Skip to main content

37th International Symposium on Intensive Care and Emergency Medicine (part 2 of 3)

Brussels, Belgium. 21-24 March 2017

P171 The use of extracorporeal membrane oxygenation for ventricular septal rupture complicated by refractory cardiogenic shock

D Rob, R Špunda, J Lindner, J Šmalcová, O Šmíd, T Kovárník, A Linhart, J Bìlohlávek

First Faculty of Medicine, Charles University and General University Hospital, Prague, Czech Republic, Prague, Czech Republic

Introduction: Ventricular septal rupture (VSR) is an unusual mechanical complication of myocardial infarction (MI) in the era of reperfusion therapy, but the mortality rate of patients who present with cardiogenic shock (CS) remains extremely high. Whereas current American and European guidelines recommend urgent surgical repair regardless of hemodynamic status, promising outcomes have been repeatedly reported with the use of circulatory support, enabling hemodynamic stabilization and delaying repair after consolidation of the infarct scar. Therefore, we analyzed our experience with the use of Veno-Arterial Extracorporeal Membrane Oxygenation (V-A ECMO) in post-infarction VSR.

Methods: We conducted a retrospective search of institutional database of all patients presenting with post-infarction VSR from January 2007 to June 2016. Data of 33 consecutive patients were retrospectively reviewed and analyzed.

Results: In our center, 7 out of 33 patients with post-MI VSR and refractory CS (despite vasopressor and intraaortic balloon pump therapy) received V-A ECMO support. V-A ECMO improved end-organ perfusion with lower lactate levels 24 hours after implantation (7.514 vs. 1.514, p < 0.005), normalized arterial pH (7.25 vs. 7.40, p < 0.036), improved mean arterial pressure (64 mm/Hg vs. 83 mm/Hg, p < 0.001) and lowered heart rate (115/min vs. 68/min, p < 0.001) in all patients. Mean duration of ECMO support was 12 days, 5 out of 7 patients underwent surgical repair, 4 were successfully weaned from ECMO, 3 survived 30 days and 2 survived more than 1 year. The most frequent complication (5 patients) as well as the cause of death (3 patients) was bleeding.

Conclusions: Our experience suggest that V-A ECMO support in patients with VSR and refractory CS improves end-organ perfusion, provides hemodynamic stabilization and increases time for cardiovascular team decision. Bleeding complications are an important limitation of this method.

P172 Feasibility of cerebral circulatory arrest diagnosis by TCD in VA ECMO patients

MM Marinoni1, G Cianchi2, S Trapani2, ML Migliaccio2, L Gucci2, M Bonizzoli2, A Cramaro2, M Cozzolino2, S Valente2, A Peris2

1University of Florence, Florence, Italy; 2Careggi Teaching Hospital, Florence, Italy

Introduction: The aim of our study is to investigate the feasibility of Transcranial Doppler (TCD) in Veno-arterial (VA) ECMO patients for confirmation of Cerebral Circulatory Arrest (CCA) in Brain Death (BD) diagnosis.

BD can occur in VA ECMO patients [1] and TCD is an accepted technique for BD confirmation in many countries and also in Italy [2]. In these patients, presence of Intra-aortic Balloon Pump (IABP) and residual cardiac contractility can influence TCD patterns.[3,4]

Methods: In this monocentric retrospective study TCD was performed in 5 patients evolved to BD. Left Ventricular Ejection Fraction (LVEF) values and the presence or absence of IABP were taken into account.

Results: Haemodynamic conditions of the sample study are summarized in Table 1. TCD diagnostic patterns of CCA were found in all patients in all cerebral arteries. In 2 patients TCD evaluation was available before and in CCA.

Conclusions: In the case of BD, TCD seems to be a reliable instrumental test for CCA diagnosis in patients on VA ECMO treatment with a pulsatile flow (native or IABP support).


1. Pokersnik JA et al. J Card Surgery 27: 246–252, 2012

2. National Transplantation Council Guidelines, 2009

3. Yang F et al. J Transl Med 12:106, 2014

4. Kavi T et al. Journal of Stroke and Cerebrovascular Diseases, 2016

Table 1 (abstract P172). Haemodynamic and TCD findings of patients evolved to BD

P173 A dysbalance in t-cell response predicts in-house mortality in VA-ECMO patients

E Grins1, E Kort2, M Weiland3, N Manandhar Shresta2, P Davidson3, L Algotsson1, S Fitch2, G Marco2, J Sturgill2, S Lee2, M Dickinson2, T Boeve2, A Khaghani2, P Wilton2, S Jovinge2

1Scania Univ Hospital Lund, Lund, Sweden; 2Spectrum Health Hospitals, Grand Rapids, MI, United States; 3Van Andel Institute, Grand Rapids, MI, United States

Introduction: ECMO treatment has continuously been associated with high mortality. CD4/CD8 T-cell ratio has been used to monitor loss in immune function HIV patients and high ratio has been reported in allograft connected to worse outcome.

Methods: Patients (n = 51) eligible for VA ECMO treatment at Meijer Heart Centre in Grand Rapids Michigan were consented through by themselves or their Legal Representatives. Blood was drawn before the patient was cannulated.

Results: Non-survivors had higher CD4/CD8 p < 0.0266 than non-survivers. IN ROC regression CD4/CD8 ratio performed second best AUC 0.72

Conclusions: A more aggressive T-cell activation as reflected by CD4/CD8 ratio is related to mortality.

Fig. 1 (abstract P173).

CD4/CD8 ratio Survivors vs Non-Survivors

Fig. 2 (abstract P173).

ROC Regression plot predictors of ECMO Mortality

P174 ENCOURAGE…ing results for veno-arterial ECMO in myocardial infarction

AN Ahmad, R Loveridge, S Vlachos, S Patel, E Gelandt, L Morgan, S Butt, M Whitehorne, V Kakar, C Park, M Hayes, C Willars, T Hurst, T Best, A Vercueil, G Auzinger

King´s College Hospital, London, United Kingdom

Introduction: The ENCOURAGE score is a mortality risk score for acute myocardial infarction (AMI) patients treated with veno-arterial extracorporeal membrane oxygenation (VA-ECMO) using pre-ECMO parameters. [1]

Only 41% of patients in the ENCOURAGE dataset survived six months with 20% of them requiring VAD or heart transplant.

This study aims to evaluate our 6-month outcomes against those predicted by the score.

Methods: All consecutive patients receiving VA-ECMO post AMI between 2012 and 2016 had data prospectively collected for analysis and an ENCOURAGE score calculated, using the seven parameters that comprise the score: age, sex, BMI, GCS, creatinine, lactate and prothrombin activity.

Patient outcomes using six-month follow-up data were compared and a standardized mortality ratio (SMR) calculated.

Results: 12 patients were included. Median ENCOURAGE score was 27 (IQR 6) with 83% in the worst risk classes. Predicted survival of our cohort according to the score was 18%.

42% were initiated as eCPR, 92% were after cardiac arrest (45% OHCA, 55% IHCA), mean lactate was 11.9 (SD + - 6.7), median SOFA 16 (IQR 4).

75% were ECMO survivors and 50% were alive at six months without VAD or transplantation, all with good Cerebral Performance Category scores.

The ENCOURAGE SMR is 0.611 (95% CI 0.25-1.27).

Conclusions: Risk prediction models that use pre-ECMO criteria can be used to assess the performance of ECMO centres [2,3] and this technique can be extended to sub-group analysis.

Previous studies have suggested poorer outcomes (36% [4], 21% [5]), in line with the ENCOURAGE dataset but the service exceeds predicted outcomes, and we recommend against restrictive criteria for the use of VA-ECMO.

The UK also needs to consider destination therapy for those surviving ECMO but not ICU.


[1] Muller G et al. Intensive care medicine. 2016;42(3):370–8.

[2] Chaddock et al. EuroELSO 2016

[3] Loveridge et al. Crit Care 20:94. 2016

[4] Boqambar et al. EuroELSO 2016

[5] Francis et al. EuroELSO 2016

P175 Veno-arterial extracorporeal membrane oxygenation for cardiac support: a single center experience

B Adibelli, N Akovali, A Torgay, P Zeyneloglu, A Pirat, Z Kayhan

Ankara Baskent Hospital, Ankara, Turkey

Introduction: Veno-arterial extracorporeal membrane oxygenation (VA ECMO) ensures end-organ perfusion while fully replacing heart and lung function to allow time for possible heart recovery or may bridge patients to heart transplantation or ventricular assist device (VAD) implantation. We review our 4-year experience regarding VA-ECMO use for cardiac support including early, midterm outcomes and survival.

Methods: This is a retrospective analysis of patients undergoing VA-ECMO for cardiac support from January 2012 to December 2015. ECMO was performed through the femoral vessels percutaneously with ultrasound guidance in 39 of them and through the right atrium to ascending aorta in 7 of them during cardiac surgery. VA-ECMO was applied during or after cardiac surgery in 30 (65%), not related to surgery in 16 (35%) out of 46 patients. Extracorporeal CPR was done in 6 patients.

Results: A total of 46 patients were supported with VA-ECMO. Mean age of patients was 47.7 ± 20.9 years with 67.7% males. The leading diagnoses were dilated cardiomyopathy, heart failure and coronary artery disease. Mean duration of VA-ECMO support was 226.9 ± 283.4 hours. Overall, 14 (30.4%) patients were successfully weaned off ECMO and survived, 32 (69.6%) patients died due to multifactorial complications. Overall, 12 patients were bridged to heart transplantation and 8 patients to left VAD. The 30-day and 6-month survival rates were 41% and 20%, respectively. Requirement for mechanical ventilation before and after VA-ECMO implantation was significantly less in patients who survived (p < 0.05).

Conclusions: In cases of cardiomyopathy refractory to medical treatment, failure to wean off cardiopulmonary bypass and refractory shock post cardiac arrest, circulation can be supported with VA-ECMO. Our experience among 46 patients implanted with VA-ECMO for cardiac support, survival to discharge was 30%. Survival was significantly better in those patients who did not require mechanical ventilation before and after VA-ECMO implantation.

P176 Use of automated chest compression devices after out-of-hospital cardiac arrest in Sweden

SS Schmidbauer1, J Herlitz2, T Karlsson3, H Friberg1

1Skåne University Hospital, Lund, Lund, Sweden; 2PreHospen, University of Borås, The Pre-hospital Research Centre of Western Sweden, Borås, Sweden; 3Health Metrics, Institute of Medicine, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden

Introduction: Although shown not to increase survival rates, automated chest compression (ACC) devices are frequently utilised after out-of-hospital cardiac arrest (OHCA) in Sweden. With no implementation guidelines available, it is not known how these devices are put to use. In this retrospective observational study, we evaluated the utilisation of ACC devices in Sweden between the years 2011-2015. The association between ACC-CPR and 30-day survival was also assessed.

Methods: The Swedish Cardiopulmonary Resuscitation Registry is a prospectively recorded nationwide registry of modified Utstein-style parameters with a coverage of close to 100% of OHCA cases where resuscitation was attempted. Propensity score matching (PSM) and logistic regression with multiple imputation (MI) were used to study the association between ACC-CPR and survival.

Results: During the study period, 25898 patients were identified in the registry. After exclusions, 24316 were included in the study population. Of these, 32.4% received ACC-CPR. Overall, unadjusted 30-day-survival was 6.3% in the ACC-CPR group, 12.8% in the manual CPR group and 10.7% for the entire study population. Male gender and an initial shockable rhythm were factors associated with ACC device use, whereas crew witnessed status was associated with manual CPR. Administration of adrenaline and antiarrhythmics was also more prevalent in the ACC-CPR group, and so was intubation (Table 2).

The odds ratio for 30-day survival regarding ACC device utilisation was 0.72 (95% CI 0.62-0.84), p < .001 by means of PSM (n = 13922). Similar results were seen using stratification on the PS (n = 20633) as well as logistic regression with MI (data not shown).

Conclusions: Automated chest compression devices are frequently used after OHCA and predominantly so for patients with a more refractory condition. Their use might be associated with lower survival rates.

Table 2 (abstract P176). baseline characteristics and treatment data

P177 Mechanical ventilation during CPR

R Knafelj, P Radsel

Rihard Knafelj, Ljubljana, Slovenia

Introduction: Guidelines for mechanical ventilation (MV) during cardiopulmonary resuscitation suggest low frequency (f) with normal tidal volume (Vt) avoiding hyperventilation and hyperinflation. During inhospital cardiac arrest in patients that are already mechanically ventilated, optimal ventilatory strategy is not known. We hypothesized that using CPR ventilation mode results in better Vt and (f) compared to bag ventilation, volume (V-AC) or pressure (P-AC) ventilation.CPR mode has pre-configured settings (FiO2 1.0,PEEP 5/20 cmH2O, f 12/min), alarms are deactivated, ventilation is synchronized with chest compressions

Methods: 40 tests were performed. Ventilation with V-AC (450 mL, PEEP 5 cmH2O, max pressure limit 30 cmH2O) was set (Elisa 800,HL,Germany). After chest compressions started group 1remained in V-AC, group 2 was switched to bag ventilation, group 3 to BIPAP, group 4 to CPR mode. During chest compressions changes in settings but not in mode were allowed. Vt, f of ventilation and chest compressions rate were measured in all groups

Results: During CPRf of chests compression did not differ across groups (118 ± 4, 110 ± 12, 113 ± 10, 112 ± 8 group 1, 2, 3, 4 respectively). Vt were significantly lower in groups 1, 3 and 4 (87 ± 47, 48 ± 71, 268 ± 46 mL respectively) compared to group 2 (1139 ± 133, p < 0.005) due to reached high pressure limit (group 1, 3, 4) or excessive bagging (group 2). Group 4 received higher Vt compared to group 2 and 3 (p < 0.005). Ventilation f was higher in groups 1, 2 and 3 (18 ± 3, 21 ± 4, 20 ± 3 respectively) compared to group 4 (12 ± 0p, <0.005)

Conclusions: CPR ventilation mode prevents hypo/hyperventilation and hyperinflation compared to other modes. Better guidelines compliance was demonstrated for chest compression rate. Clinical impact of newly implemented CPR ventilation mode warrants further studies.


Soar J et al. ALCS. Resuscitation.2015.95:100–47

Fig. 3 (abstract P177).

CPR mode

Fig. 4 (abstract P177).

Vt during CPR

P178 New system to control FDO2 with bag valve mask for premature infants

F Duprez1, T Bonus1, G Cuvelier2, S Mashayekhi1, M Maka1, S Ollieuz1, G Reychler3

1Epicura, Hornu, Belgium; 2Condorcet, Tournai, Belgium; 3UCL, Bruxelles, Belgium

Introduction: According to the recommendations of the European Resuscitation Council (ERC), Cardio Pulmonary Resuscitation (CPR) in premature infants must be made with a fraction of delivered oxygen (FDO[sub]2[/sub]) not exceeding 30%. Bag valve masks for premature infant (BVMp) can be used for ventilation and oxygenation during CPR. In such a case, even with a low oxygen flow rate (OFR), a BVMp delivers higher FDO[sub]2[/sub] than recommended. Indeed, with a BVMp, FDO[sub]2[/sub] rises proportionally to OFR but decreases inversely proportionally to minute ventilation (MV). Therefore, in neonatology resuscitation, controlling and maintaining the FDO[sub]2[/sub] below 30% is very difficult, even with a low OFR. To meet the ERC recommendations, we developed a new system aimed at delivering adequate FDO[sub]2[/sub] with a BVMp: the DupRey system. This system uses the Venturi effect to provides a stable air-oxygen mixture to a BVMp. The present study was aimed at evaluating the actual O[sub]2[/sub] fraction delivered between bag valve masks for premature infant used conventionally or used with the DupRey system.

Methods: On a bench study, a BVMp (Laerdal™ for premature infant type 850150) was connected to a test lung (Maquet™ VA800 - compliance 0,02 L/cm H[sub]2[/sub]O - resistance 20 cm H[sub]2[/sub]O/L/sec. With the BVMp, two MV (0.7 L/min and 1 L/min) were generated. A metronome gave the frequency of insufflations. The BVMp was tested both with and without oxygen reservoir (OR) and the pop off valve was closed. Two OFR: 0.6 and 1 L/min were analyzed and compared to the DupRey with Venturi 24% and 28% (OFR: 5 L/min). OFR were analyzed by a thermal mass flow meter Vogtlyn™ Red Y. The FDO[sub]2[/sub] and MV measurements were made using an analog iWorx® acquisition system (GA207 gas analyzer associated with digital IWx 214) and LabScribe II ® software.

Statistical: One Way Repeated Measures Analysis of Variance followed by Holm Sidak method.

