and benefit of MTH.

The survey consisted of nineteen multiple choice questions, ranging from asking how many times the participant had initiated MTH; to parameters for the protocol; to how it impacts survival. The surveys were completed by: ten neurology residents and five neurology attendings; twelve EM residents and two EM attendings; and twenty IM residents.

All of the neurology residents and EM physicians surveyed had been the primary provider for a post-arrest patient who underwent MTH. The neurology residents unanimously agreed that MTH after resuscitation from a shockable rhythm is standard of care, however only 86% of EM physicians and 63% of IM residents agreed. 38% of EM physicians and 15% of IM physicians answered that MTH may be initiated in cases presenting after either a shockable or a non-shockable rhythm. 60% of the participants acknowledged that ventricular fibrillation portends the most favorable outcome. Nearly 80% of participants agreed that ideal ROSC is less than thirty minutes. Three-quarters of physicians indicated the goal temperature as 32-35°C; however, half of the neurology residents and 33% of neurology attendings answered this incorrectly.

In conclusion, this survey has revealed a general understanding of MTH, however, each specialty has its deficiencies. We can now educate each subset of physicians in a problem-focused manner. 

Early quantitative assessment of non-contrast brain computed tomography (CT) using specialized software correlates with outcomes of cardiac arrest survivors. 1 The proposed algorithm compared Hounsfield units (HU) in the putamen (PU) to the posterior limb of the internal capsule (PLIC), but the work has not been validated in patients treated with therapeutic hypothermia (TH) or using standard software and equipment.

We included CA survivors treated with TH who underwent CT in the first 72h after resuscitation (ROSC). HU were averaged bilaterally at two levels in the PU and PLIC, and the PU/PLIC ratio calculated by a board-certified radiologist using a GE LightSpeed VCT 64 slice scanner and AGFA PACs system. Receiver-operator characteristic (ROC) curves were constructed, evaluating PU or PU/PLIC to predict poor outcome (CPC 3-5) at hospital discharge (HD) and 6 months (6M).

84 patients had median age 58 years, 66% male, 81% out-of-hospital CA, 68% witnessed, 41% VT/VF, 32% PEA, and 26% Asystole. Median (IQR) time to ROSC was 17 (10-28) minutes. 59/84 (70%) patients had PO. When stratified by outcome, CT performed 2.4 (1.2-10) hrs after ROSC showed similar HU measurements for PLIC (26.3 PO vs 26.7 GO, p=0.6) but lower HU in PU (32.7 vs 34.1, p=0.01) and PU/PLIC (1.24 vs 1.27, p=0.02). HU values for PU and PU/PLIC both predicted outcome: ROC area under the curve (AUC) for PU = 0.68 (95%CI 0.55-0.80) and PU/PLIC = 0.66 (0.54-0.79). Among 78 patients with 6M outcome data, PU predicted outcome (32.7 PO vs 34.0 GO, p=0.035) with AUC =0.66 (0.53-0.80), but PU/PLIC did not.

Early after CA, Hounsfield unit measurements in the putamen, and the PU/PLIC ratio were lower among patients with poor outcome, but the magnitude of the differences was small, and clinical utility uncertain. Additional study is warranted.

Global cerebral edema following aneurysmal subarachnoid hemorrhage (aSAH) is associated with 50% in-hospital mortality. Therapeutic hypothermia (TH) is recommended for reduction of intracranial pressure (ICP) based on class I evidence; however safety in prolonged states remains poorly studied.

We retrospectively reviewed all cases of refractory ICP elevation at the Mayo Clinic Florida neurointensive care unit (NICU) from 2008-2012 who received adjunct TH for more than 48 hours. Primary safety endpoints were QTc prolongation, development of bacteremia, and coagulopathy. Additional outcomes included in-hospital mortality, hospital/NICU length of stay, and functional status at 6 months.

17 patients with aSAH and/or intracerebral hemorrhage underwent adjunct TH. Median age was 44; 8 were male. On admission, median APACHE2 was 16, and WFNS was higher than 3 in 14, all being modified Fisher 3-4. 10 required barbiturates in addition to sedation, paralysis and hyperosmolar therapy. TH was initiated on a median of hospital day 2 and continued for a median of 7 days (minimum=4, maximum=23). Mean ICP over 24 hours prior to TH was 15.6mmHg(SD=3.8; range 10.6-24.2), decreasing to 11.1mmHg(SD=3.9, range 6.1-21.1) over the first 24 hours of TH. 16 patients had external ventricular drains placed and 10 required decompressive craniectomy on average day 6 hospital stay (range 0-13). Safety data showed Torsades-de-pointes in 1, mean QTc prolongation of 59 with mean lengthening of aPTT by 2.7. 4 patients had bacteremia on admission with new infections (urine, sputum, blood) documented in 6 during TH. Overall, 11 (65%) survived to discharge. Median NICU/hospital length of stay was 27/32. Average modified Rankin score at follow up was 3.4.

Hypothermia greater than 48 hours as an adjunct to standard ICP reducing therapies appears feasible in patients with refractory intracranial hypertension. However, definitive safety of prolonged TH would require direct comparison with similar cohort. 

Refractory raised intracranial pressure (RICP) secondary to intracerebral hemorrhage (ICH) and severe subarachnoid hemorrhage (SAH) is a life threatening condition. Treatment for RICP typically induces hypothermia (TH) and decompressive hemicraniectomy (HCT). However, direct comparison of the efficacy of these two therapies is lacking. Data from this study may help determine the sequence of therapies that might improve outcomes in this patient population. In the present study using retrospective design, we tested the hypothesis that for patients with RICP, TH is as effective in reducing ICP as HCT, using functional outcome at discharge as defined by modified Rankin Scale (mRS) as the primary outcome.

We retrospectively reviewed all adult patients admitted to the Neurointensive care unit from 2009 to 2012 with SAH and ICH with resultant elevated ICP, who survived the first 24 hours after admission. Exclusion criteria included: pupillary anisocoria, limitation of care within 24 hours of admission; or hemicraniectomy or craniotomy with clot evacuation prior to ICP monitoring were excluded.

Initial review included 11 patients (TH=4 and HCT=7). Based on univariate analysis, admitting GCS score was higher with HCT (5 vs 8, p=0.13), but other baseline demographic and clinical characteristics were similar. TH group had longer ICU LOS (19 vs 9), LOS ventilation (19 vs 9), and higher cost. However, discharge mRS (5 vs 5,p=0.46) was similar.

Our initial analysis indicates longer ICU care and overall cost with TH, but similar functional outcomes at discharge. Subsequent analysis will include inclusion of additional patients, ICP comparison and adjustment for baseline characteristics.

Malignant middle cerebral artery(MCA) infarction is devastating ischemic stroke, which the mortality rate is up to 80%. Therapeutic hypothermia is one of the most promising neuro-protective therapies. Successful result of hypothermia for cardiac arrest renewed interest in therapeutic hypothermia for stroke. The purpose of this study was to assess whether therapeutic hypothermia can reduce the cerebral edema and can improve the functional outcome in patients with malignant MCA infarction.

We reviewed retrospectively patients with malignant MCA infarction presented within 24 hours of symptom onset in a single center hypothermia registry. After informed consent, patients who had refused decompressive hemicraniectomy were treated with therapeutic hypothermia and monitored in the neurocritical care unit for complications. A modified Rankin Scale(mRS) and National Institutes of Health Stroke Scale(NIHSS) were obtained at 3 months after symptom onset.

Eleven patients with a mean age of 70±8 years and an NIHSS score of 18.5±4.3 were treated with therapeutic hypothermia(33±1 ). Seven of eleven patients were MCA infarction, and four was ICA T-occlusion. The mean time from symptom onset to initiation of hypothermia was 33.0±21.5 hours and the total duration of hypothermia was 103.1±66.6 hours. Noncritical complications included shivering(n=11), bradycardia(n=7), hypertension(n=7), pneumonia(n=6), and arrhythmia(n=5). Electrolyte imbalances were common during the hypothermia (hypernatremia;n=4, hypokalemia;n=9, hypophosphatemia;n=11). Mortality rates was 18%(n=2) and the mean NIHSS at discharge was 19.2±13.5. The mean mRS at 3 months was 4.5±1.4 in all patients and 4.2±1.4 in survivals.

This result shows that therapeutic hypothermia can prevent the progression of cerebral edema and improve functional outcome in acute malignant MCA infarctions and ICA T-occlusion. Long duration hypothermia more than 3 days appears feasible and safe in these patients. Therapeutic hypothermia may be a good alternative therapeutic option to early decompressive hemicraniectomy. Large clinical trials are needed whether hypothermia will be a best treatment to improve functional outcome.

Therapeutic hypothermia (TH) is know to cause immune suppression. Determining the degree of immune suppression at the bedside is often difficult or impossible. Immune Cell Function (ICF) measures the concentration of ATP from circulating CD4 cells following in vitro stimulation with phytohemagglutinin (PHA) as an indicator of immune cell function. ICF is often used in solid organ transplant programs to modulate the immunosuppressive treatment. We propose the use of IFC to determine the degree of immune depression in the patient treated with TH.

Immune Cell Function, Cylex Inc, Columbia, MD was obtained in three populations of patients: Group 1: Patients treated with TH, IFC obtained while at target temperate, 33 degrees C Group 2: Patiients that were admitted to the care of the Neurocritical care team, requiring ICU care. Group 3: Patient from Sanford Renal transplant program with stable immunosuppressive therapy.

Group 1, 13 patients, average ICF: 195 Group 2: 12 patients, average ICF: 316 Group 3: 22 Patients, average ICF, 197.

Patient being treated with TH have a profoundly depressed ICF. The level of immunosuppression is equal to if not greater that those with solid organ transplants. According to the Cylex data a level of less than 225 represents an immune suppressed state. This does not appear to be a phenomenon of the critically brain injured patient since those without TH had a normal ICF While further studies are in process, this data has effected out practice. We now treat patients on TH as immunosuppressed patients.

Therapeutic hypothermia (TH) has become a first-line therapeutic modality in patients suffering from traumatic brain injury and cardiac arrest. Shivering induced by TH reduces the ability of the cooling device to achieve target temperature. This can lead to increased intracranial pressure (ICP) and increased metabolic demand. The Bedside Shiver Assessment Score (BSAS) has been validated in identifying and grading shivering. However, the BSAS cannot identify microshivering which is visually undetectable shivering that is thought to have the same detrimental physiologic consequences as shivering. Continuous 21 channel EEG (cEEG) can detect microshivering but is labor intensive, requires specialized training to interpret results and is expensive. We propose that the Philips EEG with Compression Spectral Array 2 lead (Philips 2) can be utilized to detect microshivering as effectively as cEEG but is more cost effective.

The Philips 2 was placed by the bedside nurse. The lead placement varied depending on underlying injuries. Patients were assessed utilizing the BSAS and the Philips 2. If high frequency activity increased on the Philips 2, the patients were assessed using the BSAS. If the BSAS was 0 then 5-10mg of Vecuronium was given to intubated, sedated patients. Both patient temperature and water temperature were recorded.

Two patients with TBI were evaluated. The water temperature decreased and the patient's temperature increased during the periods of high frequency activity on the Philips 2. After vecuronium, the high frequency activity ceased, water temperature increased and core temperature returned to the previously set level.

The Philips 2 is a relatively low cost device when compared to cEEG that can be applied and monitored by the nursing staff to detect microshivering. Additionally, we were able to validate that control of microshivering improved the TH device's ability to achieve and maintain the patient's temperature goal.

Therapeutic hypothermia is widely accepted as a standard of practice for Out of Hospital cardiac arrest (OHHCA). However, its implementation is still highly variable in different hospital settings. Most of the current data comes from centers of excellence. We wanted to evaluate performance of implementation of "Hypothermia Protocol" (HP) including its complications and outcomes in our large referral community based hospital.

We conducted retrospective chart review of 30 patients who underwent HP from 2008-2011. Data collected included demographics, time of cardiac arrest, time of arrival to ER and time to induction of HP, methods used for induction, complications and outcomes.

Out of the 30 patients, 16 patients (53%) had pulse less electrical activity (PEA), 13 (43.3%) patients had ventricular tachycardia/fibrillation, and 1 (3.33%) had complete heart block as the initial rhythm. Average time to arrive to ER was 30 minutes. Almost 60% of patients had HT induction in ED, 35% (%) in ICU and 6.66% outside of the hospital. Average time to initiate HT from the initial event was 1hour and 56 minutes. Average time to achieve the target temperature from the initial event was 2 hours. Inner cool was the most common modality used in 46.6 (%). Lactic acidosis (19.7%) was the most common complication encountered, followed by Hypotension (12%), Coagulopathy (12%) and Seizure (12%) Trend of improved outcomes with less renal failure, coagulopathy, seizure was observed with shorter induction times. Time to achieve target temperature had no effect. Initial rhythm, age and gender also had no impact on the outcome.

Shorter induction time appears to decrease complications and improve outcomes. Using multiple cooling modalities also appeared to have better outcomes. However larger studies are needed to confirm this observation. Earlier induction of Mild Therapeutic hypothermia improves survival and neurological outcome and decreases incidence of some of the complications.

Introduction Secondary brain injury after aneurysmal subarachnoid hemorrhage (aSAH) is a major cause of mortality. Mild hypothermia (32-33 0 C) may protect against cerebral ischemia and edema in aSAH patients. The aim of this study is to describe the use of CT perfusion (CTP) characteristics to initiate re-warming in patients with secondary brain injury after aSAH.

We performed a retrospective review of all patients admitted to a large comprehensive stroke center between 2000 and 2012 with aSAH who were treated with hypothermia and received CTP imaging. Mild hypothermia (32-33 0 C) was started because of severe vasospasm, increased intracranial pressure or cerebral edema. Baseline characteristics, including clinical severity grading by Hunt Hess (HH) and Fisher scales, were collected. Clinical outcomes were measured by discharge modified Rankin score (mRS) and disposition. CTP was performed with a 64-slice scanner.

Twenty patients fulfilled inclusion criteria. In 18/20 (90%) patients, re-warming was based on favorable CTP characteristics and in 2/18 (10%) based on favorable TCD findings. The mean duration of hypothermia was 10.5 days. Five patients were re-warmed due to normal CTP, despite TCD findings suggesting moderate to severe vasospasm. 15 patients, re-warming was initiated given improving TCD findings and despite less favorable CTP data (most showing "matched" abnormalities of decreased CBV, CBF and increased MTT). Clinical outcomes were worse in this group; mRS

Better outcome was seen in all patients in whom re-warming was initiated based on normal CTP. In these patients, there was a discrepancy between CTP and TCD data. Poor outcome was associated with abnormal CTP regardless of TCD findings. CTP may be a useful tool to guide treatment of aSAH patients receiving hypothermia.

Diagnosis of pediatric brain death (PBD) continues to be a significant challenge. New guidelines for PBD diagnosis were published in Pediatrics in 2011. We recently conducted a mailed survey to assess current understanding of these new guidelines and general perspectives about PBD among a convenience sample of Midwest USA physicians.

We developed a 23 item survey. Items included 6 demographic questions, 1 question about familiarity with the guidelines, and 16 questions concerning perceived discrepancies and other attitudes toward the guidelines. We mailed our survey to 319 physicians at 10 university hospitals: 102 pediatric intensivists, 143 neonatologists, 13 adult neurointensivists, and 61 pediatric neurologists, Three weeks after the initial mailing, we followed up with a reminder by mail and/or phone. We performed Fisher's exact test to assess statistical significance of responses among different specialties.

After 10 weeks, we had a 17% response rate. Respondents included 8 pediatric neurologists, 2 neurointensivists, 26 pediatric intensivists, and 20 neonatologists. Twenty percent of respondents were unfamiliar with the new PBD guidelines (neonatologists were least familiar). Twenty-three percent stated they were 'not comfortable' making a PBD diagnosis and 85% deemed it was either preferable or essential to obtain a neurointensivist or pediatric neurology consultation for PBD assessment. There was general agreement that the current intervals for the required 2 exams were appropriate in children (delineated by age). Interestingly, 66% allowed patients to remain ventilated for a significant period of time after PBD declaration.

We found that a significant number of pediatric physicians are not familiar of the new PBD guidelines and there remains some variability in the assessment of these patients. Pediatric neurologists or neurointensivists are still considered an important part of the process of PBD determination.

The mid-position fixed pupil (MPFP) is an imperfect reference to the mid-size pupil that occurs with the complete loss of neural influence from devastating midbrain injury (primary or secondary) and death (brain and cardiopulmonary). For this reason, proper recognition and interpretation of the MPFP is critical to the neurological localization/diagnostic process and a vital element to the clinical verification of brain death. While the description of the size range of the MPFP has been dogmatically passed down from numerous classical texts (4-7 mm) for decades, it has not been accurately quantified. Modern pupillometry offers accurate quantification of pupil size.

Using a portable infrared pupillometer (Forsite, NeuroOptics Inc., Irvine, CA), within 24 hours after death, we evaluated the pupil size of dead patients who did not have any previous eye surgery, known eye disease, or use of eye medications.

50 pupils were evaluated in 25 dead patients (mean age 66) an average of 13 hours after death. The pupil size range was 2.26 -5.46 mm, with a median size of 4.21mm (SD of 1.8mm). 30/50 pupils (60%) were <4 mm and none were > 5.5 mm. 8/25 patients (32 %) had a side-to-side difference of at least 0.5 mm. Thankfully none were reactive!

The MPFP is generally smaller than classically described and 95% fall between 2.31 and 6 mm. 60% of MPFP's are less than 4mm. We never found any MPFP's more than 5.5 mm. Subtle but frequent side-to-side asymmetry (> 0.5 mm) existed in approximately 30% of the dead patients. With our continued work we can finally achieve a more quantitative description of the important finding of the MPFP so that it can be incorporated into our definitive texts, enveloped into our understanding, and applied to our clinical practice.

Brain death diagnosis is clinical in Uruguay. It is defined as the irreversible loss of brain stem functions. Ancillary tests are needed as confirmatory tests in selected cases: 1) impossibility or contraindication to perform clinical testing (barbiturates, facial trauma, etc.); 2) non demonstrable structural lesion; 3) unknown coma etiology; 4) difficulties to wait for a second clinical test. The most used confirmatory test is Transcranial Doppler (TCD) ultrasonography. Objectives: To study 1) the clinical characteristics of patients in whom brain death could not be diagnosed clinically; 2) TCD ultrasonographic patterns; 3) Number of cases in which TCD aided in management.

Epidemiologic and observational study. Patients included: those in who brain death was suspected but the clinical examination of brain stem reflexes and/or apnea test could not be performed for different reasons. Period: From 2000 to 2010. The variables studied were demographic and clinical characteristics, TCD sonographic patterns. Cerebral Circulatory Arrest was diagnosed when the patterns found were systolic spikes, reverberating flow, and no-flow (if a previous study had demonstrated ultrasound permeability of skull windows) in bilateral anterior sectors and posterior sector. Continuous flow or systolic peaks were negative for the diagnosis of CCA.

445 patients in who the clinical diagnosis of brain death was not possible or needed to be confirmed. 75% adults. 62% were men, with an average of 38 y.o. in adults, and 5 y.o.in children; structural etiology 85%. Etiology: traumatic 45%; vascular 30%; anoxic-ischemic 15%, infectious 6%, toxic-metabolic 1%, other 3%. CCA was confirmed in 75%, systolic spikes in 84%. CCA was discarded in 20%. In this group the study was repeated in 9%, confirming CCA in 75%. It was not concluding in 5%.

DTC helped in the decision to how to continue the management of the patient in 95% of the cases, diagnosing CCA in 81%.

There is an awkward physician and cross-institutional variability in the approach to brain death (BD) diagnosis and all of its ramifications; physiological, logistical, and psychosocial. Physician variability is related, in part, to a basic knowledge deficit and inexperience. However, public confidence in the reality of BD relies on consistent and accurate diagnosis and the physician's facility with the management of its implications.

Our full-day (8 hour) brain death simulation workshop (BDSW) was designed to enhance confidence with BD diagnosis and management. It included a didactic lecture and seven learning stations: case study analysis (recognizing brain death mimics), a high fidelity mannequin simulation (BD examination including cold water calorics and apnea testing, hemodynamic management, and diabetes insipidus management), family discussions with professional actors trained to provide feedback, and four relevant content stations. Each participant was observed by a neurocritical care expert, each receiving one-on-one and group feedback.

23 physicians participants from 3 continents participated in the BDSW with 12 expert faculty. All participants felt much more confident with brain death diagnosis and management. At least 60% were humbled by the station on "discussion of brain death with families", recognizing their need to practice communicating about brain death effectively. 90% felt better equipped to contemporize their local policies and advocate for enhanced uniformity of practice.

Our BDSW provides a model comprehensive training experience that had a favorable impact on trainee confidence and their interest and capacity to advocate for better uniformity and training of peers. 2. The BDSW can be part of a future tiered approach to credentialing experts in this important clinical area. 3. We are conducting the 2 nd BDSW on November 12, 2012 with improvements based on the 1 st workshop. Neurocritical care experts must embrace the primary responsibility for preserving the integrity of BD diagnosis and educating our colleagues.

The use of carbogen in apnea testing to declare brain death may facilitate achieving the prerequisite PCO2 needed to confirm apnea testing by establishing a target end point that is typically reached faster and has been shown to limit adverse effects. As the use of Extracorporeal Membrane Oxygenation (ECMO) in adults increases, so does the need to perform apnea testing while on ECMO. However, traditional apnea testing on critically ill patients is compounded by lung derecruitment and hemodynamic instability rendering an aborted apnea test or worse, cardiac arrest and death. The literature on apnea testing of patients on venous-arterial (VA)-ECMO is minimal.

Per hospital protocol, a carbogen mixture (97% oxygen and 3% carbon dioxide) was delivered through the ventilator for an apnea test on a 25 year old female on VA-ECMO. The ventilator's mandatory rate was set at 2 breaths/minute to adequately deliver the carbogen mixture through the artificial airway. A carbogen formula was used to calculate a target end-point of an EtCO2 of 70 mmHg for a positive apnea test. An ABG was drawn prior to the apnea test and again once the target EtCO2 was achieved.

Pre-apnea ABG: 7.40/36/322/98%. The EtCO2 goal was reached within 8 minutes and the post-apnea ABG was drawn: 7.22/61/86/93%. The patient remained hemodynamically stable throughout the apnea test which was confirmed as a positive apnea test.

The use of carbogen in apnea testing on a patient receiving VA-ECMO demonstrates the possibility of performing a successful apnea test for declaration of brain death. Although more investigation is needed, this case demonstrates the ability to perform apnea testing on critically ill and unstable patients while maintaining hemodynamic stability which preserves the option for organ donation.

Drowning victims have historically been eliminated from consideration for lung donation as aspiration may cause direct pulmonary toxicity, often confounded by significant neurogenic pulmonary edema. A significant minority of these patients (10-15%), aspirate only minimal amounts of water into their lungs, protected by severe-persisting laryngospasm (dry drowning), but progress to brain death due to significant anoxic injury. Historically, even with limited evidence of aspiration,transplant centers do not consider evaluating drowning victims as lung donors. However, as the division between the number of eligible recipients and available donor organs continues to grow, criteria for acceptable donor organs are expanding. Once an absolute contraindication for lung donation, this practice has persisted on a per case basis but is reported infrequently with somewhat mixed results.

We analyzed the UNOS registry of Donors for Lung and Heart-Lung Transplant from January 1, 1991 to December 31, 2011 (n=53), and then examined survival outcomes from lung transplant recipients from donors who suffered drowning between 2001 to 2011 recipients (n = 39) to outcomes previously reported from lung transplant recipients during that period.

For recipients of lungs from donors with drowning as cause of death, unadjusted survival at one 

Drowning victims, even when initially resuscitated, often suffer significant anoxic injury and death by neurologic criteria. While the management of drowning victims as organ donors may present additional challenges, with proper donor selection, the use of lungs recovered from carefully screened donors after drowning appears to be a safe option for the expansion of the donor pool.

Racial disparity in health care utilization and outcomes is an area of substantial concern. A study performed in the 1990's in our Neuro-ICU found that nonwhites were half as likely to withdraw life-sustaining therapy (WLST). This may be explained by differences in socioeconomic status (SES), cultural preference, lack of end-of-life planning, or trust in the health-care system. To better understand the basis and evolution of this disparity, we analyzed it over two more recent epochs (determining whether it has improved over time), while specifically accounting for SES.

We extracted data from a prospective Neuro-ICU database on all ventilated patients with GCS of 8 or less between 2002 and 2009. We analyzed how the rate of WLST was affected by age, race, gender, insurance and socioeconomic status (quintiles based on median household income of residence zip code), marital status, receipt of surgical/ICU interventions, GCS and APACHE II. We then compared SES-adjusted disparity for WLST (non-whites vs. whites) in 2002-2004 with 2007-2009 .

Non-whites accounted for 715 of 2062 patients (36%) and were younger, less likely to be married (29% vs. 54%), insured (28% vs. 37%), and reside in upper-income zip codes (all p<0.001). Rate of WLST was lower in non-whites (21% vs. 30%, p<0.001), despite comparable overall hospital mortality. After controlling for SES and other confounders, non-white race was still associated with lower odds of WLST (aOR 0.61, 95% CI 0.47-0.79). This disparity was prominent in the earlier epoch (aOR 0.45, 0.28-0.74) while race was no longer a statistically significant marker in the more recent cohort (aOR 0.75, 0.52-1.09).

Race appears to influence the likelihood of WLST in severely brain-injured patients independent of SES. This disparity, which has been attenuated over the past decade in our ICU, may be related to cultural differences or barriers relating to end-of-life planning or trust.

Multiple parameters have been associated with outcome in comatose post-cardiac arrest patients. Anecdotal observations suggest that patients who are cooler upon ED arrival tend to have poorer outcomes; if arrival temperature correlates with outcome, it may serve as an additional tool for patient prognostication.

We performed a retrospective analysis of a prospectively collected data set from 80 comatose post-cardiac arrest patients to determine if a relationship exists between arrival temperature and outcome. Of the 80 patients, 38 patients (48%) with out-of-hospital cardiac arrests and with arrival temperatures recorded prior to initiation of hypothermia treatment were included and divided into those with good outcomes (16 subjects; mRS =>3) or poor outcomes (22 subjects; mRS =<4 or death) at 6 months; 6 subjects (3 poor outcome survivors and 3 who progressed to brain death) remained when patients whose poor outcome (death) was due to withdrawal of care were removed from the poor outcome group.

Analysis using a two-tailed unpaired t-test on 16 subjects with good versus 6 with poor outcomes demonstrated a significant difference in temperature on ED arrival: mean temperature of patients with good outcomes was 36.2 o C (SD=0.9 o C), while that of patients with poor outcome was 34.8 o C (SD=1.6), p=0.01. When patients who died due to withdrawal of care were included in the analysis, a strong trend in difference between the two groups remained, but was not statistically significant (p=0.07).

Low body temperature upon ED arrival correlates with poor outcome in post-cardiac arrest patients and may serve as an additional prognostic variable. Cooler temperatures may merely reflect longer lapsed time before return to normal circulation; alternatively, they may be a result of poor temperature regulation in more severely brain injured patients. Further investigation of this issue with a larger patient pool is warranted.

Diencephalon injury (DI) has been described in neurocritical care. Consciousness alterations (CA), dysnatremia, hemodynamic instability, fever, muscle dystonia are signs of DI. These symptoms are non-specific. The goal of the study was to describe structure of acute diencephalon dysfunction syndrome (ADDS) on the model of isolated acute DI.

This retrospective study evaluated all patients operated in 2006-2010. Inclusion criteria: adult patients in stable preoperative condition; sellar region tumors (SRT); complicated postoperative period. Exclusion criteria: intra-cranial complications, not related with direct DI (epi-, subdural hematomas, brain ischemia). Organ dysfunctions and dysnatremia were registered.

83 patients were included, 7 excluded. All had CA and dysnatremia. Hemodynamic dysfunction developed in 52 patients, respiratory dysfunction in 50 patients, ileus in 51 patients, thrombocytopenia in 21 patients, renal dysfunction in 3 patients, hepatic dysfunction in 1 patient. There were 5 groups. First (n=12) had CA, dysnatremia. ICU LOS was 8.5 days. Glasgow Outcome Scale (GOS) 4 had 11 patients; GOS 3: 1 patient. Second group (n=11) had CA, dysnatremia, one somatic organ dysfunction (SOD). ICU LOS was 13 days. GOS 4 had 5 patients, GOS 3: 6 patients. Third group (n=12) had CA, dysnatremia, two SOD. ICU LOS was 23.5 days. GOS 4, 5 had 3 patients; GOS 3: 5 patients; GOS 1: 4 patients. Fourth group (n=25) had CA, dysnatremia, 3 SOD. ICU LOS was 22 days. GOS 4, 5 had 6 patients; GOS 3: 8 patients; GOS 1: 11 patients. Fifth group (n=12) had CA, dysnatremia, 4 SOD. ICU LOS was 40.5 days. GOS 4 had 4 patients; GOS 3: 1 patient; GOS 1: 7 patients. Sixth group (n=4) had CA, dysnatremia, 5 SOD. ICU LOS was 7.5 days. All died.

ADDS consists of CA, dysnatremia, and at least one SOD. Severity of ADDS depends on number of SOD.

Intracranial pressure (ICP) monitoring is widely used in the management of patients with traumatic brain injury. ICP monitoring may also be useful in other situations characterized by high ICP, including cardiac arrest survivors (CAS) after return of spontaneous circulation (ROSC). However, no prospective study has examined the incidence of raised ICP among CAS. This pilot study will examine the feasibility of screening for elevated ICP in CAS admitted to the Toronto Western Hospital (TWH) in 2013 -using the non-invasive technique of optic nerve ultrasonography (ONUS) --to identify patients with elevated ICP, who might benefit from invasive ICP monitoring to optimize their management after they survive cardiac arrest.

Evidence of elevated ICP will be examined by 2 blinded ultrasonographers(USF) who will measure the optic nerve sheath diameter (ONSD) in both eyes of all CAS every 12 hours from ROSC. All findings will be defined in a dichotomous method (elevated/not elevated).

Primary outcome: Incidence of major protocol violations, defined as the inability to attain 5 of 6 ONUS recordings during first 72 hours at the specified time point (every 12 hours) by each USF. For every major protocol violation, an audit will be done to understand the reason for the violation and tailor the protocol to improve compliance in future studies.

Advances in resuscitation medicine have demonstrated an improvement in patient outcomes in CAS by the implementation of TH. The exact mechanism of action of TH is not well understood and has been postulated to partially involve a decrease in ICP. No prospective data currently exists linking TH with ICP. Using ONUS as a non-invasive modality, we have designed a single centre feasibility study to assess the ability of ONUS to measure ICP in CAS, as well as to aid in sample size calculations for a larger multicentre prospective cohort study. A preliminary study demonstrated that >10% of whole brain volume with an apparent diffusion coefficient (ADC) <650x10 -6 mm 2 /sec identified poor outcome (death/vegetative state) with 100% specificity and 81% sensitivity. We aimed to validate this threshold in an external dataset.

A multicenter retrospective observational study of DWI MRIs of comatose post-cardiac arrest patients obtained between 25 and 120 hours post-arrest was performed. Poor outcome was defined as death or persistent coma at day 14. Imaging was processed in a blinded fashion using Medical Image Processing, Analysis and Visualization program (MIPAV). The brain was semi-automatically outlined on the b0 images using a levelset algorithm. The ADC values of each voxel within the brain were determined. Outcomes were assessed blinded to quantitative DWI information. Treating physicians were not blinded to the MRI scans, but they were unaware of the quantitative DWI analysis.

Data from 111 patients from five US centers were included: mean age was 59±16 years, 36% female, arrest time 24±19 minutes, 59% of patients received hypothermia, and MRIs were obtained at 68±24 hours post-arrest. Thirty-two percent had a good outcome. The median (IQR) percentage of brain tissue with ADC<650x10 -6 mm 2 /sec was 3.7% (2.1-5.8) in good and 29.1% (9.5-47.0) in poor outcome patients (P<0.001). An ADC<650x10 -6 mm 2 /sec >10% was 93% (95% CI 76-99) specific and 73% (95% CI 61-81) sensitive for poor outcome with a positive predictive value of 97% (87-99) and a negative predictive value of 56% (41-70). The odds ratio of having a poor outcome if >10% of brain had an ADC<650x10 -6 mm 2 /sec was 38 (95%CI 8-171).

Quantitative DWI MRI in comatose post-cardiac arrest patients holds great promise as a prognostic adjunct between 2 and 5 days after arrest. 

Predicting outcome for comatose post-cardiac arrest patients is challenging and compounded by the use of therapeutic hypothermia and sedative agents. Brain MRI is a potential attractive prognostic adjunct not affected by drugs or metabolic derangements; however, most proposed methods require image post-processing. We assessed the prognostic value of color apparent diffusion coefficient (cADC) maps.

Consecutive post-cardiac arrest patients remaining comatose after resuscitation were prospectively enrolled. cADC maps were created by assigning ADC values to 8 colors ranging from red to blue. The treating teams did not see these maps. Two raters independently and blinded reviewed the cADC maps and predicted 3 month outcome as poor (Glasgow Outcome Scale (GOS) 1-2), or good (GOS of 3-5). Both raters were "trained" by viewing examples. The agreement between raters and the predictive performance of the cADC maps were assessed.

112 cADC maps of 94 patients (56% with poor, 44% with good outcome) were reviewed: age 59±15 years, 36% females, 69% underwent therapeutic hypothermia, median (IQR) arrest duration 20 min (14-30), and time between the arrest and MRI 82 hours (60-141). Kappa for agreement on predicting favorable vs. unfavorable outcome was 0.76. For the two reviewers, the sensitivity for predicting poor outcome was 0.85 (95% CI 0.73-0.92) and 0.78 (0.66-0.87), the specificity 0.81 (0.66-0.90) and 0.74 (0.59-0.86), and the true positive predictive rate 86% (74-93%) and 81% (69-89%), respectively. For MRI scans acquired between 25-120 hours after the arrest (i.e. the time-interval when ADC changes are most apparent), the specificity improved to 0.87 (0.68-0.96) and 0.77 (0.57-0.89), respectively.

MRI color ADC maps are easy to interpret and may be useful for predicting outcome of comatose post-cardiac arrest patients in the first 5 days after the arrest. Color ADC maps do not require post-processing and can be created in realtime.

There are few reports of outcome in patients with fat embolism syndrome with diffuse MRI abnormities.

We report the outcome of 2 patients with fat embolism syndrome.

Case 1 A 21-year-old previously healthy gentleman had a right femur fracture from a motor vehicle accident. He had acute respiratory failure 24 hours later requiring intubation. Chest x ray showed bilateral lung infiltrates. Neurological examination showed patient comatose with intact brainstem reflexes and extensor posturing. On day 3, he had fever, tachycardia, profuse sweating, and diffuse petechial rash. MRI brain showed diffuse restricted diffusion lesions. He started to open his eyes in 2 weeks and underwent tracheostomy and feeding tube placement. At 6 month follow up he only had mild memory problems. Case 2 A 42-year-old previously healthy gentleman had a gun shot in the left foot. Over the next 10 hours he became stuporous. X ray showed multiple fractures including calcaneus, soft tissue swelling and subcutaneous emphysema. Over the next 7 hours he worsened and displayed extensor posturing. MRI brain showed diffuse innumerable tiny infarcts. Patient was noted to have episodic fever, profuse sweating, and severe tachycardia. Patient had spontaneous eye opening next day and underwent tracheostomy and gastrostomy. He was transferred to a long term facility. Patient improved substantially and at 6 months follow up he was independently living at home with minor neurologic deficits.

Substantial improvement may occur in comatose patients with fat embolism syndrome despite paroxysmal sympathetic hyperactivity syndrome and significant MRI abnormities. 

Malignant pertussis is a rare life-threatening illness characterized by severe respiratory failure, extreme leukocytosis, and pulmonary hypertension. During 2011, an outburst of Whooping Cough was experienced at Montevideo, Uruguay. We present the cases of two infants, 5 and 7 months old, suffering malignant pertussis, admitted to a university pediatric intensive care unit (PICU) for severe acute respiratory failure associated with severe leukocytosis. Both children showed signs of profound coma and bilateral arreactive dilated pupils while being aggressively treated. Both of them showed a transcranial pattern of cerebral circulatory arrest (CCA) on transcranial Doppler (TCD).To our knowledge, a pattern of CCA has not been previously reported like mode of death secondary to neurologic injury in this disease.

Both cases were very similar: A 5-month-old boy, incomplete vaccinatinon, malnourished. A 7 month-old girl, vaccinated. Both had suffered at one and 4 months-old, severe bronchiolitis caused by respiratory syncytial virus, both needed 7 days of mechanical ventilation. Both were admitted to ICU with cough, fever, increased work of breathing, hypoxemia and were mechanically ventilated. They presented respiratory acidosis, hipoxemia, extreme leukocytosis greater than 100000, bilateral hyperinsuflation in chest x-ray. Echocardiography: pulmonary hypertension, 50-60 mmHg SPAP, circulatory failure, anuric renal failure. Bordetella Pertussis was diagnosed with PCR of airway secretions Treatment: blood exchange transfusions, milrinone, maximum dose inotropic drugs, peritoneal dialisis. After one week arreactive dilated pupils and profound coma were evident. Brain death was suspected, sedation and muscle blockers were interrupted.

Neurologic exam confirmed brain death. TCD showed sysytolic spykes in bilateral middle cerebral arteries, basilar artery, confirming CCA. Necropsy performed in case 2 showed bilateral pneumonia, small pulmonary artery branches thrombosis, neuronal necrosis, with brain edema, and renal tubular necrosis.

The mode of death in these two cases was brain death, with CCA. The probable pathophysiologic mechanisms were related to hyperviscosity and cardiac failure. 

DAVF's can be associated with benign or aggressive symptoms based on location and venous drainage. Cerebral venous ischemia is a reversible process emphasizing the importance of early recognition and treatment of dAVF's. 

In a geographically isolated region with limited neuroscience intensive care unit (NSICU) capacity, neurointensivists are often challenged to allocate resources and triage intracerebral hemorrhage (ICH) patients. We sought to assess the factors impacting the neurointensivists' triage decision for NSICU admission after ICH.

Consecutive patients hospitalized for ICH between 2006 and 2010 at a tertiary center that has the only 8-bed NSICU for the state, geographically isolated from the nearest NSICU (>2,000 miles away), were studied. Multivariable logistic regression models were used to test for predictors of NSICU admission, adjusted for each component of the ICH Score, transfer from another hospital, initial systolic blood pressure (SBP) >180 mmHg, and early do-not-resuscitate (DNR) order.

Among a total of 397 consecutive patients hospitalized for ICH, 204 patients (51%) were admitted to the NSICU while 193 patients (49%) were admitted to a non-NSICU unit. The ICH patients were more likely to be admitted to the NSICU if they had hematoma volume >30 cm 3 (OR 3.37, 95% CI 1.85-5.81), intraventricular hemorrhage (OR 2.35, 95% CI 1.39-3.96), Glasgow Coma Scale (GCS) score of 5-12 (OR 4.45, 95% CI 2.47-8.02), GCS score of 3-4 (OR 8.75, 95% CI 3.25-23.54), infratentorial hemorrhage (OR 3.25, 95% CI 1.56-6.77), transfer from another hospital (OR 2.82, 95% CI 1.51-5.28), and SBP >180 mmHg (OR 2.22, 95% CI 1.35-0.51, 95% CI 0.27-0.96) and early DNR order (OR 0.34, 95% CI 0.16-0.72).

The triage decisions for NSICU admission after ICH were based on clinical severity, age and early DNR status. A prospective study is needed to help establish a safe triage algorithm for ICH patients in a region with limited neurocritical care capacity. using a semi-automatic threshold based volumetry algorithm. Neurological status (NIHSS) was recorded daily and outcome was assessed at discharge using the modified Rankin scale (mRS).

The difference of PHE volumes between day 1 and day 5-8, representing the edema growth (PHE delta ), correlated significantly with the mRS at discharge (p=0.034; F=14). This correlation was still significant, when ICH volume on admission was controlled. Other factors that showed a significant association with outcome at discharge were NIHSS (ANOVA: p>0.0001, F=15.36) and ICH volume (ANOVA: p>0.0001, F=37.99) on admission. In a multivariate regression model only the initial NIHSS remained a significant predictor of functional outcome. PHE growth showed a weak trend towards significance (p=0.1).

PHE growth at the first days after symptom onset may influence early functional outcome after spontaneous ICH. Treatment strategies aimed at reduction of PHE burden after ICH may take advantage of this finding.

Assess the use of a 3-factor prothrombin complex concentrate (PCC, Profilnine®), compared to fresh frozen plasma (FFP) in establishing hemostasis in warfarin associated intracranial hemorrhage (ICH).

dmitted to UNC Health-Systems between 4/1/10 and 8/31/11 that received PCC, FFP, or both in conjunction with phytonadione for the treatment of warfarin associated ICH. Patients who received a factor product other than Profilnine®were excluded. Data collection included hematoma expansion, achievement of INR reversal (INR <1.5), 30-day mortality and endpoints related to safety (thromboembolic events, infection, and transfusion related acute lung injury).

Of the 29 patients included, 5 patients received PCC alone, 9 patients received PCC plus FFP and 15 patients received FFP alone. Hemorrhage expansion occurred in 4 of 5 patients (80%) in the PCC group, 4 of 9 patients (44%) in the PCC plus FFP group and 4 of 15 patients (27%) in the FFP group (PCC versus FFP, p=0.018; PCC plus FFP versus FFP, p=0.262). INR reversal occurred in 100% of patients in the PCC alone group, 100% of patients in the FFP alone group and 7 of 9 patients (78%) in the combination group.

This study assessed the impact of Profilnine®, FFP, or the combination, on achieving hemostasis based on hematoma expansion. Profilnine® achieved INR reversal but appeared to be less effective than FFP in preventing hemorrhage expansion. 

Fever after ICH is common and associated with poor outcome. However, the impact of therapeutic temperature modulation (TTM) to treat fever after ICH is unclear.

We performed a case-control study of TTM in ICH patients with fever. Patients undergoing TTM with advanced temperature modulating devices were prospectively enrolled in our TTM Database from 2006-2010 (TTM Group). Target temperature was 37C in all cases. Controls were matched in severity by ICH score and retrospectively obtained from a period (2001) (2002) (2003) (2004) before our routine use of TTM for ICH. Primary outcome was discharge modified Rankin Score.

We enrolled 40 patients in each group. Median ICH Score was 3 (range 1-4 

Therapeutic normothermia is associated with increased length of mechanical ventilation and NICU stays, but is not associated with improved discharge outcome.

