key: cord-015126-cyhcbk1j
authors: nan
title: PS 0036-0344
date: 2007-08-25
journal: Intensive Care Med
DOI: 10.1007/s00134-007-0820-y
sha: 
doc_id: 15126
cord_uid: cyhcbk1j

nan

In our 21-bed ICU-cum-HDU of a 550-bed tertiary referral cancer centre, medical oncology admissions increased from <5% of total admissions to over 35% in the last 3 years. We audited outcomes in these patients to determine prognostic factors that may aid patient selection and management. METHODS. 263 consecutive admissions (170 males, 93 females, age > 16years) from February 1, 2006 to February 28, 2007 were prospectively studied. The total SOFA score on Day1 (SOFA1), the highest SOFA score of the first three days (MAX3) and the change in SOFA score between Day2 and Day1 (Delta1) and between Day3 and Day1 (Delta2) were calculated. Predictors of outcome were identified using univariate and multivariate binary logistic regression. RESULTS. 58 patients had solid tumours , 89 had leukemia, 84 lymphoma, 17 myeloma, and 15 had other diagnoses. Mean age was 44.8±16 years and APACHE II score was 21.2±7.9. ICU mortality was 71% and hospital mortality 78.7%. 67/74 patients (90.5%) with ICU stay < 1 day died. Overall length of ICU stay was 3.7±4.3days. In survivors vs. nonsurvivors, SOFA1, Delta1 and Delta2 (median, interquartile range) were 3.0(1.0 to 6.0) vs. 8.0(5.0 to 12.0; p<0.000), -1.0(-2.0 to 0) vs. 1.0(-1.0 to 3.0; p<0.000) and 0(-2.75 to 1.0) vs. 1.0(-1.0 to 5.0; p<0.02), respectively. Several factors were associated with mortality on univariate analysis (Table 1) . On multivariate analysis, only need for vasopressors (OR 9.14, p=0 .002) and MAX3 (OR 1.52, CI 1.19-1.93, p=0.001) were independently associated with hospital mortality, while type of cancer and leucopenia were not. For 141 patients staying >2 days, no factor predicted hospital mortality, but SOFA1 (OR 1.64, CI 1.01-2.68, p=0.047), Delta1 (OR 1.57, CI 1.15-2.14, p=0.004) and Delta2 (OR 1.25, CI 1.01-1.60, p=0.04) predicted ICU mortality. 

Usually, the CGR transfused in our ICU are old (about 80 % of RBC are stocked more than 14 days). ICU outcome is independently associated with the number of RBC transfused, but not with their age. This result in contradiction with previous report could possibly be explained by the systematic leucodepletion performed before storage in France, contrary to precedent studies where RBC were not leukodepleted. We compared them with ≥70 years old and an ICU stay < 30 days patients, the differences in ICU mortality, Apache II, age, gender and the necessity for renal replacement therapy (RRT) were not significant (see table) . The survivors patients (≥70 years old and an ICU stay ≥ 30 days) were more older and 21(65'62%) were still alive one year later. When we analyzed the overall patients, according their stay < or ≥ 30 days, did not find statistically significant differences between both groups in the mortality (p=0'610 CONCLUSION. ICU mortality rates in elderly patients with a stay < or ≥ 30 days at ICU were comparable. The 1 year-survival of elderly patients with a long-term intensive care unit stay was high.

RESULTS. Seventy patients were admitted to our ICU with the diagnosis of acute pancreatitis during study period and 35 of them were later confirmed as having SAP. The average ICU length of stay in patients with SAP was 17 days compared to 5 days in patients with mild form of the disease. Pancreatic infection was present in 8 patients. The mortality rate in the group with SAP was 31% compared to 2,8% in the group with mild acute pancreatitis, p< 0,05. The most common etiology of patients with SAP was biliary and this was similar both in survivors and non-survivors. The most common cause of death in the group with SAP was multiple organ dysfunction/failure syndrom(MODS/MOF) in 80% followed by bleeding complications in 20%. Twelve patients with SAP (34%) underwent the surgical intervention. Mortality in the group of patients who underwent a surgical intervention was 25% (3 patients). 55,0+/-17,5 56,3+-14,08 APACHE II score( mean+/-SD) 20,2+/-11,3 10,0+-6,9* Necrotising form (%) 90 88 Infected necrosis (%) 20 24 CT guided FNAB (%) 30 32 * p > 0.05 CONCLUSION. The patients with mild form of acute pancreatitis had low mortality rate (similar to general ward population) despite positive ICU admission criteria in our case series with fifty per cent development of severe form with organ dysfunction/failure later on. APACHE II score was better predictor of mortality in patients with SAP than presence, extent or infection of pancreatic necrosis. Patients with higher risk for development of severe form of acute pancreatitis should be admitted to multidisciplinary ICU prior to definitive diagnostic evaluation of pancreas. Further studies are warranted. CONCLUSION. ABSI is an aprropiate score for estimating the probability of death in critical brun injury patients. Preexisting cardiac and liver diseases have a little influence on mortality and its addition to the ABSI variables don't predict mortality more accurately.

Poisoned patients constituted up to 10,5 % of all ICU admissions in our hospital. Demographic data and specific poisons have been presented at the table. The total poisoned mortality rate was 10,2 %. Methyl alcohol poisoning has a higher mortality than others poisoning. CONCLUSION. Childhood poisoning is usually accidental and is usually associated with a low morbidity and mortality. In adults, self-poisoning is usually deliberate suicide or parasuicide) and has a higher morbidity and mortality rate.

(1) The most important part of the poisoned patient's care are the general supportive management and specific antidotes therapy. 

It has abundantly been demonstrated that duration of mechanical ventilation can be reduced by the use of protocols for weaning and sedation [1, 2] . Utilization of the required sedation scales and adherence to protocols, however, is poor in daily practice, as has been shown in recent studies [3, 4] . It has been proposed to use daily checklists to improve the quality of care [5] .

To improve adherence to the established guidelines for weaning and sedation in our ICU, we included two questions in a checklist printed on patients' charts which had to be answered daily by the physician on duty: CONCLUSION. The checklist as a daily reminder to observe established weaning and sedation protocols may have significantly accelerated weaning from mechanical ventilation.

We carried out a prospective and descriptive study in patients admitted to our ICU from 1993 to 2004. We defined tolerance as the need to use more than 300 mg/h, at least for four hours, or the need either to use or to change to other sedatives to obtain a 3 to 5 level on the Ramsay scale (1). The appearance of tolerance in the first 48 hours was considered as tachyphylaxis or early therapeutic failure to this sedative. In our sedation protocol we use propofol preferably in patients who need frequent neurological consciousness evaluations, or in patients whose sedation is expected to have short to medium duration, and who have haemodynamic stability. Also, we use propofol as a sequential strategy when early weaning from ventilation is expected. All patients received analgesic drugs.

During this time, we admitted 5277 patients, 2519 of them needed mechanical ventilation and in 2005 patients we administered continuous analgesic and sedative infusions. Continuous propofol infusions were administered in 1651 patients at some point of their sedative strategy, and 683 (34% of the sedated patients) received propofol for more than 48 hours. Tolerance development was observed in 48 patients, 3% of the patients sedated with propofol.

In thirty-seven of them, this situation was present in the first 48 hours (early therapeutic failure).

CONCLUSION. In our sedative protocol for propofol use, the incidence of tolerance in patients sedated with this drug was 3%, which is substantially less than the usual described midazolam tolerance. Most of these cases (77%) happened in the first 48 hours. 

Diabetes mellitus (DM) with its chronic and acute complications puts patients suffering from the disease at increased risk. None of the scoring systems used for risk prediction in intensive care units accounts for diabetes as a risk factor, although, in everyday practice, patients with DM admitted to ICUs may be recognized as those with higher risk. Not much data is available on how much risk can be attributed to diabetes. We have compared course and outcome of patients with DM with non-diabetics to try to answer this question.

We have analyzed data from the "croicu.net", national pilot-project which collects data on patients from ICUs in Croatia. Data collected during the first 26 months (Nov 2004 -Dec 2006 have been analyzed. Adult patients from 14 ICUs in university hospitals were included; three most frequent admission diagnoses were selected for comparison of diabetic and non-diabetic patients. The diagnosis of DM had to be established prior to admission according to the usual criteria. ICU mortality and ICU length of stay (LOS) were primary outcome measures. Incidence of organ failure was a measure of disease course.

In the analysed period there were 6456 admissions to the analysed ICUs, 1341 (20.8%) with documented DM prior to admission. Patients with MD did not differ significantly from non-diabetics in age or sex distribution. Overall mortality was higher for DM patients (15.1% vs. 12.7%), as was LOS (5.8 vs. 4.1 days). Three most frequent diagnoses were: sepsis (N=723; 11.2%), pulmonary oedema (659; 10.2%) and myocardial infarction (N=555; 8.6%). Patients with diabetes had significantly higher mortality and higher LOS in all three subgroups. In the sepsis subgroup, patients with diabetes had higher incidence of organ failure and higher number of failing organs. In the other two subgroups, the differences were not significant. In multivariate analyses which was performed separately for all three diagnoses and included DM, age, APACHE II score and SOFA score, diabetes mellitus was shown to be an independent predictor of mortality and LOS in all three cases.

Although some chronic effects of diabetes mellitus can be included in multiparameter scoring systems such as APACHE II score, the disease itself is not scored. We have shown on three most common diagnoses in ICUs of university hospitals that diabetes mellitus is an independent predictor of mortality and LOS and that it has significantly higher incidence of organ failure in sepsis. Patients with DM should be given appropriate attention as high risk patients in the ICU.

INTRODUCTION. Neuromuscular abnormalities are common in critically ill patients with systemic inflammation and organ failures. We assessed the incidence of a clinically diagnosed critical illness polyneuro-myopathy (CIPM), and its potential impact on mortality and long-term neurological outcome.

METHODS. 39 consecutive critically ill patients on mechanical ventilation for 48 hours and with the presence of 2 or more SIRS criteria were prospectively studied. Based on daily clinical neurological examinations, CIPM was defined as symmetric limb muscle weakness [2 or more muscle groups, M3 or less (MRC)] without other explanation than CIPM in patients with normal neurology at ICU admission. A Barthel index (score for activities in daily living) was performed at day 28 and 6 months after ICU discharge. After 6 months a neurological examination was also performed.

. CIPM was diagnosed in 13 patients (33%). Patients with suspected CIPM had a prolonged ICU stay and a high mortality. The Barthel index was significantly lower in this group at day 28 but improved over the next six months. 20 of 23 patient who survived could be reached 6 months after discharge and 12 of them were clinically examined. At this time the most compromised activity in daily living is climbing stairs. Patients with a clinical diagnosis of CIPM have a high mortality. If they survive, they are severely limited in simple daily activities one month after ICU discharge, but improve later.

Host infection by pathogens triggers innate immune response leading to a systemic inflammatory response, often followed by a paradoxical compensatory antiinflammatory response. This immune dysfunction can impair the eradication of primary infections and favor the emergence of nosocomial sepsis. Dendritic cells (DCs) have a central role in initiation and control of innate and adaptative immune responses to infectious challenges. DCs might contribute to sepsis-induced immunodepression. Indeed, depletion of DCs has been reported in secondary lymphoid organs of patients who died from sepsis and in animal models of lethal sepsis. In order to investigate the mechanisms of sepsis-induced immunodepression, we studied quantitative and functional features of DCs in a murine model of sublethal sepsis.

We developed a sublethal murine model of polymicrobial sepsis through cecal ligature and puncture followed by short course of antibiotics and volume resuscitation. We isolated splenic DCs by immunomagnetic procedure and generated bone marrow-derived DCs (BMDCs) by 6-day culture of medullar progenitors in the presence of GM-CSF before stimulation with LPS to induce maturation. We counted spleen DCs and studied the following functional features of spleen DCs and BMDCs in the early (day 1) and late (day 8) phases of sepsis : maturation (expression of MHCII, CD40 and CD86 through FACS analysis), production of cytokines (TNF-alpha, IL-12, IL-10) and priming of CD3-positive T-cell lymphocytes (3H-thymidine proliferation assay in allogeneic mixed lymphocyte reaction).

Upon anesthesia induction with isoflurane sepsis was initiated by cecal ligation and double puncture in 3 groups of 4 C57BL/6J-mice per group [18G, 22G, 26G] (CLP). Control mice underwent laparatomy and manipulation of the cecum only (Sham). 24, 48 and 96 hrs post-surgery in 26 and 22G mice and 36 hrs post surgery in 18G mice single cell suspensions of thymus and spleen were analyzed by means of cell surface staining and flow cytometry. Fluorescence-labeled antibodies included CD3, CD4, CD8, B220, IgM, IgD, CD25, CD69. Data are presented as Mean+SEM.

RESULTS. Similar to previous results, thymi primarily demonstrated a time-dependent reduction of CD4+CD8+ double-positive cells which was more pronounced during severe sepsis (18+22G). At 96hrs post-CLP CD4+ cells and CD8+ T-cells recovered to values of sham mice in 26G animals, which previously recovered fastest with highest survival rates of about 90%. In contrast CD4+ cells and CD8+ cells, respectively, raised to maximum levels at 96 hrs in 22G animals. Concerning Spleocytes CD4+ and CD8+ cells were similarly reduced to about 80% and 70% after 48hrs and to 60% and 50% in 22G and 26G mice compared to sham mice. Splenocytes of 18G-treated mice, which could only be investigated at 36 hrs postCLP showed no difference to sham mice. As far as B cells are concerned no significant differences between the groups or different time points could be detected. Relative numbers of peripheral T cells expressing the early activation marker CD69 or CD25 were clearly more pronounced at 24hrs compared to 96hrs in 26 and 22G mice. In 18G treated mice CD69 and CD25 positive T cells were significantly higher at 36hrs compared to sham mice.

A mild CLP model is more appropriate to study during murine sepsis. The rapid occurrence of peripheral activated T cells suggest a very early function of the adaptive immune system during sepsis. Considering a milder disease course of 26G mice they seem to more efficiently use their T cells to fight the infection. Thymocyte data suggest a block in lymphopoiesis from CD4-CD8-to CD4+CD8+. B cells are not likely to play a major role in polymicrobial murine sepsis. Further studies have to be performed to elucidate the turnover and the homing of lymphocytes during sepsis. 

Endotoxaemia is associated with intestinal perfusion deficits and gut barrier failure. Regional sympathetic blockade by means of thoracic epidural anaesthesia (TEA) has been shown to positively affect intestinal microcirculation during endotoxaemia. This study tests the hypothesis that the microvascular changes observed with TEA go along with an increase in overall gastrointestinal blood flow. In addition we investigated whether the use of TEA influences gut barrier function.

After approval of the animal care committee rats were anaesthetised (urethane/ketamine), hemodynamically monitored and mechanically ventilated with room air. Lidocaine 2% or normal saline were administered as a bolus (30 µl) and subsequent continuous infusion (30 µl x h −1 ) via an epidural catheter (tip at T7/8, spread T4-T10). Organ blood flow (n = 30 rats) was measured by the fluorescent microspheres technique at baseline, 30 min after epidural infusion, and 60 min and 120 min after the infusion of endotoxin (E. coli lipopolysaccharide, 1.5 mg x kg-1 x h −1 ) or normal saline. For assessment of gut barrier failure rats (n = 27) received a bolus infusion of endotoxin (50 mg x kg-1) or normal saline and epithelial permeability to low molecular fluorescein isothiocyanate-dextran (4kD) was quantified using a ligated loop of terminal ileum after 5 hours of normotensive endotoxaemia. 

In hypodynamic shock models pure O2 breathing was shown to redistribute blood flow in favour of hepato-splanchnic organs and to improve survival. In contrast, this therapeutic approach has not yet been evaluated in hyperdynamic septic shock, since an increased production of O2 radicals, which is directly related to the increased O2 partial pressure, is considered as harmful. Therefore, we investigated the effects of pure O2 breathing on hepato-splanchnic macro-and microcirculation, energy balance and tissue cell death during porcine fecal peritonitis.

After induction of fecal peritonitis, pigs were randomly ventilated for 24h with 100% O2 (n=10) or an FiO2 adjusted to yield a SaO2>92% (n=10). Before as well as at 12 and 24 h of peritonitis we measured cardiac output as well as hepatic artery and portal vein (pv) flows (ultrasound flow probes), microcirculation in the intestinal wall (Laser Doppler flow), intestinal wall oxygenation, portal and hepatic-venous acid-base status, and lactate/pyruvate (L/P) ratios. Apoptosis was analysed post-mortem in liver biopsies with the TUNEL assay. Within group effects were analyzed using a Friedman ANOVA on ranks, intergroup differences with an unpaired rank sum test.

At the end of experiment the contribution of both pv and total liver blood flow to cardiac output was significantly higher in the hyperoxic animals than in the control group (Qliver/CO 20 (6;25)% vs. 25 (13;34)%, p=0.022; Qpv/CO 18 (6;23)% vs. 23 (13;32)%, p=0.043, respectively), which was concomitant with attenuated regional venous metabolic acidosis and lower hepatic-venous L/P-ratios. Intestinal wall microcirculation and oxygenation did not significantly differ between the two groups. The hyperoxic animals presented with a markedly reduced number of apoptotic cells in the liver.

Our results show that early 100% O2 ventilation redistribute blood flow in favour of the hepato-splanchnic system even in peritonitis-induced hyperdynamic septic shock. Furthermore, the hepatic energy balance is improved and the morphologic integrity of the liver better maintained under these conditions. GRANT ACKNOWLEDGEMENT. Supported by the Eli Lilly-ESICM Sepsis Elite Award, the Alexander-von-Humboldt-Stiftung, and the Deutscher Akademischer Austauschdienst

Glucocorticoids are known as strong modulators of immune response that play an important role in patophysiology of sepsis and inflammation. They have strong influence on the development of immune system, its effector functions, and trafficking of immune cells.The biological activity of glucocorticoids depends not only on their plasma concentration, the number of receptors and the responsiveness of the target cells but also on the local metabolism of glucocorticoids that is predominated by 11b-hydroxysteroid dehydrogenase (11HSD). Two isoforms of 11HSD are known. The isoform 11HSD1 operates in vivo predominantly as a NADPH-dependent reductase that locally increases glucocorticoid concentration (cortisol, corticosterone) by reduction their 11-oxo derivatives (cortisone, 11dehydrocorticosterone) . The isoform 11HSD2 is a sole NAD+-dependent dehydrogenase that inactivates biologically active glucocorticoids to their inactive 11-oxo derivatives. The aim of this study was to investigate peripheral metabolism of glucocorticoids in immune cells and tissues in experimental model of sepsis and inflammation.

Sepsis was induced in BALB/C mice and Wistar rats by intraperitoneal administration of lipopolysaccharide or pooled fecal inoculum. In these animals and in healthy controls we measured expression and activity of 11HSD1 in lymphatic nodes, peripheral blood leukocytes and alveolar macrophages. Activity was measured by incubation with corticosterone and 11-dehydrocorticosterone, following HPLC determination. The abundance of 11HSD1 mRNA was measured by semi-quantitative real-time RT-PCR.

For years etomidate has been known to cause adrenal insufficiency in the critically ill and is a confounder when studying corticosteroids in septic shock.

Subgroup analysis of a prospective, randomized, placebo-controlled study of corticosteroids in septic shock. Patients underwent a short high dose ACTH test before study drug administration. Patients received 11d treatment with hydrocortisone (HC) or placebo (P). The affects of etomidate administration on ACTH responsiveness and 28d mortality were studied.

RESULTS. 499 patients were enrolled. Overall 33.5% patients died in the HC group and 31% in the P group (p=0.57). In total 27% of patients received etomidate. 101 received etomidate before baseline [21% HC group + 20% P group] and in 36 after baseline [8% HC group + 6% P group]. Overall, more of the patients receiving etomidate were ACTH nonresponders [58% vs 42%] . No mortality differences was seen between patients receiving etomidate at any time during study and those who did not receive etomidate [34.3% vs. 31 .5%](p=0.61). There was a possible trend towards a difference in mortality between patients who received etomidate in the 72 hrs before randomisation [45% HC vs. 38% P] or not receiving etomidate during this time period [31% HC vs. 30% P](p=0.052).

Etomidate was commonly used in patients in the CORTICUS study. Etomidate was associated with an increased likelihood of adrenal hyporesponsiveness in all patients. There was no increase in mortality associated with etomidate administration at any time, there was a trend towards increased mortality in those who received it in the 72 hours before trial baseline. This result comes from an underpowered subgroup and should be considered exploratory.

D. Pestaña* 1 , E. Martinez-Casanova 1 , A. Buño 2 , R. Madero 3 , A. Criado 1 1 Anestesia-Reanimación, 2 Análisis clínicos, 3 Bioestadística, Hospital Universitario La Paz, Madrid, Spain INTRODUCTION. Steroids are indicated in septic shock patients when relative adrenal insufficiency is suspected. Our aim was to study if the measurement of total proteins (1) and eosinophil count (2) improves the accuracy of cortisolemia to predict the hemodynamic response to steroid treatment in this setting (3).

We analysed data from 66 consecutive surgical patients with criteria of septic shock receiving steroid treatment. Four criteria were chosen to define hemodynamic improvement based on the combination of noradrenaline (NA) withdrawal (at 24 and 48 h) and an increase of the hemodynamic index (HI = mean arterial pressure/NA dose) of 150% at 24 h and of 350% at 48 h. The accuracy of the baseline cortisolemia to predict the hemodynamic response to steroid treatment following the four criteria was determined by ROC curve analysis.

The largest area under curve (AUC) was found for the noradrenaline withdrawal or an increase of the HI > 350% at 48 h after starting the steroid treatment (Table 1) . This criteria was met by 35 patients (53%) and was associated with a lower mortality (25.7% vs 67.7%, p=0.001, 70% sensibility and 72.2% specificity). However, no clear cortisolemia cut-off value for the diagnosis of adrenal insufficiency based on the hemodynamic response could be found. Neither the basal proteins nor the eosinophils improved the accuracy of cortisolemia to predict a hemodynamic improvement. Mortality was also related to age (p=0.017), APACHE II (p=0.004) and SOFA score (p=0.005). Neither basal cortisolemia nor lactate were related with ICU mortality. 

Twelve septic shock patients admitted to the ICU < 96 hours after family consent were enrolled. We excluded all patients in use of steroids in the preceeding 6 months, etomidate, espironolactone, oestrogens, oral contraceptives, ketoconazole or any other drug known to suppress adrenal function; AIDS, pregnancy, history of disease of the hypothalamic-pituitaryadrenal axis, shock of other etiologies. After a baseline serum cortisol was obtained, a LD (1 ug) corticotropin stimulation testing was performed. Subsequently, serum cortisol at 30 and 60 min was measured. Four hours later, another BC was obtained. Then, a HD (249 ug) corticotropin stimulation testing test was performed and serum cortisol was again measured after 30 and 60 min.

RESULTS. Both Baseline serum cortisols were similar. Delta HD cortisol was higher than Delta LD cortisol (19.1±15.3 vs. 8.8±6.6ug/dl, p=0.009). Five patiens had a BC < 25 ug/dl, but only one showed RAI in both tests. Concordance between LD and HD tests was 66% (8/12). It was strong for responders (86%,6/7) but weak for non-responders to LD test (40%, 5/2).

The preliminary results of our study suggest that a LD test is a more sensitive test than a HD test. A further study comparing treatment of RAI defined by a LD or a HD test is still needed. 

The potassium channels (Kc), ATP-sensitive K+ (KATP) channels and calcium-activated potassium (BK) channels, may be implicated in shock induced vasoplegia. The aim of our study was to demonstrate that the potassium channels are overexpressed in experimental shock independently of the etiology.

Three rats models of shock were used : peritonitis by caecal ligation and perforation (CLP, n=14) observed at 18h, ischemia-reperfusion model (hemorrhagic shock + resuscitation + laparotomy, n=9) observed at 18h, and pressure fixed hemorrhagic shock (n=6) observed at 4h. These three models were compared to a control group. We performed quantitative real-time PCR (LightCycler technology -Roche -and SYBR Green -Sigma) and Western Blot on aorta and mesenteric arteries. We studied the expression of the vascular smooth muscle KATP channels -Kir 6.1 and SUR 2B subunits -and BK channels -BK alpha subunit. We assessed the inflammatory syndrome in studying iNOS expression.

We were able to detect Kir 6.1, SUR 2B, BK alpha and iNOS ARNm in both vessels.

Quantitative real-time PCR results (reference gene : beta-actine) CLP CLP IR IR HS HS aorta mesenteric aorta mesenteric aorta mesenteric iNOS 3.7 ± 0.7* 12.1 ± 3.5* 4.2 ± 1.2* 21.8 ± 11.6* 14.2 ± 3.1* 22.4 ± 7.5* expression Kir 6.1 1.7 ± 0.2 4.8 ± 1.2* 20.5 ± 5.2* 7.1 ± 1.6* 19.5 ± 2.0* 11.5 ± 4.8* expression SUR 2B

1.4 ± 0.2 3.2 ± 0.6* 6.8 ± 1.5* 2.3 ± 0.6* 5.3 ± 1.3* 1.6 ± 0.4 expression BK alpha 1.9 ± 0.2* 1.9 ± 0.4* 2.8 ± 0.4* 2.7 ± 0.6* 2.1 ± 0.5 1.1 ± 0.5 expression * : p< 0.05 vs control group CONCLUSION. Various potassium channels are activated and up-regulated during shock independently of the etiology. Thus, potassium channels likely play a major role in sepsis but also in prolonged and severe hemorrhagic shock and in ischemia reperfusion. (CARS) . A predominantly anti-inflammatory reaction induces immunosuppression with impaired host defense. Application of GM-CSF to patients with major surgery or sepsis has been proposed to improve host-defense. In this study we investigated the differential effects of GM-CSF production in an ex-vivo model. and LPS on the TNF-a.

Whole blood of 40 healthy donors (Age 16-72 years, mean 54 years) was used to determine optimal concentrations and incubation time for LPS. The immunomodulating properties of GM-CSF (Leukine ® (sargramostim), Berlex)) were investigated in whole blood of 28 healthy donors (36-65 years, Mean 51 years) and 12 ICU patients suffering from sepsis. Six of the patients had immunoparalysis as defined according to local standards by a monocytic HLA-DR expression of < 150 MFI and an ex-vivo stimulation test of < 175 pg/ml after LPS incubation (DPC Biermann, Bad Nauheim , Germany), whereas the other 6 displayed a HLA-DR expression of > 150 MFI and a ex-vivo stimulation test of > 175 pg/ml. Samples were primed either with GM-CSF, GM-CSF simultaneously or LPS prior to incubation. TNF-a and IL-8 concentrations were determined with the IMMULITE chemoluminescence immunoassay system (DPC-Biermann, Bad Nauheim, Germany). Leukocyte phenotyping was performed by dual-colour flow cytometry using whole blood lysis technique and monoclonal antibodies.

In healthy donors, ex-vivo stimulation with LPS leads to a massive increase of TNF-a production. However, if whole blood is incubated with GM-CSF 3 hours prior to the LPS challenge, the TNF-a production is significantly increased. The simultaneous incubation with LPS and GM-CSF leads to a significant decrease in TNF-a levels in the same patient population. GM-CSF stimulation of whole blood 3 hours after the production. In patients LPS challenge causes no significant change in TNF-a levels of with sepsis and endogenous TNF-a < 30pg/ml, GM-CSF pre-incubation production, whereas patients leads to a significant increase in ex-vivo TNF-a had a blunted ex-vivo reaction to LPS with higher endogenous levels of TNF-a stimulation.

Both the sequence of stimulation with either GM-CSF or LPS and the presence or absence of systemic TNF-a determine the ex-vivo cytokine response of whole blood. Hence, it may be speculated that 1. the administration of GM-CSF prior to the inflammatory stimulus would be most efficient, and that 2. the lack of stimulation effect in patients with high endogenous TNF-a may mirror Endotoxin tolerance. 

The most common acquired causes of weakness and muscle wasting in the critically ill patient in the intensive care units (ICU) are critical illness polyneuropathy and critical illness myopathy. There is significant clinical and neurophysiologic overlap between the two conditions, such that the term critical illness polyneuropathy and myopathy (CIPNM) is often used.

Over a 12-mo period, 22 critically ill patients who needed prolonged intensive care were studied. Clinical manifestations include delayed weaning from the respirator not explained by pulmonary complications, muscle weakness and prolonging of the mobilization phase. Included patients were classified as having MOF, SIRS and sepsis according to established consensus definitions. The occurrence of a positive EMG for CIPNM, as defined by an electrophysiologist who was blinded for treatment allocation, was analyzed during ICU stay. Variables recorded at baseline and during follow-up included patient demographics, principal diagnosis, routine blood tests and microbiological culture results. Levels of TNF-alpha, IL-6, IL-10, IL-4, procalcitonin (PCT) and C-reactive protein concentrations were repeatedly measured by ELISA.

All patients were divided in: patients without CIPNM at any time (group A, n=7), with a positive EMG during ICU stay (group B, n=8), and with a diagnosis of CIPNM since the admission (group C, n=7). EMG testing demonstrated severe acute denervation with striking involvement of proximal muscles in 15 patients. 6 patients died of complications of sepsis. Critically ill patients without CIPNM showed serum IL-6 levels lower (p < 0,05) than those with a diagnosis of CIPNM while no differences were found as concerned serum IL-10 levels. IL-4 and TNF-alpha did not show any difference between the two groups. IL-6 levels resulted higher in groups A and B (p < 0,05) while IL-10 levels were higher in group A (p < 0,01). In the group B, we observed a characteristic pattern of IL-6 and IL-10 serum concentrations that may be important for clinical outcome. IL-6 levels were higher than IL-l0 in patients with worse clinical outcome. The opposite pattern was observed in those with a good prognosis. No differences in clinical and laboratory variables were observed between patients with and without CIPNM. PCT appeared to be most helpful in differentiating patients with sepsis from those with SIRS (p < 0,001), exhibiting a greatest sensitivity (83%) and specificity (96%).

CONCLUSION. The analysis of the serum cytokines IL-6, IL-10, TNF-alpha and IL-4 to standard indicator did not improved the predictive power of detecting CIPNM but may contribuite to explain its pathogenesis. 

High dose glucocorticoids are known to induce muscle weakness. We investigated in a pilot study the occurrence of CIP/CIM in septic shock patients treated with low dose hydrocortisone (HC).

Patients were enrolled in the randomized controlled study of HC in septic shock (CORTICUS) and received HC (50 mg q 6h for 5 days, tapered until day 11) or placebo (PL). Electrophysiological testing (EP) consisted of the assessment of compound muscle (CMAP) and sensory nerve action potentials (SNAP), spontaneous activity (SPA), and muscle membrane excitability investigated by direct muscle stimulation (DMS). Clinical muscle weakness was defined by a Medical Research Council scale (MRC) below 4. CMAP and SNAP were categorized based upon normal age related values. EP results were categorized as unspecific (CIM or CIP or both) when CMAPs and SPAs were pathological in >/= 2 muscles. Presence of CIP was defined by pathological SNAPs in >/= 1 nerve, and CIM by DMS values < 3 mV. Data are shown as mean and 95%CI, Chi square test and Mann-Whitney-U-test were performed for statistical analysis.

From Jun 03 -Feb 05, 20 patients were enrolled in 9 sites: 9 HC and 11 PL. Median time for EP assessment was 12 days (4 -38) after study enrolment. 10 PL and 7 HC patients had unspecific electrophysiological signs; 6 PL patients, but only 1 HC patient had reduced SNAPs indicating CIP. In 16 patients DMS could be performed, 7/11 PL and 3/5 HC patients showed reduced muscle membrane excitability indicating CIM. In 13 patients (PL 7, HC 6) evaluation of MRC score was possible. Muscle strength did not differ between placebo [3.8 (3/4.5)] and HC group [4 (3.2/4.7)]. None of the parameters reached statistical significance.

CONCLUSION. The frequency of CIP/CIM diagnosed by electrophysiological examination was higher in patients who received placebo. The clinical diagnosis of muscle weakness assessed by MRC scale was not different in both groups. With limitations of the small sample size, this first prospective evaluation showed no impact of HC on the development of CIP/CIM in this cohort of patients with septic shock. 

Surviving sepsis campaign guidelines recommend treatment with hydrocortisone in septic shock patients requiring vasopressor support. However, the association of fludrocortisone remains controversial. The objective of the study was to determine if the association of fludrocortisone in patients with septic shock and adrenal insufficiency treated with hydrocortisone is related to an improved outcome.

From a database including 106 patients with septic shock requiring vasopressor support, we retrospectively studied 87 patients who fulfilled criteria for adrenal insufficiency (baseline cortisol less than 20 µg/dl and/or an increase after injecting 250 µg Synacthen less than 9 µg/dl). All patients included received treatment with hydrocortisone (H) or hydrocortisone plus fludrocortisone (H+F) for at least 24h. Data are presented as mean ± standard deviation. Groups were compared by using Student's t test for continuous variables and Chi-Square test for categorical variables. Long rank test and Kaplan-Meier curves were used to analyze time to shock reversal and mortality.

Forty-eight patients received hydrocortisone (H group) and 39 hydrocortisone plus fludrocortisone (H+F group). Overall mortality was 63% (55 patients). Both groups were comparable in baseline clinical and demographic characteristics. No differences were found in age (mean age 61±14), gender, weight (75±15 vs 73±13, p 0,48) (kg), infection site and severity scores: SAPS II (50±14 vs 47±14, p 0,36), APACHE II (22±6 vs 21±6, p 0,57) and SOFA max (14±3 vs 13±2, p 0,06). Both groups presented no differences regarding baseline (24±6 vs 25±13,p 0,88), stimulated (31±16 vs 31±15, p 0,99) and delta cortisol values (6,8±5,4 vs 6,6±5,2, p 0,76)(µg/dl). We did not find differences between both groups in norepinephrine(NE)maximal dose received(µg/kg/min), time to shock reversal (days of NE use), time of mechanical ventilation, ICU and in-hospital length of stay (days) and mortality ( 

Prospective, randomized, double-blind, placebo-controlled study of 28-day mortality in patients with septic shock for less than 72 hr who underwent a short high dose ACTH test in 52 centres in 9 European countries. Patients received 11-day treatment with HC (50 mg q 6h for 5 days, q 12h for 3 days, q 24hr for 3 days) or placebo (P). Serum electrolytes levels were obtained at baseline, day1 (D1), day2 (D2), day3 (D3) and day 7 (D7) from randomisation.

From Mar 02 -Nov 05, 499 patients were enrolled. Baseline serum sodium were 139 (6) mmol/l and 140 (6) mmol/l in the HC and P group respectively. Serum sodium peaked at D3 (143 mmol/L) and remained elevated up to D 7 (143 mmol/l) in the HC group. In the placebo group, serum sodiumpeaked at D2 (142 mmol/l). The mean change in serum sodium were, in HC treated and P treated patients respectively, at D 1: 1.3 (3.9 SD) vs 1.3 (3.7) mmol/l; D 2: 2.4 (5.4) vs 1.7 (5.5) mmol/l; D3: 3.1 (6.7) vs 1.6 (6.5) mmol/l; and at D 7:3.0 (8.0) vs 0.5 (7.5) mmol/l. The difference between groups reached statistical significance at Day 7 (p=0.003). There were no significant changes in mean potassium levels over time between the two treatment arms. 

According to the guidelines for the management of severe sepsis and septic shock, low doses of steroids are recommended in septic shock patients requiring vasopressors, despite adequate fluid replacement. The aim of this retrospective case control study was to assess the effectiveness of low doses of hydrocortisone in patients with late septic shock and MODS.

The study was held in a 19 bed multidisciplinary ICU of a tertiary hospital. Twenty four Norepinephrine dependent (> 0.5γ /kg/min) patients, fulfilling the criteria of septic shock, were enrolled in the study. Patients were divided in 2 Groups according to the continuous administration of 300 mg Hydrocortisone for 7days (Group A:12 pts) or conventional treatment (Group B:12 pts). End points of the study were, the within 7 days vasopressors weaning, evolution of MODS and 7-day as well as 28-day survival. MODS was described by SOFA score. Statistics : Statistical analysis was computed by using paired t-test and linear regression analysis.

Groups were similar regarding demographics (57±17 vs 64±15 y), initial SOFA score (10±3 vs 9,5±2), initial Norepinephrine dose (1.9 ± 0.7 vs 1.13 ± 0.6 γ /kg/min) and mean elapsed time from the onset of shock (3.7± 3.1 vs 3.5±2.5 days). An early and significant decrease in Norepinephrine dose (p<0.005), was observed in all Group A pts, while no difference was detected in Group B pts. This decrease was associated with hemodynamic stability. On days 3 and 4 mean ABP was significantly higher in Group A pts (p<0.001, P<0.005). Weaning from vasopressors within 7 days was achieved in 5 pts in Group A (41.6%) and 1 pts in Group B (0.8%). Seven day mortality was 16.6% in Group A vs 50% in Group B while 28-day mortality was 50% and 91% respectively. In the treatment group a positive correlation between the within 7 days shock reversal and survival (cor coeff = 0.657, r 2 = 0.432, p=0.02) was found. There was no relation between the time elapsed from the onset of shock to the steroid administration and survival (p=0.66). Oxygenation parameters (FiO2/PO2), SOFA score and creatinine did not differ between groups. WBC in Group A pts were significantly higher (p<0.005) only on day 3. No significant adverse effects were detected.

In late septic shock patients with MODS the administration of low doses of hydrocortisone is associated with decreased vasopressors requirements, hemodynamic improvement and beneficial effect on survival. The within 7 days shock reversal was a good predictor of survival. INTRODUCTION. Early microcirculatory impairment followed by mitochondrial dysfunction may combine to produce multi-organ failure in sepsis. We recently reported that tissue oxygen tension (tPO2), the balance of local O2 supply/demand, is variably affected in four different organs (kidney cortex, liver, muscle, bladder) at 3h' post-endotoxin challenge (1). We seek to measure temporal changes in tPO2 in these organs in a resuscitated rat model for up to 72h following the onset of faecal peritonitis. Here we present our 6-hr timepoint results with assessment of the impact of fluid loading. METHODS. Male Wistar rats (approx 300g weight) with tunnelled right jugular venous cannulae in situ received i.p. injection of faecal slurry. Fluid (1:1 mixture of 5% glucose/6% hetastarch; 10 ml/kg/h) was started 2h later. At 6h, rats were anaesthetised with isoflurane, and then instrumented with a left common carotid arterial line and tissue PO2 probes (Oxford Optronix, UK) sited in thigh muscle, between right and left liver lobes, in the left renal cortex and within the bladder lumen. After 30-min stabilisation, recordings were made of BP, tPO2, and end-diastolic volume (EDV) and cardiac output (CO) by echocardiography (Vivid 7, GE Healthcare, Bedford, UK). This was performed before (BI, baseline instrumented) and after fluid challenge (F) of 25 ml/kg bolus of 6% hetastarch given to optimise LV filling. Comparisons were made against sham-operated animals that underwent instrumentation but received no i.p. injection. 

Notwithstanding considerable volume resuscitation beforehand, left ventricular filling and output were significantly reduced at 6h in this faecal peritonitis model. Despite the 30% reduction in output, baseline tPO2 values were similar in bladder and renal cortex compared to sham animals but showed a decreased trend in muscle and a significant reduction in liver. Fluid loading restored cardiac output to control values, however only muscle and liver tPO2 increased, albeit not significantly. These data suggest a combination of microcirculatory and mitochondrial dysfunction with each predominating in different organ beds at this timepoint. Confirmation is required using complementary techniques. 

Microcirculatory dysfunction leads to inadequate tissue oxygenation and multi organ failure during sepsis or septic shock. Aim of this study was to compare non-invasive assessment of tissue oxygen saturation (StO2) with systemic oxygenation using mixed venous oxygen saturation (SvO2) as an indicator in an established model of porcine septic shock.

In a prospective animal study 20 anaesthetised, ventilated pigs (28.2 ± 2.1 kg) were investigated. Animals received 1g/kg/body weight faeces into abdominal cavity to induce sepsis and were observed over 8 hours. Volume therapy was administered to maintain a central venous pressure of 12 mmHg. SvO2 measured by CO-Oxymetry (Radiometer, Copenhagen) was obtained hourly after induction of sepsis. At the same time quadriceps muscle StO2 was measured by near-infrared spectroscopy (NIRS) (InSpectra TM , Hutchinson, USA). Correlation was analyzed by linear regression analysis.

A total of 136 measurements were performed in 20 animals. StO2 was significantly correlated with the SvO2. r = 0.52 (r 2 = 0.27) (p<0.01) and y = 0,32x + 43,6. Comparing the change in StO2 and SvO2 of two successive measurements reveals a correlation of r = 0.19 (r 2 = 0.035) (p<0.05). Changes in StO2 and SvO2 were parallel in 47% of two successive measurements (both measurements changed at the same time in the same direction).

Although there is a significant correlation between StO2 and SvO2 in our experimental septic shock model, paired StO2 and SvO2 changed in the same direction only in 47%. Thus, SvO2 may not be estimated on the basis of StO2 in treatment of experimental septic shock and tissue oxygenation may not be estimated on the basis of SvO2 either. Whether a combination of StO2 and systemic oxygenation measurements is a useful monitoring approach in sepsis needs to be revealed.

GRANT ACKNOWLEDGEMENT. Inspectra device was provided by Hutchinson. 

Systemic immune response syndrome (SIRS) frequently develops in critically ill patients and may lead to multiple organ dysfunction or failure even in the presence of normal or normalized global hemodynamic parameters, mainly due to tissue dysoxia and microvascular dysfunction. Near Infrared Spectroscopy (NIRS) is a validated method for the assessment of tissue oxygenation but its accordance with routine parameters has not yet been sufficiently studied. Aim: To compare NIRS parameters to routine monitoring parameters of the critically ill.