Results: Statistical differences (p < .001) were found between A5-A6 / B5-B6 and all other columns. However, no statistical differences were found between: B2 and A2, A6 and B6, B5 and A5.

Conclusions: For an oxygen flow ranging from 0.6 to 1 L/min and two MV analyzed, a Bag Valve Masks for premature infant with an OR delivers very high FDO[sub]2[/sub] (>86%). Without OR, at same OFRs, FDO[sub]2[/sub] decreases but they maintain high values (>44%). The DupRey delivers FDO[sub]2[/sub] < to 30% regardless MV. The DupRey system is easily accessible for medical teams who do not have access to modern technology.


WyllieJ. ERC Guidelines for Resuscitation 2015: Resuscitation. 2015 Oct. 95: 249–63.

Fig. 5 (abstract P178).

Fraction delivered in O2 between BVMp (with and without oxygen reservoir) and DupRey system

P179 Could echocardiography have prognostic value in patients after successful resuscitation?

R Mosaddegh, S Abbasi, S Talaee

Iran University of Medical Sciences, Tehran, Iran

Introduction: After successful cardio pulmonary resuscitation, many patients have poor outcome, because of current illness that causes cardiopulmonary arrest, or complications of resuscitation or any other problem. Researchers want to know if echocardiography have prognostic value in these patients.

Methods: In this case series the researchers enroll 50 patients with successful resuscitation in three general hospitals. Echocardiography was done for all patients without considering the duration of resuscitation in 24 hours after ROSC and ventricular and septal wall motion was observed for hypokinesia, akinesia or dyskinesia. Ejection Fraction (EF), E-Point Septal Separation (EPSS), Inferior Vena Cava (IVC) diameter and presence of Hepatic Portal Vein Gas (HPVG) and 24 hour's survival was measured. The relation between outcome and these echocardiographic findings were observed then.

Results: Twenty eight participants were survived and 22 were died. The median age of participants was 55.52 (SD: ±23.57) years. The mean EF for all participants was 26.74 ± 18.26 percent. The mean EF was 27.82 (SD ± 15.79) in survivors and 25.89 (SD ± 20.23) without statistically significant difference (P < 0.05). Ventricular wall motion, HPVG presence and IVC diameter has no statistically significant difference in both survivors and non survivors (P < 0.05).

Conclusions: Based on the results of this study it seems that echocardiographic findings in the first 24 hours of post resuscitation period could not help to predict the prognosis of survivors of cardio-pulmonary arrest, and it's not reasonable to send the post resuscitative patients to the other wards to do echocardiography in the first 24 hours after resuscitation.

P180 Diagnostic and therapeutic value of coronary angiography and ct-scan after extracorporeal cardiopulmonary resuscitation (eCPR) – a single center registry study

VZ Zotzmann, DS Staudacher, TW Wengenmayer, DD Dürschmied, CB Bode

Heart Center, Freiburg, Germany

Introduction: Implantation of a venoarterial extracorporeal membrane oxygenation (ECMO) in patients with ongoing cardiopulmonary resuscitation without return of spontaneous circulation (eCPR) can stabilize hemodynamics. Further diagnostic work up is needed in order to diagnose and treat the cause of the collapse. Patients after eCPR compromise a heterogeneous population with more severe underlying pathologies when compared to patients with return of spontaneous circulation. Aim of this study is the evaluation of the diagnostic value of coronary angiography and CT-scan after eCPR.

Methods: All patients after eCPR treated at a single tertiary referral hospital between December 2010 and November 2015 were included in a retrospective registry study.

Results: A total of 123 patients were considered (age 59.5 ± 15.3 years, low-flow time 59.0 ± 28.2 min, survival 11.4%). 52 patients presented with non-shockable rhythm (age 63.8 ± 16.1 years, low-flow duration 51.0 ± 23.1 min, survival 15.4%) while 71 patients presented with either a shockable rhythm or ST-elevation (age 56.3 ± 14.0 years, low-flow duration 64.8 ± 30.2 min, survival 8.5% p < .01, <.01 and .03, respectively).

Coronary angiography was performed significantly less frequent in patients with non-shockable rhythm (59.6% vs. 93.0% p < .01), see Fig. 6. A lesion deemed responsible for collapse however was found at similar rates in both groups (71.0% with non-shockable rhythm vs. 83.3% with ST-elevation or shockable rhythm, p = .18).

CT-scan was performed at similar rates in both groups (65.4% vs 50.7%, p = .14). Pathologies deemed responsible for collapse however were found more often in patients with non-shockable rhythm (23.5% vs. 0%, p < .01). CT-Scan yielded findings relevant to the further treatment frequently in both groups (92.2% vs 91.6%, p = 1). Cause of collapse could be detected by CT-scan at significantly lower rates when compared to coronary angiography in both groups (23.5% vs.71.0% and 0% vs. 83.3%, both p < .01).

Conclusions: Coronary angiography yielded a significantly better diagnostic value than CT-scan after eCPR disregarding initial rhythm or presence of ST-elevation. Considering the potential therapeutic option, a coronary angiography first approach might be preferable. A routine CT-scan however might be reasonable in all patients since findings relevant to the further treatment are frequent.

Fig. 6 (abstract P180).

Diagnostic Of coronary angiography and CT-scan in patients after eCPR

P181 Comparison of FiO2 50% or 100% on brain oxygenation and cardiac mitochondrial function in experimental cardiac arrest

A Nelskylä1, J Nurmi2, M Jousi2, A Schramko1, E Mervaala2, G Ristagno3, M Skrifvars1

1Helsinki University Hospital, Helsinki, Finland; 2Helsinki University Hospital and University of Helsinki, Helsinki, Finland; 3Istituto di Ricerche Farmacologiche “Mario Negri”, Milan, Italy

Introduction: Guidelines advocate 100% oxygen during CPR. We hypothesized that 50% oxygen during CPR maintains cerebral oxygenation and compared to 100%, alleviates cardiac mitochondrial injury.

Methods: With Finnish National Animal Experiment Board (ESAVI/1077/04.10.07/2016) approval, ventricular fibrillation (VF) was induced electrically in anaesthetized adult pigs and left untreated for 7 minutes, followed by randomization to mechanical CPR (LUCAS) with 50% or 100% oxygen. Defibrillation was performed at 13 minutes and if unsuccessful, CPR continued for 20 minutes with defibrillation and 1 mg adrenaline every 2 minutes. Cerebral oxygenation was measured with near-infrared spectroscopy (rSO2, INVOS™ 5100C Cerebral Oximeter) and invasive brain tissue oxygen (pbO2) with an intraparenchymal probe (NEUROVENT-PTO, RAUMEDIC) in the frontal cortex. A heart biopsy was obtained 20 minutes after ROSC with analysis of mitochondrial respiration (OROBOROS Instruments Corp., Innsbruck, Austria) and compared to 4 control animals. Data are shown as mean with standard deviation (SD). Brain rSO2 and pbO2 were compared between groups over time with mixed linear model with mean arterial blood pressure (MAP) as covariate. Mitochondrial respiration was compared with analysis of variance.

Results: Of 20 pigs, one had a breach of protocol and were excluded leaving 9 in the 50% group and 10 in the 100% group. Groups were similar regarding rSO2 and pbO2 before CPR. With a median time of 15 minutes, 6 pigs achieved ROSC in the 50% group and 8 pigs in the 100% group. During resuscitation time was significantly associated with rSO2 (p < 0.001) and pbO2 (p < 0.001). The rSO2 (0.012) was lower with FiO2 50%, but pbO2 was not (p = 0.43). Mean arterial pressure was associated with pbO2 (p = 0.045). After ROSC the rSO2 (p < 0.001) and pbO2 (p < 0.001) increased significantly over time without difference between oxygen groups (p = 0.37, p = 0.18). Compared to controls mitochondrial respiration was decreased with adenosine diphosphate (ADP) levels of 57 (17) Pmol/sec/mg compared to 92 (23) Pmol/sec/mg (p = 0.008), no difference was seen between oxygen groups (p = 0.79).

Conclusions: Use of 50% oxygen during CPR results in lower oxygen content in blood, but brain tissue oxygenation can be maintained with efficient CPR. After ROSC brain oxygen increases rapidly. Cardiac arrest results in disturbed cardiac mitochondrial respiration but this is not alleviated with the use of 50% oxygen.

P182 Pediatric extracorporeal cardiopulmonary resuscitation experiences of a pediatric intensive care unit

G Ozsoy, T Kendirli, E Azapagasi, O Perk, U Gadirova, E Ozcinar, M Cakici, C Baran, S Durdu, A Uysalel, M Dogan, M Ramoglu, T Ucar, E Tutar, S Atalay, R Akar

Ankara Univercity, Ankara, Turkey

Introduction: Extracorporeal membrane oxygenation (ECMO) has been used during cardiopulmonary resuscitation (CPR) to improve the outcomes in selected patients. According to the international registry of the Extracorporeal Life Support Organization (ELSO), more than 1232 (41%) children have benefited from ECMO-assisted CPR (ECPR) since its introduction. We want to summarize our ECPR experiences in our unit.

Methods: This is a retrospective study performed in the 15-bed tertiary Pediatric Intensive Care Unit (PICU) at Ankara University Hospital. The children who underwent ECPR from September 2014 to August 2016 were assessed.

Results: Eight children underwent ECPR in our unit. Their median age and weight were 80.5 months and 35 kilograms respectively. They were all hospital-arrest. Their primary diseases were 62.5% cardiac and 37.5% non-cardiac. The causes of cardiac arrest were dysrhythmia (37.5%), heart failure (25%), sepsis (12.5%), bleeding (12.5%), low cardiac output syndrome (12.5%) and airway disease (12.5%). ECPR was performed in PICU (6) and in operation room (2). ECMO was started in the 78 minutes (median) of CPR. The time of ECPR was 85 minute (median). ECMO cannulas were placed in femoral vein-femoral artery (5), central (2), and internal jugular vein-carotid artery (1). The cannulas were not able to place in two patients. ECMO did not work in two patients which cannulated central and femoral vein-femoral artery. ECMO succeeded in 4 (50%) patients. The range of the ECMO time after ECPR was 12 hours and 9 days.

Conclusions: ECPR improves survival after cardiac arrest, especially in the patients who might benefit from this treatment.

P183 Midazolam is an independent risk factor for prolonged awakening after cardiac arrest

M Kamps1, G Leeuwerink2, J Hofmeijer2, O Hoiting3, J Van der Hoeven1, C Hoedemaekers1

1Radboudumc, Nijmegen, Netherlands; 2Rijnstate, Arnhem, Netherlands; 3Canisius Wilhelmina Ziekenhuis, Nijmegen, Netherlands

Introduction: Neurologic prognostication after cardiac arrest is a delicate process with lots of confounders that influence the outcome of these patients. Sedation is known as one of the major confounders. We assessed the hypothesis that prolonged awakening after cardiac arrest is mainly caused by the use of long-acting sedation.

Methods: This is a retrospective, multicentre, cohort study. We studied patients with ROSC after cardiac arrest in three different hospitals. We studied variables such as the cumulative dosage of sedatives, time of awakening after cessation of sedation, renal function, targeted temperature and GOS score after six months. Renal function was defined as the RIFLE score at admission. Early awakening was defined as awakening within 48 hours after cessation of sedation. The Glasgow coma scale was used to score awakening, with a motor score of 6 with an eye score of 3 or 4 defined as awakening. The patients with good neurologic outcome were divided in an early and late awakening group. Good neurologic outcome was defined as a GOS score of 4 or 5 after six months.

Results: We studied 122 patients with a good neurologic outcome after six months. Demographic variables such as age, weight, ROSC and targeted temperature were similar in the early and late awakening group. In the late awakening group 92% of the patients were treated with midazolam at day 0 compared to 57% in the early awakening group (p = 0.021). At day 1 83% of the patients with late awakening were treated with midazolam vs 56% in the early awakening group (p = 0.063). The cumulative dosage of midazolam did not differ between both groups (169 mg/24 hr in the early awakening group and 188 mg/24 hr in the late awakening group P = 0.613). Renal function was similar in both groups, RIFLE 0 in 81.8% in the early awakening group vs 67.7% in the late awakening group (P = 0.269) The patients in the early awakening group were significantly more often treated with propofol 74.5% vs 41.7% the late awakening group (P = 0.017) The dosage was significantly higher in the early awakening group. (2466 mg/24 hr in the early awakening group vs 643 mg/24 hr late awakening group p = 0.05)

Conclusions: In conclusion the use of midazolam, independent of the dosage or renal function, is a risk factor for prolonged awakening after cardiac arrest. Thereby midazolam could be a major confounder in the prognostication of neurologic outcome. The current protocol advises the use of opioids and hypnotics. We suggest that, if possible, patients after cardiac arrest are treated with short-acting sedation only to prevent inaccurate neurologic prognostication due to sedation effects.

P184 The influences of ketamine or morphine on hemodynamics, acid-base status and early survival in rats after asphyxia cardiac arrest: a pilot study

A Konkayev1, V Kuklin2, T Kondratyev3, M Konkayeva1, N Akhatov1, M Sovershaev4, T Tveita3, V Dahl2

1Astana Medical University, Astana, Kazakhstan; 2Akershus University Hospital, Oslo, Norway; 3The Arctic University of Norway, Tromsø, Norway; 4University Hospital of Northern Norway, Tromsø, Norway

Introduction: Acute hypoxia results in uncontrolled release of glutamate and the consequent stimulation of NMDA receptors, which affects the whole ionic homeostasis and finally activates apoptosis of neurons [1]. A potential therapeutic approach to prevent this sequence of events is a blockade of NMDA receptors. Meanwhile, in different models of acute hypoxia, activation of delta-opioid receptors by morphine demonstrates cardioprotective effect with a consequent increase in animal survival [2]. Thus, we aimed to test the effects of morphine or ketamine on hemodynamics, acid-base status and early survival in rats after asphyxia cardiac arrest (ACA)

Methods: After instrumentation under anaesthesia with Thiopental sodium (60 mg/kg, i.p.), Wistar rats (n = 21) weighing between 350–400 g were randomly assigned to three groups where: 1. Morhpine 5 mg/kg iv (n = 7) was given 10 min before ACA; 2. Ketamine 40 mg/kg iv (n = 7) was given 10 min before ACA; 3. Control (n = 7), the same amount of NaCl 0,9% iv was given 10 min before ACA. The rats were asphyxiated by clamping the tracheostomia tube at the end of expiration for 5 min. Resuscitation included epinephrine (0.02 mg/kg, iv), manual thoracic compressions (180 per min) and mechanical ventilation (21% O2, 80 breaths/min). Invasive MAP was recorded at the baseline (BL), every 1 min during ACA and every 5 min in post-resuscitation (PR) period. Blood gas samples were taken at the BL and 10 min at the PR period. Early survival was determined at the 20 min after ACA.

Results: No differences in MAP between the rats was found at the BL period. The rats pre-treated by ketamine got significantly higher MAP during PR period (133.9 ± 30.4 vs 52.7 ± 37.3 and 60.0 ± 26.2 mm Hg, respectively, p < 0.002) and had significantly lower production of lactate (11.8 ± 2.2 vs 13.5 ± 1.5 and 15.6 ± 1.2 mmol/l, respectively, p < 0.002) when compared to the rats treated by morhpine and only saline. Six of the seven rats survived at the 20 min after ACA in the ketamine group while four of the seven and two of the seven rats survived in the morhpine and Control groups respectively (P = 0.122).

Conclusions: Pre-treatment with ketamine attenuated significantly disturbances in hemodynamics and lactate after ACA, but it did not improve early survival when compared to the rats pre-treated by morphine or saline.


1. Choi DW. Neuron 1988;1:623–34

2. Endoh H, et al. Crit Care Med 2001;29:623–7



P186 Ubiquitin c-terminal hydrolase l1 as a predictor of neurological outcome after cardiac arrest and resuscitation

L Wihersaari1, MB Skrifvars2, S Bendel3, KM Kaukonen4, J Vaahersalo4, J Romppanen5, V Pettilä4, M Reinikainen1

1North Karelia Central Hospital, Joensuu, Finland; 2Australian and New Zealand Intensive Care Research Centre, School of Public Health and Preventive Medicine, Monash University, Melbourne, Australia; 3Kuopio University Hospital, Kuopio, Finland; 4Helsinki University and Helsinki University Hospital, Helsinki, Finland; 5Eastern Finland Laboratory Centre, Kuopio, Finland

Introduction: Ubiquitin C-terminal hydrolase L1 (UCHL1) is an enzyme present in central nervous system neurons. We aimed to assess UCHL1 as a predictor of neurological outcome after cardiac arrest in comparison with neuron-specific enolase (NSE) in this FINNRESUSCI substudy [1].