Spontaneous intracerebral hemorrhage (sICH) is a dynamic process with significant growth in over one-third during the first 24 hours. Catheter-based evacuation of sICH plus recombinant tissue plasminogen activator (rtPA) is a novel surgical approach for which optimal timing of stereotactic catheter placement and clot aspiration are not known.

We investigated factors associated with significant ICH expansion (>33% or 12.5cc volume increase) on prerandomization CT scans of 117 patients meeting criteria for the MISTIE trial, a multi-center phase II clinical trial, evaluating safety and efficacy of minimally invasive surgery plus thrombolytic to treat ICH. Subjects randomized to surgery underwent stereotactic clot aspiration followed by injections of rtPA through the hematoma catheter every 8 hours, up to 9 doses, or until a clot reduction endpoint.

Median diagnostic ICH volume was 35.6cc (iqr 24.7). Overall, 18.8% of patients exhibited significant hematoma expansion at a median of 20.6 (iqr 13.3) hours from symptom onset. Predictors of hematoma growth were smaller diagnostic ICH volume (OR 0.88; P=0.001), longer interval from symptom onset to pre-randomization CT (OR 1.07; P=0.008), non-lobar location (OR 0.22, P=0.05), lower initial platelet count (OR 0.98; P=0.03), and lower initial hematocrit (OR 0.84; P=0.002). Age, gender, admission blood pressure, initial coagulation parameters, hematoma shape and density scores did not predict hematoma expansion. End of treatment hematoma expansion occurred in 3/117 (2.6%) patients of whom 2 had early ICH expansion and 2 underwent surgical intervention.

Stabilization of hematoma growth can be anticipated within 24 hours of symptom onset in patients considered for minimally invasive surgery using the MISTIE protocol. Smaller initial ICH size, deep location and lower hematocrit and platelet counts were independent determinants of significant ICH expansion before surgery. Patients with early expansion may represent a group at higher risk for re-expansion with clot aspiration and thrombolytic therapy.

Financial Support: Daniel F. Hanley received funding from NIH grant R01 NS046309. 

Thin-section noncontrast CT (NCCT) provides a measure of thrombus composition based on Hounsfield Units (HU) and may predict resistance to thombolytics in acute ischemic stroke. Hematoma composition may affect thrombolytic efficacy of Tissue Plasminogen Activator (TPA) in acute intraventricular hemorrhage (IVH). We assessed the value of hematoma HU quantification as a predictive marker of IVH clearance after intraventricular TPA administration.

Serial NCCT was performed on 52 patients who received intraventricular TPA as part of the CLEAR IVH trial (Clot Lysis: Evaluating Accelerated Resolution of IVH) and 12 controls with IVH treated with external ventricular drainage (EVD) alone. A blinded investigator calculated HU values for IVH volumes on admission, day 3-4 and day 6-9 NCCT for each patient.

Median IVH volume on admission for TPA-treated patients was 38.3(iqr 34.1)cc, and decreased to 4.9(14.5)cc at day6-9. Mean(SD) HU for IVH was 52.1(4.3) on presentation and decreased significantly to 50.1(4.5) on day3-4, and to 45.1(4.95) on day6-9. IVH HU count was significantly correlated with IVH volume at all CT timepoints: admission:p=0.002; day3-4:p<0.001; day6-9:p<0.001. There was no correlation between admission serum platelet count, fibrinogen level or hemoglobin and clot HUs. Only CSF protein was positively correlated with IVH HU (p=0.03). Total IVH HUs were significantly lower in TPA-treated (vs. control) patients at day6-9 (p=0.001), but not at day3-4. Change in IVH volume from admission to day3-4 was positively correlated with higher initial HU in TPA-treated patients (p=0.04), but HU was not significant after adjustment for IVH volume and TPA treatment.

Hounsfield Unit counts of IVH decrease significantly over the first week on NCCT and the decrease is greater in TPAtreated patients. Unlike thrombus HUs in large intracranial vessels, IVH HUs are not associated with erythrocyte or platelet concentrations. Higher HU is not an independent predictor of success of intraventricular thrombolysis. 

Although neurocardiogenic myocardial injury is well described among patients with spontaneous intracerebral hemorrhage (sICH), it has not been investigated systematically in patients with acute subdural hemorrhage (aSDH). We sought to investigate the prevalence and characteristics of myocardial injury in patients with aSDH.

Consecutive adult patients with a diagnosis of aSDH admitted to the Rush University Neurosciences Intensive Care Unit were analyzed. Myocardial injury, defined as troponin I elevation (> 0.09ng/ml) on admission or during hospital course, was identified. Electrocardiograms (ECG) and echocardiograms, obtained within the first 24 hours and read by a cardiologist blinded to the patient's history, were retrospectively reviewed.

A total of 107 patients were admitted with aSDH between 12/2009 and 11/2011. The mean age was 66 years (SD 17 years), and 63% were male. Comorbidities included hypertension (78%), diabetes mellitus (32%), coronary artery disease or prior myocardial infarction (37%), congestive heart failure (10%), coronary artery stent or bypass procedure (21%). ECGs were normal in 24%. Non-specific ST-T morphologic changes, QRS changes or sinus arrhythmias were seen in 71%. ST-elevations suggestive of myocardial infarction were not seen. Of 17 patients with elevated troponin, 13 had known severe cardiac disease, and 4 had severe medical complications (sepsis, renal failure, hepatic failure and acute lymphoma crisis). Diffuse ECG changes ("cerebral T waves") and echocardiographic findings suggestive of neurogenic stunned myocardium (reversible wall motion abnormalities, apical ballooning) were not seen.

Although we found ECG changes to be common after aSDH, myocardial injury was only observed in the context of concomitant cardiac or medical disease. Classic neurogenic cardiac findings (cerebral T waves, neurogenic cardiomyopathy) were not observed. While myocardial injury in sICH often is attributed to neurocardiogenic causes, these are unlikely prominent mechanisms in patients with aSDH. Other cardiac or medical causes are common and should be excluded.

Prognostication in intracerebral hemorrhage is complex and mortality remains high. While tools such as the ICH score have been developed to assist with prognostication, physicians clearly use additional parameters in clinical practice. Though do-not-resuscitate orders do not indicate the withholding of any treatment other than CPR, they are associated with increase risk of death in patients after ICH. We sought to understand early DNR (<72hours) designation in patients with hope of moving toward more precise tools for prognostication.

Patients admitted to the Neurological Intensive Care Unit from July 2007 to December 2010 with a diagnosis of supartentorial ICH were identified. Data for all patients were collected retrospectively. Patients without a DNR order throughout their admission were compared to patients who received a new DNR order in the first 72 hours of admission. Patients with pre-existing DNR orders were not included.

157 subjects were included in the study with 18.4% made DNR within 72 hours of admission to our NICU. Factors showing a significant correlation with a new DNR order included advanced age, Caucasian race, or residence in a skilled nursing facility. History of malignancy, atrial fibrillation, current use of antihypertensives or warfarin, or alcohol abuse predicted a DNR order. ICH resulting in a low admission Glasgow Coma Score, high ICH score, intraventricular extension and blood volumes greater than 30ml correlated with a new DNR order.

While individual elements of the ICH score correlate with a new DNR order in this population, other characteristics were also associated with an early DNR order. Early DNR orders may create a self-fulfilling prophecy if limitations of support are instituted without a clear understanding of who may benefit from aggressive care. Thus, identification of factors that providers believe to be life-limiting may serve as a starting point to avoid early limitations in aggressiveness of care. 

Intracranial hypotension is caused by low cerebrospinal fluid (CSF) pressure, clinically distinguished by orthostatic headaches and neurologic signs. Subdural effusions and even hemorrhage can be a secondary effect. Known causes include dural punctures as well as spontaneous CSF leaks. Treatments are guided towards repairing the cause of the hypotension.

A 74 year-old man on Coumadin for a mechanical aortic valve was transferred to our institution for evaluation of bilateral SDH. He presented twenty-four hours earlier with sudden-onset severe headache and normal neurological exam. Upon transfer, he was noted to be drowsy, with a left third cranial nerve palsy. He endorsed a postural headache that worsened upon standing. MRI of the brain showed small subdural effusions with subacute hemorrhage and minimal mass effect, as well as severe distortion of the midbrain with narrowing of the ventricular system, crowding of the basal cisterns and extensive pachymeningeal enhancement. MRI of the spine did not demonstrate a CSF leak but showed a small perineural cyst at T11.

The patient's headache and exam findings initially improved with lying flat. A trial of elevating his head of bed failed, with the patient further developing decreased level of arousal, frontal release signs, and recurrent left third nerve palsy. Two attempts at lumbar epidural blood patch (EBP) did not lead to sustained improvement, but a thoracic-directed EBP led to durable and complete resolution of the headache and neurologic deficits.

Intracranial hypotension should be considered as a cause for subdural hemorrhages in the absence of trauma. Clues include postural headaches and clinical evidence of brainstem dysfunction and radiographic evidence of brainstem distortion out of proportion to the size of the SDHs. EBPs directed at suspected CSF leaks can be effective when nondirected lumbar EBPs fail.

Multiple guidelines recommend the reversal of oral anticoagulation when a patient experiences an intracranial hemorrhage (ICH). Both activated prothrombin complex (aPCC) and recombinant factor VII activated (rFVIIa) have been utilized to reverse warfarin-associated coagulopathy. However, there have been no direct comparisons of these agents.

This was an IRB approved, retrospective cohort analysis of patients with ICH. Patients were included if they received either aPCC (at least 500 units) or rFVIIa (at least 1 mg), if they had a discharge diagnosis of intracranial hemorrhage, and if they received warfarin prior to admission. Patients were excluded if they were less than 18, or did not have documented pre-treatment and post-treatment INRs. The primary endpoint for this study was the change from pretreatment INR and post-treatment INR in the aPCC group and rFVIIa group. Secondary endpoints included change in CT measured hemorrhage volumes, ICU length of stay (LOS), hospital LOS, mortality, ICU discharge GCS, and thromboembolic adverse events.

A total of 60 patients were included in the analysis. Of those, 35 received aPCC and 25 received rFVIIa. Baseline demographics were comparable; however, patients in the aPCC group had a higher rate of atrial fibrillation (68% vs. 40%; p=0.028). When compared over time, both aPCC and rFVIIa significantly reduced the INR (p<0.005); however, there was no difference in the amount of change observed between the two groups (p=0.72). In addition, we saw no significant differences with regard to ICU LOS, hospital LOS, mortality, ICU discharge GCS, or thromboembolic adverse events. Over time, an increase in ICH volume was observed in both groups.

In patients with intracranial hemorrhage, aPCC and rFVIIa are associated with rapid reversal of warfarin-associated coagulopathy; however, these agents may not slow ICH growth. 

Accurate prognostication of patients with intracerebral hemorrhage (ICH) is critical because it may affect aggressiveness of care and patient outcome. ICH prediction models help stratify patients according to their chance of a good or poor outcome. We compared the accuracy of neurointensivists' prediction of functional outcome to outcome prediction by the ICH score.

Adult spontaneous ICH patients with an admission GCS >5 were prospectively enrolled. The treating neurointensivist predicted the 3-month modified Rankin scale score (mRS) within 2 days of hospital admission. None of the neurointensivists used the ICH score routinely to help predict outcome. Patient outcomes were dichotomized to good (mRS 0-3) and poor (mRS 4-6). Neurointensivists' predictions were compared to the ICH score using the actual 3-month mRS as the reference.

Of 116 prospectively enrolled patients, 101 were included: 2 withdrew consent and 13 were lost to follow-up. Neurointensivists' overall accuracy was 79%, which was higher than the accuracy of the ICH score at a cut-off of >1 (67%; p=0.06) or >2 (58%; p=0.001). At a cut-off >1, the sensitivity for poor outcome prediction did not differ, but the neurointensivists' specificity for poor outcome was greater (92% vs. 66%; p<0.001). Conversely, at an ICH score cut-off >2 the specificity for poor outcome prediction did not differ, but the neurointensivists' sensitivity for poor outcome was greater (67% vs. 25%; p<0.001). The results were similar if, instead of the original ICH score, a modified ICH score was used as the comparison that had been developed on the same patient cohort.

Neurointensivists at our institution predict ICH outcome overall with 79% accuracy. Generally, predictions for poor outcome are more accurate than those for good outcome. Outcome predictions for the individual patient by the treating neurointensivist are more accurate than those based on ICH prediction models. 

There is continued controversy regarding glycemic control and its effect on outcomes for patients with ICH as well as other ICU patient populations. The relationship between ICHsize and glycemic control has not been clearly defined.

A retrospective review of 87 patients with supratentorial ICH and no history of diabetes between 2003 and 2007 was performed. Admission blood glucose (BG) as well as BG at 24, 48 and 72 hours was measured while all patients were maintained on the same sliding scale insulin regimen. Statistical analysis was performed to compare admission ICH size to admission BG and subsequent BG control.

BG>120mg/dL (mean 172 ±60) and mean ICH size 87.4 ± 78.1. Average BG levels over 72

Average BG levels over 72 hours were 106 ±35mg/dL. Elevated admission BG was significantly correlated with admission ICH size (p=0.001). Average BG levels over 72 hours trended towards, but were not significantly correlated with admission ICH size (r =0.19, p=0.078).

In nondiabetic patients, elevated admission glucose is significantly associated with ICH size. Poor outcomes associated with elevated glucose may be associated more with extent of cerebral insult than with glycemic control.

The benefits of marriage on health have been known for over 150 years. More recently, married couples have been found to have a lower risk of cancer, dementia, and heart disease. We aimed to explore the effects of marital status on outcome after intracranial hemorrhage.

A prospective study was conducted between 2008-2011 of patients with subarachnoid hemorrhage (SAH, N=118), intracerebral hemorrhage (ICH, N=130) and subdural hemorrhage (SDH, N=133), admitted to the neuro-ICU at a tertiarycare academic hospital. Marital status was coded as married versus single, widowed, divorced or separated. Modified Rankin Score and Barthel Index were compared between the two groups at 3-months using multiple logistic regression analysis.

Of 371 patients, 198 (53%) were married, 84 (23%) were single and 89 (24%) were widowed, divorced or separated. Women were less likely to be married, and smoking was less common among the married (both P<0.01). There was no difference in age, insurance or employment status, race, education, days to diagnosis, or history of heart disease, diabetes, hypertension, trauma or coagulopathy. After adjusting for age, admission GCS, APACHE physiologicalsubscore, gender, tobacco and bleed type, marriage was significantly protective against death or severe disability (mRS 4-6; adjusted OR 0.5, 95%CI 0.3-0.9, P=0.028) and predicted better activities of daily living (Barthel Index), (aOR 2.0, 95%CI 1.1-3.9, P=0.032). There was no difference in discharge disposition, length of stay or hospitalization costs.

Marriage is protective against death or severe disability and predicts better activities of daily living among patients with intracranial hemorrhage.

Warfarin associated intracranial hemorrhage leads to poor outcomes. We studied the influence of a standardized emergent warfarin reversal protocol incorporating prothrombin complex concentrates (Profilnine SD®) on time to achieve

The protocol was implemented in 2010. Sixty three patients (25 pre and 38 post protocol) from 2009-2011 with intracranial -protocol patients received recombinant Factor 7 a (rFVIIa); post protocol patients with INR 1.6 -4 and >4 received 25 and 35-50 units/kg of Profilnine SD® respectively. Hemorrhage volumes were measured on consecutive CT scans using MIPAV semi-automated software.

Groups were similar for baseline median INR (2.5 Vs 2.7), NIHSS (1 Vs 2), follow-up CT time (7.6 Vs 7.4 hours) and hemorrhage volumes (7.5 Vs 5.2 cc) but differed in hemorrhage type: ICH (36% Vs 53%) and SDH (48% Vs 18%), p=0.03. Treatments also differed: Vitamin K (80% Vs 97%, p=0.03), Profilnine SD® (0% Vs 61% p=<0.001), rFVIIa (24% Vs 3% p=0.01) and number of plasma units (5 Vs 3, p=0.001).Time to target INR was similar (13.3 Vs 10.1 hours) driven by pre-protocol rFVIIa use (rFVIIa used Vs not, 13.3 Vs 22.5, p=0.03) and this led to INR rebound in <24 hours. Excluding rFVIIa, the post protocol group normalized INR faster (22.5 Vs 10.4 hours, p=0.036). The post protocol group had less absolute (6.3 Vs 1 cc p=0.01) and relative hemorrhage growth (95% Vs 29% p=0.03) without any thrombotic events. Despite comparable mortality (24 Vs 28%), post-protocol survivors more frequently achieved mRS 0-3(66 Vs 52% p=NS).

A standardized emergent warfarin reversal protocol is not only safe but leads to faster INR normalization, less hemorrhage growth, plasma conservation and possibly better neurological outcomes. 

Perihemorrhagic edema (PHE) after intracerebral hemorrhage(ICH) may exceed the initial hematoma volume by 100 to 600% respectively and thereby lead to increased intracranial pressure (ICP), clinical detoriation or even herniation. Intravenous hypertonic saline (HTS) has been shown to reduce PHE formation after ICH. Clinical data suggest that HTS may be superior to mannitol in lowering ICP. EUSI and ASA guidelines recommend the use of intravenous mannitol up to a serum osmolality (SO) of 320 mosmol/kg or HST in order to reduce elevated ICP. We aimed to investigate the effect of mannitol and SO on the evolution of PHE after ICH.

Nineteen patients with supratentorial spontaneous ICH treated with 20% intravenous mannitol solution (125-250ml every 6h) for 5-10 days and 19 controls who did not receive mannitol or any other osmotic agents during the course of treatment were identified retrospectively from our institutional ICH database. Patients treated with mannitol and controls were matched for ICH-volume (±5ml). PHE volume was calculated on CT scans using a semiautomatic threshold based volumetric algorithm. Diagnostic CT scans and follow-up scans performed on days 1, 2-3,4-6,7-9 and 10-12 were analyzed. SO, concentration of sodium and glucose were obtained from patient records.

The matching resulted in similar ICH-volumes in both groups (mannitol: 24.9±14.4ml, controls: 22.6±22.8ml). Mean age was 71(53-83) years in the mannitol group and 64(31-85)years in controls (p=0.1). Initial relative PHE did not differ significantly in both groups (mannitol: 1.25±0.6, controls: 1.01±0.7, p=0.24). There was no effect of mannitol treatment on the course of PHE (F=0.83,p=0.37). There was no significant correlation between SO and relative PHE at any timepoint of follow-up.

We found no effect of mannitol use and SO on the evolution of PHE. Other underlying mechanisms may explain the shortterm effect of mannitol boluses on ICP in patients with spontaneous supratentorial ICH. 

When operating at maximum intensive care unit (ICU) bed capacity where allocation of critical care resources is required, physicians may be pressured to initiate do-not-resuscitate (DNR) orders in patients with intracerebral hemorrhage (ICH). We sought to assess the relationship between early (<24 hours from admission) DNR orders and neuroscience intensive care unit (NSICU) bed capacity in patients admitted with acute ICH.

We retrospectively studied consecutive patients hospitalized for ICH between 2008 and 2010 at a tertiary center that has the only 8-bed NSICU for the state, geographically isolated from the nearest NSICU (>2,000 miles away). Multivariable logistic regression models were used to test for predictors of early DNR orders, adjusted for each component of the ICH Score and NSICU bed census on admission. NSICU bed census was dichotomized to full (all beds occupied) vs. not full (at least 1 available bed).

Among 206 

Maximum ICU bed capacity on admission is not associated with the decisions to initiate early DNR orders in ICH patients. This suggests that physicians were not preferentially initiating early care limitation when critical care resources were becoming scarce. 

Dural arteriovenous fistulas (DAVF) are rare, acquired cerebrovascular lesions consisting of abnormal vascular connections between arteries that normally supply the dura and veins that drain the brain parenchyma -that is to say, arteries not associated with the brain parenchyma manage to drain via the dural venous sinus system. The clinical consequences of these lesions are typically hemorrhage, seizure, or venous congestion. Venous congestion may present acutely with hemorrhage or subacutely with signs and symptoms such as progressive cognitive decline, seizures, or encephalopathy. Parkinsonism, tinnitus, and intracranial hypertension have also been described.

Case report with review of literature.

We describe a 61-year-old man with no past medical history who developed subacute onset dementia with bithalamic T2 hyperintensity on MRI without restricted diffusion. Subsequent intraventricular hemorrhage resulted in emergent transfer to our institution's neurocritical care unit for an emergent diagnostic cerebral angiogram of a Borden II/Cognard IIb DAVF with immediate angiographic embolization and obliteration.

DAVFs are lesions with significant risk of aggressive neurologic devastation related to venous congestion and subsequent hemorrhage. The severity of DAVFs requires clinicians to be aware of these lesions and of their common and uncommon presentations. 

Little is known about the ability of prognostic scores to predict outcome in patients with secondary intraparenchymatous intracranial hemorrhage (IPH). Our objective was to describe the clinical characteristics, ICH scores at presentation and prognosis in patients with secondary IPH.

We performed a post-hoc analysis of prospectively collected data of consecutive patients admitted to a tertiary hospital with IPH. The characteristics of patients with secondary IPH were compared to those of patients with spontaneous IPH. Patients with secondary IPH had either a positive underlying vascular lesion identified as the IPH etiology or impaired coagulation at presentation (a platelet count <50,000 per cubic mm, an INR > 3.0, or an aPTT > 80 than seconds).

A total of 71 patients with IPH were admitted to our hospital from January-2009 to January-2012. Of those, 10 patients (14%) had a secondary IPH (2 cavernomas, 5 arteriovenous malformations, 1 dural fistula, 1 reversible vasoconstriction syndrome and 1 sacular aneurysm 

Patients with secondary IPH had lower ICH scores at presentation and lower in-hospital mortality than patients with spontaneous IPH. Despite lower ICH scores at admission, patients with secondary IPH had similar functional outcomes when compared to patients with spontaneous IPH. Larger studies should focus on specifically developing better prognostic tools in such patients. 

A large number of studies in traumatic brain injury patients have shown efficacy of hypothermia for control of ICP and if used for prolonged duration, has shown to improve mortality and functional outcome. For other Neurologic catastrophes, due to a risk of rebound edema during re-warming, medical complications and other factors, it has either not been commonly used or been used when most of other options are exhausted.

This is a retrospective analysis of 8 patients with massive ICH (blood volume of > 60 ml), of non traumatic etiology, dominantly in brain parenchyma. All patients had intracranial pressure monitoring via external ventriculostomy catheter. Hypothermia was induced and maintained at target temperature via non-invasive, surface cooling pads. Modified rankling score (mRS) was recorded at 4 months after the ictus in all survivors.

Patient ages ranged from 37 to 53 years. Cause of ICH was hypertension in 7 patients and ruptured aneurysm in one patient. Duration of treatment ranges from 5-14 days. Target temperature required to adequately control ICP ranged from 32-34 C. Two patients (25%) survived with good recovery (mRS of 2), one (12.5%) with moderate disability (mRS of 3), two (25%) with moderately severe disability and three (37.5%) died. Most common side effect of hypothermia was hypotension requiring pressors in five (62%), electrolyte imbalance in 4 (50%), pneumonia in 3 (37%), thrombocytopenia in 3 (37%) patients. All complications were successfully treated and major complications of treatment (bleeding diathesis, septic shock syndrome, death) were not observed.

Controlled hypothermia for up to 14 days is safe and feasible for the treatment of cerebral edema and intracranial hypertension in young patients with massive (> 60 ml of blood volume) non traumatic ICH. However, prolonged duration of treatment may be required for definitive control of ICP. This study serves as a template for future efficacy trials. 

Intracerebral hemorrhage (ICH) accounts for 10% to 15% of strokes and is associated with substantial morbidity and mortality. It remains controversial whether surgical intervention or a conservative approach is the best option for treating ICH. We assessed the hypothesis that early surgical intervention in patients with primary supratentorial ICH may serve to improve 30-day outcome.

A total of 97 patients with primary supratentorial ICH, in whom surgical intervention was indicated for hematoma removal according to the guidelines, were admitted to our hospital during a continuous 12-month observation period. Patients with the consent to the surgical intervention (n=35) underwent surgery within 24 hours of symptom onset and the others (n=62) were given the conservative treatment. The outcome was the proportion of patients who had an unfavorable outcome (persistent vegetative state or death), as assessed on the basis of the Glasgow outcome scale (GOS) at 30 days.

The 30-day mortality rate was 51.5% (standard error, 5.1%). There was no significant difference in outcome between the two treatment groups. After adjustment for other significant covariates, although a lower unfavorable outcome was found in surgical group but the difference was not significant (odds ratio = 0.725). Among the confounding factors, presence of intraventricular hemorrhage (IVH) and low Glasgow Coma Scale (GCS) score on admission were independently associated with poor outcome 30 days after ICH (both P <0.001).

We found no benefit for early surgical intervention over conservative treatment in patients with primary supratentorial ICH. Presence of IVH and low GCS score were strong predictors of poor outcome in these patients.

Given the high morbidity and mortality associated with intracerebral hemorrhage (ICH), family members and healthcare providers base early supportive management decisions, at least partly, on expected prognosis. In the comatose patient with ICH, this short term prognosis is most overtly characterized by regaining of consciousness.

A retrospective consecutive cohort of 51 patients, between 2006 and 2011, with ICH and admission Glasgow Coma Scale that were associated with regaining of consciousness after coma in ICH. Variables associated with awakening in univariate analysis were tested in multivariable logistic regression.

The group that awakened had higher initial GCS scores, smaller ICH volumes, and less IVH, but was similar in other baseline characteristics. Early DNR orders, in the first 24 hours, tended to be used more frequently in patients who ultimately remained comatose, but the difference was not statistically significant. Admission GCS, volume of ICH, and presence of IVH identified in univariate analysis were tested along with age and gender as potential confounders of outcome in multivariable analysis. Higher admission GCS score was associated with an increased likelihood of awakening from coma (OR 4.9 [95%CI 1.9-13] per category, p=0.001). 87% of patients with initial GCS of 7-8, 40% with initial GCS of 5-6, and 20% with initial GCS of 3-4 regained consciousness. Awakening from coma, in the cohort of 27 patients who regained consciousness, occurred in 59% of patients by day 2, 89% by day 7, and 96% by day 9.

GCS score is the single most important predictor of early awakening in patients who present in coma after ICH. Patients who regained consciousness typically did so within the first 9 days of hospital admission.

Intracerebral hemorrhage (ICH) is an infrequent but severe complication in pregnant women with hypertension, it accounts for 50% of all deaths related to cerebral complications in this group.

A-32-year-old female, G2P1 at 34 weeks of gestational age, with prenatal care, no relevant past medical history, presented for a follow-up visit. She was admitted with BP 150/90 mmHg, and treated with IV labetalol; the preeclampsia work-up was negative, BP range between 150/90 and 180/110 mmHg. Approximately 10 h after admission, she complained of diffuse headache, nausea, vomiting, and epigastric pain. Headache symptoms increased follow by focal seizure and progression to generalized tonic-clonic seizures. Magnesium sulphate and phenytoin were administered to control the seizure, immediate blood analysis revealed DIC. The diagnosis of eclampsia was made, and emergency Csection followed. The airway was secured with rapid sequence technique; a healthy infant was delivered under general anesthesia. The patient remained comatose 1 hour after surgery with GCS 4, 90 minutes later she demonstrated a decerebrate posture with non-reactive pupils. A non-contrast CT-scan revealed an intracerebral hematoma. DIC was treated, and neurosurgeon performed a right frontotemporal craniotomy. A postoperative CT scan confirmed the resolution of the ICH. The patient opened her eyes and started responding to commands by the third day, on day 7 she was extubated and the GCS was 9. By the 3rd week, the patient was transferred to rehabilitation, where she remained for 4 weeks. At 3 years, she regained a full cognitive recovery.

This case emphasizes that even short time hypertension should be treated aggressively to prevent ICH. The prompt intervention of a multidisciplinary team (obstetric, neurosurgery, and anesthesiology) is required to ameliorate the devastating effects of eclampsia and ICH.

Although hypertension is the commonest cause of non-traumatic intracerebral hemorrhage (ICH), it is important to rule out other causes. Most patients with ICH have an elevated BP on presentation but many are unaware if they have longstanding hypertension. Echocardiographic abnormalities may be revealing in such circumstances. We studied the incidence of echocardiographic abnormalities and their usefulness in determining the etiology of ICH in these patients.

We conducted a retrospective study of echocardiographic abnormalities in ICH patients admitted to a tertiary university hospital between Jan 2008 to Oct 2011 who also had a cerebral angiogram. Subjects with and without underlying vascular location (categorized as typical hypertensive location or not), history of hypertension and the presence of the following echocardiographic abnormalities: left ventricular hypertrophy (LVH), diastolic dysfunction (DD), systolic dysfunction (SD), hyperdynamic ventricular function (HVF), wall motion abnormalities, atrial enlargement (AE) and valvular abnormalities using Chi-square test and Fisher exact test. We then conducted a multivariate logistic regression analysis including variables with a p<0.25 in the univariate analysis.

A total of 303 subjects were admitted with an ICH. 159 subjects had an echocardiogram and of these, 130 also had an angiogram (conventional angiogram: 72, CT angiogram: 32, MR angiogram: 26). The echocardiogram was abnormal in 89.2% (93.2% with a history of hypertension p=0.006). Common abnormalities were: LVH (43.1%), DD (42.3%), HVF(18.5%), SD (13.1%) and AE (12.3%). Of these, only DD (p=0.049) was significantly associated with absence of underlying vascular abnormalities on a univariate analysis. On multivariate analysis, none of the echocardiographic abnormalities showed a significant association.

Echocardiographic abnormalities, mainly LVH and DD are commonly seen in ICH patients, however none of these abnormalities are independently associated with an absence of underlying vascular anomalies. 

Stroke in the HIV+ population is a growing problem, though it is unclear whether HIV is an independent risk factor. We describe a series of HIV+ patients with intracerebral hemorrhage (ICH).

We reviewed records of all patients with diagnoses of ICH and HIV/AIDS admitted to an academic, inner-city hospital between 2005 and 2011. Patients with traumatic hemorrhage, ischemic stroke with hemorrhagic conversion, hemorrhagic neoplasms, toxoplasmosis with hemorrhage, subarachnoid hemorrhage, and extra-axial hemorrhages were excluded. We reviewed demographics, risk factors, laboratory tests, and neuroimaging. Outcomes at 90 days were determined by modified Rankin scale (mRS).

Six HIV+ patients (83% male, mean age 51) met inclusion criteria, with 1 patient having recurrent hemorrhages; 50% were Black, 33% Hispanic, and 17% of other racial groups. All patients met criteria for AIDS. Risk factors included: prior stroke (17%), diabetes (33%), hypertension (33%), smoking (50%), and illicit drug use (33%). Only 1 patient was taking antithrombotic medication. The co-prevalence of HCV was 50%. Admission blood pressure was >140/90 in 3/6 patients. Laboratory evaluation demonstrated 1 patient with a prolonged INR (>1.3) and 3 patients with thrombocytopenia (<150). The hemorrhages were lobar in 4/6 and deep in 2/6. Only 3 patients had vessel imaging; one had an AVM and none demonstrated aneurysm or vasculitis. At 90 days, four patients were deceased and the two survivors had mRS of 4 and 5.

In this cohort, HIV-associated ICH occurred only in AIDS patients. Outcomes were uniformly poor, with 100% of patients having a HTN and unexpected predominance of lobar hemorrhages in younger patients, suggesting a distinct mechanism of ICH. 

In 1988 Gregory Call and Marie Fleming reported four patients with what appeared to be a reversible form of cerebral vasoconstriction. Since then a number of authors have reported reversible cerebral vasoconstriction syndromes (RCVS), often in association with potential etiological precipitants. The major complication of RCVS is ischemic stroke, but hemorrhagic strokes can also occur, eventually leading to permanent sequelae and even death. Recent reports and case series have suggested that intracranial hemorrhages may be frequent in RCVSand its presentations may range from cortical subarachnoid hemorrhages to intracerebral hemorrhages and subdural hemorrhages.

We report two cases of RCVS in middle age women, with hemorrhagic strokes caused after the prescription of dipirone, isometheptene and anhydrous caffeine, with putaminal hemorrhage, and lobar frontal hemorrhage.

Both cases showed complete reversion of arterial vasoconstriction weeks later by the transcranial Doppler.

Despite the reversibility of the vascular constrictions that characterize RCVS, brain lesions are observed in up to 81% of the patients.Most of these lesions are of ischemic nature; however hemorrhagic phenomena are not uncommon and have only been reported in 5%-12% of the cases. Isometheptene has been described as a trigger for RCVS in only a handful of patients, all of whom were women in the postpartum period. Even though RCVS diagnosis demands evidence of complete reversibility of the vasospasms, differential diagnosis with SAH can be made by the identification of classic RCVS triggers and assessment of the vascular patterns brain arteries.

Magnesium (Mg) has been hypothesized to have a neurprotective effect against cerebral ischemia. Several ongoing studies are examining the effect of exogenous magnesium in reducing disability and maintaining normal cerebral function. We examined initial endogenous Mg levels in patients with spontaneous intraparenchymal hemorrhage (IPH), in order to determine if higher Mg blood serum levels would confer neuroprotective benefit.

This is a retrospective study on patients admitted to a university affiliated community hospital. Demographic data were obtained from a prospectively collected registry database. Initial Magnesium levels were gathered retrospectively from the registry database. We included all patients with IPH in our analysis. For evaluating the severity and outcome of the patients with IPH we used the University of California San Francisco intracerebral hemorrhage (UCSF ICH) score on admission / 24 hours to quantify stroke severity and mRS on discharge to measure outcome. We employed correlation coefficients (Spearman's rho) and the Mann Whitney test for analysis of the data. SPSS version 11 was used for data processing.

Our review identified 74 patients with a diagnosis of IPH. The Serum Mg levels in patients with IPH negatively correlated to UCSF ICH score on admission (p=0.01, r= -0.37) and at 24 hours (p=0.007, r= -0.40). There was a trend towards better outcomes at discharge in patients with higher Mg levels (p=0.1, r= -0.25).

Higher levels of endogenous serum Mg were found to confer reduction in IPH severity and progression. Initial Serum Mg levels could serve as an early predictor of IPH severity. A larger prospective study is warranted to study the effect of endogenous Mg on outcomes in patients with IPH.

Spinal dural arteriovenous fistulas (DAVFs) account for 70% of all vascular spinal malformations. The incidence is 5-10/million/year in the general population although it is generally under-diagnosed. Men are affected five times more often than women and the mean age at the time of diagnosis is 55-60 years. Spinal DAVFs generally do not present acutely and are very rarely located in the cervical region. We present a case of atypical acute spinal cord infarct secondary to a cervical DAVF.

Case report and extensive literature search carried out to understand spinal DAVFs.

This 60 year old gentleman presented to our Neurocritical Care Unit with bilateral upper extremity weakness and right lower extremity weakness proceeded by upper back and neck pain. The patient rapidly deteriorated to near quadriplegia and respiratory failure requiring prolonged artificial ventilation. Initial studies included normal MRI of the brain and CT angiogram of the head and neck. MRI of the spine revealed abnormal signal intensity within the anterior cervical cord from C3-C7 levels in the distribution of the anterior spinal artery. There were no flow voids to suggest dilated perimedullary vessels. However, given the clinical picture, a spinal angiogram was obtained and demonstrated a cervical DAVF supplied by a dural branch vessel originating from the left vertebral artery.

Understanding spinal vascular anatomy is important for diagnosis of spinal DAVFs. Our case is unusual because 1) acute evolution of quadriplegia and respiratory failure, 2) lack of any abnormal vessels seen on MRI, and 3) ischemic changes restricted to the anterior spinal artery distribution. The case emphasizes the importance of proceeding with spinal angiography if the clinical suspicion of DAVF is high. Early detection and management can lead to improved functional outcome. 

Although coma is a syndrome commonly associated to catastrophic brain injury, this patient population remains poorly characterized. The chief goal of therapy is aimed at reversal of coma. Despite this urgency, there is paucity of data regarding the factors that predict emergence. We characterize a population of patients with new onset of coma in the Neuro-ICU and describe clinical and structural factors that predict emergence.

Prospective longitudinal consecutive cohort of patients, enrolled in an intensive care setting.

Three hundred patients met investigation enrollment criteria between May 2010 and July 2012. A brain lesion was identified at the onset of coma in most patients (92%). Frequent etiologic factors were cerebrovascular (65%), seizures (18%), trauma (15%), CNS infection (12%), or other (7%). The most frequent cerebrovascular factors were any ICH (40%), IVH (34%) and SAH (33%), either alone or in combination. Emergence from coma was predicted by a higher initial GCS (emergence=9[3-13] vs. no emergence=5.5[3-11] p<0.0001), seizures as presenting disorder (emergence=26% vs. no emergence=9% p=0.025), and a trend to lesser frequency of ICH component (emergence=31% vs. no emergence=50%, p=0.06). The importance of mass effect as measured by midline shift reversal and cisternal compression resolution is presented in a separate poster. Mortality in this cohort is 45%.

The population of patients with acute coma is highly heterogeneous. However clinical and structural factors predict emergence. A higher initial GCS predicted recovery of coma. Structural cerebrovascular lesions with less ICH component had a tendency toward higher rates of recovery. Non-structural treatable causes of coma such as seizures were associated with higher rates of recovery. Mortality and disability remain dismal in this population. 

Optimal blood pressure (BP) control in intracerebral hemorrhage (ICH) patients remains controversial. Aggressive BP reduction may limit hematoma expansion, but may also cause hypoperfusion and ischemia. We investigated the relationship BP lowering in the first 24 hours and the presence of diffusion weighted imaging (DWI) lesions on MRI.

We prospectively enrolled consecutive patients presenting with an acute spontaneous ICH. Brain MRIs were reviewed for the presence of lesions with reduced diffusion attributable to tissue compression, vessel compression, or hypoperfusion ipsilateral to the hematoma. BPs were recorded on hospital presentation, and at 6, 12, 18, and 24 hours.

Of 160 eligible patients, 136 met inclusion criteria: age: 63±17 years; hematoma volume: 10 mL (IQR 4-33); admission NIHSS: 6 (IQR 2-16); ICH onset to maximal BP reduction 18 hours (IQR 12-18); and ICH onset to MRI: 37 hours (IQR 25-75). DWI lesions were detected in 50% of patients: 79% of patients had lesions attributed to tissue compression, 21% to vessel compression, and 12% to hypoperfusion (some patients had multiple lesion types). DWI lesions were associated with larger hematoma volumes (32 vs. 12mL, p <0.001); higher admission mean arterial pressures (MAP) (125 vs. 113mmHg, p=0.006); and greater average MAP reductions (25 vs. 17%, p=0.006). After controlling for ICH volume using logistic regression: for every 10% of MAP reduction, the risk of DWI lesions increased (OR 1.28, 95% CI: 1.01-1.62); for each 30% reduction in MAP the risk of DWI lesions more than doubled (OR 2.3, 95% CI 1.05-4.83). The proportion of patients with DWI lesions increased as the maximum percent MAP reduction increased in a dose dependent fashion.

Ischemic brain lesions in patients with spontaneous ICH are common and associated with hematoma volume and BP lowering. Aggressive BP lowering may contribute to ICH associated ischemic lesions.

Financial Support: Sources of funding: This research was supported by the NIH (R01 NS034866) to CACW, and the Stanford School of Medicine Medical Scholars program to JTK. 

Coma is a major cause of death, disability and economic burden to the health care system. Acutely comatose patients with primarily neurologic injury are at risk to develop neurologic and systemic complications. In this study, we seek to identify the timing of medical complications and their impact on mortality in acutely comatose patients admitted to neurocritical care unit.

One hundred patients with acute coma for at least 12 hours or longer were enrolled prospectively in the study from May 2010 to Jan 2011. Major neurologic and systemic complications were identified prospectively and the frequency and timing of each major complication was established.

Of the 100 patients studied, mean age was 59±18.4 years and 53% were females. A mean of 3.3±2 complications occurred. 

In this cohort of patients with coma, there were more non-neurological (55%) versus neurological (45%) complications. Most complications (58%) were noted in the 3-4 day interval. Further characterization of these complications is essential to the care of comatose patients in the NCCU.

Pathophysiology of brain dysfunction associated with sepsis is still poorly understood. Our purpose was to study the metabolic alterations and mithocondrial dysfunction in a clinically relevant model of septic shock.

Twelve anesthetized, invasively monitored, and mechanically ventilated pigs were allocated to a sham procedure (n = 5) or sepsis (n = 7), in which peritonitis was induced by intra-abdominal injection of autologous faeces. Animals were studied until spontaneous death or for a maximum of 24 hours. In addition to global hemodynamic and laboratory assessment, intracranial pressure and cerebral microdyalisis were assessed 6, 12, 18 and 24 hours after sepsis induction. After death, brain were removed and brain homogenates were studied to assess mithocondrial dysfunction.

All septic animals developed a hyperdynamic state associated with organ dysfunction. In the septic animals, there was a progressive increase in L/P ratio and glycerol, as well as a progressive decrease in brain glucose concentration during the study period. The comparison between control and septic animals and the analysis of brain homogenates are undergoing.

In this model of peritonitis, cerebral metabolism was derranged, with increasing levels of L/P ratio and decreasing levels of brain glucose during study period. These alterations may play a role in the pathogenesis of sepsis-associated encephalopathy.

At Sanford USD Medical Center, Neuro Critical Care (NCC) patients are frequently treated with continuous infusions of 3% sodium chloride. It has been observed that this patient population often develops iatrogenic hyperchloremic metabolic acidosis, frequently managed with intravenous sodium bicarbonate. Upon notification of a nationwide intravenous sodium bicarbonate shortage (March 18 th , 2012), our NCC providers were forced to explore other potential options for managing this acidosis. It was decided that our NCC patients would be initiated on enteral sodium bicarbonate at the time continuous 3% sodium chloride was started.

A retrospective chart review of NCC patients 18 years and older, initiated on continuous 3% sodium chloride with enteral sodium bicarbonate tablets from March 18 th , 2012 to June 4 th , 2012 were evaluated. Data collected included demographics and the following while in the Intensive Care Unit (ICU): baseline serum sodium, chloride, and bicarbonate; type of injury; acidosis defined as serum bicarbonate level <20, or lower than baseline; and volume of 3% sodium chloride and bicarbonate administered.