Thirty two consecutive critically ill patients (age=57±17 years, male/female=18/14, length of ICU stay=17±12 days) were enrolled. All patients were evaluated with NIRS and the occlusion technique within 24 hours of ICU admission. All patients were mechanically ventilated and 26 were sedated. Routine hemodynamic parameters (mean arterial pressure=83±15 mmHg, central venous pressure=8±3 mmHg, heart rate=85±16), full blood analysis (hemoglobin=11.3±4.4 g/dL, white blood cells=11,345±4,602 /dL) and arterial blood gases analysis were recorded. SOFA, APACHE II and SAPS III (55±13) scores were assigned on ICU entry day. Tissue Oxygen Saturation (StO2%) was continuously monitored before, during and after 3-min occlusion of the brachial artery via pneumatic cuff inflated up to 50 mmHg above measured systolic arterial blood pressure. (ELWI) has been demonstrated to predict mortality and to correlate to PaO2/FiO2-ratio and to the compliance of the lungs in patients with sepsis and ARDS. However, with an increasing number of obese patients, there is the question which body weight should be used for indexation of ELWI. Therefore it was the aim of our study, to investigate the correlation of ELWI to PaO2/FiO2-ratio and Oxygenation Index (mean airway pressure* 100 / PaO2) using different weight parameters for indexation.

In 25 patients of a medical ICU with a body mass index >25kg/m2, 260 measurements of extravascular lung water were performed using the PiCCO system (Pulsion, Munich; 7.0. software). Extravascular lung water was indexed using the actual body weight (ABW), predicted (PBW), ideal (IBW) and adjusted body weight(AdBW) , respectively. These data were correlated to PaO2/FiO2-ratio and Oxygenation Index. Spearman correlation, SPSS-software. The highest correlation to PaO2/FiO2-ratio was found using AdBW, the highest correlation to Oxygenation Index for ELWI adjusted to PBW. 3.) Although the extent of correlation varied within smaller limits (-0,438 to -0.510 and 0.446 to 0.578, respectively), the distribution of the patients within "normal", "modestly elevated" and "significantly elevated" ELWI would have changed markedly using different indices. 4.) With regard to impaired respiratory function in the patients of our study, PBW, IBW and AdBW seem to more accurately reflect "functional" extravascular lung water than ABW with 71% of the patients in the normal range. 

Our objective is to analyse the hemodynamic profile and the extravascular lung water in the first stages of severe acute pancreatitis (SAP) that are admitted at the Intensive Care Unit (ICU), through the collected data by transpulmonary thermodilution.

Observational and prospective study, in which 13-SAP-diagnosed patients consecutively admitted at the ICU were analyzed. All of them were monitorised at their admission with continuous cardiac output system PICCO ® (Pulsion Medical Systems). Demographic variables, general (APACHE II and SOFA) and specific (Balthazar) severity scores as well as the development or not of respiratory failure, were collected. The ordinary hemodynamic parameters [Heart rate (HR), mean arterial pressure (MAP), cardiac index (CI), vascular resistances (SVRI)] were determined on days 0, 1, 4 and 7 as well as preload parameters [intrathoracic blood volume index (ITBI), global end-diastolic volume index (GEDI)], extravascular lung water index (ELWI) and pulmonary vascular permeability index (PVPI) according to PICCO ® methodology. The results are expressed as means±SD and percentages. The non-parametric Mann-Whitney test for quantitative variables was performed and statistical significant level was established at p<0.05.

Age was 54±21 years with a majority of males (54%). The biliar was the most frequent cause (46%). APACHE II=13±5 and SOFA=6±3. All patients showed an alteration determined by CT scan (Balthazar Grading System) degree C or higher. Seven patients (54%) needed mechanical ventilation in the first 48 hours. Hospitalary mortality was of 38%. On day 0, the CI (3.5±0.8 l/min/m 2 ) and the RVSI (1885±717 din.seg.cm -5 .m 2 ) were at normal parameters and only 3 patients needed vasopressor support. However, on days 0 and 1, the preload parameters were low (ITBI= 765±162 ml/m 2 and GEDI =612±130 ml/m 2 ) and improved on the 4th day (ITBI= 870±195 ml/m 2 and GEDI =706±172 ml/m 2 ). Patients with respiratory failure and mechanical ventilation showed neither higher ELWI nor higher PVPI than the rest (day 1, ELWI: 7.1±1.8 vs 5.8±0.7 ml/kg; PVPI: 1.7±0.5 vs 1.6±0.3; p=NS).

In our population, certain hypovolemia degree in the first stages of the disease was found, corresponding to the development of the third space. The respiratory failure associated is not mainly due to an extravascular lung water increase or to a permeability increase. 0.6 (0.5 -1.5) 0.85 (0.8 -1.0) 0.025 CPO after dobutamine (W) 1. 2 (0.8 -1-2) 0.6 (0.5 -0.9) 0.008 POAP: Pulmonary occlusion arterial pressure, SWI: Stroke Work Index.

CONCLUSION. CPODelta after dobutamine challenge is a good predictor for mortality in SS. 

Septic shock is a common disorder with a high mortality. Recent guidelines for the haemodynamic management of severe sepsis have emphasized the importance of aggressive volume resuscitation in the initial phase. Central venous pressure (CVP) and pulmonary capillary pressure (PCP) are common end-points for volume resuscitation, however these cardiac filling pressures are poor predictors of fluid responsiveness in septic patients. 1 Right ventricular end diastolic volume index (RVEDVI) is a better predictor of preload, and it allows the identification of patients with right ventricular (RV) dysfunction and dilation (> 100-130 ml/m 2 ), as well as predicting mortality. We correlated RVEDVI with PCP, CVP and hypoperfusion variables during septic shock initial management.

Longitudinal, prospective and observational study. Demographic, haemodynamic (RVEDVI, PCP, CVP) and hypoperfusion (lactate, base deficit) variables were obtained. Descriptive statistics with mean ± SD (numerical variables) and frequencies and percentages (categorical ones). Comparisons between groups with U Mann-Whitney Test and X 2 and Fisher exact test as needed (statistically significant value if p<0.05).

RESULTS. 30 patients (mean age 64±17)were divided in: survivors n=21 (RVEDVI 147±16 ml/mt 2 ) and non-survivors n=9 (RVEDVI 111±15 ml/mt 2 ). Early dilation of RV predicts survival with a sensibility of 100% sensibility and specificity of 67%. METHODS. Ten patients with severe sepsis 60±6 yr, 25 patients with septic shock 63±8 yr and 10 polytrauma patients with hemorrhagic shock 40±7 yr, who remained in ICU more than 24 hours were included in the study. Serial BNP measurements were performed for at least 5 days. Consecutive hemodynamic measurements were done using a right ventricular Ejection Fraction (RVEF) thermodilution catheter (Edwards). Transthoracic echocardiography was performed in the first two days.

. BNP values (1st day) was dramatically elevated in septic shock (1110±356 pgmL-1), significantly elevated in severe sepsis (312±115 pgmL-1), but within normal limits in hemorrhagic shock (52±23 pgmL-1) (p<0.001). Inotropes (noradrenaline) were similar in patients with septic or hemorrhagic shock on day 1. BNP levels did not correlate with pulmonary arterial wedge pressure, right atrial pressure, RVEF or left ventricular EF (LVEF) measured by echocardiography. Eleven patients with septic shock, 2 with sepsis and 1 with hemorrhagic shock died during 28 days. BNP decreased gradually in survivors from septic shock after day 2. Septic shock survivors had lower APACHE II, and increased RVEF and LVEF compared to non-survivors (21±4, 39±4 and 73±9 vs 27±5, 32±7 and 65±7 respectively, all p<0.05), but not BNP (960±297 vs 1322±422 pgmL-1).

In conclusion, BNP is significantly elevated in sepsis, mainly in patients with septic schock, probably indicating the level of inflammation severity. Inotropes, shock and myocardial stretch, as it is expressed from hemodynamic parameters, do not seem to be implicated to BNP release. 

Sepsis and septic shock are major causes of mortality and morbidity in the ICU. If inflammatory mediators responsible of sepsis remain elevated or if there is a poor cardiac function, septic myocardial dysfunction may occur, increasing morbidity and mortality. Brain natriuretic peptide (BNP) is an adequate biomarker for cardiac failure 1 so our objective was to determine its utility in predicting myocardial dysfunction in septic patients. 

The role of hemofiltration, its dose and biological effects in sepsis remain a contentious issue. Although some beneficial effects on systemic hemodynamics and reduced vasopressor requirement were reported, the potential of hemofiltration to prevent sepsis-related disturbances of microcirculation and energy balance has not been evaluated. Therefore, we investigated the effects of standard hemofiltration (HF, ultrafiltration rate 35 ml/kg/h) and high volume hemofiltration (HVHF, 100 ml/kg/h) during 22h hyperdynamic porcine septic shock.

In 21 mechanically ventilated and instrumented pigs fecal peritonitis was induced by inoculating autologue feces. 12h after induction of sepsis pigs were randomly assigned to three groups: 1) controls (n=7), 2) HF (n=7), 3) HVHF (n=7). Before, 12,18 and 22h after the induction of peritonitis we measured, in addition to systemic and regional hemodynamics, ileal mucosal and renal cortex microvascular perfusion (OPS and laser Doppler flowmetry). Energy balance was determined by measuring arterial lactate pyruvate (L/P) and hepatic venous ketone body (KBR) ratios.

In the control group hyperdynamic septic shock resulted in a progressive deterioration of intestinal mucosal and renal cortex microvascular perfusion despite well-maintained regional blood flows. Altered microcirculation was paralleled by gradually increased L/P and KBR indicating disturbed energy balance. Compared to six animals in the control group, only three and two pigs required noradenaline support in HF and HVHF group, respectively. However, neither HF nor HVHF blunted the sepsis-induced alterations in microvascular perfusion and cellular energetics.

In this clinically relevant model of septic shock, the protective systemic hemodynamic effects of early hemofiltration did not translate into the improved microvascular perfusion and energy metabolism. HVHF did not confer any additional benefit. The value of hemodynamic improvement as a surrogate marker for efficacy of HF is therefore ambiguous. 

Patients in prolonged septic shock show enhanced pressor sensitivity to vasopressin(VP) yet decreased response to norepinephrine(NE). As both act via G protein-coupled receptors and activate the inositol phosphate cascade to increase vascular smooth muscle(VSM) Ca 2+ levels, the reason for this disparity is uncertain. We postulate that these drugs may have diverse effects on different Ca 2+ mobilisation pathways during sepsis. We investigated this using specific modulators of Ca 2+ release and influx on contractile responses to VP and NE in mesenteric arteries from septic and sham-operated rats.

Sepsis was induced in 6 awake, fluid-resuscitated Wistar rats by ip injection of fecal slurry. Paired sham controls received no injection. Rats were sacrificed after 24h, and mesenteric arteries mounted on a wire myograph to measure isometric tension responses to VP and NE. The contributions of sarcoplasmic reticulum(SR) Ca 2+ release and Ca 2+ entry through the store-operated channel(SOCC) were assessed by removing and returning extracellular Ca 2+ respectively. The contribution of the voltage-gated Ca 2+ channel(VGCC) was assessed by applying VP/NE in the presence of nifedipine.

Contractions were significantly enhanced to VP but depressed to NE in septic vessels . In all arteries, constriction to both agonists relied predominantly on extracellular Ca 2+ influx rather than SR Ca 2+ release. NE responses were more sensitive to extracellular Ca 2+ removal in septic vessels. The Ca 2+ influx in response to NE was almost entirely VGCC-mediated, with a negligible contribution from SOCCs in both sham and septic arteries. SOCCs contributed significantly to VP contraction however, and SOCC-rather than VGCC-mediated influx of Ca 2+ predominated in septic arteries. 

Patients in prolonged septic shock show enhanced pressor sensitivity to vasopressin (VP) yet decreased responsiveness to norepinephrine (NE). We have reproduced this pattern in ex-vivo contractile responses of resistance arteries taken from rats subjected to a clinically realistic septic insult (1). We hypothesise that an underlying mechanism is VP-mediated sensitisation of the vascular smooth muscle contractile apparatus to calcium. To investigate this, we performed simultaneous wire myography and fluorescence microscopy to examine the relationship between contractile response and intracellular calcium concentration ([Ca 2+ ]i).

Sepsis was induced in conscious, tethered, male Wistar rats by intra-peritoneal injection of faecal slurry. Paired sham controls received no such injection. Both groups received 10ml/kg/hr of intravenous fluid. After 24 hours, animals were sacrificed, and 3rd order mesenteric arteries dissected and mounted on a wire myograph (Danish Myo Technology). Arteries were loaded with a fluorescent calcium indicator (fura-2, 10muM) for 1 hour and imaged by fluorescence microscopy. [Ca 2+ ]i and isometric tension kinetics were measured simultaneously in response to VP (3nM) and NE (10muM).

]i was higher in arteries taken from septic rats. Tension responses to VP were significantly enhanced in septic arteries, however the associated increases in [Ca 2+ ]i were comparable in septic and sham groups. Tension responses to NE were significantly decreased in septic arteries, with a similar degree of depression in delta [Ca 2+ ]i. Data were analysed for statistical significance using un-paired t tests. CONCLUSION. The higher baseline [Ca 2+ ]i in the vascular smooth muscle of septic arteries suggests an abnormality of intracellular calcium storage. The ability of VP to produce a greater contractile response in septic compared to sham arteries, despite an equivalent degree of [Ca 2+ ]i elevation, implies sensitisation of the contractile apparatus to the effect of VP. There was contractile hyporesponsiveness to NE in the septic vessels and no evidence of calcium sensitisation to this agonist. These findings provide one potential explanation for the hypersensitivity to VP observed in patients with septic shock. 

Mitochondrial dysfunction and compromised cellular energetic status are associated with poor outcome in septic patients [1] . Maintenance of mitochondrial function is mediated in part by activity of transcription factors NRF-1 and NRF-2, the transcriptional co-activator PGC1-alpha and mitochondrial transcription factor alpha (Tfam). These markers of mitochondrial biogenesis were elevated in a rodent model of endotoxaemia [2] . In an ongoing study in critically ill patients, we have investigated the relationship between cellular energetics and mitochondrial biogenesis.

With ethics approval and appropriate consents, critically ill patients were recruited within 24h of ICU admission. Age-matched control patients were undergoing elective hip surgery. Muscle biopsies were taken from vastus lateralis. ATP and creatine compounds were determined by HPLC of perchloric acid extracts and standardised to total creatine [total Cr = phosphocreatine (PCr) + creatine (Cr)]. mRNA levels for PGC1-alpha, NRF-1 and Tfam were determined by RT-PCR and standardised to 18S mRNA. Data were analysed for significance using one-way ANOVA.

The ratio of PCr/Cr was significantly decreased in both survivors and non-survivors. mRNA levels of the mitochondrial biogenesis markers PGC-1alpha and NRF1 increased in survivors but not in non-survivors. A similar pattern was observed with the mitochondrial transcription factor Tfam, although statistical significance was not reached. (1) The decreased PCr in both survivors and non-survivors indicates increased demand for ATP in the acute phase of critical illness. (2) Increased levels of markers of mitochondrial biogenesis in survivors indicate that maintenance of mitochondrial function, specifically ATP synthesis, may be crucial to recovery. Failure to maintain adequate mitochondrial function through biogenesis may contribute to ATP depletion and mortality. 

Local metabolic changes are not well investigated in sepsis and SIRS. Our aim was to describe subcutaneous metabolic changes using microdialysis (MD) concurrently with systemic hemodynamics over 7 days in patients with sepsis/SIRS and circulatory failure.

METHODS. 25 patients with severe sepsis/SIRS were recruited. At inclusion, all patients had circulatory failure despite resuscitation according to the Rivers concept. Cardiac Index (CI), Intrathoracic Blood Volume Index (ITBVI), Extravascular Lung Water Index (EVLWI), Blood Lactate (P-Lac), MD Lactate (MD-Lac) and MD Lactate-Pyruvate ratios (MD-Lac/Pyr) were analysed 4-6 hourly. Data were tested for differences over time using ANOVA. Patients were subdivided into sepsis and SIRS groups, and intergroup differences were tested using the Rank Sum test.

Mean APACHE scores were 26&24 for sepsis & SIRS respectively. SOFA decreased from 11.5 to 4.5 with no difference between sepsis & SIRS. CI increased over time and ITBVI, EVLWI, P-Lac & MD-Lac decreased. MD-Lac & P-Lac were maximal at Day1. Lactate concentrations were generally higher in MD than in blood, and in the sepsis group. 

Severe sepsis and septic shock have been recognized as a serious clinical problem that shows an increasing incidence and that is responsible for substantial morbidity and mortality in intensive care units. Sepsis has been defined as the systemic host response to infection with an overwhelming systemic production of both pro-and anti-inflammatory mediators. Continuous hemofiltration has been suggested as possible therapeutic option that may remove the inflammatory mediators. On the other hand, hemodialysis and hemofiltration were reported to influence cardiac electrophysiological parameters and to increase the arrhythmogenic risk. Therefore, in this study we have investigated the effects of hemofiltration on electrophysiological properties of the septic pig heart.

METHODS. 40 pigs of both sexes were divided into 5 groups: 1) control group without hemofiltration; 2) control group with conventional hemofiltration (35 ml/kg/hour); 3) septic group without hemofiltration; 4) septic group with conventional hemofiltration (35 ml/kg/hour); 5) septic group with high-volume hemofiltration (100 ml/kg/hour). In septic groups, the sepsis was induced by fecal peritonitis and maintained for 24 hours. Hemofiltration was applied for the second 12 hours of this period. ECG was measured just before and after 24-hours period of sepsis in septic groups and at the same time points in non-septic groups. Action potentials were recorded in isolated ventricular preparations obtained from the hearts at the end of experiments.

. RR and QT intervals were significantly shortened by sepsis in all 3 septic groups, in non-septic groups they were not influenced by the experiment. Action potential duration (APD) was also significantly shortened by sepsis (septic group without hemofiltration vs. control group without hemofiltration) at all cycle lengths tested (500, 1000, 2000 ms). Both conventional and high-volume hemofiltration in septic groups shortened APD further at slow pacing rates. Hemofiltrate obtained in septic groups by both conventional and high-volume hemofiltration prolonged significantly and reversibly APD at all pacing rates. Substitution solution alone had no effect on APD. Neither hemofiltration nor hemofiltrate in control, non-septic groups influenced APD.

We conclude that the hemofiltration in septic groups and the septic hemofiltrate influence significantly the electrophysiological properties of the heart, probably due to removal/content of various inflammatory mediators in the septic hemofiltrate. INTRODUCTION. The precise mechanism by which multiorgan failure develops in severe sepsis and septic shock remains unclear. Potential mechanisms include alterations of microvascular flow distribution, mitochondrial dysfunction and treatment effects. We investigated the effects of LPS and different catecholamines on oxidative respiration of rat skeletal muscle fibers and hepatocytes.

Muscle fibers (M. gastrocnemius) were isolated from anesthetized male Wistar rats (200-300 g). Human hepatocytes (HepG2 cells) and human monocytes (MonoMac 6 -MM6) were also used. To avoid systemic effects of endotoxin and catecholamines, experiments were performed in vitro using the skinned-fiber technique. The mechanically dissected muscle fibers were incubated with LPS (10 µg/ml) for 2 h. After 1 h of LPS incubation, norepinephrine, dopamine, and dobutamine (100 µM each) were added. Monocytes and hepatocytes were treated with different concentrations of LPS only. Mitochondrial respiration was determined after permeabilization with saponin, using a Clark type electrode (Oxygraph 2K, Orobros Instruments, Innsbruck, Austria). 

Septic shock is associated with severe cardiac dysfunction, whose mechanisms remain only partly defined. Recent data suggested that it might be triggered by the direct action of microorganisms and their products on the heart itself. We previously shown that flagellin (FLAG), the protein monomer from bacterial flagella, is a potent activator of NF-κB-dependent pro-inflammatory signaling in cultured cardiomyocytes. Therefore, the aim of the present study was to evaluate whether FLAG might induce such an inflammation in the heart in vivo and contribute to cardiac dysfunction.

H9c2 cardiomyocytes were stimulated with recombinant Salmonella muenchen FLAG (1-100 ng/ml, 10 min to 24 h). In vivo, BALB/c mice were injected (tail vein) with 1-5 µg FLAG (30 min to 6 h). The effects of FLAG were evaluated by its ability to activate NF-κB, and to induce transcription of TNFα and MIP-2 cytokines. In vivo, cardiac neutrophils recruitment was evaluated by myeloperoxidase (MPO) activity. The expression of the FLAG receptor TLR5 was also determined. In vivo physiological measurements: left ventricular pression-volume curves. A microtip pressure-volume (PV) catheter (SPR-839; Millar Instruments) was inserted into the left ventricle (LV) via the right carotid artery. The pressure and volume signals were continuously recorded and heart rate, cardiac output, end-systolic and end-diastolic volumes, stroke volume, ejection fraction and end-systolic and end-diastolic pressures were measured. Load-independent indices of LV systolic and diastolic functions were determined by the slope of the end-systolic, respectively end-diastolic PV relationships in conditions of rapidly reduced preload (transient compression of the vena cava).

. FLAG activated NF-κB in cardiomyocytes in vitro and in vivo, and also upregulated the transcription of TNFα and MIP-2. FLAG also increased cardiac neutrophils recruitment. FLAG induced significant increases in end-systolic and end-diastolic LV volumes, indicating cardiac dilation, and a significant reduction of the load-independent indices of LV systolic function (end-systolic PV relationship, ESPVR, and maximal elastance, Emax), indicating significant LV systolic dysfunction. In contrast, no change in the slope of the end-diastolic PV relationship (EDPVR) was noted.

Bacterial flagellin induces a prototypical inflammatory response in cardiomyocytes in vitro and in the myocardium in vivo. These effects are associated with a profound alteration of the LV systolic function in vivo, suggesting that flagellin may represent a critical mediator of cardiac dysfunction in septic shock. 

Current guidelines recommend either dopamine (DA) or norepinephrine (NE) as the initial vasopressor in septic shock (SS), but the management of moderate to severe SS is still controversial. To explore this issue is important, because pharmacodynamic differences between vasopressors might be irrelevant in mild cases, but could potentially affect outcome in more severe patients. Beside clinical implications, there are also economical considerations since these drugs are not cost-equivalent. This subject may be specially important for developing countries. The aim of our study was to compare NE vs DA as the exclusive vasopressor for established moderate to severe septic shock (requirements of > 0.1 mcg/k/min of NE or > 10 mcg /k/min of DA to maintain MAP 70 to 80 mmHg)

Multicentric RCT involving nine polivalent ICUs from Argentina, Brazil and Chile, randomizing moderate to severe SS patients to NE or DA titrated to target MAP or maximal dose of 2 mcg/k/min NE or 100 mcg/k/min DA. After inclusion patients were switched blindly to the assigned drug. The study could be stopped if severe hypotension or arrhythmias developed. Epinephrine was used as a rescue drug. Main outcome criteria were 28 day mortality, organ dysfunctions and adverse effects (AE).

The study was stopped early after randomizing 50 patients because of low enrollment rate. Only 45 patients were evaluable. Main results are shown on the table. Adverse effects with DA were 4 cases of atrial fibrillation (AF) and 6 supraventricular paroxysmal tachycardia (SPT), which were considered serious in 3 cases. AEs with NE were two AF and one SPT, which resolved with no drug suspension. AEs occurred more frequently with higher doses of DA. CONCLUSION. The use of dopamine as exclusive vasopressor for established moderate to severe septic shock appears to be associated with a worst outcome and more adverse effects. This should be explored in a future better powered RCT. 

Although arterial blood pressure (ABP) is a widely used guide for hemodynamic therapy in sepsis, few data exist on its association with mortality and on critical ABP limits that should be maintained.

In this retrospective cohort study, clinical, hemodynamic, and laboratory parameters were extracted from a prospectively collected database in 274 sepsis patients. The severity and duration of hypotension was calculated by the area under the curve (AUC) of systolic arterial blood pressure (SAP), mean arterial blood pressure (MAP), and mean perfusion pressure (MPP = MAP -central venous pressure). Laboratory parameters included the most aberrant variables during the ICU stay. Urine output per hour during the first 24 hours and need for renal replacement therapy were recorded. The Sepsis-related Organ Failure Assessment (SOFA) score was calculated from given clinical and laboratory parameters. Binary and linear regression models were corrected for the severity of disease by inclusion of the SAPS II (excluding SAP count) as a covariate and were used to examine the association between ABP and 28 day-mortality or organ function. Similarly, a binary logistic regression model including SAPS II as a covariate was used to determine the best discriminating cut-off limit of ABP in regards of 28 day-mortality. The goodness of fit of each limit was assessed by the r2-value according to the Nagelkerke method.

. SAP and MAP were recorded for 21.6±3.6 hours, MPP for 18.9±5.2 hours. There was a significant association between 28 day-mortality and the AUC of SAP (p<0.001, r2=0.262), MAP (p<0.001, r2=0.269), MPP (p<0.001, r2=0.295). The area under MAP 60 mmHg and MPP 45 mmHg was associated best with 28 day-mortality. One or more episodes of MAP <60 or MPP <45 mmHg increased 28 day-mortality by 2.96 (CI 95% 1.06-10.38, p=0.041) and 2.97 (CI 95% 1.19-7.76, p=0.01), respectively. There was a linear association between time under the critical MAP and MPP limit and 28 day-mortality. While ABP was significantly associated with the SOFA score, arterial lactate levels, and renal function, no association with liver function or troponin I was observed. The critical MAP and MPP limits for the need for renal replacement therapy were 75 mmHg (r2=0.127, p<0.001) and 60 mmHg (r2=0.124, p<0.001), respectively.

During early sepsis, ABP is associated with 28 day-mortality and organ function. MPP shows the best association with mortality and may be a new resuscitation target. 

Animal models of traumatic brain injury (TBI) are used to elucidate sequelae underlying human head injury in an effort to identify potential neuroprotective therapies. Although human TBI is a highly complex multifactorial disorder, animal trauma models tend to replicate only single factors involved in the pathobiology of clinical head injury and may thus partly underlie the discrepancy between preclinical and clinical trials of neuroprotective therapeutics. We here present our experience with a large animal model of TBI which was designed to closely resemble the forces impacting the brain in e.g. traffic accidents.

Anesthetized, mechanically ventilated instrumented sheep (n=26) were placed in prone position with the head resting on a support to allow free lateral movements of the head. A left-temporal head impact was then delivered by mechanical stunning device (MK 1200, Schermer, Germany), which is approved for euthanasia of domestic lifestock. A captive bolt with a mushroom-shaped head is propelled from the muzzle of the stunner against the skull by the discharge of blank cartridge inserted in a chamber behind the proximal end of the bolt. Depending on the charge and the positioning of the stunner, this device delivers an intracranial atmospheric pressure of approximately 11 bar in sheep at a bolt velocity of approximately 49 ms-1. To prevent skull fractures, a steel plate was attached to the left temporal fossa. A fiberoptic intracranial pressure (ICP) catheter and a brain tissue oxygen (PbrO2) probe were introduced in the parietal white matter. Unilateral ultrasound flowprobes were attached to the internal carotid artery to measure cerebral blood flow. After measurements, sheep were killed and the brains removed for neuropathological examination.

Brain injury was characterized by a marked increase in ICP from 10 ± 3 to 23 ± 13 mmHg (mean values ± standard deviations) 2 hours after head impact. Intracranial hypertension was accompanied by a significant decrease of cerebral blood flow. PbrO2 significantly decreased from 19 ± 10 to 13 ± 12 mmHg. The decrease in sinus venous oxygen saturation did not reach statistical significance. In instrumented Control animals (n=2), parameters remained unchanged. Neuropathological examinations revealed the presence of multifocal traumatic subarachnoid hemorrhage in 18, and diffuse axonal injury in 23 out of 24 animals. While interstitial brain edema was found in all sheep brains, contusion zones were present only in a minority of the animals.

The pathobiological characteristics of the head impact model presented here closely resemble the alterations frequently found in human TBI. The relatively high variability of neuropathological changes after head impact may be seen as a disadvantage of this model. 

Non-neurologic organ dysfunction triggered by infection represents a frequent and independent predictor of poor outcome in traumatic brain injury (TBI) patients admitted to intensive care units (1). Because TBI itself significantly increases susceptibility to infection (2) and infection is a potentially modifiable risk factor, we developed a combined experimental model of TBI and sepsis in the rat.

Controlled cortical impact (CCI) was produced in left parietal cortex by using a 3 mm diameter tip (velocity 5 m/sec; depth 2 mm). Sepsis was induced contemporarily by cecal ligation and puncture (CLP). The outcome was evaluated in terms of mortality, neurological function (via the Morris Water Maze (MWM) and Beam Balance (BB) tests)and histologically.

Rats were subdivided into 4 groups: sham, CCI, CLP, and CCI + CLP. 15-day mortality was 0% in sham, 4% in CCI and 16% in CLP group respectively. Adding CLP to CCI increased mortality up to 40% (p<0.01 vs CCI and p<0.05 CLP alone). At 48h and 1 week post-injury MWM and BB test performance was significantly worse in CCI and CCI + CLP than in sham and CLP groups (p<0.05). Lesion volume was similar in injured groups. CA3 cell loss in left hippocampus was unaffected in the sham and CLP groups, while it was 25% in CCI and 34% in CCI + CLP groups (p<0.05 CCI vs CCI + CLP).

Our results show that the occurrence of systemic sepsis exacerbates mortality and cerebral damage in rats subjected to traumatic brain injury. 

T. J. P. Lieutaud* 1 , J. Rhodes 2 , P. J. D. Andrews 2 1 Anesthesiology and intensive care medicine, hospices civils de Lyon, Lyon, France, 2 Anesthesiology and Intensive care medicine, University of Edinburgh, Edinburgh, United Kingdom INTRODUCTION. Human recombinant erythropoietin (EPO) appears promising in different brain injury models but its cellular mechanisms remain poorly understood. Following brain trauma injury (TBI), inflammation (IL-1B) and chemokine expression (MIP-2, Neuropath Appl Neurobiol 1998) are important. The aim of this study was to measure the effects of acutely administered rhEPO on IL-1B and MIP-2 after TBI. METHODS. With Home Office approval, under isoflurane anesthesia 9 rats SD were subject to lateral fluid percussion TBI (1.6-1.8 atm) (Dixon J Neurosurg 1987) of the left parietal cortex. EPO (1000, 3000 or 5000 iu/kg) or placebo were injected in a random and double blinded manner by the intra-peritoneal (ip)route. The ipsi-and contra-lateral cerebral cortices were removed 4h later and homogenized. IL-1B and MIP-2 were measured in the surnageant using ELISA kits. Results are expressed as pg/mg of protein (mean ± SEM).

There was a significant increase in IL-1B and MIP-2 in the ipsilateral cortex in comparison with the contralateral side for both proteins analyzed. Neither 1000 nor 3000 and 5000 iu/kg rhEPO did not exhibited any significant effect (figure 1).

CONCLUSION. This study confirms that inflammation is important and occurs early after LFP-TBI. EPO did not display significant effects on two of the main inflammation mediators. 

The purpose of this study was to evaluate the effects of agmatine on histopathological damage following traumatic injury using a clinically relevant model of diffuse axonal injury (DAI) on the rat.

A total of 24 male Sprague-Dawley rats weighing 175-200 g were anaesthetized and subjected to head trauma using Marmarou's impact-acceleration model. The rats were then separated into two groups; one group was treated with agmatine and the other group was treated with saline for up to four days immediately after the head trauma. Rats from both groups were killed one, three or eight days post-injury. The brains were examined histopathologically and scored according to the neuronal, vascular and axonal damage.

There were no significant histopathological differences between the control and agmatine-treated group after one or three days (p>0.05), but evaluation after eight days revealed a significant improvement in the group treated with agmatine (p<0.04).

Our data indicate that agmatine has a beneficial effect in diffuse axonal injury and should be tried for therapeutic use in the management of this condition.

D. Morii*, Y. Miyagatani Critical Care department, National Hospital Organization Kure Medical Center, Kure, Japan

The disadvantageous effect of haemorrhagic shock on head trauma related mortality are well known. Thus, efficacious shock treatment is a surely significant measure against the development of secondary brain damage. The small volume resuscitation by hypertonic saline has been shown to promote systemic and cerebral haemodynamic benefits. Similarly, many clinical studies have demonstrated the effects of long-term mild hypothermia on outcome of traumatic brain injury. In this study, we evaluated the new strategy consisting of therapeutic mild hypothermia and hypertonic saline therapy to the multiple trauma patient with severe traumatic brain injury.

Severe multiple trauma patients (ISS>=33, Head AIS>= 3) were studied to evaluated the efficacy of therapeutic hypothermia (35.5˚C for 48 h) and hypertonic saline therapy (Na+: 200 mEq/L for the first 24 h , 180 mEq/L for 24 to 48 h, 160 mEq/L for 48 to 72 h) which were applied to them in parallel with massive blood transfusion . We evaluated Glasgow Coma Scale (GCS), Injury Severity Score (ISS), the probability of survival (Ps), the volume of blood transfusion, infusion and urine volume during the first 3days, and Glasgow Outcome Scale (GOS). We monitored the extent of brain swelling by head CT.

Four male patients (age: 39±16 y.o., mean±SD) were examined. The characteristics of injury mechanism were explosion 1, MVA 2, fall 1. On admission, GCS, Head AIS, ISS and Ps were 8.75±4.5, 4.5±1, 46±11 and 0.67±0.29, respectively. The sum of blood transfusion, infusion, and urine volume during the first 72 h were 2200±1773 ml, 16313±5503 ml, 6985±4109 ml. No patient was died and their GOS on posttrauma day 90 was 3.8±0.5.

The combined therapy of therapeutic hypothermia and hypertonic saline to multiple trauma patients with brain injury may lead to good outcome in spite of the necessity of a large quantity of blood transfusion and infusion. 

Recent data suggest that commonly used anaesthetic agents, e.g. propofol, cause neurodegeneration in the developing brain. The intention of our study was to investigate the effects of propofol on primary neuronal cultures referring to the cell survival rate.

Primary cortical neuronal cultures were prepared from Wistar rat embryos at 18 days gestation. To test the effect of propofol on neuronal survival, cultures were exposed to 100 µl Gibco Neurobasal-A medium per well with propofol at a concentration of 1 mg/ml for 3, 6, 9, 12, 24, and 48 hrs. Cell viability was assessed using the methyltetrazolium method (MTT) and was related to untreated cells as controls. All cells were kept in normoxia.

After three and six hours of exposition to propofol cell viability values of the propofol treated cells were significantly higher (150.1±12.1%, p=0.0033 and 122.6±8.5%, p=0.0297, respectively) compared to untreated control cells (100%). After 9 hours, values were decreasing to levels of the control cells (106.6±9.8%). After 12, 24 and 48 hours of exposition to propofol, in contrast, cell viability was significantly reduced (82.1±6.9%, p=0.0285, 42.8±5.9%, p<0.0001 and 22.4±4.0%, p<0.0001) compared to controls.

At high concentrations, propofol has a time-dependent effect on the viability of primary cortical neurons. During the first 6 hrs propofol has a potential neuroprotective effect, whereas it seems to cause neurodegeneration in the period of 12 to 48 hrs of exposition.

E. Paramythiotou* 1 , J. Papanikolaou 2 , P. Ntagiopoulos 2 , A. Armaganidis 1 , A. Karabinis 2 1 ICU, Attikon University hospital, 2 ICU, George Gennimatas hospital, Athens, Greece

Multiple trauma patients constitute a significant majority of admissions in a general ICU. Brain injury is often present in those patients. The aim of our study was to investigate demographic, clinical and management characteristics in trauma patients suffering a brain injury in a five year period.

In a retrospective study all trauma patients hospitalized in the 10 -bed multivalent ICU of a 700 bed -tertiary hospital between 1st Jan 2002 and 31th Dec 2006 suffering a traumatic brain injury were enrolled. Recorded data included age, gender, cause of the injury, ICU length of stay, initial Glasgow Coma Score (CGS), submission or not to an emergent neurosurgical intervention, all cause mortality and neurological outcome.

A total of 260 trauma patients were hospitalized during the study period. TBI was present in 176 patients (67.7 %). Among them, 35 were women (20%) and 141 (80%) were men. Their mean age was 37.6 years (range 16 -83). ICU length of stay (LOS) ranged between two and 300 days (mean 35.4days). Traffic road accidents were the cause in 156 cases (88.6 %) while 14 TBIs (8 %) were due to fall from a height on the ground which happened either accidentally or as a result of a suicide attempt. The rest 6 cases (3.4%) were due to accidents during work. Mean Glasgow coma score was seven (range 3-13). An extradural hematoma was present in 9p and a subdural one in 30p. Intracerebral hemorrhage was noticed in 5p, hemorrhagic contusions in 21p (with or without diffuse axonal injury) and a traumatic subarachnoid hemorrhage in 32 p. Twenty nine patients were submitted to craniotomy and 37 p were submitted to unilateral or bilateral decompressive craniectomy. Mean LOS was 37.8 d for p submitted to a surgical intervention versus 39 d for the other group. Barbiturates were used in 16 p (9%). A total of 138 patients survived (78.4%). Death was due to neurological cause (herniation of brain stem and subsequent cerebral death) in 25 p. Other causes of death included sepsis, multi organ failure, severe injury in other organs, and hemorrhage from upper gastrointestinal tract. A poor neurologic outcome (mean Glasgow outcome score < 7) was noticed in 10% of patients.

Almost two thirds of trauma victims suffer from a cerebral injury. Most of them are young males, victims of traffic road accidents. The injury is often severe and one third of patients are submitted to a neurosurgical operation. Though overall mortality is rather low, long duration of treatment is often required and severe disability is present in a not negligible number of patients. 

In the majority of the Intensive Care Units (ICU), several of the admissions involves patients with primary nervous system illnesses. A great progress of the technologies used in the ICU in the last few decades had reduced neurological illnesses mortality and morbidity.

Since september of 2006 we had beginning an longitudinal e prospective coort study verifying the characteristics of the patients 18 years older that had been admitted in the ICU for primary neurological cause (clinical or surgical). The study occurred in a private hospital ICU with 46 beds. We recorded 148 patients until the moment.

The number of neurological patients corresponds 10% of the admissions in the unit. The average age of this group of patients is significantly lesser of the remain ICU patients (65 vs. 72 years), however does not have difference estatistically significant between APACHE II (12 vs. 11) and the mortality (11 vs. 12%) of the neurological patients and others. The stay of length in the unit is bigger (8,6 vs. 4,8) . We also recorded mechanical ventilation time length (26% ventilated patients with for average time 14 days). In ventilated patients, 41% was tracheostomyzed (on average in 9 days). 19% developed sepsis (11% with septic shock). The patients were divided and analised in several goups (for example: trauma, surgery, central nervous sistem infection, vascular disease,...). 

Neurology was one of the most benefited specialties with the Intensive Care Units progress and evolution. However, high mortality and morbidity caused by the neurological illness, and the social and economic impact that its sequels cause, still deserve the attention of the involved professionals cares of these patients in the acute illness.

N. Baffoun* 1 , W. Gdoura 1 , H. Ouragini 1 , K. Baccar 1 , M. Lamourou 1 , T. Chaoua 1 , R. Souissi 1 , C. Kaddour 1 , N. Ben Romdhane 2 , S. Mahjoub 2 1 Anesthesia and intensive care, National institute of neurology, 2 Departement of haematology, CHU La Rabta, Tunis, Tunisia

Trauma victims develop frequently various degrees of haemostatic disorders. The severity of such post traumatic coagulopathie is considered to be major detrimental factor of outcome. The aims of our study were: to identify the origin of such disorders, time course and their correlation with mortality. Our aim was identification of coagulopathy disorders and relation to outcome in severely head injured.

Prospective study,June 2003-March 2004. Included:critically ill isolated closed severe head trauma. Collected data:Demographics,Management prior and during ICU hospitalization (sedation, catecolamin drug use, blood product transfusion, intra-cranial pressure monitoring, neurosurgical emergency surgery etc.),CT-Scan results, Daily worst Glasgow coma scale, admission Simplified Acute Physiology Score II. We inserted an arterial catheter for invasive pressure monitoring, a central venous catheter and a unilateral jugular bulb in front of the most damaged brain hemisphere(cf. CT-scan). Jugular bulb thrombosis was prevented by continuous infusion of 2ml per hour isotonic serum without heparin. Blood samples were obtained simultaneously from the central venous line(K) and jugular bulb(B) at admission, 6th, 12th hour, and then in case of neurological aggravationt or daily till 3th day. We measured platelet count,prothrombin time (PT),activated partial thromboplastin time (ACT),fibrinogen concentration (Fib), prothrombin fraction1+2 (F) and thrombin anti-thrombin complex (TAT). During the study only central venous blood samples (PT, ACT, Fib and Platelet count) could be available if necessary. Otherwise blood samples were centrifuged and preserved refrigerated for post hoc analysis. Statistical analysis by Student's t test, paired t test for paired results and analysis of variance. Significance set as p<0,05.

RESULTS. n=19; 9 survivors(S) and 10 deaths (NS). No differences between S and NS in demographics,management modalities, admission GCS(7±3), CT-scan,SAPS II (27±10 vs 30±17, p=0,69). B vs simultaneous K Platelet count was significantly lower in all drawn blood samples,with a trend to decrease overtime. S vs NS at Day1 and Day2: 191±60 vs 125±35 (p=0,017). Admission B thrombin fractions was higher in NS(1000±209 vs 460±294, p=0,014). B Day1 TAT was higher in NS:45±20 vs 9,6±12 p=0,02. No difference for other tests between B vs K and S vs NS for different paired tests.

Pro-coagulant factors (F and TAT) are valuable prognostic factors at Day1 in closed isolated severe head trauma. 

Severe traumatic injury is a multisystemic disease where normal homeostatic mechanisms are lost. This situation involves an increase in physiological needs. Usually these patients present anormalities in the hypothalamic-hypophyseal axis, which become neuroendocrine dysfunctions with deteriorated physical or neuropsychological secuelae. The aim of this study is to improve our knowledge about this part of the axis in acute phase of politraumatism.

METHODS. An observational prospective study was carried out, with 27 patients who were admitted to our ICU with a critical traumatic injury, for six months. Demographic and epidemiological data were registered. APACHE-II (Acute physiology and chronic health evaluation system) and APACHE-III scores during the first three days were measured. TISS (Therapeutic intervention scoring system) score during the hospital stay was recorded. Also GH (grown hormone), IGF-1 (insulinlike-grown-factor-1) levels and Nitrogen urinary losses in the first three days after traumatic event were measured. Statistical data were analysed with the SPSS 13.0 program.