Methods: We prospectively collected data on 249 patients who were admitted to 21 intensive care units after out-of-hospital cardiac arrest (OHCA) between March 1, 2010 and February 28, 2011. Of these patients, 177 (71%) had a shockable initial rhythm. We measured serum concentrations of UCHL1 and NSE at 24 h and 48 h after cardiac arrest. UCHL1 concentrations were analysed by a commercial ELISA kit and NSE concentrations were measured by electrochemiluminescence immunoassay. We evaluated the ability of these biomarkers to predict poor outcome (defined as Cerebral Performance Category 3-5, indicating death or severe neurologic deficits) at 12 months after cardiac arrest using the area under the receiver operating characteristic curve (AUROC).

Results: Overall, 121 patients (49%) had a poor outcome at 12 months. For both UCHL1 and NSE, the concentrations were higher for patients with poor outcome than for those with good outcome (Table 3). The NSE concentration at 48 h was the best predictor of poor outcome (AUROC 0.72).

The median time from cardiac arrest to return of spontaneous circulation (ROSC) was 20 min. The prognostic performance of NSE at 48 h was particularly good for patients with the time to ROSC longer than 20 min, with AUROC 0.80 (95% CI, 0.71-0.89). For patients with ROSC under 20 min, the AUROC for NSE at 48 h was only 0.53 (0.41-0.66), but the AUROC for UCHL1 at 24 h was 0.70 (0.59-0.81).

Conclusions: Post-cardiac arrest UCHL1 concentrations are higher in patients with poor outcome than in those with good outcome, but the ability of UCHL1 to predict long-term outcome is weaker than that of NSE in the overall population of ICU-treated OHCA patients. However, for patients with ROSC under 20 min, the prognostic performance of UCHL1 at 24 h was satisfactory.


1. Vaahersalo J et al. Therapeutic hypothermia after out-of-hospital cardiac arrest in Finnish intensive care units: the FINNRESUSCI study. Intensive Care Med 39: 826–37, 2013.

Table 3 (abtsract P186). Concentrations (ng/ml) presented as means ± standard deviations.

P187 Time to awakening after cardiac arrest and target temperature management

A Lybeck1, T Cronberg1, N Nielsen2, H Friberg1

1Skane University Hospital, Lund University, Lund, Sweden; 2Helsingborg Hospital, Lund University, Helsingborg, Sweden

Introduction: In this post hoc analysis of the target Target Temperature Management-trail (TTM-trail) [1], we investigate the time until awakening and its relationship to target temperature and neurological outcome. Sedation is also compared.

Methods: The TTM-trial randomized 950 patients to a target temperature of 33 °C (TTM33) or 36 °C (TTM36) in 36 hospitals in 12 countries, with no difference in survival or neurological outcome between groups. Awakening was defined as Glasgow Coma Scale motor score (GCS-M) 6 in the ICU. Neurological outcome was assessed using the cerebral performance category scale (CPC) at 180 days. Cumulative doses of sedative drugs (propofol, midazolam, fentanyl, morphine, remifentanil) were retrospectively collected at 12, 24 and 48 hours. There was a strict protocol for prognostication and withdrawal of care.

Results: 496 patients had registered awakening in the ICU. Day of awakening occurred later in TTM33 (median 4, IQR 3-6) vs TTM36 (median 4, IQR 3-5), p < 0.0021 (Mann-Whitney-U). The latest recorded awakening was at day 22. We found a correlation between day of awakening and neurological outcome (Spearmans correlation coefficient 0.20, p < 0.0001), but there was no difference in neurological outcome between treatment groups (p = 0.21, Chi-squared). Doses of sedative drugs were available for 352 patients from 20 trial sites. We found no difference in doses of sedative drugs at 12, 24 or 48 hours between TTM33 and TTM36.

Conclusions: Time to awakening was longer in TTM33 than in TTM36. Day of awakening correlated with neurological outcome. In patients who awoke, there was no difference in neurological outcome between treatment groups.


1. Nielsen, N et al. N Engl J Med 369(23): 2197–206, 2013.

P188 Target temperature management in comatose survivors of cardiac arrest - comparison of endovascular, esophageal and surface cooling

M Rauber, K Steblovnik, A Jazbec, M Noc

University Medical Centre Ljubljana, Ljubljana, Slovenia

Introduction: Target temperature management represents important part of post-resuscitation care in comatose survivors of out of hospital cardiac arrest (OHCA). Early induction of hypothermia, tight maintenance and prevention of hyperthermia during rewarming appear to be essential to maximize neuroprotection. The aim of our study was to compare endovascular, esophageal and surface cooling.

Methods: Endovascular cooling (ENC) using InnerCool Accutrol Catheter (Philips Healthcare, San Diego, CA, USA), esophageal cooling (ESC) using dedicated device (Advanced Cooling Therapy, Chicago, IL, USA) with concomitant 0.9% saline (0-4 °C, 30 ml/kg in 30 min) and surface cooling (SFC) with ice packs and concomitant cold saline were compared. Target temperature was 32-34 °C. After 24 hours of maintenance, gradual rewarming targeted at 0.1-0.2 °C/h was performed. Core body temperature was measured continuously by thermistor placed intravesically.

Results: A total of 22 OHCA patients were included. ENC was used in 7 patients, ESC in 5 patients and SFC in 10 patients (Fig. 7). By extrapolating temperature curves, mean time from initiation of cooling to 34 °C was 1.1 h in ENC group, 2.8 h in ESC group and 2.1 h in SC group (p = 0.08). Variation of temperature during the maintenance phase expressed as mean hourly standard deviation was 0.26 °C in ENC, 0.42 °C in ESC and 0.76 °C in SFC groups, respectively (p < 0.001). The percentage of patients with post-rewarming hyperthermia (>38 °C) at any measurement was 43% in ENC group, 20% in ESC group and 50% in SFC group (p = 0.3). The post-rewarming hyperthermia in ENC and ESC groups was linked with device removal prior to 72 hours post OHCA.

Conclusions: Our non-randomized comparison indicates that ENC provides the fastest induction of hypothermia and best temperature maintenance. Significant proportion of patients still experience temporary hyperthermia during rewarming regardless of the cooling method.

Fig. 7 (abstract P188).

Mean patients’ temperatures after ICU admission for each cooling method.

P189 Influence of an esophageal cooling and warming device on patient temperature in the operating room

P Kalasbail1, F Garrett2, E Kulstad3

1Cleveland Clinic, Cleveland, OH, United States; 2Garrett Technologies, Northbrook, IL, United States; 3Advocate Christ Medical Center, Oak Lawn, IL, United States

Introduction: Temperature management is important for a number of clinical conditions, and a variety of methods exist to effect changes in body temperature. In general, these methods are divided into external (surface) or internal/core (intravascular) approaches. A new device (the ECD, or Esophageal Cooling/Warming Device) utilizes a closed circulation of cooled or warmed water via the esophageal route to achieve patient heat transfer, offering core temperature control without the need to access the intravascular space. We sought to measure the influence of this device on the temperature of patients undergoing surgical procedures in the operating room.

Methods: This was a secondary analysis of data from a prospective interventional study of patients undergoing elective non-cardiac surgery at the Cleveland Clinic, USA. The primary outcome was quantification of heat transferred to or from patients. Under this IRB approved study, after written informed consent, patients were subjected to two distinct 30-minute periods of either warming or cooling (with the order randomized), during which inlet temperature, outlet temperature, and flow rate through the ECD was measured, providing the data necessary for heat balance calculations. In this secondary analysis we examined patient temperatures recorded during the study, and determined mean temperature change over each 30 minute interval. Temperatures were recorded using zero-flux cutaneous thermometry (3 M SpotOn).

Results: Nineteen patients were enrolled in this study from April to November, 2016. During the warming cycle, mean patient starting temperature was 35.5 C +/- 0.52 C, increasing to 35.8 +/- 0.62 C over 30 minutes. The mean warming rate was 0.56 +/- 0.64 C/hr. During the cooling cycle, mean patient starting temperature was 35.7 C +/- 0.61 C, decreasing to 35.3 +/- 0.61 C over 30 minutes. The mean cooling rate was 0.88 +/- 0.87 C/hr. The finding that cooling was faster than warming may be attributed to the greater difference in core-to-perfusion temperature (water temperature of 7 C cooling, 42 C warming); however, extraneous factors such as patient exposure, body habitus, external warming, and room temperature, may also have influenced performance.

Conclusions: Although patients differed slightly in their rates of temperature change, these data suggest that patient temperature modulation utilizing the esophageal route is effective, and that clinically meaningful core patient temperature change can be attained in both cooling and warming, without the need to access the intravascular space.


Markota A, Fluher J, Balazic P, Kit B, Sinkovic A: Therapeutic hypothermia with esophageal heat transfer device. Resuscitation 2015, 96:138.

P190 Outcomes and interventions among out-of-hospital cardiac arrest patients transported to hospital with ongoing cardiopulmonary resuscitation

DJ Bergström, HR Olsson, S Schmidbauer, H Friberg

Skåne University Hospital, Lund, Lund, Sweden

Introduction: The introduction of automatic chest compression (ACC) devices has made it possible to transport patients with ongoing cardiopulmonary resuscitation (CPR) following out-of-hospital cardiac arrest (OHCA). In the region of Skåne in Sweden, all ambulances are equipped with an ACC device and local guidelines encourage prompt transportation of a majority of patients to hospital, regardless of whether field return of spontaneous circulation (ROSC) is achieved. In this retrospective registry study, we investigated the outcomes for patients transported with ongoing CPR between 2010-2015. We also studied the frequency of hospital-bound interventions against suspected reversible causes of cardiac arrest.

Methods: A local register containing Utstein-style data from all OHCA patients transported to SUS Lund was used and additional data were retrieved from medical records. If and where patients achieved ROSC was recorded. Interventions against suspected reversible causes of cardiac arrest were also noted and categorised per the “4Hs & 4Ts”, outlined in resuscitation guidelines. Patients with in-ambulance arrest were excluded.

Results: During the study period, 639 patients were transported to hospital following OHCA. A total of 160 patients achieved sustained ROSC before admission to hospital, 72 of whom (45%) survived to hospital discharge. Seventy-eight patients were excluded due to in-ambulance arrest (n = 49), arrest at hospital grounds (n = 11) and missing data (n = 18). The remaining 401 patients were transported with ongoing CPR of whom 52 (13%) eventually had sustained ROSC and admitted to the ICU. Eight patients (2%) survived to hospital discharge, 4 did not receive any further in-hospital intervention while 4 received an intervention that could not have been performed in the prehospital setting. These interventions were: angiography with epinephrine injected in the aortic root (n = 1), percutaneous coronary intervention (n = 1), insertion of a trans-venous pacemaker (n = 1), intubation and suction of aspirated fluids (n = 1). Three of these four survivors had intermittent ROSC in the field. The initial rhythm was shockable in 5 of 8 survivors and 3 had pulseless electrical activity.

Conclusions: The prognosis for patients transported to hospital with ongoing CPR is poor and the extra treatment potential in-hospital are rarely utilised. No patient with asystole without ROSC in the field survived.

P191 Hypoxic and hyperoxic preconditioning in organ protection against ischemia-reperfusion injury: the experimental study

I Mandel1, S Mikheev2, Y Podoxenov2, I Suhodolo3, A Podoxenov2, J Svirko2, A Sementsov2, L Maslov2, V Shipulin2

1City Clinical Hospital No 83 of FMBA of Russia, Moscow, Russia; 2Cardiology Research Institute, Tomsk, Russia; 3Siberian State Medical University, Tomsk, Russia

Introduction: Preconditioning with a moderate hypoxia and hyperoxia is an effective drug-free way to mitigate organ dysfunction to ischemia-reperfusion injury [1, 2].

Methods: The prospective study included 20 rabbits divided into four groups: hypoxic preconditioning (HypP), n = 5; hyperoxic preconditioning (HyperP), n = 5; hypoxic-hyperoxic preconditioning (HHP), n = 5; and control group, n = 5. Study was approved by local ethics committee. All animals were anesthetized by sevoflurane and mechanically ventilated via nasotracheal tube. In HypP group we exposed rabbits to two series of 10% oxygen for 10 min with 5 min reoxygenation. In HyperP group rabbits were exposed to 80% oxygen for 30 min. In HHP group rabbits were exposed to two series of 10% oxygen for 10 min with 5 min reoxygenation followed by 80% oxygen for 30 min. Then we started CPB and induced acute myocardial infarction by ligation of left coronary artery. After 45 min of ischemia reperfusion was performed for 60 min. We investigated myocardial slices and measured ischemic area (IA) and risk area (RA) [3] and calculated IA/RA ratio; also we conducted light microscopy of gut mucosa, kidneys, liver, spleen and lungs.

Results: IA/RA decreased in HypP group by 23%, in HyperP group 26%, in HHP group by 32% in comparison with control group (p = 0.009, Kruskal-Wallis test) [3]. Acid-based status, blood lactate and glucose levels were stable during all types of preconditioning. Single ventricular arrhythmia was observed more often than multiple ventricular arrhythmia in preconditioning animals and the opposite in control group. Incidence of ventricular fibrillation were lower in HHP group (differences did not reach statistical significance, X2 test). Light microscopy of myocardium revealed less damage in HHP group as compared to other groups. Light microscopy of kidneys revealed marked edema of cortical and medullar substances in the control group. Gut mucosa and liver parenchyma had enlarged capillaries, sometimes filled with erythrocytes. Microscopic structure of kidneys, small intestine and liver was less affected in HHP group.

Conclusions: Hypoxic-hyperoxic preconditioning provided the highest tolerance of the myocardium and splanchnic organs to the effects of ischemia-reperfusion injury.


1. Petrosillo G. et al. Free Radic Biol Med. - 2011. - Vol 50(3). – p.. 477–483.

2. Xu K. et al. Adv Exp Med Biol. - 2014. - Vol. 812. – p. 309–315.

3. Mandel I. et al. Journal of Cardiothoracic and Vascular Anesthesia. - 2016. - Vol. 30 (Supplement 1), p. S6–S7.

P192 Diabetes in an animal model worsens neurological outcome following cardiac arrest

LV Vammen1, SR Rahbek2, NS Secher1, JP Povlsen3, NJ Jessen4, BL Løfgren1, AG Granfeldt1

1Aarhus University Hospital, Aarhus C, Denmark; 2Regional Hospital of Randers, Randers, Denmark; 3Regional Hospital of Horsens, Horsens, Denmark; 4Aarhus University, Aarhus, Denmark

Introduction: Cardiac arrest carries a poor prognosis. The average cardiac arrest patient is comorbid and retrospective studies suggest that diabetes mellitus is an independent risk factor for increased mortality after cardiac arrest. Despite this, cardiac arrest animal studies are conducted on healthy young animals, limiting our knowledge regarding the post-cardiac arrest organ dysfunction and the impact of type 2 diabetes mellitus (T2DM).

We hypothesize that T2DM, in a rat model of cardiac arrest, is associated with increased brain injury and reduced left ventricular function following resuscitation.

Methods: We used the Zucker Diabetic Fatty (ZDF) rat as an animal model of T2DM. The ZDF rats (n = 13), non-diabetic Zucker Lean Control (ZLC) rats (n = 15), and healthy Sprague Dawley (SprD) rats (n = 8) underwent asphyxia-induced cardiac arrest and were resuscitated and monitored for 180 min. after return of spontaneous circulation (ROSC). Brain injury was evaluated by neuron specific enolase (NSE) and left ventricular function was measured as fractional shortening (FS) by echocardiography both measured at baseline and 180 min. after ROSC.

Results: Total asphyxia time was 639 s (SD 24) in the ZDF group which was significantly longer than 592 s (SD 18, p < 0.0001) in the ZLC group, but no different from the SprD group at 634 s (SD 20, p = 0.6). There were no differences among groups in baseline NSE or FS. 180 min. after ROSC median levels of NSE were significantly increased in the ZDF group, 10.8 ng/mL [25%Q;75%Q: 7.6;11.3], compared with the two control groups: ZLC group 2.0 ng/mL [25%Q;75%Q 1.7;2.3, p = 0.0004] and SprD group 2.8 ng/mL [25%Q;75%Q 2.3;3.4, p = 0.0004]. At the end of experiment mean FS was significantly higher in the ZDF group at 36% (SD 6), compared with the ZLC group, 22% (SD 3, p = 0.0031), and the SprD group, 24% (SD 6, p < 0.0001). At baseline lactate was 1.9 (SD 0.9) in the ZDF group, which was comparable with 1.4 (SD 0.4, p = 0.512) in the ZLC group and 1.2 (SD 0.5, p = 0.052) in the SprD group. At the end of experiment lactate concentration in the ZDF group was at 8.2 mmol/L (SD 3.1) significantly higher than the control groups: ZLC group 2.6 mmol/L (SD 1.7, p = 0.001) SprD group 1.8 mmol/L (SD 0.9, p < 0.0001).