Of the 11 patients identified, 27.3% developed iatrogenic metabolic acidosis during ICU stay. Average duration of continuous 3% sodium chloride infusion was 6 days (range 1 -22 days) with an average volume of 3% Sodium Chloride dispensed of 5,000 mLs (range 500 -32,000 mLs). Three patients evaluated developed an acidosis during ICU stay, of which 2 were hyperchloremic at the time of acidosis. Only 1 patient required intravenous sodium bicarbonate, however the patient had been off hypertonic saline for more than 1 day.

Enteral sodium bicarbonate appears to be an effective method at preventing iatrogenic hyperchloremic metabolic acidosis when initiated along with continuous infusions of hypertonic saline in NCC patients. This may be a method to conserve intravenous sodium bicarbonate during drug shortages. Further studies are needed.

Fever is common in neurocritically ill patients. It can be from central causes, inflammatory, infectious, and other conditions. A method to differentiate infectious from non-infectious fever would allow for appropriate initiation of empirical antimicrobial therapy. Apart from avoiding unnecessary antimicrobial usage, this approach can save health-care costs and limit the development of antimicrobial resistance. Procalcitonin is a peptide precursor of the hormone calcitonin. Procalcitonin levels rise as a proinflammatory response to bacterial infections. Numerous studies have evaluated procalcitonin levels utility in the initiation and discontinuation of antibiotics in the inpatient setting; however, there is a paucity of studies regarding the use of procalcitonin levels in neurologically ill patients. This study examines the effectiveness of a procalcitonin-guided algorithm in a neurocritical care unit.

A modified PRORATA trial procalcitonin algorithm was developed and utilized prospectively. Patients that met criteria of 1) admission into the neurocritical care unit 2) age > 18 years 3) temperature > 101 F in the last 24hrs 4) no obvious source of infection were enrolled. Depending on the procalcitonin level and the presence of new SIRs criteria antibiotics were initiated per our algorithm. Radiographical, microbiological, laboratory, and clinical outcomes were recorded to determine the accuracy of the procalcitonin algorithm in the decision to initiate or modify patient's antibiotic therapy.

Results from the first 20 enrolled patients found the procalcitonin algorithm had 67% sensitivity and 88% specificity in predicting bacterial infections as the etiology of fever with a positive predictive value of 89%. The study population included intracranial hemorrhage (55%), ischemic stroke (20%), and others (25%). 60% of the study population had infectious fever while 40% had non-infectious fever.

Interim results suggest a procalcitonin-guided algorithm may be a valuable tool in differentiating infectious from noninfectious fever in the neuro-ICU. Further research is needed; data collection is ongoing.

Posterior reversible encephalopathy syndrome (PRES) is defined by acute neurologic symptoms caused by vasogenic cerebral edema. Recurrence of PRES is thought to be rare and has not been well described.

Patients prospectively diagnosed with PRES from 2005-2012 were pooled with retrospectively identified patients diagnosed with PRES from 1999-2005 at an academic referral center. Detailed clinical information and radiologic imaging results were collected. Patients without clinical or radiographic resolution between episodes and patients without brain imaging available for review were excluded.

Of a total of 189 patients with PRES, 12 (6%) had recurrence. One patient had four episodes, one patient had three, and ten patients had two episodes each, resulting in 27 total PRES episodes. Seven patients (58%) had an autoimmune disorder. The average time between episodes was 14 months. Acute hypertension was present in 22 of 27 episodes. Of these, mean blood pressure was 198/116 mmHg. Etiologies of PRES included hypertension (n=15), cytotoxic medications (n=2), sepsis (n=2), and multifactorial (n=5). Renal failure was present in 13/27 episodes, and was acute in 5. Clinical symptoms included headache (n=17), seizure (n=17), visual disturbances (n=9), encephalopathy (n=9) and focal deficits (n=1). Only one patient (8%) had the exact same clinical symptoms with recurrence. Ten patients had MRI at each episode of PRES. Vasogenic edema affected the same brain areas at each episode in 5 patients. In the rest, some affected regions were similar, but additional regions were different between PRES episodes. None had entirely new areas of involvement.

PRES recurred in approximately 6% of our patients. In the majority, clinical symptoms differed at recurrence compared to the initial episode. In all patients, radiologic patterns of vasogenic edema in the repeat episode were similar to those in the initial PRES episode, but also affected other brain regions in approximately 50%.

Ventilator-associated pneumonia (VAP) is a common complication in comatose patients. Diagnosis in this population is unreliable despite physician training and validated criteria leading to potential misdiagnosis and inappropriate antimicrobial use.

We investigated clinical features associated with misdiagnosis of VAP and excess antibiotic days (EAD). Ventilated comatose patients (Glasgow coma scale motor score <6) suspected of having VAP were prospectively identified in a Neurocritical Care Unit in 2011. VAP was retrospectively diagnosed using Centers for Disease Control (CDC) criteria by two neurointensivists and an infection control practitioner. Appropriateness of the NCCU team's VAP diagnosis and therapy was performed using clinical, microbiologic and radiographic data.

Of 173 comatose patients, 34 cases were treated as possible VAP by the NCCU team. Of these, 13 patients had VAP by CDC criteria. VAP and non-VAP groups did not differ in age, admission GCS, total ventilator days or mean total antibiotic days (11.38±9.0 (VAP) vs. 13.23±7.9 (non-VAP); p=0.53). Clinical features significantly associated with VAP (vs. non-VAP) were change in sputum character, tachypnea, oxygen desaturation, persistent infiltrate on chest xray and positive sputum microbiology. Two-thirds (66.7%) of non-VAP patients received pneumonia targeted antibiotics for >8 days vs. 30.1% of VAP patients (p=0.04), contributing 88 EADs, including 66 vancomycin days, 22 piperacillin-tazobactam days and 44 cephalosporin days. Median days from intubation to starting antibiotics was 1 (non-VAP) vs. 4 (VAP) days (p = 0.13). No pre-specified factors were associated with inappropriate continued VAP treatment.

Inappropriate diagnosis and treatment of VAP resulted in a cumulative 141 EADs in one year in the NCCU. Clinical differences between patients without VAP who had antibiotics continued or discontinued were minimal, suggesting that clinician behaviors contribute to unnecessary prescribing. Strategies to improve the diagnosis of and antibiotic use for VAP in comatose patients is needed. 

Management of hyponatremia in patients with acute brain injury can be challenging. The oral vasopressin receptor antagonist has been studied extensively in other disease process but not in acute brain injury. We report our experience regarding the efficacy and safety of tolvaptan, an oral vasopression V 2 -receptor antagonist, for the correction of hyponatremia in acutely brain-injured patients.

tolvaptan for the correction of euvolemic or hypervolemic hyponatremia.

Baseline serum sodium concentration was 128.9 ± 3.1 mEq/L. Seven patients received 15 mg of tolvaptan once (singledose-users), and 11 patients received another 15 mg on the next days (double-dose-users). 24 hours after tolvaptan administration, serum sodium concentration increased by 5.5 ± 2.1 mEq/L in single-dose-users (p = 0.016) and 5.3 ± 5.0 mEq/L in double-dose-users (p = 0.004). 48 hours after administration of first dose of tolvaptan, serum sodium increased by 7.4 ± 2.2 mEq/L in single-dose-users (p = 0.016) and 9.9 ± 5.5 mEq/L in double-dose-users (p = 0.002). During four days of observation, the increases in the average area under the curve of the serum sodium concentration was 5.2 ± 1.9 mEq/L in single-dose-users (p = 0.016) and 7.8 ± 4.2 in double-dose-users (p = 0.002). Urine output increased by 2.3 ± 1.9 L during the first 24 hr in single-dose-users (p = 0.038). No significant changes in fluid balance, serum creatinine and Glasgow Coma Scale were observed. Of four patients who underwent neuro-monitoring, intracranial pressures, cerebral perfusion pressures and mean arterial pressure did not change significantly compared with their baseline values.

Tolvaptan was effective and well-tolerated for the correction of hyponatremia in patients with acute brain injuries. Validation can be done with further studies.

Patients with acute brain injury but normal lung function often undergo intubation and subsequent tracheostomy for the concern of airway protection. We previously described patients with primary brain injury and encephalopathy who fail extubation demonstrated signs of disrupted ventilation usually with periods of prolonged hypoventilation. We examined the clinical characteristics of patients with a tracheostomy who are readmitted to the ICU with respiratory decompensation.

Retrospective review of patients admitted to the neurocritical care unit (NCCU) of a tertiary care hospital who underwent tracheostomy from September 2009 to June 2011.

Of 90 patients who received tracheostomies during their admission to the NCCU, 38 (42%) were successfully transferred to the floor, 22 (25%) were readmitted to the ICU, and 30 (33%) had other dispositions such as discharge to rehabilitation and withdrawal of care. There were a total of 28 readmissions, 12 due to respiratory decompensation and 3 due to cardiopulmonary arrest. 

Hypoventilation is commonly seen in neurological patients who receive a tracheostomy. Potential predictors of respiratory decompensation and readmission of these patients include their brainstem reflexes and respiratory patterns as assessed by the FOUR score as well as their duration of mechanical ventilation.

Twenty-two percent of neurocritically ill patients may become hypernatremic. Moderate-severe traumatic brain injury (msTBI) patients may develop hypernatremia possibly from diabetes insipidus (DI) or osmotherapy for brain edema treatment. Retrospective studies suggest that hypernatremia (serum sodium [sNa] >160 mMol/L) may be associated with an increased risk of death. However, these studies failed to adjust for DI and the use of osmotherapy. We examined the impact of mean and peak sNa on in-hospital mortality and functional outcome at hospital discharge, adjusted for these important variables.

In a prospective observational study, 144 consecutive msTBI patients from a single Level I trauma center between 11/2009-2/2012 were analyzed. Poor outcome was defined as Glasgow Outcome Scale (GOS) 1-3. Multivariable logistic and ordinal regression was utilized to adjust for admission characteristics, injury severity, ICU length-of-stay, brain edema, osmotherapy (mannitol/hypertonic saline), and DI. Firth's Method was used in logistic regression models to accommodate small sample sizes.

The mean age was 51 years, 70% were men, and the median Glasglow coma scale and injury severity scores were 5 and 32, respectively. Higher mean and higher peak sNa were significantly associated with worse outcomes, both when using the dichotomized (OR 1.3; 95% CI 1.1-1.5 for mean and OR 1.1; 95% CI 1-1.2 for peak sNa) and ordinal GOS (OR 0.8; 95% CI 0.7-0.9 for mean and OR 0.9; 95% CI 0.86-0.98 for peak sNa). For every 5 mMol/L increase in mean sNa and every 12 mMol/L increase in peak sNa, patients worsened by one GOS category.

Our results suggest that higher sNa values are associated with worse neurological outcome, independent of osmotherapy, brain edema and DI. It will be important to determine which sNa might be "too high". 

While autonomic instability occurs as part of anti-N-methyl D-aspartate (anti-NMDA) receptor encephalitis, anti-NMDA receptor encephalitis is not a recognized cause of the clinical syndrome of paroxysmal sympathetic hyperactivity (PSH).

We present a case of anti-NMDA receptor encephalitis in which PSH was a cardinal feature.

A 31-year-old woman had a generalized tonic-clonic seizure, and then developed progressively worsening neuropsychiatric symptoms, including mania, hallucinations, echolalia, and suicidal ideation. Diagnostic work-up revealed anti-NMDA receptor antibodies detected in the serum and in the cerebrospinal fluid (CSF). One week after symptom onset, the patient experienced intermittent episodes of sinus tachycardia, hypertension, tachypnea, diaphoresis and extensor posturing. The episodes were both spontaneous and stimulus responsive (for example, during endotracheal suctioning). The episodes, consistent with PSH, were initially treated with dexmedetomidine, which was titrated to effect. Gabapentin and propranolol were added later for symptom control, but eventually weaned off as her symptoms abated approximately six weeks into the illness.

We believe that the autonomic instability associated with anti-NMDA receptor encephalitis may often be PSH. PSH often goes unrecognized in patients outside of the setting of TBI, thus specific PSH management strategies may be overlooked in other contexts. Anti-NMDA receptor encephalitis may represent the functional companion to the structural lesion encountered in TBI. Recognition of PSH in this setting is important to guide the management of the autonomic instability, but may also have mechanistic implications.

A 34-year old male with history of motor vehicle accident s/p frontal sinus surgery was admitted with streptococcus pneumoniae meningitis and altered mental status. Upon admission, he was febrile with leukocytosis. Head CT showed left sinus opacification and CSF studies were consistent with bacterial meningitis. Despite broad-spectrum antibiotics and interval improvement in his head CT and CSF studies, his mental status continued to decline. Shortly after ICU admission, he became lethargic with a new right-sided hemiparesis. CTA revealed diffusely narrowed intracranial arteries most compatible with vasospasm and MRI was consistent with multiple areas of infarction. His neurological exam continued to deteriorate necessitating intubation for airway protection. TCDs showed bilateral MCA vasospasm.

Initially, vasospasm was managed with nimodipine, hypertension, and euvolemia. Systolic blood pressure was artificially elevated with vasopressors, inotropes, and ultimately with methylene blue. Despite aggressive medical management, there was little improvement clinically. Therefore, he received four sessions of angiography with intra-arterial verapamil.

After the final intra-arterial verapamil treatment, he demonstrated angiographic and clinical improvement. We conclude that patient's cerebral vasospasm was a direct complication of streptococcus pneumonia meningitis. Intra-arterial verapamil appears to be effective in treating pneumococcal meningitis induced symptomatic cerebral vasospasm. However, there is limited data to predict its vasodilatory sustainability and optimal treatment intervals.

Pneumococcal meningitis is the leading cause of bacterial meningitis beyond the neonatal period. Clinical and experimental research had demonstrated that vascular alterations are common in bacterial meningitis and are associated with stroke. Despite the introduction of the pneumococcal vaccine, availability of effective antibiotics, and advances in adjunctive strategies, mortality and morbidity rates associated with arterial complications secondary to pneumococcal meningitis remain high. This case is noteworthy because to our knowledge this is the first reported case of pneumococcal bacterial meningitis induced vasospasm that has been successfully treated with intra-arterial verapamil.

Xuemei Cai 1, 3

Osmotic myelinolysis is a life threatening problem associated with rapid correction of chronic hyponatremia. The brain cannot readily restore organic osmolytes; thus rapid correction of serum osmolality leads to cellular shrinkage causing axonal dissociation from myelin sheaths. Current guidelines state that serum sodium (SNa) should be corrected at a rate not exceeding 6-10mEq/L/day but when extracellular volume depletion is the cause, vasopressin suppression after saline treatment increases risk of rapid overcorrection. There is no standard of care that directs treatment once osmotic myelinolysis occurs. We report a case of a patient who developed clinical symptoms of osmotic myelinolysis syndrome who was successfully treated with re-induction of hyponatremia which led to complete neurological recovery.

A 71-year-old woman on thiazide treatment for hypertension developed protracted vomiting and diarrhea for several days followed by confusion and lethargy. In the emergency department, SNa was 107mEq/L. She received isotonic saline and over the next 12 hours, SNa rose 21mEq/L. On hospital day two, her neurological condition deteriorated rapidly with development of mutism, increased tone in all extremities, hyperreflexia. Osmotic myelinolysis syndrome was diagnosed on clinical grounds. She was given desmopressin with 5% dextrose in water (D5W) to rapidly lower her lower her SNa. Her neurological status improved at a SNa of 115mEq/L. Thereafter, SNa was slowly uptitrated with desmopressin and 3% normal saline. She made a complete neurological recovery. MRI performed at discharge and one month later showed no abnormalities.

Overcorrection of SNa in chronic hyponatremia is a common iatrogenic problem which can lead to osmotic myelinolysis syndrome, a highly morbid and oftentimes fatal neurological condition. Our case supports immediate re-induction of hyponatremia in patients with symptoms suggestive of osmotic myelinolysis at a time when imaging may be unremarkable and complete neurological recovery is achievable.

Posterior reversible encephalopathy syndrome (PRES) is manifested by acute neurological findings with evidence of vasogenic edema on brain imaging possibly due to cerebral vascular endothelial dysfunction. The epidemiology of PRES in pediatric critical care has not been well described and it may be under recognized and thus prompt treatment delayed.

We performed a retrospective review of all patients with diagnosis of PRES over 18 month period (January 2010 to June 2011) in a pediatric critical care unit (PCCU) at a tertiary care university hospital. Data from hospitalization and 6 month follow up were reviewed.

There were 1221 admissions to PCCU and 141 neurology service consultations during the study. Six patients were diagnosed with PRES (Incidence -1 in 203 PCCU admissions) with median age 17 years (Mean±SD; 14.8±5.74 years).

All patients presented at onset with generalized tonic-clonic or clonic type seizures that lasted up to 12 hrs and returned to baseline mental status in 2-7 days. Other clinical features were headache and visual impairment. Risk factors preceding the onset of PRES included anemia [hemoglobin 8.6±1.6 g/dl], azotemia, hypertension, hypernatremia, hypocalcemia, hypomagnesemia, and recent use of chemotherapy (azathioprine, cyclophosphamide, tacrolimus and mycophenolate mofetil). Brain MRI demonstrated increased T2/FLAIR signal within the parieto-occipital white matter in all 6 patients, frontal lobe changes in 5 patients and vertebro-basilar system changes in 2 patients. No regions of restricted diffusion were seen on diffusion weighted imaging. At 6 month follow up, no patients had residual neurological deficits from PRES and neuroimaging revealed significant resolution of white matter signal changes.

PRES is associated with multiple disease states including systemic lupus erythematosus, sickle cell disease, sepsis, recent use of cytotoxic medications and renal failure. Knowledge of the risk factors associated with PRES, its clinical presentation, and characteristic MRI findings may lead to more rapid recognition and treatment. 

Adults with neurological injury are at increased risk for tracheobronchial foreign body aspiration. This report will present a case of silent foreign body aspiration in a patient who presented to the emergency department with status epilepticus.

Case report and review of the literature.

An 87 year-old African American man presented to the emergency department with status epilepticus. Seizures were controlled with intravenous lorazepam and fosphenytoin, and the patient was intubated for airway protection. On day four following admission to the Neurosciences Critical Care Unit, a routine magnetic resonance imaging (MRI) scan demonstrated susceptibility artifact from a metallic focus which completely obscured the spine structures at C4-C5. Upon review of the patient's previous imaging, numerous abnormalities were reported on daily chest x-rays and a foreign body was identified within the trachea on a thoracic CT from admission. A bronchoscopy was performed which revealed a watchband within the trachea and right mainstem bronchus.

Tracheobronchial foreign body aspiration should be considered in patients with unexplained respiratory symptoms, and a high degree of clinical suspicion should be maintained in patients with neurologic impairment. Abnormalities on chest xray and computed tomography should prompt an early pursuit of the diagnosis in high-risk patients. MRI, although generally considered to be a safe imaging modality, could be potentially harmful to patients with unidentified foreign bodies.

Hypokalemic periodic paralysis (HypoPP) is a disease characterized by muscle weakness or paralysis secondary to low serum potassium levels. Neurogenic diabetes insipidius (DI) is a condition where patient excretes large volume of diluted urine due to low level of Anti-Diuretic Hormone (ADH). Here, we report a case of HypoPP in a patient with neurogenic DI.

A 30 year-old right-handed Hispanic male was admitted for seizures after developing a dental abscess. This patient had a history of pituitary adenoma resection at the age of 12 with subsequent pan-hypopituitarism for which he was on hormonal supplementation. On hospital day three, he developed sudden onset of quadriparesis with motor strength in upper extremities 0/5 bilaterally and 1/5 in both lower extremities and absent deep tendon reflexes throughout. His routine laboratory studies showed severe hypokalemia of 1.6 mEq/dl. Nerve Conduction Study (NCS) revealed absent compound motor action potentials with normal sensory potentials. Electromyography (EMG) revealed no abnormal insertional activity or spontaneous activity. Some muscles demonstrated no volitional motor units and a few others had decreased recruitment in distal small motor units. Following aggressive correction of the hypokalemia he regained his full strength and repeat EMG showed normal motor units, normal recruitment, but no myotonic discharges. NCS showed return of compound motor action potentials in all nerves tested.

HypoPP remains an important differential in an acute case of paralysis and acute management is important. 

We report a case of a 57-year-old Caucasian male who presented to a community hospital with complaints of flu-like symptoms. He underwent pulmonary-vein isolation for chronic atrial fibrillation thirty days prior to admission. His history includes left frontal and right parietal ischemic infarcts, mitral valve repair, coronary artery bypass grafting, patent foramen ovale closure, and coronary artery disease. Approximately 30 hours prior to arrival, he developed nausea, vomiting, fatigue and confusion. He was febrile and appeared encephalopathic. A telemedicine stroke consultation recommended transfer to a tertiary care facility. While the initial concern was for acute cerebral ischemia, he did not meet exclusion criteria for thrombolytic therapy. The patient received aggressive initial hydration and broad spectrum intravenous antibiotics for coverage of meningitis. Blood cultures, complete blood count, comprehensive metabolic panel, urinalysis, stool culture and a lumbar puncture were performed. Interestingly, his blood cultures remained persistently positive for gram positive cocci in chains and clusters. Occult stool was positive and his oral gastric tube demonstrated bloody drainage. The remainder of his laboratory work was unremarkable. CT scan of his head revealed old ischemic infarcts without hemorrhage or hypodensity. The patient continued to decompensate in the Neurointensive care unit where he eventually required intubation. A CT scan of the chest was highly suspicious of a left atrial-esophogeal fistula.

Cardiothoracic surgery was notified of the atrio-esphogeal fistula and he was taken to the operating room for a right thoracotomy with repair of the fistula and intercostal muscle flap. Post-operative MRI brain demonstrated innumerable air emboli and diffuse areas of ischemic infarction. Atrio-esophageal fistula is a very rare complication following pulmonary vein isolation, and because prognosis is dependent upon prompt surgical correction, neurointensivists should be aware of this entity

Financial Support: None

Propofol infusion syndrome (PRIS) is a rare but devastating complication of high dose administration of diprivan in children and young adults which presents with metabolic acidosis, rhabdomyolysis and fatal cardiac dysrhythmias.

We report a case of PRIS in a 20-year old, previously healthy, postpartum female who received a high dose diprivan infusion for 48 hours at an outside institution for the treatment of presumed refractory convulsive status epilepticus.

Patient received diprivan100mcg/kg/min for the first 24 hours. Diprivan was increased to 150mcg/kg/min to achieve burst suppression on the electroencephalogram. Diprivan was stopped after 48 hours due to lactic acidosis. Subsequently patient developed renal failure and elevation of CK up to 140,000.She was transferred to our institution for continuous hemofiltration and possible extracorporeal membrane oxygenation (ECMO).After transfer she developed atrial fibrillation, ventricular tachycardia and fibrillation. An ECMO catheter was placed when she was in ventricular fibrillation for 60 minutes. After starting ECMO the patient developed asystole for 8 hours, requiring a transvenous pacemaker. Her cardiac dysfunction improved rapidly and ECMO was discontinued after 5 days. The patient started to follow commands consistently at 14 days after the onset of fulminant PRIS. MRI of the brain showed a subacute right posterior cerebral artery infarct attributed to cardiac embolism. The patient left intensive care unit after 3 weeks.

Close metabolic and cardiac monitoring should be applied when a patient is on high-dose of diprivan(>80mcg/kg/min). Diprivan should be stopped as soon as unexplained metabolic acidosis, rhabdomyolysis and cardiac dysrhythmias are noticed, and transfer the patient to a center with continuous hemofiltration and ECMO capabilities should be considered. ECMO can be a lifesaving intervention in patients with fulminant PRIS.

Postpartum cerebral angiopathy (PCA) is a rare pregnancy complication. PCA is often a benign condition that resolves spontaneously, but can lead to stroke or death. The purpose of this case study is to describe events that transpired in the care of a patient with severe persistent PCA, for whom unconventional treatment was initiated because conventional treatment failed.

Retrospective and current chart reviews were conducted, including relevant medical history. Objective data related to the patient's condition were reviewed. We examined the evolution of medical and nursing care as the patient's condition deteriorated despite aggressive conventional therapy, and reviewed ensuing events: multidisciplinary collaboration to search for other viable treatment options, consultation with colleagues from another major medical center regarding their experience with nicardipine and recommendations on off-label use for PCA, and decision-making including the family about whether or not to administer intraventriuclar nicardipine.

Multiple disciplines (i.e., doctors, nurses, and pharmacists) and family members contributed to the complex decision to initiate unconventional treatment. We administered 2 mg intraventricular nicardipine every eight hours for seven days. Using transcranial dopplers, cerebral arteriograms, and clinical assessment data, we evaluated the effectiveness of this unconventional treatment. After seven days, we discontinued the nicardipine, while continuing standard treatment to maintain hypertension and hypervolemia. Currently, the patient is expected to make a full recovery with few residual stroke deficits.

A multidisciplinary approach, including the family in the decision-making process, enabled creative problem-solving for a challenging clinical situation. When conventional methods failed, our team collaborated to think outside the box and take a calculated risk, altering the course of our patient's condition from critical toward survival and recovery.

The objective was to determine the diagnostic yield and safety of brain MRI in critically ill patients with ICU-acquired acute brain dysfunction.

Patients in the medical and surgical ICUs who developed acute brain dysfunction and underwent brain MRI were included. Patients with preexisting brain disorders and those from neurological ICUs were excluded. MRI scans were analyzed by three specialists trained in neuroimaging. Outcome variables included Glasgow Outcome Scale at discharge (1-3 categorized as unfavorable and >3 as favorable) and death.

146 patients underwent brain MRI for evaluation of encephalopathy, seizures, focal deficit. Signs of parenchymal brain abnormalities were detected in 130 patients (89%) including white matter hyperintensities in 71.2% and acute cerebral infarcts in 40.4%. Results from brain MRI led to modification of diagnosis and treatment in 53% of cases. Patients with MRI defined lesions were more likely to have an unfavorable outcome. There were no adverse events from transportation to the radiology site or from MRI performance.

In ICU patients with acute brain dysfunction, MRI is a safe noninvasive diagnostic tool that often leads to substantial modification of diagnosis and treatment. Structural brain injury contributes significantly to the pathogenesis of cerebral dysfunction during critical illness and should be taken into account even if other reasons for encephalopathy are presumed. 

Central nervous system (CNS) and intraventricular infections are a devastating complication for patient admitted to an intensive care unit. The use of intrathecal (IT) antibiotics for the treatment of CNS infections has been reported in small case studies. Our purpose was to report patients who have received IT antibiotics for intraventricular infections in our facility and discuss our findings.

Retrospective case series of patients who received intrathecal antibiotics in combination with systemic antibiotics for treatment of intraventricular CNS infection over the past 7 years. Basic demographic and clinical measures were collected from the hospital data base.

Seven patients received IT antibiotic therapy for CNS infection. Admitting diagnoses were head trauma (1), intracranial hemorrhage (4), and subarachnoid hemorrhage (1). One patient had an infected ventriculoperitoneal shunt. All of the patients received an external ventricular drainage device during admission prior to developing CNS infection. Time from hardware placement to first positive CSF culture for 1 patient was 4 days; 2 patients were positive with first CSF; 2 were within 14 days; and 1 had his VP shunt in place for 36 days prior to positive cultures. Pathogens cultured from CSF included Klebsiella pneumoniae, Acinetobacter baumannii and Vancomycin-Resistant Enterococcus faecalis in 2 patients each, and Methicillin-Resistant Staphylococcus aureus in 1 patient. The intrathecally instilled antibiotics were colistin, streptomycin, tobramycin and vancomycin. Two of the patients cleared CSF cultures in 1 day, 1 patient cleared in 3 days, 2 patients cleared in 6 days and 1 took 26 days to clear CSF.

Based on this small case series we found IT antibiotic adjunct therapy as a viable option for treating CNS infections as most of our patients cleared CSF within 6 days of treatment initiation. Further studies are warranted to support our findings.

We report a case of an esthesioneuroblastoma or Olfactory Neuroblastoma (ONB) presenting with frontal lobe dysfunction and hence depression with rapidly declining mental status resulting from hydrocephalus and stroke meningitis.

This is a 48year old man who presented with fever, headache and AMS. He had 12 months history of progressive headache, face pain, rhinorrhea, nasal congestion and depression. CT head showed destruction of the cribriform plate by a mass arising from the right nasal cavity with extension into the right inferior frontal cranial fossa. An EVD was placed emergently for elevated ICP. He was also found to have multiple strokes in the right basal ganglia and corpus callosum. An incidental mycotic aneurysm was seen at the right posterior cerebral artery. Labs showed WBC of 23, Sodium of 122, potassium of 3, bicarbonate is 30. A Lumbar Puncture was performed which showed evidence of bacterial meningitis. A diagnosis of ONB was established by histopathology and confirmed by immunohistochemistry. On staging, the mass was classified as a Kadish stage C tumor.

He underwent coiling of the pseudo aneurysm of right PCA and maxillary embolization, followed by bifrontal craniotomy and endovascular resection of tumor

ONB is a rare malignant tumor of neuroectodermal origin and is thought to arise from the olfactory epithelium. Symptoms are related to nasal obstruction, orbital extension, invasion of thecribriform plate, paraneoplastic syndromes with hypercalcemia and hyponatremia and can cause frontal lobe dysfunction. Physical examination generally reveals a vascular, polypoid mass located in the nasal cavity. MRI helps to differentiate tumor from other causes of nasal obstruction. They typically stain for neuron-specific enolase (NSE). There has been no standardized RCT done due to rarity of the tumor but traditionally the mainstay of treatment in such locally advanced patients is combinedotolaryngologic and neurosurgical craniofacial resection followed by adjuvant radiotherapy. 

We describe a case of delayed PTLD in a 51 year old diabetic patient with ESRD several years after multiple solid organ transplants; a successful pancreatic transplant and a rejected renal transplant. She initially presented with mild left hemiparesis and was found to have enhancing and non-enhancing both supra and infra tentorial lesions, without evidence of disease in the graft, skin or bone marrow. The histological diagnosis of PTLD was made after a right frontal brain biopsy. She had intercurrent worsening of left hemiparesis post biopsy due to hemorrhagic transformation of one of the lesions. The patient initially responded to a decrease in immunosuppressive medications which included Tacrolimus and Cellcept however, she eventually also required Rituximab and whole brain radiation to maintain remission.

In this case report we highlight the manifestations of CNS PTLD, dilemmas in diagnosis and various strategies for management. This can be a fatal complication of solid organ transplants if not recognized and treated early.

Dysautonomia has been well associated with Guillain Barre Syndrome (GBS). The dysautonomic effects of GBS may cause a variety of reversible clinical syndromes associated with sympathetic dysfunction including PRES and takotsubo cardiomyopathy. PRES can be a presenting feature following GBS treatment with intravenous (IV) immunoglobulins or may present later in recovery. Dysautonomia resulting from GBS is the most likely explanation for this assocication while another possible mechanism can be the influence of cytokines, produced in the context of GBS, on the permeability of blood brain barrier. In this abstract we highlight a self limited case of PRES presenting as an early complication of GBS.

Case: Our patient was a 57 year old female with Hypertension who presented to an outside hospital with alteration in mental status. She had developed bilateral lower extremity weakness and difficulty ambulating for 3-4 days prior to admission. She reportedly had an upper respiratory infection about 2 weeks prior to presentation. At the time of transfer to our hospital the patient had a generalized tonic clonic seizure and was started on Keppra. She had a fluctuating mental status from being awake to stuporous. Bilateral lower extremity power was 2/5 in all muscle groups. Initially, deep tendon reflexes were 1+ in lower extremities but after a few hours she became areflexic in lower extremities with 1+ reflexes in upper extremities and downgoing plantars. MRI brain T2/FLAIR images showed lesions consistent with PRES. CSF showed cyto-albuminologic dissociation and diagnosis of GBS was made. She was started on a 5 day course of iv IG. She was discharged to a rehab facility with some improvement in her paraparesis and no recurrent seziures.

This case report illustrates that patients can develop PRES as a complication of GBS perhaps due to dysautonomia but PRES may be self limited in this setting.

Data exists describing the outcomes of critically ill patients with specific conditions in specialty intensive care units (ICU) versus general ICUs. Severe sepsis and septic shock(SS/SH) outcomes have not been robustly evaluated in community hospitals between specialty ICUs. We chose to evaluate whether patients admitted to ICUs with SS/SH would have higher mortalities in Neuroscience (NS) and Cardiac (Cards) ICUs versus general medical surgical ICUs (MSICU).

Intensivists. The variables collected include age, time to antibiotics, intravenous fluids given, central line placements, code status, vasopressor requirements at 6 hours and mortality. Chi-square analysis was used to compare mortality rates.

ICUs who were directly admitted from the ED were NS 39% (n =12), Cards 36%(n=34), MSICU 33% (n=34) p=.38.

The mortality rate for patients admitted with SS/SH was similar independent of the type of the ICU the patients received care in. A multivariate analysis needs to be done to confirm these outcomes

Neurocrit Care (2012) 17:S1-S337 S193 

Thromboembolism is a known and feared complication of administering prothrombin complex concentrates (PCC) but the true incidence is unknown. Most data is in regards to MI, DVT, PE and DIC with little reported on ischemic stroke. This is the first known report in the literature of acute basilar thrombosis after reversal of anticoagulation with PCC.

We present a 77 year old women with acute basilar thrombosis after reversal of anticoagulation using PCC (Profilinine SD). She was admitted with a hemodynamically stable lower GI bleed with a supratherapeutic INR of 10. She was taking Coumadin for a recent pulmonary embolus. Anticoagulation was reversed using Profilnine SD 4950 Units (47 U/kg) and vitamin K 10 mg intravenously. 2 hours later she developed left facial weakness, quadraparesis and anarthria. CT brain showed no early ischemic changes. CT angiogram showed occlusion from the mid-basilar to the basilar apex with normal vertebral arteries from the origin to the site of occlusion. Factor II activity was elevated with normal activity of factor VII, IX and X. TTE showed normal wall motion and ejection fraction without evidence of thrombus or shunt.

PCC protocols for reversal of anticoagulation are used with increasing frequency, even in non-emergent situations. Thromboembolism is a known complication of administration, even with modern formulations of PCC which include anticoagulants. Risk of thromboembolism increases with doses above 40 U/kg and with repeated dosing. The cause of thrombogenicity remains uncertain. Accumulating data indicates the importance of factor II (prothrombin) which has a linear relationship with thrombin generation. Our case suggests that given potentially fatal thromboembolic complications, PCC administration should be weighed against the need for rapid correction of coagulopathy. More discussion is needed regarding complications of PCC administration, optimal dosing and uniform production of PCC products on the market.

Endovascular reperfusion reduces infarct volume to improve clinical outcome; however treatment effect may be diluted by subsequent care. An exploratory analysis was done to determine if discharge disposition impacted 90 day mRS after definitive reperfusion therapy. 

In our study, patients discharged to SNF & AR after thrombectomy have similar medical & neurological severity at admission and similar final infarct volumes at discharge. Despite these similarities, patients discharged to SNF had a significantly lower probability of achieving a good neurological outcome. Further study is required to determine if AR should be considered in more patients to improve clinical outcomes. 

Patients with acute ischemic stroke develop respiratory failure due to airway compromise from loss of protective reflexes or cerebral swelling. In such patients, traditional weaning parameters poorly predict successful extubation. Failure of extubation increases complications, prolongs hospitalization and increase cost of care. We hypothesize that predictive factors can be identified in determining ischemic stroke patients with respiratory failure who can be successfully extubated.

Between January 2007 to December 2010, 166 consecutive patients admitted to a metropolitan academic stroke center with acute ischemic stroke and were mechanically ventilated within 48 hours of admission were reviewed after IRB approval. Patients who were intubated for procedures only, extubated within 48 hours, or placed on comfort measures were excluded, leaving 42 patients for analysis. Statistical analysis was done using SAS 9.1 and univariate or multivariate logistic regression was performed when appropriate.

Of the included patients, the average age was 65.3 ± 11.5 years, and 27 (64.3%) were male. The median admission NIHSS was 13.5 and majority of patients had cardioembolic (20) or large vessel atherosclerotic (20) strokes. 17 patients had posterior circulation stroke (40.5%). Eleven patients failed extubation (26.2%). Acute basilar occlusion was found to be a strong predictor of extubation failure (OR=21.3 95%CI: 2.7-167 p=0.004) when adjusted for age, stroke severity and duration of mechanical ventilation. Increasing age and higher NIHSS showed trend toward increased risk for extubation failure but did not reach statistical significance. Hospital length of stay doubled, ICU length of stay tripled, and total hospital cost doubled in patients who failed extubation.

Patients with respiratory failure due to acute stroke from basilar occlusion were more likely to fail extubation. Patients who fail extubation had longer ICU and hospital stay doubling the cost of care. Further studies are needed to determine whether preemptive tracheostomy may be beneficial in this group of patients.

Early detection of patients likely to develop malignant middle cerebral artery (MCA) infarction (mMCAI) is essential to enable timely decision for promising interventions (e.g., decompressive hemicraniectomy). This study was designed to evaluate whether quantitative EEG (qEEG) could predict mMCAI within 6 hours of stroke onset.

This prospective, observational cohort study enrolled 67 patients with a MCA infarct. All of them underwent EEG monitoring within 6 hours after symptom onset. Subsequently, their raw EEG data were quantitatively analyzed and the qEEG parameters including (delta+theta) / (alpha+beta) ratio (DTABR) and brain symmetry index (BSI) were computed based on the power spectral density. Patients were classified in the mMCAI group if they had decline of consciousness with radiological signs of space-occupying brain edema, whereas the others were allocated into the non-mMCAI group.

For the 2 groups, we compared the above qEEG parameters, and clinical and imaging variables. Univariate and multivariate discriminant analysis was used to determine the most accurate predictors of mMCAI.

Of the 67 patients included, 37 developed mMCAI. Univariate analysis showed that the values of DTABR and BSI, the NIHSS scores on admission and a hypoattenuation on admission cerebral computed tomography (CCT) scans > 50% MCA territory were significant predictors of mMCAI. The further logistic regression analysis identified BSI > 0.49 (odds ratio [OR] 10.32, 95% confidence interval [CI] 6.78 to 16.19; P = 0.001) and the infarct size > 50% MCA territory on CCT scan at admission (OR 5.17, 95% CI 1.76 to 11.51; P = 0.009) as independent predictors, and BSI > 0.49 was the better predictor, which achieves a positive likelihood ratio (LR) of 11.3 (95% CI 3.81 to 33.67) and a negative LR of 0.11 (95% CI 0.04 to 0.33).

Quantitative EEG allows the early prediction of mMCAI, and can help in the selection of patients for decompressive hemicraniectomy.

Financial Support: None

The modified Rankin Scale (mRS) is a 6-level outcome scale used to assess level of function in neurological disease. Its utility is underscored by widespread use in stroke outcomes assessment, but the basic levels of function encoded by the mRS are not specific to stroke. Still, poor interobserver reliability and the requirement for expert and face-to-face interviews are problems in determining an mRS score. We have developed a 9 question "yes/no" questionnaire, the mRS-9Q, and an online mRS calculator to quickly and accurately determine the mRS. We hypothesize that (1) the mRS-9Q has acceptable interobserver reliability, (2) the mRS-9Q can be administered equally well in person or over the telephone, (3) the mRS-9Q can be administered accurately by personnel without clinical expertise, and (4) the mRS-9Q allows application of the mRS to a broad range of Neurological conditions.

The mRS-9Q was administered by form or telephone. A web-based tool calculated the mRS and performed error checking. Part 1 compared the mRS-9Q to an mRS Structured Interview (n=80). Part 2 compared mRS-9Q administration by telephone and by paper form (n=80). Part 3 compared administration by an expert interviewer with administration by a non-expert (n=83). Part 4 examined reproducibility over 2 weeks (n=84).

Agreement was very good in all Study Parts. In Part 1 (mRS-9Q vs. mRS-SI), k was 0.80 and k w was 0.96. In Part 2 (Telephone vs. Paper), k was 0.83 and k w was 0.95. In Part 3 (Expert vs Non-Expert), k was 0.72 and k w was 0.93. In Part 4 (Reproducibility), k was 0.76 and k w was 0.93.

The mRS-9Q can reliably determine the mRS by paper survey or over the telephone. Importantly, the mRS-9Q survey does not require the participation of trained experts-excellent results are obtained when non-medical study personnel administer the survey.

Potentially inappropriate medications (PIMs) are medications that may increase cognitive burden and impact clinical outcomes in elderly ICU patients. This study evaluates the use of PIMs and outcomes in elderly stroke patients.

This is a retrospective study of p July 2011. Number of PIMs, length of stay (LOS), and changes in GCS and RASS scores were evaluated. Fisher's Exact test was used to compare groups.

Of 112 

a significantly longer NSICU LOS and worse outcomes.

Introduction AIS patients often have acutely elevated BP requiring IV antihypertensives (IVAH). Previous work shows AHA/ASA recommended antihypertensives used to reduce BP in AIS commonly results in polypharmacy and its consequences: overshoot hypotension and increased mortality. This study evaluates the association between IVAH polypharmacy and both clinical and economic outcomes in AIS.

Premier, a US hospital administrative database. Patients with MS-DRGs 61 to 66 and a primary AIS ICD-9 code (433.X1 or 434.x1) were included. Patients were matched in a 1:1 fashion utilizing propensity score methodology controlling ICU admission, baseline characteristics, and pre-existing conditions.

From January to December 2009, 9203 study patients received at least one IVAH on day one or two of hospitalization and 28.2% of those received more than one IVAH. After matching, 1886 patients remained in each group. Patients in GP1 had a lower mortality rate than GP2 (9.6% vs 12.9%, p=0.0012), lower vasopressor use (4.3% vs 5.8%, p=0.0313), shorter LOS (median 4 days vs 5 days, p<0.0001), and lower total hospital costs (median $8,138 vs $9,622, p<0.0001). t-PA use was similar between groups (14.1% vs 15.0%, p=0.46).

Polypharmacy to treat acute hypertension is associated with worse clinical and economic outcomes in AIS regardless of tPA administration. Recent evidence suggests precise and reliable BP control is critical during the entire stroke pathway of care. Currently recommended IVAH do not reliably manage BP as single agents. In order to avoid polypharmacy and improve outcomes and costs, the ideal IVAH drug needs to reliably manage and maintain precise BP control as monotherapy.

Financial Support: Authors are employees of The Medicines Company which markets an IV antihypertensive agent.

Mean corpuscular hemoglobin concentration (MCHC) is a red blood cell indicie that is obtained as part of a complete blood count (CBC). MCHC values reflect individual red blood cell (RBC) Hemoglobin (Hgb) content, and are directly affected by changes in Hgb production and DNA synthesis. Recently another hematologic indicie: the Red cell Distribution Width (RDW), has been shown to be an independent predictor of outcome in patients with stroke. We sought to determine if MCHC on admission could be predictive of clinical outcome.