In our study 85,2% (23 cases) were men and 14,8% (4 cases) were women. The average age was 48,15 years old. The hypothalamic-hypophyseal-somatotrophic axis role in the first three days was characterized by a progressive increase in GH levels and a progressive decrease in IGF-1 levels. Connections between average hormonal levels in the first three days and APACHE-II, APACHE-III and TISS scores during this time were studied. A good inverse connection between IGF-1 and prognosis was shown SI (Spearman Index) -0,524, -0,631 p value 0,006 and 0,001 respectively with APACHE-II and APACHE-III. This appropriate connection could not be shown with TISS score SP -0,039 p value 0,848; but the connection between GH and TISS was better, SP 0,37 p value 0,063.

CONCLUSION. GH levels increase and IGF-1 levels decrease in the first three days after acute trauma. Lower IGF-1 levels can mean a worse prognosis. There are no connections between IGF-1 and sanitary resources used (TISS score) but these connections seem to get better when GH levels are higher. 

Trimodal distribution of deaths and the golden hour concepts are in part responsible for the genesis of all modern trauma systems but these concepts have been challenged recently. Our aim was to describe distribution of death in trauma using data from a trauma system and discuss what can be done from the organizational point of view to improve outcome.

All traumatic deaths occurring between 2001 and 2005 in a trauma system were. Data on age, gender, time and place of injury, time of first and second hospital arrival, cause of trauma and type of accident, hospital characteristics, dominant injury and time of death were collected for this study. For mortality distribution the variable time was transformed applying a natural logarithm.

RESULTS. 1436 deaths occurred over a period of 53 months. 52% at the scene, 18% in the level I trauma centre, 21 % in level III trauma centre and the remaining in level IV/V trauma centre. Death distribution using a logarithmic scale in minutes showed four peaks: deaths at the scene, deaths in the first hours, deaths in the first two days and finally deaths in the second week that we referred as 2 minutes, 2 hours, 2 days and 2 weeks peak (Image1). We found statistically significant differences in age and dominant injury concerning timing of death.

A tetramodal pattern of death distribution could be described. Our data support the need to focus on the treatment of severe head injuries namely in the intensive care environment. 

Anaemia is usually detected in critically ill patients. Red Bloos cell Transfusion is not free of risk. We want to start an alternatives to transfusion protocol but fist we tryed to dercrive our critically ill patients anaemia. Our objectives were to: Study the red blood cell and iron metabolism in the ICU patients at admission. Observe changes in these parameters across the first seven days after admission. Observe RBC transfusion and his relation whit morbidity and mortality. Find transfusion predictors at the admission moment.

During tree mounths of 2006, we include all the admissions in a Trauma and Neurocritical ICU of our hospital that stay in unit more than 24 hours. At the moment of admission we determinated Haematocrit, (Hto), Haemoglobin (Hb), and Reticulocytes (%retic) levels, Iron metabolism, Folic Acid, B12, EPO and Creatinin (Kr) We repeated determinations seven days after admission if patient was still in ICU. Adverse events occurred during ICU stay were also registered (mainly infections) together the number of RBC transfusions (with Hb levels before and after administration).

We included in the study 73 patients. Severe Traumas (41%), Neurocritical patients (38%), Tumoral Neurosurgery (6,8%) and other patients (13%) . Average age was 54.25 years and APACHEII 14.6 ± 6.9 points. 76% were males . Results of admission blood determinations and seven days after are exposed in Table I . There is a tendency to decrease in Hto and Hb parameters, but not significant. The only parameter we observe difference statistically significative was the Reticulocites rate (%retic), significative lower 7 days after admission. (p<0.05) In Graphic 1 we describe anaemia groups in admission and the evolution of anaemia groups seven days after admission. We appreciated that no anaemia group suffers a severe decrease. 30% of patients were transfused during their fist week stay. Average levels of pre-transfusional Hb were 8.01 g/dl . We analysed transfusion predictors. Hto and Hb levels at admission predict transfusion. There is no other analytical parameter at admission that predicts transfusion. We also detected Tracheal intubated patients at admission and patients with inotropic drugs perfusions at admission were significative more transfused (p<0.02 and p<0.05). CONCLUSION. The high mortality rate in our patients is related to the initial GCS and cranial CAT at the moment of admission. It is necessary to continue the study to determine the influence of the rest of the variables in the mortality rate of these patients. INTRODUCTION. Traumatic brain injury, subarachnoid hemorrhage (SAH) and spontaneous intracerebral hemorrhage (ICH) are associated with systemic inflammatory response syndrome (SIRS). Early diagnosis of sepsis versus SIRS is frequently difficult in neurointensive critical care units. Procalcitonin (PCT) has been used as a predictor marker of bacterial infection in different groups of patients. There is variable and scarce information about PCT in neurocritical patients. The aim of this study was to evaluate the utility of serum PCT in the early diagnosis of fever from bacterial infectious origin in patients with acute brain hemorrhage.

We made a prospective diagnostic study between July 2004 and January 2005. We analyzed serum level of PCT and C-reactive protein (CRP) on consecutive patients with diagnosis of SAH, ICH or TBI who have fever during the Intensive Care Unit admission. We excluded patients with antibiotic therapy previous to admission. PCT and CRP were blindely measured from samples of serum extracted within 48 hs of fever onset and within 24 hs of antibiotic administration. Blinded to PCT and CRP results and according to previously defined criteria patients were classified in two groups: proved bacterial infection (PBI) and non proved bacterial infection (NPBI). Serum PCT was measured by immunochromatographic semiquantitative method BRAHAMS PCT-Q (Brahams Diagnostica, Berlin, Germany). Its sensitivity is 0.5 ng/ml. We analyzed sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of serum PCT and CRP for diagnosis PBI. We defined negative serum PCT as <0.5 ng/ml and negative CRP as <100 mg/L.

We studied 17 patient, 6 with SAH (35%) and 5 with ICH (29%). Ten patients had PBI (59%, 95%CI 33-82%). PBI were pneumonia (6), urinary tract infection (3), meningitis (2) and central line associated blood infection (2). Two patients had simultaneous infection sources. There were 2 bacteremic infections. PCT was positive in 4 patients in PBI group (3 pneumonia and 1 bacteremic central line associated blood infection) and in 0 of NPBI. Sensitivity was 40% (95%CI 12-74%), specificity 100% (95%CI 59-100%), PPV 100% (95%CI 40-100%) and NPV 54% (95%CI 25-81%). CRP was positive in 10 pts, 5 PBI and 1 in NPBI. Sensitivity was 50% (95%CI 19-81%), specificity 85% (95%CI 42-99%), PPV 83% (95%CI 33-99%) and NPV 55 (95%CI 23-83%).

In this study serum PCT had an adequate PPV to diagnose PBI, without false positives results. However, it has a low negative predictive value to diagnose PBI. Due to the results obtained, we consider that the quantitative PCT assay with a sensitivity limit of 0.1 ng/ml should be used for the future study to evaluate the role of PCT as a predictor marker of acute bacterial infection in patients with acute brain hemorrhage. 

In different published series cerebral infarction occurs in 30-60% of patients with symptomatic vasospasm after subarachnoid hemorrhage (SAH) despite maximal therapy. Standard triple-H treatment is associated with life-threatening side-effects (such as myocardial ischemia and pulmonary edema) and has not been properly validated. Milrinone, a phosphodiesterase IV inhibitor, has few side effects and exhibits inotropic, vasodilatory and immunomdulatory properties besides inhibiting platelet aggregation and thromboxane A2 synthesis. We present our experience using our M&H protocol (milrinone and homeostasis) in patients with vasospasm. It consists of CVP-guided normovolemia (maintain CVP=or>6), aggressive temperature control, maintenance of normal serum sodium and step-wise interventions based on symptoms (milrinone 0.05-0.1 mg/kg bolus plus infusion, levophed and angiogram plus intra-arterial milrinone).

We retrospectively reviewed the charts and imaging studies of 60 patients diagnosed with symptomatic vasospasm based on the development of focal symptoms and the results of angiographic and doppler studies. Cerebral infarction was defined as a new hypodensity on CT scan appearing at least 2 days after aneurysm clipping or coiling. CONCLUSION. Among the different physiological scores, the SAH-PDS was most strongly associated with the major outcomes and the H&H score was better than the other aneurysmal bleed scores. The strong association of physiological scores with outcomes suggest that interventions targeting physiological derangements may improve outcomes in SAH patients. 

Contrast Induced Nephropathy (CIN) is the acute deterioration of renal function due to parenteral administration of radio-contrast media. CIN is defined as an increase in serum creatinine concentration of >44 µmol/L (0.5 mg/dL) or 25% above baseline within 48 hours after contrast administration.

[1]Epidemiologic data in neurosurgical patients undergoing endovascular coiling are sparse and only one study in stroke patients reported figures of 5% prevalence. [2] CIN is associated with increased morbidity, length of hospital stay and costs. Pre-existing renal failure and the dose of contrast media are known risk factors for the development of CIN in cardiac patients where the condition is well-described.

[3]Although the pathogenesis of CIN is not entirely clear, several mechanisms for contrast-induced renal injury have been proposed, including alterations in renal medullary perfusion, direct cytotoxicity and oxygen-free radical generation.

[4]

We conducted a twelve month retrospective electronic patient record based review of data from patients presenting to the hospital for endovascular coiling. Renal dysfunction was based on increase in serum creatinine of 44 µmol/L (0.5 mg/dL) or 25% above baseline within 48 hours after contrast administration; the incidence of contrast induced nephropathy was investigated. Peri-operative care and post-operative management were analysed. A multi-variate analysis of risk factors was conducted and statistical tests done using Microsoft Excel.

. 149 patients visited our hospital neurosciences unit and underwent endovascular coiling over a one year period (Sept 2005-Sept 2006). The incidence of contrast induced nephropathy was 9%. 1.3% had pre-existing renal disease and 0.6% needed haemofiltration on intensive care for renal failure post-operatively. The odds ratio for developing CIN in patients with Diabetes Mellitus was 3.4 (0.8-14.2) p=0.2. The odds ratio for developing CIN with pre-existing kidney disease was 1.89 (0.37-9.4) p=0.54. The development of CIN did not show any correlation with patient age, emergency or electively performed procedure or the number of coils used. No anti-oxidants were given for prophylaxis and no protocol for peri-operative hydration was used though fluids were administered intra-operatively.

CONCLUSION. CIN is a common cause of acute renal functional impairment and accounts for significant morbidity in patients undergoing endovascular coiling. Patients with pre-existing renal failure are at high risk; other predisposing factors should be identified. There is some evidence regarding use of peri-procedural hydration and anti-oxidants and, therefore, management protocols should be developed. 

Open prospective observational study. Were studied patients treated by embolization after spontaneous intracranial aneurysm rupture. Included: Embolization complicated by rupture of aneurysm during the obliteration procedure. Rupture was ascertained by extravasation of contrast.

Current results of period ranging from July 2006 till October 2006. Thirty two patients embolized for 32 aneurysms. One patient pesented a rupture during the embolization: She was a 41 y.o female; she came to our institution's emergency suffering from acute headache, nausea vomiting and a mild meningism. She got no neurological defect (WFNS Grade I). CTscan showed a mild SAH (Fisher Class 1). An angiography followed, confirming presence of a 3 mm ruptured pericallosum aneurysm. During embolization procedure, a sudden hemodynamic instability (bradycardia, unstable blood pressure) was noticed and rerupture of aneurysm confirmed by extravasation of contrast medium. This complication occurred during placement of the first coil. The procedure continued successfully and aneurysm was completely obliterated by three coils. CTscan performed immediately after end of the procedure showed no massive cerebral haemorrhage (Class 2 Fisher). The patient was thereafter transferred to our ICU where she was extubated. She developed a transient neurological defect (right hemiparesis). She was discharged alive without any disability.

Aneurysmal perforation during embolization seems to be a rare event. In our case it doesn't cause much damage, but clinical severity is variable and far from being predictable. Re-bleeding can result in severe intracranial hypertension and ultimately brain death. Aneurysm thrombosis complicated 2 procedures, and was fatal for both (respectively 3 and 5 days after embolization) due to massive ischemia (aneurysm of the internal carotid artery) and refractory intracranial hypertension (aneurysm of anterior communicating artery). Those two patients got respectively WFNS Grade / Fisher Classification: III/3 and II/2. The patient with WFNS Grade IV got a successful uncomplicated procedure 14 days after the initial insult and partial clinical recovery. He continued to improve and was discharged alive from hospital without major neurological disability (GOS: good, modified Rankin scale = 2).

CONCLUSION. Endovascular coiling could be an efficient therapeutic tool. Incidence and outcome of procedures complications is still to be determined. Strategy in patients with high WFNS grade is certainly try embolization because of too risky surgery. Right management timing is still to be determined.

The quantitative estimation of blood loss helps in the choice of the best treatment tactics. The purpose of the study is to evaluate the ability of Central Blood Volume Index (CBVI, volume in heart and lungs and large vessels divided on body weight) and Total End Diastolic Volume Index (TEDVI, sum of the end-diastolic volumes of the atria and ventricles divided on body weight) to reflect the magnitude of a hemorrhage.

Normo-volumic values of CBVI and TEDVI were measured in 14 cardiac ICU cardiac patients, 2 pigs and 8 rats with weight range of 0.3 kg to 95 kg. Blood loss in the order of 25-35 ml/kg (3-4 steps) was applied in 8 rats and 2 pigs. Ultrasound dilution technology utilizes the decrease in blood ultrasound velocity caused by injecting isotonic saline, and can be used in species of any size. Cardiac index (CI), CBVI and TEDVI were measured by HCP101 (Transonic Systems Inc., USA) before and after blood loss. A disposable extracorporeal AV loop filled with heparinized saline was connected between an existing artery catheter and central venous catheter. Reusable ultrasound sensors were clamped on to the arterial and venous limbs of the loop. A peristaltic pump (Nipro, Japan) was used to circulate the blood from the artery to the vein at 8-12 ml/min for 5-7 min. Measurements were obtained by injecting 0.5-1 ml/kg (max 25 ml) of isotonic saline. At the conclusion, the AV loop was flushed with heparinized saline.

In normo-volemic situations indexes are in the range of CBVI = 9-15 ml/kg and TEDVI = 5-9 ml/kg, despite 30 times differences in weight. A dramatic blood loss of 25-35 ml/kg in experimental animals produces the same magnitude 48-50% decrease in CBVI and TEDVI. Severe dysphagia associated with silent aspiration and the danger of asphyxia requires translaryngeal intubation or tracheostomy. The aim of the study was to apply the clinical screening test (CST) and fibrooptic evaluation of swallowing test (FEST) to determine the best method of upper airway protection.

It was a prospective cohort study during the period of 2003-06. It included 1643 patients operated for FPT. All patients were delivered to ICU intubated and mechanically ventilated after operation. After full recovery from anesthesia, returning to consciousness and passing spontaneous breathing test (SBT) (if not -mechanical ventilation continued) they underwent CST of 6 points. The patients who passed CST without deficit were considered to have none or low level of dysphagia. The patients who passed CST with some deficit were considered to have dysphagia. All the patients were extubated and underwent FEST. In patients with poor CST, ICU crew was ready to perform translaringeal intubation immediately if necessary. Patients with severe cases of dysphagia underwent tracheostomy and received cuffed tracheostomy tubes to prevent aspiration and ensure free air passage. On the next day after performing tracheostomy, swallowing rehabilitation therapy began. Tracheostomiesd patients underwent FEST every week. After passing FEST with blue dye, decanulation was possible.

RESULTS. 345 patients of total group who did not recover consciousness or did not pass SBT in 48 hours after operation were determined for prolonged artificial ventilation and were excluded from further study. The 1155 patients who passed CST without any deficit were successfully extubated and showed absent or mild dysphagia in FEST. 143 patients passed CST with deficit and after FEST were divided into three groups by the level of dysphagia 52-mild, 39-intermediate and 52-severe. The regress of swallowing disorders was evaluated by FEST every week. In the first group the earliest recovery was in three days, in the two other groups none recovered earlier than after three weeks. The latest recovery was determined after a year of swallowing rehabilitation therapy. Two patients were not decanulated at all. Postoperative recovery made possible to reduce RS. But insufficient RS exhaust the patient and may result in secondary impairment of the brain. The aim of the study was the analysis of different respiratory strategies in these patients to choose the best.

It was a prospective cohort study of 101 patients after removal of PFT with complicated postoperative period during 8 and no significant difference in±2005-06. The age of the patients was 34 severity of complications and neurological status. All patients included into study demanded RS after operation because of low RD. All patients had bulbar palsy syndrome (BPS). 64 patients with BPS were tracheostomiesed. After full recovery from anesthesia and returning to consciousness ventilation modes were SIMV+PS or CPAP+PS (ventilator PB 7200). RR (respirator and patient), Tv, PS, FiO2, PEEP, paO2 and paCO2 and neurological status were evaluated and registered daily. The criteria of readiness to wean were determined as: PaO2/FiO2>200, PEEP<5, PS<8-10, SpO2>90%, RR<30, FiO2<30%, GCS>10. Weaning was successful if patient could breathe spontaneously for more than 48 hours without neurological deficit arise.

Patients were divided into 3 groups: 1. SIMV+PS ventilation (respirator RR 50-80% of total RR) -41 cases; 2. CPAP+PS ventilation -33 cases; 3. Failed extubation in first 72 hours -27 cases. All patients of the 3 group were ventilated in SIMV+PS after reintubation. The patients of the 2 group were extremely unstable and the modes of ventilation were corrected 4-10 times per day. Duration of ventilation was minimal in the 1 group with maximum replacement of spontaneous breathing with artificial ventilation -SIMV+PS (Table 1 ). In this group was minimal number of breathing disorders (minimal number of ventilator mode corrections) and patients were most stable. In first group was tendency to regress of BPS (73 %) and there wasn't cases of arising neurologic deficit. But in the 2 group there was increase of bulbar palsy syndrome in 46% cases and no regression. 

C. A. Eynon* 1 , P. Collins 2 1 Neurosciences ICU, Wessex Neurological Centre, Southampton, 2 Wessex Regional Transplant, Queen Alexandra Hospital, Portsmouth, United Kingdom

The management of severe brain injury in the UK is undergoing significant change. National recommendations are that all severely brain-injured patients are referred to specialist centres. Protocolised guidelines for the management of brain injury have resulted in improvements in mortality and morbidity. With this has come a reduction in the numbers of brainstem dead patients suitable for solid organ donation. However, there still exists a group of patients for which continued treatment is felt to be futile and who may be suitable as solid organ donors following death by cardiorespiratory criteria.

All deaths during a 24-month period were audited prospectively. When patients did not fulfill the requirements for brainstem testing, futility in continuing medical treatment was determined by the supervising consultant neurosurgeon, neurointensivist and senior nurse. In such patients, treatment other than comfort care was withdrawn. Patients (<75 yrs) where medical treatment was to be withdrawn were considered for NHB organ donation.

. 75 patients died during a 2-year period. 13 patients had death confirmed by brainstem tests of which 5 became solid organ donors. 28 patients were potential NHB donors. NHB donation was considered in 26 cases and offered to the family in 23. In one case the next of kin were untraceable, in one case the coroner refused permission. Consent for donation was obtained from the family in 15/23 cases. NHB organ donation occurred in 9 cases. In the remaining 6 cases, 4 patients died outside the time window for organ retrieval, in one the next of kin withdrew permission and in one the coroner did not grant permission. 3 of the patients who died outside the time window for NHB organ donation, subsequently donated tissue. A total of 18 kidney transplants, 5 liver transplants and one double lung transplant were performed from NHB donors.

CONCLUSION. The number of brainstem dead patients is declining in the UK. Patients in whom continuation of medical care is felt to be futile can provide a source of solid organs suitable for tranplantation. Successful transplantation of solid organs from potential NHB donors occurs in a significant proportion of cases. Feedback from family members has been supportive regarding the decision to donate. 

The studies on treatment of patients with head injury and brain damage, with sudden cardiac arrest due to various reasons revealed, that it is very useful to introduce neuroprotective therapy in those patients. It allows to decrease the consequences of local and global brain ischemia. The aim of the study was to present the efficacy and tolerance of treatment with amantadine sulphate (Amantix, Merz, Germany), as a neuroprotective therapy.

In the Intensive Care Unit, between 2004 and 2007 we monitored a group of 67 patients with consciousness disorders, in the age of 65.85 +/-12.85, with average BMI of 26.48 +/-3.49. The level of coma's deepness and its reasons were different. The examination plan, methods used, choice and classification of patients were carried out based on previously prepared protocols. The minimal period of treatment with intravenous infusion of amantadine sulphate was 7 days, however, if possible, the therapy was continued for 14 days. After this period the patients received Amantix in tablets. Many additional therapeutic measures from different groups were used in those patients. An endotracheal intubation and ventilation were necessary in all of the patients. Amantix was used as treatment's supplementation in the dose of 2x200mg/day. At the admission the patients were classified with the use of GCS (Glasgow Coma Scale). In order to evaluate the effects of use of the preparation, some specific function of the patients were examined before the use of Amantix and after finishing of the therapy. The examination was carried out by the Intensive Care Unit doctors, neurologist and nurses taking direct care of the patients. The results were compared with the control group of 55 patients, age 48,35 +/-25,52. Those patients were treated with the use of standard methods. All of the collected data were worked up statistically.

The authors revealed statistically important difference in GCS grading between the groups. The average GCS score in Amantix group at the admission was: 5.55 +/-1.25, and at the discharge: 12.46 +/-1.21. Analogically, in the control group the admission score was: 4.82 +/-2.11, and at the discharge: 8.76 +/-4.21. In 5 patients using Amantix we have noted the presence of side effects, usually it was hiperactivity. 55 patients were transferred to different wards. 12 patients died. The average hospitalization period in the Amantix group was: 20.08 +/-8.05, and in the control group: 25.58 +/-7.25 days. . This has been fuelled by increasing evidence demonstrating either sub-optimal care or poor end of life decision making as antecedants to cardiac arrest calls on acute wards. Outreach and Medical Emergency Teams have developed as a result, but their effectiveness remains unproven [2] . At Southend, development of a Critical Care Outreach service began in 2002. The aim of this study was to establish the trends in cardiac arrest call rates from the acute wards in the years prior to, during and after the introduction of the Outreach team, to assess any potential impact this may have had.

Hospital switchboard records were analysed retrospectively to provide data relating to the date, time and location of ward cardiac arrest calls occurring between January 1 2001 and December 31 2006. Arrest calls to all acute wards except the Critical Care Unit were included. The data collected was then related to hospital inpatient activity (in terms of completed in-patient consultant episodes, supplied by the Hospital's Information Department) to enable meaningful interpretation of the observed trends. Table 1 summarises the results from the Medical and Surgical wards separately and then together to present data for the hospital's acute wards as a whole. The data shows an upwards trend for the years prior to and during the establishment of the Outreach service, and a falling trend subsequently. CONCLUSION. The establishment of a comprehensive Outreach service that promotes all aspects of Outreach Critical Care (expediting appropriate and preventing inappropriate critical care admissions, following up patients post critical care discharge and promoting critical care skills throughout the hospital) is likely to lead to a reduced frequency of cardiac arrest calls. However, this effect may take years and not months following introduction to be manifest. We suggest all Outreach services should collect and present this simple data locally to demonstrate the potential impact of their activities. 

Intracerebral haemorrhage (ICH) represents 10-30% of all strokes. The acute and subsequent blood pressure management presents a therapeutic dilemma. It is necessary decrease high systolic blood pressure, but there is the risk of decrease cereb. Objective: Can the regional cerebral oximetry helps us to determine individual adequate blood pressure? ral perfusion pressure and risk of ischemia developing. METHODS. Regional oxymetry is the method of measurement the cerebral oxygen content based on near-infrared spectroscopy, which is carried out by means of the INVOS device (In Vivo Optical Spectroscopy). This method is non-invasive, delivers continuous information and it allows the possibility of emergency therapeutic response. rSO2 is transcutaneous monitoring of regional cerebral saturation with hemoglobin oxygen (rSO2) in mixed blood in the frontoparietal regions, which represents interface beetween the basin of the anterior and middle cerebral arteries. The normal value of rSO2 is beetwen 55-75% in a majority of the population, and every change from the baseline in both directions by more than 20-25% signifies the risk of ischemia for the observed tissue. During a twelve-month period all pacients admitting with ICH in our neurointensive care unit (NICU) were managed by regional cerebral oximetry (n = 20). Arterial blood pressure was monitored and was corrected farmacologically. The functional outcome of patients when discharged from the NICU and after six month were evaluated by the Glasgow Outcome Scale, Barthel index and Modified Rankin Scale. Data was collected retrospectively for comparison with pacient which didn't monitor by rSO2.

We found correlation between discovery of patological rSO2 values and age, initial GSC and volume of ICH. There are less septic and hemodynamic complications in the group with monitoring rSO2.

Using this method, the probability of successful improving outcome all patients with intracerebral haemorrhage will be estimated. There is the need for guidelines regarding the blood pressure managemet of these patients. 

Elaborated data are available on ICcollege.be. 39 of 134 (28,3%) ICU directors, representing 637 ICU beds completed the extended query. Main findings were: visits limited < 2h/day (1h36 + 1h56) ; HCP dedicated to family (68%) children admitted from 10 y of age (89%) ; family accompanied by HCP during resuscitation (74%) ; no witnessed resuscitation procedures (5%) ; scare possibilities for family to stay during night (42%) ; insufficient bad news delivery (65,79%) ; poor team psychological support (5-11%). 147 ICU physicians completed the follow-up simplified query. Main findings were: psychological support for family (59,86%) and team (27,89%) ; post-resuscitation debriefing (31,97%) ; identification of dedicated HCP (59,86%) ; use of (90,48%) and written (96,60%) DNR-orders ; comprehension of (98,64%) and family witness (97,28%) of patients' will ; structured bad news delivery (97,96%) ; witnessed resuscitation (4,76%) and invasive procedures (2,04%) ; children accepted < 10 y (4,08%).

In Belgium, although there's obvious concern from the majority of ICU's to communicate with relatives, recommendations for psychological team support, teaching bad news delivery, schedule of visits and witnessed procedures are made. 

Sudden death constitutes an important sanitary problem. Early diagnosis and advanced cardiorrespiratory live support are considered the most important factors related with short term prognosis. The objective of this study was to analyze the prognosis, clinical characteristics and evolution of patients who initially recovered after an episode of out-of/hospital or in-hospital cardiac arrest and who were admitted to a medical-surgical intensive care unit (ICU).

Sixty three consecutive patients were included and retrostectively studied when they were admitted to a medical-surgical ICU.

For two years, from April of 2004 until April of 2006, sixty three consecutive patients were included. Eighteen of the patients were women (28.6%) and 45 were men (71%). CPR was given out of hospital to 17 patients, and 28 patients suffered sudden death on a conventional hospital ward and 18 patients in special units (surgery, coronary, emergency room, etc.). The etiology of the arrest was considered to be of probable primary cardiac origin in 60% of the episodes and the rest of the origin of arrest was considered secondary to other pathologies (respiratory, sepsis. . . ). Mortality in ICU was 52.4% and 48,6 % were discharge alive but of that percentage of patients only 20% were released without important neurological damage. Patients recovering following cardio-pulmonary arrest out of hospital and hospital ward had greater mortality than those who suffered an event in a monitored area (25%).(p<0,01) The lengthy resuscitation times (greater than 15 minutes), elevated Apache II scores and advanced age is associated with greater mortality.

Recovered cardiac arrest is a pathology with high mortality and morbidity in intensive care. In our series only 20% were released alive without severe neurological damage. The existing condition of the patient and the excessively long resuscitation times were decisive factors in these results. 

We conducted a retrospective case-note study in a six-month period at an innercity district hospital (distant from any international airport), and report three patients who deteriorated about the time of overseas travel by air.

RESULTS. Case 1. A retired gentleman of 74-years with progressive idiopathic pulmonary fibrosis requiring home oxygen therapy travelled by air without a medical escort. He deteriorated shortly after his arrival at the family home in the UK. He presented to the Emergency Department in respiratory failure requiring non-invasive ventilatory support. He died during prolonged hospitalization. Case 2. A 51-year old woman with obstructive sleep apnoea reduced her diuretic prescription without her physician's knowledge prior to a long-haul flight. She deteriorated with acute shortness of breath shortly after her arrival at the family home in the UK. She was brought by her family to the Emergency Department where she was found to be in cardiogenic pulmonary oedema, requiring non-invasive ventilation. She survived hospitalization and was discharged with home oxygen therapy. Case 3. A 39-year old man collapsed in the street explaining to passers-by that he had swallowed some packages. He had a travel ticket from the airport in his possession but was able to give no other history. He was taken to the Emergency Department and required intubation due to extreme agitation. He was found radiographically to have ingested multiple wrapped packets. He required laparotomy to remove 36 differently coloured packs some of which had ruptured releasing their contents. Urinalysis revealed cocaine metabolites. He subsequently made an uneventful recovery after extubation and transfer to a surgical ward.

Patients may present to hospitals distant from international airports with clinical deterioration consequent upon risks associated with long-distance air travel. (2000) 

Prospective observational study of a cohort including every septic patient admitted in a medical ICU of an university hospital from May 2004 to December 2005. Demographic, clinical, laboratory and therapeutic variables were registered. A clinical examination assessing motor deficit and tendon reflexes was daily performed in order to check CIPNM criteria. Univariate and multivariate logistic regression tests were used.

. 378 septic patients were included with age 60±17, APACHE II score 22±7, maximum SOFA score 9.7±4, ICU mortality 37%, in-hospital mortality 44%. 290 patients survived at least 12 days. 93 patients did not require MV and none of them developed CIPNM. Finally the analysis was performed with the 197 patients who survived at least 12 days and required MV, with a CIPNM incidence of 39%. 73 variables were included in the univariate analysis. After multivariate analysis, it was found that several variables were significantly related with risk for the development of CIPNM (odds ratio, OR; 95% confidence interval, IC; signification level of change in 2 log likelihood, p): 1. MV length (days): OR 1. 

Patients in the ICU often develop an acute neuromuscular disorder characterised by difficulty of weaning from mechanical ventilation and associated with variable degrees of muscular weakness including quadriplegia [1] . Often associated with steroid treatment, neuromuscular blocking agents (NMBA) and septic patients, the pathogenesis of CIM is poorly understood [1] . Originally thought to be neuropathic in nature, however, today myopathy is more often diagnosed [2] . To further clarify this point we present a series of 12 patients.

Between 1994 and 2002 a retrospective study was carried out on 12 patients diagnosed with CIM and whose muscle samples were analysed in the Dept. of Neuropathology of CHUVI, Spain. In the clinical studies special attention was paid to the neuromuscular status APACHE II, and treatments with steroids, NMBA,total parenteral nutrition (TPN)and insulin. All patients underwent electromyographic studies and biopsy and in those with sensitive neurography an abnormal nerve biopsy.

Of the 12 patients, 4 were women and 8 were men, all aged between 46 and 86, (mean 66±11). In three of the patients admission to the ICU was not necessary. All save two received prolonged high doses of steroids and two were on chronic treatment of steroids. Only one was treated with NMBA for more than 15 days. Two patients were diabetic with no electromyographic signs of neuropathy. Seven needed insulin to control glucemia during the critical period. 5 received TPN, and 9 had sings of sepsis. Muscle biopsy showed signs suggestive of CIM (atrophy of both types, alteration of the intermiofibrilar pattern) and in some cases miofagia and thick filament loss. In two cases there was discrepancy between neurophysiologic and biopsy findings (muscle and nerve). The seven patients that survived the acute illness showed neuromuscular symptoms on release from hospital. Follow up was possible on three patients for 5, 8 and 9 years respectively. All recovered muscle strength, the electromyography normalized and currently have normal independent daily life activities. The aim of this clinical trial is to study CIP in ICU patients (pts) after surgical procedures.

We enrolled retrospectively 64 ICU pts (49 men (76.6%), 15 women (23.4%) who underwent at least one surgical procedure under general anaesthesia and developed CIP. All of them were mechanically ventilated and stayed > 12 days. Underlying diseases: multiple trauma 43, complicated surgery 19, pancreatitis 2. Mean age: 42.8±18.1 years. Operation sites: Abdomen 28, CNS 26, orthopaedics 12, thorax 4, other 3. Mean anaesthesia time: 129±24 min. In all pts an electromyogram was performed twice, as well as daily neurological examination. We analyzed several parameters predisposing to CIP. CONCLUSION. 1) Sepsis predisposes to CIP, but CIP can be appeared without sepsis (32.8%). 2) Age and serum albumin values do not predispose to CIP (p<0.1); however the early implementation of a nutritional protocol is useful. 3) Although not well correlated, we try, if possible, to avoid neuromuscular agents. 4) High PGL predispose to CIP (p<0.01); it is important to maintain PGL < 110 mg%. 5) CIP prolongs LMV (p<0.01), LOS in ICU (p<0.01) and LOS in hospital (p<0.001), but does not increase MR significantly (p<0.1).

S. Kjaergaard* 1 , S. E. Rees 2 1 Intensive Care, Anaesthesia and Intensive Care, Region North Jutland, Aalborg, 2 Center for Model-based Medical Decision Support, Aalborg University, Denmark (1) is accepted as the gold standard method of describing pulmonary gas exchange. In the clinical setting, if any, only very simple one-parameter models are used. The parameters of these varying upon changing the FiO2. In a previous paper we have compared the MIGET with a simpler model, and shown that this simpler model is a good fit to the inert gas data obtained from the MIGET experiment (2) . This study explores whether the simpler model can reproduce oxygenation data in an oleic acid lung damage model upon changing the FiO2 and compared these results with those obtained using the MIGET.

Seven pigs were used for the study. Lung damage was induced by an intravenous infusion of oleic acid. Six inert gases were infused to estimate the distribution of V/Q-ratios of the MIGET model and dead space, shunt and a parameter describing V/Q mismatch, i.e. fA2, of the simpler model (1,2). Measurements were taken at five different ventilator settings. The two models were then used to simulate arterial oxygenation data when the model-parameters along with measurements of mixed venous blood gases at different values of FiO2 were given as input to the models.

Both models can be used to simulate SaO2 at varying FIO2. This is shown in the figure where the models have been used to simulate SaO2 at varying values of FiO2 (MIGET "+", simple "squares") ranging from 0.21-1.0. It shows that the models simulate identical values of SaO2 with a mean difference = -0.2 +/-1.7.

Since the MIGET and the simpler models provide both equally good fit to the inert gas data (2) and precise predictions of arterial oxygenation, they might be interchangeable in a clinical setting where only a limited amount of data are accessible. In addition, the parameters of the simpler model can be obtained quickly and non-invasively (3). The model could therefore have applications a clinical situation. 

Ethanol may be used in the management of toxic alcohol poisonings 1 , or as sedation in alcohol withdrawal. Ethanol may be a component within drug formulations, for example nimodipine infusion or chemotherapeutic agents 2 . Ethanol flush has also been used to restore the patency of occluded catheter lumens 3 . In clinical practice, ethanol should only be infused via a pCVC and not a peripheral venous cannula, as the high osmolality of ethanol can cause thrombophlebitis. Given anecdotal reports of pCVC deterioration during ethanol infusion 2,4 , this study applied a bench testing method and statistical modelling to develop clinical practice guidelines at our institution.

The test solutions used were: dextrose (D) 5%; ethanol (E) 5%, 10%, 25%, 50%, 70% and 100%. Each test solution was perfused through 3 pCVCs. A total of 21 pCVCs were perfused. (b) 24 Hour Perfusion. The test solutions used were: D5%, E5%, E10% or E25%. Each test solution was perfused through 3 pCVCs. After perfusion, the strength of all pCVCs was assessed. The pCVC was attached to a force gauge. A known force was applied to the pCVC and the pCVC length was measured. This was repeated for increasing forces until the pCVC broke. Length-force relationships were plotted and were described statistically using linear mixed effects models.

. This bench test model produced reproducible data. The pCVCs were not directly traumatised by the testing apparatus. (a) 30 Minute Perfusion. pCVCs perfused with E50% , E70% or E100% perished with obvious structural deterioration. Two distinct length-force relationships were described on linear mixed effects models: E50%, E70% or E100% weakened the pCVCs , whilst D5%, E5%, E10% and E25% had no effect upon pCVC structure (p<0.05) (b) 24 Hour Perfusion. The pCVCs did not perish. On linear mixed effects models, E10% and E25% weakened the pCVCs, whilst D5% and E5% had no effect (p<0.0001).

CONCLUSION. This model quantifies the effect of ethanol infusion upon pCVCs. This has not been demonstrated previously. The infusion of E50% E70% or E100% via pCVCs should be avoided. Infusion of E10% and E25% for 24 hours weakens pCVCs. Nimodipine and other drugs using ethanol as a carrier vehicle should be infused via pCVCs with caution. These potential hazards should be outlined in individual pCVC package inserts and drug product information leaflets. (2005) 

In septic shock patients tissue microcirculation is altered despite an increased tissue oxygen tension (1). Microcirculatory distress could be one of the earliest stages in the progress of sepsis to multiple organ failure, and microcirculatory shunting could be an important contributing factor to this development (2) . SOFA score has been suggested to clinically assess the level of organ dysfunction(3). We've done a prospective observational study to determine if changes in the rate of thenar muscles tissue deoxygenation during stagnant ischemia in patients with severe sepsis and septic shock are related to changes in organ dysfunction using the SOFA score.

Fourteen septic shock patients were included in a preliminary study during the first days of sepsis evolution. , Hutchinson?Thenar muscle StO2 was measured noninvasively by NIRS (InSpectra Technology, USA) before and during upper limb ischemia. StO2 decrease (downslope) after limb ischemia were analyzed during first and fifth day after ICU admission. Changes in StO2downslope, SOFA score, cardiac output, lactate and the use of vasoactive drugs between first and fifth days were recorded.

We found good correlation between ∆StO2downslope and ∆SOFA between the first and the fifth day. (Spearman's rho = -0,693; p<0,01). Our results are in accordance with those reported by Pareznik(4) wich correlated isolated values of StO2 with SOFA in septic shock patients but moreover we show that changes in both variables during evolution are also correlated.

In septic shock patients, thenar muscle ∆StO2downslope is well correlated with changes in ∆SOFA, a clinically accepted tool to measure organ dysfunction evolution during sepsis. ∆StO2downslope monitoring could be not only a good marker of microcirculatory state but also a good indicator of organ dysfunction evolution during sepsis and consequently a potentially therapeutic objective. 

One of the important tasks that the anesthesiologist should perform is to monitor the functions of body organs; lung airway pressure is among the most important ones. A real-time continuous monitoring device which would be designed in a small volume and is portable could be used by anesthesiologists for this purpose. So, this device could improve the quality of anesthesia care while being efficient and cost containing.

The device consists of four consisting parts as follows: sensors (Pressure transmitter and Gas velocity transmitter), processors (two AVR microprocessors), monitor and software. Software simulation: The performance of the monitor was controlled through a simulation process with Matlab-Simulink software (the Mathworks Inc. MA, USA),(1).

The monitoring device demonstrated acceptable results, both clinically and at the lab assessments.

The study demonstrated this device as an effective, reliable and cost containing device. 

A. Rodríguez Salgado* 1 , A. Socias 1 , B. Comas 2 , A. Llompart 2 , I. Losada 3 , P. Ibáñez 1 , M. Borges 1 1 Intensive Care Unit, 2 Emergency Department, 3 Internal Medicine, H. Son Llàtzer, Palma de Mallorca, Spain

Since we have a global computerized system on our hospital we used it to develope an integral and multidisciplinary working protocol for the early recognition of sepsis and its appropiatte therapy.

Prospective study conducted in a four-hundred bed teaching hospital with medical and surgical areas and the support of a global computerized system and on line internet conexion among areas. A computerized protocol to improve management of sepsis was developed. It automatically produces an annotation on the medical chart and a serie of analytics forms when activated. Additionally clinical guidelines on sepsis management can be consulted. It was started on January 2006, and here we present all patients included until January 2007.

During the study period 313 patient were included in the protocol, with a mean age of 63,92 (16,10) y, 65,4% were male. We have observed an ascending tendence in the number of patients included in the protocol, having arised from 8 patients on January 2006 to 50 on January 2007. The protocol was activated at the ICU in 129 (51,8%) cases, at the emergency department in 99 (39,8%) and at hospitalization units in 21 (8,4%). Two-hundred and two (65,4%) patients were admited at the ICU. Though initially the protocol was exclusivelly directed to patient with severe sepsis or septic shock, lately some patients with sepsis have been included. So, 25 (8%) had sepsis, 155 (49,8%) severe sepsis and 131 (42,1%) septic shock. Only 158 (50.5%) had fever and 179 (57.2%) had arterial hypotension at the protocol entry. Sepsis was community-adquired in 223 (72,4%) cases, nosocomial-non ICU adquired 61 (19,5%) cases and ICU adquired in 24 (7,8%). The the most frequent site of infection was the lung in 134 (43,1%) patients, followed by the abdomen in 80 (25,7%) patients. Isolation of the causal microorganism was achieved in 219 (70%) patients. Blood cultures were positive in 192(61.9%) cases. Forty seven (15%) had 1 organ disfuntion (OD), 81 (25.9%) 2 OD, 68 (21.7%) 3 OD and 63 (20.13%) 4 or more OD. Mean lactate levels were 2,66 (2,21) mmol/l, 2,30 (2,11) mmol/l and 2,01 (1,99) mmol/l at the activation moment, at 6 and a 12 hour respectively. Mean C-reactive protein levels were 78,8 (22,89) mg/l. Eighty-five (27,2%) patients deceased, of whom 2 (8%) had sepsis, 30 (19,4%) severe sepsis and 53 (40,5%) septic shock at the moment of activation. CONCLUSION. It is possible to implement a global multidisciplinary computerized protocol for identification and management of the sepsis, although this is a laborious and continual process.