Conclusions: Cardiac arrest in an animal model of T2DM results in increased brain injury, while in contrast left ventricular function was increased when compared to non-diabetic animals.

P193 Increased energy required to reach target temperature in post-cardiac arrest patients is associated with better outcomes

A Grossestreuer1, S Perman2, P Patel1, S Ganley1, J Portmann1, M Cocchi1, M Donnino1

1Beth Israel Deaconess Medical Center, Boston, MA, United States; 2University of Colorado, Denver, CO, United States

Introduction: We hypothesized the amount of energy required by the surface device to reach target temperature (Ttarget) in post-cardiac arrest patients treated with therapeutic hypothermia (TH) may be associated with outcomes by serving as a proxy for patient thermoregulatory ability and may modify the relationship between the time to Ttarget and outcomes. Some studies have shown that TH-treated post-arrest patients who reach Ttarget quickly have worse outcomes than those who cool more slowly. However, the ischemia-reperfusion insult of cardiac arrest may cause temperature derangements that affect the time trajectory of TH independent of external cooling factors.

Methods: Adult patients with sustained return of spontaneous circulation treated with TH between 2008-2015 with serial temperature data were included. Time to Ttarget was defined as time from TH initiation to the first time the patient temperature was < =34C. Patients with Ttarget >34C were excluded. The energy required to bring a patient to Ttarget (“energy units”) was calculated as average inverse water temperature x 100 x hours between initiation and Ttarget. Primary outcome was neurologic status (measured by Cerebral Performance Category [CPC] score); secondary outcome was survival, both at hospital discharge. Univariate analyses were performed using Wilcoxon rank-sum tests; multivariate analyses used logistic regression. P < 0.05 was considered statistically significant.

Results: Of 205 patients included, those with CPC 3-5 required less energy to reach Ttarget (median 8.1 (IQR: 3.6-21.6) vs median 20.0 (IQR: 9.0, 33.5) energy units, p = 0.001) and reached Ttarget quicker (median 2.3 (IQR: 1.5, 4.0) vs median 3.6 (IQR: 2.0, 5.0) hours, p = 0.01) than patients with CPC 1-2. Patients who did not survive required less energy than survivors (median 8.1 (IQR: 3.6-20.8) vs median 19.0 (IQR: 6.5, 33.5) energy units, p = 0.001) and reached Ttarget quicker (median 2.2 (IQR: 1.5, 3.8) vs median 3.6 (IQR: 2.0, 5.0) hours; p = 0.01). Controlling for average water temperature between initiation and Ttarget, the relationship between outcomes and time to Ttarget was no longer significant. Controlling for location, witnessed arrest, age, initial rhythm, and neuromuscular blockade use, increased energy was associated with better neurologic (aOR: 1.01 (95%CI 1.00-1.03), p = 0.039) and survival (aOR: 1.01 (95%CI 1.00-1.03), p = 0.045) outcomes.

Conclusions: Increased energy requirement during TH initiation is associated with better outcomes at hospital discharge and may affect the relationship between time to Ttarget and outcomes.



P195 Clinical outcomes of patients with different co-morbideties witnessed in cardiac arrests inside the intensive care unit

Y Nassar, S Fathy, A Gaber, S Mokhtar

Cairo University, Giza, Egypt

Introduction: We aimed for clinical assessment of cardiopulmonary resuscitation (CPR) procedures of witnessed cardiac arrests inside the intensive care units (ICU) and follow up of patients surviving to discharge.

Methods: Data were collected prospectively from patients who were witnessed in cardiac arrest and underwent CPR inside an adult medical ICU of Cairo University in the period from Jan. 2013 to Feb 2015. Resuscitation protocol was done according to the latest recommendation of the European society of cardiology. Clinical data were recorded and surviving patients were clinically followed daily until hospital discharge.

Results: The study included 110 patients: 41 females (37%) and 69 males (63%). There were 26(24%) patients under 50 years and 84(76%) patients above 50 years. ROSC occurred in 60(55%) and 22(20%) survived to discharge.

According to underlying illness:

  • ROSC increased with CNS comorbidities (p0.05), Shock (p0.008), Low MPM0-III (p 0.015) While ROSC decreased with and Respiratory failure (p 0.01)

  • Long term survival increased with Low MPM0-III score (p0.018), Low Sofa score (p 0.01), and Rapidly correctable causes (Hypoxia, Hypovolemia, Hydrogen ion acidosis, Hypokalemia, Hyperkalemia, Hypoglycemia, Hypothermia, Toxins, Cardiac Temponade, Tension pneumothorax, Thrombosis and Trauma) (p0.004), while Long term survival decreased with CNS comorbidities (p0.02), Shock (p,0.01), Respiratory Failure post-arrest (p0.02) and Mechanical ventilation post-arrest (p < 0.001)

According to CPR procedures:

  • ROSC increased with AF Rhythm (p0.03),less duration of CPR (p0.03), number of cycles of CPR < 2 (p < 0.001), number of DC shocks <2 (p0.02), EF >50% (p0.01), Low HCO3 pre-arrest (p0.009), low HCO3 during arrest (p0.03) and Noradrenaline post-arrest (p0.003).ROSC decreased with high PaCO2 pre-arrest (p 0.002) and high PaCO2 during arrest (p 0.001).

  • Long term survival increased with RBBB (p < 0.001) and frequent PVCs (p < 0.001). Long term survival decreased with Asystole (p 0.01).

Conclusions: CPR in the ICU may achieve variable rates of short and long term survival depending on the associated comorbidities

P196 Outcomes of non-traumatic out-of-hospital cardiac arrests witnessed by layperson

YC Chia

Tan Tock Seng Hospital, Singapore, Singapore

Introduction: To look at the outcomes of patients who suffered a non-traumatic OHCA, witnessed by layperson.

Methods: This is a retrospective case record review. Inclusion criteria included all patients who suffered a non-traumatic OHCA conveyed by emergency medical services (EMS) to our Emergency Department (ED) from 1st Aug 2012 to 31st Aug 2014. Exclusion criteria included traumatic OHCA and all patients declared dead at scene. EMS data were extracted from National Cardiac Arrest registry. Data of patients admitted were extracted from inpatient electronic case records. Patient discharged from hospital were followed up for 30 days.

Results: There were 888 OHCA. 529 were witnessed. Among witnessed, 279 by family, 134 by layperson, 76 by EMS, and 40 by healthcare providers.

134 OHCA witnessed by layperson, 47 occurred in night-time between 2000-0800 hrs. 87 occurred in daytime between 0800-2000 hrs. 4 happened in healthcare facility, 40 in a residential area, 8 in industrial area, 13 in places of recreation, 43 in commercial / public area, 20 in the streets or highway and 6 in brothels.

57 received bystander CPR while 77 did not. Among 57 who received bystander CPR, 9 also had community AED applied. The rhythms on the AED were 4 unknown, 4 asystole and 1 PEA. None was shockable. The initial rhythms by EMS on those 57 patients who received bystander CPR were 24 asystole, 8 PEA, 10 unknown and 15 VF. 3 of the patients with VF were subsequently discharged from hospital alive with CPC 1.

None of the 77 patients who had a witnessed OHCA by a layperson and did not received bystander CPR survived.

134 OHCA witnessed by layperson, 21 were admitted to hospital. 4 were discharged alive with cerebral performance category (CPC) 1. Among these 4 patients with good outcome, 3 had VF as initial presenting rhythm, while 1 had PEA. All happened in public area in daytime, all had bystander CPR, all did not have community AED applied but all had field ROSC

Conclusions: Our results showed patients with OHCA witnessed by layperson tend to occur in public areas during daytime and most did not received bystander CPR or application of community AED. Some of these arrests were witnessed by laypersons, however, none survived when CPR was not started. This warrants further study to find out the reasons for not initiating CPR. There should be more efforts directed at community CPR programs. Technology can help to direct rescuers to OHCA victim so that early CPR can be initiated.

P197 Reducing the risk of a poor outcome: considering the ‘other’ victim of out-of-hospital cardiac arrest (OOHCA) following unsuccessful resuscitation

R Lewis-Cuthbertson1, K Mustafa2, A Sabra2, A Evans2, P Bennett1

1Swansea University, Swansea, United Kingdom; 2Abertawe Bro Morgannwg Health Board, Swansea, United Kingdom

Introduction: Early intervention and prevention of psychological disorders could help to significantly reduce costs to the NHS. Relatives who experience the sudden and unexpected death of a relative following OOHCA may be at high risk of developing both psychological and physical health problems. As many of these deaths occur within the Emergency Department(ED), it is important to understand the support needs of relatives to help minimise risk of a poor outcome.

Methods: Next of kin of deceased non-traumatic, non-paediatric OOHCA patients were invited to take part in face-to-face interviews 3 months’ after death. Twelve male and female participants who experienced the death of a relative following an OOHCA and who had either i)witnessed the event and provided CPR, ii)witnessed the event and did not provide CPR, and iii)those who did not witness the event completed audio-recorded interviews lasting up to 90 minutes. Audio-recordings were transcribed verbatim and subjected to inductive thematic analysis.

Results: Three major themes were identified. 1)Negative psychological and physical health outcomes: Post-traumatic stress symptoms were reported including vivid and intrusive re-experiencing of the event. Flashbacks, recurring traumatic images, and experiencing physical symptoms associated with the heart were common. Intimations of mortality were associated with hyperarousal symptoms and health anxieties. Coping techniques included avoidance behaviours and emotional numbing, often masking their distress and support needs. 2)Am I to Blame: Self-critical thinking regarding one’s own actions in relation to this event were evident. Relatives blamed themselves for not noticing sooner that something was wrong, particularly when an underlying heart condition was identified as cause of death. Many were pre-occupied with the thought: ‘Could I have done more?’. 3)Information and Support Needs: Many felt uncertain and uninformed about what was happening creating feelings of anger, frustration and confusion. Seeking information was important for relatives to help both try to make sense of what happened and exonerate feelings of guilt and self-blame.

Conclusions: Findings suggest that the psychological impact of experiencing the sudden death of a relative following an OOHCA may be profound. Information provision is crucial to help relatives make sense of their experience and exonerate feelings of guilt and self-blame. Support of relatives needs to be a more serious consideration to help minimise risk of poor psychological outcome and reduce the health economic burden this may pose.



P199 he use of the bispectral index and suppression ratio to predict poor neurological outcome in post-cardiac arrest patients treated with targeted temperature management at 33°c

W Eertmans, C Genbrugge, W Boer, J Dens, C De Deyne, F Jans

Ziekenhuis Oost-Limburg, Genk, Belgium

Introduction: Bispectral index (BIS) monitoring has been considered as a promising electrophysiological tool for early prognostication after out-of-hospital cardiac arrest (OHCA). In recent years, a broad range of BIS thresholds has been put forward at diverse time points to predict neurological outcome in OHCA patients. This study aimed to reach consensus about the optimal time point and threshold for predicting poor neurological outcome after OHCA using the BIS monitor.

Methods: A prospective, observational study was performed during TTM at 33 °C in 77 successfully resuscitated OHCA patients. After admission to the ICU, BIS and Suppression Ratio (SR) monitoring was started using the BIS VISTA™ (Aspect Medical Systems, Inc. Norwood, USA). BIS and SR values were continuously recorded during the hypothermic and rewarming phase. During this time period, mean BIS and SR values per hour were calculated and used for analysis. The Cerebral Performance Category (CPC) scale was used to define patient’s outcome at 180 days after OHCA (CPC 1-2: good - CPC 3-5: poor neurologic outcome). Receiver operator characteristics curves were constructed to determine the best cut-off value and time point to predict poor neurological outcome.

Results: At 180 days post-cardiac arrest, 38 patients (49%) had a good neurological outcome (CPC 1-2), while 39 patients (51%) had a poor outcome (CPC 5). There were no patients with a CPC 3 or 4. Patients with a good neurological outcome had higher BIS and lower SR values than non-survivors. Using a mean BIS value below 25.5 at hour 12 as threshold criteria, poor neurological outcome was predicted with a sensitivity of 49% (95% CI 30-65%) and specificity of 97% (95% CI 85-100%) (AUC: 0.722 (0.570-0.875); p = 0.006). With a cut-off value of SR above 2.5, the optimal sensitivity (74%, 95% CI 56-87%) and specificity (92%, 95% CI 78-98%) for poor neurological outcome was obtained at hour 23 (AUC: 0.836 (0.717-0.955); p < 0.001).

Conclusions: This prospective, observational study confirmed that mean BIS values at hour 12 can be used to predict poor neurological outcome. In addition, we showed that the predictive ability of the SR might be even higher as compared to the one of BIS. Overall, our results suggest that BIS and SR monitoring can be used to assist with early neuroprognostication after OHCA.

P200 Retrospective analysis of trends in outcome following Out of Hospital Cardiac Arrest from a UK regional cardiac arrest centre, with a focus on haematological parameters

A Skorko, M Thomas

Bristol Royal Infirmary, Bristol, United Kingdom

Introduction: The Bristol Royal Infirmary is a tertiary out of hospital cardiac arrest (OHCA) centre, serving a population of 1,000,000 in the South West of England.

Debate is ongoing regarding the optimal antiplatelet strategy for survivors of OHCA, given the concerns of both bleeding and clotting complications in this population. Studies vary in the rates of bleeding seen in this population, from 0 to 56% [1].

To understand why OHCA survivors are at risk of bleeding and whether this impacts survival to hospital discharge, we retrospectively analysed laboratory parameters of coagulation(PT and APTT), platelet count, and haemoglobin then stratified these by survival to hospital discharge, over the period January 2009 to August 2016.

Methods: Routinely collected electronic data was used to identify patients for this study. Data was extracted for those coded as ‘Anoxic or ischaemic coma or encephalopathy’ with ‘Acute myocardial infarction’ or ‘Ventricular tachycardia or fibrillation´.

Parameters of; status at hospital discharge, platelet count, APTT and haemoglobin within 24 hours of admission were extracted from the database.

Results: Comparing the relative risks; An INR of >2 gave a risk ratio of death of 1.29 (95% CI 0.828- 2.01, p =0.257). A platelet count of <50 gave a risk ratio of death of 1.30 (95% CI 0.866- 1.97, p =0.301).

Conclusions: Our data demonstrates that platelet count, hemoglobin, PT and APTT do not differ between survivors and non-survivors of OHCA.

However, absolute values do not give an indication as to the function of platelets or of the coagulation cascade. An analysis of platelet function and the coagulation cascade as a whole may provide better insights into the risks of bleeding in this population. We therefore propose to carry out a prospective analysis of thromoelastometric coagulation assessment and platelet inhibition, and correlate this with bleeding events and administration of antiplatelet therapy


1. Nolan JP et al. Increasing survival after admission to UK critical care units following cardiopulmonary resuscitation. Critical Care. 2016;20.

Table 4 (abstract P200). median and interquartile ranges (IQR) for haematological parameters
Fig. 8 (abstract P200).

Number of patients treated by year and % survival to hospital discharge

P201 Organ donation after brain death in refractory cardiac arrest treated with extracorporeal CPR

M Casadio1, A Coppo2, A Vargiolu2, J Villa1, M Rota1, L Avalli2, G Citerio1

1University of Milano-Bicocca, Monza, Italy; 2San Gerardo Hospital, Monza, Italy

Introduction: Cardiac arrest (CA) is a catastrophic event with a high rate of mortality, often resulting in devastating brain injury that might evolve to brain death (BD)[1]. Organ donation from BD after CA patients with ECMO support is still a poorly explored field[2].

Methods: We retrospectively enrolled all patients admitted to our hospital between January 2011 and September 2016 after refractory CA treated with eCPR.

Results: In the study period 112/215 CA patients received eCPR (52.09%). 30 eCPR-subjects (26.78%) survived at 6 months (85.71% with good cerebral performance, CPC 1-2). 82 died in ICU (25 BD, 22.32% and 57 for other causes, 50.89%) [Fig. 9]. Deads vs. alives differed in age (p = 0.02), comorbidities (p = 0.001), CA (intra or extra-hospital p = 0.03), low flow time (p < 0.0001), mean arterial pressure (p = 0.004), glycemia (p = 0.04) anemia (p = 0.02), renal function (creatinine p < 0.0001, urea p = 0.0005), and early neurological evaluation (CT scan p < 0.001, EEG recording p < 0.001, brainstem reflexes p < 0.001, presence of somatosensory potentials p = 0.03 and GCS p < 0.0001). BD and dead from other causes patients differed in early neurological evaluation (CT scan p < 0.0001, EEG p = 0.004, brainstem reflexes p = 0.02), thrombocytopenia (p = 0.008), coagulation derangement (p = 0.01), inotropic support (p = 0.03). Tab1 shows characteristics of eligible patients at the time of donation. Rate of donation in BD patients was 56% (refusal based on organ biopsy or evaluation in the operation room) with 39 donated organs (23 kidneys, 12 livers, 4 lungs, 89.74% with good functional recovery).