This is a retrospective study on patients admitted to a university affiliated community hospital. Initial MCHC data were gathered retrospectively from the registry database. We included both ischemic and intraparenchymal hemorrhage (IPH) stroke patients in our analysis. For evaluating the severity and outcome of the patients with ischemic strokes we used NIHSS on admission and MRS on discharge respectively. In IPH patients, we utilized the University of California San Francisco intracerebral hemorrhage (ICH) score on admission / 24 hours to quantify the severity of the stroke and mRS on discharge to measure the outcome. We used correlation coefficients (Spearman's rho) and the Mann Whitney test for analysis of the data. SPSS version 11 was used for data processing.

Our review identified 74 patients with a diagnosis of IPH and 543 with a diagnosis of ischemic stroke. The MCHC values in the IPH group positively correlated to UCSF ICH score on admission (p=0.01, r=0.29) and at 24 hours(p=0.006, r=0.31), as well as to mRS at discharge (p=0.02, r-0.27). The MCHC levels for ischemic stroke patients correlated weakly and negatively to NIHSS on admission (p=0.001, r= -0.14) and D-mRS (p= <0.001, r= -0.17).

MCHC levels on admission correlate significantly with clinical measures of stroke severity and disability. MCHC could serve as an early predictor for outcome in different stroke subtypes.

In the absence of specific guidelines, there is considerable variance in pre-procedural intubation practices for endovascular treatment of acute ischemic stroke. The purpose of this study is to understand and characterize the variance in pre-procedural intubation practices and identify the reasons that influence the choice of pre-procedural intubation practices among treating physicians.

We selected 10 random cases from a prospective database of patients undergoing endovascular treatment for acute ischemic stroke and prepared a case summary providing pertinent demographic, clinical, and imaging data. Twenty clinicians independently reviewed the case summaries and responded to whether they would intubate any of the 10 patients and identified the reasons for their choices. Clinicians were also asked to identify their training background (Neurology, Neurosurgery or Radiology trained endovascular specialist, vascular neurologist or neuro-intensivist).

Reasons for intubation and agreement between clinicians for each case were ascertained.

The decision to intubate the patient was made in 63 of 200 total clinical scenarios. The major reasons identified by the physicians for pre-procedural intubation were high National Institute of Health (NIH) stroke scale scores on admission 26.9% (n=17), labored breathing or desaturation 23.8% (n=15), less than optimal respiratory status of patients combined with drowsiness or reduced level of consciousness 14.3% (n=9), inability to follow command such as due to aphasia 12.7% (n=8), seizures 1.6% (n=1), and no reason 20.6% (n=13). Overall agreement between clinicians regarding decision of pre-procedural intubation among the 10 case scenarios was 30. 

The decision of pre-procedural intubation varies widely among clinicians. Due to recent data that suggests that decision of pre-procedural intubation may impact on patients' outcomes, better standardization of such practices is required.

Hyperglycemia has been shown to be associated with worse outcomes, increased hemorrhage rates, and increased mortality in patients with acute ischemic stroke (AIS). We evaluated the effect of admission hyperglycemia on 90-day functional outcome, mortality, and hemorrhage rates in patients undergoing multimodal endovascular therapy (MET) for AIS.

Retrospective review of glucose on admission was performed in 638 patients undergoing MET between 1996 and 2012 in a tertiary care academic medical center. Demographic data, diabetic status, NIHSS score, radiologic studies, and recanalization TIMI grade were analyzed, amongst other known predictors of hemorrhage and poor outcome.

Mean age was 66.5+13.5 and mean NIHSS 16.9 + 6.7. Hyperglycemia was present in 310 ( 

Admission hyperglycemia in patients undergoing MET is associated with poor 90-day functional outcome and higher rates of in-hospital death and HI. In non-diabetic patients, hyperglycemia was only associated with increased mortality and HI. Despite equivocal results for induced normoglycemia, this data justifies a prospective trial for moderate glycemic control in this patient population. 

Previous studies suggest that low cholesterol levels are associated with higher rates of hemorrhage after acute ischemic stroke (AIS). We studied the effect of serum lipoproteins and premorbid statin use on the rate of hemorrhage in AIS patients treated with multimodal endovascular therapy (MET).

Retrospective review of statin use and lipoprotein levels on admission including LDL, HDL and total cholesterol (TC) was perfomed in 601 patients undergoing MET between 1996 and 2012 in a tertiary care academic medical center. Demographic data, NIHSS score, radiologic studies, and recanalization TIMI grade were analyzed, amongst other known

LDL <100 mg/dL was associated with a higher incidence of HT (OR 1.67, 95% CI 1.21-2.07, P= 0.025). HDL >60 was associated with higher rates of PH (OR 1.32, 95% CI 1.07-1.76, P=0.045). TC levels and premorbid statin use were not associated with higher rates of hemorrhage. Statin use, LDL, HDL and TC were not independently associated with functional outcome at 3 months. Patients with hemorrhage and TC <200 had significantly higher rates of good functional outcome compared to those with TC >200 (OR 1.33, 95% CI 1.05-1.55, P=0.035). There was no significant association between statin use and rates of hemorrhage or functional outcome in patients presenting with LDL <100.

Low LDL and high HDL levels are associated with increased rates of hemorrhage after MET for AIS. Statin use had no effect on post-intervention hemorrhage or functional outcome regardless of admission lipid levels. Despite the association between low LDL and hemorrhage, statin use in patients with a low LDL was not associated with poor outcomes. This data justifies further study of the effect of continuation and early initiation of statin therapy in this patient population.

Mexican Americans (MAs) have shown lower post-stroke mortality compared to non-Hispanic whites (NHWs). Limited evidence suggests race/ethnic differences exist in intensive care unit (ICU) admissions following stroke. Our objective was to investigate the association of ethnicity with admission to the ICU following stroke.

Cases of intracerebral hemorrhage and acute ischemic stroke were prospectively ascertained as part of the Brain Attack Surveillance in Corpus Christi (BASIC) project for the period January, 2000 through December, 2009. Logistic regression models fitted within the generalized additive model framework were used to test associations between ethnicity and ICU admission and potential confounders. An interaction term between age and ethnicity was investigated in the final model.

A total 1,464 cases were included in analysis. MAs were younger, more likely to have diabetes, and less likely to have atrial fibrillation, health insurance, or high school diploma than NHWs. On unadjusted analysis, there was a trend toward MAs being more likely to be admitted to ICU than NHWs (34.6% versus 30.3%; OR=1.22; 95% CI 0.98-1.52; p=0.08). However, on adjusted analysis, no overall association between MA ethnicity and ICU admission (OR=1.13; 95% CI 0.85-1.50) was found. When an interaction term for age and ethnicity was added to this model, there was only borderline evidence for effect modification by age of the ethnicity/ICU relationship (p=0.16).

No overall association between ethnicity and ICU admission was observed in this community. ICU utilization alone does not likely explain ethnic differences in survival following stroke between MAs and NHWs. The Medicines Company, Parsippany, NJ, USA

The relationship between blood pressure variability and inpatient outcomes and costs following AIS is not well understood.

Using data from >100 US hospitals (Cerner Health Facts®), we identified all admissions between 1/1/2009 and 12/31/2010 of -9-CM diagnosis codes 433.X1, 434.X1). In patients with principal diagnoses of AIS, time of initial clinical presentation was designated "index time"; for those with secondary diagnoses of AIS, index time was in-hospital onset of stroke symptoms. We calculated blood pressure variability (BPV) as maximum difference (MD) (i.e., highest -lowest recorded BP) in the 3-hour period following index time. Patients were igh"] vs <median value ["low"]). We examined rates of hemorrhagic transformation, in-hospital mortality, and total in-hospital charges in the two groups.

A total of 322 admissions satisfied all selection criteria (high BPV: 166; low BPV: 156); patients were comparable with respect to age (mean age, 71 years), gender, and comorbidities. Patients with high BPV were nominally more likely to experience hemorrhagic transformation (1.8% vs 1.3%; p=0.70) and to die in hospital (4.2% vs 3.2%; p=0.63), respectively. They also had significantly higher total in-hospital charges ($32,536 vs $23,856; p<0.01).

Highly variable blood pressure during the 3 hours following AIS is associated with increased total in-hospital charges. Further research is needed to better understand the degree to which better early hypertensive management may improve patient outcomes.

Financial Support: Our research was funded by The Medicines Company. Mr. Berger, Mr. Moynahan, and Dr. Oster are employed by Policy Analysis Inc., an independent contract-research organization with previous and ongoing

Previous work from our group revealed an association between lower end-tidal carbon dioxide (CO 2 ) values with worse outcomes in intubated, mechanically ventilated patients undergoing intra-arterial procedures for acute ischemic stroke. Based on that observation, we hypothesized there might also be an association between lower arterial CO 2 and outcomes in this patient group.

We retrospectively reviewed the charts of participants in a national, multi-center mechanical thrombectomy trial. All patients were intubated and mechanically ventilated at the time of their intra-arterial intervention per protocol at our institution. Arterial blood gasses (ABG) were typically obtained post-procedurally at admission to the ICU as part of admission laboratory panel. Demographics, hospital course data, arterial blood gasses and outcomes (90-day Modified Rankin Score) were collected. Descriptive statistics (median and intraquartile range for continuous variables; counts and percentages for categorical variables) were used to describe clinical data. Categorical variables were compared using Fisher's exact tests, and ordinal comparisons were made with the Wilcoxon-Mann-Whitneytest. Significance was set at : mRS0-4; vs. unfavorable: mRS 5-6.

Seventy-five patients were identified as part of the registry, though 14 were excluded from this analysis (13 had no recorded ABGs, 1 lost to follow up). Patients with an unfavorable outcome were older (73 [59-81] v. 58 [53-70] , P=0.01). In the unfavorable group, median pCO2 levels were statistically significantly lower than those with a good outcome (38 [35-41] v. 42 [39-43] , P=0.01).

Our results suggest that there may be an association between lower pCO2 values and outcome in intubated, mechanically ventilated ischemic stroke patients. One possible explanation is that hypocapnea facilitates cerebral vasoconstriction, worsening ischemia and possibly enlarging the ischemic penumbra. This association should be further interrogated in a prospective, controlled trial.

Although ischemic brain tissue should appear particularly dependent on the main oxygen carrier hemoglobin (Hb), optimum Hb and hematocrit (Hct) levels in neurointensive care unit (NICU) patients with severe ischemic stroke (IS) are still unknown. Research thereof and of the best Red Blood Cell (RBC) transfusion regimen is neglected. Here, we aimed to investigate associations of Hb/Hct dynamics and transfusion activity with NICU variables and outcome in NICU patients with ischemic stroke.

Patients with IS were retrospectively assessed if admitted to the NICU for at least 3 days and mechanically ventilated. Assessment involved clinical characteristics, RBC transfusion events, and laboratory parameters such as Hb, Hct, Creactive protein, blood count, creatinine. Mortality and functional outcome (by mRS) were assessed prospectively by telephone interview at 3, 6 and 12 months after discharge. Statistics were descriptive and explorative, i.e. involved correlation / variance tests for associations between Hb, Hct levels, transfusion events and outcome (linear regression analysis, t-test).

109 patients were analyzed. 62% had anterior circulation, 38% posterior circulation ischemic stroke with a median NIHSS of 17 and a median APACHEII of 22. Median NICU stay was 14 days, median ventilation time 244 hours. Baseline Hb and Hct were 13 g/dl and 0.38, mean decrease over NICU stay was 30% leading to minimum Hb and Hct of 9 g/dl and 0.26, respectively. mRS at day 90 was 3 (6%), 4 (14%), 5 (46%) and 6 (=dead, 34%). A first preliminary analysis did not reveal any consistent associations between severity of anemia or transfusion activity and outcome.

Anemia is frequent in NICU patients with ischemic stroke. At present, this retrospective analysis does not support a more aggressive RBC transfusion practice. More detailed subgroup analysis results will be presented at the time of the meeting.

Although not clinically validated, physicians frequently use imaging to help predict neurologic and cognitive outcome after hypoxic-ischemic injury to the brain following cardiac arrest. We systematically reviewed the current literature investigating the accuracy of imaging in predicting outcome in adult patients who are comatose after cardiac arrest.

The electronic databases of Medline via PubMed, EMBASE via OVID, and the Cochrane Database of Systematic Reviews were searched using combinations of the search terms "cardiac arrest," "cardiopulmonary resuscitation," "brain hypoxia," and "neuroimaging." Eligible studies were reviewed in detail and classified by level of evidence (LOE) and methodological quality as defined by the International Liaison Committee on Resuscitation (ILCOR).

802 articles were identified, 72 of which met inclusion criteria and were reviewed in detail. 65 articles found imaging to be helpful in accurately predicting outcome: 15 were categorized as LOE P3 (10 good, 3 fair, 2 poor quality), 30 LOE P4 (6 good, 12 fair, 12 poor quality), and 20 LOE P5 (5 good, 3 fair, 12 poor quality). 8 articles found imaging to not be helpful in accurately predicting outcome: 2 were categorized as LOE P2 (1 good, 1 fair quality), 3 LOE P3 (all poor quality), 3 LOE P4 (1 good, 1 fair, 1 poor quality). 1 article was found to be equivocal, categorized as LOE P4 (fair quality).

No LOE P1 trials have been performed and although two LOE P2 studies exist, they do not compare different modalities and have limited generalizability in the clinical setting. There is insufficient evidence to recommend routine use of imaging to predict outcome in comatose cardiac arrest patients. Further studies are needed to determine the utility of imaging in this setting, and should include comparison with already validated methods of prognostication such as the clinical exam and electrophysiology.

In acute ischemic stroke, factors affecting blood pressure (BP) levels in the first few hours after onset are not well known. The rise in BP levels can be a mechanism to compensate for vessel occlusion and to ensure the viability of penumbral brain tissue (PBT).

Between January 2006 and December 2009, data of 206 consecutive acute ischemic stroke patients admitted to Florence Nightingale Stroke Unit within the first 12 hours and evaluated with stroke MRI were analyzed. Potential early correlates of BP levels were analyzed with multivariate logistic regression technique. Age, gender, hypertension (HT) history, treatment for HT, onset to door time, NIHSS, level of consciousness, stroke type and stroke etiology, PWI/DWI mismatch, index proximal vessel occlusion (iPVO), were included in the analyses. A PWI/DWI mismatch, a potential indicator of PBT, was considered as present if estimated at least to be approximately 20% by eyeballing the lesion. Index PVO was defined as symptomatic occlusion or 70-99% stenosis of carotid or vertebrobasilar artery system vessels. High BP level was defined as SYS>140mmHg or DYS>90 mmHg at admission.

Two hundred-six patients (119 male; mean age 69; range 20-96 years) were evaluated. Hundred and four patients (50, 5%) had high BP at admission. In univariate logistic regression analysis, women (p: 0,023), age (p: 0,083), TACS (p: 0,035), hypertension history (p: 0,054), iPVO (p: 0,081) were associated with high BP values. Only TACS (OR: 2, 667; 95% CI: 1,290-5,513) was independently associated with high BP readings at admissions in multivariate analysis.

We did not find any argument to state that high admission blood pressure is a compensatory response following brain tissue ischemia.

Intravenous recombinant tissue plasminogen activator (IV r-tPA) has revolutionized the management of acute ischemic stroke. However, symptomatic intracranial (6%) and severe systemic (1.6%) hemorrhagic complications after thrombolysis remain a concern.

We present a rare complication of r-tPA and underscore the importance of close monitoring after thrombolysis. The clinical history, laboratory, and imaging studies were reviewed.

A 46 year old man with psoriasis and morbid obesity presented with acute aphasia and right hemiplegia. He had fallen as a result, striking his right eye. His examination demonstrated right periorbital ecchymosis without ptosis, expressive aphasia, leftward gaze deviation and corresponding hemianopsia, and right facial weakness and hemiplegia. His summated NIH Stroke Scale (NIHSS) was 13. Initial cranial imaging demonstrated no blood, though did show an abnormal hyperdensity within the proximal left middle cerebral artery territory. He received IV-tPA 2 hours from symptom onset with significant neurological improvement within 20 minutes of thrombolysis. 30 minutes after initiation of IV r-tPA, he rapidly developed periorbital edema with ecchymosis leading to complete ptosis of the right eye. Repeat cranial imaging showed an enlarging retro-orbital hematoma. An emergent lateral canthotomy was performed of the right eye to rapidly decompress the optic nerve. Within 3 days of thrombolysis and successful orbital decompression, as visualized on repeat cranial imaging, he made near full neurologic and visual recovery.

To our knowledge, this is the first reported case of a near catastrophic hemorrhagic ocular complication after IV r-tPA therapy for acute ischemic stroke. Despite the suspected trivial nature of injury, thrombolytic treatment should proceed with caution in the setting of any trauma. This report highlights the importance of careful inspection and maintaining a high index of suspicion and vigilance for unanticipated complications after thrombolytic therapy.

In the setting of acute or evolving stroke, outcome may be dependant on the urgent re-establishment of cerebral perfusion.Options for restoring cerebral blood flow include the intra-venous or intra-arterial administration of thrombolytic agents, mechanical thrombolysis, and urgent carotid endarterectomy. There is very limited experience with emergency extracrannial-intracranial (EC-IC) bypass in this setting.

We reviewed the medical records and neuroimaging studies of 20 consecutive patients who underwent urgent EC-IC bypass in the face of acute cerebral ischemia. None were considered appropriate candidates for endovascular therapy. Ages ranged from 21 to 74 years, average 48.5 years. Average follow-up was 6.2 years. Preoperative angiographic evaluation identified critical narrowing of the supraclinoid ICA in 8, the M1 segment of the middle cerebral artery in 10, and the cervical/petrous ICA in 2. All had progressive, refractory symptoms associated with enlarging areas of ischemic changes on diffusion-weighted MRI despite maximal medical therapy including anticoagulation and antiplatelet agents, blood pressure elevation, and fluid resuscitation.

All patients underwent urgent STA-MCA anastomosis. In every case, bypass resulted in stabilization of the progressive ischemic symptoms; in 3 cases, revascularization was followed by rapid, dramatic improvement of preoperative deficit. Five patients awoke with transient worsening of their preoperative neurological deficit which improved over 24-48 hours. No patient demonstrated a significant new area of ischemia on MR imaging.

Emergency EC -IC bypass for acute ischemic injury was both safe and effective in our experience. This population was characterized by relatively young patients with severely limited collateral circulation. In this series of 20 carefully selected patients, bypass was successful in arresting ongoing ischemic symptoms, and in some cases, resulted in rapid neurological improvement.

The ability for clinicians to predict outcome is of paramount importance when treating and counseling stroke patients and families. The DRAGON score is used to predict outcome in patients with anterior circulation strokes that have received intravenous tPA. We sought to determine if the DRAGON score could be applied to patients undergoing endovascular stroke therapy.

Charts for patients with interventions performed by a single operator (MFS) from January 2011 to March 2012 were reviewed. Presenting symptoms were used to derive the DRAGON score. Outcome predictability was compared to the findings in the original DRAGON score paper.

Twenty-four patients underwent endovascular stroke treatment; Fourteen patients presented with anterior circulation ischemic strokes. Five patients had only endovascular treatment, and 9 patients had both IVtPA and endovascular treatment. The total average time from onset to termination of the endovascular procedure was 367 minutes. In the 5 endovascular alone patients, 3 patients survived with a mean DRAGON score of 6.00 and mean discharge mRS of 3.00. Of the 9 patients who received both intravenous and endovascular therapy, 8 survived with a mean DRAGON score of 4.25 and mean discharge mRS of 2.25. Four of the 11 surviving patients had greater than 90% specificity for poor outcome (mRS 5-6) based on the original paper. These patients however demonstrated a good recovery with an average mRS of 2.25.

Despite the extended window for treatment and recanalization, patients who receive acute endovascular stroke therapy appear to have similar outcomes to the predicted outcome using the DRAGON score. Furthermore, our study showed that patients who were expected to have poor outcome had the potential to improve clinically. This study reinforces the benefit of endovascular stroke therapy.

Intracranial arterial stenosis are relatively common findings of stroke patients in Asia area. We reviewed stroke database to investigate clinical risk factors related to intracranial arterial stenosis, including carotid disease, and peripheral arterial disease which reflects advanced atherosclerosis.

Acute stroke patients at the National Health Insurance Corporation Ilsan Hospital from January 2008 to December 2010 with available transcranial Doppler(TCD) examination, carotid ultrasound and ankle-brachial indexes(ABI) formed the analysis cohorts. Retrospective review was performed.

A total of 642 patients were included during that period, 212 patients with incomplete TCD study due to poor insonation windows were excluded(33%). According to TCD criteria, 3 groups of intracranial arterial stenosis are defined: 0 vessel stenosis is in 220 patients(51%), 1-2 vessels in 125 patients(29%), more than 3 vessels in 85 patients(20%). As the arterial number of intracranial stenosis increased, ABI is decreased(P=0.013) and the size of carotid artery plaque is increased(P=0.011). Among the risk factors, Diabetes, age, past stroke history are increased(P=0.0000, P=0.006, P=0.05) and HDL cholesterol showed tendency of decrease(P=0.033). However hypertension, smoking, total cholesterol, LDL cholesterol, triglyceride and sex are not correlated with intracranial arterial stenosis.

Among the acute stroke patients, about a half of them have intracranial arterial stenosis and these patients tend to have higher burden of advanced atherosclerosis as evidenced by a higher prevalence of Diabetes, large sized plaques of carotid artery and peripheral arterial occlusive disease.

Dedicated neurocritical care service in an acute-ICU setting with specialized neuro nurses and physicians improves the quality of care and patient outcomes. We aimed to find out the impact of specialized focused neurocritical service as compared to a general surgical/medical ICU setting in a community hospital.

We retrospectively reviewed data from 2005-2011, on 165 patients who received endovascular treatment (IAtPA, thrombolysis, mechanical thrombectomy, with or without intra and extra cranial stenting) in order to achieve recanalization. Patients were divided into two groups: Group A (n=58) General Med/Surg ICU care in 2005-2008 and, Group B (n=107) Focused Neurocritical care 2009-2011. Functional outcome data (mRS 90 days) between the 2 groups was compared through patient records. Group A patients were cared for with general surgical/medical ICU care nurses while Group B patients were cared for by a specialized core group of 12-16 nurses specifically trained in neurocritical care.

Both groups were comparable in terms of age, sex, admission NIHSS and co-morbidities (hypertension, hyperlipidemia, diabetes, CCF, A.Fib, other). Group A mRS 0-2 (n=22) 38%, Group B mRS 0-2 (n=45) 42%. Group A mRS 3-4 (n=14) 23%, Group B mRS 3-4 (n=18) 17%. Group A mRS 5 (n=12) 20%, Group B mRS 5 (n=22) 21%. Mortality for both groups was comparable at 20% (Group A n=12; Group B n=21).

Functional outcomes of fully independent patients (mRS 0-2) improved from 38% to 42% when a focused neurocritical care nursing service was implemented as compared to standard medical ICU nursing care. Strict adherence to neurocritical care protocols and proper attention to co-morbidities is the key to improved outcomes in critically sick acute stroke patient populations.

In-hospital strokes remain a significant source of morbidity for patients. Paradoxical embolism has been implicated as a potential source for these strokes. To date, there is only minimal literature regarding paradoxical embolus as a cause for stroke in the hospitalized patient. Over a one-year period we studied in-patient stroke alerts and their etiologies at our institution. The hypothesis of this study is that strokes in hospitalized patients are caused by paradoxical emboli.

This is a retrospective analysis of prospectively collected in-hospital stroke team calls (n=93) over a one-year period. We excluded patients on the stroke service or on neurologic floors including the neurological intensive care unit. We further excluded patients that were found to have stroke-mimics by consensus. From these patients, we collected demographic information and results of transthoracic echocardiograms (TTE) and lower extremity (LE) duplex. The categorical data was analyzed using chi-square on JMP 9.0.

A confirmed acute ischemic stroke was found in 57 (61%) of the in-hospital stroke alerts. The majority of stroke alerts in our institution were from the cardiology and cardiothoracic services (71.9%). A TTE and LE duplex were available in 98.2% and 43.9%, respectively. Two patients were identified with a patent foramen ovale (PFO) and nine with a deep venous thrombosis (DVT). One patient was found to have both a DVT and PFO which was presumed as the source of embolus. Overall, there was no significant association of in-hospital stroke and presumed paradoxical embolus.

The present study shows no association of in-hospital strokes and paradoxical emboli. This study is limited by the infrequent ordering of LE duplexes in this at risk population but is strengthened by the available TTE results.

Posterior circulation stroke (PCS) is associated with high mortality and poor outcome. This single centre, retrospective analysis evaluates long-term mortality and functional outcome in PCS patients treated with/without revascularization therapy (RT).

Between January 2006 and December 2009, dataof82 consecutive PCS patients admitted to Florence Nightingale Stroke Unit within the first 12 hours were analyzed. After evaluation with MRI, eligible patients with PWI/DWI mismatch selected with eye-balling technique were treated with RT. Ninety days modified Rankin Score (mRS) and mortality were the main outcome.

Eighty-two patients (48 male; mean age 68; range 27-92 years) were evaluated. The mean onset to door time was 258 minutes (SD:243). Seventy-eight patients were examined with MRI while 4 patients were examined with CT. Twenty-one patients received RT; intravenous thrombolysis in 11, endovascular multimodal revascularization in 6 and bridging therapy in 4 patients. Mean NIHSS score was 8 (range:0-38) [treated group (TG):13(2-38); untreated group (UTG):7(0-32) p:0,009]. Arterial occlusion was present in 43(7%)(TG:61,1%;UTG:37,7%; p:0,104). Mean door to treatment time was 152 minutes (SD: 282).Mean onset to treatment time was409minutes (SD:281).Mean discharge NIHSS score was 6 (range:0-32) [TG: 9(0-30); UTG:5(0-32) p:0,009]. Discharge mRS 0-1 ratio was 47,6% (TG:23,8%; UTG:55,7%; p:0,013). The inhospital mortality rate was 8,5% (TG:14,3%; UTG:6,6%; p:0,365). First month (TG:38,6%; UTG:57,1%) and 3 rd month (TG:46,7%; UTG:50%). mRS 0-1 ratio (p:0,174 vs. 1, respectively) also 1 th month (TG:21,1%; UTG:11,9%) and 3 rd month (TG:26,7%; UTG:17,9%) mortality (p:0,286 vs. 0,696, respectively) were similar between groups.

In posterior circulation stroke, despite severe clinical manifestations at admission and hospital discharge, after long term follow up, the outcome in patients treated with revascularization therapy is similar to those patients with benign outcome and not necessitating any revascularization therapy from the outset.

We present a case series that highlights the feasibility of decompressive hemicraniectomy (DHC) in pediatric patients with ischemic stroke.

A retrospective chart review identified 3 cases of ischemic stroke at Texas Children's Hospital between 2008-2012 where DHC was performed for high intracranial pressure (ICP) after standard medical therapy failed to lower ICP. Information was obtained about patient characteristics on admission, radiological features of the stroke, surgical procedures, complications of the DHC and cranioplasty, and clinical outcomes. We also surveyed published literature on DHC for pediatric patients with ischemic stroke.

There was no mortality in this case series. Case 1 had a modified Rankin Score (mRS) of 3 at a follow up visit after 36 months. Case 2 had mRS of 1 at a follow up visit after 36 months. Cranioplasty was complicated by epidural abscess in his case. Case 3 had mRS of 3 at a 24 months follow up. Review of literature identified 7 other published case series consisting of 14 cases of DHC in pediatric patients with ischemic stroke. Detailed analysis of these cases is presented in the tabular form.

This case series highlights the fact that DHC can be performed safely and effectively in pediatric patients with ischemic stroke with potential lifesaving and improved functional outcome. Decompressive hemicraniectomy should be considered as a therapeutic option for refractory elevated ICP following large hemispheric strokes in the pediatric population.

Basilar artery occlusion (BAO) is a devastating neurological disease that can be difficult to diagnose due to its protean manifestations, and the initial CT will often not reveal an acute infarction. We present a patient with BAO who was initially diagnosed with Lyme disease.

A 67 y/o female presented with neck pain, an unsteady gait, partial facial paralysis, and mild dysarthria. She was noted to have an erythematous area on her neck that contained a tick. The initial head CT was negative. Lyme disease was diagnosed and ceftriaxone and doxycycline were initiated. Within 24 hours, her symptoms progressed to hemiparesis and aphasia. A stat MRI demonstrated the absence of flow in the basilar and left vertebral arteries with restricted diffusion in the pontine and mid-right parietal regions. The patient was transferred to a primary stroke center, but she was outside the window for stroke rescue.

Acute Lyme disease is characterized by lymphocytic meningitis, cranial neuropathy (particularly facial palsy) and radiculoneuritis. Though these symptoms usually take weeks to occur, the initial tick bite may not be recognized thus precluding an accurate evaluation of the time course. BAO may present with a similar constellation of symptoms including headache, facial paralysis, and transient paresis called the "herald hemiparesis" of BAO. The fluctuating course of early BAO may be confusing and a high index of suspicion is required. Intra-arterial lytic therapy, mechanical thrombolysis, or a combination is recommended up to 12 hours of symptom onset. Recanalization is paramount to preserving neurologic function. Unfortunately, she arrived at our institution outside the window for invasive therapy. Her symptoms continued to progress to a locked-in state and she was transferred to a LTAC facility.

The neurological manifestations of BAO may be confused with other diagnosis and a high index of suspicion is required.

Metabolic abnormalities negatively influence outcome in patients with traumatic brain injury, subarachnoid hemorrhage, hemorrhagic stroke and ischemic stroke with or without thrombolytic therapy. The prognostic value of many potentially correctable physiologic markers in stroke patients receiving thrombolysis is unknown.

Twenty-one consecutive acute ischemic stroke patients treated with tissue plasminogen activator (tPA) were retrospectively studied. Multiple metabolic and physiological variables including blood urea nitrogen, creatinine, sodium, potassium, chloride, calcium, phosphorous, magnesium and body temperature were analyzed. Independent T Test was used to compare mean scores of these variables and determine their effect on outcome. Functional status at discharge was the primary outcome measure, being fully or partially independent determined as good outcome and fully dependent or dead as poor outcome. Secondary outcome was the presence of hemorrhagic conversion.

Seventeen patients had good outcome, mean age 66, while 4 patients had poor outcome, mean age 81. Hyperthermia and admission Acute Physiology and Chronic Health Evaluation (APACHE) II score were associated with poor outcome (p<0.05). Hemorrhagic conversion occurred in 4 patients and was associated with hyperthermia, higher Simplified Acute Physiology Score (SAPS) II score and hyponatremia (p<0.05 for all).

This single-center, retrospective study suggests that mild hyperthermia, hyponatremia and higher APACHE II and SAPS II scores are associated with poor functional outcome and hemorrhagic conversion in patients with acute ischemic stroke treated with tPA. Further study is required to determine if correcting these variables influences outcome.

Alterations in electrolyte balance and other basic physiologic indicies such as glucose have been implicated in the pathophysiology of coronary heart disease. However, the relationship between the electrolyte levels and other physiologic indicies measured immediately after an acute ischemic stroke has not been clearly delineated. Objective: The aim of the present study was to test whether changes in a patient's basic metabolic panel modify the severity or outcomes of acute ischemic stroke.

The study is a retrospective study on ischemic stroke patients admitted to a university affiliated community hospital. Demographic data were collected from the data registry. Values were obtained within one hour of presentation for serum sodium (Na), potassium (K), glucose (Gluc), chloride (Cl), Magnesium (Mg), bicarbonate (HCO3), BUN, and creatinine (Cr). As well glomerular filtration rate (GFR) and temperature values were also recorded. Severity and outcome were measured using the NIHSS on admission and the mRS on discharge respectively. Correlation coefficients (Spearsman's rho) and the Mann Whitney test were employed in the analysis of the data. SPSS version 11 was utilized for data processing.

Results 543 consecutive acute ischemic stroke patients met the study criteria. Serum Ca (p=0.05, r= -0.08) and Gluc levels (p=0.02, r=0.09) were significantly correlated with the mRS. Serum Cl, Ca, Bicarbonate, temperature and BUN were significantly correlated with NIHSS on admission measurements (p=0.03, 0.002, 0.003,0.01; r=0.09, -0.13, ;0.13; 0.10 respectively). Mg showed a negative trend of correlation with the NIHSS on admission as well (p=0.06; r=-0.08), suggesting a protective effect of higher Mg levels.

The study shows that initial metabolic parameters, such as serum Mg, Ca, HCO3, BUN, and temperature may potentially allow for early prediction of the severity and outcome in patients with ischemic stroke.

Hypersomnolence is not typically appreciated as a focal neurologic finding, though bilateral thalamic infarcts may present with hypersomnolence as the only neurologic manifestation.

A 52 year old man presented with acute onset confusion, somnolence and slurred speech. His neurological examination was notable for somnolence, bilateral ptosis and dysarthria. Routine laboratory investigations and CSF analysis were unremarkable, aside from a urine toxicology screen which was positive for opiates. Initial head computed tomography (CT) with CT angiography of the head and neck were unrevealing. Magnetic resonance imaging was contraindicated as the patient had an automatic internal cardiac defibrillator (AICD). A working diagnosis of opiate intoxication was made in light of the urine toxicology results. Because the patient failed to improve over the ensuing 24 hours, a repeat head CT was obtained which revealed bilateral medial thalamic infarctions.

While hypersomnolence is often associated with toxic-metabolic disorders, it may rarely be the result of acute arterial stroke. In the context of stroke, hypersomnolence can be accompanied by other symptoms including weakness, paresthesias, memory impairment, sectoranopsia, and personality changes. The feature of hypersomnolence is usually the result of an infarct of perforators arising from the posterior cerebral artery, specifically the paramedian,and tuberothalamic arterial branches, which are involved in irrigation of the reticular, and intralaminar nuclei of the thalamus that are involved in arousal. Concomitant neurologic signs may not be present or may be difficult to elicit in this setting as patients are often unable to participate in the neurologic exam. Acute stroke should therefore be considered in the differential diagnosis of hypersomnolence. Failure to consider stroke as a potential eitiology may lead to delay in acute or secondary stroke prevention.

Metals play key roles in epigenetic events in living organisms. Zinc, cadmium, lead, selenium, calcium, magnesium, sodium, and potassium have been found to be associated with stroke risk in NHANES and other studies. The central hypothesis of this pilot study is that metals and metalloproteins may determine and distinguish stroke phenotype (ischemic vs. hemorrhagic).

Stroke patients at the University Hospital emergency department (ED) were enrolled in a plasma banking project. After IRB approval and informed consent, blood draws were performed in the ED, and demographic and clinical information recorded. We analyzed 29 plasma samples collected within 12hours of symptom onset. We used the proteomic techniques of affinity chromatography (to remove the abundant proteins albumin and IgG), followed by size exclusion chromatography (SEC -to eliminate low molecular weight compounds and fractionate the proteins), inductively coupled plasma mass spectrometry (ICPMS -to identify differentially expressed metalloproteins in plasma) and electrospray mass spectrometry (to identify the tryptic peptides known to represent specific proteins in the plasma). The areas and standard deviations of the chromatograms for the metalloproteins for stroke mimics (n=10), ischemic (n=10) and hemorrhagic (n=9) stroke patients were calculated using origin software. Differences between SEC-ICPMS peak areas of the metalloproteins for the ischemic, hemorrhagic and mimic samples were examined using two-sample t-Test and box chart statistics.

Mg, Al, Mn, Cu, Zn, Se, Mo and Pb were studied. Significantly different metals were Mg, Al, Mn, Cu and Se. Box chart statistics performed for the SEC-ICPMS metalloprotein peak area data revealed significant differences in all metalloproteins except Al. Tryptic peptide mapping identified significant differences in metalloproteins.

SEC-ICPMS detected differences in fractions of specific metal containing proteins in the plasma of stroke patients and patients who presented with a stroke mimic. Ongoing efforts are aimed at identifying potential biologically relevant stroke biomarkers from the current list of differentially expressed proteins. 

Retrospective chart review of patients admitted to the Neurocritical care unit from August 2010 to August 2011 who developed ICP crises (>20 mm Hg for >5 minutes) and were treated with 23.4% HTS. Only data for the first ever treatment with HTS were collected. Patient demographics, onset and duration of action, lowest ICP achieved and use of adjunctive therapies were recorded. Descriptive statistics and correlation analysis were performed.

Complete data were available for 19 patients. Ten subjects (52%) were female, the mean age was 52 + 18 years. Glasgow coma scale (GCS) was 6 + 3 and 13 (68%) patients concomitantly received therapeutic hypothermia and pentobarbital coma. A 67+ 29.4% reduction in ICP following administration of 23.4% HTS was observed (absolute change: 19 + 9.4 mmHg). The mean time to ICP <20 mmHg was 41 + 25 minutes and time to rebound ICP >20 mmHg post-HTS administration was 120 minutes in 75% of our cohort. Following treatment the mean improvement in GCS was 2 + 2. A dose-response curve was generated.

23.4% HTS was associated with a 67% reduction in ICP values in critically ill neurology/neurosurgery patients. Time to clinical endpoint of ICP <20 mmHg was 41 minutes and in 75% of patients the duration of action was 120 minutes. An improvement of 2 points in GCS was also observed. The first description of a dose-response curve for 23.4% HTS in humans is reported. 

Over a ten year period, we accumulated a prospective dataset of 500 severely brain injured patients with multimodality monitoring (Brain Tissue Oxygen Monitoring and Outcomes Project). Patients' data existed within individual excel files with heterogeneous fields. As different research subprojects arose, additional excel files were created to support new data extraction from clinical records. Several issues were apparent: (1) merging and querying data was time-consuming and rate-limiting in research productivity;

(2) users were unable to make uniform changes to all files; (3) different users could not simultaneously enter data, (4) auditing data entry was difficult. Our goal was to convert the dataset into a relational database, to enhance clinical research efficiency.

Microsoft Access was used to build a database with a relational backend structure and a graphical user interface (GUI) frontend. A reporting tool was built for analysis, preview, printing, and customized queries. Extract-Transfer-Load functions were programmed to create seamless data integration between the Access database and the enterprise-wide clinical data warehouse (e.g. laboratory values, radiology results).

It took approximately 640 man-hours to audit existing excel data, and to load distilled data from 500 excel files into structured database tables. It took approximately 480 man-hours for application implementation and testing. The GUI supported multiple simultaneous users' during data auditing, enforced validation rules that corrected data entry in realtime, and centralized user account management. We have provided 20 research queries to date.

Excel has limitations as a tool for clinical research informatics. A relational database that is built with pre-defined rules, fields, and tables dispenses with the time-consuming step of merging and cleaning data and makes large dataset queries and analyses more efficient. It allows straightforward integration with other relational databases such as enterprise-wide clinical data warehouses, enabling expansion of queries into other clinical information systems.

Financial Support: Elsa Lin is a data analyst whose salary was partially supported within the past twelve months by a grant from Integra (Brain Tissue Oxygen Monitoring) for the specific purpose of creating a relational

Objective of this case study is to report a case of central nervous system (CNS) Histoplasmosis presenting as an ischemic pontine vasculitis and chronic basilar meningitis. Histoplasmosis, a disease caused by fungus Histoplasma Capsulatum, primarily affects immune-suppressed patients and commonly involves the lung but occasionally can have variable CNS presentations.

A thirty-five year old Caucasian immune-competent male came with worsening of aphasia and confusion after having presented four weeks prior with dysarthria, gait ataxia and bilateral upper extremity weakness. He was diagnosed with bilateral pontine ischemic strokes secondary to small vessel vasculitis and but had limited response to high dose steroids.

Cerebral spinal fluid (CSF) examination showed elevated protein, low glucose and elevated cells suggestive of meningitis and he was started on empiric antibiotics and trials of repeat intravenous (IV) steroids. Follow-up imaging revealed obstructive hydrocephalous and he underwent successful ventriculo-peritoneal (VP) shunt placement. His CSF culture came back positive for H. Capsulatum. CSF Histoplasma antigen and urine antigen were also positive. He was initially treated with ambisome but changed to voriconzaole secondary to renal insufficiency and was eventually continued on itraconazole. At one year, the patient good clinical improvement and follow-up cultures were negative.

While pulmonary involvement of Histoplasmosis in immune-suppressed patients is common, systemic presentation of this fungal infection in immune-competent patients is exceeding rare. Clinicians should consider CNS Histoplasmosis on the differential diagnosis in atypical stroke cases, particularly those with chronic basilar meningitis.

There is increasing incidence of dengue fever in our country and encephalopathy is the most common neurological manifestation of severe infection. However, recent studies have shown that there is increasing evidence for dengue viral neurotropism. Dengue encephalitis, a distinct clinical entity have been found to be associated with the neurovirulence involving serotypes DEN-2 and DEN-3. The objective of this study is to report the clinical course, laboratory, and radiographic findings of Dengue Encephalitis that did not go through the usual state of dengue fever. Management of this specific viral infection will likewise be discussed.

Case presentation and report with literature review.

A20-year old Filipino male, methamphetamine and marijuana user was admitted to our hospital because of seizures preceded by headache and fever. He was managed as a case of viral meningitis supported by Cranial MRI findings and CSF studies. After nearly 5 days, he clinically deteriorated initially from a very agitated, restless and combative state progressing to frank stupor. Body temperature was uncontrollably high. Repeat CSF studies revealed elevated pressure, lymphocytosis, normal protein and sugar, and positive IgM Dengue Virus. Serum study for Dengue virus IgM Capture Elisa was also positive. Other significant tests ruled out Malaria, HIV and NMDA antibody as source of encephalitis. After intravenous steroids were started, on top of antipsychotics, clinical symptoms were noted to eventually resolve.

We theorize that dengue encephalitis should be considered in the differential diagnosis of acute viral meningoencephalitis though the classical manifestations of dengue may not exist. While dengue infection may be endemic in Asian countries, this should be considered in other parts of the world especially when patients rapidly deteriorate in the course of the disease. Immunecompetence definitely play a vital role in the recovery. Steroid therapy may be life saving in very severe cases.

Intracranial monitors can help guide the care of patients with severe brain injury. The devices are invasive and so may be associated with complications. Furthermore, accurate interpretation of the monitors' data is needed to be of potential benefit. In this study we asked whether experience influences "device failure" or interpretation.

Retrospective analysis was performed on a prospective database that included 500 patients (median age 48; range 14-90) with severe brain injury and who received intraparenchymal multimodality monitoring through a triple lumen bolt (Licox IMP#).