T. Kyprianou* 1 , G. Panayi 2 , D. Zeinalipur-Yazti 3 , M. Dikaiakos 3 1 Intensive Care Unit, Nicosia General Hospital, 2 Ngo, Intensive Care Forum, 3 Dept of Computer Science, Universiy of Cyprus, Nicosia, Cyprus INTRODUCTION. The physiological condition of ICU patients is marked by rapidly evolving and frequently life-threatening derangements as well as 'silent' yet important alterations in homeostasis. Reliable monitoring i.e. the capability to collect, store, process, and share inpatient monitoring data along with physicians' remarks can bring tremendous benefits to all aspects of Intensive Care Medicine (practice, research, education). Currently, Grid infrastructures assemble an extensive collection of resources and expertise (EGEE Grid: 200+ sites around the world with more than 30,000 CPU's -5PB of storage, adequate for storing and managing ICU-related data. We present the design and implementation of the Intensive Care Window (IC-Window), a software tool that enables the retrieval and integration of data from patient-attached medical sensors.

IC-Window follows a modular design to retrieve data from different patient monitoring devices. The tool includes a full-edged interaction protocol and graphical user-interface to interact with the Phillips IntelliVue MP70 monitor. IC-Window is implemented in the context of ICGrid (Intensive Care Grid), a novel data-grid framework that utilizes the EGEE infrastructure to enable the seamless integration, correlation and retrieval of 'clinically interesting episodes' across Intensive Care Units clusters.

We present preliminary data from software's use in 10 ICU patients.

CONCLUSION. IC window belongs to a new generation of tools that could improve dramatically Intensivist's capabilities as offers virtually unlimited storage capacity for every possible type of patient's data. In the future we plan to extend the IC-Window application to communicate with other medical devices found within the ICU. This will provide an open platform for the aforementioned applications. INTRODUCTION. Strict glycemic control by lowering blood glucose levels to 80 -110 mg/dL reduces the intensive care unit (ICU) mortality, morbidity, duration of the hospital stay, and overall medical care costs. To provide an intelligent system for tight glycemic control, the EU-project "Closed Loop Insulin Infusion for Critically Ill Patients (CLINICIP)" was started in January 2004. Three different sensor technologies -two based on an enzymatic reaction with immobilised glucose oxidase using either amperometry or fluorimetry as transducer and another based on reagent-free infrared spectroscopy -have been developed to continuously monitor the glucose levels in the subcutaneous interstitial body fluid.

Monitoring of the subcutaneous interstitial fluid is realized using a microdialysis catheter CMA60 from CMA Microdialysis AB as a body interface to all glucose sensors. Experiments were carried out at the Center for Medical Research (Graz, Austria), lasting up to 28 h with the probands starting under fasting condition, but receiving later their normal diet. After microdialysis probe implantation, the perfusate (either 5 % mannitol solution or ELO-MEL) flow rates were around 1 µl/min. For reference measurements, dialysate samples were collected. In parallel, blood glucose concentrations in venous blood samples, collected under arterialised conditions with the arm resting in a hot box, were determined using a glucose analyzer from Beckman Instruments.

A Clarke Error Grid analysis of the results from all three sensors has shown all values in clinically acceptable zones. The blood reference and sensor measurements were further compared using Bland-Altman plots. Owing to the tubing connecting the catheter outflow and sensor, the lag times for the sensor readouts were between 10 and 30 min. For the electrochemical and infrared sensors a simultaneous micro-dialysis recovery rate determination has already been implemented for improving the correlation of the sensor readout to the whole blood levels. 

Some observational studies suggest that the use of pulmonary-artery catheters to guide therapy is associated with increased mortality.

We performed a randomized trial to study outcome benefit of using pulmonary artery catheter (PAC) in ARDS patients when compared to standard care using central venous catheter (CVC). The subjects were 26 ARDS patients on mechanical ventilator who were assigned either to PAC (PAC group), or CVC (CVC group). The base-line characteristics of the two treatment groups were similar.

The primary outcome was ICU and in-hospital mortality from any cause. The PAC group had a significantly lower ICU mortality than the CVC group (2 vs 3, p value= 0.004) but there was no difference between the 2 groups in in-hospital mortality (one case mortality in CVC group). There were no significant differences between PAC and CVC groups in urine output (1.7± 0.89 vs. 1.7± 0.56), use of vasopressors (0.46 ± 0.51 vs. 0.61± 0.51), and length of hospital stay (4.1 ± 2.1 vs. 4.46 ± 2.7) respectively.

Our findings suggest that PAC can be used in ARDS patients for better hemodynamic assessment that may result in reduced ICU stay and mortality rate. 

Ethanol may be used in the management of toxic alcohol poisonings1, or as sedation in alcohol withdrawal. Ethanol may be a component within drug formulations, for example nimodipine infusion or chemotherapeutic agents2. Ethanol flush has also been used to restore the patency of occluded catheter lumens3. In clinical practice, ethanol must be infused via a pCVC, as its high osmolality can cause peripheral thrombophlebitis. Given anecdotal reports of pCVC deterioration during ethanol infusion2,4 , this study applied a bench test and a statistical model to develop clinical practice guidelines at our institution.

Each 20cm triple lumen pCVC was perfused with a single test solution only. (a) 30 Minute Perfusion. The test solutions used were: dextrose (D) 5%; ethanol (E) 5%, 10%, 25%, 50%, 70% and 100%. Each test solution was perfused through 3 pCVCs. A total of 21 pCVCs were perfused. (b) 24 Hour Perfusion. 12 additional pCVCs were perfused with D5%, E5%, E10% or E25%. After perfusion, the strength of all pCVCs was assessed. The pCVC was attached to a force gauge. A known force was applied to the pCVC and the pCVC length was measured. This was repeated for increasing forces until the pCVC broke. Length-force relationships were plotted and were described statistically using linear mixed effects models.

. This bench test model produced reproducible data. The pCVCs were not directly traumatised by the testing apparatus. (a) 30 Minute Perfusion. pCVCs perfused with E50% , E70% or E100% perished with obvious structural deterioration. Two distinct length-force relationships were described on linear mixed effects models: E50%, E70% or E100% weakened the pCVCs , whilst D5%, E5%, E10% and E25% had no effect upon pCVC structure (p< 0.05). (b) 24 Hour Perfusion. The pCVCs did not perish. E10% and E25% weakened the pCVCs (p< 0.0001).

not been demonstrated previously. The infusion of E50% E70% or E100% via pCVCs should be avoided. Infusion of E10% and E25% for 24 hours weakens pCVCs. Nimodipine and other drugs using ethanol as a carrier vehicle should be infused via pCVCs with caution. These potential hazards should be outlined in individual pCVC package inserts and drug product information leaflets. (2005) INTRODUCTION. Inadvertent esophageal intubation may lead to serious complications such as hypoxia, cardiac arrythmias and death. Auscultation of breath sounds may be an inaccurate method to determine correct endotracheal tube placement of endotracheal tube placement.1 Vibration response imaging (VRI) is a novel non-invasive technology that measures vibration energy of lung sounds during respiration. As air moves in and out of the lungs, vibrations propagate through lung tissues and are recorded by 36 sensors spacially distributed on the patient's back over the lungs and a dynamic image is created.

A 68 year old female patient presented with lung cancer. Plain chest radiograph and CT scan revealed a large left lung mass comparable for a neoplasm. She was admitted for left lung lobectomy. After informed consent was obtained, she underwent VRI before and after intubation. The esophagus was inadventently intubated and recognized immediately after the VRI recording was obtained. The patient went on to have a successful operation.

Analysis of the VRI data obtained during esophageal and tracheal ventilation are compared along with a normal VRI image. During esophageal ventilation most of the vibrations (60%) were detected by the upper sensors and the least by the lower sensors (8%) (Fig. 1) . Following the endotracheal intubation as well as in a normal image, the vibrations were more evenly distributed with the sensors from the middle region receiving more vibrations.

Quick detection of inadvertent esophageal intubation is crucial to prevent serious complications but commonly used methods of confirmation such as auscultation and plain chest radiograph are inaccurate or do not provide timely results. VRI is a novel technology that offers the potential to quickly identify inadvertent esophageal intubation in the OR and perhaps other settings. 

The Acapella ® is a small hand-held vibratory device that combines the resistive features of the positive expiratory pressure (PEP) and the vibratory features of a flutter valve to mobilize secretions in the airway. VRI is a novel dynamic imaging technique that measures vibration energy of lung sounds generated during respiration. In this study, our aim is to determine, using the VRI, what regions of the lungs receive the most vibrations when the Acapella is being used.

A 20 second VRI recording was performed on a healthy volunteer during normal breathing (first three breaths) and while using the Acapella device (last four breaths). The VRI recordings were obtained in 20 second periods of respiration. Dynamic digital images and numerical raw values for vibration energy are analyzed and compared any regions of interest.

. VRI images at maximal expiration while using acapella show increased total vibration intensity. When the distribution of expiratory vibration is examined, it appears that vibration from the acapella goes more to the lower lung regions (Figure 1 and 2) . 

Asymptomatic catheter-related central vein thrombosis (CVT) which is diagnosed by venographic studies is mentioned to be as high as 66 %. Moreover, when thrombosis occurred, the risk of catheter related sepsis was declared to be 2.6 % higher. In this prospective study we aimed to diagnose CVT early as possible, its incidence and risk factors.

ICU patients (pts) that needed a central venous access for at least 48 hours without chemotherapeutic agents administration were included in this prospective study. The catheters were inserted via internal jugular or subclavian vein at bedside under aseptic conditions using the Seldinger technique. Diagnosis of vein thrombosis was detected by color Doppler ultrasound examination performed in less than 24 h after catheter removal (Picture). The protocol was approved by the ethic committee.

Three hundred and thirty eight pts (154 F, 184 M), mean 58.8 years old (17-91 years), were included in the study. Catheters mean duration time was 11.62 days and duration of insertion mean time was 7.16 min (4-40 min). In 289 pts catheter insertion was performed with a single puncture, in 31 pts with double and in 15 pts with three and more punctures. Catheter localization was : in 256 pts right subclavian vein, in 61 pts left subclavian vein, in 15 pts right internal jugular vein and in 6 pts left internal jugular vein. Catheter related thrombosis was diagnosed in 23 pts (6.8%) while catheter infection was seen in 13 patient (4.1%) (table). Generally the chemotherapeutic agents administered via the central vein catheter have thrombogenic effect. When we study our CVT diagnosed 23 pts we found out that all of them were over 60 years old, the mean catheter duration time was 11.28 days (Table) . But these results were not statistically significant when compared with the other pts under 60 years old and more than 11.62 days of mean catheter duration time. Out of 275 pts who were not under anticoagulant therapy 14 had CVT while out of 62 pts under anticoagulant therapy 9 had CVT diagnose which was found statistically insignificant (p>0.05). Our results show that patients under anticoagulant therapy have a three fold more CVT risk ratio than the others who are not using this anticoagulant therapy.

Patients under anticoagulant therapy have to be followed more closely regarding to CVT.

The provision of good glycaemic control is thought to have some beneficial aspects in critical care patients. We have previously described the introduction of a web-based insulin dose calculator program to support the control of blood glucose in critical care.

The aim of this study is to describe a modified version of a calculator program based on Van de Berghe's studies. This allows nursing staff to enter blood glucose values together with the insulin infusion rate into a calculator. The calculator then provides a recommended insulin infusion rate to control blood glucose with the added ability to recommend small bolus doses of insulin when appropriate, store blood glucose concentrations, insulin rates, bed number and the date and time of calculation. We also modified our feeding protocol to restrict the target enteral feed from 1800 kcal to 1500 kcal per day and removed the night time rest period. We studied the data stored by the program which was used for all patients admitted to a 16-bedded intensive care unit (approximately 40% of whom have neurological injuries) between June 2005 and May 2006.

Overall there were 661 patients admitted (mean APACHE II score 16 [SD +/-7], with a mean age of 51 years [SD +/-17]. 146 patients died prior to ICU discharge. There was a total of 5573 patient days with 13029 recorded calculation data points. The mean blood glucose concentration was 6.7 mmol (95CI 3.7 -12.1). There were 34 episodes of treated hypoglycaemia of which 11 were on an insulin infusion. There were two troughs in the time of data entry that corresponded with staff handover. There was no diurnal variation in blood glucose concentration or in insulin infusion rates, although this did peak slightly in the early morning. The mean value of the insulin infusion rate was 4.3 units / hr (SD +/-3.7). In normal subjects there is a decreased level of endogenous insulin in the early morning, that is only partly lost with constant nutrition.

From this study we concluded that the web based insulin calculator facilitates the dosing of insulin in critical care in an economic manner. The lack of diurnal blood glucose concentration variation, suggests that once daily estimation of blood glucose may be an acceptable method of monitoring blood glucose concentrations in critical care. Systemic inflammatory response syndrome (SIRS) is a common entity in the intensive care units. Early institution of an appropriate antimicrobial regimen in infected patients is associated with a better outcome. Both C-reactive protein (CRP) and procalcitonin (PCT) are accepted sepsis markers. However, there is still controversy concerning the correlation between serum concentrations, infection and sepsis severity. Objective:to determine the clinical aplication of procalcitonin (PCT) and C-reactive protein (CRP) plasma concentrations in the detection of SIRS related to infection and sepsis and the assesment of severity of sepsis.

Desing: Prospective observational study. Setting: Medicosurgical intensive care unit. Patients: Over a period of 2 months (January-February 2007), forty seven consecutive adult patients admitted in a intensive care unit for an expected stay >24 hrs.and SRIS symtoms and signs. Informed consent was obtained from all patients. Measurements: PCT and CRP plasma concentrations and white blood cell counts , APACHE II Y SOFA within the first 24 h . Each patient was examined at the time of enrollment and was classified in one of the following four categories according to the ACCP criteria: SIRIS and sepsis group (sepsis, severe sepsis and septic shock). Statistical analysis: Were performed with SPSS 12.0. Differences in continuous variables between infected and non infected patients were compared with the nonparametric Mann-Whitney test. and lineal.regressión.

PCT levels were significantly higher in the severe sepsis(p=0,01) and shock septic group (p<0,01). PCT and CPR levels no weren found differences between sepsis of less gravity group and noninfectious SIRS. PCT and CRP levels are significantly correlated to the severity of organ dysfunction (SOFA y APACHE II). 

PCT and CRP levels were significantly higher withing short space of time in patient with infection than in patients with non-infectious SIRS, but for sepsis of less gravity, PCT and CRP plasma values not differentiate between sepsis and non-infectious SIRS. 

Investigators have reported microcirculatory alterations in critically ill patients using various techniques. Persistent microvascular alterations might be associated with the development of organ failure and death. In this study, microcirculatory blood transit time was measured in intensive care patients using micro-channel flow analyzers and related to the severity score and mortality.

Thirty-one patients were included in this study. Mean APACHE-II score was 16.8. Patients were divided into two groups, group L (APACHE-II<17, n=18) and group H (APACHE-II>=17, n=13). In both groups, blood transit time was measured using microchannel flow analyzers (MC FANs). The micro-machined silicon chip is utilized in these instruments to simulate human capillary blood flow. Microcirculatory alteration was presented as a blood transit time (second) of heparinized blood through micro-channel array under the pressure difference of 20 cmH2O. Hematocrit, white blood cell (WBC) count, platelet count, and labolatory data were obtained at the same time.

Blood transit time was significantly longer in group H comparing that in group L (105.2+/-20.3 sec, 57.6+/-9.5 sec, P<0.05). WBC count was larger in group H comparing that in group L (15800+/-2300 /ul, 10400+/-1300 /ul, P<0.05). Triglyceride (TG) and immunogloblin (IgG/M/A) levels were significantly higher in group H comparing these in group L. None of the group L patients died, however, hospital mortality rate was 53.8? in group H.

Blood transit time through micro-channel array was prolonged in patients with high APACHE score 2)WBC, TG, and immunogloblin levels might be associated with patients blood fluidity. 3) Micro-channel flow analysis may become a valuable tool to monitor microcirculation in critically ill patients. 

A. Roman* 1 , T. El Mahi 2 , C. Hanicq 1 , D. Gnat 2 , F. Vertongen 2 , E. Stevens 1 1 Intensive Care, 2 Clinical Chemistry, CHU Saint-Pierre, Brussels, Belgium

Bedside glucose monitoring is mandatory for ICU patients under tight glycemic control. Point-of-care (POC) glucometers are based on glucose-dehydrogenase coupled with pyrroloquinoline-quinone/ferricyanide (GD/PQQ)or phenanthroline-quinone/NAD (GD/PQNAD), or glucose-oxydase/ferricyanide (GO) enzymatic methods for whole blood measurements. The laboratory reference method is hexokinase for measuring the plasma glucose levels. Some drugs and metabolites can interfere with POC methods. The aim of this study was to evaluate the effect of the uric acid levels on the accuracy of these bedside methods.

In this prospective observational study, arterial blood glucose was measured simultaneously on the Accu-chek Inform Roche (GD/PQQ), on the Precision PCx Abbott (GD/PQNAD), on the RapidLab1265 Bayer (GO) and each value was compared with the reference laboratory result.363 measures were done in 75 adult ICU patients. Uric acid was obtained only once a day. A Bland-Altman analysis was done. Biases were expressed as the POC minus the laboratory result. Data were also analysed using linear regression. Spearmann's Rho squares were calculated to evaluate the uric acid level effect on the difference between POC and laboratory methods.

The uric acid level range was 1.2 to 15.8 mg/dL. The biases, the 95% limits of agreement between each POC method and the reference method, the R 2 of Spearmann for the correlation between uric acid level and the difference of result glucose level for each POC method are shown in table 1. The Accu-chek Inform overestimates moderately the glucose level while the Precision PCX and the Rapidlab underestimate it slightly. The Wilcoxon ranked test with Bonferroni correction gave a p < 0.001 for comparing the bias from the Accu-chek to the bias from the Precision PCx, p < 0.001 when compared to the bias obtained for the RapidLab. No statistical difference between the Precision PCx bias and Rapilab was found. The R 2 of Spearmann correlating the effect of the uric acid level and the difference between the Accu-chek and the reference method was 0.247. The weak effect of the uric acid level of the patient on the overestimation of the glucose measured by the Accu-chek can be summarized as : Glucose difference(Accu-chek-Laboratory) = 2.53 x Uric Acid (mg/dL) -2.3 . For the other POC glucometers, such correlations were absent. 

A patient presented with severe acidosis, point-of-care (POC) lactate of 42 mmol/L, suspicion of mesenteric ischemia and potential need for laparotomy. However, plasmalactates was <4mmol/L, and ethylene glycol (EG) ingestion was subsequently diagnosed. We, therefore, wished to determine why discrepant lactates occur and if this "lactate-gap" could be clinically useful.

We phlebotomized blood, added various concentrations of EG metabolites, and tested with the five most common lactate analyzers. 

The pressure-volume(P-V)curve of the respiratory system defines the mechanical properties of the lung and the chest wall by relating airway pressure(Paw)in no-flow conditions with lung volume at the same pressure level. Objective:To evaluate a new technique for P-V curve tracing.

Two P-V curves were obtained in 10 ALI/ARDS patients using the continuous positive airway pressure (CPAP) method and an automated system built into a commercial ventilator (P-V Tool 2, Galileo, Hamilton). For the CPAP method, ventilators were switched to CPAP and pressure was raised from 0 to 35 cmH2O in 5 cmH2O steps and then decreased while respiratory inductive plethysmography measured lung volume. For the automated method, we selected the automatic PV mode(Galileo, Hamilton)with flow 3L/m and maximum pressure of 35 cmH2O. Lung-volume and airway-pressure data were recorded. P-V pairs were fitted to a mathematical model. Lower (LIP) and upper (UIP) inflection points on the inspiratory limb and maximum curvature point on the deflation limb were obtained. Correlation between methods was calculated using bias and 95% agreement limits for LIPs and UIPs and the intraclass correlation coefficient (ICC) for absolute agreement for each pressure level.

No adverse events were observed. P-V curves were equivalent for each method, with ICC >0.75 for each pressure level. Bias and precision for LIP and UIP were:LIP 0.51±1.90cmH2O and UIP 0.53±3.04cmH2O.

The automated method for tracing P-V curves is equivalent to the CPAP method. Easily applicable at the bedside, it avoids ventilator disconnection and can obtain both inspiratory and deflation limbs of P-V curves.

INTRODUCTION. Hypoxic hepatitis (HH) is a common cause of acute hepatic impairment. However, few is known about the degree and duration of the reversal of the liver impairment. Therefore we assessed the liver function by indocyanine green (ICG) clearance via LiMON (PULSION Medical Systems, Munich, Germany) in patients with HH.

ICG clearance was assessed in 17 critically ill patients fulfilling the criteria of hypoxic hepatitis. Mean APACHE III score was 81 ± 31. Nine patients were male. ICU survival was 47%. ICG -plasma disappearance rate (PDR) (normal range: 18-25 %/min) and the retention rate of ICG extrapolated to 15 minutes (R15) were obtained on the day of development of HH and till day five. Nine patients with decompensated liver cirrhosis Child C requiring intensive care therapy served as control group.

RESULTS. ICG-PDR and R15 expressed as mean ± standard deviation were 5.79 ± 1.89 %/min and 42.12 ± 10.13 %, respectively (17 patients), on the day of development of HH. ICG-PDR and R15 were 6.07 ± 2.67 %/min and 42.81± 14.50 %, respectively, in the control group and was comparable to the HH group (p=NS). ICG-PDR and R 15 improved continuously from time of development of HH to day five (7 patients alive and at ICU) and were comparable to the course of laboratory data during observation period (Table 1) . 

Exhaled breath condensate (EBC) is a non-invasive means of collecting samples of airway lining fluid from the lower respiratory tract and monitoring respiratory diseases. We have used EBC acidification to study the effects of mechanical ventilation.

EBC was collected (30-40 minutes at -20 o C: EcoScreen, Jaeger). Immediately after collection and as soon as the sample returned to room temperature, we measured conductivity and pH before and after deareation with helium (10 minutes). Results are expressed as median (interquartil range). We have applied SPSSwin with Spearman correlation and Mann-Whitney test. 

Our earlier evaluations of a decision support system for tight glucose control (TGC) in the critically ill utilising model predictive control (MPC) documented clinically acceptable performance with hourly BG sampling. The MPC advises on insulin infusion based on blood glucose (BG) measurements and carbohydrate content of parenteral and enteral nutrition. In the present study, we evaluated an improved version of the MPC (v 1.04.01 to 1.04.03), which extends the advice by suggesting the time of the next BG measurement in the range from half-to four-hourly to reduce nurse workload.

Patients were admitted at one medical (MUG; N=21) and two surgical (KUL: N=13; CUP: N=19) ICUs. Patients were followed for a minimum of 48 hours and up to 72 hours. We evaluated safety of TGC (hypoglycaemia frequency), efficacy (mean BG; hyperglycaemic index, HGI; and time spent in the target range 4.4-6.1mM), and efficiency (time between BG measurements). Nonparametric statistical tests evaluated differences among ICUs.

One hypoglycaemia (BG < 2.2 mM) occurred in one subject at MUG and in another at CUP. There was no hypoglycaemia at KUL. BG was within the target range but differed among ICUs with values of 5.8 (5.5-6.2), 5.9 (5.6-6.2), and 6.4 (5.9-7.1) mM [median ( 

Strict glycemic control of plasma glucose has become general practice in most ICUs. Frequent glucose control is required to titrate the amount of insulin infused and detect episodes of hypoglycemia. For practical reasons bedside glucometry is often used. Aim of our study was to determine the accuracy of several glucose point-of-care (POCT) devices in critically ill ICU patients.

Arterial blood samples from unselected ICU patients were collected and glucose measurements were performed on a bloodgas analyzer (glucose-oxidase; RapidLab bloodgas analyzer, Bayer Diagnostics) and three different POCT devices (GDH-PQQ, Accu-Chek Sensor, Roche Diagnostics), GDH-NAD+ (Precision, Abbott Diagnostics) and modified GDH (Hemocue). Results of paired measurements were compared in three ways. Paired values were plotted on a Bland-Altman plot. The Pearson correlation coefficient (r) between the different methods was determined by linear regression. Each pair was also analysed using the International Organization for Standardization (ISO) criteria: -Glucose > 4,1 mmol/l value within 20% of reference -Glucose ≤ 4,1 mmol/l value within 0.8 mmol/l of reference.

Comparison between Accu-Chek and Rapidlab 860 of samples from unselected ICU patients (n=212) showed a good correlation (r2=0.9431). Bland-Altman analysis and analysis by ISO criteria revealed clinical significant differences in 13.2% of pairs. In all cases the POCT values were higher than the values from the bloodgas analyzer. Comparable results were found using the Precision and Hemocue: although correlation was high, analysis by ISO criteria showed differences in 9/81 (11.1%) and 3/81 (3.7%) of pairs. 

A clinically important inaccuracy was found between POCT devices and bloodgas glucose measurements in critically ill ICU patients. In the most cases values from POCT devices were false high, increasing the risk of hypoglycemia. In the context of an insulin infusion protocol for aggressive glucose control in sedated ICU patients POCT devices are potentially dangerous and should be avoided.

Acute hyperglycaemia associated with insulin resistance is common in critically ill patients. Acute tight control of blood glucose is considered important, although difficult to perform in routine care. We developed a software to implement tight glycaemic control (CGAO): after each glucose level measure, the CGAO advises a new insulin pump rate and the schedule for the next glucose control, gives indication for correcting any hypoglycaemia episode, and presents numerous parameters describing the quality of glycaemic control.

In a retrospective case control study, we compared the software CGAO (LK2, Igny, France) used routinely in our unit since may 2006 with our previous method for glycaemic control based on daily medical prescriptions. Patients without CGAO (group PRES) were randomly selected from our prospective intensive care database (admission after January 1, 2004) and matched 2:1 for sex, age, Simplified Acute Physiologic Score (SAPS II), medical or surgical category, history of type 2 diabetes, and length of stay (LOS) with patients for whom we used CGAO. Type 1 diabetic patients or patients with LOS < 3 days were excluded. Endpoints were average glucose level, hyperglycaemic index calculated above 6.1 mmoles/l, fractions of time (FT) resp. with normoglycaemia [4.2 -6.1 mmoles/l] and hyperglycaemia [> 6.1 mmoles/l], cumulative duration of hypoglycaemia [< 4,2 mmoles/l], average insuline requirements per day, and mean sampling interval for glucose control.

We included 150 patients (mean age: 64 ± 17 years, SAPS II: 42 ± 18, surgical: 26 %, type 2 diabetic: 12%), permitting to compare 50 CGAO patients with 100 PRES patients. 

A. Sigalas*, D. W. Patch, A. K. Burroughs, J. P. O'Beirne Liver Transplantation and Hepatobiliary Medicine, Royal Free Hospital, London, United Kingdom

Recently a number of studies have reported that relative adrenal insufficiency (RAI) is common in critically ill cirrhotics. Depending on the definition used the prevalence of RAI in critically ill cirrhotics has been reported to be 50-60%, whilst in patients immediately post liver transplantation the incidence of RAI has been reported to be 93%. Given the high prevalence of RAI in critically ill cirrhotics and patients undergoing liver transplantation, we hypothesised that adrenal function impairment may be a feature of chronic liver disease per se. The aim of this study was to define the prevalence of impaired adrenal function in patients with stable cirrhosis. We also examined whether the use of the 1µg or 250 µg ACTH tests was associated with different responses.

METHODS. 34 patients with biopsy proven cirrhosis (or compatible imaging and biochemistry) underwent adrenal function testing with the 1 µg (n=20) or 250 µg(n=14) short synacthen tests (SST). Patients were those with stable cirrhosis undergoing evaluation for transplantation or assessment for TIPS insertion for refractory ascites. Patients with a recent history of infection or bleeding were excluded.

. 34 patients underwent adrenal function testing. The median age of the group was 54 (IQR 48-60). The commonest cause of cirrhosis was alcohol in 40%. Disease severity was measured by MELD and Childs-Pugh scores. The median MELD was 16 (IQR 13.5-20.5) and the median Childs-Pugh score was 11 (IQR 8-13). 11 patients (32%) showed a baseline cortisol < 250 nmol/L and an increment < 250 nmol/L following SST. 19 patients (56%) had an increment in cortisol < 250 nmol/L following SST. 23 patients (67%) had a baseline cortisol < 250 nmol/L. Overall abnormalities in the SST (low baseline, peak or increment) were seen in 31 patients (90%). There were no significant differences in the frequency of abnormalities in the SST between the 1 µg or 250 µg SST groups. In multivariate analysis only MELD score significantly predicted abnormalities in the SST.

The above data suggest that adrenal dysfunction is a frequent finding in patients with stable cirrhosis and is correlated with liver disease severity. The underlying mechanism of this finding is unknown but may account for the very high frequency of RAI in critically ill cirrhotics. The direct relation between glucose and lactate levels in critically ill patients has hardly been studied. We studied the relation between glucose and lactate in general and during hypoglycemia.

Intensive insulin therapy was performed with the nurse-centered GRIP computer system that aimed at a glucose level of 6.5 mmol/L or less. Glucose and lactate were routinely measured together. All hypoglycemias detected over a 20-month period at the surgical ICU were analyzed. Hypoglycemia was divided in mild (3.4 thru 3.9), moderate (2.3 thru 3.3) and severe (<=2.2 mmol/L) hypoglycemia.

. 30,000 glucose/lactate measurements were analyzed in 727 patients. Glucose and lactate both were not normally distributed. After taking these distributions into account no evident relationship between simultaneous measurements of glucose and lactate was seen. 230 hypoglycemias were identified (141 mild;79 moderate; 8 severe). Lactate showed a with a nadir value two hours after the hypoglycemia. The magnitude of hypoglycemia was not related with lactate response. 

Evidence accumulates that improved glucose control in intensive care patients results in better outcome. Improved glucose control requires rapid point of care glucose measurement. However, the reliability of point of care glucose measurements has been questioned. This study was done to evaluate the accuracy of AccuCheck point of care glucose measurement in intensive patients as compared to glucose measurement by the central hospital laboratory.

The unit is a 10 bed mixed closed format ICU. Glucose regulation is performed by nurses for all patients using a computerised protocol(1). For this study, paired glucose measurements were randomly done in patients in the ICU, only when glucose measurement was clinically indicated and only if workload permitted the extra task. The AccuCheck Inform device (Roche Diagnostics) measures whole blood glucose in a single drop of blood. The central laboratory uses glycoseoxidase Vitros 950 to measure glucose in serum.

From 40 patients 246 paired measurements were obtained (table 1) . Central laboratory glucose measurement was generally higher than AccuCheck glucose measurement. The mean difference was 0,6 mmol/L. Correlation coefficient r was 0,935. The difference was more than 1,0 mmol in 24% of cases. Blood samples were mostly (98%) derived from arterial lines. The correlation and Bland Altman plots are presented in figure 1 . 

Related literature was examined for benchmarking purposes. Data collection was carried out over a one month period, two days a week, in the ICU. Each blood sugar level (BSL) was recorded and ensuing action chosen on adjusting the insulin infusion rate, and resultant information analysed. A survey was carried out on nursing staff regarding their views on the protocol. Statistical analysis was carried out using Microsoft Excel ® .

The BSLs were in the target range of 5.5-6.9mmol/L 39.4% of the time (n=863). The proportion of BSLs that complied with the Surviving Sepsis Guidelines target of less than 8.3mmol/L was good at 79.7%. The incidence of severe hypoglycaemia, defined as less than 2.2mmol/L, was low at 0.2%. Compliance with the action chosen on adjusting the insulin infusion rate was high at 76.3%. Total compliance (action and timing) with the protocol was 22%, and a relationship between compliance and achieving target BSLs was shown. In general, a positive view of the protocol was obtained from the nursing staff regarding the protocol.

The AMNCH ICU insulin infusion protocol is effective at achieving tight glycaemic control in a safe manner. The low incidence of severe hypoglycaemia and high proportion of BSLs complying with the Surviving Sepsis Guidelines illustrates this. Compliance with the protocol is achievable, demonstrated by the high level of compliance on action taken on the insulin infusion rate and the survey responses. However the timing of BSL checking needs to be addressed in future drafts of the protocol, as this is an area that needs improvement in terms of feasibility and compliance. Further changes and auditing of the protocol are necessary to ensure consistency and improvement of the tight glycaemic control.

INTRODUCTION. Intensive insulin therapy might be able to reduce mortality and/or morbidity in critical patients. Besides adherence to strict protocols this strategy implies multiple, accurate measurements of glycemia. Gold-standard laboratory assessment isn't able to provide immediate readings and capillary or arterial blood samples may differ too much when bedside reflectance meters are used, particularly in shock patients. Our aim was to assess the accuracy of two methods of blood glucose analysis (bedside "glucometer" using capillary and arterial blood) in two groups of critical ill patients (shock and non-shock).

Prospective non-randomized, cohort study, in a university hospital general ICU. A group of 18 consecutive ICU patients with shock syndrome and vasoactive amines and another 18 contemporary patients without shock, were included (shock-SG and non-shock-NSG groups). For each patient 1 to 4 "triplets" of blood samples were collected in a 48h period, and included concomitant samples of blood drawn from fingerstick (cap) and non-heparinized arterial line(art). Drops of capillary and arterial blood were analyzed with a bedside glucometer (GlucoTouch ® , Lifescan), and a sample of arterial plasma was sent to laboratory for glycemia determination (lab).

. Total group had a median age of 57 years, mean SAPS II of 53,4. SG was older (median age -64 vs 48 ys) and more ill (mean SAPS II 63,2 vs 43,7) than the NSG. Total mortality was 33,3% (SG-55,6%; NSG-11,1%). In the SG 61,1% had septic and 27,8% cardiogenic shock. In the NSG 44,4% had politrauma and 16,7% pneumonia. A total of 102 "triplets" were studied. Non Parametric Wilcoxon test was applied to test agreement between cap-lab and art-lab paired samples. Although we've found a highly significant correlation (Spearman r>0,8) between cap-lab and art-lab values, agreement were rejected by 2-tailed Wilcoxon signed ranks test, both in total, SG and NSG (p=0.000). An error grid-analysis using ISO 2003 for blood glucose determination showed that 19,1% of cap and 23,4% of art determinations had a deviation more than 20% the reference lab value in the SG. In the NSG 31% of cap and 38% of art samples had more than 20% deviation.

This study show that the glucometer we used had an unacceptable accuracy, both in shock and non-shock patients, far from the ISO criteria that imposes only 5% of values can be more than 20% apart the reference value. Glucose control is a major issue in the ICU and standard procedures for its determination are still lacking. INTRODUCTION. Arginine (ARG) is a precursor of the vasodilator nitric oxide (NO), while asymmetric dimethylarginine (ADMA), derived from proteolysis of methylated ARG residues, is a NO synthase inhibitor. Accumulation of ADMA is related to oxidative stress, impairing its degradation, and to renal-and liver failure. Accumulation is associated with increased mortality (1). Aim of this study was to evaluate the relation between plasma ARG, ADMA, ARG/ADMA ratio, organ failure and survival in patients with shock.

We measured plasma concentrations of ARG, ADMA and lactate, SOFA scores and hospital mortality in septic (SS) or cardiogenic shock (CS) patients on d1, d2 and d5 of ICU admission. Patients were enterally fed with Impact (ARG-enriched). Values are presented in mean ± SD or median (IQR). For regression analysis, ARG, ADMA and ARG/ADMA were log transformed.

Of the 44 patients, 27 had SS, 17 CS. Mean age was 66 ± 14 yrs, SOFA 9 ± 3, APACHE II 26 ± 7.4. Hospital mortality was 36%, predicted mortality was 56 ± 25%. At d1, median (IQR) of ARG was 30 (22-48) mumol/l (normal range 30-145 mumol/l), ADMA 0.42 (0.33-0.55) mumol/l, ARG/ADMA 80 (49-116) and lactate 4.6 ± 3.2 mmol/l. ARG and ARG/ADMA at d1 were inversely related to lactate (R = 0.59, p < 0.001, for ARG; R = 0.65, p < 0.001 for ARG /ADMA), and to SOFA scores. The table presents the relation between ARG and ARG/ADMA to SOFA score during sampling, and of ARG and ARG/ADMA on day 1 to maximum SOFA score. 

Apneic oxygenation (AO) is apllied during several operations in thoracic surgery and some procedures in th ICU. Retention of CO2 often leads to hypoxemia, limiting the tolerable time in AO. This experimental study was designed to evaluate the effects of recruitment maneuver on oxygenation, CO2 retention and survival times AO.

Following the Ethic Committee approval, 15 male Sprague-Dawley rats were anesthetized, tracheostomized, cannulated via the a. carotis and ventilated with pressure controlled ventilation (peak pressure: 15 cmH2O, frequency: 25/min, 0 cm H2O PEEP) for 15 minutes. Following the basal (T0) arterial blood gas sample, they were randomized into 3 groups and disconnected from the ventilator: In Group 1 (n=6), rats underwent AO with a cannula inserted to carina (O2-flow:0.2 L/min), in Group 2 (n=6), recruitment maneuver (40 cm H2O (PEEP) ventilation pressure during 20 seconds) was performed before AO. In control group (Group 3, n=3), data were recorded after apnea (this group was stopped after the first 3 subjects have died during the study period). Further arterial blood gas samples were drawn in 1st, 3rd and 6th minutes, and pH, pO2, pCO2, HCO3 and BE values were recorded. Survival times after the initiation of AO were also investigated. Kruskal-Wallis test was used to compare the values in different times, and Mann-Whitney-u the values in different groups.

There were no significant difference in T0 values. Compared to T0 values, there was a significant decrease in pO2 and a significant increase in pCO2 during 3rd and 6th minutes in all subjects, with a less change in G2. There was a significant difference between G1 and G2 in pO2 after 3 and 6 minutes p<0.05; Table 1 ), the difference in pCO2 was not significant. Survival time in G2 was significantly longer (G1:10,3±2,3 min; G2:14,3±3,6 min; p<0.05). 

To investigate potential prognostic factors and to predict extent of risks for postoperative pulmonary complications by logistic regressive analysis, and to evaluate the role of non-invasive ventilation in reducing the incidence of complications in elderly patients.

Stair-climbing test was carried out with ASA score, FEV1, changes of SpO2 and HR et al were noted at the same time. Logistical regressive analysis based on the parameters above were used to assess the relation between potential prognostic factors and postoperative complications. Patients with limited pulmonary reserves were selected using the equation, and protective effect of non-invasive ventilation on these patients was assessed.

Incidence of postoperative pulmonary complications for high-risk patients with non-invasive ventilation was 33.2%, and incidence of pulmonary complications for high-risk patients without non-invasive ventilation was 67.7%. There was not a significant difference between these two groups with low-risk (P>0.05).

CONCLUSION. The mathematical model of logistic regressive analysis using stair-climbing testing combined with other parameters is a simple, reliable method to predict the cardiopulmonary reserved function in elderly patients. Non-invasive ventilation can effectively reduce the incidence of postoperative pulmonary complications for high-risk patients, but it has no effect on patients with low-risk. 

Continuous epidural analgesia (EA) and intravenous analgesia (IA) are widely used for postoperative thoracic pain control. The aim of this study is to compare the advantages and the disadvantages of both analgesic techniques. Ropivacaine 0.2% to 20 mg/h using thoracic epidural catheters (EA) vs intravenous analgesia with Remifentanyl 0.055 µgr/Kg/min (IA).

One hundred patients, undergoing pulmonary surgery, were recruited and divided, after randomization into 2 groups. Patients included in EA group had an epidural thoracic catheter placed at Th4 -Th6 space, received Ropivacaine 0.2% by continuous infusion (rate 10 ml/h). Patients included in IA group received an ev continuous infusion of Remifentanyl (rate 0.055 µgr/Kg/min for 24 hours). Rescue medication consisted of morphine 2 mg ev at patients demand. Analgesia at rest and while coughing as evaluated by visual analogue scale (VAS). Haemodynamics, motor blockade (Bromage scale) and side effects such as nausea, vomiting and pruritus were observed. The follow-up took place after weaning and every hour to 24 hours at rest and coughing. Data are reported to media ± standard deviation (SD). Analgesic effects were compared by using Chi square statistics (p<0.05). Both groups showed good analgesic effects. Remifentanyl seems to decrease the incidence of side effects and the need of rescue analgesia.

CONCLUSION. 1)Our data show that both analgesic techniques are able to guarantee a good pain relief after thoracotomy. 2)Epidural analgesia was more difficult to perform and it showed less acceptance by patients. 

Non-invasive ventilation (NIV) has become an effective treatment to reduce morbidity and mortality in patients with acute respiratory failure. Its application has been restricted to critical care o intermediate care areas, and little data is available on its usefulness in the post-anaesthesia care units (PACU). The aim of this study is to document our experience after eight patients treated in the PACU.

We undertook a retrospective audit of patients treated with NIV between 1 October and 31 December 2006. Data of past medical history, age, ASA physical status, surgical procedure, anaesthesia modality, type of respiratory failure, ventilatory mode, and time of NIV were recorded. We also recorded side effects related to NIV application. Descriptive statistical analysis was used.

Eight patients were included. The mean age was 59.75±19.44 (SD) years. Five patients were classified as ASA 3 (62.5%), two as ASA 4 (12.5%), and one as ASA 1 (12.5%). Three patients had morbid obesity, two chronic heart failure, and two chronic obstructive pulmonary disease. General and regional anaesthesia were employed in 6 and 2 cases respectively. Type of surgery was thoracic (25%), urologic (25%), and plastic (25%). There was one case of abdominal surgery and another one of oral surgery. Hypoxemic failure was detected in three patients (37.5%), and CPAP was applied in these cases. BIPAP was applied in cases of hypercapnic (12.5%) or global (50%) respiratory failure. The mean time of NIV was 89.37±26.78 (SD) minutes. No complications related to NIV occurred. No patient required either intubation or transfer to the ICU. All of them were transferred to the surgical wards the same day.

CONCLUSION. NIV can be safely applied to selected patients in the PACU, to treat respiratory failure after either general or regional anaesthesia. It is an effective method to avoid intubation and ICU stays, with minimal side effects. Further studies should be conducted to analyze the clinical and economic impact of NIV in the PACU. 

The routine use of volatile anesthetics in intensive care medicine has been limited so far due to technical difficulties and the need for an anaesthetic machine. The new Anesthetic Conserving Device (AnaConDa)can provide a safe application of isoflurane or sevoflurane under intensive care conditions. This system is a modified heat and moisture exchanger which includes activated carbon fibres and works as a miniaturized vapor with recirculation. We studied the effectiveness of sevoflurane sedation in operative intensive care patients undergoing mechanical ventilation.