Conclusions: eCPR patients might become BD and be considered potential resource for organ donation with a similar success rate as organs retrieved from patients deceased from other causes.


1. Sandroni C et al Int Care Med 42(11):1661–1671,2016

2. Citerio G et al Int Care Med 42(3):305–15,2016

Table 5 (abtsract P201). See text for description
Fig. 9 (abstract P201).

See text for description





P204 Effects of long-term post-ischemic treadmill exercise on gliosis in the aged gerbil hippocampus induced by transient cerebral ischemia

JB Moon, JH Cho, CW Park, TG Ohk, MC Shin, MH Won

Kangwon National University, Chuncheonsi, South Korea

Introduction: Therapeutic exercise is an integral component of the rehabilitation of patients with stroke. The objective of the present study was to investigate effects of post-ischemic exercise on neuronal damage or death and gliosis in the aged gerbil hippocampus after transient cerebral ischemia using immunohistochemistry.

Methods: Aged gerbils (male, 22 to 24 months) induced by ischemia were subjected to treadmill exercise for 1 or 4 weeks. Neuronal death was apparently found in the stratum pyramidale of the hippocampal CA1 region and in the polymorphic layer (PoL) of the dentate gyrus (DG) using cresyl violet and Fluoro-Jade B histofluorescence staining.

Results: In addition, no significant difference in neuronal death was found after 1 or 4 weeks of post-ischemic treadmill exercise. However, post-ischemic treadmill exercise apparently affected gliosis (activation of astrocytes and microglia). GFAP immunoreactive astrocyte and Iba-1 immunoreactive microglia were activated in the CA1 and PoL of the DG of the group without treadmill exercise. However, 4 weeks after treadmill exercise significantly alleviated ischemia-induced astrocyte and microglia activation, although the gliosis was not alleviated in the animals with 1-week exercise.

Conclusions: These findings suggest that long-term post-ischemic treadmill exercise after transient cerebral ischemia could not influence neuronal protection, however, it could effectively alleviate astrocyte and microglial activation in the aged hippocampus induced by 5 min of transient cerebral ischemia.

P205 Predictors of need for critical care support after stroke thrombolysis in an intensive care unit

P Papamichalis1, V Zisopoulou1, E Dardiotis2, S Karagiannis1, D Papadopoulos1, T Zafeiridis1, D Babalis1, A Skoura1, I Staikos1, A Komnos1

1General Hospital of Larissa, Larissa, Greece; 2University Hospital of Larissa, Larissa, Greece

Introduction: One of the most promising interventions for acute ischemic strokes is intravenous thrombolysis (IVT). In our hospital it is performed in the Intensive Care Unit (ICU), with a 10 year experience for the intervention and participation at international studies [Safe Implementation of Thrombolysis in Stroke – MOnitoring STudy (SITS-MOST)] [1] and registries [Safe Implementation of Treatments in Stroke - International Stroke Thrombolysis Register (SITS-ISTR)]. The aim of the study was to evaluate which factors can predict the need for critical care support after thrombolysis.

Methods: Retrospective study including 124 patients with acute ischemic stroke, with mean age 65 years and National Institutes of Health Stroke Scale (NIHSS) at admission 11/range 2 – 28. They all fulfilled the international inclusion criteria [2] and received IVT with alteplase. Demographic data and severity scores [Simplified Acute Physiology Score (SAPS) II and NIHSS] were recorded. Patients were divided to those who demanded advanced life support and neurocritical care interventions after IVT (n = 14) and those who did not (n = 110). Comparison amongst the two groups was performed with application of proper statistical tests.

Results: The need for critical care support was significantly greater for patients with higher SAPSII, higher NIHSS after 2 hours, at 24 hours and at 7 days (Mann-Whitney U test for the above mentioned comparisons) and patients with history of vascular disease (Fisher’s exact test) (p < 0.05 for all comparisons). Off-Label Thrombolysis, NIHSS <5, age >80 years, sex, age, NIHSS at admission, aggregate thrombolysis time, history of: diabetes mellitus, arrhythmia, hypertension, smoking, hyperlipidemia, former ischemic stroke did not significantly correlate with critical care need.

Conclusions: In accordance with previous studies [3,4] higher severity scores (SAPSII, NIHSS) and presence of vascular disease can serve as markers for prediction of need for critical care support after IVT. Our report is the first one in the international literature of SAPSII correlating with thrombolysis results.


1) Wahlgren N et al, Lancet 369:275–282, 2007

2) The European Stroke Organisation (ESO) Executive Committee and the ESO Writing Committee, Cerebrovasc Dis 25:457–507, 2008

3) Faigle R et al, PLoS One 9:e88652, 2014

4) Mazya M et al, Stroke 43:1524–1531, 2012

P206 A life threatening emergency: PRES - cases series and literature review

S Silva Passos, F Maeda, L Silva Souza, A Amato Filho, T Araújo Guerra Granjeia, M Schweller, D Franci, M De Carvalho Filho, T Martins Santos, P De Azevedo

University of Campinas, Campinas, Brazil

Introduction: To describe comorbidities, clinical presentation, diagnostic method, treatment and outcome of 6 patients with Posterior Reversible Encephalopathy Syndrome (PRES). PRES is consequence of a reversible subcortical brain edema in patients with acute neurological symptoms, such as headache, impaired sensorium, visual abnormalities, nausea/vomiting, cerebellar syndrome, focal neurological deficits and seizures. Causes include hypertension, eclampsia/pre-eclampsia, sepsis, autoimmune disease, immunosuppressive agents, chemotherapy and renal failure. Radiological findings on computed tomography (CT) and magnetic resonance imaging (MRI) include abnormalities of white and grey matter, predominantly affecting parietal and occipital lobes. The treatment is based on the removal of the underlying cause.

Methods: Retrospective and descriptive study of 6 medical records of patients hospitalized with diagnosis of PRES at the tertiary hospital of the University of Campinas, São Paulo, between 2015 and 2016.

Results: Four patients had lupus, 2 were solid-organ transplant patients (1 kidney and 1 liver). Immunosuppression and headache were found in all patients, 5 presented hypertensive emergencies, 4 had seizures, 3 showed decreased level of consciousness and nausea and vomiting, 1 had status epilepticus. The diagnosis was made clinically and with CT in 4 cases and MRI in 2 cases (Fig. 10). Treatment was performed with intravenous blood pressure lowering agents and antiepileptic drugs. The length of stay ranged from 22 days to 66 days. Five patients showed full recovery, and 1 died of intracranial hemorrhage.

Conclusions: Autoimmune disease, use of immunosuppressant and hypertension are important risk factors for PRES. Patients usually have a good recovery after prolonged stay in hospital, but death and neurological disability may occur. Therefore, early recognition and appropriate treatment may change patient´s outcome.

Fig. 10 (abstract P206).

Multiple hyperintense areas with mass effect in the high convexity

P207 A case report: chronic lymphocytic inflammation with pontine perivascular enhancement responsive to steroids (CLIPPERS)

R Wall, I Welters

Royal Liverpool University Hospital, Liverpool, United Kingdom

Introduction: Chronic lymphocytic inflammation with pontine perivascular enhancement responsive to steroids (CLIPPERS) was first described in 2010 [1]. Symptoms include diplopia and gait ataxia. Pontine lesions are a commonly seen on magnetic resonance imaging. Patients display a favourable response to glucocorticosteroid therapy.

Methods: A 54 year-old man with no significant past medical history presented to the emergency department with a 10-day history of vomiting and feeling generally unwell. He suffered from dizziness, ataxia, diplopia, bilateral nystagmus, taste and hearing disturbances. Within 24 hours of admission he developed right sided weakness. GCS dropped to 8 and he required intubation. He had downgoing plantar reflexes, myoclonic jerks, tonic-clonic movements of lower limbs and spasticity of both upper and lower limbs. Seizures occurred daily and were terminated with lorazepam. Two weeks after admission, the patient developed pyrexia of 420 C and required cooling for 7 days.

Results: The patient was treated for infective encephalitis with amoxicillin, ceftriaxone and acyclovir. Cerebrospinal fluid showed a white cell count of 96 cells/mcl (40% polymorphs and 60% lymphocytes). A second lumbar puncture a week later had a white cell count of 72 (5% polymorphs and 95% lymphocytes). All cultures, viral PCR and cryptococcal antigens were negative. The first MR head showed high T2 signal and swelling in the pons, which extended into the cerebral peduncle on the left. This area and the middle cerebral peduncles had high signal on the FLAIR sequence. There was restricted diffusion within the pons and middle cerebellar peduncle. A second MR scan identified several irregular regions of enhancement within the midbrain and pons, leading to a diagnosis of CLIPPERS. High-dose methylprednisolone was commenced intravenously. The patient’s clinical condition improved rapidly. He was alert and followed commands within days. Nystagmus persisted, but he had no further spasticity. A third MR after five days of steroid treatment described significant improvement in the pontine lesions, but a small focus of high signal remained. Midbrain oedema present on previous scans had resolved.

Conclusions: CLIPPERS is a diagnosis of exclusion, with patients often initially treated for stroke or encephalitis. Pyrexia has not been reported in cases of CLIPPERS. Central pyrexia is common in neurological conditions. An infective cause was not identified, however, with no response to antimicrobials in the absence of positive cultures, central pyrexia is likely. The rapid response to steroids supports a diagnosis of CLIPPERS.


1. Pittock S et al. Brain 2010;13:2626

P208 Comparison of successful rate between ultrasound guided lumbar puncture and surface landmark method for difficult LP patients in ED; a randomized controlled trial

P Tansuwannarat, P Sanguanwit

Ramathibodi hospital, Bangkok, Thailand

Introduction: Lumbar puncture (LP) is the main procedure to obtain the diagnosis of meningitis and subarachnoid hemorrhage. However the success rate could be compromised in difficult LP patients such as patients with obesity or scoliosis. Ultrasound-guided LP (UGLP) was proposed as a potential method to improve the success rate of this procedure.We aimed to investigate the success rate of UGLP compared to surface-landmarked LP (SMLP) in patients who visited an Emergency Department at Ramathibodi Hospital, Bangkok, Thailand.

Methods: This is a prospective randomized controlled trail from August 2015 to July 2016. All adult (>18 years old) patients with a diagnosis of suspected meningitis were included. Difficult LP patient was defined as patient with bone mineral density of 25 kg/m2 or more and scoliosis. Primary outcome was the success rate of first LP attempt. Secondary outcome included number of attempts, time to complete the procedure, and post-procedure complication.

Results: There were 40 patients included with a mean age of 60 + 19.34 years and 53% were male. The majority of patients were suspected for meningitis (82%).There were 20 patients in each group. Success rate at the first LP attempt was greater in UGLP group compared to SMLP group [16(80%) vs 7(35%), p 0.009]. Median time (minutes) to complete the procedure was shorter in UGLP compared to SMLP group [(5 vs. 13.5), p 0.002]. Post-procedure complication (blood-contaminated CSF) occurred less in UGLP compared to SMLP group [(2(10%) vs. 6(30%), p 0.235

Conclusions: UGLP significantly improve the success rate of first LP attempt and decrease time to complete the procedure in difficult LP patients.


UGLP significantly improve the success rate of first LP attempt and decrease time to complete the procedure in difficult LP patients.

Table 6 (abtsract P208). Generalized Characteristics From All Enrolled Patients
Table 7 (abtsract P208). Compare sucess rate in 1st attempt and complication between USLP VS SLLP

P209 Acid-base characteristics of the cerebrospinal fluid of patients with subarachnoid hemorrhage and in control subjects according to Stewart’s approach

T Langer 1, M Carbonara2, A Caccioppola1, C Ferraris Fusarini2, E Carlesso1, E Paradiso1, M Battistini1, E Cattaneo1, F Zadek1, R Maiavacca2, N Stocchetti1, A Pesenti1

1University of Milan, Milan, Italy; 2Fondazione IRCCS Ca’ Granda, Ospedale Maggiore Policlinico, Milan, Italy

Introduction: The pathophysiology of the acid-base equilibrium of the cerebrospinal fluid (CSF) is important, as it influences respiration [1]. Aim of the present study was to describe CSF acid-base of patients with subarachnoid hemorrhage (SAH) and compare them with control subjects.

Methods: In patients with SAH, a CSF sample was taken from the external ventricular drain simultaneously with an arterial blood sample to measure electrolytes, albumin, phosphates, PCO[sub]2[/sub] and pH. A similar procedure was performed in patients without significant comorbidities undergoing spinal anesthesia for elective surgery. For each sample the Strong Ion Difference (SID[sub]CSF[/sub]) and the total concentration of weak, non-volatile acids (A[sub]TOT[/sub]) were calculated using standard formulae. Furthermore plasma SID and its difference with SID[sub]CSF[/sub] was computed (Δ SID). Comparison between groups was performed via t-test or Rank Sum Test, as appropriate.

Results: Ten patients with SAH (55 ± 16 years) and 7 controls (56 ± 11 years) were enrolled. Acid-base results are summarized in Table. In SAH patients, SID[sub]CSF[/sub] was lower, mainly due to a higher lactate (3.2 ± 1.5 vs. 1.5 ± 0.3 mEq/L, p < 0.001) and chloride concentration (125 ± 2 vs. 120 ± 2 mEq/L, p < 0.001). Despite the acidifying effect of lower SID[sub]CSF[/sub], a lower CSF PCO[sub]2[/sub], with unchanged CSF A[sub]TOT[/sub] led to CSF pH values similar to controls. Finally, despite a lower plasma SID in the SAH group, Δ SID was significantly higher in these patients.

Conclusions: The CSF of patients with SAH, as compared to control subjects, has lower SID and PCO[sub]2[/sub], but similar pH. In these patients, the reduction in SID[sub]CSF[/sub] is not coupled with a similar reduction in plasma SID, leading to a higher Δ SID, i.e. a more pronounced acidification of CSF as compared to plasma.


1. Langer T et al. Intensive Care Med 42(3):436–9, 2016

Table 8 (abstract P209). SAH = subarachnoid hemorrhage; Data presented as mean ± standard deviation.

P210 Polyuria and natriuresis after aneurysmal subarachnoid haemorrhage

A Ramos1, F Acharta1, J Toledo1, M Perezlindo1, L Lovesio1, A Dogliotti2, C Lovesio1

1Sanatorio Parque, Rosario, Argentina; 2Grupo Oroño, Rosario, Argentina

Introduction: Natriuresis and polyuria are common events after aneurysmal subarachnoid haemorrhage (aSAH). A relationship has been found between polyuria, cerebral salt wasting syndrome (CSWS) and vasospasm [1]. The aim of this study is to determine the relationship between creatinine clearance and natriuresis, and to identify variables related to natriuresis.

Methods: During 2 years (2014-2016) 29 patients with aSAH and polyuria were identified. The tomographic characteristics and neurological clinical scores were considered. 24-hour urine was obtained in individuals with polyuria (>3 liters / day). CSWS was defined as the 75th percentile (>923 mEq / l urinary sodium). Symptomatic vasospasm was defined as clinical deterioration confirmed with cerebral angiography.

Results: 29 patients were included, eight of them (27.6%) developed CSWS. No patient presented hyponatremia. No predictors of CSWS were found (Fig. 11). Neither sodium in 24 hours (AUC: 0.57, 95% CI: 0.37-0.75 p = 0.5), nor the volume of diuresis predicted symptomatic vasospasm (AUC: 0.57, 95% CI: 0.37-0.75 p = 0.5). There was no correlation between creatinine clearance and natriuresis (p = 0.17) (Fig. 12).

Conclusions: We found no relationship between polyuria, natriuresis and symptomatic vasospasm.


1. Brown RJ et al. Polyuria and cerebral vasospasm after aneurysmal subarachnoid hemorrhage. BMC Neurology. 15:1–7, 2015.

Fig. 11 (abstract P210).

Multivariate analysis

Fig. 12 (abstract P210).

Correlation between creatinine clearance and natriuresis

P211 Coma in late night amsterdam; do not forget the travel history

N Schroten1, B Van der Veen2, MC De Vries2, J Veenstra2

1VUMC, Amsterdam, Netherlands; 2OLVG, Amsterdam, Netherlands

Introduction: Unconscious young patients are admitted to hospitals in Amsterdam nearly on a daily basis, mostly due to intoxications. However, the following case underscores that routine laboratory and imaging investigations do not replace a detailed interrogation.