A total of 698 triple lumen bolts were placed during an 8-year period. Device failure was defined as: 1) broken or bent (n=20; 2.87%); 2) improper placement (N=11, 1.57%); and 3) ineffective (no response to O 2 challenge n=11, 1.57%). There were 27 (3.8%) devices thought to provide "incorrect" data but subsequently were found to be accurate, i.e. improper data interpretation. There was a decline in device failure over the entire study period. Each calendar year was divided into quartiles. Device failure incidence was 30%, 10%, 55% and 5% per quarter, i.e. was greatest during the third quarter during the time of academic and staff changeover (OR 10.77; p = 0.0131). In addition, improper data interpretation was greatest during the 3 rd quarter.

Our data suggest that experience with multi-modality monitors is associated with a reduced incidence of device failure or improper data interpretation. Educational efforts may reduce the need for device replacement.

Financial Support: Peter LeRoux Funding from Integra for research.

While efforts to "go green" and promote sustainability are well-established in many sectors, there has not been an adequate push toward such practice in the healthcare and medical fields. Healthcare accounts for 11% of all commercial energy use, 4 bil pds of waste, and 8% of greenhouse gas emissions in the US. These figures requires significant for efforts to be implemented; we each subscribe to, "First, do no harm" demands that these negative environmental impacts be addressed and mitigated immediately.

The intent of this report is to investigate and analyze the opportunities the healthcare industry has to embark on sustainable practices. We analyzed green architecture for new healthcare campuses and renovation of outdated facilities, submit efficiency and cost analyses of disposable versus reusable textiles, and offer observations on innovative technologies being developed to promote sustainability. This study was conducted after an extensive review of published literature, verified statistical reports presenting the cost-effectiveness and improved efficiency of pursuing an sustainable model of healthcare delivery. In cluded is a cradle-to-grave analysis of multiple facets of the healthcare/sustainability field, and addresses a number of specialist-specific avenues, including critical care and anesthesiology.

Energy-efficient building options -including rooftop gardens and alternative power sources -can cut energy consumption by 60%. Healthcare providers in all fields are making efforts toward lowering the carbon footprint of hospitals by reducing greenhouse gas emissions and utilizing resourcing, second use and extensive recycling techniques and efforts. Extensive Life Cycle Assessments (LCAs) prove that reusable medical textiles and tools are dramatically less expensive environmentally and financially than their disposable counterparts.

While efforts are being made to promote sustainability in healthcare, more must be done. The evidence is clear: environmentally-conscious endeavors save money and help lessen the stress placed on the environment. For such a heavy-hitting culprit of consumption, the healthcare industry simply must begin implementing "green" practices based on already-present data.

Standard metabolic prediction equations have been validated in general critical care populations, but have not been well studied in the neurologically critically ill. We sought to determine whether: 1) standard prediction equations accurately predict caloric requirements in neurocritical care patients; 2) variation in resting energy expenditure (REE) exists among different subpopulations of neurocritical care patients; and whether the same factors influence REE among different neurocritical care subpopulations.

Indirect calorimetry measurements were retrospectively reviewed for 58 mechanically-ventilated patients admitted to the Neuro ICU from January 2009 to June 2012. The measured REE data were compared to the predicted basal energy expenditure (BEE) calculated with the modified Penn State University (PSU-m) equation. Patients were classified into 3 neurological subtypes, stroke (n=27), status epilepticus (n=14), and other (n=17). Traumatic brain injury (TBI) patients were not included.

Of the entire cohort, median measured REE was 1432 (IQR 1215-1689) kcal/d and median predicted BEE was 1519 (IQR 1318-1708) kcal/d. The predicted BEE correlated well with the measured REE (coefficient 1.00; p<0.001) in the overall cohort. There was no significant difference in the predicted calorie requirement for stroke or status epilepticus. However, there was a suggestion that patients with status epilepticus were relatively hypometabolic (defined as REE <90% of the predicted BEE) compared to other subgroup populations [OR=3.2; 95% CI (0.95-10.8); p=0.06]. Factors significantly associated with REE include: maximum 24 hour temperature, administration of intravenous sedation, body mass index (BMI) and sex. Age and hospital day of REE were not predictive of energy expenditure.

The PSU-m predictive equation accurately estimates caloric needs for patients with non-TBI neurological injury. Patients with status epilepticus may be hypometabolic relative to other neurologically injured patients, which may be due to use of multiple sedatives in this subpopulation. Further research is needed to confirm these findings.

The American Society of Anesthesiology provides guidelines for preoperative fasting for healthy patients undergoing elective procedures. These guidelines are often extrapolated to the critically ill population for procedures and extubation. We tested the hypothesis that NPO practice differs between subspecialty, institution and practitioner-type.

After IRB approval, we conducted surveys of the memberships of the Society of Critical Care Medicine (SCCM), Neurocritical Care Society (NCS), and American Burn Association (ABA) regarding their NPO practice in critically ill patients. Survey questions included frequency of use of nasogastric (NG) vs. nasoduodenal (ND) tubes, NPO time prior to procedures, and NPO time prior to extubation. Responses were analyzed with Stata 12.0, using a one-way analysis of variance by ranks.

We received a total of 1402 responses (9% response rate) encompassing practitioners from medical, surgical, neurosurgical (130 responses), pediatric, cardiac, burn, trauma, and multidisciplinary ICUs. 106 respondents (7.6%) report 100% use of NG tubes, whereas 62 (4.4%) report 100% use of ND tubes. Excluding responses from pediatric ICU practitioners, the NPO practice in NICUs for intubated and non-intubated patients with ND tubes undergoing procedures is similar to respondents from other ICUs except the burn ICU (p<0.001). There is no difference in NPO practice of patients with NG tubes undergoing procedures across all ICUs. NICU respondents report the most commonly used NPO time prior to procedures is 6 hours for intubated patients with ND tubes (33.1%) and 8 hours for those with NG tubes (40.9%). For burn ICU respondents, the most commonly reported NPO time for intubated patients with ND tubes prior to procedures is 0 hours (54.9%), while 8 hours is reported for those with NG tubes (37.7%).

NPO practice in critically ill patients varies across the subspecialty units. Further research is necessary to develop evidence-based guidelines for NPO practices in the critically ill patients.

Patients intubated for primary neurological reasons represent a unique critically-ill population. Extubation failure rates in primary brain injury (PBI) patients are 18-38% compared to 13-18% in the general critical care population. These populations have never been directly compared. We hypothesized that intubated PBI patients would have higher rates of extubation failure compared to non-PBI patients.

Retrospective cohort of intubated patients admitted to the Medical Intensive Care Unit or the Neurocritical Care Unit in a tertiary-care university hospital between October 1, 2008 and September 30, 2010. Extubation failure was defined as requiring endotracheal intubation at 48 hours, 72 hours and one week.

Of the1684 . Failing extubation at 72 hours did not put patients at increased risk for VAP. Total ventilator days were similar between PBI and non-PBI patients. PBI patients who failed at 72 hours did not have a significant increase in ventilator days, intensive care unit days or mortality.

Our data indicates PBI patients are at increased risk for extubation failure compared to non-PBI patients. Future prospective study is warranted to determine predictors of extubation failure at 72 hours in PBI patients.

Peripherally inserted central catheters (PICC) is been routinely used instead of Central Venous Catheter (CVC) in our intensive care unit (ICU) patients, that includes critical Neurologic/Neurosurgical patients. There are a number of studies has been done to evaluate risks of PICC placement in general medical and surgical ICUs.

A retrospective analysis to determine risk of Large Vein thrombosis due to PICC in Neurologic sub-population of patients in a general medical/surgical ICU. Charts and venous ultrasound studies of patients admitted to ICU primarily for neurologic condition were reviewed.

Out of 183 consecutive patients, 54 underwent PICC insertion. 10 (18.5 %) had clinical and ultrasound evidence of large venous thrombosis attributed to PICC. The presence of a PICC line conferred a relative risk of 4.78 for the development of a DVT. Patients with PICC lines had a longer duration of stay in the ICU (mean days = 12.2 +/-8.3) when compared to patients without PICC lines (mean days= 3.2 +/-2.4) t(57) = 7.9, p <.001.

Routine placement of PICC instead of CVC is associated with increased risk of thrombotic events in large veins in Neuro Critical sub population of a general ICU, which may be associated with longer ICU stay. More caution should be exercised before routinely using PICC instead of CVC.

There are many potential obstacles to guideline adoption and compliance in clinical practice. The purpose of this research was to develop a computer-readable format for clinical pathways, guidelines, and research protocols such that they could be rapidly distributed, displayed at the bedside, and driven by patient context. The goal is to increase guideline compliance and reduce errors made at the bedside.

We collected 12 institutional clinical practice guidelines from the abstract authors, 7 guidelines from professional societies (including the Neurocritical Care Society), and one multi-center research protocol (BOOST-II). We analyzed each to look for common constructs that would form the basis of a computer-readable care path "language". We also reviewed previous attempts at computer-readable guidelines to discover what might be applicable to our system.

The analysis showed considerable variation in the way guidelines are put to practice at the bedside. Despite this, we found a set of generalized patterns that were used to develop a care path representation (language) that could encapsulate the content of the guidelines. Structured goal-oriented steps, alarm and time couplers, and a "monitoring cycle" were designed and represented in an XML-based language. A scripting method for decision logic also was developed. Software was written to read the XML script, display the care path "flow-chart", provide interaction with the health care provider, and links to related instructional content. Integration with real-time multimodal monitoring data allows the care path to be driven by the context of the patient.

This abstract outlines the first part of a larger project to develop an open-standard guideline format and display software that will decrease the time to adoption of neurocritical care guidelines and increase compliance in clinical practice.

Financial Support: Funding received by Moberg Research from NIH/NINDS and US Army/TATRC to carry out this work. One of the authors (R Moberg) is President and Owner of Moberg Research.

The objective of this study was to develop empiric treatment guidelines for patients admitted to the Neurosciences Intensive Care Unit based on unit specific antimicrobial surveillance.

A prospective chart review was performed from October 2011 to April 2012 of all adult patients admitted to the Neurosciences Intensive Care Unit with positive cultures from any site. In addition to culture data and antimicrobial sensitivities, time of admission, diagnosis, placement of an external ventriculostomy device (EVD), duration of cefazolin prophylaxis and risk factors for healthcare-associated infections (HAI) were collected. Hospitalization within 30 days, residency in an extended care facility or hemodialysis at the time of admission were considered HAI risk factors. Cultures were analyzed as those occurring before or after day 7 of NICU stay. Patients residing in the unit as a result of overflow were excluded.

A total of 50 patients and 130 positive culture results were included and analyzed by duration of ICU stay <7 days (n=69) or > 7 days (n=61). EVD placement and cefazolin prophylaxis were present in 65% of patients for a mean of 8.2 days. At <7 days, methicillin-resistant Staphyloccous aureus (MRSA) was the most common pathogen in patients with risk factors for HAI. At <7 days without risk factors, the most common pathogens were methicillin-sensitive Staphylococcus aureus (MSSA) (n=8) and Enterobacter (n=5) in the sputum and Enterococcus (n=4) in the urine. A further analysis revealed theseisolates emerged after day 5 of admission in patients receiving cefazolin prophylaxis. Beyond 7 days, sputum isolates predominated and consisted of gram negative pathogens (n=18), MSSA (n=11) and MRSA (n=12).

Selective pressure from cefazolin prophylaxis was apparent in unit surveillance and emerged at or after 5 days. Based on these results, institutional empiric antibiotic treatment regimens were adjusted to cover these pathogens after day 5 of NICU stay.

The direct thrombin inhibitor dabigatran etexilate is approved for the prevention of stroke and systemic embolism in patients with non-valvular atrial fibrillation. Despite the clinical benefits of dabigatran, hemorrhage remains a feared complication due to the lack of reversal agent and limited experience with interventions to reverse dabigatran's anticoagulant effect. In addition, reliable laboratory tests to measure the degree of anticoagulation associated with dabigatran are not widely available.

An interprofessional team developed institutional protocols for the management of dabigatran and dabigatran-associated hemorrhages. Clinical and neuroimaging data was collected from four patients with dabigatran-related subdural hematoma, subarachnoid hemorrhage and/or intracerebral hemorrhage who were treated between November 2011 and April 2012. Data collected includes age, gender, past medical history, renal function, coagulation and hematology parameters, computed tomography findings, blood products or clotting factors administered, and hemodialysis parameters, if applicable.

The patients ranged in age from 79-90 years. All patients were inappropriately prescribed dabigatran due to age over 75 years, renal insufficiency, increased bleeding risk, and/or an unlabeled indication. Serial evaluation of each patient's coagulation assays was conducted in order to quantify the degree of anticoagulation. Three of the four patients received emergent hemodialysis and one patient received recombinant activated factor VII. Two patients received blood products, including FFP and platelets, with no observed clinical change. All patients survived to hospital discharge.

Though this case series is small, it demonstrates the importance of thoroughly evaluating a patient's renal function, bleeding risk, and concomitant medications to determine the appropriateness of dabigatran therapy. It is imperative for clinicians to understand dabigatran's pharmacokinetics and recognize the major factors that increase dabigatran exposure. Increased age and renal insufficiency seemed to play a significant role in the hemorrhagic cases we encountered.

Post-surgical cerebral venous sinus thrombosis is extremely rare. The management of this complication is challenging for neurointensivists; since anticoagulation may increase the risk of bleeding after craniotomy.

A previously healthy, 21 year-old male was found to have a right cerebellopontine angle mass on brain magnetic resonance imaging (MRI) during headache evaluation. He underwent a prolonged surgery with retrosigmoid craniotomy and resection of an acoustic schwannoma in left lateral decubitus position.

Immediately post-operatively, the patient had a seizure. Brain CT-scan showed hyperdensity in the right transverse sinus suggestive of thrombus. Cerebral angiogram confirmed occlusion of the superior sagittal sinus (SSS) torcula and bilateral transverse sinuses. Intravenous heparin was initiated; however, due to further deterioration with brain herniation, endovascular administration of tissue plasminogen activator (rt-PA) assisted with thromboaspiration with Penumbra catheter was performed, followed by continuous infusion of rt-PA (1 mg/hr) via microcatheter in the SSS for 2 days. A repeat angiogram showed near complete recanalization of the sinuses. Heparin was continued, but he developed heparininduced thrombocytopenia, and was switched to bilvalirudin. His hospital course was complicated with intraventricular hemorrhage, acute respiratory distress syndrome, methicillin-resistant staphylococcus aureus (MRSA) bacteremia, and Takasubo's cardiomyopathy. He had residual right facial nerve palsy and hemiparesis related to pontine ischemia. His prothrombin gene mutation was positive for one copy. He was ambulating with assistance prior to discharge to acute inpatient rehabilitation.

Cerebral venous sinus thrombosis is a rare complication of retrosigmoid resection of CPA tumor. Aggressive treatment with endovascular rt-PA administration into the venous sinuses may be life-saving, but carries significant risks in the fresh post-operative period.

Many lives have been lost due to the loss of the airway in critically ill patients. The introduction of the video laryngoscope has been a useful tool that has saved many lives in recent years. One of the limitations of the video laryngoscope is that despite being able to see the vocal cords and the airway beyond it may be difficult to advance the endotracheal tube into the airway. We present a novel approach using a bougie to simplify this problem.

A cohort of ten patients in a community critical care unit with difficult airways were intubated using this new technique in a nonrandomized fashion over a period of six months. A video laryngoscope was used in each case with patients sedated and paralyzed with usual agents used for rapid sequence induction. Historically, the bougie when used with standard laryngoscopes is introduced into the airway by line of sight and the endotracheal tube is advanced over the bougie. With a video laryngoscope a direct line of sight is not available and passing the bougie is challenging because of a 90 degree angle from the open mouth to the airway. This new technique involves lubricating both the metal stylet with the 90 degree turn and a 15 french x 70 cm bougie and advancing them to the end of the endotracheal tube. The bougie is then advanced into the airway through the endotracheal tube under direct vision through the video laryngoscope. The endotracheal tube is then advanced into the airway over the bougie.

All ten patients were intubated without difficulty and without complication.

This new technique should be considered as an option in securing the airway in critically ill patients. Further validation testing by other investigators is warranted regarding this new technique to determine if a randomized controlled trial is justified.

Francis R. 

Ventilator-associated pneumonia (VAP) remains a problem in traumatic brain injury and high-risk surgery patients. We use early non-bronchoscopic broncho-alveolar lavage (screening-BAL) in the surgical intensive care unit (SICU) to identify ventilated patients with bronchiolar bacteria prior to 48 hours. We reviewed results of these screening-BALs in neurotrauma patients from 3/2011 to 6/2012.

All ventilated patients in the SICU underwent screening-BAL 36-48 hours after intubation; quantitative cultures (>10 4 CFU/mL) were used to identify positive specimens. Clinical pneumonia was defined as clinical pulmonary infection score (CPIS)>6 and subsequent positive diagnostic-BAL. Continuous and dichotomous data were compared from the screening-BAL results and clinical diagnosis of pneumonia.

Screening-BALs were performed in 47 neuro-trauma patients (mean ISS 19.9 ± 6.11) with an average head abbreviated injury score (HAIS) of 3.83 ± 0.94. Thirty-three of these were positive for organisms (70%). Twenty-four clinical pneumonias were diagnosed and in 20 of these patients the causative organism identified was the same organism in the screening-BAL (83.3% agreement; Kappa 0.357; p = 0.02). One patient with a negative screening-BAL developed clinical pneumonia. The median day to develop pneumonia was 3.5 (2, 13). The HAIS was higher in patients with a positive screening-BAL (4.06 ±0.7 vs. 3.29 ±1.2; p = 0.03). There were no significant differences in the age, ICU length of stay, ISS, or HAIS in patients with a positive screening BAL vs. the patients that developed a clinical pneumonia.

Positive screening mini-BAL results are associated with the development of VAP by the same organism. Screening BAL in neuro-trauma patients may be a mechanism to identify patients who are at-risk for developing pneumonia later in their hospitalization and early identification of the causative agent. Further studies are warranted to determine if intervention on these results changes clinical course. 

Human rabies is a relatively rare disease in the United States, with approximately 2 cases diagnosed annually. The most common exposure in the U.S. relates to bats, however canines and other animals have also been implicated. The typical incubation period from exposure to development of symptoms is 2-3 months, while periods of up to 6 years have been described. We present an atypical case of human rabies presenting after a prolonged incubation period in the United States.

We describe a case of a 40 year old Brazilian man without prior medical history who presented with progressive sensory symptoms leading to encephalopathy and ultimately death. Extensive workup revealed no other causes of his symptoms, and brain tissue samples sent to the CDC at the time of his death confirmed a diagnosis of rabies by direct fluorescent antibody testing. In addition, sequencing of the virus confirmed a variant found in canines in Brazil. The patient had not traveled to Brazil in over 8 years, and had no confirmed exposure other than an encounter with a wild dog in Brazil without reported bites or scratches before immigrating. Because the viral genotype has not been previously identified among animals in the United States, this case represents the longest confirmed human rabies incubation in the United States to date. Characterization of the illness revealed loss of evoked potentials, electroencephalography amplitude attenuation, MR spectroscopy changes of the deep nuclei, and an atypical inflammatory response on pathologic testing. We speculate that either an atypical immunologic response or the patient's recent anabolic steroid use may have mediated delayed progression.

This case underscores the importance of keeping human rabies in the differential diagnosis of rapidly progressive encephalomyelitis, even without an exposure history, or with a remote exposure history.

The Full Outline of UnResponsiveness (FOUR) score has been validated as an alternative to the Glasgow Coma Scale (GCS) in the evaluation of stuporous and comatose patients and predicts long-term outcomes. The utility of serial FOUR score and GCS by nurses in detecting changes in neurologic exam in the Neurocritical Care Unit (NCCU) and whether high frequency monitoring after the first assessment is beneficial has not been studied.

The electronic charts of 50 consecutive patients with surgical and non-surgical brain pathology admitted to a NCCU were reviewed, yielding 2881 observations of GCS, FOURscore, and cranial nerve assessments. Changes in neurologic exam promoting notification of a provider were abstracted from nursing notes.

Of 50 patients (M:F:28:22, Age: 54+2 yrs), 25 had semi-elective neurosurgery, 8-ischemic/hemorrhagic stroke, 6encephalopathy/infection, 4-subarachnoid hemorrhage, 4-traumatic brain injury, 2-seizures, 1-other. Admission median GCS was 14(iqr-5); median FOURscore was 16(4). Comparison of 2881 q2-4h FOURscore vs. 311 qdaily FOURscore readings showed no significant difference in FOURscore by frequency of measurement (p=0.61). In 23 occurrences of change in neurologic exam resulting in provider notification, changes in mean FOURscore and GCS from 4 hours prior to the event were 1.2(SD-2.4) and 1.2(2.1) respectively; p=1.00). From 2 hours prior to event, changes in mean FOURscore and GCS were 0.9(2.1) and 1.1(2.1) respectively; p=0.41). In one cerebral herniation event, neither scheduled FOURscore nor GCS changed.

Use of the FOURscore for serial monitoring and early detection of worsening of neurologic condition performs similarly to GCS and is less sensitive than subjective assessment of trained NCCU nurses. The utility of incorporating the FOURscore into the on-going nursing assessment paradigm of all NCCU patients requires further evaluation. There may be subsets of patients or conditions (with lower sumscores than in our cohort) for which daily or more frequent monitoring has predictive value.

A technique for real time, non-invasive blood flow monitoring would be a major asset to clinicians in neurocritical care. We studied the ability of a new hybrid technology employing ultrasound tagged near infrared spectroscopy (UT-NIRS) to detect changes in cerebral blood flow (CBF) as compared to measurements by 133 Xenon single photon emission computer tomography ( 133 Xe-SPECT).

Twelve healthy volunteers were enrolled in the study. A CerOx monitor (Ornim Medical Ltd. Israel) provided continuous UT-NIRS monitoring of regional tissue oxygen saturation (StO 2 ) and regional cerebral blood flow index (CFI). 133 Xe-SPECT (Ceraspect; DSI, Waltham, MA, USA) was then used to measure CBF at baseline, 15 minutes and 60 minutes after acetazolamide injection.

Ten subjects completed the study. Significant increases in CBF as measured by both UT-NIRS CFI and 133 Xe-SPECT CBF were noted 15 minutes after acetazolamide injection. At 60 minutes following injection, 133 Xe-SPECT CBF had returned to baseline while UT-NIRS CFI remained elevated compared to baseline. A significant correlation between UT-NIRS CFI and 133 Xe-SPECT CBF values was found at 15 minutes but not 60 minutes after acetazolamide injection. Specificity and sensitivity for detecting an increase in CFI following acetazolamide injection were calculated using a receiver operating curve (ROC), with an area under the curve of 0.95 (+/_ SEM 0.001). No statistically significant changes in UT-NIRS StO 2 were noted following acetazolamide injection.

UT-NIRS CFI can detect increased CBF following acetazolamide injection, correlates with a gold standard, 133 Xe-SPECT, and the ROC curve analysis demonstrates excellent discrimination. The difference in the measurements at 60 minutes may be explained by different ratios of gray matter to white matter in the regions of interest as assessed by the two techniques. UT-NIRS CFI can be more sensitive to changes in cerebral perfusion than simple regional tissue oximetry.

Financial Support: Dr Gress is a member of the Scientific Advisory Board of Ornim Medical Ltd and holds stock options in the company.

Level of coma has traditionally been measured clinically (e.g. Glasgow Coma Scale, Four Score, etc.), or with neurodiagnostic tests (e.g. EEG). Developing more objective, longer term measures of coma could improve quantitation of arousal and modification of response to therapy. We used a post-cardiac arrest (CA) rodent coma model to test 3-D bodily acceleration as a wireless, continuous measure of early movement during coma arousal, and compared it to EEG based markers validated previously.

Five adult Wistar rats (Male, 300-350gms) underwent EEG electrode implantation 1wk prior to asphyxia-induced 7min CA. Four hours after resuscitation, rats were attached to a wireless EEG-accelerometer system. Wideband and sub-band EEG were analyzed to yield IQ, an entropy based and previously validated measure of coma arousal. We defined activity as the variability in 3-D acceleration as quantified by the standard deviation of acceleration.

We found a significant positive linear correlation between accelerometer activity and full band EEG IQ (R=0.60± 0.23, mean ± SD). When EEG sub-bands were divided into two categories (0.3-50Hz and 50-150Hz), accelerometer activity had better correlation with higher frequency sub-bands (R=0.27±0.10 vs. R=0.46±0.21). During individual sub-band analysis, we were able to find a moderate correlation with the higher frequency IQ 50-90Hz (R=0.49±0.20).

These results suggest that 3-D acceleration based activity, measuring early subtle movements during coma arousal, correlates with EEG IQ. This relationship was stronger for higher frequency sub-bands. This suggests that subtle motor activity quantitated by an accelerometer may be an acceptable indirect measure of arousal. Such accelerometer-based systems also have the advantage of being more objective and affordable while also offering longer term monitoring. Therefore, accelerometer-based monitoring for coma arousal may have clinical applicability in intensive care units. 

Recent literature emphasizes the impact of vancomycin concentrations on patient outcomes, especially in serious infections such as central nervous system infections (CNSI) and pneumonia. Achieving adequate concentrations is challenging in the critically ill due to changes in volume of distribution and clearance. We investigated the impact of a PVDS in our neurologic units.

Retrospective chart review comparing outcomes of PVDS (RPh-group) to pre-implementation control group (MD-group). Adult inpatients receiving vancomycin on neurologic units (Neuro ICU and floor) were included in a 6 month pre/post period. RPh-group patients receiving vancomycin not consulted to PVDS were excluded. Outcomes evaluated number of vancomycin levels and proportion within goal range (10-25 mcg/mL).

In MD-and RPh-groups, 113 and 123 patients were enrolled, respectively. RPh-group had a higher percentage of patients with weight >100 kg and CrCl >50 mL/min. ICU patients accounted for 47% and 61% of the MD-and RPh-groups, respectively. Common indications were CNSI and pneumonia in both groups. 209 levels were drawn in MD-group versus 228 levels in RPh-group. A higher percentage of levels were within goal range in RPh-group (68%) versus MD-group (52%, p = 0.0005). Amongst patients with CNSI, RPh-group had a higher percentage of levels within goal range (65% vs. 51%, p = 0.08). ICU patients in RPh-group had a higher percentage of levels within goal range (69 vs. 59%, p = 0.108). In ICU patients, younger age (p = 0.083) and CrCl >50 mL/min (p = 0.061) trended toward initial subtherapeutic levels despite receiving ~25 mg/kg/day of vancomycin.

Implementation of the PVDS in neurologic units resulted in higher attainment of therapeutic concentrations. In ICU patients, addition of a loading dose or higher daily doses of vancomycin may need to be employed by the PVDS to ensure achievement of target concentrations.

Intraventricular therapy (IVT) with polymyxin B (PolyB), an antibiotic with similar pharmacological action to colistin (PolyE), by external ventricular derivation (EVD) has the main goal of offering major bioavailability of the drug, since its use by intravenous and direct action are restricted by the blood-brain barrier, with penetration of only 25%. The patient of the present report had arterial venous malformations followed by hemorrhagic stroke, which caused elevated intracranial pressure. The objective is to show an example of the effect of IVT PolyB in a patient with meningoencephalitis infection by multidrug-resistant Gram-negative bacteria (A. baumannii and P. aeruginosa), that are common in the ICU. A literature review was made on the subject of therapy with PolyB about the pharmacological characteristics, nephrotoxicity and neurotoxicity. A comparative table of the resistance profile of the strain treated in this study was created, with the intrinsic resistance of the species. Also, the development of liquor evolution (culture and routine) of the patient before the treatment was monitored, until negative liquor. The effectiveness of EVD, the colonizer germ and monitoring of the serial aspects of the liquor were analyzed.

The patient was treated with intravenous and intrathecal administration of PolyB (IVT) from November 14th to November 28th.

On 11/14/2008, therapy with intravenous PolyB was started: 1500000UI(20.000UI /Kg/d) once a day, on every day of treatment; and IVT by EVD: 50000UI in solution once a day during the first three days, and on alternate days during all the treatment. As a result of the use of intrathecal PolyB associated with intravenous, effectiveness was proven in the routines of liquor negative for such germs, not showing any reports of neurotoxity and nephotoxity.

IVT PolyB proved to be very efficient on treating meningoencephalitis quickly. No toxic effect was associated with the drug.

Enhancing the level of alertness in comatose patients after acute brain injury is a very challenging problem. the use of alerting agents like Modafinil is reasonably established for TBI patients in the chronic phase but not in the acute settings. We retrospectively reviewed the use of these agents at our center over a five year period to determine efficacy and use patterns in the acute brain injury settings.

A chart review for patients who were admitted to the NICU at DUMC during (2007) (2008) (2009) (2010) (2011) and treated with an enhancing agent (Modafinil, Methylphenidate) for decreased level of alertness secondary to an acute brain injury. Electronic records were then reviewed to confirm the intended use of the agent, and a number of clinical data elements was recorded.

99 patients were found to meet study criteria and data elements were extracted. 85 patients received Modafinil, 14 received Methylphenidate. The average GCS was 8 on admission and 9 at discharge. Average delay in trialing alerting agents was (10.3) days and in most cases the agents were used within a few days of withdrawal of care or discharge to hospice. Outcomes varied widely with (50.5%) going to nursing home, (24.2%) going to rehab, 10.1% going home and 16.2% to hospice or death. SAH was the most common injury (33.3%) followed by ICH (23.2%), SDH (7.1%) and TBI (6.1%). Review of documented GCS during acute hospitalization showed no significant changes during the period of alerting agent trial for any diagnosis other than TBI. In TBI a significant 3 points improvement was seen on average.

Our data showed that starting Methylphenidate or Modafinil for the purpose of improving the level of consciousness in acute brain injury patients is not effective except for patients with traumatic brain injury. Based on these observation alternative agents like L-Dopa should be explored.

Nurses in the Neurocritical Care Unit (NCCU) are responsible for performing serial neurological exams to establish baseline and potentially detect patient deterioration. Nurses spend considerable time doing frequent neurological checks but the current neurological exam is open to subjectivity. We want to quantify the agreement between nurses doing these exams.

Over the course of one week we tracked the neurological exams of 51 patients admitted to the Neurocritical Care Unit. We compared exams between the off-going and on-coming nurses. Each exam consists of 13 single elements, LOC, orientation, right and left pupil size, reaction and description, characteristics of speech/communication and motor response in all four extremities. Grouping right and left pupils gave 10 element-groups. We examined 186 change of shift (CoS) opportunities. When there was more than one variation of an element-group a thorough chart review was performed to identify clinical indicators, such as medications given, to determine if there was a true clinical explanation for the variation.

CoS exams were the same between nurses 49% of the time whereas 20% of exams had a single variation and 31% contained 2 or more single variations. Of the 42 CoS opportunities with multiple variations in element-groups only 7 exams showed a clinical reason for the change. That leaves 35 exams with multiple unexplained variations. This accounts for 19% of overall total exam opportunities.

Nearly 20% of the time nurses do not agree on the neurological exam of a patient when examined before and after CoS. Inconsistency in terminology and methods between nurses may hinder accurate communication. A comprehensive literature search did not reveal a standard neurological exam for NCCU nurses. Further discussion needs to take place between neuro-nurses across the nation with the goal of defining terms and developing a national standard for the serial neurological exam performed by nurses.

Electrical Impedance Spectroscopy (EIS) is novel, portable, easy-to-implement device that aims to provide rapid, affordable point-of-care detection, assessment, and monitoring of acute brain injury. An adaptation of "passive" electroencephalography (EEG), EIS relies on non-invasive measurement and modeling of the conduction of minute electrical currents applied transcranially across a spectrum of frequencies. Our purpose was to test of the feasibility of EIS to distinguish the impedance differences between normal subjects and brain injury attributable to acute/subacute intracranial hemorrhage or subacute ischemic stroke.

We performed a prospective, observational, proof-of-principle study of 9 patients admitted to our Neurosciences Intensive Care Unit for ischemic stroke or intracranial hemorrhage, and 4 healthy volunteers. 15-minute EIS recordings were obtained for each patient. The EIS device delivered a small "white-noise" alternating current through a pair of stimulation electrodes; voltages were recorded across three bilaterally symmetric electrode pairs in an EEG montage. Log-log plots of impedance (y-axis) as a function of current frequency (x-axis, range 100Hz-100kHz) were produced for each set of electrodes per patient.

Mean age was 68 years (range 42-90); 61% (8/13) were female. Of these 9 brain-injured patients: (a) among all 4 patients with subacute hemorrhage (days old), impedances dropped at higher current frequencies; (b) among all 4 patients with subacute ischemic stroke (days old), impedances increased at higher frequencies; and (c) in one patient with acute hemorrhage (hours old), impedances were not significantly different at higher frequencies but evolved to the subacute hemorrhage pattern (a) at a day-6 follow-up recording. All brain-injured patients were distinguishable from normal control volunteers.

EIS is a noninvasive, portable diagnostic modality that has potential for clinical applications in multi-modal neuromonitoring and far-forward battlefield/ambulance arenas for diagnosing and monitoring acute and subacute brain injured patients. Future development requires clinical validation, standardization, hardware and software optimization, and graphical user interface development.

Financial Support: This work is supported by National Institute of Biomedical Imaging and Bioengineering Point of Care Center for Emerging Neurotechnologies (POC-CENT), subaward 5U54EB007954-04 and by an "Innovation Gra

Hypertonic saline (HS) improves cerebral edema, blood flow, and is inexpensive. However, use of HS is complicated by reports of induced renal dysfunction and associations with increased blood-stream infection. We hypothesize HS alters renal perfusion leading to a state of relative renal insufficiency.

With institutional review board approval, we retrospectively reviewed our hospital's use of HS since March of 2005, and prospectively since October 2010. Comparisons were made between admission diagnoses, changes in creatinine (Cr), and formulation of HS received (3% NaCl, 3% NaCl/Sodium Acetate mix, and 23.4% NaCl) to patients receiving normal saline or lactated ringers. Intervariable associationswere calculated between using Pearson's correlation coefficients. 1329 patients of the retrospective portion were identified. The data presented represents the first 230 patients with data.

There were significant differences in the APACHE II scores and Glasgow Coma Scale (GCS) scores between the different formulations of HS. The overall correlation of chlorine (Cl -) and sodium (Na + ) with creatinine (Cr), and within each of the saline types, were not significant. When dichotomized by the diagnosis, significant correlations appear. Traumatic brain injury (TBI) patients demonstrated moderate correlation between Na + & Cr of 0.45. Stroke patients demonstrated small correlations between Na + & Cr, and C l-& Cr (0.19 for both). Patients receiving HS outside the neurocritical care unit (NCCU) demonstrated a small but significant correlation between Cl and Cr at 0.29.

Patients receiving HS have lower GCS and higher APACHE II scores. Elevations of Na + or Cl in stroke, Na + in TBI, and Cl in non-NCCU patients correlating with elevations in Cr. As reductions in renal function predict mortality, therapies precipitating kidney injury are concerning. Cl -, a potent renal vasoconstrictor, reduces renal blood flow. Prospective comparisons of HS formulation and renal function are needed to further assess if formulation affects outcome and cost.

First recognized after rapid initiation of nutrition in prisoners of war during World War II, refeeding syndrome (RS) is the manifestation of fluid and electrolyte disturbances precipitation systemic dysfunction. Here we report a case of RS in a patient with Duchenne's muscular dystrophy (DMD).

A case report and literature review.

A 23-year-old male with past medical history of DMD, chronically ventilated and feed via a percutaneous endoscopic gastrostomytube, presented with pneumonia, sepsis, and status epilepticus. He was treated with broad spectrum antibiotics, early goal-directed therapy, and 18 hours of electrographic seizures suppression with a midazolam infusion. Admission labs demonstrated a minimally low albumin (3.4 g/dL), mild hypokalemia (3.1mmol/L), and the presence of urinary ketones. Enteral nutrition was started post-admission day (PAD) one. PAD 2 found elevated serum glucose and precipitous drops in potassium, phosphate, calcium, and magnesium refractory to replacement. PAD three attempts to wean the patient to his home ventilator setting failed, and he remained encephalopathic. Enteral nutrition was changed to a more elemental, peptide-based formulation, and multivitamin with thiamine was added. Electrolyte abnormalities persisted. PAD 4, it was learned the family had reduced the patient's daily enteral nutrition by approximately half over six months to have him fit within his wheelchair. Learning this, enteral feeds were reduced by half, advanced at a reduced rate reaching goal in 2 days, and electrolyte abnormalities resolved commensurately. Over the next three days, the patients mental status returned to baseline and ventilation improved. No cardiac or hemodynamic complications occurred, but his infections resolved slowly.

A significant concern in the critically ill, the constellation of problems associated with refeeding syndrome have systemic implications. These are centered on increased cellular uptake of phosphorus following the reintroduction of carbohydrates. The role of DMD in refeeding syndrome is uncertain, and has not been previously reported.

To determine hospital mortality and complication rates associated with surgical clipping and endovascular coiling of cerebral aneurysms in children, and to evaluate the trend of utilization of these procedures over the recent years in various US hospitals.

From the Kid's Inpatient Sample database for the years 1998 through 2009, we identified a cohort of children admitted with the diagnoses of intracranial aneurysms and aneurysmal subarachnoid hemorrhage. Hospital-associated complications and in-hospital mortality were compared among the clipping and coiling treatment groups. A multivariate logistic regression analysis was used to identify independent variables associated with hospital mortality. Cochrane-Armitage test was used to assess the trend of hospital utilization of these procedures in various hospital subtypes.

After data cleansing, 1120 children were included in the analysis. Two hundred (18%) children had aneurysm clipping and 920 (82%) had endovascular coiling procedures. The coiled group was younger (11.29 ±6.57 versus 14.49 ± 4.88)and had even gender distribution. Hospital mortality was higher in the clipped population, 6.09% versus 1.65% (adjusted odds ratio 3.02; 95% CI 1.24, 7.36; P = 0.002). In addition, hydrocephalus, status epilepticus and pulmonary complications were higher in the clipped population (P <0.001). Lastly, the length of hospital stay as well as the hospital charges was higher in the clipped population (P <0.0001). The rate of hospitals' use of the endovascular coiling has increased in various types of hospitals over the years included in this study (P <0.0001). The trend in mortality rates among the clipped population remained higher (5.4%-6.2%) compared to the coiled group (1-2.4%).

Endovascular coiling of cerebral aneurysms in children is associated with fewer deaths and complications, shorter hospital stay, and less hospital charges compared to clipping. The trend of hospitals' utilization of coiling procedures has increased during the recent years.

Understanding and managing complex physiologies is a critical, but difficult, problem in the Neruologicial-ICU. Most of the information that must be assimilated in the ICU exists at the level of raw data, individual test results and observations, and individual clinician notes. This mass of data obscures a holistic view of the patient, hides the development of trends, makes it difficult for clinicians to notice interactions between different variables. Graphical displays and patient summaries enhanced or outperformed traditional text displays in numerous studies (Elson & Connelly, 1997; Balas et al. 1991) , but this work hasn't yet been extended to support intracranial pressure (ICP).

The aim of this effort was to develop an interactive ICP-specific data visualization using cognitive engineering principles. The visualization is designed to transform and consolidate complex multimodal physiological data into integrated interactive displays.

We have developed a drill-down interactive visualization to enable clinicians to manage ICP and identify blood pressure target goals that will ensure adequate cerebral perfusion and thereby create and maintain an optimal physiologic environment for the comatose injured brain to heal. Using high-resolution physiologic monitoring data, this drill-down screen depicts the status of cerebral autoregulation using methods well described in the clinical literature (Czosnyka, Smielewski et al. 2001; Jaeger, Schuhmann et al. 2007 ) Additionally, the drill-down provide graphical display of bloodpressure, intracranial pressure, and brain oxygen tension over time.

With this interactive visualization, along with medication and lab data, the clinician can determine the target brain oxygen tension for a specific patient and whether to intervene on blood pressure, intracranial pressure or a combination of both in order to achieve a brain oxygenation goal (i.e., goal-directed therapy). The next step in this project is to conduct an experiment comparing this visualization against standard methods.

NICOM is a novel technique of monitoring hemodynamic status which is based on bioreactance technology. Ventricular outflow causes changes in the phase of radiofrequency waves as they cross chest. Measuring the phase shift enables calculation of flow. Technique is entirely non-invasive.

Retrospective analysis of collected data.

We describe the use of NICOM in a tertiary care neuroscience intensive care unit. 41 patients were monitored on the NICOM from January 2011 until June 2012 for an average of 8 days. Diagnoses of patients monitored on NICOM were: SAH -11, ischemic stroke -9, ICH 6, TBI -5, SDH-3, Brain Tumor-2, spinal surgery-2 and others. 48% of patients were on mechanical ventilation, 31% were treated with pressors. In the first 48 hrs of monitoring, there were 32 PLR (passive leg raising) tests and 2 fluid challenges performed to measure fluid responsiveness. 22 patients (53%) were fluid responsive and 16 (39%) had an intervention. Selected cases will be presented

NICOM system is safe and can be useful in the NeuroICU setting. It can be used in intubated patients with sepsis, unexplained hypotension, hypertensive therapy in SAH or during hypothermia therapy. It is also useful in non intubated, alert patients, were fluid status has to be monitored closely. Although NICOM is a seemingly simple-to-use technology, there were multiple clinical challenges including education of the staff, proper test performance and consistent charting. Inconsistent machine calibration, use of compression stockings during a PLR, and untimely sensors changes were the main problems. In the neuroICU patients with increased ICP, use of fluid challenge can be safer than PLR. Repeated staff training resulted in more consistent data.

Limited information is available regarding the current state of informatics in various NCC units. We sought to assess the current state and needs for informatics infrastructure to help determine priorities and future directions of informatics research in Neurocritical Care.

A survey instrument was developed and with the support of the Neurocritical Care Research Consortium chair, distributed to the 152 participants/registrants of the 2 nd Neurocritical Care Research Conference.

A response rate of 47% (72 of 152) was achieved. Most responders worked in an academic medical center (93.1%), level 1 trauma center (70.4%) and/or mixed multi-bed (mean=18.7) neuromedical/neurosurgical ICU (72.2%), commonly treating ICH (97.2%), SAH (95.8%), Ischemic stroke (86.1%), and Traumatic Brain Injury (69.4%). 

Acquiring, integrating, storing and analyzing MM data in a comprehensive informatics architecture for clinical and research use is stated as important but is rarely achieved due to financial and technical barriers. A centralized dissemination of technical assistance and a societal statement prioritizing informatics to advance NCC research may help facilitate future adoption.

Access to Neurocritical Care Units (NCCUs) in the Mountain West is geographically limited. We evaluated practice patterns among providers in this region and hypothesized that hospital size and distance from NCCUs impact decisions to transfer patients with critical neurological illness.