We included 40 ventilated patients (neurosurgery, septic patients) in our retrospective analysis. The Anaesthetic Conserving Device (AnaConDa-System) replaces the common heat and moisture exchanger in the ventilator circuit. The volatile anaesthetic is continuously applied in liquid status via a syringe pump to the minivapor where the anesthetic is vaporized. The expired anaesthetic gas is stored in the carbon filter and about 90% are resupplied into the breathing circle.

First experiences with sevoflurane at our institution with a mean application time of 90.1 ± 56.3 hours per patient, showed a mean dose of 5.5 ± 1.3 ml sevofluran to achieve the individually targeted sedation level. 13.6 ± 7.1 minutes after the end of sevoflurane application, the patients could be neurologically evaluated or transferred to spontaneous breathing or extubated. No relevant side effects like nausea, vomiting or elevated enzymes were observed.

We could demonstrate a safe application route, no development of tolerance as well as short wake-up times after long-term sedation with sevoflurane. The current literature suggest that volatile anaesthetics present an alternative for long-term sedation on intensive care units, providing optimized pathways from a medical as well as from an economical viewpoint. 

Safety and effectiveness of sedation and analgesia in permanent pacemaker implant (PPM) is of special concern, due to age and comorbidity of the implanted patients. Remifentanil pharmacological properties appear to be of interest in this setting. To date, there are no reports describing the use of remifentanil in this procedure, without the use of mechanical ventilation.

Consecutive patients in whom a PPM or other procedures, such as pacemaker battery change, was scheduled were included. A sedation and analgesia protocol for PPM implantation was performed: metoclopramide premedication, remifentanil infusion (20 mg/ml), local anaesthesia with mepivacain 2%, magnesic metimazol administration at the end of procedure, and remifentanil infusion withdrawal 20 minutes later. Remifentanil infusion was initiated at a rate of 2 mcg/min, increasing the rate to attain a sedation Ramsay scale grade 2 or 3, to a maximum of 6 mcg/min. Remifentanil failure was defined as the need to administer a different sedation after the maximum dosage was attained. Adverse effects, lenght of infusion and dosage were recorded.

.Two hundred and thirty-six consecutive patients were included. The men age was 77,5 ± 13,1. Procedures: bicameral pacemaker 23,7%, unicameral 64,3%, battery change 5,6%, other 7,3%. Infusion description and adverse effects are showed in tables 1 and 2. Serious adverse effects were resolved with remifentanil infusion withdrawal. All the procedures were completed. 

Remifentanil is safe and effective as sedation and analgesia for PPM implantation, even for old patients, with the dosages used in our protocol. Nausea is the most frequent adverse effect. Serious adverse effects are uncommon and can be resolved with infusion withdrawal.

Glass PSA, Gan TJ, Howell S. A review of the pharmacokinetics and pharmacodynamics of remifentanilo. Anesth Analg 1999; 89: S7-S14. 

Peripheral arterial occlusive disease (PAOD) can cause intense neuropathic/ischemic limb pain in patients (pts) with end stage renal disease (ESRD). Although fentanyl may be an excellent choice in ESRD due to the absence of active metabolites, the use of fentanyl as PCA in ESRD has never been reported. We used IV Fentanyl PCA for ischemic lower extremity pain in 11 ESRD patients (9M, 2F), 10 of whom were scheduled for amputation.

Pts received IV Fentanyl PCA via a Gemstar (Abbott) pump. Initial settings were 25mcg bolus, 20 min lockout, no basal, and dose was adjusted as needed to achieve Visual Analogue Scale (VAS) score < 4. PCA started 48 hours preamputation and continued postoperatively for 48 h in 6 pts (4 pts had epidural postoperative analgesia and one terminal cancer pt did not have surgery). Pain was assessed twice daily with VAS. The McGill Pain Questionnaire (MPQ) -total ranked rating index (PRI(R)), was administered immediately before and 24 h after PCA started. Sedation was assessed twice daily on a four-point scale: 1) agitated, 2) awake, 3) roused by voice and 4) unarousable. Pain scores were compared with paired t-test. Group data are presented as Mean ± SD. Mean sedation score was 2 in men and 3 in women. We did not observe respiratory depression in any patient. 

The aim of this study was to determine risk factors for relapse, and for ICU-mortality in patients with ventilator-associated pneumonia (VAP) related to nonfermenting Gram negative bacilli (NF-GNB).

Retrospective case-control study based on prospectively collected data. VAP diagnosis was based on clinical, radiographic and microbiologic (endotracheal aspirate ≥ 10 6 cfu/ml) criteria. Patients with monobacterial VAP related to NF-GNB were eligible. Patients with subsequent superinfection or persistent pulmonary infection were excluded. Patients with relapse of NF-GNB VAP were matched (1:2) with patients without relapse according to duration of mechanical ventilation before VAP occurrence. Univariate and multivariate analyses were used to determine risk factors for relapse, and for ICU-mortality in cases and controls.

. 276 patients were eligible. 26 patients were excluded for superinfection. No persistant infection was diagnosed. 30 (11%) patients developed a relapse of NF-GNB VAP, and were all successfully matched with 60 controls. Pseudomonas aeruginosa was the most frequently isolated bacteria (55%), followed by Acinteobacter baumannii (34%) and Stenotrophomonas maltophilia (10%). No significant difference was found between cases and controls with regard to age (63±17 vs 64±14), male gender (90% vs 72%, p = 0.061), and surgery (13% vs 27%). However, SAPS II at ICU admission (40±10 vs 48±14, p = 0.015) was significantly lower in cases than in controls. Duration of adequate antibiotic treatment for first VAP episode was significantly shorter in cases than in controls (12 ± 3 vs 14 ± 4d, p = 0.031). Inadequate initial antibiotic treatment was the only variable independently associated with relapse of VAP related to NF-GNB (OR [95% CI] = 8.1 

Inadequate initial antibiotic treatment is independently associated with relapse of VAP related to NF-GNB and with ICU-mortality. ∆ radiologic score and SAPS II at day 7 after VAP diagnosis are independent risk factors for ICU-mortality in these patients.

S. Blot* 1 , J. Solé-Violán 2 , J. Blanquer 3 , J. Almirall 4 , A. Rodriguez 5 , J. Rello 5 1 ICU, Ghent Univ Hosp, Ghent, Belgium, 2 ICU, Dr Negrin Hosp, Gran Canaria, 3 Respiratory Care, Clinic Hosp, Valencia, 4 ICU, Mataró Hosp, Barcelona, 5 ICU, Joan XXIII Univ Hosp, Tarragona, Spain

Practice guidelines suggest processes of care such as timely pulse oximetry monitoring and antibiotic therapy, as quality indicators for the management of communityacquired pneumonia (CAP). The objective of this study was to determine whether postponed initial processes of care such as pulse oximetry monitoring delays initiation of antibiotic therapy and adversely affects intensive care unit (ICU) survival in patients with severe CAP.

A prospective observational multicenter study was conducted including 529 patients with CAP admitted to the ICU in 33 hospitals. A secondary analysis was conducted to evaluate processes of care and ICU survival.

Postponed blood culture sampling, arterial blood gas sampling and pulse oximetry monitoring was predictive for delayed antibiotic administration (P<0.001). Linear regression analysis demonstrated that a delay of >1h in blood culture sampling was associated with a delay of 4.7h (95% confidence interval [CI], 2.5-7.0) in antibiotic therapy, a delay of >1h in blood gas sampling with a delay of 5.2h (95% CI, 2.9-7.6), and a delay in pulse oximetry monitoring of >1h with a delay of 4.2h (95% CI, 1.7-6.7). A delay in antibiotic administration of >6h was associated with increased mortality in univariate analysis (relative risk [RR], 1.85; 95% CI, 1.15-2.97), but not after adjustment for disease severity. A delay in pulse oximetry monitoring of >4h was associated with increased mortality in univariate analysis (RR, 1.97; 95% CI, 1.11-3.51) and after adjustment for disease severity (hazard ratio, 1.95; 95% CI, 1.22-3.11).

In patients with severe CAP timely executed processes of care are associated with a short time to antibiotic administration and reduced risk of death. 

Appropriateness of antibiotic therapy is associated with reduction of bacterial load. C-reactive protein (CRP) is a valid biochemical surrogate. Our objective was to determine the correlation of bacterial load, measured by quantitative tracheal aspirate (QTA), with CRP as an indicator of inflammatory response in episodes of lower respiratory tract infection. To evaluate whether appropriateness of antibiotic treatment influences microbiologic (QTA), biochemical (CRP) and clinical resolution criteria (temperature, WBC, SOFA and PO2/FiO2 fraction).

Prospective cohort study. Sixty-five intubated patients with monomicrobial lower respiratory tract infection were included. CRP and bacterial load variation were evaluated through the ratio between D4 and D0 measures.

A QTA was performed on lower respiratory tract onset (D0) and 96h afterwards (D4). Its logarithm value (logQTA) was recorded. LogQTA correlated positively with CRP, temperature and WBC. LogQTA has decreased significantly more from D0 to D4 in patients receiving appropriate empirical antibiotic therapy compared to those with inappropriate treatment (LogQTA ratio 0.77 vs 1.04, p<0.05). Mean CRP levels showed a similar pattern, decreasing from D0 to D4 in patients receiving appropriate empirical antibiotic treatment, but not in episodes with inappropriate treatment (CRP ratio D4/D0 0.54 vs 1.36, p<0.05). ANCOVA showed that CRP level on D4 was significant lower in patients with appropriate antibiotic treatment compared to inappropriate empiric treatment (103±10 mg/L vs 192±14 mg/L, p<0.05). The best cut-off to predict appropriateness of antibiotic therapy is a CRP levels reduction of 80% on D4(AUC=0.87).

CONCLUSION. C-reactive protein correlates with bacterial load and is a valid biochemical surrogate of bacterial burden in lower respiratory tract infection. Follow-up measurements of CRP anticipate the appropriateness of antibiotic therapy. 

A. Günther* 1 , P. Schenk 2 , M. Maggiorini 3 , A. Betbesé 4 , P. F. Laterre 5 , N. Fedorovskiy 6 , F. J. H. Taut 7 , R. G. Spragg 8 1 University of Giessen, Lung Center, Giessen, Germany, 2 , Medical University Vienna, Vienna, Austria, 3 , Universitätsspital Zürich, Zurich, Switzerland, 4 , Hospital Sta Cruz y San Pablo, Barcelona, Spain, 5 , Hôpital Saint Luc, Brussels, Belgium, 6 , City Clinical Hospital N67, Moscow, Russian Federation, 7 ALTANA Pharma AG, a member of the Nycomed Group, Konstanz, Germany, 8 , UC San Diego, San Diego, United States

The formal diagnosis of ARDS requires the acute onset of a severe impairment in oxygenatio(PaO2/FiO2 <= 200 mm Hg), exclusion of a hydrostatic cause, and the presence of diffuse bilateral opacities. Pneumonia is one of the most common underlying reasons for development of ARDS, but when only unilateral opacities are present, these patients fail to fulfil ARDS criteria. It is currently not known whether fulfilment of the formal ARDS criteria has any impact on 28-day mortality in patients with pneumonia suffering from severe gas exchange abnormalities.

The VALID study, a randomised, double-blind study in intubated and mechanically ventilated patients with severe respiratory failure (PaO2/FiO2 <= 170 mm Hg) due to pneumonia or aspiration of gastric contents investigates the effect of rSP-C surfactant (Venticute ® ) on mortality. The study does not require a formal diagnosis of ARDS for patient enrolment. However, the presence or absence of ARDS is documented. We conducted univariate and multivariate logistic regression analyses using preliminary blinded data from the first 443 patients randomised with a diagnosis of pneumonia. The prognostic value of the formal diagnosis of ARDS was determined.

Univariate logistic regression analysis failed to identify a significant correlation (p=0.13) between the formal diagnosis of ARDS and mortality at day 28. PaO2/FiO2 was more likely to be associated with mortality (p=0.01) as was the number of quadrants on chest radiograph that showed opacities (p=0.02). Age and APACHE II score were highly associated with mortality (p<0.0001). Multivariate logistic regression identified age (p<0.0001), the number of involved quadrants (p=0.002), and APACHE II (p=0.06) as independent factors affecting 28-day mortality.

CONCLUSION. The prognosis of ventilated patients with pneumonia is not dependent on the formal diagnosis of ARDS. Instead, age, APACHE II score, and the number of lung quadrants with radiographic opacities are more predictive of outcome. Bernard GR et al. Intensive Care Med. 1994; 20:225-232 . 

To determinate the clinical-epidemiological characteristics and risk factors for postsurgical pneumonia (PSP) after lung cancer resection in a university hospital. A retrospective case-control paired study (1:2) was performed in 604 cases of lung cancer collected from 1999 to 2004. Definition of PSP case was a new or changing radiographic infiltrates with two or more of the following criteria: fever > 38 o C, WBC>10000 mm3 or/and purulent secretions. Control group was formed by patients matched by age and lung cancer stage.

. 81 patients were evaluated (22 PSP and 62 controls). Overall, data of both groups were: Age 64 ± 10 yr, males 79 (94%), smoking habit (active or past smokers) 81 patients (96%), COPD 64 patients (76%) and weight loss over 10 kg in 5 patients (6%). Incidence of PSP was 4%, crude mortality rate and attributable mortality estimated for PSP was 32% and 24%, respectively. In the PSP group, we found the following isolates (64%): P. aeruginosa (18%), S. viridans (14%), H. influenzae (14%) S. pneumoniae (9%) and undeterminated (36%). PSP was associated with low BMI (p=0.007), low FEV1 (p=0.012), stage IIIA (p=0.004), anaesthetic time (p=0.047), pneumonectomy (p=0.017), thoracic pain (p=0.046), reintubation (p=0.001) and haemorrhage (p=0.018).

CONCLUSION. The incidence of PSP in our series is low but with a high mortality. Identification of risk factors (some of them suitable for medical intervention) may improve the management of lung cancer patients treated with surgery. 

J. Karhu* 1 , H. Syrjälä 2 , P. Ylipalosaari 2 , J. Laurila 1 , P. Ohtonen 3 , T. I. Ala-Kokko 1 1 Anesthesiology, Division of Intensive Care, 2 Infection Control, 3 Surgery, Oulu University Hospital, Oulu, Finland INTRODUCTION. SCAP (severe community acquired pneumonia) and HAP (hospital acquired pneumonia) requiring ICU treatment have been shown to be associated with significantly higher mortality compared to those not requiring ICU treatment (1, 2). We compared pneumonias acquired outside the ICU to that acquired in the ICU, during mechanical ventilation (ventilator-associated pneumonia, VAP).

Patients admitted into a mixed university level ICU during a 14 month period whose ICU stay was longer than 48 hours were included. The occurrence of SCAP, HAP and VAP were prospectively assessed. The following information was collected: age, severity of underlying disease on admission, underlying malignancy and recent use of immunosuppressive therapy. The length of ICU and hospital stay as well ICU, hospital and 28 day mortalities were recorded.

A total of 335 patients fulfilled the inclusion criteria during the study period. There were a total of 156 pneumonias. Majority of the pneumonias were SCAP (86/156), while there were 44 HAP and 26 VAP cases. Patients with HAP tended to be older (61.7, P=0.07) and a larger proportion of them had malignancy (34%, P< 0.001), compared to VAP (56 years, 27 %) or SCAP (24 years, 7 %). There were no significant differences between the mean admission APACHE II scores (SCAP 23.6 vs. HAP 22.6 vs. VAP 22.2) . The ICU length of stay was longest in VAP; while the hospital stay was longest in patients with HAP (Table 1 ). The survival rates were highest in HAP, although this did not reach statistical significance. In APACHE II and age adjusted multivariate logistic regression analysis VAP (OR 3.8, 95% CI 1.39-10.47, P=0.009) and SCAP (OR 3.3, 95% CI 1.71-6.41, P<0.001) remained significant risk factors for hospital mortality together with immunosuppression (OR 2.3, 95% CI 1.25-4.42, P=0.008). 

Heart surgery in infants is often associated with pulmonary inflammatory process. At the same time, the blood level of pro-inflammatory factors: Interleukin-6 (IL-6) and

Interleukin-8 (IL-8) is increased. The number of polymorphonuclear leukocytes (PMN-elastase) and neutrophils is raised as well. A qualitative evaluation of the factors, cellular composition analysis of nonbronchoscopic trachebronchial lavage (NTL) combined with clinical findings can help early diagnose pneumonia. The objective of the study was to reveal the peripheral blood level of pro-inflammatory cytokines (Il-6, Il-8), the activity of PMN-elastase and α1antiprotease inhibitor (α1-PI), as well as examine the NTL cellular composition and cytokine level in infants before and after heart surgery.

We studied 24 infants aged from 4 days to 11 months, weighting between 2.3 and 11 kg. 15 patients underwent cardiopulmonary bypass surgery, 9 patients were operated on without cardiopulmonary bypass. In 7 cases a clinical diagnosis of pneumonia was made between 2 and 5 days postoperatively. Early postoperative survival was 100%.

The peripheral blood cytokine concentration in operated infants pre-and postoperatively is presented in the study (Table 1) . A significant increase in pro-inflammatory factors after surgery can be observed. We examined the NTL of 9 infants who underwent heart surgery and who did not develop pneumonia. We noticed that the number of neutrophils increased significantly in all patients after cardiopulmonary bypass surgery, sometimes reaching 80%. We consider it as a sign of pulmonary inflammatory process. The number of nonviable alveolar macrophages before and after surgery exceeded 50%. It indicates a decrease in cellular pulmonary protection. The PMN-elastase peripheral blood activity was 273.2±75.00 IU/ml preoperatively and 339± 55.00 IU/ml postoperatively; the α1-PI level was 32.1±7.0 IU/ml and 32.5±15.00 IU/ml, respectively. CONCLUSION. Thus, an increase in the peripheral blood level of pro-inflammatory cytokines was observed in infants who underwent heart surgery. At the same time, the NTL relative number of neutrophils was increased. An early detection of the mentioned factors appears to be a diagnostic marker of the pulmonary inflammation reaction onset. 

All colistin resistant gram-negative isolates from patients hospitalized in a 7-bed ICU during one-year period were retrospectively recorded. Demographic data, the underlying disease, prior antimicrobial therapy, microbiological data and the clinical and bacteriological response to treatment were recorded. The antimicrobial susceptibility of the isolates was determined using the disk-diffusion (Kirby-Bauer) method, the VITEK II system and the Etest method (AB Biodisk, Solna-Sweden). Interpretation of the susceptibility results was in accordance to the Clinical and Laboratory Standards Institute (CLSI).

Nine patients with infections caused by colistin resistant gram-negative isolates were recorded. All patients had prolonged ICU stay, were under mechanical ventilation and had a significant exposure to antibiotics including colistin for MDR gram-negative bacteria. Three K.pneumonia isolates producing metallo-beta-lactamases (MBL), two K. pneumonia isolates producing extended spectrum b-lactamases (ESBL) and MBL, two Acinetobacter baumannii isolates susceptible to tetracyclines, one pandrug resistant (PDR) Acinetobacter baumannii and one PDR Pseudomonas aeruginosa were recorded. The bacteria were isolated from bronchial secretions in four cases and from the blood stream in five patients. In five patients antibiotic treatment was based on susceptibility tests, with clinical and bacteriological success. Antibiotic combinations including colistin plus meropenem or colistin plus cefepime were provided in patients harbouring PDR isolates. These patients failed to respond to treatment and had a fatal outcome. The overall clinical success and survival rate was 55.5% at 14 days.

CONCLUSION. The development of colistin resistant strains with increasing mortality rates urges for the continuous surveillance on these highly resistant organisms and the strict implementation of infection control practices. 

Ventilator-associated pneumonia (VAP) is one of the most severe infections in the ICU, continuing to complicate a high percentage of the patients receiving mechanical ventilation and leading to increased morbidity and mortality, especially when it is due to highrisk pathogens. Our aim was to study the incidence and outcome of VAP due to MDR bacteria in our ICU.

Prospective, epidemiological study, in a mixed ICU of a tertiary care hospital. All patients admitted from August 2004 to March 2007 were included. Lower respiratory tract samples of all patients with suspicion of VAP were cultured. Standard diagnostic criteria were followed. Statistical analysis was performed with SPSS v.12.

During the 32 months period of the study 330 patients were admitted. Their mean age was 66 years and 61% of them were male. Their mean APACHE score was 18 and the average duration of stay in the ICU was 18 days. Forty-two episodes of VAP due to MDR bacteria were recorded in 39 patients. The bacteria isolated from lower respiratory tract samples were 21 Acinetobacter baumanii, 16 Pseudomonas aeruginosa, 4 Klebsiella pneumoniae and 1 Enterobacter cloacae, while in 13 cases concomitant bacteremia was recorded. The mean time from admission to the ICU to diagnosis of VAP was 30 days. Positive outcome was noted in 62% of cases and was found to be reversely related to the APACHE II score (p=0.02), to days of stay in the ICU (p=0.011) and to multi-organ failure (p=0.002). Of the 39 patients with VAP, 25 had normal renal function before the lung infection. Of these, 15 developed renal failure due to the lung infection and had to be started on renal replacement therapy. The mortality of these patients was significantly higher than for the patients who did not develop renal failure (p=0.02). Regarding the crude mortality of patients with and without VAP, this was found to be 53.8% and 32.5% respectively (p=0.012). (Pa) is not a frequent pathogen in this setting but could be associated with poor prognosis. In our population of patients undergoing CS, we compared risk factors and prognosis of Pa-EOP with EOP due to others micro-organisms.

This retrospective study performed on 2 years (2005-6) involved 1504 patients (pts) who underwent CS with cardiopulmonary by-pass. Diagnostic of pneumonia was based on clinical and laboratory criteria: T˚>38.4, purulent tracheal secretions, WBC>11,000/mm3, chest X-ray changes and microbiological criteria (Broncho-alveolar lavage>104 cfu/ml). Pre, per and postoperative risk factors, empiric antibiotic, and prognosis of Pa-EOP were compared with those obtained for EOP due to others germs. The 2 groups were compared using chi-square. P<0.05 was considered significant.

Over the studied period, EOP occurred in 62 pts (incidence 4 %), including 14 pts ( CONCLUSION. In our experience, Pa-EOP following CS seems to be more frequent than what was previously reported. Criteria for prediction of Pa-EOP remain to be assessed. In case of Pa EOP, empiric antibiotic is often inappropriate with a possible increased risk of mortality. These results lead us to modify our empiric broad-spectrum antibiotic treatment and to take into account Pa, especially in severe forms of EOP and in COPD pts. 

Antibiotic exposure and timing of pneumonia onset influence ventilatorassociated pneumonia (VAP) isolates. The first goal of this investigation was to evaluate whether trauma also influences prevalence of microorganisms.

A retrospective, single-center, observational cohort study.

. VAP isolates in a multidisciplinary ICU documented by quantitative respiratory cultures and recorded in a 42-month database were compared, based on the presence (T) or absence of trauma (AT). Causative microorganisms were classified in four groups, based on mechanical ventilation duration (>5 days), and previous antibiotic exposure. One hundred eighty-three patients developed 196 episodes of VAP (98 trauma). Methicillin-sensitive Staphylococcus aureus (MSSA) was more frequent (34.5% vs 11.5%, p<0.01) in trauma, whereas MRSA was more frequent (2% vs 11.5%, p<0.01) in nontrauma. No significant differences were found between trauma and nontrauma patients regarding prevalence of other microorganisms. In trauma patients, MSSA episodes were equally distributed between early and late-onset VAP(51% vs 49%) but no MRSA episode ocurred in the early-onset group.

CONCLUSION. Trauma influences the microbiology of pneumonia and it should be considered in the initial antibiotic regimen choice. Our data demonstrate that patients with trauma had a higher prevalence of MSSA, but the overall prevalence was sufficiently high to warrant an S. aureus coverage for both groups. On the other hand, since no MRSA was isolated during the first 10 days of mechanical ventilation on trauma patients, MRSA coverage in these patients is only necessary after ten days of admission. 

A retrospective study of a HIV patient's cohort that stays in ICU with acquired community pneumonia in the period between January 2002 and December 2006. Data analyzed included age, clinic stage, years of disease evolution, antiretroviral therapy, CD4 levels and viral charge at the hospitalization, positive HCV and/or HBV, severity scores and microorganism isolated. Chi-square analysis was used to compare categorical data. Continuous data was compared using Student's t-test. Prognostic factors of mortality were studied by multivariate logistic regression analysis.

. fifty-three patients were studied. 66% were males. The average age was 40±9 years. The most frequently risk practice was intravenous drug addiction (68% 

We prospectively collected data regarding demographics and microbiology of bacteremias. Blood cultures were obtained on clinical suspicion of bacteremia and followed up on days 3, 7, and 14th. Severity of illness scores, APACHE and SOFA were recorded at baseline and days 3, 7, and 14th. Improving hand hygiene is a cost-effective way of decreasing hospital-acquired infection rates.

In this study we recorded opportunities for and compliance to hand hygiene in our ICU.

Four trained nurses and a doctor monitored opportunities for hand hygiene performance (hand antisepsis and glove use) as well as compliance to the CDC guidelines in our ICU for 10 days. The procedure was anonymous, involved all ICU personnel and was performed in 15-min sessions, throughout all shifts.

We collected 829 opportunities for hand hygiene, mostly related to nurses (62%). Compliance to hand antisepsis was 45%, higher in nursing and assistant staff (48% and 53%, respectively) compared to doctors (34%). Compliance was lowest before contact of healthcare staff with a patient or his inanimate environment (23% and 16%, respectively). The activity index (=the need for hand antisepsis performance) for the nursing staff was high (12 opportunities per hour per nurse in the morning shift, ie 96 opportunities per shift). However, no significant correlation was found between compliance rate and activity index of the staff (r=-0.17, p=0.21). Alcohol-based hand-rub was used in 62% of the cases. Technique of antisepsis performance was uniformly poor and mean duration of the procedure was low (6.3 seconds). Compliance with glove use guidelines was 91% and was high in all staff categories and all types of opportunities. is an aerobic non-fermenting gram negative bacillus. It is generally considered an opportunistic pathogen. S. maltophilia is increasingly recognised as a cause of nosocomial infection among ventilated and immunocompromised patients, and in those receiving broad spectrum antibiotics. S. maltophilia infections are commonly resistant to multiple antibiotics including beta lactams, quinolones, aminoglycosides and carbapenems. Reported mortality rates for patients with bacteraemia due to S. maltophilia vary from 10-60%. The Mid Western Regional Hospital, Limerick, Ireland, is a 500 bed hospital located on three sites. The intensive care unit(ICU) is a seven bed medical and surgical unit with approximately 450 admissions per year. The S. maltophilia clusters prompted epidemiological investigation, restriction fragment-length polymorphism typing (RFLP) of genomic DNA of outbreak strains, and finally, instituting revised infection control measures to limit spread.

We conducted a retrospective chart review of affected patients noting admission APACHE II scores, medical co-morbidity, immunocompetence, antibiotic history, and patient outcome. We collected cultures of ICU cubicle/ room surfaces, sinks, ventilatory equipment, and water sources. Patients and environmental isolates were examined by RFLP typing. 

This preliminary analysis suggests that PCT can be use to accurately early identify sepsis only at levels above 2 ng/ml and then use them to decide to rapidly beginning the use of antibiotic. In patients with PCT below 2 ng/ml we cannot use them to exclude the diagnosis of sepsis. With the cutoff 0,5 ng/ml we found the same analysis. Other studies with more samples are necessary to confirm this conclusion. 

During these three years 402 patients were hospitalized in total. One hundred and thirty one (29.8 %) were hospitalized less than 72 h and were excluded. A total of 99 bacteremias were observed. Forty -four bacteremias were catheter related bloodstream infections. Fifty five were due to gram negative microorganisms (Pseudomonas aeruginosa 23%, Acinetobacter baumanni16%, Klebsiella pneumonia16%). In the following table, resistance to broad spectrum antimicrobials is presented during these three years. 

Infection in patients with severe stroke is an important problem and the sensitivity and specificity of its diagnosis with clinical criteria are deficient. Fever is a common event and, as leucocytes or C-reactive protein, its specificity is very low in this kind of patient.

Our objective was to evaluate the utility of a biological marker such as procalcitonin (PCT) in the diagnosis of infection in patients with severe stroke.

We followed 27 patients with severe stroke receiving mechanical ventilation because of coma. During the first 4 days of evolution NIH and APACHE II scales were registered, we measured PCT and C-reactive protein on days 1 and 3 and if infection was suspected microbiological samples were collected. Infection was diagnosed if the patient fulfilled the CDC criteria. Mann-Whithney U and X-square tests were used.

Twenty-six cases corresponded to haemorrhagic stroke. Baseline characteristics were: mean age 57 years, 55% males, Glasgow scale 6 (3-12), NIH scale 30 (13-37), APACHE II 20 (9-32), temperature 36.9 o C (33-39.5), leucocytes 10341/mm3 (2400-24800), PCT 0.404 ng/ml (0.022-1.311) and C-reactive protein 37.3 mg/dl (3.7-121). On the third day of evolution 3 cases of ventilator-associated pneumonia were diagnosed. When compared with the noninfection group there were no differences in baseline characteristics and on the infection day we only found differences in PCT, 3.505 ng/ml in front of 0.552 ng/ml; p < 0.001. Seventeen (70%) of the patients without infection presented a temperature 3 38 o C sometime during the follow-up and in all cases PCT did not show any change. 

These results indicate that PCT is a useful tool in the diagnosis of infection in patients with severe stroke.

The ongoing challenge of accurately diagnosing infection in the ICU motivates a search for novel molecular diagnostics. We reported recently that microarray analysis of circulating leukocytes can be used to derive a "riboleukogram", which captures the dynamics of the host response to and recovery from ventilator-associated pneumonia (VAP). In the current study, we tested the hypothesis that the informational content of circulating leukocytes differs, thereby allowing one to rank leukocyte populations on their potential to contribute to RNA diagnostics for pneumonia.

Sixteen patients (10 male, 6 female) at risk for VAP were entered into our IRBapproved study that collects blood and clinical data every 48 hours for up to 21 days. Four of the sixteen patients developed VAP as diagnosed and treated by the attending ICU physician.

Previously reported blood protocols were used to isolate buffy coat, enriched neutrophil, and enriched monocyte populations by using negative selection. Cellular purity was assessed by FACS for one of the 4 VAP patients. Genome-wide expression analysis was performed on RMAnormalized signal from Affymetrix U133 2.0 Plus GeneChips. EDGE software (FDR=0.10) was used to determine changes in mRNA abundance over time for each cell population.

During the 5-day window in which each of the four patients (all males) developed VAP, significant changes in gene expression were observed (Table) , but the information content (number of genes altered) varied across leukocyte populations. These differences were not due to signal variance (coefficient of variation, CV) or differences in the number of samples available for analysis. Moreover, only 0.6% of the monocyte gene list overlaps with the neutrophil list, arguing that neutrophil contamination of monocyte populations is insufficient to explain the 40-fold difference in gene number. 

The aim of the present study was to evaluate the relationship between the cytokine expression in bronchoalveolar lavage fluid and bacterial burden in mechanically ventilated patients with suspected pneumonia. Mechanically ventilated patients with suspected pneumonia admitted in ICU from November 2004 to January 2006 were prospectively enrolled. Fiberoptic bronchoalveolar lavage (BAL) was performed with 150 ml of sterile isotonic saline in 3 aliquots of 50 ml, local anesthetic were not used. BAL samples for microbiologic quantitative cultures and BAL cytokines: interleukin (IL) 6, IL 8, tumor necrosis factor-alpha (TNF-alpha), granulocyte colony stimulating factor (G-CSF) and granulocyte-monocyte colony stimulating factor (GM-CSF) were measured.

. 59 patients were included, most of the patients (79.7%) were with prior antibiotic therapy. 22 patients (37.2%) had a positive bacterial culture defined than a diagnostic threshold of > 104 colony-forming unit/ ml. The concentration of TNF-alpha was significantly higher in the group of patients with positive BAL (table 1) . It has been demonstrated in a swine model that therapeutic hypothermia (30˚C) facilitated transthoracic defibrillation. However, the mechanisms leading to reduced defibrillation threshold (DFT) remain unclear. We hypothesized that therapeutic hypothermia promotes the wavefront organization of ventricular fibrillation (VF), therefore facilitating defibrillation. METHODS. By using a two-camera optical mapping system, epicardial activation patterns of VF were studied in 7 isolated rabbit hearts at baseline (37˚C), 10-min therapeutic hypothermia (30˚C), and 10-min rewarming (37˚C). In 12 additional hearts, DFT50 (voltage required to achieve 50% probability of successful defibrillation, n=6 hearts) and APD (action potential duration)/conduction velocity (CV) restitutions (n=6 hearts) were determined at these 3 stages. RESULTS. Comparing with at baseline (35±5%) and rewarming (34±7%), there was a higher percentage of VF duration containing organized repetitive activities during hypothermia (73±10%, p<0.001). However, there was no significant difference of DFT50 among these 3 stages (151±34, 141±38, and 146±46 V, p=0.556). The electrophysiologic characteristics of ventricles at these 3 stages were summarized in table 1. In brief, hypothermia prolonged APD, decreased CV, and subsequently shortened wavelength. Hypothermia also failed to flatten the slope of APD restitution. Furthermore, APD dispersion at the epicardial surfaces of both ventricles and CV heterogeneity among 4 epicardial lines were all enhanced by hypothermia. (pt) with Acute Coroanry Syndrome (ACS) at admission is a associated with a high mortality. The mechansims are poorly understood. We sought to determine an interrelation between no coronary reflow after percutaneous coronary intervention (PCI), the likelihood of developing cardiogenic shock, death in hospital and plasma glucose level at admission.

We performed a prospective analysis of 161 consecutive pt presenting with an ACS in our emergency room. We recorded basis data (gender, age, BMI), cardiovascular risk factors, burden of coronary artery disease (CAD), coronary blood flow after PCI, Killip-classification, left vetricular ejection fraction, probabilty of developing cardiogenic shock and the likelihood of dying in-hospital. 

Our findings suggest that elevated BS at admission is a useful risk marker to identify pt with a high risk to develop coronary no reflow-phenomenon after PCI. This may be due to increased inflammatory activity and hypercoagulability. If one dies in cardiogenic shock, these pt present always with elevated BS at admission.

Prull MW, Trappe HJ. Activation of blood coagulation in NSTEMI: Does diabetes mellitus matter? Intensivmed 2007. 

We measured serum cortisol levels before and 60 minutes after a 0,25mg corticotropin stimulation test in 15 pts with CS following acute myocardial infarction (MI) and in a control group of 8 pts with uncomplicated MI at day 0, 1, 2, 3, 5, and 7 after onset of shock/MI. RAI was defined by an increase in serum cortisol levels in response to corticotropin of less than 9µg/dl. Data were correlated to vasopressor-need and interleukin (IL) levels (IL1,IL6,IL8,IL10).

Baseline cortisol levels in pts with CS were significantly higher than in control pts especially on day 0 (35±22 vs 15±9, p=0.006). In 5 CS-pts the test-series were stopped at day 1 to 3 because the physician in charge started a therapy-trial with hydrocortisone due to increasing vasopressor need. Three other pts died within the seven day period. RAI was observed only at day 0 in 5 of the 15 CS-pts but in none of the control pts (p=0.06). These CS pts with RAI had higher Il-6 and IL-10 levels at baseline ( 

During tidal mechanical ventilation, an end-expiratory pause abolishes the cyclic increase in intra-thoracic pressure. This may produce a transient increase in cardiac preload and then in cardiac output in volume responsive patients. Our objective was to test whether the effects of an end-expiratory pause on cardiac index and pulse pressure may help in detecting fluid responsiveness in patients with acute circulatory failure.

In 30 mechanically ventilated patients with an acute circulatory failure and no spontaneous ventilator triggering who were deemed at volume expansion, we performed a 15-sec end-expiratory pause. We continuously measured the systemic arterial pressure and the pulse contour-derived cardiac index (PiCCO device) at baseline, during the 10 last seconds of the end-expiratory pause and after a 500mL saline administration.

Volume expansion induced an increase in cardiac index ≥15% in 20 patients (classified as responders). In these patients, volume expansion increased the cardiac index by 41±35% from 2.1±1.1 L/min/m2. Before volume expansion, the end-expiratory pause had induced an increase in cardiac index by 11±11% and in pulse pressure by 14±16% as compared to the baseline values. By contrast in the 10 non-responders, before volume expansion the cardiac index and the pulse pressure did not change during the pause as compared to baseline (2±2 % and 1±3% increases, respectively). Importantly, an increase in cardiac index ≥5% during the end-expiratory pause predicted fluid responsiveness with a sensitivity of 84% and a specificity of 90%. A pause-induced increase in pulse pressure ≥4% detected fluid responsiveness with similar sensitivity and specificity (90% and 90%). In responders, a second end-expiratory pause was performed again immediately after volume expansion. In 16 patients, the increases in cardiac index induced by this second pause induced had dropped below 5%. In the 4 remaining responders, the second pause induced an increase in cardiac index still higher than 5% (12±2%). In these patients, the pause-induced increase in cardiac index was abolished by a second 500mL saline administration.

CONCLUSION. An increase in cardiac index and in pulse pressure during an end-expiratory pause enables to detect fluid responsiveness in critically ill patients with mechanical ventilation and acute circulatory failure. , and tissue Doppler imaging measurements of the mitral annulus velocities like early (Ea) peak diastolic velocity. The aim of the study was to examine which echocardiographic index is the best marker of preload by making the hypothesis that a good measure of preload should increase with fluid-induced increase in stroke volume (SV) but not with dobutamine-induced increase in SV. 

Comparison of the capacity of the intra thoracic blood volume index (ITBVI) and the central venous pressure (CVP) to predict fluid responsiveness in critically ill patients with acute circulatory failure (systolic blood pressure < 90 mmHg or vasopressor requirement).

METHODS. This prospective interventional study performed in a surgical Intensive Care Unit of a tertiary University Hospital included 35 (21 males) mechanically ventilated and sedated patients with acute cardiovascular failure requiring cardiac output measurement (transpulmonary thermodilution technique)and a fluid challenge. Intervention: Fluid responsiveness was defined as an increase in stroke index (SI = cardiac output/heart rate/body surface area) ≥ 15%. Receiver operating characteristic (ROC) curves were generated for ITBVI and CVP.

In 48 eligible patients, 10 could not be included because of cardiac arrhythmia (n = 8) or moribund status (n = 2) or protocol violation (n = 3). The cause of acute circulatory failure was septic shock in 24 (69%) patients, haemorrhagic shock in 2 (6%) patients, and systemic inflammatory response syndrome in 9 (26%) patients. Fluid challenge induced an SI increase ≥ 15% in 18 (51%) patients (responders(R). No statistical difference was shown between responders and non responders for CVP and ITBVI. The areas under the ROC curves of ITBVI and CVP were 0.64 [95% CI: 0.46-0.80], and 0.68 [95% CI: 0.50-0.83], respectively, without any statistical difference (p = 0.73). The best cut of value for CVP and ITBVI were 9 mmHg (sensitivity = 61 %; specificity = 82%) and 928 ml.m-2 (sensitivity = 78 %; specificity = 53 %), respectively. The relative changes in SI and CI were correlated with relative changes in ITBVI (r = 0.59, p = 0.001; r = 0.66, p = 0.0001 respectively) but no correlation was found between relative changes in SI and CI and relative changes in CVP (r = -0.07, p = 0.70; r = 0.10; p = 0.54).

CONCLUSION. ITBVI is similar to CVP to predict fluid responsiveness in critically ill patients with acute circulatory failure. 

The pulse pressure variation (PPV) is used to predict fluid responsiveness in mechanically ventilated patients. Nevertheless false positive of this parameter have been reported especially in patient with right ventricular dysfunction. The peak systolic velocity of tricuspid annular motion (Sta) assessed by Doppler echocardiography (DEC) is a parameter of right ventricular systolic function. The aim of the study was to find out whether Sta can discriminate between false and true positive of VPP.

METHODS. 35 mechanically ventilated patients were prospectively included. All patients had a measurement of PPV>12%. A DEC was realised before and after infusion of 500 ml of colloid solution. Patients were separated into 2 groups as they were responders (R) (at least 15% increase in stroke volume (SV)) or non-responders (NR) to fluid infusion. All data are expressed as mean [standard deviation]. The comparison of demographic, hemodynamic and echocardiographic parameters in R and NR patients was performed using a t-test. A p value < 0.05 was considered statistically significant. ROC curves were plotted. A threshold value of Sta was calculated with ROC curve. 

In the resting patient, pulse pressure (PP = systolic -diastolic pressure) is mainly related to arterial stiffness and stroke volume index (SVi). The dynamic effects of fluid loading on PP are poorly documented and were studied in the critically ill using arterial tonometry. We tested the hypotheses that i) arterial stiffness was unchanged after fluid loading, ii) PP changes paralleled SVi changes such that PP increased in fluid-responders only, and iii) aortic PP was more indicative of SVi changes than radial PP.

Twenty-two critically ill patients (9F), mean age(SD), 53(15) years, were prospectively included. Radial pressures were calibrated from brachial cuff pressures. Radial applanation tonometry (Sphygmocor ® ) allowed us to estimate aortic PP, left ventricular ejection time, and the augmentation index which quantifies wave reflection. The SVi was calculated by transpulmonary thermodilution. The arterial stiffness was estimated from the aortic pressure curve using standard formula. Fluid challenge (500 mL saline 0.9%) was required by the patient's hemodynamic status. Data were obtained before and immediately after fluid loading. Responders had increases in SVi > 15 %.

Baseline mean values were as follows: SVi = 38(13) mL.m-2, heart rate= 89(15) bpm, mean arterial pressure (MAP) = 75(15) mmHg, radial PP = 51(18) mmHg, aortic PP = 32(11) mmHg. After fluid loading, SVi increased from 38(13) to 42(15) mL.m-2 and MAP increased from to 80(17) mmHg (each p < 0.01). Arterial stiffness was unchanged (1.28(0.71) vs 1.33(0.89) mmHg.mL-1. m2 ) as well as heart rate, left ventricular ejection time, radial and aortic PPs and augmentation index. There was a positive linear relationship between the SVi changes and the changes in radial PP (r2 = 0.50) and aortic PP (r2=0.61) (each p < 0.01), not MAP (r2 = 0.006). When responders (n=8) and non responders (n=14) were compared, the increases in MAP were similar while the changes in PP were higher in responders (radial: 13 mmHg, 24%; aortic: 10 mmHg; 35% ) than in non responders. (radial: -1mmHg, -0.5%, aortic: -1mmHg; -0.6%) (each p< 0.01).