Methods: A 17-year old woman was found unconscious in the early morning. Her family mentions she had complained about headaches, but no other complaints or fever. At the emergency department her temperature was 37 °C, respiratory frequency 31/min and pulse 120/min. She was unresponsive and had uncontrolled jerking movements, without signs of lateralisation and with normal stem reflexes. Laboratory analyses showed C-reactive protein 36 mg/l; Leukocytes 6,4x10^9/l; Haemoglobin 5,4 mmol/l; Mean corpuscular volume 86 fl; Thrombocytes 156x10^9/l; LD 318 U/l, bilirubin 28 μ mol/l. Cerebral computer tomography and analysis of cerebrospinal fluid were unremarkable. Empiric therapy with broad-spectrum antibiotics, acyclovir and dexamethasone were started.

Results: At the intensive care unit she developed a fever up to 40 °C and haemoglobin levels decreased to 3,8 mmol/l. Further interrogation of the parents reported a visit to Ghana for 2 weeks 3 months before. She had taken mefloquine prophylactically. A thick smear was positive: plasmodium falciparum with a parasitaemia of 0,3%. After artesunate 2,4 mg/kg twice daily intravenously the patient recovered rapidly. On follow up she had minor concentration problems.

Conclusions: Cerebral malaria is a diffuse symmetric encephalopathy. Children are at a higher risk than adults. Focal signs are unusual. CT scans and cerebrospinal fluid analysis are usually unremarkable. It is important to note that cerebral malaria may have a prolonged incubation time, especially in patients using malaria prophylaxis, like in the current case, and may be lethal even at a low parasitemia. Early treatment is vital. The authors confirm they have received informed consent to publish from the patient.


1. Jakka SR, Veena S, Atmakuri RM, Eisenhut M. Characteristic abnormalities in cerebrospinal fluid biochemistry in children with cerebral malaria compared to viral encephalitis. Cerebrospinal Fluid Res. 2006;3:8.

2. Giha HA1, A-Elbasit IE, A-Elgadir TM, Adam I, Berzins K, Elghazali G, Elbashir MI. Cerebral malaria is frequently associated with latent parasitemia among the semi-immune population of eastern Sudan. Microbes Infect. 2005 Aug-Sep;7(11-12):1196–203.

P212 Healthcare-associated infections in the neurological intensive care unit: 6-year surveillance study at a major tertiary care center

YB Abulhasan1, S Rachel1, M Châtillon-Angle1, N Alabdulraheem1, I Schiller2, N Dendukuri2, M Angle1, C Frenette1

1Montreal Neurological Institute and Hospital, Montreal, Canada; 2McGill University, Montreal, Canada

Introduction: To report incidence rates, pathogens distribution, and patient related outcomes of healthcare-associated infections (HAIs) in a neurological intensive care unit (Neuro-ICU) patient population over a 6-year period.

Methods: We are presenting a prospective cohort study of all patients admitted in a 14 bed Neuro-ICU part of a highly-specialized referral center from April 1, 2010 to March 31, 2016. Surveillance for HAIs was carried by infection control professionals who reviewed laboratory results and targeted specific clinical indicators to match National Healthcare Safety Network infection criteria. Rates were calculated per 1,000 patient days and per 1,000 device days. Differences in infection rates were analyzed by emergency neurocritical care diagnostic categories. We studied the association between: i) primary diagnosis and infection using Cox proportional hazards model, ii) infection and length of stay using linear regression, and iii) infection and mortality using Cox proportional hazards model. Yearly objectives were set to reduce HAIs with implementation of targeted infection control measures.

Results: There were 6,034 Neuro-ICU admissions resulting in 20,845 Neuro-ICU days. A total of 228 HAIs were identified. Pooled mean HAI incidence rates were pneumonia 17.1, UTI 3.7, ventriculostomy-associated infection (VAI) 0.8, central line-associated blood stream infection (CLABSI) and primary bacteremia 0.2, Clostridium difficile-associated diarrhea (CDAD) 0.6, and other HAIs 0.7 per 1,000 Neuro-ICU days. For device-associated infections, which accounted for 80.7% of HAIs, pooled mean rates were 18.5 ventilator-associated pneumonia (VAP), 5.0 catheter-associated urinary tract infection (CAUTI), 4.0 VAI, and 0.6 CLABSI episodes per 1,000 device days. Among the various diagnostic categories, intracerebral/intraventricular hemorrhage (ICH/IVH), subdural hematoma, seizure/status epilepticus, and subarachnoid hemorrhage were associated with the highest pooled HAIs with incidence rates of 22.2, 21.3, 16.5, and 15.2 per 1,000 Neuro-ICU days respectively. Pathogen frequencies were S. aureus (27%) and Klebsiella species (12.2%) for pneumonias, E. coli (44.8%) for UTIs, S. epidermidis (57.9%) for VAIs. Among patients with HAIs, all-cause 30-day case mortality proportion was 9.7% and occurred a median of 14 days after the HAI. Prolonged Neuro-ICU length of stay was strongly associated with all HAIs (P < 0.05).

Conclusions: Pneumonia, UTI and VAI are the commonest HAIs among our cohort of neurocritical care patients with pooled rates of HAIs most pronounced among ICH/IVH patients.

P213 Paradoxical cerebrovascular hemodynamic changes with nicardipine

S Lahiri1, K Schlick1, SA Mayer2, P Lyden1

1Cedars-Sinai Medical Center, Los Angeles, CA, United States; 2Mount Sinai Medical Center, New York, United States

Introduction: IV nicardipine is commonly used for blood pressure reduction. Few studies have described its effects on cerebrovascular hemodynamics as measured by transcranial Doppler (TCD) waveform analysis and pulsatility index (PI).

Methods: The data presented are from patients who underwent TCD monitoring before, after, or during nicardipine administration.

Results: TCD waveforms during nicardipine infusion are characterized by a prominent systolic peak and dicrotic notch. Systolic deceleration is more pronounced and PIs are significantly elevated in patients on nicardipine (p < 0.05).

Conclusions: This study provides first evidence of paradoxical intracranial vasoconstriction associated with nicardipine. This is a consistent finding in patients treated with IV nicardipine and is contradictory to what is expected from a vasodilator and anti-hypertensive.

Fig. 13 (abstract P213).

Left panel without nicardipine, right panel with nicardipine

P214 Clinical manifestations and diagnosis of patients with cerebral venous thrombosis: retrospective study and literature review

M Akatsuka1, J Arakawa1, M Yamakage2

1Japanese Red Cross Kitami Hospital, Kitami, Japan; 2Sapporo Medical University School of Medicine, Sapporo, Japan

Introduction: Cerebral venous thrombosis (CVT) is a rare neurovascular disorder with a highly variable presentation that accounts for only 0.5% of all cases of stroke [1]. It can lead to neurologic impairment. There have been some studies and case reports about CVT; however, there is a lack of information on the incidence and clinical features of CVT due to its rarity. We, therefore, conducted this study to clarify important aspects of the epidemiology, diagnosis, and prognosis of CVT.

Methods: We carried out a retrospective observational study of all patients with a diagnosis of CVT in our hospital between January 2003 and October 2016. From the electronic patient data management system, we obtained information on the patient• fs age, gender, symptoms, risk factors, location of thrombosis, and outcome as well as images. We also performed a systemic review of the literature for CVT using a multiple web research platform (PubMed) from 2000 to 2016.

Results: Four patients were diagnosed with CVT in our hospital during the study period. A total of 17 articles were found, from which 1220 cases were determined eligible for review and the author• fs cases were added. The mean age of the patients was 39.3 years (standard deviation [SD]: 17.4 years), and 70.8% of the patients were female. The most frequent symptom was headache (884 cases, 72.2%). The use of an oral contraceptive was one of the most frequent predisposing risk factors (290 patients, 33.5%). For diagnosis of CVT, the majority of patients underwent computed tomography (CT) (80.5%), magnetic resonance imaging (MRI) (83%), and magnetic resonance venography (MRV) (66.8%). The most frequent site of involvement was the superior sagittal sinus (543 cases, 44.4%). The hospital mortality rate was 5.4%. There was no association between hospital mortality and location of thrombosis. The results of CT or MRI were normal in some cases, but the results of MRV were abnormal in all cases and lead to the diagnosis of CVT.

Conclusions: Patients presenting with headache, especially patients taking an oral contraceptive, should be examined carefully. Not only CT but also MRI in combination with MRV has high sensitivity and specificity for establishing a diagnosis of CVT. Although there is no typical red flag symptomatology, recognition of CVT is important.


[1] Bousser MG et al.: Lancet Neurol 6: 162–70, 2007

P215 Diaphragmatic electric activity during apnea testing for brain death determination

J Rubio1, JA Rubio Mateo-Sidron2, R Sierra1, M Celaya1, L Benitez1, S Alvarez-Ossorio1

1Hospital Universitario Puerta del Mar, Cadiz, Spain; 2Hospital Xanit, Benalmadena, Malaga, Spain

Introduction: Absence of respiratory control reflexes in the brainstem in response to hypercapnic stimulation through the observation of the thoracic and abdominal movements, positive apnea testing (AT), is a key component in the clinical assessment of brain death (BD). False negative results of AT may occur due to ventilatory auto-triggering which could hamper and delay BD determination. Electrical activity of the diaphragm (EAdi) reflects the neural respiratory drive. We hypothesized that EAdi monitoring would add accuracy and safety to AT procedure.

Methods: We performed a single centre prospective observational study of adult patients admitted to the ICU with devastating acute brain injury and clinical examination consistent with BD. All patients were mechanically ventilated using the Servo-i® and the conventional nasogastric tube was replaced with an EADi catheter (16Fr/125 cm; Maquet Critical Care, Solna, Sweden) designed to be used with neurally adjusted ventilation. Stable patients on ventilator control mode were switched to Pressure Support during the period of apnea. Respiratory movements, arterial pressure, hearth rhythm and SpO2 were continuously monitored and airway pressure, airflow and EADi were also saved using a Ventilation Record Card for later analysis. Study variables were recorded at 3 time points: basal (T1), start of AT (T2) end of AT (T3). AT duration and complications during the procedure were recorded.

Results: We included 8 patients in the study. EADi signals recorded during AT were tonic and with very low voltage. EADi amplitude values at T2 and T3 ranged between 0,2 ± 0,15 μV to 0,5 ± 0,52 μV and 0,2 ± 0,07 μV to 0,55 ± 0,52 μV respectively. The median procedure duration was 12 ± 2,5 min. Sensitive auto-triggering was observed in a case. The AT was positive and completed in all patients and no severe complications were noted.

Conclusions: We found that EADi monitoring and analysis is a safe and effective tool for diagnosing apnea during BD confirmation, avoiding false negative diagnosis based on direct observation. This strategy of rapid implementation, operator-independent, minimally invasive and low cost, may be introduced in the AT protocols.

P216 Usefulness of a method for doing apnea testing during brain death determination

J Rubio1, JA Rubio Mateo-Sidron2, R Sierra1, A Fernandez1, O Gonzalez1

1Hospital Universitario Puerta del Mar, Cadiz, Spain; 2Hospital Xanit, Benalmadena, Malaga, Spain

Introduction: Apnea Testing (AT) is a key element for Brian Death (BD) determination but accepted guidelines including the procedure itself do not exist. All current methods of AT have as main purpose to elevate CO2 and observe the patient for any spontaneous effort by close observation of respiratory movements that may sometimes be subtle and doubtful. Moreover, commonly AT practice involve disconnection of ventilator circuit, adding an adjustable CPAP valve to the distal extremity of the T-piece extension in order to try to preserve oxygenation. This study was aimed to evaluate an alternative AT method that allows the patient to stay connected to the mechanical ventilator circuit during AT procedure and avoid the disadvantages of the current methods.

Methods: We performed a single centre prospective observational study of adult patients admitted to the ICU with devastating acute brain injury and clinical examination consistent with BD. All patients were mechanically ventilated using the Servo-i® ventilator (Maquet Critical Care, Solna, Sweden). After 10 minutes of preoxygenation, controlled mode ventilation was switched to Pressure Support (PS) during the whole test, setting level of PS to 8 cmH2O above PEEP 5 cmH2O, Trigger Pressure -2 cmH2O and FiO2 100% with backup ventilation off. Patients were closely monitored throughout the procedure by clinical observation. Airway pressure, airflow and volume waveforms were displayed on the ventilator screen. Patient and screen data were saved to a Ventilation Record Card for later analysis. Duration of AT, invasive arterial pressure, heart rate and SpO2 were recorded. Physiologic, ventilatory mode and settings, respiratory mechanics and arterial blood gases were collected at 3 time points: basal (T1), start of AT (T2), end of AT (T3). After AT ventilator was switched to prior settings.

Results: Nine patients were studied. Main data collected are showed in the table. Analysis of pressure and flow waveform tracings showed absence of airflow and maintenance of PEEP at stable levels. A case of sensitive auto-triggering was clearly observed due to the selected PS level but the procedure did not have to be aborted. There were no complications or discontinued procedures.

Conclusions: This method appears to be safe and effective for AT during BD assessment. We speculate that these findings could have a significant impact if the strategy used really contributes to increase numbers of organ procurements and under the best possible conditions.

Table 9 (abstract P216). Results

P217 Effect of HHH therapy on CBF after severe subarachnoid hemorrhage: regional cerebral blood flow studied by bedside Xenon-enhanced CT

H Engquist, E Rostami, P Enblad

Uppsala University, Dpt of Neuroscience/Neurosurgery, Uppsala, Sweden

Introduction: Detection and management of delayed cerebral ischemia (DCI) after severe subarachnoid hemorrhage (SAH) is difficult, and tools are lacking to guide the therapy. In the present study, bedside Xenon-enhanced computerized tomography (XeCT) was used to assess regional cerebral blood flow (rCBF) prior to clinical suspicion of DCI and during treatment to resolve DCI.

Methods: Patients diagnosed with SAH and requiring mechanical ventilation, were prospectively enrolled in the study. XeCT, using inhaled stable Xenon as an inert contrast agent, was scheduled at day 0-3, 4-7 and 8-12. Patients clinically diagnosed with DCI received standard five-day treatment to augment CBF by hypervolemia, hemodilution and hypertension (HHH-therapy). XeCT at 0-2 days before start of HHH was considered as baseline, and next XeCT was performed during the therapy. Corresponding data were collected for non-DCI patients with XeCT measurements in matching time-windows (day 2-4 and 5-8 respectively).

Results: Nineteen patients who later developed DCI were included, and twenty-six patients without DCI were identified as a comparison group. There were no significant differences in the systemic hemodynamic parameters or pCO[sub]2[/sub] before vs during HHH, although a slightly elevated systolic blood pressure (SBP) was noted. Median global cortical CBF for the DCI group increased from 28.0 (IQR 24.6-34.5) to 37.8 (IQR 26.9-41.7) ml/100 g/min, P = 0.008. Median rCBF of the worst vascular territory increased from 19.8 (IQR 14.8-26.5) to 27.5 (IQR 17.4-34.5) ml/100 g/min, P = 0.033. For the group with no DCI, global CBF at baseline was higher and there was no significant change at the XeCT in the second time-window.

Conclusions: The initial low global CBF found in patients diagnosed with DCI, increased significantly during HHH-therapy despite modest changes of SBP. A concomitant increase in rCBF was also found in the vascular territories with worst rCBF. The increase in CBF may be related to the HHH-therapy, but a time-dependent natural recovery of CBF cannot be ruled out. XeCT may be helpful in managing poor grade SAH patients.

Fig. 14 (abstract P217).

CBF at baseline and during HHH-therapy in DCI

P218 Spontaneous primary intracerebral hemorrhage: factors influencing poor outcome. A two-centers series

J Toledo1, A Ramos1, F Acharta1, L Canullo1, J Nallino1, A Dogliotti2, C Lovesio1

1Sanatorio Parque, Rosario, Argentina; 2Grupo Oroño, Rosario, Argentina

Introduction: Patients presenting intracerebral hemorrhage (ICH) have commonly been reported to have a poor prognosis. We prospectively analyzed data from 98 patients with spontaneous primary ICH to evaluate possible predictors of poor outcome defined as a Glasgow Outcome Scale (GOS) < =3 at hospital discharge after ICH.

Methods: Between june 1, 2013 and september 30, 2016; 98 patients with ICH were treated. On admission data including patient characteristics, clinical findings, radiologic features, and functional neurologic outcome were assessed and further analyzed.

A multivariate analysis was performed to identify predictors of poor outcome.