Surveys were sent to hospital providers with varying degrees of access to NCCUs in the Mountain West, to examine what factors influence decisions to transfer patients with critical neurological illness. The survey queried location, hospital size, locally represented specialties, patterns of transfer, frequency of illness presentation, influences for and against decisions to transfer such as timeframes and perceived futility, and awareness of NCCUs and services they provide.

151 responses were received. Responses were grouped by distance from the closest NCCU and by hospital size. Results showed that futility in outcome has a strong influence on decisions against transfer for smaller hospitals and hospitals that require air transport (P<0.05). Notably, distance required to transfer is not a strong factor in the decision to transfer patients (P=0.168). For larger hospitals and hospitals within ground transport range of a NCCU, patient condition, patient risk during transfer, and specialized intensivist support are less influential in transport decisions.

Patient transfer for critical neurological illness originates from hospitals with varying size and geographic access to NCCUs. While distance required to transfer does not appear to be a significant limitation, perceived futility in outcome is a strong influence against deciding to transfer. Among providers in smaller hospitals at greater distance from NCCUs, significantly more providers have never heard of NCCUs or services provided. These findings suggest that therapeutic nihilism regarding critical neurological illness in smaller hospitals at greater distances from NCCUs influences patient outcomes. Patients and providers in these locations may be significantly impacted by further education about Neurocritical Care and implementation of Tele-Neurocritical Care services.

Neurocritical care is a multidisciplinary specialty whose participants originate from diverse medical backgrounds. Review of the growing body of literature is essential for clinicians and strategies for continuing education may be expected to be unique for this field. This exploratory survey aims to define how the neurocritical care team (NCCT) educates itself.

A fifteen question survey was sent to all Neurocritical Care Society members and responses were gathered over a one month period. Basic statistical analyses of rates and comparisons of response rate proportions were conducted.

187 surveys were returned (17%). 62% of respondents were physicians, 27% were non-physician team members, and 11% were physicians in training. Regardless of background or training, individuals seek published literature through a combination of electronic-print media outlets (72%) rather than a singular approach. However, 85% spend the most time reading journal articles. 59% of NCCT members review the same journals monthly and allocate individual manuscript time contingent upon interest. Neurocritical Care (91%), Critical Care Medicine (67%), and New England Journal of Medicine (54%) are the most commonly reviewed journals. 38% of NCCT members do not attend a journal club. Academic neurointensivists (75%) and fellows (77%) are most and nurses are least (31%) likely to attend. Participation in NCC subspecialty (33%) or general critical care (32%) clubs is more common than neurology (12%) or neurosurgery (10%). Responders rate national meetings (39%) as their most influential educational experience. Attending physicians (34%) are more likely than trainees and non-physicians (14%) to consider personal literature review most valuable (p <0.05). 62% of all NCCT members attended last year's NCS annual meeting, compared to SCCM (32%) and regional conferences (35%). NCCT members infrequently attended (<20%) general topic neurological or neurosurgical national conferences.

Despite diverse backgrounds, NCCT members seek continuing medical education through common subspecialty specific methods.

Financial Support: None

The contributions and perceptions of staff regarding nurse practitioners (NPs) and physician assistants (PAs) in neuroscience ICUs throughout the country are not well known. The objectives of this study were to determine the impact of neuroscience NPs and PAs and assess demographics of ICUs.

All members of the Neurocritical Care Society were asked to complete a survey to obtain their perception regarding the addition of NPs and PAs to the ICU team. Participants rated the abilities of NPs and PAs to promote a team environment, anticipate or prevent neurologic deterioration, address patient or staff concerns in a timely manner, safety, and communicate effectively on a 1-5 Likert scale. In addition, members were asked to provide basic demographics and background information on the type and size of ICU, type of providers in charge, and the role of NPs and PAs in their ICU, including procedures performed, documents written, and number of patients per provider. Both quantitative and qualitative data was collected and analyzed. A Mantel-Haenszel Chi Square and ordinal logistic regression model were used to determine the relationship between the background information and the perception of the abilities of NPs and PAs.

The study cohort composed of 15% of NCS members. Additional responsibility of NPs and PAs was associated with higher scores in safety, ability to promote a team environment, address patient or staff concerns, communication, and most importantly the ability to anticipate or prevent a neurologic deterioration (p<0.0001 for all). Number of NPs and PAs, number of years of employment of NPs and PAs, number of procedures, and amount of documentation also positively affected safety.

Additional responsibility of NPs and PAs has strong potential to improve staff, patient, and family satisfaction, safety, and prevent neurologic deterioration. NPs and PAs should be utilized to the full extent of their role.

We conducted a survey study in an academic, co-managed Neuro ICU to explore family satisfaction regarding the care of their surviving loved ones and compared results with concurrent data from the hospital's closed medical ICU (MICU).

Over 38 days, we administered the Family Satisfaction-ICU instrument to Neuro ICU and MICU patients' families at time of ICU discharge. Those whose loved ones passed away during ICU admission were excluded.

The capture rates of families from the Neuro ICU and MICU were 65.3% (79 surveys) and 62.5% (45 surveys 

In our Neuro ICU, patients' families could be more satisfied with several aspects of care. Further study is needed to determine (1) whether a closed Neuro ICU model improves family satisfaction and (2) whether instituting a system in which the neurointensivist team regularly meets with all available families daily improves perceptions of shared decision making, even in routine situations. 

Non-funded prospective patient registries at any given institution rely largely on volunteer clinical personnel. Presupposing that an all-inclusive database would be self-defeating in this type of environment, we designed and implemented a quality improvement (QI) database with intentional iterative design.

Neurointensivists identified by consensus the injury/disease related events and procedures that were most important to track for QI and for judging clinical intensity of our unit. We compiled a list of syndromes that were either commonly studied by principal investigators or were common primary diagnoses in our unit. For each syndrome, we identified commonly accepted grading or intensity scores. The Clinical and Translational Science Awards electronic data entry module, REDCap, facilitated data collection. Consecutive patients in our ICU were entered upon discharge. Weekly meetings served to adjudicate disease classification, grading scores (frequently based on consensus imaging review), and discharge disposition. Opportunities to enter free-text items were allowed to enhance the intentionally iterative design.

In quarterly reviews, we removed items that were consistently left blank and added standardized items corresponding to consistently annotated free-text items.

Since its implementation in January 2011, the Neurocritical Care QI Patient Registry has accrued 2589 separate entries. Consensus-driven iterative changes to the registry have resulted in complete data entry. Participation at weekly registry meetings has been consistent and enthusiastic, routinely drawing 5-6 physicians (2-3 fellows, 2-3 attendings). 3 QI projects have been enabled to date.

Resource limitations may be a practical hindrance to achieving all-inclusive databases outside of funded clinical studies. An iterative design driven by consensus in the described approach can result in a rich database with complete data entry and continued volunteer participation. Future incorporation of supplemental information sources via enterprise-wide clinical data warehouses may achieve more complete databases that comply with standardized ideals such as the Common Data Elements.

Many neurology residency programs have begun implementing mandatory rotations through neurocritical care (NCC) as part of the curriculum. The added experience was thought to be beneficial for residents after graduating the program; however, we wondered how it might affect residents and patients during residency. We thought to survey residents about their programs and the amount of time they spend in NCC rotations. We also wanted to know how they felt the extra time spent in these rotations affected their consulting habits, and therefore their ability to manage cases on their own.

All neurology residents in the United States were the target population for this survey. A list of 126 neurology residency programs was obtained from the American Medical Academy's FREIDA database. The names and email addresses of program directors were generated, and they were contacted by email with a link to an online survey.

The respondents were 101 neurology residents (37 PGY-2, 37 PGY-3 and 27 PGY-4 residents). Of the respondents, 73.7% stated that NCC was a mandatory rotation in their program while the remaining 26.3% said that it was not. Of those who had mandatory NCC rotations, 47.9% said they were for 6-8 weeks duration, while 55.1% agreed they should be 6-8 weeks long. When asked how often they ask for consultations from other specialties, residents who had mandatory rotations through NCC were more likely to say they usually do not consult other specialties, while those who did not have mandatory rotations were more likely to consult for all non-neurological issues.

The survey results demonstrated that neurology residents who have mandatory rotations in NCC are more confident in their abilities to manage their own patients. This is thought to promote continuity of care and may reduce medical errors as well as healthcare cost. 

A botulism epidemic in a maximum-security prison cell-block posed numerous logistical dilemmas for which telemedicine served as a bridge to management.

Inmates in a high-security prison cell-block brewed batches of "pruno" by fermenting fruit, raw potato, and granulated sugar in reusable bags that were passed throughout the cell-block. One of the batches was contaminated with Type A botulism. Twenty-nine inmates were potentially exposed, but the actual exposures were initially indeterminable due to the inmates' fears of incrimination.

The index case developed nausea, emesis, diplopia, and ptosis approximately six hours after exposure and presented to the emergency department (ED) two days later with generalized weakness, dysarthria, dysphagia, hypophonia, and dyspnea. He required intubation and was admitted to the Neurocritical Care Unit (NCCU). Four additional inmates presented with similar symptoms within several hours of the index case. Two required intubation and all were admitted to the NCCU. Within twenty-four hours of admitting the first five cases, nine additional inmates developed symptoms. Five were evaluated in the ED; three were admitted to the NCCU and two were discharged to the prison infirmary and monitored using telemedicine. Two patients were initially evaluated and monitored with telemedicine at the prison. The remainder of the cell-block was evaluated by prison infirmary staff. Botulinum toxin type A was confirmed with bioassay and cultures in these patients, but classic electrodiagnostic findings were absent. The eight inmates admitted were treated with hepavalent botulinum antitoxin (H-BAT). Obtaining the antitoxin required collaboration with the CDC for transport from several sites around the country. Inmates were followed post-discharge using telemedicine and showed improvement.

This botulism epidemic presented a logistical logjam. Initial telemedicine evaluation and subsequent monitoring played a key role in managing NCCU access and optimizing security resources for the prison, ED, and NCCU. 

Intrahospital transport of neurocritical care unit (NCCU) patients is associated with accidental line removal, unplanned extubation, and hemodynamic instability. Further, because patients must be accompanied by a nurse during intrahospital transport, there is an inherent reduction in home unit staffing which reduces direct patient care and monitoring for other NCCU patients. The purpose of this project was to assess the impact of a Neurocritical Care Transport Nurse (NTRN) on patient safety, improved direct patient care time and improved staff satisfaction.

The 3-month NTRN pilot program was initiated in our 16 bed NCCU. For three months, the NTRN worked five 8-hour shifts per week. The NTRN accompanied patients during intrahospital transports, assisted with admissions, functioned as resource nurse in the NCCU, and relieved nurses for meal breaks. Data was collected in real time and included time-inmotion data, adverse event records, and a pre-post work-flow surveys.

The NTRN completed 103 intrahospital transports with were zero safety events. The mean length of time for intrahospital transport prior to the pilot was significantly greater than transport by the NTRN (87 vs. 28 minutes; p<.001). The mean time it took nurses to stabilize a new admission/post-op patients was reduced from 85 minutes to 28 minutes. Staff surveys were overwhelmingly positive with 89% of nurses reporting the NTRN saved them time; 24% reported increased opportunity for meal breaks, and 71% attributed reduced overtime due to the NTRN program. Individual nurses reported that the NTRN program saved them an average of 47.5 minutes each shift (8.7 hours per shift).

The NTRN pilot program was associated with fewer safety events, increased staff satisfaction, more rapid attention to patient needs and reduced overtime. The program should be implemented full time and evaluated for potential costsavings.

Many factors are associated with time delays to reperfusion in endovascular treatment for acute ischemic stroke (AIS).

We assessed if a prototypical neurointensive care unit layout where both the angio suite and CT scanner are inside the unit can reduce times to reperfusion. We compared time from CT to groin puncture (GP) in patients that were transferred from outside hospitals (OSH) directly to the NICU versus those who went through our emergency department (ED).

We retrospectively reviewed 230 patients from a prospectively maintained database from October 2010-June 2012 who underwent endovascular therapy for AIS. A univariate analysis was performed to compare the patients' characteristics between the two populations and to identify differences in time intervals between CT imaging and GP.

A total of 195 patients were included in our analysis. 125 (64%) patients were from OSH. Patient characteristics in both groups were similar except for OSH patients had significantly less history of hypertension (67% vs 81%, p<0.02) but had longer time intervals from last known normal to GP (median 310 mins vs. median 221mins, p <0.01) and lower pretreatment ASPECTS (48% <8 vs. 25% <8, p <0.01). Patients' transferred from OSH had significantly lower times from inhouse CT to GP as compared to patients from the ED (44.2 +/-33mins vs. 114 +/-58mins). Although there was an increased number of non-contiguous multimodal imaging studies performed on ED patients compared to those from OSH (37% vs 5%, p<0.01), exclusion of these patients still resulted in a significant shorter time frame between CT to GP (41.9+/-27 mins vs. 95.8+/-40.3 mins) among OSH transfers.

The design of an integrated biplane angio suite within the NICU reduces the times from CT imaging to GP, thereby lowering the times to reperfusion, and potentially, patient outcome.

Sepsis is a challenge for the Intensive Care Unit (ICU), being the main cause of death during hospitalization.

It was performed a longitudinal and individualized intervention authorized by the HSJA Ethics Committee applying the campaign 'Simple Actions Save Lives' in which 105 educational adhesives worked as a guide for washing hands and flags for high contaminated locations. A decontamination routine of monitors, control panels, ventilators and infusion bombs was established every 12 hours; and continued education for the health team was intensified during the intervention. Two groups were created, patient enrollments in periods of 45 days before and after the intervention, more than 24 hours of hospitalization: group A with 18 patients and group B with 15 patients.

The hospital infection incidence decreased by 40% and VAP by 39.6%. Urine culture was positive in 33,3% of those patients (n=5) in group A and in 16.7% (n=1) in group B (a 50.1% decrease ). The cultures of catheter tip were positive in 68.8% (n=22) of catheters in group A, which used 32 catheter in total, and none in group B, which used 13 catheters. The sepsis incidence decreased by 39.6%. Septic shock was detected in 16.6% (n=3) of patients in group A. There was a drop of the costs between groups (R4,479.28, 10.5%). The cost of campaign material was R$50.00.

This intervention was a simple form to decrease the related number of infections in the Neurovascular ICU, having spent irrelevant values when compared to treatment of these clinical tables.

Intracranial pressure (ICP) management guidelines have been established; however there is no data documenting actual ICP management practices in the United States, or the degree to which clinicians comply with existing guidelines. The primary aim of this study is to explore nursing and medical practice patterns associated with ICP monitoring and management.

A prospective multi-center non-randomized observational design was used.The study sample was composed of consented nurse/patient dyads, with 2 dyads enrolled per study site. Study patient subjects included were over age 18, had ICP monitoring in situ, and were diagnosed with intracranial pathology. Nurse subjects included were those assigned to the patient, who routinely worked in the unit, and had completed their orientation training. Each dyad consented to a 2 hour observational period, where data was collected on nurse interventions for ICP management.

Dyads (N=28) were enrolled at 16 hospitals between August 2009 and May 2012. Patients were primarily male 68%, mean age of 47 years, and non-Hispanic. Nurses were primarily female 90%, non-Hispanic, and a mean of 7.7 years of critical care experience. We observed 11 distinct nursing/medical interventions hypothesized to reduce ICP. Although CSF diversion and limiting stimulation were the most frequently used interventions, there was not a consistent hierarchical approach to initiating ICP reduction interventions. Wide variances in nursing and medical treatment patterns were observed for ICP treatment threshold, first-line therapy, and the order in which interventions were initiated.

Despite established guidelines, variability exists throughout the U.S. in how physicians and nurses monitor and manage ICP. More research is needed to compare intervention techniques to determine the impact these differences have on outcomes in patients requiring ICP management.

Administrative data are being increasing being used to measure quality of care, for public reporting, and in pay for performance. Administrative data are inexpensive, readily available, and target clinical outcomes.

The aim of this study was to evaluate the use the use of administrative data in identifying potentially preventable events and iatrogenic complications in patients admitted to an academic medical center with a primary diagnosis of acute stroke. Administrative data for all adults patients (> 18 years of age) with a discharge diagnosis of stroke 431, 433.01, 433.10, 433.11, 433.21, 433.31, 433.81, 433.91, 434.00, 434.01, 434.11, 434.91, and 436) were evaluated from January 2010-December 2011 for complications based on administrative data by looking at secondary diagnoses that were not present on admission using data from the University Healthsystem Consortium Database. Both the Agency for Healthcare Research and Quality (AHRQ) Quality Indicators (QIs) for inpatient conditions and known codes for other complications such as catheter associated urinary tract infection, pneumonia, and deep vein thrombosis or pulmonary embolus (not associated with surgery) were used to identify potential events.

Of the 600 cases reviewed, many cases had at least one complication. The leading cause of potentially preventable events were related-to-infection (central line associated bloodstream infection, sepsis, catheter associated urinary tract infection, and aspiration and/or hospital-acquired pneumonia). The AHRQ QI only captured a fraction of the events. Patients with subarachnoid hemorrhage had the highest mortality, followed by intracranial hemorrhage. Several of the deaths occurred in low risk patients and had at least one potentially preventable complication. When reviewing flagged records, a small number of events reflected opportunities to improve documentation and/or coding, with the majority of remaining events associated with opportunities for improvement.

Administrative data may be a useful adjunct to quality improvement efforts.

Financial Support: Co-deputy project lead for the AHRQ QI project (AHRQ sub-contractor).

Continuous video-EEG (cVEEG) monitoring is often utilized in the evaluation of impaired consciousness. Nonconvulsive seizures may be distinguished from metabolic disorders when triphasic waves (TW) are recorded. However, rhythmicity detected on cVEEG may call into question the presence of electrographic seizures. The following case describes the transient resolution of rhythmic TWs after acute administration of glucose in a patient with hypoglycemic encephalopathy.

Case report reviewing clinical, laboratory and electroencephalographic features of a patient with metabolic encephalopathy secondary to hypoglycemia.

A 66-year-old woman with type-1-diabetes and treated osteomyelitis of the foot presented with altered mental status. She was intubated and stuporous. Cranial nerves were intact. All four extremities withdrew to noxious stimulation. Plantar responses were flexor. MRI brain exhibited leptomeningeal enhancement consistent with meningitis. Serum BG=392mg/d, and CSF glucose=16mg/dL. 1 month after antibiotic treatment, she was following commands. Repeat MRI revealed complete resolution of leptomeningeal enhancement. During recovery, she developed sudden onset stupor with left facial movements, and underwent cVEEG monitoring. EEG showed generalized, polymorphic delta/theta slowing intermixed with TWs, without electrographic correlate of facial movements. During cVEEG, TW activity increased in rhythmicity and frequency, coinciding with worsening hypoglycemia, with a nadir BG=14mg/dL. Electrographic activity was not induced or exacerbated by stimulation. Administration of 50 ml of D50 (25 g D-glucose) resulted in transient resolution of TWs within minutes, which corresponded to a BG=64mg/dL. However, background slowing remained on cVEEG, with gradual reemergence of infrequently occurring TWs despite normoglycemia. Mental status returned to baseline approximately 72 hours after BG stabilization.

Rhythmic triphasic wave activity due to hypoglycemia may be distinguished from electrographic seizures after acute correction of BG, with corresponding transient resolution of triphasic waves. However, clinical response to correction of metabolic dysfunction may be delayed for up to 72 hours.

Continuous-IV-midazolam (cIV-MDZ) is recommended for treatment of refractory status epilepticus (RSE) but doses are controversial. Here we compare a historical cohort (N=29) treated with low dose to a subsequent cohort of 100 patients treated with high dose cIV-MDZ for RSE.

Following the analysis of the historical cohort (1996 -2000 Neurology 2001,57(6) :1036-1042) we changed our protocol for RSE allowing for higher cIV-MDZ doses and collected 100 consecutive cases (2002) (2003) (2004) (2005) (2006) (2007) (2008) (2009) (2010) (2011) . Exclusion criteria: cardiac arrest; prior treatment with a different cIV-AED. We collected data on baseline characteristics, cIV-MDZ doses, seizure control, complications, hospital course, and outcome. High dose was compared to low dose cIV-MDZ on an intention to treat basis using logistic regression analysis with the significance level set at P<0.05.

Baseline characteristics were similar between groups. Median maximum cIV-MDZ dose was 0.4 mg/kg/h (IQR 0.2, 1) for the high and 0.2 mg/kg/h (IQR 0.1, 0.3) for the low dose group (p<0.001), but duration of cIV-MDZ was the same between the two groups (median 72 hours). "Withdrawal seizures" (within 48 hours after cIV-MDZ discontinuation) were less frequent in the high dose group (15% vs 64%, OR 0.10; 95%-CI 0.03-0.27). "Breakthrough seizures", "ultimate cIV-MDZ failure", and complications were not different. Discharge mortality was lower in the high dose group (40% vs 62%, OR 0.30, 95%-CI 0.11-0.81) after controlling for age, etiology, and APACHE-2 scores. At 3 months, mortality was similar between the two groups. Lower death or vegetative state rate for those treated with high dose cIV-MDZ was seen at 3 months (74% vs 96%; OR 0.14; 95%-CI 0.02-1.07), but this finding is limited by missing 3-month functional outcome data in both groups.

High dose cIV-MDZ treatment for RSE can be performed safely in an ICU setting and may be more efficacious in controlling seizure activity. Outcome data are promising and warrant further prospective study.

The clinical utility of free valproic acid (VPA) levels is unclear, and the actual free fraction (FF) of VPA in hospitalized patients is not well established. Our goal was to assess and compare the total levels, free levels, and FF of VPA in inpatients and outpatients and to determine factors that may influence the FF.

Retrospective chart review of 437 paired total and free VPA levels in 220 inpatients and 41 outpatients. Demographical, laboratory, and concomitant interacting medication data were collected and analyzed. Paired total and free levels were categorized based on their status in regards to the therapeutic range (i.e., subtherapeutic, therapeutic, or supratherapeutic) and whether the paired levels were concordant or discordant (e.g., both levels in therapeutic range, or mismatched). Linear regression was used to assess the impact of variables on the FF. Logistic regression was used to determine if variables predicted the likelihood of having discordant paired levels.

Inpatients had a significantly higher median FF compared to outpatients (31.1% vs. 15.5% respectively; p <0.0001). Total levels were found to be a poor predictor of free levels (R 2 = 0.34) in hospitalized patients. Inpatient free levels were discordant with the therapeutic status indicated by the total level 68% of the time. In a linear regression model, albumin (p <0.0001), total protein (p <0.0001) and co-administration of phenytoin (p=0.003) and carbapenems (p=0.001) were found to significantly and independently impact the FF. Multiple logistic regression indicated albumin as a significant predictor of the total and free levels being discordant in regards to therapeutic status (OR 0.13 [95%CI 0.07-0.25], p<0.001).

Inpatients had a significantly higher FF compared to outpatients. Inpatient free levels were frequently discordant with the total levels in terms of the therapeutic status. Decreasing albumin was a significant predictor of discordance between the free and total levels.

Increased continuous EEG utilization in the ICU has generated an interest in faster acquisition and interpretation of EEG data. Limited electrode arrays (LEA) coupled with quantitative algorithms have been leveraged for this purpose. However, previous studies with LEA's have suggested an inherent error rate produced by a reduced number of electrodes. The aim of the current project was to test a novel LEA and determine if multiple montages could correct any error rate.

With approval from our IRB, 250 short de-identified EEG segments were retrospectively collected from clinical cEEG archives. Segments contained one of five primary findings: normal, diffuse slowing, periodic epileptiform discharges (PEDs), seizure and burst suppression. All files were reformatted into an 8 electrode array containing a lateral chain and central electrode bilaterally. Segments were distributed to four experienced neurophysiologists in two phases. In phase 1, segments were interpreted in a single anterior-posterior bipolar montage and compared to the original read. In phase 2, fifty frequently misread segments from phase 1 were reinterpreted using four additional montages.

In phase 1, 1000 EEG interpretations were reviewed yielding a sensitivity of 73% for seizure and 70% for PEDs, burst suppression, and normal. The specificity was greater than 90% in all cases. The sensitivity and specificity for diffuse slowing was 90% and 75%. In phase 2, 150 EEG interpretations were collected with no significant improvement noted in the detection of any EEG finding.

In agreement with past studies, this trial suggests that LEAs contain a base error rate engendered by the reduced number of electrodes. This error rate is maintained regardless of the number of available montages. The implication of these findings suggests that studies examining the use of LEA's for use in seizure detection and neurophysiologic algorithms should calculate an error rate specific to the electrode array before algorithm testing.

The incidence of nonconvulsive status epilepticus (NCSE) and other electrographic features in comatose post-cardiac arrest syndrome patients treated with therapeutic hypothermia (TH) is still under investigation. The objective of this study is to determine the incidence of NCSE and other electrographic features and correlate with neurologic outcome and survival.

Review of consecutive subjects treated with TH and receiving continuous EEG (cEEG) monitoring between May 2009 and December 2010. Demographic data, survival, and functional outcomes using cerebral performance category (CPC) scale were prospectively recorded.

Forty eight patients were included, with mean age of 63 years (SD 15), majority were males (n=26, 54%) and experienced out-of-hospital cardiac arrest (n=28, 58%). Ventricular fibrillation was the initial cardiac rhythm in 20 patients (42%). All patients received TH. Twenty seven patients (56%) died. Seventeen patients (35.4%) had good neurologic outcome (CPC 1 or 2). NCSE occurred in 2 patients (4.2%), both of whom died. Periodic epileptiform discharges occurred in 6 patients (12.5%), 5 (83%) of whom had poor neurologic outcome or death (CPC 3-5) compared to 62% poor outcome in whom periodic epileptiform discharges did not occur (nonsignificant). Burst suppression occurred in 15 patients (31.2%), all (100%) of whom had poor neurologic outcome or death compared to 48.5% poor outcome in whom burst suppression did not occur (p <0.05); and severe background attenuation occurred in 10 patients (21%), 9 (90%) of whom had poor neurologic outcome or death compared to 58% poor outcome in whom severe background attenuation did not occur (p<0.05).

NCSE occurred in 4.2% of post-cardiac arrest patients undergoing therapeutic hypothermia. Outcomes are poor in postcardiac arrest patients undergoing therapeutic hypothermia with NCSE, burst suppression or severe background attenuation. Larger prospective studies are needed to further evaluate and characterize cEEG findings in comatose postcardiac arrest patients undergoing TH.

Encephalopathy is a frequent occurrence in the critical care setting. Previously, we have shown that patients with a primary neurologic injury and encephalopathy are at high risk for cEEG seizures. Patients with a presumed metabolic etiology of encephalopathy have been poorly characterized. The purpose of this study was to identify the frequency and underlying etiology of cEEG seizures that occur in critically-ill patients with a presumed metabolic etiology.

We retrospectively reviewed prospectively collected cEEG and clinical data on consecutive patients monitored from January 1, 2007, to December 31, 2011. We identified those patients with cEEG seizures (n=570) and included in this study only those patients with metabolic etiologies. EEG seizures were defined as evolving rhythms in frequency, distribution, and/or morphology at 2 Hz or greater for more than 10 seconds duration. Statistical analyses were performed with JMP 9.0.

Sixty-six (11.6%) patients were identified as having metabolic causes for cEEG seizures with the most common etiology being sepsis (50.0%) which linearly increased (R 2 =0.64) in detection from 3 in 2007 to 17 in 2011. Other etiologies included liver failure (15.2%), posterior reversible encephalopathy syndrome (PRES; 13.6%), electrolyte/glucose derangement (10.6%), drug overdose/withdrawal (9.1%), and renal failure (1.5%). 80.3% of the cEEG seizures were without clinical signs. A linear increase in cEEG seizures occurred with a decrease in level of consciousness (R 2 =0.98). The majority (56.1%) of patients were eventually discharged for rehabilitation, but 36.4% expired prior to discharge.

This retrospective study shows an increase of cEEG detected seizures in patients with a presumed metabolic etiology from 2007 to 2011. This increase in seizures is likely due to increased targeted monitoring. This highlights the value of using cEEG database information to target at risk populations. Our results should guide the use of cEEG monitoring in the metabolic patient particularly those with septic encephalopathy.

Medically induced burst suppression on EEG is often seen in critically ill patients who are sedated for treatment of status epilepticus, cerebral edema, and in patients with anoxic brain injury or post cardiac arrest undergoing hypothermia treatment. Previous studies have demonstrated that the majority of these patients have poor prognosis. We decided to investigate if specific EEG patterns during burst suppression in these patients would correlate with different outcomes.

We retrospectively identified 40 patients with medically induced burst suppression out of 732 patients who had continuous EEG monitoring (cEEG) from January 2010 through December 2011 in our Neuro ICU. Neonates and children were excluded. All EEG tracings were independently reviewed by two electroencephalographers and classified into discrete seizures, status epilepticus (SE), interictal epileptiform discharges (IED), burst suppression, and epileptic bursts defined as burst suppression with IED within the burst activity. Primary outcome was Cerebral Performance Categories (CPC) at hospital discharge.

Of the 40 patients, 20 were identified to have epileptic bursts-one SE, eight anoxic brain injury, two ischemic stroke, 4 hemorrhagic stroke, five other medical conditions. The mortality rate of patients with epileptic bursts was 70% compared to 55% for those without. Only 10% in each group had good neurologic recovery defined as CPC score of 1-2. Patients with epileptic bursts on average had longer duration of monitoring (7 days versus 4) due to refractory seizures and, subsequently, increased number of AEDs (2.9 vs 1.8) used.

Similar to previous findings, the patients in our study had poor prognosis. Our findings additionally show that epileptic bursts in this patient population correlated with more refractory seizures and a higher mortality rate. The presence of epileptic bursts may be used as an adjunctive indicator for prognosis in patients who are in medically induced burst suppression. Larger population study is underway.

Epidemiologic studies in epilepsy using large administrative databases depend on accurate ICD-9-CM classification. We sought to determine the accuracy of ICD-9-CM code 345.3 (grand-mal status) for diagnosing status epilepticus (SE) after hospital admission.

A case-control study at an academic institution was conducted. Twenty-one subjects with discharge ICD-9-CM code 345.3 (grand-mal status) and 42 consecutive admissions without the code of interest were randomly selected. SE was defined as neurologist documentation of continuous clinical seizure activity for five minutes or longer and/or two or more discrete clinical seizures without inter-ictal return to baseline (clinical diagnosis) and/or EEG consistent with SE by board certified neurophysiologist interpretation (EEG diagnosis).

All 21 cases and none of the controls met our pre-defined criteria of SE. Therefore, the sensitivity, specificity, positive predictive value (PPV), negative predictive value, and accuracy of the code was 100%. When the diagnosis relied on clinical criteria alone, the sensitivity decreased to 82% with PPV of 100%. When the diagnosis was made by EEG criteria alone the sensitivity decreased to 55% and PPV of 100%.

The ICD-9-CM code 345.3 is both accurate and specific for the diagnosis of SE after admission at an academic institution. Clinical definitions of SE and the prevalence of the disease may affect the sensitivity and PPV of ICD-9-CM code 345.3 for the diagnosis of SE. The results of our study require further validation in other cohorts.

Refractory status epilepticus (SE) has been linked to significant morbidity and mortality. When pharmacological treatment fails, ketogenic diet has shown to suppress seizure activity in children and is gaining acceptance as an adjunctive treatment in adults. While case reports exist, there are no standard guidelines for implementing ketogenic diet in adult neurocritical care patients. The purpose of this abstract is to demonstrate a standard guideline for ketogenic diet utilization in a neurocritical care unit.

A performance improvement project was undertaken to standardize ketogenic diet administration in enterally fed neurocritical care patients with medically refractory SE. The guidelines include patient selection, team communication, patient monitoring, family education, patient transitions out of intensive care and measures for patient outcome from this treatment. Patients admitted with SE are initiated with standard pharmacologic treatment; if treatment does not result in SE cessation, then ketogenic diet is reviewed. Discussion with the patient's family is required to review long-term implications and potential lifestyle choices related to diet after critical illness. A standard checklist within the guidelines assures communication to all necessary organizational departments including appropriate consults. Daily monitoring and discussion in patient rounds evaluates daily patient progress. Team communication is focused on diet tolerance, medication carbohydrate content, concurrent pharmacologic SE management and patient progression.

Since 2008, after implementation of the standardized guideline, seven adult SE patients have been treated with ketogenic diet. Mean age was 43.3 years; range 26-69 years; two patients were male. Ketosis was achieved in six of seven patients and five of six patients sustained resolution of SE after ketosis was achieved.

Our organizational experience indicates that coordinated team care, family education, goal planning and a standardized guideline contribute to successful implementation of ketogenic diet. Further research is needed to determine overall effectiveness of this therapy.

Status epilepticus (SE) is a potentially life-threatening condition that is frequently under-recognized, may be refractory to initial treatments, and often requires admission to general intensive care units (ICUs) We hypothesized that admission of patients with SE to the Neurosciences ICU (NICU) vs the Medical ICU (MICU) might correlate with surrogates for improved patient outcome.

We performed a single-center, retrospective cohort study of patients with SE admitted to the NICU vs the MICU in our institution between 2005-08. Admission to either ICU depended on bed availability and Emergency Medicine preference. Clustering methods were used for analyses, taking into account multiple visits of the same patient.

There were 168 visits for 151 patients with definite or probable SE [46 (27%) in the NICU and 122 (73%) in the MICU]. APACHE II scores were significant higher in the MICU group (17.5 vs 13.4, p=0.003). More continuous EEGs were ordered in the NICU (85% vs 30%, p<0.001). cEEG was ordered more frequently in complex partial/non-convulsive and less in convulsive clinical presentations. The NICU had a higher rate of complex partial/non-convulsive SE and the MICU of generalized convulsive SE (41% vs 21% and 57% vs 76%, p<0.033). Admission diagnoses differed, with the NICU having a higher rate of stroke and the MICU a higher rate of toxometabolic etiologies (39% vs 12% and 11% vs 21%, p<0.002). After adjusting for covariates, no difference was found in the ICU or hospital length-of-stay and modified Rankin Scale at discharge.

Management differences occurred in MICU vs. NICU-managed SE, possibly based on variabilities in presentation and etiology. However, no reduction in length-of-stay or different discharge outcomes between the ICUs was found.

Hongki Song 1 , Taechon Kang 1 , Dongjin Shin

Although levetiracetam(LEV, 2S-(oxo-1-pyrrolidinyl)butanamide, Keppra®, UCB Pharma) has been reported to be well tolerated and effective in SE refractory to benzodiazepine (BDZ), there was little preclinical or clinical data concerning the outcomes of LEV in comparison to DZP, and VPA in SE-induced neuronal death.

To address this relevant lack of information, we have performed the preclinical study to investigate the effect of diazepam (DZP), valproate (VPA), and LEV alone, and the efficacy of LEV as an add-on treatment with DZP on the SE-induced neuronal death.

DZP and VPA. However, it is noticeable that LEV as an add-on drug with DZP could not alleviateSE-induced neuronal damage as compared to effective to protect neuronal damages from SE, as compared to DZP. In contrast to LEV, VPA(50 and 100 mg/kg) as an add-on drug with DZP significantly reduced SE-induced neuronal damage as compared to DZP alone, and showed the similar effect of VPA (150 mg/kg) alone.

These findings indicate that, unlike VPA, LEV may negatively interact with DZP, and suggest that LEV may be more effective to prevent SE-induced neuronal death as a first line drug than as a second line therapy after BDZ treatment, and that LEV as an add-on drug with BDZ may not provide any additional benefit to outcome of SE.

Temkin and colleagues found that phenytoin exerted a beneficial effect by decreasing the rate of seizures by 73% during the first week after a traumatic brain injury. The purpose of this study was to determine the need for monitoring and titrating to therapeutic free phenytoin levels in patients receiving phenytoin for prophylaxis within 7 days following a traumatic brain injury.

This was a retrospective study of 93 patients for a traumatic brain injury (TBI), who met the inclusion criteria and received phenytoin for seizure prophylaxis for 7 days following injury. Eligible patients were divided to two arms: patients with phenytoin levels (n=34) and patients without levels (n=59). The primary outcome measure was the incidence of seizures in those that were monitored for free phenytoin levels and those that were not monitored for free phenytoin levels. The secondary outcome measure was the appropriateness of phenytoin dosing in regards to initial loading and maintenance dose.

A total of 2 seizures occurred in the entire study population. Both seizures transpired in patients with phenytoin levels. Patient 1 was diagnosed with a seizure event on Day 2, with free phenytoin obtained on Day 3 at a therapeutic level of 1.85 mg/L. Patient 2 had a witnessed seizure on Day 6, with free phenytoin level obtained on Day 5 also within therapeutic range at 1.61 mg/L. There was no incidence of seizure in patients who were not monitored for phenytoin levels. Inconsistent phenytoin loading and maintenance doses were identified.

This study suggests that monitoring phenytoin to therapeutic levels for seizure prophylaxis did not demonstrate a decrease in the occurrence of seizures. We are unable to make recommendations given the inherent limitations of our study. A large prospective, randomized trial is needed to clarify the need for monitoring phenytoin to therapeutic levels.

Seizure prophylaxis for nontraumatic intracerebral hemorrhage (ICH) and aneurysmal subarachnoid hemorrhage (SAH) is common practice in the intensive care unit(ICU). Typical antiepileptics include phenytoin (PTN) and levetiracetam (LVT). Previou studies have suggested worse long term outcomes with ICU PTN use, but such data is lacking for LVT. In addtion, few studies have compared LVT to PTN for seizure prophylaxis in ICH or SAH patient in the ICU setting. We hypothesize that seizure prophylaxis with LVT, as compared to PTN, for patients admitted with ICH and SAH will result in similar outcomes at hospital discharge as measured by the modified Rankin scale (mRS).

This study is a single center retrospective review from 2009-2012, to ultimately include approximately 250 adult patients with the diagnosis of SAH or ICH who received seizure prophylaxis with either LVT or PTN. Basic demographic, past medical history, severity of illness scales; length of mechanical, ICU and hosital length of stay; seizure occurrence, use of continuous electroencephalogram, data will be collected, in addition to other variables. Patients with prior seizure history or seizure on presentation, do-not-resuscitate hours within 24 hours of ICU admission, will be excluded.

To date, our analysis includes 42 patients (LVT =31 and PTN =11). Comparing PTN to LTR, univariate analysis of demographics, baseline clinical characteristics and outcomes were similar between the two groups (all p>0.05).

In our initial univariate analysis, functional outcome at discharge was similar between PTN and LVT when used for seizure prophylaxis in patients admitted with ICH or SAH. Subsequent analysis will include additional patients (approximately 250) with multivariate adjustment.

Cerebral microbleeds (CMBs) are commonly found in patients with microvascular pathology such as primary intracerebral hemorrhage, cerebral amyloid angiopathy, and ischemic stroke. However, to our knowledge, there have been no reports of CMBs or their acute appearance in patients with status epilepticus (SE).

Here we describe two patients admitted to our neuro-intensive care unit with generalized tonic-clonic seizures. Laboratory tests were unremarkable except for mild pleoc onset and did not showed abnormal findings. Seizures continued despite multiple anti-epileptic drugs including phenytoin, valproic acid, topiramate, clonazepam, pregabalin, lacosamide, phenobarbital, levetiracetam, and continuous infusion of propofol, ketamine and midazolam (up to 1.4 mg/kg/hr in the first patient and 2.9 mg/kg/hr in the second patient). Followup 3.0-tesla susceptibility-weighted imaging revealed 63 new CMBs (44 lobar [9 frontal, 12 parietal, 19 temporal, 2 occipital, and 2 insular], 13 deep [3 corpus callosum and 10 deep/periventricular white matter], and 6 infratentorial [5 brainstem and 1 cerebellum]) in the first patient (performed 29 days after initial imaging) and 14 new CMBs (11 lobar [3 frontal, 4 parietal, 2 temporal, and 2 occipital], and 3 deep [1 corpus callosum and 2 deep/periventricular white matter]) in the second patient (performed 41 days after initial imaging). Multimodal neuromonitoring was available between initial and follow-up imaging in the second patient and suggested metabolic distress (lactate-pyruvate ratio >40), cerebrovascular dysautoregulation (pressure reactivity index >0.2), brain tissue hypoxia (brain tissue oxygen partial pressure <15 mmHg), and fluctuations of blood pressure (variance, 163 mmHg) and cerebral perfusion pressure (variance, 180 mmHg).

CMBs may develop acutely in patients with refractory SE, which may point towards microvascular disturbances in refractory seizures. Further prospective studies are necessary to explore the pathophysiology and clinical implications of new CMBs in SE.

Synthetic cannabanoids, often sold as "spice" and various other labels, are a popular product sold in incense shops and through the internet. When inhaled, consumers often report experiences similar to marijuana use, and have thus become a popular street substitute for marijuana. Unfortunately, with increasing use, there has been an increase in the number of patients presenting to emergency departments due to toxic effects of these products.

We describe a 24 year old gentleman with history of bipolar disorder but no history of neurological disease who presented to the emergency department with altered mental status and tachycardia who subsequently had a witnessed tonic-clonic seizure. Patient received appropriate workup for his potential toxicity. We also performed a literature search on "spice" incense found in his backpack on presentation.

Patient had admitted to smoking "spice" incense on questioning. Patient's negative drug screen, negative workup, as well as symptomatic improvement on phenytoin supported the source of his seizure as the toxic effect of inhaled "spice". We also on literature review discovered several other cases similar to this patient's case.

"Spice" or synthetic cannabanoid-induced toxicity is an emerging etiology of new-onset seizure and does not appear on conventional drug screens. Critical care professionals should be aware of this product to recognize and appropriately treat this toxicity.

Refractory status epilepticus (RSE) is associated with high morbidity and mortality. Etiological heterogeneity and refractoriness to treatment remain a challenge for the treating intensivist. Here we present a patient with RSE and folic acid (FA) deficiency.

Brain metabolism was hourly analyzed using cerebral microdialysis (CMA600-analyzer; CMA71-catheter). FA concentrations of brain extracellular microdialysate (FAMD-EC) and serum (FAserum) were analyzed using ElecsysFolateIII® -assay. In vitro recoveryof FA was calculated using cerebrospinal fluid (CSF).