Given the unchanged arterial stiffness throughout the fluid infusion, the changes in aortic PP (and slightly to a lesser extent radial PP) paralleled the changes in SVi. Both radial and aortic PPs increased in responders but not in non responders, while MAP similarly increased in the two groups. The capability of arterial PP changes to track SVi changes during fluid loading appears promising but deserves a further large scale study. 

New device may be used in intensive care unit to measure Cardiac Output (CO) by arterial pulse pressure waveform analysis , but comparative studies with CO thermodilution in cardiac surgery have shown large bias between the methods 1. Aim of this study is to evaluate in critical ill patients not submitted to cardiac operation 1-Cardiac Output (CO wave) obtained using Flo Track TM Vigileo .2 -The correlation with CO obtained by Thermodilution (CO therm).

METHODS. 35 critical care patients admitted to a general intensive care were enrolled in the study . All patients were mechanically ventilated ( TV 6-8 ml /Kg Pl press < 30 cmH20) and connected to an integrated monitoring system ( Flow Trac TM / Vigileo TM , Ewdards Lifescience ,Irvine ,CA, USA ) that attaches to an arterial cannula . A central venous catheter and a PAC ( Thermodilution catheter ; Arrow International , Inc ., reading ,PA,USA ) was inserted via the jugular internal vein . After haemodynamic stabilization CO wave was calculated from an arterial pressure based algorithm that utilises the relationship between pulse pressure and stroke volume , primarily based on the standard deviation of the pulse pressure waveform. At the same time a CO therm. determination was performed by triple injection of 10 ml of iced isotone NA CL into the central line of the PAC. Every patients had two CO determination at two time point. For each measurement of CO therm corresponding simulataneous CO wave was documenteted . A regression analysis and Bland Altman analysis was used to compare the two methods of CO determination.

A total of 70 CO determination was performed in 35 patients . CO Vigileo correlated CO thermodilution with r = 0.68 , p< 0,001. At Table 1 are reported the Bland Altman's results. 

The left ventricular ejection fraction (LVEF) as measured by echocardiography is considered as the reference estimate of the LV global contractility at the ICU bedside. The transpulmonary thermodilution technique (PiCCO system) continuously provides a measure of the cardiac function index (CFI), which is the ratio of cardiac output over global end-diastolic volume. Thus it could be considered as a marker of cardiac global contractility and could enable a continuous monitoring of this key parameter. We tested whether CFI could actually behave as an indicator of LV systolic function by testing if it fulfilled the following criteria: (i) increase with inotropic stimulation, (ii) no alteration by fluid loading, (iii) correlation with the echographic LVEF and (iv) ability to track the changes in LVEF during inotropic stimulation.

In 33 patients (40 cases) with an acute circulatory failure, we simultaneously measured the echographic LVEF (transthoracic 4-chambers apical view) and the CFI at baseline, after a 500mL saline administration in a group of 22 cases and after 15-min of dobutamine administration in a group of 18 cases.

Volume expansion did not alter LVEF significantly (47±11% vs. 46±10% at baseline) nor CFI (4.4±2.2 vs. 4.5±2.2 min-1 at baseline). By contrast, dobutamine infusion induced a significant increase in LVEF from 30±10% at baseline to 39±10%(+33±29%) and in CFI from 3.2±1.7 at baseline to 4.0±1.9 min -1 (+27±22%). Considering the whole set of CFI:LVEF pairs of measurements (n=80), a significant correlation was observed between CFI and LVEF (r=0.65, p<0.0001). Importantly, a CFI value <3.2min -1 predicted a LVEF value higher than35% with a sensitivity of 76% and a specificity of 87%. In patients receiving dobutamine, there was a significant correlation between the changes in CFI and the changes in LVEF induced by dobutamine infusion (r=0.69, p<0.0001).

Our study demonstrates that CFI fulfilled the criteria that are required from a bedside indicator of LV contractile function: it was increased by inotropic stimulation while it was not altered by volume expansion, it was fairly correlated with the echographic LVEF and it was able to track the changes in echographic LVEF with reliability. This suggests that the continuous monitoring of CFI provided by transpulmonary thermodilution could help in assessing the effects of inotropic therapy and could alert the physician in case of abrupt LV contractile deterioration.

Passive leg raising (PLR) is a predictive test of preload responsiveness in patients with acute circulatory failure. It could predict fluid response to fluid loading in mechanically ventilated patients. Critically ill patients have an increased risk of lower extremity deep venous thrombosis. Elastic compression stocking (ECS) is frequently used in association with unfraction or low molecular weight heparin. The aim of this study was to evaluate the effect of the elastic compression stocking on the PLR test variations.

METHODS. 20 patients undergoing cardiac surgery were included. All of them were anaesthetised and mechanically ventilated (tidal volume ≥ 8ml/kg). Pre-operative left ventricular ejection fraction was > 50% for all patients. They were monitored with central venous pressure (CVP), invasive blood pressure and esophageal doppler. Hemodynamics parameters were obtained before and after PLR, without and with elastic compression stocking respectively (SSV = systolic stroke volume, CO = cardiac output, PPV = pulse pressure variation and SBP = systolic blood pressure). Results are presented as median [inter quartile range](IQR) and compared with Mann Whitney test.

. table represents hemodynamics variations after PLR without and then with elastic compression stocking. Second table represents hemodynamics effects of the elastic compression stocking in supine position (SP). CONCLUSION. This study shows a clear improvement in gut permeability after surgery. The effects of early feeding shall be assessed in a future study. METHODS. descriptive-prospective study. Pre and post-class 12 question -survey (administered one week before and after). The Transplant co-ordination team gave informative classes in secondary schools, 2005-06 / 2006-07.

. surveys collected; 713 pre/ 581 post-class: 58% of 4ESO (16 years old), 38% 2Bachiller (18 years old) and 4% Ciclo formativo (18 years old) / 581 post-class: 36% 4ESO, 59% 2Bachiller and CF 4% . 98% had some prior awareness and 67% broad knowledge. Massmedia is usually sole information channel (53%), ticked in all cases. Other sources were: family, school and peers. Regarding attitude to donation: we found no differences in refusals between own donation or relatives'(25%); or in doubts 46% -42%. Related to transparency and parity of the health system: 26% believed equality did not exist and 60% had doubts. 55% felt this inequality was worse abroad. 64% are convinced that organ trafficking exists and 26% assume it is possible. Pre-course standpoint by course is showed in figure 1 . 9% had prior knowledge about Spanish Transplant law. Following classes the students claim higher awareness (87%). In general they maintain their standpoint on donation,45% have reconsidered their previous attitude. Regarding transparency and equality, 26% maintain doubts and 15 % are convinced of its absence. On trafficking: 72% assume it is possible, 13% occurs exclusively abroad, uniform group distribution. Post-course attitudes by course are in figure 1 . Despite an in-depth discussion about the law and its consequences (presumed consent), they generally disagree and some consider this too extreme , refusing to accept that donation is an obligation (only 8% agree) and believing that it should be an optional act of solidarity (81%). CONCLUSION. Knowledge about donation and transplant in urban areas is slanted, due to information sources ( usually mass media ) and a warped (TV-dominated) perception of the health system's transparency and equality. A considerable number of students still refuse donation or maintain their scepticism, despite a decrease following classes. However, our desire is not to convince them to become donors, we simply wish to provide decision-making tools. Generally college students ,without gender differences, are the most resistant to the process, having the greatest incidence of refusals and doubts about transparency, equality and organ trafficking. 140 (115-186) 0.18c (PaO2/FiO2) / PEEP day 1 12.5 (7.9-25.19) 10.7 (6.2-20.7) 0.33c (PaO2/FiO2) / PEEP day 3 14.8 (9.6-20.3) 11.2 (8.4-14.7) 0.01c (PaO2/FiO2) / PEEP day 7 13.3 (10.8-27.5) 10.6 (7.8-17.9) 0.02c CONCLUSION. The PaO2/FiO2 ratio on day one is useful to predict mortality, but not in the subsequent days. The (PaO2/FiO2)/PEEP index is a better predictor in later days, specially on the third and seventh day of MV.

A. ROCH* 1 , L. Fouché 1 , J. Forel 1 , D. Blayac 2 , C. Aglioni 3 , D. Lambert 2 , J. Carpentier 3 , L. Papazian 1 1 Réanimation Médicale, 2 DAR, Hôpitaux sud, 3 Réanimation, Hôpital Laveran, Marseille, France INTRODUCTION. General anesthesia promotes atelectasis of the dependent parts of the lung. We evaluated the differential effects of neuromuscular blocking agents (NMBA) on consolidation formation in healthy or injured lungs.

METHODS. 30 pigs (37±2 kg) were anaesthetized with Pentobarbital, Fentanyl and Ketamine in order to prevent spontaneous ventilation and ventilated using volume controlled ventilation (Vt 10 ml/kg, FiO2 0.4) for 6 hours after randomization into 5 groups: healthy lungs ventilated without (HZEEPno) or with NMBA (cisatracurium, HZEEPnmba), healthy lungs ventilated with NMBA and PEEP 5 (HPEEPnmba) and injured lungs ventilated without (TweenPEEPno) or with NMBA (TweenPEEPnmba). Lung injury was induced using instillation of 1.5 ml/kg of 7.5% Tween 20. Injured lungs were ventilated with PEEP 8 , FiO2 0.8 and Vt 10 ml/kg. After lung removal, six sections of equal thickness were obtained from the right lower lobe and 4 from the upper. Sections were photographed and analyzed using a software (Sigmascan pro 5, SPSS Inc). The areas of consolidated, edematous and normal parenchyma were measured on each section and then added to obtain the percentage of consolidated lung.

. NMBA use induced a two-fold increase of the consolidation (from 20±6 to 39±8 %)that was totally prevented by PEEP 5. The deleterious effect of NMBA on derecruitment did not occur in injured lungs. Consolidation was located to the dependent parts in healthy lungs and NMBA extended consolidation towards more cephalad parts. In injured lungs, consolidated parenchyma was diffuse and its cephalo-caudal distribution was not affected by NMBA. PaO2 to FiO2 ratio was affected neither by NMBA nor by PEEP. * p<0.001 vs HZEEPno and HPEEPnmba; **p<0.001 vs HZEEPno and HZEEPnmba.

CONCLUSION. NMBA increase dependent lung consolidation during volume-controlled ventilation of healthy lungs. This effect is prevented by a moderate PEEP level. In contrast, NMBA do not increase the extent of pathologic lung areas in injured lungs ventilated during a 6-h period.

20th ESICM Annual Congress -Berlin, Germany -7-10 October 2007 S81 0306

M. Amigoni* 1 , M. Scanziani 1 , G. Bellani 1 , G. Balconi 2 , E. Zanotto 1 , S. Masson 2 , N. Patroniti 1 , R. Latini 2 , A. Pesenti 1 1 Dept of Experimental Medicine, Milano-Bicocca University, Monza, 2 Cardiovascular Research, Istituto di Ricerche Farmacologiche Mario Negri, Milano, Italy INTRODUCTION. Surfactant dysfunction seems to play a pivotal role in the deterioration of gas exchange and lung mechanics that occurs in ALI/ARDS following aspiration pneumonitis. We investigated the effects of exogenous surfactant administration in a murine model of unilateral acid-induced lung injury.

We instilled 1.5 ml/kg BW of 0.1M hydrochloric acid in the right bronchus of anesthetized and mechanically ventilated mice (Vt 8-10 ml kg-1 BW, RR 130 min-1, FiO2 1 and PEEP of 2.5 cmH2O). Mechanical ventilation was stopped 10 minutes after injury; animals were then placed in an oxygenated chamber (FiO2 0.5). After 10', 1hr or 6hrs from acid instillation, the mice were reintubated and received a single bolus of surfactant in the injured lung at a low or high dose. Each animal was again mechanically ventilated for 10 minutes, placed in oxygenated chamber until full awakening. Acid-injured mice instilled at the same time and with the same volume (1ml/kg BW) of sterile saline (0.9% NaCl) were used as controls. Lung mechanics, blood gas analysis, and lung myeloperoxidase activity (MPO) were assessed 24hrs after acid aspiration.

No effect of surfactant administration was present upon oxygenation 24hrs after the injury. At the opposite the high dose group showed a significantly better compliance at 24hrs, when compared to both the low dose and control groups. This effect was present only in the late (6hrs) administration group. MPO activity did not change after surfactant treatment in the right (injured) lung while in the controlateral, it tended to be lower in both low and high dose when treatment administration occurred at 6hrs (n=4/group: N right lung 38±15 left lung 40.2±19.7; S(low dose) right lung 39.2±17.8 left lung 28.7±15.8; S(high dose) right lung 40.1±16.5 left lung 30.3±13.6). 

Pulmonary aspiration is associated with significant morbidity and mortality 1 . Several risk factors for aspiration have been highlighted in the literature 2 . The aims of this study were to: (i) identify specifically which patient factors predispose to aspiration and (ii) determine the outcome of patients admitted to our inner city hospital intensive care unit (ICU) with a diagnosis of aspiration.

We identified 27 patients with a diagnosis of pulmonary aspiration on our ICU over a 2 year period (August 2004-06), by using our institution's ICNARC (Intensive Care National Audit and Research Centre) database. 19 of these patients' case notes were able to be retrieved and reviewed in detail. Patient demographics, risk factors for aspiration, number of ventilated days, ICU & hospital length of stay and mortality were analysed. We also looked at any documented signs that supported the diagnosis of aspiration.

Median age of the patients was 58 years (range 30-83). 13/19 patients (68%) were male. The main risk factor was a reduced Glasgow Coma Score (19/19 patients, 100%): the median score was 5 (range 3-13). The following risk factors were also identified: obesity (5/19 patients, 26%), excessive alcohol intake (5/19, 26%), acute cerebrovascular event (4/19, 21%) and cardiorespiratory arrest (4/19, 21%). The following signs were most frequently observed: perioral vomitus (12/19 patients, 63%), acute hypoxaemia (16/19, 84%) and a new radiographic infiltrate (10/19, 53%). One patient exhibited all three markers. All 19 patients required mechanical ventilation. The median duration of ventilation was 11 days (range 1-33). The median length of ICU stay was 12 days (1-41) and the median length of hospital stay was 31 days (1-256). ICU mortality was 27% (5/19 patients) while hospital mortality was 37% (7/19).

Patients who presented to our inner city ICU with aspiration had risk factors that included impaired conscious level, obesity, a recent cerebrovascular event or cardiorespiratory arrest. Signs that supported the diagnosis of aspiration were the presence of perioral vomitus, acute hypoxaemia and a new radiographic infiltrate. ICU and hospital length of stay were both prolonged, but ICU and hospital mortality were no higher than our institution's overall rate. A high index of suspicion should be applied to these patients at risk of aspiration, to facilitate the early initiation of appropriate care. REFERENCE(S). 1. Hickling K. A retrospective survey of treatment and mortality in aspiration pneumonia. Int Care Med 1998; 14: 617-22. 2. Kozlow J. Epidemiology and impact of aspiration pneumonia in patients undergoing surgery in Maryland, 1999 -2000 . Crit Care Med 2003 31: 1930-7. 

T. Tagami* 1 , S. Kushimoto 2 , T. Atsumi 2 , R. Oyama 1 , K. Matsuda 3 , M. Kawai 2 , H. Yokota 2 , Y. Yamamoto 2 1 Surgery, Tokyo Metropolitan Saiseikai Central Hospital, 2 Critical Care Medicine, Nippon Medical School, Tokyo, 3 Critical Care Medicine, Yamanashi Prefectural Central Hospital, Yamanashi, Japan INTRODUCTION. Restoration of intravascular volume by massive fluid administration without pulmonary edema formation is one of the biggest challenges in the early treatment of burn shock. Although it is not easy to predict the development of the respiratory failure before the treatment, the hallmark of the edema is increased capillary permeability which may be possible to measure by the pulmonary vascular permeability index (PVPI). The aim of the present study was to clarify whether the PVPI is predictable indicator of pulmonary edema formation in patients with burn.

We studied 11 mechanically ventilated patients with burn involving more than 25% of the body surface area that were treated at intensive care burn unit between July 2004 and January 2007. All patients had a central venous catheter and a thermistor-tipped arterial thermodilution catheter (PiCCO system) for hemodynamic management. We measured the extravascular lung water index (EVLWI) and the pulmonary vascular permeability index(PVPI) as soon as the PiCCO catheter was inserted. Infusion volume was calculated according to the Parkland formula. Only crystalloid fluid (lactated Ringer's) was infused during the first 12 hours after the thermal injury. We investigated the medical records and defined the respiratory failure during the period of burn shock as a clinical syndrome of acute respiratory distress associate with pulmonary rales and radiographic evidence. Inclusion criteria were: 1)acute onset and rapid progress, 2)oxygenation index (PaO2/FiO2 ratio<200 and 3) bilateral infiltrates on Chest X-ray. Those are the part of the standard criteria of acute respiratory distressed syndrome.

The PVPI was significantly higher in the patient with respiratory failure (n=4 PVPI: 2.65±0.98) than in patient without respiratory failure(n=7 PVPI:1.43±0.26) before the fluid treatment. There was no significant difference between the groups in terms of EVLWI at the beginning (7ml/kg VS 6.1ml/kg). Although the EVLWI increased after 48 hours in the patient with respiratory failure, it did not change in patient without respiratory failure(18.2ml/kg VS 6.0ml/kg).

The PVPI increased before the EVLWI increased in patient with respiratory failure. The PVPI is considered to be the predictable value to identify the risk of respiratory failure during the period of burn shock. 

Ultrasonography allows observation of diaphragm. In healthy subjects, a correlation was found between its excursion and the tidal volume. In addition, diaphragm thickness variation measured in the zone of apposition has been used to evaluate paralyzed diaphragm. We assessed the accuracy of these indexes to assess diaphragmatic function and respiratory workload.

Five patients were studied in spontaneous ventilation (SV) and during noninvasive ventilation at different levels of pressure support (PS). Diaphragmatic excursion (E) was carried out subcostally. Diaphragm thickness was measured in the zone of apposition and the thickening fraction (TF) was calculated as TF = (thickness at inspiration -thickness at expiration)/thickness at expiration. Diaphragmatic pressure time product per breath (PTPdi) was measured by assessment of esophageal and gastric pressure.

PTPdi and TF both decreased as the level of pressure support increased (fig 1 and  2) . A positive correlation was found between PTPdi and TF(r=0.68; p=0.01; fig 3) . In addition, there was also a significant correlation between tidal volume and E (r=0.91; p<0.01; fig 4) .

Ultrasonography of the diaphragm could be applied in intensive care to assess diaphragmatic function. TF and PTPdi decrease as the level of pressure support increases. These results suggest that TF could help to assess diaphragmatic contribution to respiratory workload.

REFERENCE(S). (1) Fantus G. Metformin's contraindications: needed for now

Frequency of inappropriate metformin prescriptions

Systematic review and meta-analysis of studies of the timing of tracheostomy in adult patients undergoing arteficial ventilation

Catheter infection is a common concern in the intensive care unit (ICU). Recent works have pointed that the site of catheters is related to this problem. We analysed data obtained from our data base to confirm the results of previous works. METHODS. Catheters were inserted in a surgical-medical ICU, along five years. Semiquantitative cultures were obtained if the catheter was kept in place more than 48 hours and it was no longer necessary, the catheter was withdrawn because of fever of unknown origin or an infection was suspected at the point of insertion. Every catheter site, culture and germ was registered in our patient data base. We studied the following variables: type of catheter, site and results of cultures. Statistical analysis: variables were compared by Chi-square. A p< 0,05 was considered statiscally significant. RESULTS. A total of 2.407 catheters were registered (venous catheters 1307, arterial 1100). Rate of germs was as follow: gram-positive 67,4%, gram-negative 28,1%, fungi 3%, contaminated flora 1,5%. Site and germs were not statistically associated. Table 1 shows type, site and rate of infection of cultured catheters. Femoral arteries were more frequently cultured than radial arteries (p< 0,001); no differences were found for cultured venous catheters. Femoral arteries were infected more frequently than radial (p< o,001); yugular and femoral venous catheters were more frequently associated to infection. (SC) in non neutropenic patients is increasing with a high cost and mortality. We define the clinical and epidemiological profile of patients admitted to our ICU and the microbiological aspects of the pathogen. Mortality analysis was done, including Sevilla score system (SSS).

We include 36 patients admitted in ICU from 2002 to 2007 with Candidas ssp (CD) positive blood cultures (BACTER system). We analysed demographic factors, reason for admission to the unit, associated risk factors, need of multi-instrumentation or parenteral nutrition, value of APACHE II, and length of stay in the ICU. The kind of CD diagnosed, its sensitivity profile, and the existence or not of previous wide spectrum antibiotic or antifungic therapy were determined. The Sevilla score system was applied and correlated with mortality. Chi square, t-test and multivariant analysis were made.

There were 69.4% male patients, with 58 years old median age and with a length of stay longer than 14 days. The reason for admission was sepsis (27%), surgery (22.2%), acute respiratory failure (16.7%) and trauma patiens (11%). APACHE II median was 22.8 points.Risk factors related with fungal infections were diabetes (19.4%), neoplasia (11%), steroid therapy (7,7%), a length of stay longer than 7 days (44%) and antibioticoterapy. None had neutropenia. 94% of patiens received antibioticoterapy previous to diagnosis, 69.4% parenteral nutrition and 88% of them underwent multi-instrumentation. Patient isolation was achieved in 36% of them (90% in period 2005-7). Candida Albicans was isolated in 44.4% of cases against 55.6% of Candida nonalbicans, specially C. Parapsilosis 30,6%. First antifungal therapy was Fluconazole (61%), Caspofungin (16.7%) and lipid amphotericins (5.6%). We found a significant increase of SC cases along the years, (36% in 2002-4 vs 63.9% in 2005-7, p<0 .049), being unresponsive to azoles 3.8%. Mortality was specially high (41.7%), unrelated with CD type; those with high/moderate SSS risk had a significative higher mortality (p<0.035). Candida albicans was more frequently found in septic patients while Candida nonalbicans was gaining place in patients under parenteral nutrition (C.parapsilosis).CONCLUSION. 1) Systemic candidiasis affects men admitted with sepsis or surgery, with a high APACHE II index, multiple organ failure, multi-instrumentation and more than two weeks intensive care unit stay. 2) We observe a progressive incidence of non albicans candidiasis (C. Parapsilosis). 3) Type of Candida ssp did not affect mortality. 4) C. Albicans was more frequently isolated in septic patients, while Candida nonalbicans was predominant in cases with parenteral nutrition. 5) Mortality was greater in moderate/higher SSS risk group.

F. Alvarez-Lerma* 1 , M. Palomar 2 , P. 

Objetive: To present changes of multiresistance markers in ICU-acquired infections.

A prospective, cohort, multicenter study. All patients admittted to the participating spanish ICUs between the years 2002 and 2006 were included. Patients were followed until discharge from the ICU or up to a maximum of 30 days. The following infections were studied: mechanical ventilation-related pneumonia (MV-P), catheter-related urinary tract infection (CR-UTI), and primary bacteremia (PB). Markers of multiresistance were those defined by the CDC (1) 

Of a total of 39,937 pacientes included in the study,4,050 (10.1%) developed 5,546 infections (13.9%) during their stay inthe ICU, in which a total of 6,034 pathogens were identified.Multiresistance markers are shown in Table 1 . 

Pulse pressure variation greater than 12% predicts fluid responsiveness in patients ventilated with large tidal volumes. The aim of this study is to evaluate the influence of a low tidal volume on the capacity of pulse pressure variation (DeltaPP) to predict fluid responsiveness.METHODS. This is a prospective interventional study that took place in a 33-bed university hospital medico-surgical ICU. The study included eighteen mechanically ventilated critically ill patients with a low tidal volume (6-7 ml/kg) requiring fluid challenge. Fluid challenge was performed with 1,000 ml crystalloids or 500 ml colloids.

Complete hemodynamic measurements including DeltaPP were obtained before and after fluid challenge. Overall, the cardiac index increased from 3.04±1.54 to 3.80±2.32 l/min/m2 (P <0.05). It increased by more than 10% in 11 patients (responders). Pulmonary artery occluded pressure was similar (14.9±5.5 vs. 14.5±6.0 mmHg, P=0.90) but DeltaPP higher in responders than in non-responders (8±7% vs. 5±7%, P=0.46). Fluid responsiveness was equally predicted by DeltaPP (ROC curve area 0.61±0.14), pulmonary artery occluded pressure (0.58±0.14) and right atrial pressure (0.66±0.13) (P=NS). The best cutoff value for DeltaPP was 4% with a sensitivity of 46% and a specificity of 86%.

The preliminary results suggest that DeltaPP is not a better predictor of fluid responsiveness then PAOP or RAP in mechanically ventilated patients when tidal volume is 6-7 ml/kg. If used, a lower critical value may help to predict fluid responsiveness. SVV and PPV are proven influenced by the different airway pressures due to depth of tidal volume and PEEP. The effect of respiratory rate or respiration frequency on SVV and PPV is however unclear. Aim of this study was to evaluate the effect of respiration frequency on SVV and PPV in mechanically ventilated patients.

After obtaining informed consent, 6 (coronary bypass grafting) patients were studied immediately after surgery. Cardiac Output (CO), SVV and PPV were assessed by arterial pulse contour analysis (LiDCO, LiDCO Ltd). All patients were ventilated in pressure controlled mode (settings: FiO2 0.40, tidal volume 8 ml/kg, PEEP 5 cmH2O, frequency 12min-1) and sedated with propofol. In this study SVV and PPV were evaluated with fixed ventilator frequencies of 8, 12 and 16min-1. This protocol was repeated 4 to 5 times (before and after volume loading of 500 ml) in each patient. During the study the mean airway pressure was maintained constant by adjusting inspiration time. 135 collected data points are described in means (SD) and evaluated using ANOVA.

In six patients (female/male ratio 1/5) after coronary bypass grafting, mean age 62(±11.7) years [range 46-74 years], 135 data points by fixed respiratory frequencies could be analysed (29/8, 30/16 and 76/12). All measurements were performed in hemodynamically stable conditions, HR mean 84(±11.4) min-1, MAP 84.7(±9.5)mmHg, CVP 10.2(± 2.5)mmHg and CO 4.7(±0.84) L/min (p for all NS). Mean airway pressure 9.7(±0.81)mbar (Levene statistics, p = 0.605), for resp-f8 9.3(±0,71)mbar, resp-f12 9,8(±0,79)mbar and resp-f16 10.1(±0.79)mbar. On fixed respiratory rates SVV and PPV were unchanged: for SVV (resp-f8) 8.0(±3.3)%, (resp-f12) 7.3(±3.0)%, (resp-f16) 8.2(±4.8)%, p = 0.349, for PPV (resp-f8) 9.4(4.6)%, (resp-f12) 8.6(±3.8)%, (resp-f16) 9.6(±4.8)%, p = 0.499.

In ventilated cardiothoracic surgical patients, SVV and PPV were not influenced by forced changes in respiratory frequencies between 8 and 16min-1. (SVV) has been studied as a dynamic preload marker to predict fluid responsiveness in critically ill patients. Patients undergoing major abdominal surgical procedures with the aid of pneumoperitoneum may have a difficult preload management, due to either a preoperative hypovolemic status or an excessive intraoperative fluid loading to maintain an adequate volume and tissue perfusion. The aim of this study was to use the SVV to optimize the fluid management in patients undergoing major abdominal robot-assisted laparoscopic surgery.

METHODS. 20 patients (ASA score 3-4; mean age 69.3 +/-9.9) were prospectively enrolled. Cardiac index (CI), stroke volume variation (SVV), and central venous saturation (ScVO2) were calculated with the Vigileo system. Gastric carbon dioxide pressure (PgCO2) was measured with a gastric tonometer. Before the induction of anesthesia, 6 ml/kg normal saline solution was administered. Later, colloids were infused whenever a SVV >15% resulted. Hemodynamic variables and PgCO2 were measured before, during, and after the end of surgery. The total amount of intraoperatively administered fluids (IAF) was calculated. Subsequently, the IAF was compared with theoretical IAF using the formula proposed by Miller. Analysis of variance and Student's t-test were applied.

Mean surgery time was 3.6 +/-0.6 hours. CI ranged from 1.8 to 5.5 liters/min/m2. ScVO2 ranged from 62% to 88%. The PgCO2 ranged from 4.1 to 15.2 mmHg. Anova did not show significant variations of CI, ScVO2 and PgCO2. Mean baseline and postoperative SVV% were 17 +/-4.1 and 9.1 +/-4.5, respectively. With respect to preoperative values, anova showed a significant reduction for SVV%. Moreover, at the end of surgery the SVV% resulted less than 15% for each patient. The total amount of fluid was 14.9 +/-2 vs 18.7 +/-2.4 ml/kg per hour (calculated vs theoretical, respectively. P<0.05). No patient showed signs of hypoperfusion. No complication or death occurred.ONCLUSION. The Vigileo system seems to be a reliable tool to provide indications for fluid administration and volume responsiveness. It could be useful especially in major surgical procedures at risk of fluid overfilling. SVV continuously monitored may help physicians to avoid fluid overloading in patients undergoing major abdominal robot-assisted laparoscopic surgery. 

Recently, the preload parameters global enddiastolic volume GEDV and intrathoracic blood volume ITBV measured with transpulmonary thermodilution were convincingly shown to be superior to the historically used central venous pressure 1 . The extravascular lung water EVLW was shown to be a prognostic marker in critically ill patients 2 . However, in our clinical experience, we failed to achieve the proposed normal ranges for GEDV/ITBV indexed to body surface area in a substantial number of patients. As hypothesis, we investigated the dependence of transpulmonary thermodilution parameters on the patient's age.

We retrospectively analyzed the transpulmonary thermodilution data in a series of 128 patients treated on our neurosurgical intensive care unit. Diagnosis was predominantly severe subarachnoid hemorrhage, but included traumatic brain injury and polytrauma, too. ITBVI and GEDVI were measured with the PiCCO ® system (Pulsion Medical Systems AG, Munich, Germany). Measurements were performed with 20cc iced saline injected repeatedly in a central venous line. All data was stored online and pooled for analysis.

Mean patient age was 54.2 (SD 14.4) years. 2587 pooled thermodilution measurement sequences consisting of 9405 single injections were analyzed. Mean GEDVI was 721 (SD 173) ml/m 2 , mean ITBVI was 898 (SD 213) ml/m 2 and mean EVLWI was 8.0 (SD 3.7) ml/kg. Younger patients had lower mean values calculated by linear regression, with an increase of 4.6 ml/m 2 for GEDVI and 5.8 ml/m 2 for ITBVI per patient year. EVLWI was independent of age.CONCLUSION. The thermodilution data from our patient collective contrasts the use of fixed age-independent normal values for GEDVI and ITBVI but not for EVLWI. This data set, however, comprises a neurosurgical patient collective and may not be validly extrapolated to other clinical surroundings.

1. Michard F., et al.: Chest 2003; 124: 1900 -1908 2. Sakka, S., et al.: Chest 2002 122: 2080 -2086 

Thirty mechanically ventilated patients with severe sepsis or septic shock (age 60±15; APACHE-II score 31±8; 18 male) requiring invasive hemodynamic monitoring due to cardiovascular instability were included in a prospective observational trial. The study was performed in a university hospital setting with a 24-bed medical intensive care unit (ICU) and a 14-bed anaesthesiological ICU. Volume-based hemodynamic parameters were assessed using the single-pass thermal-dye transpulmonary dilution technique. Simultaneously, IVC diameter was measured throughout the respiratory cycle by transabdominal ultrasonography.

We found a statistically significant correlation of both inspiratory and expiratory IVC diameter with central venous pressure (p=0.004 and p=0.001), extravascular lung water index (p=0.001, p<0.001), intrathoracic blood volume index (p=0.026, p=0.05), the intrathoracic thermal volume (both p<0.001), and the paO2/FiO2 oxygenation index (p=0.007 and p=0.008, respectively).CONCLUSION. Sonographic determination of IVC diameter is useful in the assessment of volume status in mechanically ventilated septic patients. This approach is rapidly available, non-invasive, inexpensive, easy to learn and applicable in almost any clinical situation without doing harm. IVC sonography may contribute to a faster, more goal directed optimisation of fluid status and may help to identify patients in whom deleterious volume expansion should be avoided. It remains to be elucidated whether this approach influences the outcome of septic patients. 

A severe burn injury is associated with hypermetabolism and catabolism that has been shown to persist for over 12 months post injury. Propranolol has been shown to reduce hypermetabolism during the acute hospital course. The effect of propranolol, a nonselective beta blocker, on respiratory variables in children with severe burns has not been established. Beta-blockade is associated with a known risk of bronchoconstriction in children with hyper-reactive airway disease, but it is not known whether the effects are also seen in severely burned children. The purpose of this study was to determine the effect of propranolol, given during acute hospitalization, on respiratory variables.

Forty-six patients with burns >40% total body surface area (TBSA) were enrolled into the study and randomized to receive propranolol at 1.0 mg/kg/day (n=23) or placebo (n=23). Administration of propranolol was started the day following the first operation and continued for three weeks. Respiratory variables were measured by a Flow Transducer attached to a Bicore CP respiratory monitor. All patients were breathing spontaneously and non-intubated. Study variables included respiratory rate (RR), minute ventilation (MV), tidal volumes (VT), and peak inspiratory/expiratory flow rates (PIFR/PEFR). Baseline measurements were taken at rest before the drug or placebo was initiated. Follow-up measurements were performed at the end of the study period. Data were analyzed using paired t-test within groups and un-paired t-test between groups. Data are reported as mean ± SD. Significance was accepted at p<0.05.

The mean age in both groups was 10 ± 4 years. As expected, heart rate was reduced by approximately 20% in the propranolol group compared to placebo (p<0.05). There was a significant increase in PEFR from 0.57 ± 0.26 to 0.74 ± 0.36 L/s in the propranolol treated group (p=0.03). In contrast, neither placebo nor propranolol significantly affected RR, VT, VE or PIFR.

Results indicate that short term administration of propranolol showed significant effects on PEFR suggesting increased pulmonary conductance. Further studies on the effects of propranolol on gas exchange and lung compliance are needed. GRANT ACKNOWLEDGEMENT. Funded by NIH grants P50-GM06338 and KO1-HL70451

A. Storesund* 1 , E. Wallestad 1 , L. Rygh 2 1 Postoperative Section, Surgical Department, 2 Surgical Department, Haukeland University Hospital, Bergen, Norway

International studies point out that to work with agitated children, described as restless and disorientated are particularly stressful for the child, parents and caregiver. This project is based on the assumption by nurses in the post anaesthetic unit (PAU) that there was a noticeable post anaesthetic agitation difference between the children who received long-term opioids initially and in the end of the operation (refill, A) compared to those who only got long-term opioids in the beginning of the operation (no refill, B). The main purpose of this project was to examine whether there were any difference in postsurgical agitation between the refill and no refill group. Further, this project seeks to uncover if there are any factors that can be improved per-and postoperative for these patients.

We observed 35 post anaesthetic children, Lip-(n=8), Cleft-(n=0), and Palateclosure (n=12), Adeno-(n=5), & Adeno-tonsillectomy (n=10). These children were recruited using a convenience sampling strategy at the PAU at Haukeland University Hospital, Norway, over a 19 week period in 2006-07. A pilot-tested fixed cross sectional designed questionnaire was utilised by the nurse responsible for each patient. Several statistical tests by the use of SPSS made it possible to analyse and answer the research question: Are children who only get long-term opioids in the initial anaesthetic phase (B) of the operation more agitated than those who where also given a refill of long-term opioids (A)?

We found that 20/28 got refills of long-term opioids (A), 8/28 did not get refills (B), 20% were recorded as missing values. T-test result = 0,735 is greater than 0,05, hence there is no statistically significant difference between the two groups. Levene's test tells us that the two variances are not significantly different (Levene's test sig=0,167). There were no significant relationships between the parameters recorded. However, there was a tendency that more preoperative anxious children got refills (5/5) compared to non-anxious children (6/10) (Fisher's exact test p=0,15). The latter results may conceal the agitation-scores in the two groups; refill and non-refill-group. This possible bias may have been eliminated if the patients had been randomized to either refill or non-refill.

The present study confirms previous observations by others indicating no singular factors can explain why some children experience agitations and others do not. Analysis of the parameters studied did not discover any statistical significant relationships. Thus, how to minimise the cohort of children who experience post anaesthetic agitation still remains a recurrent challenge. 

Pulmonary hypoplasia with severe cardiorespiratory dysfunction is often the leading cause of death in neonates with congenital renal disease and oligo-anhydramnios. Aim of the study was to determine whether iNO is effective to improve respiratory function in these critically ill neonates.

We retrospectively reviewed the charts of all newborns who were admitted between February 1996 and September 2002 with the diagnosis of oligo-anhydramnios of renal origin. During this period all patients were treated according to a standardised algorithm. They were intubated either if post CPR or if FiO2 had to be increased above 0.4. 100mg/kg of bovine surfactant were applied for improvement of ventilation. Pre-and postductal oxygen saturation were measured simultaneously with target values of 85-93%. If FiO2 remained above 0.5 a transthoracic echocardiography was performed. The presence of a ductal or atrial right to left shunt or a difference in oxygen saturation between the pre-and postductal measurements of >15% led to the diagnosis of pulmonary hypertension and to the initiation of iNO therapy. Further, iNO was applied as a rescue therapy if oxygen saturation remained below 80% despite a FiO2 of 1.0 and optimization of ventilator settings and therapy with catecholamines. All patients had informed parental consent.

The patient population (n=20) included 7 children receiving iNO of whom 5 suffered from obstructive uropathy and two had polycystic kidneys, whereas 13 patients did not receive iNO treatment. In this group there were 11 children with obstructive uropathy and 2 born with polycystic kidneys. All data are presented as median (range). We concentrated on the group receiving iNO. In this group mortality was 57.1%. Therapy was started at an age of 10.9 (1-28) hrs. Initial dose of iNO was 14.9 (4-40) ppm with peak dose of 20.3 (5-57) ppm. INO led to a decrease of oxygenation index (OI) from 28.4 (3.4-44.4) to16.6 (3.2-70.7). Five children suffered from obstructive uropathy. Three of them had a favourable long-term outcome, one child died immediately, whereas one child was initially stabilized but finally succumbed to its underlying disease. Two children demonstrated genetically determined pulmonary hypoplasia due to the presence of polycystic kidneys. Both children died within the first three days despite iNO treatment.

Children with obstructive uropathy and severely impaired oxygenation seem to benefit from iNO therapy. Patients suffering from a hereditary renal and pulmonary hypoplasia did not respond favourably to iNO therapy and had a fatal outcome.

A. Khaldi*, K. Menif, A. Bouziri, A. Hamdi, S. Belhadj, N. Ben Jaballah Pediatric Intensive Crae Unit, Children's Hospital, Tunis, Tunisia

The use of high-frequency oscillatory ventilation (HFOV) and iNO resulted in a decline in the need for extracorporeal membrane oxygenation (ECMO) in near-term and term neonates with persistent pulmonary hypertension (PPHN). Association of HFOV and iNO is actually an accepted treatment modality even in non-ECMO centers. However, because not all neonates respond to HFOV + iNO, identification of factors related to a poor response is very important for prognosis and for early transfer to ECMO canters if possible. The objective of this study was to identify the risk factors predicting poor shortly outcome in near-term and term neonates with PPHN treated with HFOV and iNO in a tertiary care pediatric intensive care unit in a university hospital.

We conducted a prospective clinical study including all neonates with gestational age ≥ 34 weeks with echocardiographic signs of PPHN. Patients with pulmonary hypoplasia or congenital diaphragmatic hernia were excluded . Patients were ventilated with conventional mechanical ventilation (CMV) with iNO (10-20 ppm). HFOV were instituted if patient required, on conventional ventilation (CMV)+iNO, a fraction of inspired oxygen (FIO2) 0.5, and a mean airway pressure > 10 cm H2O to maintain adequate oxygenation or a peak inspiratory pressure > 24 cm H2O to maintain tidal volume between 5 and 7 mL/kg of body weight. HFOV were used in association with iNO in seventy infants (gestational age, 37 ± 1,9 weeks), after a mean duration of CMV of 4 ± 7 hours. Arterial blood gases, oxygenation index (OI), and alveolararterial difference in partial pressure of oxygen (P[A -a]O2) were recorded prospectively before and during HFOV.

There were a rapid and sustained decreases in mean airway pressure (MAP), OI, and P[A -a]O2 during HFOV (p ≤ 0.01). This improvement, along with decreased need for oxygen, was sustained through the subsequent course of HFOV. Sixty-six infants (93%) were weaned successfully from HFOV. Five infants (7%) were classified as meeting treatment failure and died from their underlying disease. Treatment failure was associated with lack of improvement in P[A -a]O2 and OI at 4 hour of HFOV (p < 0.01) and the presence of intractable shock requiring epinephrine or norepinephrine (p=0,01).

In near-term and term neonates with PPHN, the association of HFOV and iNO lead to a rapid and sustained improvement in gas exchange in the most cases. The magnitude of improvement of OI and P[A -a]O2 at 4 hrs can predict outcome early.

Early burn sepsis is notable for the complexity of diagnostics, malignant course and high lethality. The problem remains actual for the children who got a severe burn trauma (more than 20% body surface area). PURPOSE to define procalcitonin test (PCT) effectiveness for early sepsis diagnostics for children with thermal trauma.

During the period of time from January 2004 up to April 2007 there were 190 children in our clinic with extensive burns from 20%up to 70% body surface area (BSA) at the age from 6 months to 14 years old. 39 patients at the age from 8 months to 13 years old with the burns from 20 % to 70% BSA were included in our research. All the children got surgery in shortest time after trauma (tangential excision with authodermoplastics), antibacterial, and infusion therapy. From the moment of registration in ICU all the patients, who were suspected to have sepsis, simultaneously with traditional examinations (blood analysis, bacteriological investigation) were taken PCT analysis with the help of "PCT-express test" (BRAHMS, Germany).