Results: 98 patients were included in the study and 46% of them had an unfavorable evolution. In multivariate analysis, patients taking antiplatelet therapy at hospital admission (OR: 15.7, 95% CI: 1.03-241; p = 0.047). and elevated ICH score had worse outcome. For every point of ICH score, the probability of poor outcome increases in 43% (p < 0.001).

Conclusions: We have found that the ICH score and receiving antiplatelet agents are associated with poor prognosis in patients with spontaneous ICH.

P219 Agitation after mild to moderate traumatic brain injury in the intensive care unit

M Perreault 1, J Talic2, AJ Frenette3, L Burry4, F Bernard3, DR Williamson3

1The Montreal General Hospital, Montreal, Canada; 2University of Geneva, Geneva, Switzerland; 3Hôpital Sacré-Coeur de Montreal, Montreal, Canada; 4Mount Sinai Hospital, Toronto, Canada

Introduction: Traumatic brain injury (TBI) is a leading cause of mortality and disability worldwide. Among TBI complications, agitation is a frequent behavioural problem. Agitation leads to harm and treatment interference, unnecessary chemical and physical restraints, increased hospital length of stay, delayed rehabilitation and impedes functional independence. Agitation is reported in 70% of TBI patients in rehabilitation units but not well described during ICU. The objective of this study was to describe the frequency, timing and clinical impact of agitation in mild (GCS 13-15) to moderate (GCS 9-12) TBI in the intensive care unit (ICU).

Methods: This was a prospective observational study of mild to moderate TBI adult patients in two critical care trauma centers. Patients aged > = 18 years admitted for more than 48 hours were screened. We excluded patients with severe TBI (GCS 3-8), RASS -4 or -5 during ICU stay, pre-existing cognitive deficiency, inability to communicate in French or English and deafness or blindness were excluded. Agitation was defined as a RASS score of +2 or +3 and assessed during three time periods (00:00 to 07:59, 08:00 to 15:59 and 16:00 to 23:59). Agitated and non-agitated patients were compared using a T-test or Mann Whitney as appropriate for continuous variables and Chi-square test for dichotomous variables.

Results: During an 8-month period, 226 patients were assessed for eligibility. In total, 62 patients (74.2% men) with a median age of 58 years (IQR 33), a mean APACHE II score 11.3 ± 6.5 and mean ISS of 23.3 ± 9.3 were enrolled. At least one episode of agitation was reported in 72.6% of patients. Persistent agitation defined as at least 2 episodes on separate days was reported 54.8% of patients. Persistently agitated patients were more often mechanically ventilated (79.4% vs 46.4%; p = 0.01), had more moderate TBIs (47.1% vs 10.7%; p = 0.02) and had higher APACHE II scores (12.6 +/- 6.1 vs 9.8+/- 6.6; p = 0.09). Episodes of agitation presented at night-time, daytime and evening in 20.6%, 22.8% and 22.9%, respectively. Persistent agitation was associated with a longer length of stay (9.1 vs 4,7 days; p = 0.001), increased restraint use (82.4% vs 28.6%;p < 0.001), increased antipsychotic use (55.9% vs 7.1%; p < 0.001) and increased self-removal of arterial and central venous catheters (29.4% vs 3.6%;p = 0.009) and nasogastric tubes (20.6% vs 3.6%;p = 0.06).

Conclusions: Agitation is common in mild to moderate TBI ICU patients, occurs at all times of the day and is associated with increased use of restraints, antipsychotics and potential self-harm.

P220 Acute subdural hematoma after isolated traumatic brain injury: associated factors and prediction of lethal outcome

D Adukauskiene1, J Cyziute2, A Adukauskaite3, L Malciene4

1Hospital of Lithuanian University of Health Sciences, Kaunas, Lithuania; 2Lithuanian University of Health Sciences, Kaunas, Lithuania; 3Innsbruck Medical University Hospital, Innsbruck, Austria; 4Klaipeda University Hospital, Klaipeda, Lithuania

Introduction: The aim of this study was to determine associated factors with lethal outcome also prediction of it in case of acute subdural hematoma (ASH) after isolated traumatic brain injury (ITBI).

Methods: Retrospective analysis of 162 patients with ASH after ITBI treated in Neurosurgical Intensive Care Unit in Hospital Kaunas Clinics of Lithuanian University of Health Sciences during two years was carried out.

Results: Sixty–seven patients (41%) of 162 have died with ASH after ITBI. Twelve patients (31%) of 39 have died in age group of < = 44 years, 16 patients (37%) of 43 in group of 45 – 54 years, 14 patients (36%) of 39 in group of 55 – 64 years, but 25 patients (61%) of 41 in group of > = 65 years, p < 0.003. Twenty–four patients (26%) of 93 have died with pupillary light reflex and 43 patients (62%) of 69 without of it, p < 0.001. Five patients (25%) of 20 have died in group of Glasgow Coma Scale (GCS) score 12 – 15, 5 patients (20%) of 25 in group of 9 – 11, but 57 patients (49%) of 117 in group of 3 – 8, p < 0.002. Twenty–one patient (28%) of 75 has died with white blood cell count <10.1 x 10^9/l, but 46 patients (53%) of 87 with > = 10.1 x 10^9/l, p < 0.001. Five patients (17%) of 30 have died with glycemia 3.3 – 5.5 mmol/l and 62 patients (47%) of 132 with glycemia > = 5.6 mmol/l, p < 0.001. One patient (5%) of 22 has died in group of APACHE II score < = 10 points, 12 patients (24%) of 50 in group of score 11 – 15, but 54 patients (60%) of 90 in group of score > = 16, p < 0.001. Fifty–two patients (63%) of 82 have died in the group of estimated lethal outcome risk >25%, p < 0.001 (0.95CI: 0.53 – 0.74) and prognostic test sensitivity was found to be 78%, specificity 76%.

Conclusions: The mortality rate of acute subdural hematoma after isolated traumatic brain injury was 41%. Factors associated with lethal outcome were estimated to be age > = 65 years, absence of pupillary light reflex, Glasgow coma scale score 3 – 8, white blood cell count > = 10.1 x 10^9/l, glycemia > = 5.6 mmol/l and APACHE II score > = 16 points on the first day after trauma. Predicted lethal outcome has coincided with real mortality when the risk of lethal outcome was higher than 25%.

P221 Multimodal monitoring in critically ill polytrauma patient with traumatic brain injury (TBI)

L Luca1, A Rogobete1, O Bedreag1, M Papurica1, M Sarandan2, C Cradigati2, S Popovici1, C Vernic1, D Sandesc1

1University of Medicine and Pharmacy ”Victor Babes” Timisoara, Timisoara, Romania; 2Emergency County Hospital ”Pius Brinzeu”, Clinic of Anesthesia and Intensive Care ”Casa Austria”, Timisoara, Romania

Introduction: The most significant injury found in trauma patients is represented by traumatic brain injury which has the greatest impact on mortality. Intracranial pressure (ICP) monitoring is required in severe traumatic head injury because it optimizes treatment based on ICP values and cerebral perfusion pressure, respectively (CPP). The objective of the study was to identify the incidence of systemic complications of acute TBI and establishing a plan for diagnostic and therapeutic approach.

Methods: From a total of 64 patients admitted in the Intensive Care Unit "Casa Austria", Emergency County Hospital “Pius Brinzeu” Timisoara, Romania, between January 2016 and August 2016, only patients who received ICP monitoring (n = 10) were analysed. In the control group were analysed 13 patients. Depending on the time passed since trauma and to the time of ICP monitoring, the patients were divided into 3 categories (<18 hours, 19-24 hours, > 24 hours). There were also compared a series of clinical outcomes between the groups.

Results: Monitoring of ICP initiated after 36 hours of trauma reveals significantly higher values compared to monitoring in 24 hours ICP on demonstrating effective monitoring therapeutic management of critical patient. In monitoring patients who ICP was installed in the first 18 hours after trauma the length of stay in ICU was 12.5 ± 5 days compared to those from which the installation was done after 36 hours of trauma where the length of stay in ICU was 21.75 ± 9.64 days. After 120 hours of aggressive monitoring and therapy in accordance with the parameters analyzed, there is a decrease in the ICP and a normalization of values MAP or PPC. For patients who install monitoring was conducted in the first 18 hours after trauma, the mean of ICP was 22.99 ± 12.47 for those from which the installation was done within 19 to 24 hours after trauma the average value for ICP was 11.02 ± 2.09, and for patients whose installation was carried out after 36 hours of trauma, the mean of ICP was 29.86 ± 18.82 . It reported a relationship statistically significant between ICP and MAP values (p = 0.0008), respectively ICP and PPC (p = 0.0284), confirming that under impaired autoregulation of cerebral growth PIC TAM induce growth without significantly improving cerebral perfusion.

Conclusions: The multimodal monitoring in the management of the trauma patient a significant influence on the rate of survival.

P222 Hypothermia in management of traumatic brain injury

V Avakov1, I Shakhova2

1Tashkent Medical Academy, Tashkent, Uzbekistan; 2Université de Strasbourg, Strasbourg, France

Introduction: It is found that high body temperature in patients with high intracranial pressure is burdening association [1] due neuroexcitotoxicity [2], which further damage the nerve cells and stimulate the autoimmune processes [3]. Mechanisms of injury, exacerbated by hyperthermia, can be mitigated by mild hypothermia.

Methods: We studied 47 patients (18-72 years) with traumatic brain injury. Level of consciousness of patients at admission to hospital was 3-8 points of GCS. According to the basic conditions of management and treatment, the patients were identicals.

However, first group of patients (n = 27) during initial 6 hours has been assigned to brain hypothermia by combination of nasopharyngeal cooling with cooling of cranial vault and carotid bifurcation projection (Fig. 15).

Duration of cooling was 33,5 ± 4,2 h; brain target temperature (in the ear canal)–33-34 °C, body (axillary)–35,5-36,5 °C.

Results: Malignant hyperthermia central genesis in patients of first group was not observed, while in the second group, its frequency increased from 30 to 55%.

Duration of patient’s stay in ICU and hospital was shorter by 1.4 times in first group; GOS was the most favorable in first group-4.2 points compared with 3.7 points in second group. Mortality rate was 42% versus 65%, respectively.

Conclusions: Nasopharyngeal cooling, added to the conventional had cooling for suppress the severity of ischemic processes in brainstem and the activity of thermoregulation center, promoted neuroprotection, improved neurological outcomes, survival, and reduced duration and cost of treatment.


1. Zdravev P., 1951

2. Irazuzta J.E. et al., 1999

3. Prandini M.N. et al., 2005

Fig. 15 (abstract P222).

See text for description

P223 Citicoline in severe brain trauma: matched pair analysis suggests improved outcome

H Trimmel 1, M Majdan2, GH Herzer1

1Landesklinikum Wr. Neustadt, Wiener Neustadt, Austria; 2Department of Public Health, 91701 Trnava, Slovakia

Introduction: Goal-oriented management of severe traumatic brain injury (sTBI) from emergency site to intensive care unit can save the lives of millions of affected patients worldwide and/or improve their long-term outcome thus enhancing quality of life and saving enormous socio-economic costs. However, promising sTBI treatment strategies with neuroprotective agents, such as citicoline (CDP-choline), lacked evidence or produced contradictory results in clinical trials, some of them maybe due to inappropriate study design or insensitive methodology. As preceding evaluations provided evidence of beneficial outcome in citicoline treated sTBI patients at Wiener Neustadt Hospital (WNH), we aimed to investigate the potential role of citicoline administration in those patients.

Methods: In the course of a prehospital TBI project to optimize early TBI care within 14 Austrian Level I trauma centers, data on 778 TBI patients were prospectively collected. Unexpected superior outcome of WNH patients gave the impetus to compare patients from WNH with citicoline administration and matched patients from the other Austrian centers without citicoline use in a retrospective subgroup analysis. Patients with Glasgow Coma Scale score < 13 on site and/or Abbreviated Injury Scale of the region “head” > 2 were included.

Results: Our analysis revealed significantly reduced rates of ICU mortality (5% vs. 24%, p < 0.01), hospital mortality (9% vs. 24%, p = 0.035) and six months mortality (13% vs. 28%, p = 0.031), as well as of unfavorable outcome (34% vs. 57%, p = 0.015) and observed vs. expected ratio for mortality (0.42 vs. 0.84) in the citicoline group. Adjusted OR revealed significantly better odds for ICU survival (OR = 6.7, p = 0.014) and six months favorable outcome (OR = 2.6, p = 0.022) in citicoline treated patients.

Conclusions: Despite the limitations of a retrospective subgroup analysis the findings suggest a correlation between early and consequent citicoline administration and beneficial outcomes. Therefore, we aim to set up an initiative for a prospective randomized controlled trial with citicoline in sTBI patients.


1. Secades JJ. Citicoline: pharmacological and clinical review, 2010 update. Revista de neurologia. 2011;52 Suppl 2:S1–s62.

2. Brazinova A, Majdan M, Leitgeb J, Trimmel H, Mauritz W. Factors that may improve outcomes of early traumatic brain injury care: prospective multicenter study in Austria. Scandinavian journal of trauma, resuscitation and emergency medicine. 2015;23:53.

Table 10 (abstract P223). Demographic and baseline characteristics of the treatment groups
Table 11 (abstract P223). Outcomes at hospital discharge and six months post trauma

P224 Early measurement of low cerebral blood flow is associated with brain hypoxemia after traumatic brain injury

CS Sokoloff1, M Albert1, D Williamson1, C Odier2, J Giguère1, E Charbonney1, F Bernard1

1Hôpital du Sacré-Coeur de Montréal, Montréal, Canada; 2Centre Hospitalier Universitaire de Montréal, Montreal, Canada

Introduction: Management of traumatic brain injury (TBI) can include monitoring of brain tissue oxygenation (PbtO2) to prevent secondary brain injury. A few studies suggest measuring blood flow velocity could predict cerebral hypoxia [1]. Cerebral oxygenation measured by PbtO2 depends on blood O2 content, cerebral blood flow (CBF) and diffusion. We aimed to study the relationship between flow velocity using trans-cranial Doppler (TCD) in patients with moderate to severe TBI and PbtO2 during cerebral hypoxic episodes.

Methods. Serial TCD studies were done to assess CBF velocity of the middle cerebral artery (MCA). Measurements were done bilaterally after the insertion of PbtO2 monitoring, on a daily basis for 5 days, and during auto-regulation, vaso-reactivity and hyperoxic challenge tests when feasible. Various parameters were collected simultaneously: PbtO2, PaO2, PaCO2, Hb level, ICP, CPP and cardiac index.

Results. A total of 85 TCD studies in 17 consecutive patients were obtained. We observed 29 episodes of cerebral hypoxia (PbtO2 ≤ 20 mmHg) distributed as follow: 10 episodes of PbtO2 ≤ 10 mmHg, 5 episodes of 10-15 mmHg, and 14 episodes of 16-20 mmHg. Overall, no correlation between PbtO2 and MCA’s mean blood flow velocity (Vmean) was found. For TCD studies obtained within 24 h from trauma (N = 14), there was a weak correlation between Vmean and PbtO2 (r2 = 0.41; p = 0.035) with median values of Vmean of 42.9 ± 38.4 cm/s and of PbtO2 of 14.95 ± 8.74 mmHg. Among these, all Vmean < 40 m/s (n = 6) were associated with a PbtO2 ≤ 20 mmHg. We did not find any association with other factors.

Conclusions. Low CBF velocity (<40 m/s) measured with TCD within the first 24 h of TBI is associated with brain tissue hypoxia measured with PbtO2. Interventions to optimize CBF and O2 content immediately after TBI could help minimize early secondary injuries.


[1] van Santbrink H. et al. Serial Transcranial Doppler Measurements in Traumatic Brain Injury with Special Focus on the Early Posttraumatic Period. Acta Neurochir 2002: 144: 1141–1149

P225 The prognostic role of tractography in the management of traumatic brain injury

Z Husti, T Kaptás, Z Fülep, Z Gaál, M Tusa

Bács- Kiskun County Hospital, Kecskemét, Hungary

Introduction: The management of patients with traumatic brain injury (TBI) is a great challenge for intensive care units. In the case of TBI patients without obvious space occupying lesion and with stable cardiorespiratory condition, available conventional imaging techniques (CT, MRI) may be insufficiently sensitive to predict long- term neurological outcome. Tractography (TG) is a special technique of MRI using diffusion weighted images (DWI) which enable to visualize neuronal networks formed by connections among cortical and subcortical regions. This method be able to early detect diffuse axonal lesion that may draw attention to poor outcome unlike conventional CT or MRI. The aim of our study was to determine the role of TG in the diagnosis of TBI without elevated ICP regarding to the prognosis.