A 47-year-old male was referred to our neurocritical care unit with SE refractory to levetiracetam (3g/d) valproic-acid (1.5g/d) and 0,7mg/kg BW/h midazolam continuous infusion. The patient had a history of short bowel syndrome (SBS) after small intestine resection five months prior. Admission electroencephalography showed continuous rhythmic epileptiform activity over the right hemisphere despite adding ketamine continuous infusion (0,6mg/kg BW/h) and lacosamide (400mg/d). Neuroimaging demonstrated diffusion-weighted-imaging (DWI)-hyperintensities over the right hemisphere. CSF was normal, common causes of RSE were unlikely after extensive laboratory and CSF studies. FA Serum was found to be lower (3.4μg/l; 4.6-18.7 μg/l) at day two of RSE. After thiopental anesthesia (48 hours) and parenteral FA substitution (10mg/d), the patient was successfully weaned without electrographic or clinical seizures. Repeated imaging of the brain at day 10 showed improvement of DWI-hyperintensities. Glutamate levels in MD EC decreased overtime. The patient could be extubated and fully recovered to the functional level before RSE. FA serum increased by 241% to 11.6 μg/l, post hoc analysis of FA MD-EC revealed an increase by 77% (from 2.2 μg/l to 3.9 μg/l). In vitro recovery of FA was 21%, therefore calculated FA brain / FA serum ratio was initially 3, which is comparable to previous animal studies.

Brain extracellular folic acid can be measured using cerebral microdialysis. Although causality cannot be proven, FAdeficiency may have influenced the course of RSE in our patient. 

The management of inter-ictal EEG patterns such as SIRPIDs (stimulus-induced Rhythmic, Periodic or Ictal Discharges) in comatose intensive care unit (ICU) patients remains poorly understood whether these are secondarily injurious to brain or simply a of marker of underlying brain injury. We describe 2 cases of brain-injured patients with SIRPIDs with Ictal SPECT imaging and in regards to aggressive NeuroICU management and patient outcomes.

Case Series, N=2.

Case #1-A 67-year old female suffered a cardiac arrest and remained comatose after 3 days. Continuous ICU EEG demonstrated nonconvulsive seizures (NCSz) and status (NCSE) with up to 3Hz maximal bilateral centroparietal head spike and wave by day #6 which was refractory to initial IV levetiracetam, IV lacosamide, IV phenytoin but finally responded to IV phenobarbital load (30mg/kg) and propofol infusion. SIRPIDS were noted despite these medications with any form of tactile or auditory stimulation. We performed ictal (stimulation provoked SIRPIDs) and interictal Technetium-99-SPECT which was negative for hyperintense focus. Case #2-A 67 year old female was admitted comatose for subarachnoid hemorrhage secondary to aneurysm rupture. She received a left-sided hemicraniectomy with operative clipping of the aneurysm and drainage of a small left subdural hematoma. On postoperative day (POD) #1, cEEg showed left frontotemporal sharp waves. She was placed on leviteracetam, lacosamide, benzodiazepine, propofol infusion, and phenytion. By POD #6, cEEG revealed left frontal sharply countoured discharges when the patient was stimulated by nail bed pressure on examination, consistent with SIRPIDS. By POD #15 an ictal SPECT scan showed broad areas of hypoperfusion in the left hemisphere due to infarcts but there were no findings suggestive of a seizure focus scintigraphically.

SPECT-scan negative SIRPIDs may be helpful in terms of deescalating aggressive brain-metabolic suppressive therapies such as propofol and barbiturates, but larger, outcome-based studies are needed. 

Thromboelastography (TEG) is point-of-care test that allows for rapid global assessment of coagulation. TEG analyzes whole blood, not plasma, which better accounts for the effects of cellular components on hemostasis. We sought to determine whether there is evidence of hypercoagulability by TEG and whether it correlates with discharge outcome after aneurysmal subarachnoid hemorrhage.

Ten patients with moderate-to-severe SAH were prospectively enrolled in an IRB-approved observational study of serial thromboelastography. TEG analysis, using kaolin activated citrated samples, was performed on post-bleed days 1, 3, 5, 7 and 10. Thrombus velocity curves, including the maximal rate of thrombin generation (MRTG), time to maximal rate of thrombin generation (TMRTG), and total thrombin generation (TTG), were plotted for each patient. A hypercoagulable state was defined a priori as a G value of >11 dynes/cm2 or a maximum amplitude (MA) of greater than 70mm. Secondary outcome measures included discharge disposition.

Mean age of patients was 59.1+/-11.6 years. 7/10 patients were women and 4/10 were discharged home. The mean G parameter was within the normal range (9.59 dynes/cm2) on day 1, demonstrated a hypercoagulable profile on day 3 (11.3 dynes/cm2), peaked on day 5 (13.8 dynes/cm2), remained hypercoagulable on days 7 (12.3 dynes/cm2) and day 10 (12.0). The day 5 G value was significantly different from the day 1 value (p=0.04 

Thromboelastography may identify a transient hypercoagulable state that peaks around post-bleed day 5 in patients with SAH. This state reflects accelerated thrombin generation and correlates with discharge disposition. Defining a hypercoagulable state in patients with SAH may lead to better risk stratification and novel therapeutic interventions.

Financial Support: This study is supported in kind by Haemonetics. They supply 2 TEG machines, kits and reagents. They have neither participated in study design nor are they aware of these preliminary results.

Intravenous sedation has been associated with impaired cognitive recovery following critical illness but its influence on recovery following aSAH remains unknown.

Data from consecutive patients with aSAH admitted to Columbia-Presbyterian hospital and enrolled into the SHOP database between 08/1996-1/2009 were analyzed after exclusion of deaths and unemployment prior to hemorrhage. Employment status at 1 year was obtained through self report or through patient surrogate and trichotomized (same level, decreased level, unemployed). Proportional odds models were used to test the association between the use of continuous intravenous sedation with employment and cognitive function at 1 year after controlling for baseline demographics (age, race, occupational level, admission Hunt Hess grade) and hospital complications (pneumonia, infarction from vasospasm). Proportional hazards model was used to examine the association of sedation with time to return to work.

417 patients who had the primary outcome data of employment status at 1 year were analyzed. In multivariate analysis, exposure to continuous intravenous sedation was significantly associated with worse employment status at one year (OR=0.46, CI=0.28-0.78, p=0.004). Poor judgment (OR=0.51, CI=0.27,-0.94, p=0.03) and apathy (OR=0.34, CI=0.21-0.55, p<0.0001) at one year were significantly associated with worse employment status but not with sedation exposure.

With multivariate proportional hazards model, sedation was a significant risk factor of unemployment (HR=0.59, CI=0.38-0.91, p=0.02). Among those who returned to work within 1 year, patients who received intravenous sedation returned to work significantly later than those who did not (median 107 vs. 90days, p=0.04).

Patients who received continuous intravenous sedation following aSAH had worse one year employment status and returned to work later. Although poor judgment and apathy was associated with worse employment status, they were not associated with sedation exposure. Future studies should investigate the effects of intravenous sedation exposure on cognitive and functional recovery following brain injury.

Despite an improvement in mortality, many survivors of aSAH still have significant disability and impairment in quality of life. We investigated predictors of unemployment at 1 year among survivors of aSAH.

Data from consecutive patients with aSAH admitted to Columbia-Presbyterian hospital enrolled into the SHOP database between 08/1996-01/2009 were analyzed after exclusion of deaths and unemployment prior to hemorrhage. Employment status at 1 year was obtained through self-report or through patient surrogate and trichotomized (same level, decreased level, unemployed).Pre-morbid occupational level was trichotomized (full time, part time, housewife). Proportional odds models were used to test the association between baseline demographics, pre-morbid and discharge functional status with employment status at one year. Proportional hazards model was used to test the association of these factors with time to return to work.

A total of 417 patients had the primary outcome data of employment status at 1 year. 192 patients (47%) remained unemployed, 63 patients (15%) worked at a decreased level, while 162 patients (38%) were employed at the same level. After controlling for age, modified fisher scale, and discharge functional status, ethnicity (p=0.0001) and pre-morbid occupational level (p<0.0001) were significantly related to employment status. Hispanics (OR=0.32, CI=0.19-0.53) were less likely to be employed than other minority groups with Caucasian as the reference group. Caucasians working full time pre-morbidly provided the greatest odds for employment (OR=5.35, CI=2.69-10.64) over part time employees (OR=2.12, CI=1-4.5) and housewives (reference) Among those who returned to work at 1 year follow-up, patients who were employed at the same level returned to work sooner that those employed at a decreased level (median: 90 vs. 149 days, p=0.02).

Unemployment among survivors of aSAH remains problematic, especially among certain underrepresented minorities. Future studies should investigate modifiable factors which impede successful reintegration to the work force.

Cerebral vasospasm after aneurysmal subarachnoid hemorrhage (SAH) remains a major cause of death and disability. Delayed cerebral ischemia (DCI) after SAH is likely multi-factorial, but eventually leads to altered cerebral blood flow (CBF) and cerebral infarction. Neurointerventional treatment is used for medically refractory vasospasm, but with limited data on efficacy and impact on CBF and clinical/DCI outcomes.

Patients with SAH scheduled for neurointerventional treatment of refractory vasosasm were consented for intraprocedural CBF monitoring. We measured regional CBF using two sodium iodide scintillation scalp detectors approximating the cortical vascular territory of the treated vessel. A 1.0 mL saline bolus of 1-2 mCi of 133-Xe is injected through the coaxial catheter immediately before and after endovascular treatment. Tracer washout is recorded under stable physiologic conditions for 1.5 minutes. CBF is calculated using the initial slope index, the monoexponential slope of tracer washout from 20-80 seconds after isotope injection. Data were analyzed including standard corrections for remaining activity and physiologic parameters (Cortexplorer CBF2A, Ceretronix, Denmark). Mean arterial blood pressure, PaCO2, serum hemoglobin, and delivery of anesthetic agents were monitored. We calculated change in CBF expressed as a mean + standard deviation using repeated measures ANOVA before and after endovascular treatment.

A total of 23 SAH patients with refractory vasospasm were enrolled in the study. Moderate to severe angiographic spasm was reported in 83% of subjects. Treatment included IA verapamil in 17(74%), angioplasty only in 3(13%), and both in 3(13%). Mean change in CBF was 9 + 11 ml/100gm/min, an average of 26% change in regional CBF.

In our prospective study of 23 patients with endovascular treatment for refractory vasospasm, we detected a mean change of 26% in quantitative CBF using the intra-arterial 133-Xe washout method. Without significant radiographic evidence of large vessel change at the time of measurement, increases in CBF may be related to the microcirculatory effects of treatment.

Early detection of cerebral Vasospasm (VS), a common complication of Subarachnoid Hemorrhage (SAH) enables prompt initiation of treatment. Screening and detection of VS is done by repeated neurological examinations and Transcranial Doppler (TCD) monitoring, while angiograms are used for definitive diagnosis. This study aims to test the ability of a novel NIRS based cerebral-oximetry method to detect VS in the post SAH period.

-Hess score of 2-4 were enrolled. Patients underwent neurological examinations, TCDs and had 20-60 minute NIRS monitoring sessions daily. Whenever VS was suspected, angiography was performed. Clinical event was defined as the combined endpoint of angiographically proven vasospasm, flow velocity >130 m/s over MCA or ACA territories, or neurologic deficit manifested rformed using the CerOx3210, utilizing Ultrasound Tagged Light (UTL). Pathologic cerebral oximetry was defined as having cerebral saturation below 50% for more than 15% of recording time and AUT >500 second%.

20 patients were analyzed, 2 of whom had angiographic vasospasm. These were correctly detected by both NIRS and TCD. Of 9 combined events over the ACA territory, NIRS detected 8/9 events. NIRS also detected desaturations in 3/12 remaining cases, when no clinical or imaging event was detected. Of 12 combined events over MCA territory, 8 had an increase in desaturation AUC, and 5/8 cases with no event had increase in desaturation events. Both cases of angiography proven vasospasm were detected by NIRS as an increase in desaturation AUC, and by TCD as increase in flow velocities.

Cerebral oximetry using UTL based NIRS is comparable to TCD in detecting cerebral vasospasm, and may be superior in early detection of clinical neurologic worsening.

Extracellular fluid volume (ECFV), the main determinant of total circulating blood volume, is determined by the mass balances of Na+ plus K+ (MBNK). In patients with aneurysmal subarachnoid hemorrhage (aSAH), diminished ECFV and reduced circulating blood volume are risk factors for worsened neurologic outcomes. Maintenance of a normal ECFV based on nurse entered fluid balance (FB) has been reported to be difficult. The purpose of this study was to describe the time course of fluid and electrolyte mass balances over 10 days in a cohort of patients receiving hypervolemic or normovolemic therapy.

Data from a randomized trial were secondarily analyzed. The Intensive Management of Pressure Or Volume Expansion in Subarachnoid Hemorrhage Trial randomized 20 patients to receive either a normovolemic or hypervolemic fluid management protocol. The standardized fluid management protocol included maintenance IV fluids with rate adjustments or boluses based on 4-hourly fluid balance and CVP (when available) with a target net positive fluid balance of 1-2 L in the hypervolemia group, and <0.5 L in the normovolemia group. MBNK was calculated using published formulae. FB and estimated MBNK were compared between groups using random-effects generalized least square regression.

Baseline characteristics were similar between groups. FB was higher in the hypervolemia group than in the normovolemia group (mean difference: 713 mL/day, 95%CI:136-1290, p=0.015). MBNK was also higher in the hypervolemia group (mean difference: 90 mEq/day, 95%CI:2-178, p=0.045). Average daily FB did not reach the target in the hypervolemia group. MBNK was negative on 3/10 days in the hypervolemia group, and 9/10 days in the normovolemia group.

Hypervolemic therapy resulted in higher net FB and MBNK compared to normovolemic controls, but was relatively ineffective at generating a consistently positive FB or expanded ECFV. Our results support the notion that hypervolemia is difficult, if not impossible, to maintain in aSAH patients.

Exposure to hyperoxia is commonly seen but it is largely unknown whether hyperoxia is beneficial or harmful in patients with subarachnoid hemorrhage (SAH). We hypothesized that hyperoxia may be associated with increase in the risk of delayed cerebral ischemia (DCI) and poor 3-month outcome after SAH.

We analyzed data from single center, prospective, observational cohort database between 1996 and 2011. Patient nical ventilation, and 4) arterial partial pressure of oxygen (PaO 2 ) measurements. Patients expired within two weeks were excluded. Hyperoxia was defined as the highest quartile of an average area under the curve of PaO 2 until the development of DCI (PaO 2 mmHg) or until the post-bleed day 14 (PaO 2 three months.

Of 252 patients, no baseline characteristics were clinically contributing to hyperoxia. Ninety-seven (38.5%) patients developed DCI. Outcome data were available in 202 patients, and poor outcomes were observed in108 (53.5%) patients. The hyperoxia group had significantly higher incidence of DCI (p = 0.001) and poor outcome (p = 0.009). After adjusting for modified Fisher scale, hyperoxia was independently associated with DCI (adjusted OR, 2.70; 95% CI, 1.49-4.87; p <0.001). After adjusting for age, smoking, alcohol consumption, previous stroke, previous heart disease, Hunt-Hess scale, aneurysm size, Acute Physiology and Chronic Health Evaluation II score, serum glucose, hyperoxia was found to be independently associated with poor outcome measured at 3 months (adjusted OR, 2.80; 95% CI, 1.23-6.39; p = 0.014).

Our data suggest that exposure to hyperoxia after SAH is associated with DCI and poor 3-month outcome. Exact mechanism and the clinical implications can be explored by further investigations.

Advances in management of aneurysmal SAH (aSAH) including refinement of neurosurgical techniques, availability of endovascular options and evolution of neurocritical care have led to improved outcomes following aneurysmal SAH. We evaluated outcomes in aSAH patients admitted to our institution(s) over the past 2 decades.

Prospectively collected data of aneurysmal SAH patients admitted to the Johns Hopkins Medical Institutions between 1991-2009 was reviewed. We compared surivavl to discharge and functional outcomes at first clinic appointment post discharge (30-120 days) in patients admitted between 1991-2000 (phase 1=P1) and 2000-2009 (phase 2=P2) respectively using dichotomized GOS (Good outcome: GOS 4-5).

1134 consecutive aSAH patients were included in the analysis (P1 46.4%; P2 53.6%). There were higher rates of poor grade Hunt & Hess (P1 23%, P2 28%; p<0.05), admission GCS <8 (P1: 14%, P2 21%, p<0.005), known medical comorbidites (P1 54%, P2 64%; p=0.005), associated intraventricular hemorrhage (P1 47%, P2 55%, p<0.05) and an older population in Phase 2 (P1: 51.5, P2 53.5; p <0.05) admissions. Overall in-hospital mortality was low (18.3%) and there was no significant difference between the 2 periods in survival to discharge (p>0.05). Good outcomes were more common in Phase 2 (71.5%) compared to Phase 1 (65.2%); this difference was statistically significant after correction for other confounding factors following multivariate analysis (p<0.001) with 2-fold greater adjusted odds of good outcomes in phase 2.

Our institutional experience over 2 decades confirms that patients with aSAH have shown significant outcome improvements over time.

Hyponatremia in hospitalized patients has been associated with increased mortality, while chronic mild hyponatremia may impair attention and gait. Hyponatremia after aneurysmal subarachnoid hemorrhage (SAH) is common, yet its effect on cognitive outcome remains unclear. We aim to demonstrate the domain-specific cognitive effect of hyponatremia on patients after SAH.

We retrospectively analysed data from 1333 consecutive patients enrolled in our Columbia University SAH Outcomes Project between April 1996 and November 2010. Subjects were excluded if withdrawal of care of death occurred in the first three days. Hyponatremia was defined as a sodium level <130 mEq/L at any time during hospitalization. Univariate and multivariate analyses were performed by a Poisson regression, and a preset alpha of <0.05 was set for statistical significance.

A total of 1195 were included in the study. Hyponatremia developed in 185 subjects (15%). Their mean age was 54 years (SD+/-15), and 65 subjects were men (35%). Median time to onset and nadir of hyponatremia were 7 (IQR 3-10) and 8 days . Univariate analysis associated hyponatremia with worsened modified Rankin Scale at discharge (RR=1.107, CI 1.022-1.199), three-month Telephone Interview of Cognitive Status (TICS) (RR=0.908, CI 0.847-0.974), three-month Barthel Index (RR=0.908, CI 0.833-0.989), and three-month Lawton Instrumental Activities of Daily Living (RR= 1.130, CI 1.006-1.269). After adjustment for age, gender, Hunt and Hess grade, rebleeding, delayed neurologic ischemic deficit, and generalized cerebral edema, hyponatremia was associated with worsened three-month TICS (RR=0.932, CI 0.870-1.000). By one year, hyponatremia was not associated with either functional or cognitive impairment.

Hyponatremia-related injury after SAH appears to be associated with cognitive rather than functional impairment at three months. Early and aggressive reversal of hyponatremia may expedite cognitive recovery among survivors of SAH.

Financial Support: Dr Ortega is supported by the SPOTRIAS fellowship funded by the National Institute of Neurological Disorders and Stroke (NINDS)-P50 NS049060.Dr Mayer consults for Actelion Pharmaceuticals.There are no

Studies have shown that decreased quality-of-life (QOL) after SAH is a significant problem. The factors that predict poor QOL after SAH remain unclear. We sought to identify predictors of a poor quality of life 12 months after SAH.

We prospectively studied 12-month QOL in a cohort of 1200 patients consecutively admitted with SAH between July 1996 and May 2010. Admission clinical scores, radiographic, surgical, and acute clinical course was documented during hospitalization. Twelve months after SAH QOL was assessed using the Sickness Impact Profile (SIP). Reduced QOL was defined as two standard deviations below population-based normative values on the SIP. Univariate statistics were used to identify candidate predictors of poor QOL, and to identify significant concurrent symptoms. Backwards stepwise logistic regression was used to generate multivariable models of reduced QOL.

At 12 months, 35% of survivors who participated in the follow-up survey (272/784) reported reduced QOL. Univariate admission factors associated with reduced QOL were non-white race/ethnicity, high school education or less, poor clinical grade, loss of consciousness, hydrocephalus, pneumonia, and cerebral infarct from any cause. Multivariable analysis revealed that poor Hunt-Hess grade (OR 3.1; CI 95% 2.3-4.4), non-white race/ethnicity (OR 2.5; CI 95% 1.8-5.5), and 12 years or less of education (OR 1.5; CI 95% 1.1-2.1) were significant admission risk factors for poor QOL. Common significant co-morbidities associated with poor QOL at 12 months included greater unemployment, not currently driving, more financial difficulties, current symptoms (e.g., headaches), marital difficulties, fear of recurrent SAH, and dissatisfaction with rehabilitation.

Poor QOL affects as many as one-third of SAH survivors, and is predicted by poor admission clinical grade, non-white race/ethnicity, and lower educational status. Further research is needed to determine if improved access to support and rehabilitation services for high-risk patients groups can improve QOL after SAH. 

Biochemical mediators alter cerebral perfusion potentially resulting in neurological decline and delayed cerebral ischemia (DCI); a significant cause of morbidity and mortality following aneurysmal subarachnoid hemorrhage (aSAH). Estrogens (estrone-E1 and estradiol-E2) are mediators that have demonstrated neuroprotective properties that could play a role in DCI however few studies have evaluated the impact of estrogens on outcomes in humans following aSAH. This study sought to examine the association between cerebrospinal fluid (CSF) E1 and E2 levels and DCI following aSAH.

CSF samples were collected after hemorrhage on 36 adult aSAH patients [14-males, 11-pre and 11-post-menopausal females) admitted to the NV-ICU enrolled in a NIH study (RO1NR004339). Up to 5 CSF samples per patient were selected for analysis representing days 1-10 after hemorrhage. Samples were analyzed for E1 and E2 using liquid chromatography-tandem mass spectrometry. DCI was operationalized as radiographic/ultrasonic evidence of impaired cerebral blood flow accompanied by neurological deterioration. Statistical analysis using SAS(v9.2) included group based trajectory and multiple logistic regression.

E1 was detected in more CSF samples than E2 (83% vs 33%). Group based trajectory identified 2 distinct populations over time for both E1 (42% E1 high) and E2 (52% E2 high) values using censored normal model. Non-weighted Chisquare analysis identified differences between E1 trajectory groups by HH (p=.03) and DCI (p=.06). Using log metabolite levels, higher CSF E1 measurements were associated with higher HH (p=.03) and fisher (p=.06) scores. CSF E1 levels were not associated with DCI (p=.24). There were no differences between CSF E2 and severity of injury or DCI.There was a significant relationship between CSF E1 and E2 concentrations (p<.001).

These findings provide evidence that estrogen metabolites are measureable in CSF and may be associated with severity of injury. Future studies are warranted to further explore these findings and their association to outcomes.

High-grade spontaneous subarachnoid hemorrhage (SAH) patients are monitored in the ICU for up to 14 days, as they are at risk for complications. The diagnosis of treatable complications such as vasospasm of cerebral arteries, cardiac arrhythmias and neurogenic stress cardiomyopathy is often delayed by the limitations of monitoring capabilities. We hypothesized that changes in heart rate variability (HRV) would correlate with the onset of these conditions following SAH.

We applied computational methodology to a cohort of 25 SAH patients in a single neurointensive care unit, examining HRV profiles to identify biomarkers of vasopasm, cardiomyopathy and impending respiratory failure. HRV was quantified for individual 5 min epochs of the electrocardiogram waveform (200 Hz). QRS complexes were identified and the interbeat (RR) interval time series was constructed. Mean, standard deviation and coefficient of variation of RR intervals, as well as the ratio of low frequency to high frequency power spectral density and standard Poincare statistics were quantified.

Vasospasm occurred in 5 (20%), stress cardiomyopathy in 5 (20%) and respiratory failure in 9 (36%) of patients. In a SAH patient with Takotsubo's cardiomyopathy and respiratory failure, we found a decrease in HRV that predated the discovery of cardiomyopathy as well as the onset of respiratory distress by several hours.

The early clinical detection of vasospasm, cardiomyopathy and impending respiratory failure from on-line EKG HRV analysis would be of tremendous clinical value. In the face of changing autonomic influences in the critically ill postaneurysmal subarachnoid hemorrhage patient, the finding of an early signal prior to clinical detection of respiratory failure is encouraging. A larger and more highly annotated dataset may be required to increase the signal to noise ratio to realize the clinical potential of HRV-based biomarkers.

Retrospective analyses have found an association between transfusion and vasospasm, medical complications and mortality in subarachnoid hemorrhage (SAH) patients. Yet, none of those studies assessed the timing of transfusion, whether it occurred before, or, after vasospasm or complications. We sought to clarify whether transfusion could be considered a cause or consequence of vasospasm and complications. 

This interim analysis indicates that transfusion is not associated with vasospasm or infection when timing of transfusion is considered; fluid overload was more common after transfusion.

The most dismal sequelae of aneurysmal subarachnoid hemorrhage (aSAH) are the development of cerebral vasospasm and consecutive delayed infarctions. Their severity is linked to the clinical grade of the initial hemorrhage and the amount of blood in the basal cisterns. Together, they represent the major cause of unfavorable clinical outcome and death in aSAH patients. From retrospective data, a promising method to reduce the incidence of vasospasm is the use of a lumbar drain to remove the blood from the subarachnoid space. The recently completed LUMAS trial addressed the safety of this approach in good-grade aSAH patients (1). However, so far prospective data from subjects being at high risk for vasospasm and delayed infarction is lacking.

We present the protocol of the Earlydrain study, a prospective randomized multicenter trial comparing an intervention group with early continuous lumbar CSF drainage to a control group receiving standard neurointensive care only (2).

Eligible for participation are adults suffering from aSAH of all clinical grades who receive aneurysm treatment within 48 hours of ictus. Primary endpoint is the modified Rankin score at six months. Secondary endpoints include mortality, angiographic vasospasm, cerebral infarction, transcranial Doppler sonography (TCD) mean flow velocity and rate of shunt insertion at six months after hospital discharge.

The Earlydrain study had recently been launched and, at abstract submission, 34 patients of 300 planned were enrolled. Interim safety analysis did not reveal any concern on the use of lumbar drains after aneurysmal SAH. Up to now, ten centers in Germany, Switzerland and Canada are participating. Interested centers willing to join and contribute are still much appreciated. 

Patients with aneurysmal subarachnoid hemorrhage (aSAH) require management in centers with neurosurgical expertise necessitating emergent interhospital transfer (IHT). Our objective was to compare outcomes in aSAH IHTs to our institution with aSAH admissions from our institutional emergency department (ED).

Data for consecutive patients with aSAH admitted to Johns Hopkins Medical Institutions between 1991 and 2009 were analyzed from a prospectively obtained database. We compared in-hospital mortality and functional outcomes at first clinical appointment post-aSAH (30-120 days) using dichotomized Glasgow Outcome Scale (good outcome: Glasgow Outcome Scale 4-5) in ED admissions with IHTs.

A total of 1134 consecutive patients with aSAH were included in analysis (ED 40.1%, IHT 59.9%). Direct ED admissions had a higher incidence of poor Hunt and Hess grade (4/5) and major medical comorbidities, with no significant differences between the 2 groups in age, intraventricular hemorrhage, and hydrocephalus. In-hospital mortality for ED admissions (14.9%) was significantly lower than that for IHTs (20.5%), with 1.8 times greater adjusted odds of survival after multivariate analysis (P = .001). Emergency department admissions had nearly 2-fold greater odds of good outcomes (odds ratio, 1.89; P b .001) after multivariate analysis.

Our institutional ED SAH admissions had significantly better outcomes than did IHTs, suggesting that delays in optimizing care before transfer could deleteriously impact outcomes.

Left ventricular (LV) systolic abnormalities occur commonly after subarachnoid hemorrhage (SAH). Cardiomyopathy associated with SAH can include either predominate apical LV systolic dysfunction (CM-apical) or predominate basal LV systolic dysfunction (CM-basal). We aimed to determine if outcomes and diastolic function were different between patients with various forms of LV dysfunction after SAH.

Patients hospitalized for SAH between 2000 and 2008 were eligible for our study. Those patients with a history of heart failure, myocardial infarction, or a documented acute coronary process were excluded. Echocardiograms were reviewed and a wall motion score was provided for each of 16 LV segments. Patients were classified as CM-apical if the average wall motion score for apical segments was greater than the average wall motion score for basal segments. Patients were classified as CM-basal if the average wall motion score for basal segments was greater than the average wall motion score for apical segments.

198 of 583 patients with SAH had an echocardiogram. 142 patients had normal LV function, 32 CM-apical and 24 CMbasal. The in-hospital mortality was not different between those with no echocardiogram or those with an echocardiogram who had normal LV function, CM-apical or CM-basal. Patients with CM-apical were more likely to have shock and pulmonary complications, whereas patients with CM-basal were more likely to have sepsis. During a median follow up of 5.4 years, patients with CM-apical had the worse survival. Patients with CM-apical and CM-basal had impaired LV relaxation as compared to those with normal LV systolic function.

In-hospital mortality is not different between those patients with normal LV systolic function, CM-apical, or CM-basal. CMapical is associated with shock and pulmonary complications and a worse long term survival. Further work evaluating the response to medical intervention and the differences in hemodynamic profiles of patients with CM-apical and CM-basal is warranted.

Therapy using sodium nitroprusside (SNP) intrathecal (intraventricular) aims for a more effective approach for prophylaxis and treatment of cerebral vasospasm associated to a subarachnoid hemorrhage (SAH).

Qualitative study whose objective was to analyze clinical cases to specific approach for cerebral vasospasm related to SAH.

Two patients, the first one is a 62 years old female with aneurysm rupture of the left posterior communicating artery, SAH Fisher III, Hunt Hess 2. The second one is 46 years old male with artery rupture of the middle cerebral artery, SAH Fisher III, Hunt Hess 2, both were submitted to embolization, leading to acute hydrocephalus, in which external ventricular drainage (EVD) was established. Through the EVD, a prophylactic intrathecal protocol was instituted (2ml SNP with 10,5 ml of normal saline 0,9% solution applying 2 ml NPS through the EVD each 12 hours for 1 hour by infusion pump). Patients evolved well with no neurologic or motor sequel and with a modified Rankin scale = 0. The third patient was a 37 years old male with aneurysm rupture in anterior communicating artery, SAH Fisher III, Hunt Hess 4, severe vasospasm per operative in the left middle cerebral artery (MCA), treated by angioplasty with balloon. Starting the treatment protocol of cerebral vasospasm by lombar catheter: dosage 50 mg (2ml) SNP, solution with 4ml SNP at 6 ml of normal saline 0,9% applying 4ml through the lombar catheter each 12 hours for 1 hour by infusion pump. Patient without complication with modified Rankin scale = 1.

The cost for prophylactic therapy for 14 days was U$ 627,86; if the patient had developed clinical vasospasm, the cost for a 14 day treatment would be an average of U$ 15.287,80, having a great impact on morbidity, mortality and cost of hospital stay.

Angiography does not reveal a source of bleeding in 10-15% of those with subarachnoid hemorrhage. These patients usually have a benign course and favorable outcome, especially those with a perimesencephalic pattern of bleeding (PM-SAH); more diffuse bleeding has been associated with higher risk of vasospasm and neurological disability. We evaluated whether amount or pattern of bleeding better predicts risk of neurological complications and outcome.

Methods 114 angio-negative SAH patients were prospectively studied over seven years. Six were excluded when a vascular etiology was identified on repeat angiography. Pattern of bleeding, amount of cisternal (Hijdra score) and ventricular blood (IVH score), and ventriculomegaly (bicaudate index) were assessed. Neurological outcomes included hydrocephalus, angiographic vasospasm, and delayed ischemic neurological deficits (DIND, based on clinical deterioration). Functional outcome was assessed at 1-year using the modified Rankin Scale (mRS).

Bleeding was perimesencephalic in 60(56%), diffuse in 28(26%), cortical in 7 and CT-negative in 13. Patients with diffuse bleeding had higher Hijdra (12[IQR 7-17] vs. 5[3-8]) and IVH scores (2[0-7] vs. 0[0-2]), and bicaudate index (0.19[0.14-0.24] vs. 0.14[0.11-0.19]) than those with PM--V (18% vs. 5%, p=0.05) and require ventriculostomy (54% vs. 18%) and shunt placement (29% vs. 3%, both p=0.001). Moderate-severe angiographic vasospasm developed in 29% diffuse vs. 13% PM-SAH (p=0.08), while DIND only occurred in those diffuse bleeding (14%). Neither Hijdra nor IVH score was higher in those developing vasospasm, across or within bleeding patterns. Those with diffuse SAH were less likely to be discharged home (68% vs. 90%, p=0.01) or achieve minimal disability (mRS 0-2, 83% vs. 96%, p=0.11).

Angio-negative SAH can result in hydrocephalus, vasospasm, cerebral ischemia, and residual disability. This is more likely in those with diffuse bleeding, a disparity not explained by a greater volume of cisternal or intraventricular bleeding.

Independent of the cholesterol lowering effects of hydroxymethylglutaryl conenzyme A reductase inhibitors(statins), there has been much debate about their protective effect against delayed cerebral ischemia (DCI). Various ongoing trials are aimed at assessing their effectiveness against DCI after primary subarachnoid hemorrhage (SAH). There is scanty literature on DCI in patients who were on statins prior to the occurrence of SAH.

A retrospective chart review was done after approval from the institutional review board. Data was collected from July 2009 to April 2012 using the ICD code for SAH. Patients with SAH secondary to AVM, trauma and surgery were excluded. Demographics, baseline characteristics and occurrence of clinical DCI were collected. Admission home medication list was used to identify patients on statins prior to admission. All statistical analysis was done using SAS.

A total of 147 patients with primary SAH were included. Out of 147 patients, 23 (15.6%) were on home statin. Only 2 (9%) patients within this group developed DCI while 40 (32.2%) patients in the statin naive group developed DCI (p=0.02). This difference persisted even after correcting for age (p=0.03), sex (p=0.03), race (p=0.03), smoking (p=0.03), history of diabetes (p=0.02), stroke/TIA (p=0.03), peripheral vascular disease (p=0.04), hypertension (p=0.03), hyperlipidemia (p=0.01), home calcium channel blocker use (p=0.03) and fisher grade (p=0.02). A multivariate logistic regression analysis with backward selection further confirmed that the only significant factor affecting vasospasm was prior statin use (p = 0.03).

The above findings suggest that prior statin use reduces the rate of DCI after SAH. Though the known confounders were taken into consideration, the possibility of unknown confounders cannot be completely excluded. A larger prospective study may be required to verify these effects. The potential clinical implication of this would be to put patients with unruptured and untreated aneurysms on long-term statins.

Patients sometimes report that surviving a near-death experience results in enhanced appreciation of the preciousness and joy of life. We sought to determine how frequent the "Stroke of Insight" phenomenon occurs after SAH.

We prospectively enrolled 1200 SAH patients between 1996 and 2009 and followed up survivors with a telephone interview at 3 and 12 months. Patients were asked "do you enjoy life more, about the same, or less than you did before your brain hemorrhage?" Surrogate responses were not analyzed. Global functional outcome was evaluated with the modified Rankin scale (mRS) and QOL with the Sickness Impact Profile (SIP).

Of 764 survivors who responded to the survey, the majority (59%, N=449) reported that they enjoyed life more since the hemorrhage, whereas only 12% enjoyed life less. Enhanced life enjoyment was associated with female gender and white (versus non--Hess grade. Patients with enhanced life enjoyment were more likely to report improved marital status (18% versus 11%, P<0.0001), and were less likely to have rumination on their illness (P<0.001). Improved life enjoyment was associated with better SIP QOL scores (9.7 ± 11.5 versus 14.1 ± 14.4, P<0.001), but had no relationship with concurrent disability on the mRS (P=0.13). Remarkably, 31% of those reporting that they enjoyed life more were unable to walk without assistance (mRS 4 or 5).

The majority of SAH survivors enjoy life more after their hemorrhage. Increased life enjoyment has no relationship with physical disability and handicap, but is associated with improved QOL. Informing patients of the "Stroke of Insight" phenomenon may be a simple and effective way to set positive expectations and promote recovery after SAH and similar life-threatening illnesses.

Parenteral diclofenac infusion is commonly used in neurocritical patients and has been shown to effectively decrease body temperature after aneurysmal subarachnoid haemorrhage (aSAH). Hemodynamic side effects and in specific the effect on brain homeostasis are understudied.

Twenty-one aSAH patients with multimodal neuromonitoring of intracranial pressure (ICP), brain tissue oxygen tension (P b tO 2 ), and cerebral metabolism (microdialysis, MD) receiving parenteral diclofenac infusions were analyzed in a prospective observational cohort study. 75mg diclofenac diluted in 100cc normal saline was administered at the discretion of the attending neurointensivist. We analyzed core body (CBT) and brain temperature (BT) over 12 hours and hemodynamic (cardio-, cerebrovascular) and cerebral metabolic parameters over 4 hours after intervention. Ten-minuteaverage files of cardio-and cerebrovascular parameters and hourly files of MD datasets were analyzedusing a generalized estimating equation.A pre-intervention baseline was calculated for every parameter.

One-hundred-twenty-three parenteral diclofenac infusions over 34min (IQR 20-45min) were analyzed. CBT and BT decreased to a minimum of 37.5± 0.6°C and 37.3± 0.5°C, 7h and 6h after diclofenac infusion (baseline 38.3°C± 0.6°C and 37.9± 0.5°C, respectively, P<0.001). Hemodynamic side effects included a 10% reduction of MAP (by 10±12mmHg) and CPP (by 9±13mmHg) resulting in increased use of vasopressors in 26% of interventions (P<0.001). P b tO 2 significantly decreased from 28±15mmHg baseline by 12% (P<0.001) resulting in brain tissue hypoxia (P b tO 2 <20mmHg) in 37% of interventions and 38% (n=8) of patients. In none of the interventions with baseline P b tO 2 above 30 mmHg, brain tissue hypoxia was observed. Baseline-P b tO 2 below 30 mmHg was independently associated with brain tissue hypoxia during intervention (P<0.001). There was a trend towards higher brain tissue lactate-pyruvate ratio and lower pyruvate after

Parenteral diclofenac after SAH is associated with hemodynamic side effects and may result in brain tissue hypoxia without significantly affecting brain metabolism. The impact on outcome needs further investigation.

Delayed cerebral ischemia (DCI) is a complication of subarachnoid hemorrhage (SAH) with significant mortality/morbidity. Digital subtraction angiography (DSA) can detect cerebral vasospasm which is a surrogate marker for DCI. There is emerging data that perfusion computed tomography (CTP) is useful in detecting DCI. We have compared the utility of CTP and DSA in detecting DCI.

Patients with primary SAH admitted to two academic institutions between July 2009 and April 2012 were identified. Patients with clinical DCI who underwent DSA or CTP (image processing through Vitrea®) were included. The area of perfusion abnormality was traced out to generate cerebral blood flow (CBF), mean transit time (MTT) and cross sectional area. Abnormal CBF and MTT values were compared to normal symmetrical areas in the opposite hemisphere. DSA reports were reviewed to identify radiologic vasospasm.

Out of 154 patients, 43 had clinical DCI(27.9%). In those with DCI, 16/18 patients that underwent CTP had abnormalities (88.9%) compared to 34/39 patients that had vasospasm on DSA (87.2%; p=0.85). Median abnormal CBF was 21.6 (6.3-40.9) ml/100 gm/sec compared to 41.7 (19.3-70.7) ml/100 gm/sec in area of normal perfusion (p=0.001). Median abnormal MTT was 7.5 (5.1-15) seconds compared to the normal area of 4 (2.2-6) seconds (p<0.001). Median interhemispheric CBF and MTT difference was 14.4 (3.8-48.6) ml/100 gm/sec and 2.7 (0.9-12.3) seconds respectively. Median area of abnormal perfusion was 8.6 (2.7-52.1) cm 2 . Seventeen patients underwent CTP and DSA. A normal CTP excluded vasospasm on DSA. Perfusion abnormalities involving an area of less than 7.5 cm 2 did not have vasospasm on DSA.

CTP is a useful indicator of DCI and is comparable to DSA. In patients with clinical DCI and a normal CTP, DSA is unlikely to pick up vasospasm. As the area of perfusion abnormality increases (greater than 7.5 cm 2 in our subset of patients), DSA is more likely to show vasospasm.

Aneurysmal subarachnoid hemorrhage (aSAH) is more common in women than in men. Current knowledge on potential gender differences after an aSAH occurred is sparse, albeit of clinical relevance.

Retrospective cohort study including patients with aSAH admitted to a neurovascular center at a major academic center at the University Hospital of Bern, Switzerland. Patients below age 16 and with non-aneurysmal SAH were excluded.

We included 120 consecutive patients with aSAH between January 03, 2009 and February 26, 2011. Women were older than men (median age 57 years [interquartile range [IQR] 49-67] versus 54 [IQR 44-59], respectively, P=0.01), and progressively overrepresented with increasing age (69.2% of women for the whole cohort). Of note, in the Swiss population the proportion of both genders between 40 and 64 years is similar, with women being slightly overrepresented at older ages. Global disease severity at admission, measured by the Acute Physiology and Chronic Health Evaluation (APACHE) II score, was higher in women than in men (median score 18 points versus 14 [IQR 10-19], P=0.006) even after correction for age. The APACHE II score independently predicted an unfavourable outcome and mortality as opposed to gender. We found no differences between genders in the adopted aneurysm-securing strategy, intensive care interventions (administered drugs, rates of endotracheal intubation, tracheostomy, length of mechanical ventilation and placement of an external ventricular drainage). Women and men with aSAH confirmed to be similar in terms of medical history, clinical / radiological severity of aSAH, complications and outcome.

In conclusion, this study confirms that women with aSAH outnumbered men, especially at higher age. Global disease severity on admission is higher in women and predicts, independently from gender, unfavourable outcome and mortality. Finally, this study finds new relevant similarities between genders.

Complications of aneurysmal subarachnoid hemorrhage (aSAH) may include hypertension and neurogenic myocardial stunning. Subsequent management often involves beta blockade. High Fisher grade aSAHs may also be complicated by cerebral vasospasm, which could have pathophysiologic influence from sympathetic nervous system stimulation or inhibition.

We investigated any relationship of beta blockade to the incidence of radiographic vasospasm in aSAH by retrospectively examining 219 adults admitted to the SICU at Loma Linda University Medical Center between 8/2004 and 9/2010, excluding those who expired within 3 days of admission because of inability to assess outcomes. Three groups were isolated relevant to beta blockade: 77 were never beta blocked (No/No), 123 were started on a beta blocker after admission (No/Yes), and 18 were continued on their home beta blockers (Yes/Yes). Records were analyzed for the development of vasospasm with or without resultant infarction, death, and discharge status. Outcomes were evaluated using multivariate analysis through logistic regression and adjusted for potential confounders. Odds ratios were calculated setting the OR for No/No patients to 1.