. 17 patients (44,7%) were diagnosed sepsis, 3 children died. These patients PCT level was from 2 to 10 ng/ml; together with this all the patients had increasing quantity of leucocytes, acceleration the level of C-reactive protein, fever. 22(56%) patients had no sepsis, so PCT figures fluctuated in the bounds of 0,5 ng/ml. Among these patients traditional markers of inflammation were increased. No trustworthy difference is found as for the level of leucocytes and C-reactive protein figures between the patients without infectious complications and with sepsis. Only with the help of PCT the beginning of sepsis and SIRS manifestation can be differentiated.CONCLUSION. 1. Burn trauma itself is not the reason for PCT increase. PCT level increases in cases of burn injuries as the sign of infectious complications joining. 2. With the help of traditional sepsis markers it is difficult to differentiate SIRS manifestation and first stages of infectional complications in case of thermal trauma. 3. In cases of severe burns PCT test is a highly sensitive method of sepsis early stages diagnostics. 4. Surgery treatment at early stages after trauma allows to avoid development of severe sepsis.

H. Knoester* 1 , M. B. Bronner 2 , A. P. Bos 1 , M. A. Grootenhuis 2 1 Pediatric Intensive Care Unit, 2 Psychosocial Department, Emma Children's Hospital, AMC, Amsterdam, Netherlands INTRODUCTION. Improved survival in children with critical illnesses has led to new disease patterns due to long-term complications and effects of the original illness and its treatment. As a consequence, Health Related Quality of Life (HRQoL) has become an important outcome measure in Pediatric Intensive Care Unit (PICU) survivors. Little is known about HRQoL in PICU survivors,. HRQoL evaluation could contribute to improvement of support after discharge. The purpose of this study was to assess HRQoL in PICU survivors.

October 2005 all parents of children, acutely admitted to our PICU were invited to complete HRQoL questionnaires, 3 and 9 months after discharge. HRQoL in children from 1-6 years of age was evaluated with a Dutch validated questionnaire, the TNO-AZL Preschool Children Quality of Life Questionnaire (TAPQOL). The TAPQOL covers 12 domains of HRQoL; norm data from the general Dutch population are available. Data analyses was done by non-parametric testing (patients versus norm group) and by calculating effect sizes (difference in mean scores between the patients and the norm group divided by the standard deviation of the scores in the norm group). Effect sizes give an indication of changes in HRQoL in comparison with the norm group.

. 34 of 75 (45.3%) eligible patients were evaluated. Statistically significant differences with the norm group were found on 2 domains, 3 and 9 months after discharge (more lung problems and worse liveliness) and on 1 domain 3 months after discharge (better appetite). Moderate (0.5) and large (0.8) effect sizes were found on five respectively four domains 3 and 9 months after discharge: indicating worse HRQoL on lung problems, sleeping problems, motor functioning, anxiety, positve mood and liveliness; and indicating better HRQoL on problem behaviour. No statistically significant changes over time were found for all domains 3 and 9 months after discharge.

Our results indicate that HRQoL in young PICU survivors is decreased in some domains of physical and emotional functioning. These problems do not diminish over time. Positive evaluation by parents regarding appetite and problem behavior could be influenced by response shift (changing of internal standards and values due to confrontation with a life-threatening disease). More research is necessary because of the small study group and to determine the influence of risk factors such as length of stay, age of the child at admission, severity of illness and physical sequelae of the disease and its treatment on HRQoL. HRQoL evaluation can be a useful tool as part of screening after PICU survival to determine the necessity for follow-up care. 

Coarctation of the aorta is not an uncommon congenital heart defect. One of the possible postoperative complications is the so-called postcoarctectomy syndrome (mesenteric arteritis). The purpose of the present study is to assess the changes in gut flow through the dual sugar permeability test.

Five patients have been included in the study until now. Median age 1 month (0.3 -24) and median weight 4.1 Kg (3 -15). Premedication and anaesthesia was the same for all the patients. The test solution contains 3-O-methyl-D-glucose, D-xylose, L-rhamnose and lactulose. Patients received 2 ml/kg of the test solution after induction of anaesthesia, at 8 and 24 hours after the initial dose. Urine production is measured during a three-hour period after each instillation. The sugar content is analysed by capillary gas chromatography (Normal values L/R = 0.05, 3OMDG and Xylose 10-30%). 

A. Monsel 1 , P. Durand 2 , V. Haas 2 , C. Beaujard 3 , P. Rouleau 3 , S. El Aouadi 3 , D. Benhamou 3 , K. Asehnoune* 4 1 anesthesie reanimation, 2 reanimation pediatrique, 3 anesthesie réanimation, hopital de bicetre, bicetre, 4 anesthesie réanimation, CHU hotel-dieu, Nantes, France

Pediatric epidural anesthesia (EA) is considered to be without hemodynamic impairment in children. However, when compared with information relating to adults, little is known about the hemodynamic effects of epidural anesthesia on the cardiac output (CO) in infants.

Using transesophageal Doppler (TED) monitoring of CO, we prospectively studied 14 infants < 10 kg who were scheduled for abdominal surgery. During sevoflurane general anesthesia, TED monitoring of CO was performed before and after lumbar EA with 0.75 mL/Kg of 0.25 % bupivacaine and 1:200,000 adrenaline. CO, arterial blood pressure, and heart rate were measured before and 5, 15, and 20 minutes after performance of EA.

In patients anesthetized with sevoflurane and sufentanil, EA resulted in an increase in stroke volume by 29% (p<0.0001) and a decrease in heart rate by 13% (p<0.0001). EA also induced a significant decrease in systolic, diastolic, mean arterial blood pressure and systemic vascular resistances by 11%, 18%, 15%, and 25% respectively. Conversely, CO remained unchanged. The increase in SV observed is probably explained by optimization of afterload due to the sympathetic blockade induced by EA.

These results confirm that EA provides hemodynamic stability in infants weighing < 10 kg and support the use of EA in this pediatric population. 

Bleeding is the most frequent complication during extracorporeal life support (ECLS) after pediatric cardiac surgery. We would like to present our experience with ECLS and recirculation blood saving, volume auto-regulation system using the law of connected vessels based on converted CPB set in infants after cardiac surgery with significant bleeding.

Since 2003 to 2005 26 ECLS in the postoperative period was performed (1,9 % of all cardiac operations in this period). The significant bleeding (>5ml/kg/h) was noted in 20 pts. In most recent 11 pts the volume recirculation system was implemented, whereas in 9 previous patients blood was sucked out the circuit. The retrospective analysis of data was carried out.

There were 16 infants with single ventricle anatomy and 10 with two-ventricle anatomy. There were no significant differences with respect to age, weight and prevalence of single ventricle anatomy between groups. The indication for ECLS was cardiac arrest in 14, low cardiac output in 4, hypoxemia in 6 and sepsis in 2 patients. The overall mortality rate was 46%. The mortality did not differ significantly between groups (36,3% versus 55% in non-recirculation group; p=0,35). There was significantly lower number of blood products transfusions(p<0,05), lower number of surgical explorations(p<0,05) lower mean lactate level 8 hours after ECLS institution p(<0,05) and shorter ECLS duration (p<0,05) in the recirculation group.

The system of blood recirculation in children with bleeding on ECLS is simple, highly effective in stabilization of the haemodynamics and no-cost consuming. It can reduce necessity of chest exploration, blood product transfusions and duration of support.

T. Tunc* 1 , T. Topal 2 , M. Kul 1 ,Ö.Öngürü 3 , A. Korkmaz 2 , S.Öter 2 1 Neonataloji Bilim Dali, 2 Fizyoloji Anabilim Dali, 3 Patoloji Anabilim Dali, Gülhane Askeri Tip Akademisi, Ankara, Turkey

Necrotizing Enterocolitis (NEC) is the most common gastrointestinal emergency in the premature infant. The major risk factors in NEC include prematurity, hypoxia, enteral feeding, and bacterial colonization. These factors predispose at-risk infants to an exaggerated intestinal inflammatory response leading to ischemic bowel necrosis. Experimentally induced ischemia and reperfusion (I/R) of the intestine is a model which can be appropriately used to imitate NEC. N-acetylcysteine (NAC), erdosteine (ERD) and alpha-lipoic acid (ALA) are well-known antioxidants with similar structural properties. In the present study, the effectiveness of these three sulfur-based antioxidants against intestinal I/R-injury was evaluated.METHODS. 40 one month old male Spraque-Dawley rats were randomly divided into five groups (n = 8 for each): I/R (control), I/R+NAC, I/R+ERD, I/R+ALA and sham-operated group without I/R. Animals were operated at a temperature of 24 o C under ketamine anesthesia. Ischemia was provided by occluding the superior mesenteric artery via a microvascular clamp. Collateral vessels of the small intestine were ligated to prevent collateral circulation. 30 min of ischemia was followed by 30 min of reperfusion. NAC (100 mg/kg/day, i.p.) was administered first 30 min before operation and followed once daily for 5 days. ERD (10 mg/kg/day, oral gavage) administration was begun 2 days before operation and continued 5 daily doses. ALA (35 mg/kg/day, i.p.) was injected only one time 24 h before operation. At day 5 after operation the ileum was resected and the rats were sacrificed. Protein oxidation (carbonyl content, PCO), lipid peroxidation (malondialdehyde, MDA), superoxide dismutase (SOD) and glutathione peroxidase (GSH-Px) were measured in the ileal tissue.

Oxidative and antioxidant parameters of resected ileal segment (Mean ± SD) Groups

0 

As a clinically relevant model to NEC, our experimental I/R protocol resulted with marked rise in oxidative stress levels and fall of antioxidant enzymes activities. These changes were ameliorated with the antioxidants used. Among all, ALA presented the strongest and NAC the weakest effect. This outcome promises beneficial usage of these sulfurbased antioxidants against oxidative stress which plays an important role in NEC pathogenesis. 

A. Khaldi* 1 , K. Menif 2 , A. Bouziri 2 , A. Hamdi 2 , S. Belhadj 2 , N. Ben Jaballah 2 1 Pediatric Intensive Crae Unit, Children's Hospital, 2 Pediatric Intensive Crae Unit, Children's Hospital, Tunis, Tunisia

High-frequency oscillatory ventilation (HFOV) may significantly improve oxygenation and outcome in newborns with respiratory dysfunction and beyond the neonatal period in patients with a variety of diffuse alveolar diseases. In small airway disease like respiratory syncytial virus (VRS) bronchiolitis, HFOV is considered potentially hazardous because of the risk of air trapping. However, a few studies had reported utility of HFOV in children with acute hypoxemic or hypercapnic respiratory failure caused by VRS and failing optimal conventional mechanical ventilation (CMV). The objective of the study is to evaluate the effectiveness and safety of HFOV in pediatric patients with acute respiratory failure due to RSV and failing CMV.

We conducted, over 6-year period (October 2000 to October 2006), a prospective clinical study in a tertiary care pediatric intensive care unit. Fourteen (14) patients (ages 12 to 73 days) with acute respiratory failure due to RSV bronchiolitis and failing optimal CMV were included. Passage to HFOV was indicated for severe hypoxemia in 11 patients (median alveolar-arterial oxygen difference [P(A-a)O2]: 564 [465-624] mmHg, median oxygenation index [IO]: 28 [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] ) and for severe hypercarbia in 3 patients (median pH: 7,20 [7,11-7,24] , median PaCO2: 98 [90-120] mmHg). HFOV was instituted after a median length of CMV of 32 (4-48) hours. Ventilator settings, arterial blood gases, OI and P(A-a)O2 was recorded before HFOV (H0) and at a predetermined intervals during HFOV and compared using the one-Friedman rank-sum procedure and a two-tailed Wilcoxon matched-pairs test.

After starting HFOV, a distinct decrease in FiO2 at 4 hrs that continued to 24 hrs (p<0,05). In all patients, there were significant decreases in OI and P(A-a)O2 at 4 hrs, that were sustained up to 20 hrs (p<0,04). Target ventilation was achieved in all cases and PaCO2 significantly decreases after 1 hr of HFOV (p=0,01) and remained within the target range thereafter (40-70 mmHg). The median maximum pressure amplitude used on HFOV was 45 (35-60) cm H2O and the median maximal PAW was 23 (18-26) cm H2O. No significant complications associated with HFOV were observed. Twelve patients (86%) survived to hospital discharge without supplementary oxygen. Tow patients (14%) died from septic shock.

In pediatric patients with either hypoxemic or hypercapnic acute respiratory failure due to RSV bronchiolitis, HFOV can be used successfully and safely if conventional ventilation fails to improve gas exchange. However, randomized controlled trials are needed to identify its benefits over conventional modes of mechanical ventilation. and are influenced by numerous factors like patient's disease severity, policies of the treating unit, religious and cultural traits, education and awareness of the patient's family, financial status of the family and legal provisions. Majority of published studies on EOL reflect either European or American ethos; that is either physician's paternalistic approach about the patient or patient's autonomy and self determination,(1,2) about this sensitive process. Studies on EOL which reflect the influence and pivotal role of closely knit Indian family on EOL decision making are scant. We retrospectively analysed the EOL decisions taken by the family in our ICU as majority of the patients which merit EOL care were not in a condition of decision making.

Setting-50 bedded multidisciplinary ICU of a 400 bedded tertiary care teaching hospital in Pune in India. Case papers of all ICU admissions during one year i.e. 1st January to 31st December 2006 where EOL decision was documented, were reviewed. Data collected included demographics, underlying disease process, duration of aggressive treatment till EOL consent, duration between EOL decision and death, consenting person's relation with the patient, organ failure & level of life sustaining supports at decision and mode of payment of the treatment.

During the study period 524 patients died in our ICU of which EOL decision and consent was explicitly documented in 95 cases which constitute study population. Average age of the patient was 63 years (range 17 to 91), average duration of active treatment till EOL consent was 83.35 hours(range1 to 960),average duration between consent and death was 29.03 hours(range 1to 168). 92.7% consents were signed by close relatives( son/daughter, brother/sister, spouse, father/mother) and 7.3% were by other relatives( cousins, son in law/daughter in law). At the time of EOL decision 90.7% patients were having Glassgow coma scale 8 and below, 69.4 % patients were on mechanical ventilation, 56.84 % were on vasopressors and 6.3% were needing renal replacement. Metastatic disease (31.5%) and traumatic or vascular brain injury(22.1%) were the commonest causes of death. Only 24.05% patients had medical insurance or Employer assistance as a mode of payment for the treatment and in 75.9% cases family members were the payers.

Withholding of non beneficial life sustaining therapies as EOL process was practised in 18.12% of the total ICU deaths. All 95 (100%) EOL decisions as well as directive requests and consents were signed by patients' relatives, reflecting the importance of close family ties in Indian EOL practices. 

Our objective was to study frequencies of withholding and withdrawing treatment and time until death in a Dutch university hospital ICU.

Between October 2006 and February 2007 we collected data of all patients that died. Data were collected from patient files and during interviews with the doctors and nurses who were responsible for the patient at the time of death. We analyzed which treatments were withheld or withdrawn and calculated the time until death following withholding or withdrawal.

Preliminary results show that of 471 admissions, 74 patients died (16%). Nonsurvivor's (median age 64 years [range 15-84]) median length of stay was 3 days (range 30 minutes -5 months). In 57 patients (77%) treatment was withdrawn and in 4 patients (5%) treatment was withheld but not withdrawn. Of all patients 71 (96%) were mechanically ventilated of which 6 (8%) were weaned and extubated before death. In 4 of these patients it was decided not to intubate again and 2 other patients not to intubate at all (median time until death: 24 hours). In 45 (61%) ventilator-dependent patients mechanical ventilation was withdrawn; 35 (47%) were extubated. The median time until death after ventilator withdrawal was 30 minutes. When patients were also extubated, it was 22 minutes (P=0,16 [Mann-Whitney test]). In 6 patients mechanical ventilation was not withdrawn, but FiO2 was decreased to 0.21 (median time until death 35 minutes). In 25 patients (34%) inotropic medication was withdrawn (median time until death 30 minutes). In 20 cases, the withdrawal of inotropic medication was combined with the withdrawal of mechanical ventilation. In 8 patients (11%) it was decided not to increase inotropic support (median time until death 6:45 hours). In 45 patients (61%) the decision was made not to resuscitate in case of cardiac arrest. Median time of this decision before death was 26 hours.

In the patients that died treatment was withdrawn in the vast majority of patients. Withdrawal of mechanical ventilation and/or withdrawal of inotropic support were most often used. A considerable number of patients died within 30 minutes following withdrawal of therapy.

R. Veiga* 1 , G. Silva 1 , G. Campello 1 , C. Dias 2 , C. Granja 1 1 Intensive Care Department, Hospital Pedro Hispano, Matosinhos, 2 Biostatistics and Medical Informatics, Faculty of Medicine, Porto, Portugal

The high mortality of critically ill patients underscores the need for ICU teams to recognize end-of-life care as an integral component of critical care. Besides survival, the success of intensive care should also include the quality of lives preserved and the quality of dying. The aim of this study was to evaluate the incidence and type of end-of-life decisions in critical patients that died in an ICU.

Retrospective analysis of all patients that died in the ICU in the period of 1 January to 31 December 2006 and evaluated the following variables: demographic characteristics (age, gender); co-morbidities: (heart failure, chronic obstructive pulmonary disease (COPD), diabetes mellitus, neoplasia, chronic renal disease, HIV/AIDS, alcoholism); reason for admission; SAPS II; length of ICU stay (ICU LOS) and type of end-of-life decisions. Three concepts were defined in order to classify the end-of-life decisions: Comfort Care: a change from curative therapy to comfort care therapy; Limited therapy: maintenance of curative therapy but without escalating it (e.g. not raising rate of vasopressor agents, no renal substitution); Without previous end-of-life decisions: when no attitudes toward end-of-life care were considered. Given the diminished number of patients in the Without previous end-of-life decisions group we decided to evaluate them apart from the other two groups.RESULTS. Two-hundred and twenty seven patients were admitted in the ICU and 62 of them died (27%). Reason for admission in those who died was septic shock/ severe sepsis (43%), post-cardiac arrest (22%); cardiogenic shock (9%); acute respiratory distress syndrome (7%). The most common co-morbidity was alcoholism (19%), followed by diabetes mellitus (14%), neoplasia (13%), heart failure (11%) and COPD (11%). Forty seven patients (76%) died after Comfort care decision, eleven patients (18%) after Limited therapy decision and four (6%) patients died Without previous end-of-life decisions. Comparing the groups Comfort Care and Limited therapy we found significant differences in the following variables: hemorrhagic shock at admission (2% vs. 27%) (p=0.04); SAPS II (55 vs. 85) (p= 0.008); ICU LOS (4.9 days vs. 0.27 days) (p<0.001).

Patients in the Limited Therapy group had more admissions with hemorrhagic shock, a higher severity score and stayed less time in the ICU. This analysis suggests that end-of-life decisions in this group express their higher severity. Patients of the Comfort Care group presented less severity and stayed longer in the ICU. Their shift of curative therapy to one designated to provide Comfort care reflects an absence of a clinically favorable response. The low percentage of patients Without previous end-of-life decisions is consistent with previous reports and should be seen as a positive issue. 

Non invasive positive pressure ventilation (NIPPV) is widely accepted as an initial approach to providing ventilatory support to many patients with acute respiratory failure (ARF). Palliative approaches focused on the quality of life and comfort; represent a challenge for family's physicians and the patients. NIPPV is an attractive option to treat acute respiratory failure in end stage patients when the failure is irreversible and it is a final outcome of the primary disease. The approach to providing ventilatory support to patients with ARF, to relieve them from the sensation of dying suffocate without intubating them because they don't wish it either, is very challenging.

After institutional approval and patients consent, we conducted a prospective observational study of patients that fulfilled the criteria.12 cases received NIPPV (10 with end stage cancer and 2 with pulmonary fibrosis). When NIPPV was ordered we recorded: respiratory rate, heart rate, arterial blood pressure, neurological status and arterial blood gases, before NIPPV initiation (baseline data) and then 1st, 4th and 8th hour. At the time of initiation of NIPPV, all patients were alert and cooperative with NIPPV. Analgesia and/or sedation were used when it was necessary. PaO2, pCO2 and pH measures were analyzed using statistical methods. Percentage changes from baseline (pre-NIPPV) of these measures were used as dependent variables. (Mean value of 2 measurements at different time points was used). Dependent variables (percentage of PaO2, PCO2 and pH) were regressed on time, for each patient.

In all cases the results were statistically significant, with p-values ranging from a low of 0.0002 to a high of 0.0185. For all patients, the regression coefficient for the percentage change was positive; indicating that the percentage change was increasing with time. We can remark that PaO2 increases over time, PCO2 and pH p values > 0.05.

We believe that NIPPV via helmet CPAP is a means of potentially ensuring the highest quality of end-of-life care. NIPPV can be applied for palliative care, and it might be used to keep patients whom developed acute respiratory failure comfortable before the inevitable. 

Decisions regarding the resuscitation status of patients are among the most difficult facing healthcare professionals, patients and families. These groups often need to discuss decisions regarding resuscitation yet their understanding and expectations can differ greatly. This study sought to determine the knowledge and beliefs of doctors, nurses and the general public regarding resuscitation decisions.METHODS. An observational study was designed. Three study groups (doctors, nurses and general public) were interviewed using a face-to-face interview by a single interviewer and questionnaires completed. Questions examined opinion, factual knowledge and knowledge of the ethics surrounding hospital resuscitation attempts.

. 30 doctors, 25 nurses and 30 general public were randomly selected. 70% doctors, 24% nurses and 0% of public correctly estimated survival to discharge following in-hospital resuscitation attempt. The remainder overestimated survival. 37.9% of doctors and 76% of nurses consider resuscitation decisions to be made too infrequently. Deficiencies were identified in doctor and nurse knowledge of the ethics governing resuscitation decisions and public opinion was found to conflict with ethical guidelines. Public understanding of the nature of cardiopulmonary arrests and resuscitation attempts, and of the implications of a DNAR order is poor. 58.3% of public report television medical dramas as their primary source of information on such matters.

Knowledge regarding resuscitation principles, outcomes and ethics is poor among both healthcare staff and the general public. These knowledge differences may not be appreciated or addressed in discussions regarding resuscitation and this reduces the likelihood of meaningful discussion and acceptable decisions. There is a need for educational initiatives to address these deficiencies. Public apprehension surrounding this subject needs to be identified and corrected during discussions and this could be facilitated with a patient information leaflet. [1] Poor communication during this process may lead to unnecessary anger and a delay in the grieving process that could linger for many years to come. Giving the family the option to be present during resuscitation offers a more compassionate and family-centred approach to this crisis. This option of Family presence however is frequently met with resistance and uncertainty by health care workers who may view the family's presence as increasing their risk of making a mistake or worse, being sued. A study in the UK estimated that out of one-hundred-and-sixtytwo UK Emergency Departments Family Witnessed Resuscitation was allowed by 79% for an adult patient and 93% for a child. [2] Another US study also found that amongst patients in emergency departments, 72% preferred to have their family present during resuscitation.

[3]

A survey was conducted amongst the doctors, nurses and paramedics who work in two UK EDs to assess their attitudes and beliefs. Experience, life support training, years in practice, consent issues, ethical factors and concerns regarding medico legal implications were sought for. A 5-point Likert Scale was used and mean scores analysed using Microsoft Excel.

. 129 staff were surveyed.34% of doctors, 29% of nurses and 35% of paramedics believed in the concept in trauma FWR. In cardiac arrest patients, 55% were in favour of it, 28% opposed to it and 17% undecided. 62% of staff believed that litigation was possible with family witnessed resuscitation. 83% of respondents thought that critical incident de-briefing would be of benefit to assist staff dealing with stress. Fewer doctors believed in cardiac FWR compared to nurses (p=0.004) and paramedics (p=0.006). In trauma, difference was non-significant.

As health care professionals caring for families in the Emergency departments, we need to recognize the need for compassionate Family-centered care. With a well trained and motivated team equipped with effective, well thought out guidelines, there is considerable benefit for family members and staff in this difficult situation. 

Thorough information about the events that are going to take place in the ICU after an elective procedure might facilitate the awakening process and weaning from the ventilator, mitigating patient's anxiety and increasing their comfort. The aim of this study was to analyze the impact of preoperative information on the patient's perceptions and reactions to the usual inconveniences, such as orotracheal tube (OTT), associated with the first postoperative hours in the ICU.

Prospective, cohort study with a group of cases (A) and a control group (B). Duration: two months. Inclusion criteria: all patients undergoing elective cardiac surgery. There were no exclusion criteria. Setting: Cardiac surgical ICU of a tertiary hospital. The survey was made in the first 24 hours. The study was blinded for the doctors in charge of the patients.

The characteristics of both groups are presented as A/B with the p value into brackets. The quantitative variables are shown with the mean value and the qualitative variables as a percentage. The number of patients included was 86: 47 cases (A) and 39 controls (B). Age: 62,9/61,3 years (0,4); men: 64/74%(0,3); time receiving sedative drugs: 4,12/4,05 hours (0,7); total hours with OTT: 6,53/8,35 (0,01); hours with OTT after stopping sedation: 2,5/4,29 (0,004). The first patient's perceptions were: discomfort related to OTT in 10,6/38,5% (0,002); surgical pain in 29,8/25,6% (0,6); thirst in 19,15/12,8% (0,4); welfare or calm in 19,15/2,56% (0,01), and nothing in 12,7/17,9% (0,5). Additional sedatives were required in 6,38/20,51% (0,05). Information was considered very useful in 97,9%.

Patients valued very positively the provided information. In addition, this information had a significant impact on the tolerance to the OTT, requirement of additional sedatives, and in the sense of welfare. There were not differences in the time under sedative drugs or in the perception of thirst or pain. 

A multiparameter questionnaire was sent to 21 ICU. Each questionnaire comprised 24 informational topics groupe into 6 categories (table). One relative per patient was asked to quote (yes/no) within 4 days after admission, each item, i.e. if he would like to find information on that item in an IB. If "no" was quoted, he was asked to say why (closed answers). Demographic data on patient and relatives were correlated to the scores (nbre of "yes"), in each item category (factor analysis with varimax rotation followed by stepwise multiple linear regression).

. 246 questionnaires were analyzed (patients: age 60 ± 18 year, SAPS2: 48 ± 19, SOFA: 7 ± 7). Table: % of positive response for each item ("would you like information on this topic in an ICU booklet?") grouped into categories. "No" answers were mostly explained by "I trust the team to manage information about this" (median: 61%, range: 16-81). Mulitvariate analysis showed that demographics data describing patient condition (age, SAPS2, chronic disease) correlated (p<0.05) with "yes" score of the items comprized in "ICU rules" (table) but not with other items grouped in other information categories. CONCLUSION. Interestingly, as a whole, most items were highly wished in a booklet, suggesting that 50-70% of relatives express a plea for transparency in face of "difficult ICU issues", without taboo. Only the "yes score" to "ICU rules" items correlated with patient status whereas items from other topics did not. This sounds, as relatives visiting the most severe patients may consider visiting rules as crucial. Other items did not correlate to profiles, and may thereby be considered as societal standard requirements in terms of information. 

In 10/2003, our 15-bed medical ICU signed a convention with the ASP Iroise Association defining HV's role and presence. The Association, a member of a national network of HV associations, works with our university hospital. Four HVs took alternate turns in the ICU one afternoon per week. HV were free to meet any conscious patient or any family member who wished so; ICU staff also asked them to meet patients or families who seemed particularly distressed. HV wrote a brief commentary in a special transmission logbook which could be consulted by the staff and gave feedback about their visits whenever needed. Patients (Pts)and families (Fam)who met an HV were sent a questionnaire either in 01/2005 or in 08/2006.

1258 Pts were admitted during the period of study: the HV met 128 Pts (10,2%) and 209 families (16,6%). 56 people answered the questionnaire (23,6%): 12 Pts and 44 Fam:27 spouse,7 parents,1 sister,3 children(6 no answer). 

Ethics consultation has been introduced into the practice of medicine during the last decades as a way to help physicians and nurses come to a decision about a medical treatment where value-laden conflicts are involved. The primary goal is helping to identify, analyze, and resolve ethical problems. The aim of this study was to evaluate ethics consultation in a Dutch university hospital intensive care.

Intensivists, residents, fellows and nurses can consult a clinical ethicist specialized in intensive care for advice in value-laden situations. We evaluate ethics consultation on our ICU between 1 January 2006 and 1 April 2007.

The clinical ethicist was consulted 61 times. In 29/61 cases (48%) advice was asked before withdrawal of life-sustaining therapy. In this category 15/29 (52%) cases concerned palliative care. In 25/29 cases (86%) the independent advice was in confirmation with the physician's view. In 11/61 cases (18%) advice was sought in cases were there was doubts to proceed with intensive care therapy. In four cases relatives wanted to withdraw therapy, where the intensivist did not consider this as futile. In 9/11 cases (81%) the advice was in accordance with the treatment plan. In 6 cases (10%) questions about information asked by non-relatives. All advises were followed. 3 cases concerned triage, 2 cases withholding therapy, 4 brain death declaration, 2 a deadly iatrogenic complication and in 2 patients a question concerning emergency research. In 7 (11%) cases a lawyer specialized in health care was consulted. In the 29 cases about 'withdrawal of therapy', the advise could be given within 30 minutes in 72% of the cases.

Ethical advise by a clinical ethicist specialized in intensive care can be additional, affirmative and reassuring, and improves quality of care. In most cases advice could be given immediately. . Deferred consent has been proposed as a surrogate for a priori subject or proxy consent. The aim of this report is to evaluate the practicality and efficacy of a deferred consent procedure in an ongoing Dutch multi-centre clinical trial.

Screening logs were collected from two participating centres of a clinical trial that is currently conducted to evaluate the efficacy of early lactate-directed therapy and that uses deferred consent. Screened patients were analyzed for eligibility and reasons for exclusion. (12%) were not reported to the study investigators, 41 patients (16%) were not included for medical-ethical reasons (e.g. treating clinician deemed risk/benefit ratio of the study intervention unacceptable), in 7 patients (3%) study participation was practically impossible (e.g. unavailable study materials) and the reason was unknown in 2 patients (1%). Only 9 patients (or their relatives)(4%) refused informed consent.

In an ongoing Dutch multi-centre emergency clinical trial using deferred consent, only 4% of patients or their relatives refused informed consent. Deferred consent in emergency research is practical and facilitates a high inclusion rate. 

Adult respiratory distress syndrome (ARDS) and PEEP have been linked to right ventricular dysfunction (RVD). This has been attributed to elevated pulmonary artery pressure (PAP) and pulmonary vascular resistance (PVR) due to ARDS as well as increased intrathoracic pressure due to PEEP therapy. We wondered if RVD was a late phenomenon in ARDS or could also be detected during early PEEP treatment of hypoxia in patients with multiple ARDS risk factors. 

Pulmonary embolism is a highly prevalent disease associated with severe morbidity and mortality. Although the hemodynamic changes induced by pulmonary embolism are known, the alterations in respiratory mechanics after an embolic event are not completely understood. The aim of this study was to evaluate acute changes in hemodynamics, static and dynamic respiratory mechanics and lung histology induced by an experimental model of pulmonary microembolism.

Ten Large White pigs (weight 35-42 kg) were instrumented with arterial and pulmonary catheters and pulmonary embolism was induced in 5 pigs by injection of polystyrene microspheres (diameter ∼300 µM), in order to obtain a pulmonary mean arterial pressure (PMAP) of twice the baseline value. Five other animals were injected with saline and served as controls. Hemodynamic and respiratory data were collected and pressure x volume (PxV) loops of the respiratory system were performed by a quasi-static low flow method. Animals were followed for 12 hours and after death lung fragments were dissected and sent to pathology.

The average amount of microspheres necessary to generate microembolism was 6.7 ± 2.2 mg/kg. Pulmonary embolism induced a significant reduction in stroke volume (71±18 ml/min/bpm pre vs 36±9 post, p<0.05), an increase in PMAP (27±4 mmHg pre vs 39±6 post, p<0.05) and pulmonary vascular resistance (193±122 mmHg/L/min pre vs 451±149 post, p<0.05). Respiratory dysfunction was evidenced by significant reductions in PaO2/FiO2 ratio (480±50 pre vs 159±55 post, p<0.05), dynamic lung compliance (27±6 ml/cmH2O pre vs 19±5 post, p<0.05) and increase in dead space ventilation (20±4 pre vs 47±20 post, p<0.05). PxV curves of the respiratory system were affected by embolism, with shift of the loops to the right and consequent reduction in static compliance and pulmonary hysteresis. Pathology depicted inflammatory neutrophil infiltrates, alveolar edema, collapse and hemorrhagic infarctions.

Pulmonary microembolism induced by polystyrene microspheres is associated with cardiovascular dysfunction, as well as respiratory injury characterized by decrease in oxygenation, dynamic and static lung compliances and pulmonary hysteresis. Pathology findings were similar to those verified in inflammatory-induced acute lung injury. The similarities between respiratory and histologic features of this model and those from conditions associated with lung inflammation suggest that pharmacologic and ventilatory interventions already used to treat acute lung injury may also be tested in pulmonary embolism. 

The presence of Patent Foramen Ovale (PFO) is frequently underdiagnosed in ICU patients suffering from refractory hypoxemia. However, it is relatively common in the general population. We examined the prevalence of PFO in mechanically ventilated ICU patients with refractory hypoxemia and abnormal chest x-ray findings.

Over a period of five years, 24 mechanically ventilated patients with refractory hypoxemia and abnormal chest x-ray findings were examined with Transesophageal Echocardiography (TEE) for the presence of PFO as a contributing factor to their hypoxemia (right to left intracardiac shunt). All patients were ventilated with tidal volume 6-8ml.kg -1 and PEEP between 5-10cmH2O. Their mean PaO2/FiO2 ratio was 102±24 mmHg. The coexisting pathology consisted of: ARDS (12 cases), massive pulmonary embolism (3 cases), COPD (3 cases), CABG surgery with RV infarction (4 cases), cerebrovascular accident (1 case) and pulmonary oedema due to fluid overload (1 case 

During a two-month period we investigated the possibility of opening of the foramen ovale during a recruitment maneuver in either patients with ARDS or in patients with atelectasis and a PaO2/FiO2 ratio<200.

We enrolled 10 consecutive patients (ARDS: 6 cases, patients with atelectasis and hypoxemia: 4 cases), likely to benefit from a recruitment maneuver. Mean PaO2/FiO2 ratio was 112 and mean compliance was 40 ml.cmH2O -1 prior to the maneuver. All data regarding the mechanical properties of the lung were recorded from the ventilators monitor screen. After deficits of intravascular volume had been addressed and hemodynamics had been optimized, a baseline transesophageal echocardiographic study using contrast material was performed to rule out the possibility of a foramen ovale already patent prior to the maneuver. The recruitment inflation pressure was chosen as the lesser of 45 cm H2O or the peak pressure at 12 ml.kg -1 tidal volume. The ventilator was then adjusted to deliver this high inflation pressure for 20 secs. Five seconds after the onset of inflation, 20 ml of a contrast material were injected through a central venous line with the transesophageal probe already in place to detect the passage of the material to the left atrium. Passage of the contrast material to the left side of the circulation was detected using two dimensional echocardiography.

We found that the sustained high inflation pressure resulted in foramen ovale opening in 2 patients, whereas it did not produce such a result in 5 patients. In 3 of the 10 studied patients, the baseline transesophageal study revealed a patent foramen ovale before recruitment was attempted. No adverse effects following the recruitment maneuver were noted. Mean PaO2/FiO2 ratio was 138 and mean compliance was 40 ml.cmH2O -1 twenty minutes after the recruitment maneuver, with only one of the recruited patients showing a significant improvement in oxygenation.CONCLUSION. Patent foramen ovale may be a contributing factor of refractory hypoxemia in ICU patients. Opening of the foramen ovale is not an unlikely event during a recruitment maneuver. 

Acute respiratory distress syndrome (ARDS) remains a major problem in critically ill patients, with mortality rates of 40-50%. To date, no specific treatment has been shown to decrease mortality, but this may largely be due to the heterogeneity of the populations meeting the ARDS criteria.Objectives: To evaluate patients who died with a clinical diagnosis of ARDS and who had a postmortem examination in order to: 1-define the pathological alterations associated with the syndrome, with particular reference to the typical pattern of diffuse alveolar damage (DAD); 2-evaluate whether etiologies or precipitating factors were missed; and 3-speculate whether a lung biopsy could have guided the clinical management.

Three year (2002) (2003) (2004) review of all patients with ARDS (using the AECC criteria) who had a postmortem examination. Comparisons between ante-and post-mortem diagnoses were classified as major and minor discrepancies using the Goldman classification.

Results: Of a total of 9184 admissions, 376 patients had a clinical diagnosis of ARDS. Of these, 169 died; 69 had a postmortem examination and 64 of these had complete data for analysis. The main causes of death were multiple organ failure in 27 (42%) and refractory hypoxemia in 9 (14%). Postmortem lung examination revealed DAD in 32 (50%) patients (4 associated with a lung infection), (broncho)pneumonia without DAD in 16 (25%), invasive pulmonary aspergillosis without DAD in 5 (8%), and other diagnoses in 11 (17%). Major unexpected findings were found in 15 (23%) patients, classified as 6 Goldman class I errors and 9 class II errors. The class I errors included 4 cases of invasive pulmonary aspergillosis.CONCLUSION. ARDS as a syndrome, can be due to various pathological patterns; at autopsy, only half of patients with ARDS have typical DAD. Special attention should be paid to the possibility of aspergillosis; in this setting, lung biopsy may have a role.

G. S. Georgieva*, S. Kurata, C. Zhu, A. Bilali, T. Imai Critical Care Medicine, Tokyo Medical and Dental University, Tokyo, Japan

Development of efficient lung preservation method has been anticipated and we elucidated that positive pulmonary venous pressure (PVP) (5mmHg) prevented ischemia-reperfusion (I/R) injury in isolated mechanically ventilated rat lungs. The aim of this study is to determine whether CPAP accompanied with 5 mmHg of PVP would be effective for prevention of I/R injury.

After tracheostomy rats were ventilated at 70 strokes /min with air (5% C02) and with PEEP of 2.5 cmH20, cannulated to the left atrium and pulmonary arteries (PAs), and perfused with Krebs -Henseleit solution supplemented with albumin (4%) (0.04 ml/g/min). The lungs and heart "en block" were isolated and placed in a chamber; right and left bronchus as well as PAs were dissected which permit each lung to be ventilated and/or perfused selectively by selective occlusion of each bronchus and/or PA. After 30 min control condition, the left lung (LL) was maintained under CPAP (selective occlusion of left broncus); the control right lung (RL) was ventilated with peak airway pressure of 5 cmH20 above PEEP;perfusion to the both lungs was stopped (ischemia). Pulmonary venous outflow was elevated so as to be applied 5 mmHg to the left atrium during ischemia. After 60-min ischemia, reperfusion with 0 mmHg PVP and both lung normal ventilation were resumed for 30 min. Perfusion pressures of RL and LL was measured at the beginning and at the end of the experiment by occlusion either the left or right pulmonary artery, as appropriate. Albumin content in bronchoalveolar lavage fluid (BALF) separately for each LL and RL, and lung weight were measured. Protein content in BALF was calculated as (mg of protein)/(ml of BALF)/(g of lung dry weight). All the data were compared by Wilcoxon's rank-sum or Mann-Whitney U-test and expressed as mean +/-SD.

In I/R lung maintained at CPAP, wet/dry and BALF as well as perfusion pressure increased compared to the control RL. CONCLUSION. CPAP(2.5 cmH20) and 5 mmHg PVP cannot prevent ischemic lung injury despite constant distention of pulmonary vasculature and alveolar space. This suggests that gas exchange during ischemia would be necessary for escaping from I/R injury. 

Potential peripheral airway obstruction is of importance for the choice of ventilatory strategy in Acute Lung Injury (ALI). Use of a limited expiratory time counteracts early regional expiratory collapse but might cause hyperinflation in case of significant peripheral obstruction. The aim of this study was to assess regional expiratory time constants and gas trapping in early ALI.

Ten anesthetized pigs were ventilated in volume-controlled mode with I:E ratios of either 1:3 or 3:1 at a rate of 15 breaths per minute. Starting from the end-inspiratory level, sequential Computed Tomography (CT) exposures were performed during passive, uninterrupted expiration to the atmosphere. The procedure was performed before and after Oleic Acid-Induced Lung Injury (OAI) had been induced in the lower lobe on one side. The gas volume of bilateral dependent and non-dependent Regions of Interest (ROIs) was calculated from radiographical attenuation values. The expiratory time constant was calculated from a mono-exponential decay of ROI gas volumes during expiration. Gas trapping in injured and non-injured regions were compared.

During ventilation with I:E ratio 1:3, OAI caused overall compliance to decrease from 31 +/-5.5 to 27 +/-4.4 mL/cmH2O (p<0.05). Dependent, injured regions showed a shorter time constant and a lower volume of gas than dependent non-injured regions regardless of whether the preceding end-inspiratory volume had been increased or not by application of a limited expiratory time. In non-dependent, non-injured regions, the gas volume was similar on both sides after both patterns of ventilation. 

One of the additional approaches in the therapy of the acute respiratory distress syndrome (ARDS) is the use of a pumpless arteriovenous extracorporeal membrane oxygenator (interventional lung assist (ILA)). The aim of our study was to test the effects of an ILA system on hemodynamics and gas exchange during resuscitation and to establish whether ILA should be kept open or clamped under these circumstances.

The study was designed as a prospective experimental study. The experiments were performed on 12 pigs (41 to 58 kg body weight). The pigs were anesthetized and mechanically ventilated. One femoral artery and one femoral vein were cannulated and connected with ILA. Acute lung injury was induced by repeated bronchoalveolar lavage until arterial partial pressure of oxygen (Pao2) was lower than 100 Torr for at least 30 min during ventilation with 100% O2. Ventricular fibrillation was then induced by an indwelling pacemaker. Manual compressions of the thorax were started at once and continued for 30 minutes. In 6 animals, ILA was kept open, in the other 6 it was clamped immediately. Statistical analysis was performed using GraphPad Prism. Two-way analysis of variance was applied and significance was accepted at P values < 0.05. The data is given as mean ± SD.