One hundred and forty five patients had vasospasm, 47 consequently infarcted, and 53 died or required care in a longterm facility. Patients in the No/Yes group had significantly increased radiographic vasospasm ]. However, despite increased incidence of vasospasm, these patients had significantly fewer deaths or need for long term care [OR 0.17 (0.05-0.64)], with decreased tendency for infarcts ]. In the Yes/Yes group, there was a trend toward increased vasospasm ] that led to infarction )], with decreased mortality or need for long term care in a facility [OR 0.13 (0.01-1.30)].

The use of beta blockers in aSAH is associated with increased incidence of radiographic cerebral vasospasm. However, despite the increased rate of vasospasm, the use of beta blockers was associated with improved discharge characteristics.

Patients with subarachnoid hemorrhage (SAH) frequently undergo continuous electroencephalography (cEEG) monitoring in the ICU. We describe commonly encountered EEG patterns in SAH patients with clinical correlation.

Patients with primary SAH admitted to two academic institutions between July 2009 and April 2012 were identified. Records were reviewed to note the presence of intraventricular hemorrhagic extension (IVH), intracerebral hemorrhagic extension (ICH), location of subarachnoid blood, occurrence of delayed cerebral ischemia (DCI), patient outcomes and length of stay (LOS). EEG reports were reviewed and classified as to the presence of arrhythmic continuous slowing (aCS), rhythmic and periodic slow activity of triphasic morphology (TW), epileptiform activity (EA), and coma pattern. Patients with metabolic causes for TW were excluded.

Of 154 patients, 49 had a routine EEG or cEEG monitoring.Thirteen (26.5%) exhibited non-metabolic TW, 16 (32.7%) had EA, 18 (36.7%) had aCS, 1 patient had coma pattern and 1 had normal EEG. The presence of subarachnoid blood around the basal cisterns did not influence EEG patterns.In patients with IVH, the presence of TW patterns was significantly more common than other patterns (69.2% vs.41.7%;p=0.04). EA was associated with DCI (56.3%) as compared to non-epileptiform patterns (31.7%;p = 0.04).EA was more common in patients with ICH without statistical significance(31.3% vs.18.2%;p=0.2). Median LOS in patients with TW, EA and aCS were 27 (10-93), 24.5 (7-115) and 20.5 (11-68) days respectively without significant difference. Patient outcomes were similar among all groups.

Non-metabolic TW are scantly reported in the literature and typically associated with diencephalic and brainstem lesion. In patients with SAH, the presence of IVH and not cisternal blood was associated with non-metabolic TW. DCI was significantly associated with the generation of epileptIform activity and the presence of ICH seemed to favor an epileptiform pattern. EEG patterns did not influence LOS or outcome in our subset of patients.

Adenosine is an endogenous purine nucleoside that causes transient heart block in the AV node when administered parenterally. We describe our experience with 25 cases of severe intraoperative aneurysm rupture in which adenosine was administered to allow for control of the intraoperative bleeding.

Over a 10 year period, we have treated approximately 2000 aneurysms with open microsurgery. Two-thirds were unruptured. Severe intraoperative aneurysm rupture that could not be readily controlled occurred in 25 cases. 24 of the aneurysms had recently bled, 1 case was an unruptured aneurysm. In all cases, the amount of bleeding precluded safe application of temporary clips.

An intravenous infusion of adenosine (6 mg) was given in all cases. In 16, there was significant bradycardia and hypotension culminating in a brief cardiac pause (5-25 seconds), allowing for rapid dissection and clipping of the aneurysm. In 4 cases, there was bradycardia and hypotension, but no cardiac arrest. In 5 cases, there was limited bradycardia and hypotension, and a second dose (12 mg) was required to slow the heart enough to allow for aneurysm treatment. In such cases, the adenosine allowed us to clear the field adequately to apply temporary clips in a precise fashion, and then to clip the aneurysms properly. Poor response to the initial dosing was not related to patient size or other identifiable factor.

Adenosine has been used safely in our experience to allow for management of severe intraoperative aneurysm rupture. In most cases, there is a meaningful cardiac pause. In some instances, patients are less sensitive, and the dose must be repeated to achieve the desire effect. No adverse cardiac or pulmonary events were associated with the use of adenosine in our series.

Intraventricular hemorrhage (IVH) is an established independent predictor of poorer outcome in subarachnoid-and intracerebral-hemorrhage. Though, limited knowledge exists regarding the pathophysiologic mechanisms that may lead to cerebral injury and poorer outcome. This is the first report presenting in vivo data on cerebral perfusion and brain tissue metabolism during the occurrence of IVH and after intraventricular fibrinolysis (IVF).

A 78-year-old woman with severe subarachnoid hemorrhage (SAH), Hunt&Hess grade 3, modified Fisher Scale 4, was admitted to our neuro-critical care unit. Within the first 24 hours an extraventricular drainage was placed and a left-sided MCA aneurysm was coiled. After obtaining informed consent from the legal attorney, the patient received invasive multimodal neuro-monitoring, consisting of a cerebral blood flow (CBF)-and microdialysis-probe placed into the ipsilateral frontal white matter.

Within 8 hours after probe placement we observed a significant drop of cerebral blood flow (CBF below 20 ml/100g/min) and an increase in L/P-ratio without significant changes in cerebral perfusion-or intracranial-pressure. Imaging revealed a re-hemorrhage into the ventricular system with blockage of the foramina of Monro and acute hydrocephalus. Consequently, therapeutic IVF was undertaken with 1 mg of rtPA which lead to sufficient clot resolution. After IVF we normalization of cerebral perfusion and metabolism.

This is the first report on IVH and its potential mechanisms that may contribute to secondary injury in the human brain. A decrease of cerebral blood flow and disturbance of cerebral metabolism was documented during the occurrence of IVH, supporting existing hypotheses of global impairment. Moreover, we could document profound treatment effects of IVF leading to a restored CBF and a stable aerobic metabolism in the investigated brain tissue.

Many patients with aneurismal subarachnoid hemorrhage (SAH) present with acute, labile, hypertension and may be at risk for rebleeding. Clevidipine, a novel, ultra-short acting dihydropyridine has been used in cardiac surgery, acute hypertensive emergencies and patients with intracerebral hemorrhage, but not in SAH patients.

The CLASH study (clevidipine in aneurismal subarachnoid hemorrhage) is a prospective evaluation of the efficacy and safety of clevidipine in controlling systolic blood pressure (SBP) before the aneurysm is secured. The primary endpoint is the number of patients achieving SBP target within 30 minutes. Post-hoc, SBPs pre-infusion, during-infusion and postinfusion were compared using a generalized estimating equation.

We present the first 5 patients enrolled: 2 men and 3 women, mean H&H 2 and Fisher 3.4, 3 aneurysms coiled and 2 clipped. Mean SBP upper and lower goals were 154 ± 5.5 and 122 ± 4.5 mmHg. Analyses included 1,587 SBP data points. All patients reached SBP target within 13.3 ± 6 min using an infusion rate of 9.3 ± 8.9 mg/hour. The mean preinfusion, during-infusion and post-infusion SBPs were 165.5 ± 2.55, 146.4 ± 2.48 and 159.3 ± 11.5 mmHg (pre-infusion vs during-infusion p <0.05, pre-infusion vs post-infusion p <0.49, during-infusion vs post-infusion p <0.27). After the 1st SBP control readings, SBP was above the upper target goal 17.9 ± 18.8% and below the lower 12.3 ± 16.3% of the time. ICP did not increase during infusion (n=2). No patient rebled. In one patient the infusion was stopped temporarily 3 times due to SBP below the target range. There were 7 (0.4%) SBP values <100 mmHg and none <90 mmHg.

Clevidipine controlled SBP in all patients with SAH in <30 min and kept SBP within the selected range in 69.8% of the time without any patient rebleeding.

Financial Support: Research Grant from The Medicines Company to conduct this study.

Aneurysmal subarachnoid hemorrhage (aSAH) is associated with numerous adverse sequelae. Patients who survive the initial hemorrhage are at high risk for delayed secondary brain injury, including cerebral infarction, neuronal cell death, white matter abnormalities, and hydrocephalus. resulting in focal neurological deficits, cortical dysfunction, and both longterm cognitive and psychosocial deficits referred to as SAH-induced "delayed neurological deficits" (DNDs). Review of the literature revealed that heparin had previously been advocated to reduce complications of aSAH. Here, we report on our favorable experience with the use of heparin prophylaxis in the management of patients who are at a high risk for developing SAH-induced DNDs.

A retrospective chart review of 235 patients that presented to the University of Maryland Medical Center were reviewed between January 2006 and May 2011. Inclusion criteria were patients with Fischer Grade III SAH due to rupture of a true saccular aneurysm and were treated by surgical clipping within 48 hours of the patient's ictal event. Exclusion criteria were patients who had a localizing deficit related to an intracerebral hematoma from the ictal event. Included in this study were 43 patients that were started on an intravenous infusion of heparin and an additional 43 patients that served as matched controls.

None of the patients exhibited heparin-induced thrombocytopenia (HIT). The heparin regimen used appeared to be safe. Patients administered low-dose IV heparin experienced significantly fewer occurrences of ischemia-related CT hypodensities as well as symptomatic vasospasms than case controls.

Retrospective analysis of our clinical experience with constant IV infusion of low-dose heparin in patients at high risk for SAH-induced DNDs indicates early use of low-dose IV heparin infusion may be safe and perhaps beneficial in patients having undergone surgical clipping. Further study with a double-blind placebo-controlled trial is warranted to establish the role of heparin in the prevention of SAH-induced DNDs.

Subarachnoid hemorrhage patients (SAH) may experience cardiac biomarker elevation in serum troponin and B-type natriuretic peptide (BNP). We hypothesized that elevations in these cardiac biomarkers after SAH are predictive of increased patient mortality.

We retrospectively reviewed the medical records of all non-traumatic SAH patients admitted from March 2011 to March 2012 including medical history, modified Fisher scale on initial head CT scan, initial Glasgow Coma Scale (GCS), serum troponin T and BNP within 24hrs of admission. Survival data was dichotomized as either alive or dead by chart follow-up.

values (>91pg/ml) versus normal values against alive or dead status.

We identified 78 SAH patients, 73 with initial measured troponin, and 55 with initial measured BNP.The mean age was 58 (range 18-92) and 38% male. Modified Fisher grade was 0-2 in 47%, and grade 3-4 in 46%. The initial GCS mean was 11 (range 3-15), 91% of patients had intracranial aneurysm, while 9% were 'angiogram-negative' SAH. Twenty SAH patients died, with a mean of 19 days post SAH (range 1-107), six from cardiopulmonary or multiple organ failure, 5 from SAH, and 9 unknown/other. Elevated troponin was seen in 23% (17 of 73) with a mean = 0.46 (range, 0.1-5.23), and elevated BNP in 49% (27 of 55 patients) with a mean = 388(range, 92-2480). Patients with elevated levels of troponin had a greater chance of death (P=0.009). Patients with elevated levels of BNP also had a higher mortality (P=0.002).

The data demonstrate a statistically significant association with elevated cardiac biomarker elevation and risk of subsequent death after SAH, which occurs not only during the immediate post SAH period but after initial hospitalization.

Delayed cerebral ischemia (DCI), length of stay and Glasgow outcome scale (GOS) following angiogram-negative SAH (anSAH) are infrequently and inconsistently described in the literature. Furthermore anSAH are generally considered to have a better prognosis than aneurysmal SAH (aSAH). anSAH subgroups include benign perimesencephalic SAH (PMH) and aneurysmal-type or diffuse SAH. We report and compare outcome data of patients presented with diffuse anSAH and diffuse aSAH.

A retrospective chart review of 139 patients who presented to 2 academic institutions between July-2009 and April-2012 who met the criteria for diffuse spontaneous SAH were reviewed. The patients were further divided into anSAH (n=25) and aSAH (n=114). Delayed cerebral ischemia rates, length of stay and discharge GOS were compared and analyzed between two groups using SAS statistical software. Discharge GOS scale was dichotomized in good outcome (GOS 4-5)

Out of 139 patients, a total of 25 (17.9%) patients meet the criteria of diffuse anSAH and 114 (81.9%) meet the criteria of diffuse aSAH. Demographics and baseline characteristics including age, sex, race, hypertension, diabetes, GCS on presentation, Hunt & Hess score and Fisher grade among two groups were comparable. Overall 25% (n=5) of anSAH and 32% (n=37) of aSAH showed DCI (p=0.33). Mean length of stay was 17 days in naSAH and 19 days in aSAH. Good outcome was seen in 72% (n=18) in naSAH and 61% (n=70%) in aSAH groups (p=0.36).

In our patient cohort of anSAH, 25% of patients had DCI. Even though it is less then aSAH group it is considerably higher then previously reported in the literature. Furthermore length of stay and discharge GOS between two groups were comparable. This study indicates that diffuse anSAH is not a 'benign' condition and warrants a low index of suspicion for complications with a multidisciplinary approach to management.

Transcranial Doppler (TCD) is a common method used to measure cerebral blood flow velocities and estimate flow resistance related to intracranial pressure (ICP). We present the case of a patient with subarachnoid hemorrhage and clipped aneurysm, who, while undergoing TCDs, rebled.

A 68 year old man presented with sudden-onset severe headache and neck pain. CT of the head showed a subarachnoid hemorrhage (SAH) with intraventricular extension and obstructive hydrocephalus. An anterior communicating artery (ACOM) aneurysm was found and clipped and a ventriculostomy was placed. After surgery there was an interval decrease in the SAH. Eight days after the original event the patient re-bled during a TCD test because of clip failure.

TCD waveforms were captured before, during the bleed and post treatment with mannitol and CSF drainage from the ventriculostomy. Prior to the bleed. ICP was 8mm Hg, the left MCA flow velocity was 62 cm/sec and the pulsatility index (PI) 1.0. During the bleed the ICP increased to 54 and PI to 1.7-2.0, with the waveform showing a narrow peak and decreased diastolic and mean velocity. Mannitol 75 g was given and the ventriculostomy was opened to drain. Within 5 minutes the ICP decreased to 14 mm Hg, the PI improved to 1.4, the waveform widened and the velocities returned to previous levels (video will be provided with the Abstract showing the TCD changes). Repeat CT of the head showed increased SA blood and extensive new intraventricular hemorrhage; catheter angiogram a malpositioned clip. The ACOM aneurysm was coiled successfully.

We present this unique case of TCD capturing the dynamics of a real-time intracranial aneurismal bleed with significantly elevated ICP. Our data demonstrated the TCD PI, flow velocities and waveforms changed dramatically during the rebleeding and improved quickly with treatment.

Transcranial Doppler (TCD) is the least invasive method to detect cerebral vasospasm but is unable to interrogate vessels beyond the circle of Willis and is highly operator-dependent. We tested a novel technique whereby we record the miniscule pulsation of the skull gated with cardiac contraction and compared it to TCD in patients with subarachnoid hemorrhage.

Skull accelerometry was performed using a prototype device designed by Jan Medial, Inc. (Mountain View CA). The device has 6 highly sensitive accelerometers that couple through plastic feet to the patient's scalp, arrayed with 2 detectors over the forehead, 1 at midline occiput, 1 each over the temporal bones, and 1 on the patient's vertex. They are held in place with a plastic strap. Paired TCD recordings and accelerometry epochs (typically 15 minutes of recording) were compared in patients with and without spasm.

A total of 62 accelerometry recordings were obtained in 20 subjects with subarachnoid hemorrhage who had paired TCD recordings. This allowed 186 distinct pairings of data sets (right, left, posterior). A unique signature was identified by a fast Fourier transform waterfall technique revealing a shift in accelerometry signals to higher frequencies (representing a "bruit" of sorts) in patients with TCD identified vasospasm. An analytic model was created based on the first 124 recordings, and validated using the remaining 62 recordings. This revealed 86% sensitivity and 82% specificity for detection and localization of spasm.

Highly sensitive skull accelerometry detects a shift toward higher vibration frequency in patients with vasospasm-a cranial "bruit". This technique may be a highly sensitive tool for the detection of cerebral vasospasm following subarachnoid hemorrhage. A prospective, blinded validation study is on going to measure this novel tool's performance characteristics in a larger sample of patients.

Financial Support: Research grant from Jan Medical, Inc.

J.N. is 45 year-old Hispanic male prisoner previously healthy presented to our institution altered due to diffuse subarachnoid hemorrhage (Fisher grade IV) and a bi-lobed "Mickey Mouse" right M2 middle cerebral artery (MCA) ruptured aneurysm. Initially, J.N.'s Hunt and Hess grade level of 1 on arrival, but declined to a 4 in the ED. J.N. was intubated and an external ventricular device was placed.

The anatomy of the aneurysm was complex in nature measuring 8mm in maximal dimension with the superior lobe measuring 3.5mm and the inferior lobe measuring 3.0mm. Based on the complex anatomy of the aneurysm, a 4-vessel angiogram was planned to treat the aneurysm with a trans-arterial coil-embolization approach.

A 4X15mm septal balloon was used with a Synchro microwire, with the balloon been placed across the neck of the inferior aneurysm. The superior aneurysm was accessed with a SL-10 microcatheter and coiled in the usual fashion. The SL-10 microcatheter was then re-directed to the inferior aneurysm and coiled similarly. Post-angiographic images showed complete obliteration of the aneurysm with a small neck residual to protect en passé branches.

Evaluation of the literature is scant with reports of bi-lobed aneurysm with the classic description of "Mickey Mouse" or "Mirror" aneurysm. Trans-arterial coil-embolization provided a safe, rapid, and effective method for coiling a complex bilobed aneurysm with no major thrombo-embolic events. Trans-arterial coil-embolization is a procedure used in the treatment of gross hematoma and fistula in human and the veterinary population. To our knowledge, there is no report of trans-arterial coil embolization for the treatment of bi-lobed aneurysm posted within the usual medical research engines. Our institution is presenting a novel endovascular technique in the treatment of a classic bi-lobed Mickey Mouse aneurysm. J.N. was able to recover fully and eventually discharge to the infirmary in federal prison.

The routine practice of therapeutic hypothermia is advocated in the management of comatose survivors of out-of-hospital cardiac arrest (OHCA), particularly if ventricular fibrillation is the initial rhythm. Potential benefits of hypothermia were evaluated for comatose survivors after OHCA due to aneurysmal subarachnoid hemorrhage (SAH).

Following return of spontaneous circulation (ROSC), therapeutic hypothermia was induced for comatose SAH patients except for those with devastating brain damage on brain CT and cardiac arrest over 10 minutes. Immediately after diagnosis and evaluation of cardiac function, cooling was promptly initiated by nasogastric lavage with iced water and surface cooling under general anesthesia. The ruptured aneurysm was obliterated by surgical clipping with wide decompressive craniectomy. Core temperature was maintained at 33-Urokinase was injected via cisternal drain and nicardipine and fusdil hydrochloride were intravenously administered to prevent cerebral vasospasm. Clinical outcome was assessed according to the Glasgow Outcome Scale (GOS) 3 months later.

Six women, aged between 44 and 62 years, were eligible during the past 12 years. Their Glasgow Coma Scale was 3 after resuscitation. Electrocardiogram on arrival was asystole in 2 and pulseless electrical activity in 4 patients. Myocardial stunning was detected in 5 patients by echocardiogram. Surgery and hypothermia treatment were uneventfully conducted. Postoperative MRI revealed extensive cerebral ischemia in 2 and vasospasm-related ischemic lesion in 1 patient. Their GOS was good recovery in 1, severe disability in 2, persistent vegetative state in 1, and death in 2 patients.

Therapeutic hypothermia was feasible for OHCA patients due to SAH. Since neurogenic stunned myocardium could be a possible cause of cardiac arrest in SAH, beneficial effects of induced hypothermia are expected just like cardiogenic cardiac arrest. Appropriate prognostication methods are warranted for decision making to treat or not.

Autonomic shift (AS), characterized by increased sympathetic nervous system activation, has been implicated in neurologically mediated cardiopulmonary dysfunction and immunodepression following stroke. However direct measurement of autonomic nervous system dysfunction is difficult to obtain routinely in critically patients. We investigated the prevalence of AS defined by readily available clinical parameters and determined the association of AS with subsequent infection in a cohort of patients with aneurysmal SAH (aSAH).

Data were obtained from a single center cohort study of aSAH patients admitted from January 1, 2007 through April 1, 2012. AS was defined as at least one early routine clinical marker of neurologically mediated cardiopulmonary dysfunction (based on electrocardiogram, echocardiogram, cardiac enzyme testing or clinical diagnosis of neurogenic pulmonary edema). Exclusion criteria were beta-blocker treatment a known pre-existing abnormal electrocardiogram. Multivariable logistic regression models were developed to evaluate the association between AS and subsequent infection after adjusting for other covariates.

A total of 167 patients were included (mean age 56, 27% male). Autonomic shift was seen in 66/167 (40%), and infection was seen in 80/167 (48%). Autonomic shift was associated with subsequent infection on unadjusted analysis (OR=2.11, 95% CI 1.12, 3.97). However, on multivariable analysis adjusting for other predictors of infection, there was no significant association between AS and subsequent infection (OR 1.36, 95% CI 0.67, 2.76). Age, clinical grade, aneurysm location and presence of ICH were all identified as independent predictors of infection following aSAH.

We identified evidence suggestive of AS based on readily available clinical markers in 40% of patients with aSAH. However, AS defined by these clinical criteria was not an independent predictor of infection. Additional studies may be warranted to determine the optimal definition of AS and to determine the clinical significance of this finding.

We have previously studied the effects of falling temperature on the incidence of aSAH at our institution over 6737 days observing 1,175 aSAH. We previously reported that every degree decrease in temperature was associated with 0.6% increase in risk of aSAH [relative risk (RR), 1.006, P = 0.016]. We looked within the same data using other metrics to identify patterns in temperature changes which might result in physiological stress that increases the incidence of aSAH admissions at our institution.

We developed a mathematical equation based on the premise that 70 degrees Fahrenheit is the ideal external temperature for humans. Our formula measured the variation above or below 70° as a percentage of 70° for every day of 6737 days of observation. The relationship of absolute differences between Tmax and Tmin was examined to see if daily temperature variation was associated with increasing incidence.

The odds ratio for incident aSAH relative to 70° was 0.746 (CI 0.587 -0.948) p= 0.016. Likelihood of incident aSAH increased as the ratio of Tmax to 70° fell below zero (i.e. experienced colder temperatures). Intraday variation as measured by the absolute difference between Tmax and Tmin was strongly associated with increasing incidence, p=0.001, OR 0.987, CI (0.979-0.995). A smaller, not larger, difference between Tmax and T min was associated with increased likelihood of aSAH admission.

Colder daily maximum temperatures relative to 70° F, and smaller intraday temperature fluctuations are associated with increased aSAH admissions at our institution. Smaller daily temperature ranges correspond to seasonal periods with the least daylight in this region, and may represent sudden arrival of cold weather in warm months. Both metrics support the hypothesized increased likelihood of aSAH with falling environmental temperatures. These new methods may assist in the development of new algorithms for aSAH predictions based on temperature.

Near-infrared spectroscopy (NIRS) is a noninvasive means of measuring cerebral regional mixed arteriovenous (AV) brain oxygenation. We hypothesized that frontal NIRS would correlate against more established modes of vasospasm monitoring and systemic variables for severe aneurysmal subarachnoid hemorrhage (aSAH).

Case Report We describe a 23 year old male who presented with coma (GCS= 8, E4M3V1T) after severe aSAH (modified Fisher 4) from a ruptured giant basilar aneurysm (2.2 cm x 2 cm) who developed severe diffuse vasospasm with no change on clinical examination. Frontal NIRS monitoring was applied in addition to MAP, CPP, CBF (Hemedex TM ), cardiac output (CO), SpO2, core temperature, continuous quantitative EEG (qEEG) with alpha delta ratio (ADR) monitoring, along with daily TCD. The patient developed severe diffuse vasospasm and underwent angioplasty of the MCA, ACA, and PCA arteries and received intra-arterial verapamil. Pearson's correlation coefficient was used to analyze trends in variables pre-and post intervention.

Values were recorded over a four-day period. Calculated correlation coefficients revealed invasive CBF to right NIRS r=0.375 (p=0.002) and left r=0.162 (p=0.2) but was contralateral to the CBF probe, CO to right NIRS r=0.692 (p=0.04) and left r=0.620 (p=0.07). Coefficients with weak or negative correlation included arterial MAP to right NIRS r=-0.492 (p=0.26) and left r=-0.241 (p=0.56), noninvasive MAP to right NIRS r=-0.210 (p=0.59) and left r=0.214 (p=0.58), SpO2 to right NIRS r=-0.349 (p=0.36) and left r=-0.227 (p=0.56). Noninvasive MAP to arterial MAP r=0.713 (p=0.04), noninvasive MAP to CPP r=0.781 (p=0.01), and arterial MAP to CPP r=0.869, (p=0.005).

NIRS correlates with ipsilateral invasive CBF values (r=0.37, p=0.002) and trends with cardiac output. NIRS did not correlate with MAP, CPP, SpO2, TCD or qEEG ADR data. Larger prospective studies are needed to validate these preliminary results.

This case report describes the use of intraventricular nicardipine in a pediatric patient for the treatment of severe cerebral vasospasm following SAH from traumatic PICA dissection. Intraventricular nicardipine has been suggested as an adjuvant to standard therapies in adults with aneurysmal SAH but its use has not been described in pediatric patients.

A 12 year-old boy was transferred from an outside hospital for treatment of severe SAH following sports related head injury. He was found to have a dissecting PICA psuedoaneurysm which was treated endovascularly. Bilateral ventricular drains had been placed for hydrocephalus. His neurological examination declined on hospital day 9 and CT angiogram demonstrated severe vertebrobasilar vasospasm. Intraventricular nicardipine was administered in addition to treatment with transluminal balloon angioplasty, induced hypertension and nimodipine.

The patient received 2 mg intraventricular nicardipine twice daily for 2 days and the dose was then increased to 4 mg every 6 hours for a total of 8 days. Both ventricular drains were clamped for 30 min following administration. He tolerated doses without hemodynamic effects, elevations in intracranial pressure or evidence of ventriculitis. After improvement in clinical examination and mean cerebral blood flow velocities by TCD, intraventricular nicardipine was stopped. He was discharged to acute rehab and was ambulatory and preparing to restart school at age appropriate grade level at 3 month follow up.

Intraventricular nicardipine was safely administered in this 12 year-old patient with severe vasospasm following SAH with a good outcome. Intraventricular nicardipine should be considered as an adjuvant to standard therapies for vasospasm in pediatric patients, though further studies are needed to evaluate safety and efficacy in both pediatric and adult patients.

The benefit of early tracheostomy has been well described. Patients with aneurysmal subarachnoid hemorrhage (aSAH); however, represent a distinct population to which traditional weaning parameters may be difficult to apply. The purpose of this study is to identify admission characteristics of aSAH patients that predict need for tracheostomy.

This was a retrospective cohort analysis of 209 consecutive aSAH patients. We excluded patients with a history of symptoms longer than 72 hours prior to transfer, expired within 72 hours, or no CT scan available prior to cerebral angiography. We collected data including: demographics, co-morbidities, neurologic exam, labs, ejection fraction % on echocardiogram, modified Fisher scale, and Hijdra scale. Chi-square or Wilcoxon tests were performed where appropriate with subsequent multivariate analysis of statistically significant variables.

The data set included 35 tracheostomy patients and 174 non-tracheostomy patients. [5] [6] [7] [8] [9] [10] [11] [12] ,p=<0.0001). The modified Fisher and all components of the Hijdra scale were significantly higher in the tracheostomy group. The bicaudate index was significant (0.16vs.0.14,p=0.0224); however, presence of hydrocephalus using this index was not. In the multivariate analysis older age, lower albumin, higher pCO2 and presence of ventricular blood by Hijdra scale remained significant predictors.

Neurologic status on admission, advanced age, burden of systemic illness, and intraventricular hemorrhage are associated with increased risk of tracheostomy. Further research in this patient population on the benefits of early tracheostomy (lower mortality, less ventilator days and less intensive care unit days) is warranted.

Patients with subarachnoid hemorrhage(SAH) have variable outcomes, some of these leading to major disability. Established guidelines advocate administration of nimodipine to patients with SAH. Several recent trials have investigated the utility of statins and magnesium, however there has not been much data showing clinical benefit. We present data of patients with primary SAH who had therapy with magnesium, nimodipine and simvastatin for prevention of delayed cerebral ischemia (DCI).

Patients with primary SAH admitted to two academic institutions between July 2009 and April 2012 were identified. All patients received therapy with magnesium, nimodipine and simvastatin for DCI prophylaxis. Outcomes were categorized as good in those with Glasgow Outcome Scales (GOS) of 4-5 and poor in those with GOS 1-3. Chi-square analysis was used to compare outcomes between age, sex, race, Hunt-Hess scores (1-3 vs 4-5), presence of vasospasm and Glasgow Coma Scale (GCS) on presentation (below or above 12).

Of 149 patients identified with primary SAH, 65.77% had a good outcome. The mean age for patients with a poor outcome was 60.76 (SD 13.95) when compared to 53.95 (SD 13.12) in patients with a good outcome (p=0.004). Among those with a GCS 13-15 on admission, 84.7% had a good outcome while in those with less than 12 only 15.3% had a good outcome (p<0.0001). When comparing Hunt-Hess scales, 81.9% of those with grades between 1-3 had good outcomes, while 73.6% of patients with grades 4-5 had poor outcomes (p<0.0001). Among those who developed DCI, 24.5% had a good outcome as compared to 38.8% had poor outcomes (p=0.036). Age and race failed to show any difference in patients with good and poor outcomes.

We present our data on therapy with nimodipine, magnesium and statin for prophylaxis against DCI. Age, admission GCS, Hunt-Hess scale and occurrence of DCI were predictors of patient outcome.

Intraoperative rupture during the surgical treatment of a previosuly unruptured intracranial aneurysm is a rare event. We describe our experience with intraoperative aneurysm rupture in this setting.

We reviewed all cases of unruptured aneurysms treated by a single surgeon from July, 1997 to June, 2011 and identified those patients who suffered intraoperative aneurysm rupture.

Of 1275 unruptured aneurysms treated during this period, there were 6 instances of intraoperative aneurysm rupture (0.47%). In our experience, rupture occurred during dissection of either a perforator (2 cases) or a major efferent vessel (2 cases) from the aneurysm dome or of the dome from adherent overlying cortex (1 case). In one instance, the aneurysm ruptured during removal of the anterior clinoid process. In 4 cases, blunt rather than sharp dissection was being employed. In 3 cases, bipolar electrocautery and gentle tamponade successfully sealed the rupture point. In 2 cases, a clip placed across the bleeding site well up on the dome of the aneurysm controlled the bleeding and allowed for completion of the dissection and proper clipping. In the last case, the administration of adenosine was utilized to stop the bleeding and allow for proper clip placement. Intraoperative angiography confirmed adequate aneurysm obliteration in each case. There were no clinical consequences associated with these intraoperative ruptures.

Intraoperative rupture during elective surgery for a previously unruptured aneurysm is uncommon. In our experience, rupture was typically associated with blunt dissection on the dome of the aneurysm. The use of bipolar electrocautery, clipping of the bleeding point, or intravenous adenosine infusion were successfully used to control bleeding in our cases. The neurovascular surgeon should be prepared to address this unlikely event, should it occur.

Aneurysmal subarachnoid hemorrhage (aSAH) is a life-threatening form of hemorrhagic stroke which is more common in women than men, typically between ages 30-60. Over the course of our nursing practice, we have observed a trend of pre-menopausal aSAH female patients who experience the onset of their menses during the initial week of hospitalization. We became curious as to whether there is a correlation between aSAH and an earlier onset menses than a normal 21-28 day cycle.

Retrospective, single-center review of the medical record of pre-menopausal females ages 18-55 years who were admitted to our Neuroscience Intensive Care Unit with the diagnosis of aSAH. Chart review was specific to documentation of the onset of menses during the first week of hospitalization, medical/gynecological history with regard to last menstrual period, usual menstrual cycle characteristics, contraceptive use, past surgical history, and pertinent medications.

Over a 13 month period (June 2011-June 2012), we identified 103 aSAH patients with 58 being female. Of the female aSAH charts screened, we found 15 study patients. Nine of 15 (60%) females had documentation of starting their menses during their initial week of hospitalization for aSAH, much earlier than a normal range of the menstrual cycle of 21-28 days. One patient had menses documented on hospital day 12.

This small retrospective study suggests that aSAH may disrupt the "normal" menstrual cycle of pre-menopausal females. To our knowledge, this is the first description of this gender and disease specific phenomenon. A prospective study is planned to better understand the role aSAH has on hypothalamic-pituitary-ovarian-uterine physiology.

Introduction 10-20% of patients with spontaneous subarachnoid hemorrhage can have normal cerebral angiogram. Vasospasm, hypoperfusion or thrombosis may hide the aneurysm. DynaCT is a promising new technique which may help in these cases. We present a case report that highlights the ability of DynaCT in identifying a thrombosed aneurysm that was undetected with routine cerebral angiogram.

A 60 year old female presented with the worst headache of her life. CT scan of the brain showed subarachnoid hemorrhage (SAH) in suprasellar cistern, extending into the anterior interhemispheric fissure, bilateral perisylvian, prepontine cistern and right perimesencephalic cisterns along with extension in to the third and fourth ventricles. After the placement of an external ventricular drain, the patient was immediately taken to angio suite where a biplane cerebral angiogram showed 1-2 mm saccular aneurysm at right middle cerebral artery bifurcation and an unremarkable vasculature otherwise. Repeat imaging using DynaCT showed the presence of a 3 mm ruptured and thrombosed aneurysm at the right MCA bifurcation. The thrombosed aneurysm was visualized and clipped surgically.

This case report highlights the promising utility of DynaCT in identifying the culprit aneurysms.

Treatment of severe cerebral vasospasm in subarachnoid hemorrhage remains challenging. With failure of noninvasive therapy, endovascular modalities may be undertaken, albeit with limited efficacy; balloon angioplasty can be used only for proximal, focal spasm and intra-arterial calcium-channel blocker (CCB) bolus infusion has transient vasodilatory effects. We present a patient with severe vasospasm after subarachnoid hemorrhage, who demonstrated significant angiographic improvement with continuous infusion of intra-arterial verapamil over 24 hours.

A female in her mid-30's with sickle cell anemia presented with a Hunt and Hess 5, Fisher grade IV subarachnoid hemorrhage secondary to a ruptured right posterior communicating artery. On initial assessment, the patient was localizing with only her upper extremities. The aneurysm was completely coil embolized and standard triple-H therapy maintained. On post-bleed day 7, the patient developed left-sided hemiplegia. Angiography demonstrated critically severe, diffuse right anterior and posterior circulation vasospasm. Angioplasty could not be performed due to microwire and balloon inaccessibility of stenosed anterior and posterior circulation vessels. Subsequently, two microcatheters were positioned with their respective tips in the petrous right internal carotid artery (ICA) and V2 segment of the right vertebral artery for continuous machine controlled intra-arterial verapamil infusion. Dosing consisted of administering 2 mg/hr verapamil into the right vertebral artery and 4 mg/hr into the right ICA. The patient was placed on a heparin drip and taken to the Neurointensive care unit for monitoring.

After 24 hours of continuous IA verapamil infusion, angiography demonstrated significant improvement in right anterior and posterior circulation vasospasm, with only residual diffuse moderate stenosis. Unfortunately, no corresponding clinical improvement was noted.

Prolonged infusion of intra-arterial CCB's may provide extended angiographic improvement in severe vasospasm refractory to conservative treatment and unsuitable for balloon angioplasty. With systematic study of such techniques, optimal agents and dosing for sustained vasodilation and clinical optimization may be defined.

Vasospasm remains a significant cause of morbidity after subarachnoid hemorrhage (SAH), inducing delayed ischemic events. SAH typically results in numerous complications including severe, treatment-refractory headache. Fioricet® (acetaminophen 325mg/butalbital 50mg/caffeine 40mg) is a commonly used analgesic medication for the treatment of headache in SAH. Caffeine has been shown to reduce cerebral blood flow. The purpose of this study was to determine if there is an association between Fioricet® administration and early vasospasm.

A retrospective, medical record review was conducted, and patients were identified using the University Health Consortium (UHC) Database. Patients were included if they had an aneurysmal SAH with a presenting Hunt and Hess Grade of I-IV. Data points included occurrence of clinical vasospasm, daily amount of Fioricet® and other analgesics, daily pain scores, and patient demographics. A univariate analysis was performed to determine the association between extent of Fioricet® exposure and early vasospasm (within the first 7 days) after SAH. A multivariate analysis was performed accounting for amount of Fioricet® use, patient age, and Hunt and Hess Grade.

The population characteristics were typical of the SAH population. Patients who experienced clinical vasospasm received more Fioricet® than those who did not have vasospasm (mean 5.09 + 3.2 tablets/day versus 3.66 + 3.3 tablets/day (p=0.02)). The odds ratio for vasospasm with regards to Fioricet® use when controlled for age and Hunt and Hess grade was 1.1 (95% CI 0.96 -1.27). The multivariate analysis did not yield any statistically significant associations with vasospasm.

There was a significant association between Fioricet® exposure and vasospasm in our univariate analysis. However, when correcting for age and SAH severity, the association is not significant. Thus, the data do not currently support a clear causal association. This preliminary data will be used to support a comparative study investigating headache treatment in SAH.

Isolated complete third nerve palsy (TNP) in the setting of a subarachnoid hemorrhage (SAH) is most commonly seen secondary to a posterior communicating artery (PCOM) aneurysm. However, this same clinical picture with a negative angiogram and otherwise negative imaging studies becomes extremely rare. Although trauma has been described as one of the most common causes of isolated TNP, concomitant post-traumatic SAH and late onset isolated complete TNP has never been reported. We report a case of a delayed onset complete TNP after traumatic SAH.

Case Report.

A 61 year-old male with type-2 diabetes mellitus presented to the Emergency Department with painless diplopia and left eye ptosis three days after sustaining a fall with closed-head injury without loss of consciousness. His non-contrast head CT scan showed a Fisher 4 grade subarachnoid hemorrhage. Upon arrival and throughout his hospitalization, the patient had a Glasgow Coma Scale (GCS) of 15. His neurological exam revealed findings consistent with isolated complete thirdnerve palsy (TNP) involving the pupil. His neurological examination was otherwise normal. Diagnostic digital subtraction angiography (DSA) was negative as it was his brain MRI for aneurysm or vascular lesion. MRI did however show traumatic SAH pattern and small subdural hematomas consistent with trauma. Laboratory findings (ESR, CRP, ACE, c-ANCA and p-ANCA) did not raise suspicion of secondary vasculitic or ischemic causes of TNP. The patient was discharged five days after admission with no further complications but without any improvement of TNP signs and symptoms.

This case illustrates an atypical presentation of traumatic SAH with delayed-onset, isolated complete TNP. To our knowledge, this is the first case with these features described in the literature. His atypical presentation may represent the combination of both diabetes and traumatic injury to the cranial nerve III in the subarachnoid space, rather than either etiology alone.

Diringer Section of Neurocritical Care

Improved clinical outcomes after aneurysmal subarachnoid hemorrhage (aSAH) have been demonstrated for patients treated at high volume centers. These centers treat only 50% of all aSAH. It is common for aSAH patients to be transferred to high volume comprehensive stroke centers after presentation to a community hospital. This study aims to determine if the hospital of presentation has impact upon aSAH outcomes.

A 3-year retrospective analysis of aSAH treated in a comprehensive stroke center was undertaken. The comprehensive stroke center consisted of a neurocritical care unit, dedicated vascular neurosurgeons, and endovascular and neurocritical care specialists. Demographic and outcome data were collected on all aSAH patients who had a confirmed and secured aneurysm, survived > 10 days from admission, and completed TCD monitoring and observation for complications of vasospasm. Univariate and multivariate analyses were evaluated for differences in mortality, complications, incidence of vasospasm, discharge disposition, and length of stay.

107 patients were included (31 direct and 76 transfer). Baseline parameters known to influence outcome (age, medical complications, Glasgow Coma Scale, Fisher and Hunt and Hess grade) were similar between the two groups. Transferred patients developed ultrasound defined vasospasm more frequently (58% versus 32%; p<0.05) and had a greater delay in time to surgery (3.9 versus 2.4 days; 0<0.05). Adjusting for key predictors, direct admit patients spent 3.9 fewer days in the ICU compared to transferred patients (t=-2.7, p=0.009). Multivariate analysis showed that the likelihood of vasospasm was significantly higher for transfer patients (OR 3.46, CI: 1.2-10.3, p = 0.03). Longer in-hospital stays and decreased rates of home discharge were observed in transferred patients (p<0.01). Mortality rates were not statistically different (transfer 22.6%, direct 14.5%, p=0.31).

aSAH patients admitted directly to a comprehensive stroke center have better outcomes than those transferred from lower acuity facilities. 

Numerous advances have been made in the management of subarachnoid hemorrhage (SAH) and its complications, including symptomatic vasospasm. However, the optimal management of vasospasm in patients without neurological deficit remains uncertain. We performed an electronic survey of members of the Neurocritical Care Society (NCS) to elucidate clinical practice in this regard.

An electronic survey with ten questions about different aspects of SAH management was formulated. Our institutional review board and NCS approved the survey. Three scenarios were presented for good grade SAH patients without evidence of delayed cerebral ischemia (DCI): those with either normal TCD values, vasospasm on TCD, or vasospasm on angiography.

144 members answered the survey (response rate of 15%). Up to 45% of respondents utilized transcranial doppler (TCD) measurement to diagnose vasospasm, while 80% (95% CI, 72-86%) used clinical examination and 82% (95% CI, 75-88%) used angiography (CT or catheter). In good grade SAH patients with no evidence of DCI, 90% (95% CI, 84-95%) of respondents indicated using nimodipine in all three scenarios. In the subset with normal TCD values, 10% (95% CI, 5-16%) recommended use of hypervolemia, 4% (95% CI, 2-9%) hemodilution and 1% (95% CI 0-5%) induced hypertension. However, in the subset with vasospasm on angiography and no referable clinical symptoms, 50% (95% CI 42-58%) recommended the use of hypervolemia, 26% (95% CI, 19-34%) hemodilution, 48% (95% CI, 40-56%) induced hypertension and 43% (95% CI, 35-52%) endovascular therapy with intra-arterial vasodilators, angioplasty or stents.

From the sample above, it appears that good grade SAH patients without neurological deficit but radiological vasospasm are treated aggressively. This is not supported by current literature or guideline recommendations, which imply little benefit of aggressive therapy in such patients. Further studies are needed on the optimal management of this subset of patients, in whom the effects of vasospasm remain unclear.