With a mean systolic arterial pressure in the group with ILA open of 76 ± 28 mm Hg and 78 ± 23 mm Hg with ILA clamped and mean blood pressures of 23 ± 5 mm Hg with ILA open and 25 ± 5 mm Hg with ILA clamped the blood pressure did not differ between the two groups. Endtidal carbon dioxide decreased from 46 ± 23 Torr with ILA open and 37 ± 9 Torr before intervention to 8 ± 5 Torr and 8 ± 10 Torr, respectively. The arterial partial pressure of carbon dioxide (Paco2) was significantly lower in the group with the ILA system open (31 ± 25 mm Hg versus 80 ± 25 mm Hg at 20 minutes) and the Pao2 was higher (although significant only at 20 minutes, 191 mm Hg ± 140 mm Hg versus 57 mm Hg ± 14 mm Hg).

The blood pressure generated with thorax compressions did not differ significantly between the two groups and endtidal CO2 was also in the same range. Therefore we assume that circulation was not significantly affected by ILA and that the shunt caused by the ILA system did not deteriorate circulation. Paco2 was significantly lower in the group with the ILA system open and Pao2 was higher. Our results indicate that the ILA system was not harmful during resuscitation, it even might have a beneficial effect.GRANT ACKNOWLEDGEMENT. The study was partially supported by Novalung, Hechingen, Germany.

Respiratory failure -Miscellaneous 0317-0330 

Increased thorax rigidity and high intraabdominal pressure reduce the stretch ability of the thoracic cage and modify the regional lung function. This phenomenon is often seen in intensive care patients, e.g. with abdominal compartment syndrome. Objective of this study was to determine the effect of decreased thoracic cage compliance on regional distribution of spontaneous ventilation in different postures by the non-invasive method of electrical impedance tomography (EIT).

For this survey we examined ten healthy male spontaneously breathing volunteers (mean age ± SD: 28 ± 3 years; body weight: 81 ± 10 kg, height: 183 ± 9 cm). The compliance of the thoracic cage was restricted by external abdominal and thoracic corsets respectively. The EIT examinations were performed with the Goe-MF II EIT device (Viasys Healthcare, Höchberg, Germany). Sixteen self-adhesive electrodes (3M Red Dot 2239, 3M Health Care, Borken, Germany) were applied on the chest circumference in one transverse plane and used for rotating electrical current injection and voltage measurement. The EIT data were acquired at a rate of 25 scans/s. Impedance data and spirometry were obtained during spontaneous ventilation in three body positions (sitting, left and right side). Statistical analysis was performed using repeated ANOVA with Bonferroni's multiple comparison test and Student's t test. P values <0.05 were considered significant.RESULTS. The regional distribution of ventilation in subjects without restrictions revealed a close match with physiologically expected values. Thoracic and abdominal restrictions led to reduction of ventilation in the dependent lung areas. The non-dependent lung areas were not affected. The fractional ventilation in the dependent lung areas was reduced in the right side position from 69.3 ± 15.4% to 57.2 ± 9.8% (thoracic restrictions) and 55.7 ± 9.9% (abdominal restrictions), in the left side position from 55.3 ± 11.4% to 36.4 ± 8.4%, and 36.9 ± 8.7%.

Thoracic and abdominal restrictions of the thoracic cage reduce ventilation only in the dependent lung regions in spontaneously breathing healthy volunteers. EIT is a suitable method for non-invasive determination of regional lung ventilation. 

K. Raymondos* 1 , K. Vieweger 1 , J. Ahrens 1 , M. Przemeck 2 , M. Homann 3 , S. Piepenbrock 1 1 Anaesthesiology, Medical School Hannover, 2 Anaesthesiology, Annastift, 3 Johanniter-Unfall-Hilfe e.V., Ortsverband Wasserturm, Hannover, Germany

Germany are still performed with ambulances in that only limited monitoring and usually only volume-cycled emergency ventilators can be used. We established an intensive care ambulance system and evaluated the transfers of critically ill patients performed with this system.

We prospectively recorded 1847 interhospital-transfers. The ventilatory modes before and during the patients' transfer and further characteristics of the interhospital-transfers were evaluated. Transport ventilation was performed with the Raphael ® silver ventilator (Hamilton Medical AG, Rhäzüns, Switzerland) with that also pressure-support ventilation (PSV), airway pressure release ventilation (DuoPAP ® /APRV) and the combination of both could be used.

Indications for the interhospital-transfers included ischemic (28.1%) and other (6.7%) cardiac diseases, cerebral diseases (24.7%) of which 73% required neurosurgy, pulmonary disease (8.4%) and others (32.1%). 605 (32.8%)% of the transferred patients received ventilatory support, 712 patients (38.6%) breathed spontaneously with and 529 patients (28.7%) without oxygen insufflation. The majority of the mechanically ventilated patients received ventilatory modes supporting spontaneous breathing before (74.4%) and during the transfer (80.6%). The patients were transferred in 55 minutes (5 minutes -11 hours) over a distance of 60 km (2-970 km) (median (range)). At least 2 motor syringe pumps were needed during the transfer of 642 patients (34.8%). Monitoring during the transfer was similar or more extended compared to the monitoring in the hospital prior to transfer (ECG 93% vs. 85%, pulse oximetry 89% vs. 70%, non-invasive blood pressure 73% vs 65%, intraarterial pressure 24% vs 22% and capnography 21% vs. 9%).

Most ventilated patients received weaning techniques and most of these ventilatory modes were continued during the transfer. These ventilatory modes and a more extended monitoring including intraarterial pressure monitoring and capnography cannot be applied in emergency ambulances. The less invasive ventilatory modes and the extended monitoring enable a less invasive and safer interhospital-transfer as the intensive care treatment and monitoring prior to transfer is maintained or even extended during the transport.

A. Sánchez*, M. Palomar, R. Alcaraz, A. Socias, D. Moreira Intensive Care Unit, HG Vall d'Hebron, Barcelona, Spain INTRODUCTION. some series have shown the bad prognosis of patients with Pulmonary Fibrosis (PF) who require admittance at ICU for respiratory failure. There are doubts of the benefit of the ventilatory support if the precipitating cause is not well defined. Lung Transplant (LT) could be a therapeutical option. The aim of this study was to analyze the prognosis of the patients with PF who are admitted to an ICU of a Hospital with LT program.METHODS. case-series, observational study of patients with PF and acute respiratory failure admitted to the ICU of a third level Hospital with LT programm between January 1998 until June 2006. Information about the cause of PF, clinical course, current status, ventilatory support, length of stay, pulmonary functional tests, possibility of trasplantation, complications and mortality was collected.

. 22 patients (15 men, 7 women) with PF (14 idiopathic PF, 6 connectivopaty and 2 due to radiotherapy) were admitted for acute respiratory failure (ARF) to our ICU. Mean age was 54,5 (21-65) years. The median duration of illness from diagnosis until admittance was 1,5 (0-17) years. APACHE-II score was 15 (6-27). The precipitating cause of ARF was identified in 18 patients: bacterial pneumonia was documented in 5 patients; 1 had a pulmonary embolism; 1 fungic infection and 11 cases were due to the progression of the disease. In 4 cases the precipitating cause could not be identified. Mechanical ventilation (MV) was required by 18 patients (81,8%) during an average of 11,5 (1-51) days with a mortality rate of 77,7%. Pa O2 /Fi O2 at admittance 77 (40-260) mm Hg; and PaCO2 at admittance 54 (31-122) mm Hg. Respiratory functional studies were available in eleven patients with a FEV1 of 1.34 (0.52) l and FVC of 1.65 (0.77) l. 16 patients (72%) died during their stay at ICU. The cause of death was multi-organic failure in 7 (31.8%); refractary hypoxemia in 6 (27.3%) patients and 3 of them died while the transplantation was being performed. Mean length of stay was 14 (1-56) days. 12 patients were included in the urgent LT list and 10 were transplanted. No donor was found in 2 cases and died on the waiting list. There were performed 2 single-lung and 8 double-LT. Mean age was 48 (21-65) years. The time from the admittance until transplantation was 6 (3-19) days. 9 of them (75%) required MV with a mortality rate of 66,7%. From this group 8 (66,7%) patients died during their stay at the ICU. 3 of the patients died while the transplantation was being performed.CONCLUSION. literature shows a bad prognosis of patients with PF who need admittance to an ICU for ARF. In our experience the survival was 33% so the existance of a LT programm could offer a chance to these patients.

M. E. Lugarinho*, P. P. Souza Intensive Care Unit, Hospital de Clinicas Mario Lioni, Rio de Janeiro, Brazil INTRODUCTION. Acute kidney insufficiency (AKI) worsens the outcome in critical ill patients. We investigate whether the presence of AKI had any effect on lenght of mechanical ventilation and mortality rate.

Observational, prospective study in a 12-bed general intensive care unit (ICU) from January to December 2006. The inclusion criterion was invasive mechanical ventilation for more than 48 hours. AKI was defined as the presence of dialysis during the ICU stay. Patients were then separated into AKI and non-AKI patients (control group). The primary end point was duration of total length of mechanical ventilation and the secondary end point was the ICU mortality.

A total of 114 patients were studied: 28 with AKI and 86 non-AKI. The groups were similar in regard to age, sex, and Apache II score. The median (interquartile range) duration of mechanical ventilation 15 [6-22] versus 5 [2] [3] [4] [5] [6] [7] [8] [9] [10] days, (p<0,005). The ICU mortality rate were significantly greater in the AKI patients: 75% versus 56,4% (p<0,005).CONCLUSION. This study shows that renal insufficiency has serious impact on the duration of mechanical ventilation and morbi-mortality in critically ill patient. These data elicits the poor outcomes of mechanical ventilated patients who demands for dialytic methods. It will be useful in end of life discussions and decisions in our ICU. INTRODUCTION. 5-HT1A-R-agonist 8-OH-DPAT has been shown to counteract morphine induced ventilatory depression, while opiate antinociception remained unaffected. Repinotan-HCl, another 5-HT1A-R-agonist, is unlike 8-OH-DPAT suitable for the use in humans. It was hypothesized that Repinotan-HCl is capable to antagonize ventilatory depression without impairing anti-nociception in rat.

With approval from local animal care committee, 12 rats were anesthetized with sevoflurane and tracheotomized to record respiratory rate (RR), tidal volume (VT) minute ventilation (MV). Inguinal vessels were catheterized to monitor arterial blood pressure and apply drugs IV. Nociception was assessed by tail-flick reflex. Morphine was administered at increments of 5mg/kg until a target 70% reduction of RR was achieved. Subsequently, repinotan-HCl was added cumulatively at increasing doses (0.02, 0.2, 2, 20, 200 µg/kg, n=6). Another group received NaCl 0.9% to serve as control (n=6).

Morphine (12.5±1.8mg/kg) depressed RR to -81±4%, and TFR was abolished with first dose of morphine in any experiment. Repinotan-HCl antagonized ventilatory depression dose-dependently, 20mcg/kg repinotan-HCl re-established ventilation almost at pretreatment level (RR +1.9±14%, p<0.001, 2-ANOVA, compared to control). TFR remained absent throughout repinotan administration.

Repinotan functionally antagonized morphine-induced ventilatory depression, while suppression of nociceptive reflex sustained. 5-HT1A-R-agonists such as repinotan-HCl appear to be promising candidates to stabilize spontaneous breathing. 

A. Makowski* 1 , B. Misztal 1 , C. Plowright 1 , K. Safranow 2 1 Anaesthetics, Medway Maritime Hospital, Gillingham, United Kingdom, 2 Biochemistry, Pomeranian Medical University, Szczecin, Poland

Vapotherm's (VAP) patent pending membrane technology makes higher flows from 1 to 40 lpm possible by saturating breathing gases with water vapor at body temperature. FiO2 is ranging from 0.21-1.0. Heat and humidity allow nasal flow to be well tolerated by the patients. High flow in animal study caused small amount of PEEP. Can we achieve desired therapeutic goal in treatment of respiratory failure (RF) with this very simple, non-invasive method?

We investigated effectiveness and hospital outcome of patients with RF treated on VAP at surgical HDU between December 2005 and March 2007. Data were taken during retrospective investigations. We analysed type and reason of RF as well as respiratory rate (RR), FiO2, flow, Arterial Blood Gases (ABG). Data were collected before (BEF) VAP was commenced, 1 hour after, and every day of treatment. We also recorded length and outcome of VAP therapy and patient satisfaction. Data were analysed with Wilcoxone and also Spearman's Rank Correlation tests.

The 55 patients (49% female, 51% male) at age 33-93 (69.81±14.55) were treated 1-16 (4.5±3.13) days. We applied VAP therapy for 75.92% patients with type I RF and 24.07% with type II RF. The reasons of RF were pneumonia in 45.45%, sepsis in 21.81% pulmonary oedema in 14.5%, COPD in 9.08%, others in 9.16%. For 81.82% patients there was a sufficient and definite treatment whereas 18.18% required mechanical ventilation and ICU admission. The 87.27% of patients were satisfied with therapy. The 81.82% survived and were discharged from the hospital. 

High flow and small amount of PEEP reduce work of breathing and significantly decrease RR. After effective Vapotherm therapy we observed in ABG significant increase of oxygen saturation and PaO2. Vast majority of patients were satisfied during the treatment. 

In critically ill patients who need long-term mechanical ventilation, early tracheostomy may facilitate weaning and shorten the length of stay in intensive care (1). However, there are no clinical tests that identify patients as being at an increased risk for prolonged ventilatory support; clinicians must predict the duration of arteficial ventilation by their clinical experience.

In our surgical intensive care unit we conducted a prospective clinical study to determine if there was an association between different clinical parameters (age, body mass index, GCS, SAPS2 score, vasopressor use, PaO2/FiO2 ratio) and long-term mechanical ventilation. Furthermore, we examined the positive predictive value of clinicians' prediction; to do that, clinicians had to indicate whether they considered prolonged mechanical ventilation as the most likely (but not always certain) outcome or not. We enrolled 50 patients and collected date on days 0-4th and 7th of treatment. Prolonged meshanical ventilation was defined as at least 7 more days on respirator.

None of the examined parameters could be used alone to predict long-term mechanical ventilation. Overall sensitivity of clinicians' prediction was 72.7 %, and positive predictive value was 46.1 %. 27.3 % of patients died, 26.6 % was weaned from respirator (6.2 % extubated) within 7 days despite predicted by clinicians as having prolonged ventilatory need. Suprisingly, the best positive predictive value (64.0 %) was found on the day of admission, the worst (25.0 %) on day 7; the difference was not significant (p=0.17 with chi-square test). This result could be explained by the fact that most patients in the study group were ventilated on day 7, but only a few on day 14.CONCLUSION. Prediction of prolonged mechanical ventilation was found to be very inaccurate, and did not improve in the course of first week of treatment. However, in our department where many neurosurgical patients are treated, only a minority could be extubated within 7 days when long-term ventilatory support was predicted. As selection of patients who need tracheostomy seems not to be better after one week of treatment than at an early stage, there can be a reason for early tracheostomy if we anticipate prolonged arteficial ventilation.

N. Abidi 1 , H. Thabet* 2 , O. Béji 1 , H. ELghord 1 , N. Brahmi 1 , M. Ben Othmen 2 , N. Kouraichi 1 , M. Amamou 1 1 Intensive care medicine, 2 Emergency medicine, Centre d'assistance médicale urgente, Tunis, Tunisia INTRODUCTION. Acute exacerbation of COPD is a frequent cause of admission in ICU and usually have a poor outcome. Such a patient consume a large amount of resources particulary if they need endotracheal intubation. The aim of this study is to report epidemiological, clinical features,treatment and outcome of patients admitted in ICU for acute exacerbation of COPD.

A retrospective study was carried out of consecutive admisions in ICU over a 7 years (from january 1999 to December 2005). American thoracic society criteria are usued to define COPD. Exacerbation is defined as a worsening of COPD symptoms.

A total of 144 patients were included in this study with 164 episodes of acute exacerbation. Mean age was 66±10,5 years. The sex ratio was 2,64 (M/F: 119/45). Eighty percent were current tobacco users. Seventy two percent had one or more associated comorbities mainly cardiovascular disease. According to COPD severity 53,7% of patients were in stage III. 25,6% were receiving home oxygen and 53 (32%) were previously mechanical ventilated. On ICU admission severity score are APACHE II 21±8; IGSII 44±19. 51 patients (31%) have a Shock and 39 (23,8%) have a coma (GCS<9). Treatment consist of starting non invasive ventilation (NIV) for 49 patients (30%); 82 patients (50%) need immediate intubation and mechanical ventilation. Failure of NIV was noted for 27 patients. In the course of hospitalisation in ICU main complications were: nosocomial infection for 57 patients (34,8%), barotrauma 10 patients (6,1%) and thromboembolic complications for 3 patients (1,8%). The median ICU stay was 17,6 ± 23,5 days and Mortality was 36,6% (60 patients). The main cause of mortality were septic shock (37 cases, 61,7%) and ARDS (9 cases, 15%).

In this retrospective study patients admitted for exacerbation of COPD need a mechanical ventilation in 66,5%. Failure of NIV were 55%. Main complications were nosocomial infection (34,8% of cases). Mortality is high 36,6% but not different for patients admitted in ICU for other disease. 

It is described, that gelatin leads to red blood cell (RBC)-coating, which is protective against shear stress in extracorporeal circuits. (1) An increase of mean corpuscular volume (MCV) without an increase in mean corpuscular hemoglobin content as well as a reduction of red blood cell (RBC) counts can be assumed to reduce pulmonary oxygen transfer. Increased RBC aggregability (accelerated blood sedimentation rate, BSR), as could occur due to coating, impairs microcirculation. Since adequate oxygen delivery is important in ventilated patients to counteract metabolic acidosis, we compared RBC features in acidotic pigs undergoing hemofiltration.

Healthy pigs (male, DLxDE, 37-46kg) were anesthetized, received acid infusion (0.4 M) and low tidal ventilation with FiO2 > 0.9 resulting in normoxic acidosis (pH 7.19-2.4; PaCO2 80-85 mmHg). Tris-hydroxymethylaminomethane (THAM) was infused to titrate a pH of 7.19-7.24. Either HES 130 or GEL (n=6-7/colloid-group) was infused additionally to crystalloids (colloid to crystalloid ratio was 1:4). Samples were collected before acid and colloid infusion (BS), after induction of acidosis (baseline acidosis, BsA), and after 3h of continuous acidosis (3hA). Thereafter, acid infusion was stopped and THAM was infused with 2.1 mol/kg/h for 2h in order to normalize pH-values. Final values (FV) were taken. Parameters investigated were: PaCO2, RBC counts, MCV, and BSR. The FiO2/PaO2 ratio was also determined.

Compared to HES 130 application, GEL infusion was associated with a reduction in RBC count, an increase in MCV and an accelerated BSR from BsA until FV. Values did not recover from initial deterioration (BsA) even not after normalization of pH (FV). Based on the healthy lungs in this porcine model, these changes did not impair PaO2/FiO2 ratio. 

Whether increases in MCV were due to GEL coating or due to unhampered swelling of RBCs during acidosis could not determined. However, in acidotic pigs GEL induced unfavorable effects concerning RBC features with respect to rheology while HES 130 did not. In individuals with impaired pulmonary function and hypodynamic state the described difference between the two types of colloids could become crucial with respect to total oxygen delivery. 

Perctaneous dilational tracheostomy (PDT) has become more common procedure used in intensive care. However, several complications, such as hemorrhage, posterior tracheal wall injury, tracheal stenosis have been recently reported. The aim of this study was to confirm whether the ultrasound can easily and clearly delineate the pretracheal anatomy and identify the potential problems for PDT. We also examined the accuracy in identifying the correct puncture level between 2 and 3 tracheal cartilages blindly (by hand).

We studied 10 patients and 40 volunteers. Before ultrasound scanning, the circumference of the neck was measured and the puncture level between 2 and 3 tracheal cartilages was marked blindly in each subject. In ultrasound scanning, we examined the relationship of the thyroid to the trachea, aberrant vascular anatomy in the pretracheal region, counted the number of extrathoracic tracheal rings. The distances from the skin to cricothyroid ligament and anterior tracheal wall at the level between 2 and 3 tracheal cartilages were estimated and the relationship between depth of trachea and circumference of the neck was analyzed by simple regression. We also checked the level of trachea pointed by operator blindly was correct or not by comparing the level identified by ultrasound images.

The mean age and circumference of the neck were 39±17 years (range: 20-79) and 36±4 cm. Ultrasound examination of the trachea and thyroid was easily carried out in each subject except 2 subjects. Approximately 6 extrathoracic tracheal rings could be imaged with ultrasound. Anterior jugular veins were seen in 19 subjects (38%) and six were near the midline. The depth of trachea between 2 and 3 tracheal cartilages were varied in each subject (0.44-3.58cm) and there were stastistically relatioship between circumference of the neck and depth of trachea (R2=0.33, P=0.0047). The accurate decision of trachea level was made in 88% of the subjects.CONCLUSION. This study showed that: 1) ultrasound can delineate the neck structure and detect variations related to the complication of PDT; 2) blind identification of the puncture level for tracheostomy without ultrasound was not necessarily correct. Our results demonstrated that the routine use of ultrasound could be recommended before PDT. INTRODUCTION. Fluid therapy system of critically ill patients is very variable, and it is based in the interpreting of differents physiologic parameters with a double aim, by one hand keep an adequate perfusion of vital organs, and the other hand avoid overload volumen. Our objective was analyze changes in critically ill patients fluid therapy when we including EVLW in treatment protocol and evaluate response in short time.

Observational and prospective study in a neurotraumatological ICU. We included consecutives patients that were admited with acute lung injury/adult respiratory distress syndrome and/or septic patients who needed monitoring with central venous and arterial catheterization with PICCO system. We made a therapeutic reassessment of the fluid therapy and/or vasoactives after we knew EVLW when one of the following events in the patient evolution hapenned: hypoxemia, hypotension, olyguria/anuria, or its addition. Response in short time was also evaluated.

Our sample included 40 patients and 52 determinations( 7 patients with 2 determinations, 1 patient with 4 determination and 1 patient with 3 determination). After we knew EVLW we changed initial therapeutic plan in 55.8 %; this change affected fluids in 96.6 % and vasoactives in 30.4 %. EVLW in patients who therapeutic plan was modified was 13.76 ± 4.98 and if therapeutic plan was not modified, EVLW was 9.87 ± 4.13 (p< 0.05). Association is observed between EVLW value and decision about fluids, so when we decided increase fluids was 7.17 ± 1.47; if the decision was decrease fluid, EVLW was 12.20 ± 2.77 and in the cases that diuretics were added 16.36 ± 3.99, in all cases statistics significant was found. No differences was observed in EVLW values about vasoactives decision. We found improvement of initial event in short time after intervention in 67.3 %.CONCLUSION. EVLW determination affects in important way to fluids therapy plan in critically ill patients. We think that inclusion of EVLW contributes to a more racional management of these patients. 

Patients who had received iNO were identified from ICNARC records. Hospital notes and ICU charts were reviewed. Data collected included diagnosis, APACHE II and unit and hospital outcome. The PaO2/FiO2 ratio (in mmHg) was recorded prior to starting iNO (day 0) and subsequently on days 1-4 using the data from the time at which oxygenation was best in each 24 hour period.

. 29 patients received iNO. 28 patients received it for treatment of hypoxaemic respiratory failure, and 1 for treatment of pulmonary hypertension. Mean APACHE score was 19.2 on admission (survivors 14.4; non-survivors 22) . The mean PaO2/FiO2 ratio was 67.6 on day 0 and improved to 135.5 on day 1. In unit survivors, the mean PaO2/FiO2 increased from 62.9 to 168.3 on day 1, compared with unit non-survivors in whom it increased from 70.2 to 116.6. 19 (66%) of patients were responders to iNO (defined as a >20% increase in PaO2/FiO2 ratio). Unit and hospital survival figures for responders and non-responders are presented below. 

Hospital Surviviors (n=9) Hospital Non-Survivors (n=20) Responder (n=19) 8 (42%) 11 (58%) Non-Responder (n=10) 1 (10%) 9 (90%) Fisher's exact test (2 tailed) p=0.1 CONCLUSION. iNO was used in patients with more severe hypoxia than those included in randomised trials. (2) In this review, responders were found to have a significantly reduced unit mortality and a reduced hospital mortality compared with non-responders. We believe iNO may be a valuable therapy in ARDS patients with severe refractory hypoxaemia, and that studies in this subgroup of patients are warranted. Outcome predictors of HFOV in severe ARDS are not well studied. We prospectively evaluated the outcome predictors of HFOV in adult ARDS. METHODS. ARDS patients receiving mechanical ventilation as per the ARDSnet protocol with PO2/FiO2≤150 inspite of PEEP≥12cm and FiO2≥0.7,were considered for HFOV. Continuous Distending Pressure(CDP),Frequency ,Amplitude, Inspiratory time and Bias Flow of HFOV were optimised with the help of frequent blood gas analysis. Weaning from HFOV to pressure support ventilation was attempted once PO2/FiO2 ratio remained ≥200 with CDP≤18 cm &FiO2 ≤0.5. Responders(R) were defined as patients who were successfully weaned to a state which required no ventilatory support for > 12 hrs. Non Responders(NR)were defined as patients who could not be weaned off ventilatory assistance. RESULTS. 17 out of total 31 patients were R & 14 were NR. Both the groups were similar prior to HFOV as shown in table. Improvement in PO2/FiO2 ratio and Oxygenation Index (OI) at 6hrs &24 hrs in R group was statistically significant as compared to that in NR group. 

We could show that chaotic variation of pressure support improves pressure support ventilation (PSV), and named this new mode noisy PSV. In this work, we compared noisy PSV to conventional biphasic positive airway pressure ventilation (BIPAP), which has been claimed to be a "gold standard", in experimental acute lung injury.

After approval by the local animal care committee, 18 juvenile pigs (22.5-30.7 kg) were anesthetized and mechanically ventilated (Dräger EVITA XL 4Lab; volume controlled ventilation, VT = 10 ml/kg; FIO2 = 1.0; PEEP = 5 cmH2O). After induction of acute lung injury by saline lung lavage (30 ml/kg), lungs were recruited and a decremental PEEP trial was performed to determine the optimal PEEP according to the elastance of the respiratory system (Ers). Thereafter, spontaneous breathing was resumed and animals were randomly assigned to noisy PSV or BIPAP groups (n=9 each group). The ventilator settings were as follows -BIPAP: FIO2 = 1.0; Plow = according to PEEP of minimal Ers; Phigh = titrated to generate VT of 6 ml/kg; Thigh = 2s; Tlow = 2s -noisy PSV: FIO2 = 1.0; PEEP = according to PEEP of minimal Ers; mean PASB = titrated to generate VT of 6 ml/kg. Noisy PSV was accomplished by means of remote control of the EVITA XL 4Lab by a laptop, which generated a sequence of 600 respiratory cycles with different pressure support levels (mean = PASB; SD = 30 % of mean). Gas exchange, respiratory parameters and hemodynamics were measured at baseline, injury, after resuming of spontaneous breathing (baseline 2) and during an observational period of 4h. Statistical analysis was performed with general linear model statistics adjusted for repeated measures using baseline 2 as covariate. Significance was accepted at p<0.05.

Bodyweight, PEEP and number of lavages as well as hemodynamics did not differ significantly between groups. Oxygenation and CO2 elimination were significantly improved with noisy PSV (p<0.05 both). Analysis of respiratory parameters revealed significant lower mean airway pressures with noisy PSV as compared to BIPAP (p<0.05), as well as increased mean peak airway pressure, spontaneous respiratory rate, and mean tidal volume (p<0.05 all).CONCLUSION. This study represents the first evaluation of the recently developed noisy PSV combined with PEEP levels titrated according to lowest Ers. Noisy PSV was found superior to conventional BIPAP with regard to gas exchange and respiratory parameters. Further experimental studies are necessary to determine the potential role of noisy PSV in intensive care therapy. 

We investigated if chaotic variation of pressure support (noise) can improve the performance of pressure support ventilation (PSV) in experimental acute lung injury (ALI).

With approval of the local animal care committee, 12 pigs weighing 25 to 30 kg were anesthetized, intubated and mechanically ventilated (volume-controlled mode, FIO2=0.5, PEEP=5 cmH2O, tidal volume=12 ml/kg). Following that, ALI was induced by surfactant depletion, and biphasic intermittent positive airway pressure (BIPAP) was initiated with: lower CPAP (CPAPlow) = 5 cmH2O, higher CPAP (CPAPhigh) titrated to obtain tidal volumes of 6-8 ml/kg, respiratory rate set to obtain PaCO2 between 50-60 mmHg. Then, depth of anesthesia was decreased to allow spontaneous breathing, and animals were ventilated with two different modes (1 hour each, random sequence): 1) traditional PSV, with pressure support level set at CPAPhigh -CPAPlow; 2) noisy PSV, with random variation of pressure support and mean value set at CPAPhigh -CPAPlow, and standard deviation set at 30 % of the mean value (normal distribution). Gas exchange, inspiratory drive (P0.1) and inspiratory pressure time product of esophageal pressure (PTP) were assessed. Helical computed tomography (CT) of chest was performed at end-expiration and the hyperaerated, normally aerated, hypoaerated and non-aerated lung compartments were calculated in 8 animals. 

Patients with respiratory failure treated with VM with FIO2 0.5 were included. After 30 minutes of oxygen therapy, arterial blood gases were collected and patients were asked to quantify (from 0 to 10) three items: dyspnea, dry mouth and general confort. Then, VM was changed for HFNC (OptiflowTM, Fisher & Paykel, New Zeland) . The same variables were collected after 30 minutes using HFNC. Results are expressed as median (interquartil range). We have applied SPSSwin v13.0 with Wilcoxon test.

Patients n=10 (4 M), age 54 (33-64). In the moment of inclusion, one patient (10%) presented MODS and SOFA score was 4 (2.5-5.5). During their evolution, five patients (50%) finally need endotracheal intubation. Main results are presented in the following tables: 

A computer-driven system (CDS) has been recently used to optimise PSV to patient's needs during weaning. In some pts, the CDS fail to find a "comfort window" despite stepwise increase in pressure support (PS) levels. For these pts, CDS could further increase respiratory muscle workload. We speculate that failure to adapt respiratory rates (RR) and VT following changes in PS levels might identify a subset of pts unlikely to benefit from the CDS.To test this hypothesis, we used a bedside test before switching ventilated pts to a closed-loop algorithm of PSV.

We studied 23 pts at initiation of weaning with PSV using the smallest PS level resulting in RR≤30, VT>5 mL/kg. We collected baseline values and assessed changes in VT (DVT), RR (DRR) during 5 min after 5 cmH2O-increase and decrease in PS levels. Then, a CDS session was started at the baseline PS level. We searched for correlations between DVT, DRR, and outcome (failure/success) of the CDS sessions. A CDS session was deemed successful when the system detected criteria for separation of the ventilator or when PSV was efficiently adjusted by the CDS within 48 h after starting the session. 

In pressure support ventilation auto-PEEP is considered a major contributor to the inspiratory work of breathing. Measurement of auto-PEEP requires esophageal pressure tracings, which are not routinely available. The presence of auto-PEEP is likely, when flow is interrupted at end-expiration, a pattern well-established in controlled ventilation. We studied expiratory flow-volume relationships as substitute for detection of auto-PEEP in patients on pressure support ventilation.

In 22 patients successively admitted to our ICU respiratory mechanics were obtained from 5 consecutive breaths on pressure support ventilation. Auto-PEEP was considered present when in flow-versus-time recordings flow was interrupted at end-expiration. From flow-volume relationships expiratory time-constants were calculated and related to actual expiration times. All measurements were obtained with a NICO-computer; for analysis a computer program Analysis Plus was used (both Respironics/Novametrix, Inc.).

In 7 of the 22 patients flow at end-expiration was interrupted suggesting the presence of auto-PEEP (interrupted flow group). In the remaining patients flow was zero at end-expiration (zero flow group). In the flow-volume curves of patients in the interrupted flow group versus the zero flow group end-expiratory flows varied between 0.14 -0.33 l/s and 0.04 -0.10 l/s respectively. The expiratory time-constants ranged from 1.1 -1.6 s in the interrupted flow group and 0.5 -1.0 s in the zero flow group. The ratios between expiration times and expiratory time-constants varied between 0.8 -1.5 and 1.9 -4.1 for the interrupted and zero flow groups respectively . The means and standard deviations for both groups were:Means +/-SD 

In patients on pressure support ventilation with interrupted flows at endexpiration higher expiratory time-constants and lower ratios between expiration times and time-constants were found, suggesting the presence of auto-PEEP. These variables can be used as substitute for detection of auto-PEEP. 

Non Invasive Ventilation (NIV) is the delivery of assisted mechanical ventilation to the lungs, without the use of an invasive endotracheal airway. NIV has decreased the need for invasive mechanical ventilation and its attendant complications. Acute Cardiogenic Pulmonary Edema (ACPE) is defined as an episode of acute heart failure accompanied by severe respiratory distress and oxygen saturation <90% on room air before all treatment. Our study aimed to asses the respiratory effects of a device that delivers a continous positive airway pressure via face mask in patients with severe ACPE, the feasibility of using this technique in an emergency department (ED) and estimed the need of endotracheal intubation (EI).

We evaluated a series of 20 patients consecutively treated in our ED for ACPE, from June 2006 to December 2006. A PEEP level of 10 cm H2O delivered by CPAP-Boussignac device (Vygon, Ecouen, France) was used in all patients. FiO2 was estimed to range from 60 to 90 %. Clinical and blood gas parameters were recorded at entry and also after 30 minute and 1 hour of treatment. All patients were treated with standard medical therapy. The average of age was 67 years (52-82), 13 were male and 7 were female. The inclusion criteria for NIV were: pH <7,35 but >7,10, PaCO2 >45mmHg or an acute augment of 15-20 mmHg, respiratory rate >25/min, PaO2/FiO2 <250mmHg on room air and Score Kelly max 3.

Resolution of respiratory distress occurred from 30 to 50 minute ( media 40 minute). All patients showed an improve of clinical and emogasanalytic impairment. Only 5 patients needed EI and were transferred in ICU. 15 patients were treated in ED and after normalization and stabilization of their vital signs they were discharged in other medical departments (10 cardiology department and 5 pneumology department). The rate of EI was 25%.CONCLUSION. CPAP delivered using Boussignac device is feasible in an emergency care setting. It can quickly improve respiratory distress in ACPE patients and reduce the need of EI. 

In clinical practice NIV is being used as a sole respiratory support modality or in the weaning period in at least 50% of ARF patients admitted to emergency department. The remaining patiens need IMV as primary and secondary forms of respiratory support. Failure of NIV seems to predict higher mortality rates. As a conclusion we need both support modalities and the physician has to use them carefully according to patients condition and their expertise. METHODS. Medline, Pubmed, Cochrane, & CINAHL databases (1982 to 2006 were searched using the terms: APRV, BIPAP, Bilevel & lung protective strategy, individually and in combination. Reference lists of identified papers were also examined. Two independent reviewers determined eligibility of papers based on predefined criteria. Database searching yielded 501 citations, of which 81 were selected on review of title and abstract. Data were abstracted onto pre-designed forms from 51 experimental studies and 19 discussion articles on further review.

Of the 51 experimental studies, 31 used a randomised design, 9 were cohort studies and 11 case series. APRV was the named mode in 39 (76%) studies, BIPAP in 11 (22%), and inverse mandatory pressure release ventilation in one study. Extreme inverse inspiratory:expiratory (I:E) ratio was used in 18 (46%) APRV compared to 0 BIPAP studies (p =0.001); 8 (20%) APRV and 1(9%) BIPAP studies used mild inverse ratio (up to 2:1). A 1:1 ratio was used more often with BIPAP (7, 64% vs 12, 31%, p =0.06) as was a normal I:E ratio (3, 27% vs 2, 5%, p =0.04). In adult studies, mean inspiratory pressure was 24cmH2O (APRV) and 18cmH2O (BIPAP) (p=0.3). Mean expiratory pressure was 5.5cmH2O for both modes (p=0.9). Seven APRV studies described synchronisation, 3 (43%) stated the mode did not synchronise to patient effort. All 4 BIPAP studies that described synchronisation stated it was available.CONCLUSION. APRV assumes inverse ratio ventilation (IRV). Some studies advocate extreme IRV with short release times to improve gas exchange, haemodynamic stability, renal and splanchnic blood flow(1). Extreme IRV was used in only 46% of APRV studies, 31% described an I:E ratio of 1:1. Further, ventilator settings used for studies of APRV may be indistinguishable from BIPAP studies (2, 3) . Given the variation in ventilatory settings described, uncertainty of optimal settings may exist. Commercial ventilator branding may further add to confusion. Generic naming of ventilatory modes, as with drug prescribing, combined with consistent definitions of the parameters that define the modes, may avoid confusion, improve consistency of patient response and assist the implementation of these modes into clinical practice. PAV is intended to normalize neuro-ventilatry coupling by assisting each breath in proportion to patient effort, but requires reliable measurements of elastance (E) and resistance (R). PAV+ allows to (a) automatically and non invasively measure E and R, and (b) continuously adjust ventilatory support accordingly. Aim of our study was to test the physiological effects of PAV+ versus CMV (ARDSnet lung protective strategy) in a model of ARDS.

In 12 pigs ARDS was induced through chloridric acid inhalation (4 ml/Kg). At T0 (after damage) each pig was randomly assigned to PAV+ or CMV. Gas exchange and lung CT scan at 6 (T6) hours were compared with those obtained at T0 (Delta = T6-T0).

Data are mean +/-standard deviation; *) p < 0.05 PAV+ versus CMV CMV PAV+ ∆ Hyperinflated areas (cm2) 1 +/-1 3 +/-8 ∆ Normally aerated areas (cm2) -28 +/-46 468 +/-186 * ∆ Poorly aerated areas (cm2) 1 +/-4 48 +/-195 * ∆ Nonaerated areas (cm2) 10 +/-14 -83 +/-115 * ∆ PaO2/FiO2-37 +/-11 141 +/-66 * ∆ PaCO2 (mmHg) 8 +/-6 -15 +/-8 *

Our data suggest the ability of PAV+ to improve gas exchange, principally through an increase in normally aerated areas. The impact of PAV+ on ventilator induced lung injury deserves further investigation.GRANT ACKNOWLEDGEMENT. University of Bari. INTRODUCTION. The major advantage of high-frequency oscillatory ventilation (HFOV) to conventional mechanical ventilation (CMV) is delivery of smaller tidal volumes to an optimally recruited lung. Assuming there is a save window in the pressure volume curve of the lung between a lower zone with atelectasis and a upper zone with overdistension, surpassing this zone would result in either cyclic recruitment and decrecruitment, overdistension, or both. In diseased lungs this safe window may be too small to harbor the relatively large tidal volumes of CMV. CO2 removal (V'CO2) and therefore PaCO2 is a function of frequency (f) and alveolar delivered tidal volume (Vt): V'CO2 = f x Vt 2 . It is an inherent technical feature of all oscillators that Vt at maximal power decreases as frequency increases. In addition, pressure swings fall down the endotracheal tube and the airways. This fall in pressure swings is a function of frequency and mechanical properties of the respiratory system. As a result of both phenomena Vt delivered to the alveoli decreases substantially at higher frequencies. Up till now oscillation is set at a fixed frequency, in adults at 5 Hz, in children and neonates at 10 Hz. PaCO2 is regulated by adjusting the power, and thus the pressure swings (delta P) and the delivered volume. If the maximum power has been reached, decreasing the frequency can lower the PaCO2 further.

We calculated Vt required to keep V'CO2 constant at different oscillation frequencies and measured the delivered Vt at maximal power as function of frequency with the SensorMedics 3100A.

. Vt needed to keep V'CO2 constant and maximal delivered Vt can be plotted against oscillatory frequency. By increasing frequency, Vt needed to keep V'CO2 constant and maximal delivered Vt both decrease. However, a point is reached at which the required Vt to maintain V'CO2 equals the maximal delivered Vt. At this point Vt has its lowest possible value to maintain PaCO2. At higher frequencies the delivered volume of the oscillator is lower than required and PaCO2 would rise above the pre-arranged level.

We advocate a ventilatory strategy with the oscillator set at its maximal power and the frequency to be adjusted according to the PaCO2. With this strategy the lowest Vt is delivered to the alveoli with the largest safety margins between atelectasis and overdistension. 

Automatic tube compensation (ATC) compensates the resistance caused by the endotracheal tube. Tube resistance is defined by the equation Hagen-Poiseuille: R = (128 x x L) / π x r4. (R= resistance, = viscosity, L= length of the tube, r= radius of the tube). ATC is designed to lower the work of breathing in intubated spontaneous breathing patients by creating a higher initial flow and therefore a higher peak pressure. The aim of this study was evaluate the consequences of ATC during controlled mechanical ventilation without spontaneous breathing activity on peak pressure distal of the tracheal tube, in comparison to the set pressure. Moreover, the time needed to reach the set inspiratory pressure distal of the tube with and without ATC was assessed.

In an experimental laboratory setting using an artificial lung the maximum pressure in the ventilator (Draeger Evita 4), proximal and distal of the tube with and without 100% inspiratory ATC in a tube ID 7,0 and a tube ID 9,0 were measured. The time needed to reach the set inspiratory pressure distal of the tube with and without 100% inspiratory ATC were compared. Baseline ventilator settings were BIPAP, ASB 0, PEEP 8 mBar, I:E-ratio 1:1, FiO2 21%, rise time 0 seconds. A set of 20 measurements where performed for each of the following settings: Pressure constant group (PCG): Frequency of respectively: 10, 30 and 50 a minute at a fixed Pinsp of 25 mBar. Frequency constant group (FCG): Pinsp of respectively: 15, 30 and 45 mBar at a fixed frequency of 15 a minute.

No peak pressure were measured at any time distal of the tube regardless of frequency or set pressure. The pressure distal of the tube never exceeded the set pressure level in the ventilator. The time needed to reach the set inspiratory pressure distal of the tube was significant shorter during ATC. (see table) CONCLUSION. There is no danger of creating a higher pressure distal of the tube than the set inspiratory pressure at any time during the use of ATC 100% with the Draeger Evita 4. With the use of ATC the set inspiratory pressure at the distal end of the tube is reached more quickly. ATC creates a faster rise time on the tracheal level, resulting in a higher mean airway pressure.