Methods: Diffusion tensor imaging (DTI) TG was carried out by applying Philips Achieva 1,5 T, software 5.2.0. Tractographic images were constructed by Medlnria 2.2.3 and DSI studio 2.2016.11.02 software. TG was performed on four selected TBI patients characterized by GCS 3 with stable cardiorespiratory condition and sufficient oxygenation at praehospital period. Neither of them showed increased ICP confirmed by imaging techniques and intracranial pressure monitoring. Patients were followed- up regarding to the neurological outcome. The relationships between the severity of TG images and prognosis were investigated.

Results: DTI TG obtained from all examined patients showed radical destruction of the tracts in the white matter as seen on demonstrated pictures comparing with an anisotropic map from a healthy individual. Neurological results of patients showed as the following: The first patient remained unconsciousness in a stable vegetative condition. The second patient follows simple instructions, however, permanently aphasic. The third patient developed neurological improvement after 3 weeks from the injury but died during rehabilitation. In the fourth case where TG showed the most serious damage comparing with control recording, neurological improvement was not observed and the patient died after 10 days from the trauma. These results show that TG appropriately predicted a poor neurological outcome in patients with TBI despite obvious space occupying lesion.

Conclusions: TG may be a suitable imaging technique for predicting prognosis and can be a significant part of the decision making process during TBI management. However, further large case number of studies are needed to define the prognostic value of this method in the course of TBI management.

P226 Optimal cerebral perfusion pressure; bedside application after severe traumatic brain injury

J Donnelly1, M Aries2, M Czosnyka1, C Robba3, M Liu1, A Ercole3, D Menon3, P Hutchinson4, P Smielewski1

1Brain Physics Laboratory, Division of Neurosurgery, Department of Clinical Neurosciences, Cambridge Biomedical Campus, University of Cambridge, Cambridge, United Kingdom; 2MUMC, Maastricht, Netherlands; 3Division of Anaesthesia, Department of Medicine, Addenbrooke’s Hospital, University of Cambridge, Cambridge, United Kingdom; 4Division of Neurosurgery, Department of Clinical Neurosciences, Addenbrooke’s Hospital, University of Cambridge, Cambridge, United Kingdom

Introduction: Previously, a cerebrovascular reactivity-based method that yields continuous estimates of optimal cerebral perfusion pressure (CPPopt) has been developed using a TBI cohort. Using recent CPPopt data from patients after severe TBI, we aimed to assess the real-time clinical value of CPPopt by testing the hypothesis that bedside CPPopt data without artifact removal remains an independent determinant of patient outcome.

Methods: Single center cerebral monitoring data from severe TBI patients admitted between 2010 and 2015 were used. Treatment was guided by controlling absolute values of ICP and CPP. CPPopt was determined using a previously published curve-fitting protocol. No manual cleaning of signal artifacts was performed. The difference between current CPP and CPPopt, coupled with the current PRx value (with a threshold for impaired autoregulation (PRx) of +0.15), was used to determine whether each CPP was below the lower limit of reactivity (LLR), above the upper limit of reactivity (ULR) and within the reactivity limits. The time each patient spent below, above or within these reactivity limits was compared across GOS groups (Fig. 16).

Results: ICP monitoring data from 231 patients were available. Time spent with CPP within reactivity limits or the time spent with CPP below the LLR were both associated with GOS, independent of age, GCS motor score, pupillary reactivity and mean ICP (OR for %time CPP below LLR = 0.96 (0.94-0.98), OR for %time CPP within reactivity limits = 1.03 (1.01-1.05).

Conclusions: This observational report demonstrates that recent prospective bedside monitoring of CPPopt after severe TBI is both practical and retains independent prognostic significance.

Fig. 16. (abstract P226).

Time spent within (a) or below (b) CPP pressure reactivity limits

P227 Regional brain tissue oxygen tension normalized to arterial oxygen tension predicts outcome in traumatic brain injury

R López, J Graf, JM Montes

Clínica Alemana de Santiago, Santiago, Chile

Introduction: Regional brain tissue oxygen tension (PbtO2) depends on regional perfusion, regional oxygen extraction and arterial oxygen tension (PaO2). If PbtO2 is to be used as a brain tissue oxygen extraction variable it should be normalized for PaO2. We explored the outcome prediction potential of normalized PbtO2 [PbtO2N = (PaO2-PbtO2)/PaO2] as compared to PbtO2, intracranial pressure (ICP) and cerebral perfusion pressure (CPP) in a cohort of traumatic brain injury (TBI) patients.

Methods: Retrospective analysis of 14 patients with severe TBI and a PbtO2 probe inserted on proximity at brain parenchymal lesion. Functional outcome was classified using the Glasgow Outcome Scale-Extended (GOSE). We considered > =5 points as favorable outcome (FO). Areas under the receiver operating characteristic curves (AUC) for outcome prediction were determined for the fraction the initial 24 h (%t) with high ICP, %t with CPP <60 mmHg, %t with PbtO2 < 15 mmHg and PbtO2N at 24 h.

Results: Seven patients had FO. Table 12 shows demographic and physiological data and Fig. 17 shows time course of PbtO2 according to outcome. Patients with FO exhibited lower initial and subsequent rise in PbtO2 values. PbtO2N displays the best AUC (0.939, CI = 0.818-1, p = 0.006) among the variables tested for outcome prediction with values >0.80 associated to FO.

Conclusions: In agreement with previous reports [1], low initial PbtO2 value with gradual recovery is associated with FO. PbtO2 normalized for PaO2 at 24 h seems better than PbtO2, ICP or CPP for functional outcome prediction.


[1] Ponce LL, et al.: Neurosurg 2012; 70:1492–1503

Table 12 (abstract P227). Demographic and physiologic data according to functional outcome
Fig. 17 (abstract P227).

Time course of PbtO2 according to functional outcome

Fig. 18 (abstract P227).

ROC curves for outcome prediction

P228 Predictive value of common intensive care severity scores in traumatic brain injury

M Kenawi1, A Kandil2, K Husein1, A Samir1

1Cairo university hospital, Cairo, Egypt; 2Helal hospital, Cairo, Egypt

Introduction: Traumatic Brain Injury (TBI) causes a severe toll on society as a leading cause of mortality worldwide and the major cause of disability among young adults. The prognosis after TBI had been particularly challenging to predict, with limited availability of robust prognostic models. We aimed to evaluate the usefulness of the APACHE II (Acute Physiology and Chronic Health Evaluation II), SAPS II (Simplified Acute Physiology Score II) and SOFA (Sequential Organ Failure Assessment) scores compared to simpler models based on age and Glasgow Coma Scale (GCS) in predicting a six-month mortality of patients with moderate to severe traumatic brain injury (TBI) in the intensive care unit (ICU).

Methods: A Prospective cohort study conducted on acute TBI patients admitted to I.C.U at EL-HELAL trauma Centre and KASR AL AINI university hospital, Egypt during the period from August 2014 to April 2015. All patients were followed-up for 6 months from the day of admission. Our patients were divided into two groups (survivors and non-survivors).

Results: A total of 104 patients were enrolled. Mean age was 37 ± 17.16 years, the overall six-month mortality was 25 patients (24.4%). The univariate analysis showed that APACHE II, SAPS II, SOFA, GCS, and age had a significant statistical difference regarding mortality between both groups (P-value < 0.05) and the optimal cut-off point as mortality indicator was 14, 26, 4, 9 and 49 respectively with area under the curve (AUC) 0.88, 0.87, 0.83, 0.80 and 0.79 respectively. By Multivariate analysis using logistic regression, we found only Age and GCS had a statistically significant impact on outcome (P-value; 0.001, 0.022).

Conclusions: A simple prognostic model based only on GCS and Age displayed good predictor for six-month mortality of ICU treated patients with TBI. The use of the more complex scoring systems (APACHE II, SAPS II and SOFA) added little to the prognostic performance.

P229 Development of a process indicator-based plan-do-act-check cycle to improve quality of care in severe traumatic brain injury patients

J Heijneman, J Huijben, F Abid-Ali, M Stolk, J Van Bommel, H Lingsma, M Van der Jagt

Erasmus Medical Center, Rotterdam, Netherlands

Introduction: Treatment of traumatic brain injury (TBI) patients aims at secondary brain injury prevention. Intracranial pressure (ICP) is a crucial parameter. Process indicators, which refer to appropriateness of delivered care, can be used to monitor protocol adherence. However, use of quality indicators for TBI and ICP has been scarce. We developed and applied an ICP process indicator and additionally constructed a Plan-Do-Check-Act (PDCA) cycle describing key learning points to improve quality of care.

Methods: The study took place at an academic neurotrauma center ICU. We used focus group interviews to reach consensus on a practical ICP indicator. The indicator was applied to adult TBI patients receiving ICP monitoring (Apr.-Sept. 2016). Patients with infaust prognosis were excluded. Data on ICP levels and applied treatments were collected. Protocol non-adherence (ICP > 20 mmHg, > = 30 minutes without appropriate escalation of treatment according to protocol (Fig. 19) was first assessed by a research nurse and subsequently interpreted by an ICU-fellow and neurointensivist. Details on non-adherence were assessed and incorporated in a PDCA-cycle construct.

Results: We analysed 43 patients of whom 5 had an infaust prognosis, resulting in 38 included patients. Protocol adherence was inadequate in 3 cases (8%) (Fig. 19). In 1 patient there was an adequate intervention but not within the set time, while in 2 patients therapy was inadequate (failure to control fever, inadequate osmotic therapy). With these data we constructed a PDCA-cycle (Fig. 20).

Conclusions: We showed feasibility of an ICP process indicator for protocol adherence to construct a simple PDCA-cycle. Apart from repeating the cycle, striving for 100% adherence, future steps may include protocol adaptation, education and assessment of a possible association between ICP-indicator metrics and patient outcomes.

Fig. 19 (abstract P229).

ICP Indicator metrics and ICP management escalation stages I-III

Fig. 20 (abstract P229).


P230 Skin cover plasty after decompression craniectomy, as a method of solving intracranial hypertension – case history

RC Cihlar

Nemocnice Ceske Budejovice, C eske Budejovice, Czech Republic

Introduction: A possibility to solve intracranial hypertension by means of a skin cover plasty in patient after decompression craniectomy, which is followed by a further increase in ICP.

Methods: The case history of a man, aged 29, suffering from a severe TBI. ICP was monitored by the parenchymatous sensor inserted into the left frontal lobe in the usual manner. After conservative procedures of intracranial hypertension treatment had been exhausted, a broad decompression craniectomy was conducted on the right, which led to a decrease in ICP. During 48 hours, a repeated increase in ICP occurred in the patient, with signs of intracranial hypertension in a CT examination. After considering all possibilities, the decompression craniectomy was further enlarged by dissolving the skin suture and conducting a plastic surgery of the damage caused, by means of the combined COM 30 bandaging fabric. This is a temporary fabric cover that substitutes the skin and is used especially in the treatment of burns, as a cover of skin transplants.

Results: This procedure led to a decrease in ICP and the disappearance of CT signs of intracranial hypertension in the patient. No infectious, bleeding or necrotic complications were observed. After the remission of the cerebral oedema, COM 30 was removed after 14 days, and a suture of the skin cover was conducted above the decompression craniectomy. The patient left hospital with a satisfactory neurological finding, with GOS 5.

Conclusions: The use of the artificial skin cover in the patient with decompression craniectomy resulted in a distinct, lasting decrease in intracranial hypertension below the value of 20 mmHg. This method may be recommended as an extreme procedure to solve intracranial hypertension in patients with decompression craniectomy and a continuing increase in ICP. Informed consent to published has been obtained from the patient.

Fig. 21 (abstract P230).

Photographic documentation from the operating theatre



P232 Haemorrhagic shock in war wounded: mortality at a surgical center for war victims in Afghanistan

G Mancino, P Bertini, F Forfori, F Guarracino

Azienda Ospedaliero Universitaria Pisana, Anaesthesia And Intensive Care Department, Pisa, Italy

Introduction: A small percentage of war wounded patients is hypotensive on arrival, requiring aggressive fluid resuscitation before surgery. This population has been described, in order to identify factors related with morbidity and mortality.

Methods: Patients who presented to the ER with hemorrhagic shock and a SBP lower than 90 mmHg were enrolled. Data were collected with regards to timing and mechanism of trauma, number of organs involved, the type and amount of fluid used for the resuscitation, the number of units of whole blood transfused, the surgery performed, the body temperature and blood gas data in the immediate postoperative time. The development of complications, the duration of mechanical ventilation, the use of α dose dopamine, the mean length of stay in ICU and the in-hospital mortality were collected as well. Injury severity score (ISS) and new injury severity score (NISS) were calculated retrospectively.

Results: 60 patients with a mean ISS of 17.88 ± 5.76 were enrolled. 38.3% of the patients developed complications during the stay, and the overall in hospital mortality was 26.6% (n = 16). Patients who died were older than survivors, had higher ISS and NISS (respectively 23.50 ± 1.19 and 26.38 ± 9.65) and had received a preoperative infusion of colloids. Mortality rate was higher for those who underwent damage control surgery and who received a greater amount of ringer lactate for resuscitation (2.85 ± 0.45 liters vs 1.98 ± 0.16 liters, p = 0.0241). In the postoperative period all the variables taken into consideration showed a statistically significant correlation with mortality: survivors had higher values of body temperature, pH and hemoglobin and lower levels of lactates and base deficit. The transfusion of a higher number of units of whole blood, the infusion of dopamine and the development of complications had a strong correlation with mortality.

Conclusions: Given a relatively low mortality rate, patients with a worse outcome were those who suffered from hypothermia, acidosis and coagulopathy. The study showed that a higher amount of fluids and of units of blood transfused was associated with a higher mortality rate, maybe through its hemodiluition effect and the consequent development of coagulopathy. The data collected confirmed the association between mortality and ISS, NISS, amount of cristalloids and of blood units used for resuscitation. These results encourage the use of damage control resuscitation in selected trauma patients, in order to promptly control the hemorrhage and reduce the amount of fluid administered to patients suffering from hemorrhagic shock.

P233 Late versus early surgical fixation of femoral bone fracture could increase the neurocognitive dysfunction and postoperative morbidity of the eldery patients

D Pavelescu, I Grintescu, L Mirea

Emergency Hospital Floreasca, Bucharest, Romania

Introduction: Neurocognitive dysfunction is a particular problem in elders with 30-80% becoming delirious after major surgery and 30-40% developing early cognitive dysfuntion (POCD).

The cognitive morbidity is important, delirium and POCD were associated with longer hospital stay and costs, with higher morbidity and mortality.

Methods: After written informed consent and approval by institutional review board, 112 patients, 65-101 years old admitted in Emergency Hospital with traumatic femoral bone fracture, were enrolled in an prospective, observational study. Surgical fixation was made early in 68 patients (allocated in group A) and was delayed for at least 48 hours in the other 52 patients (allocated in group B), for different reasons (chronic anticoagulation, comorbidities). Surgical procedure for all patients was done under spinal anesthesia. All patients were evaluated at admission and on day 7 from admission. We asssess the incidence of pre and postoperative neurocognitive dysfunction using Informant Questionnaire on Cognitive Decline in the elderly, which range from 1 to 5, supplemented with MMSE. A score of more or equql of 3,38 indicates neurocognitive dysfunction. We also evaluate the presence of aquired neuromuscular weakness on day &, using Medical Research Council Score (<48), the incidence of postoperative complications (pulmonary Thromboembolism, pneumonia) and mortality. Statistical analysis was made using SPSS tools, unpaired t-test and Mann-Whytney U test, a p-value < 0,005 was considered statistically significant.

Results: No significant difference concerning neurocognitice dysfunction prior surgery between the two groups (10,8% vs 9, 2%). here is a significant higher incidence in postoperative neurocognitive dysfunction in group B (14,7% vs 66,6%), also a higher incidence of thromboembolism and pneumonia in group B. Aquired neuro-muscular weaness appear early in the first week and there is a strong correlation with late surgical intervention and delayed mobilisation. The mortality is higher in group B (5,55% vs 1,49%, p-value < 0,005).

Conclusions: Neurocognitive dysfunction is an important complication that could induce the development of aquired neuro-muscular weakness by late mobilisation, could impede the recovery and dramatically increase the postoperative morbidity and mortality.

Early surgical fixation of femoral bone fracture could improve the clinical and neurological status of the elderly patients and increase the quality of life.

P234 The effictiveness of prophylactic inferior vena cava filters insertion in trauma patients

S Alamri1, M Tharwat1

1Riyadh National Hospital, Riyadh, Saudi Arabia

Introduction: Trauma patients are at high risk of developing venous thromboembolism (VTE) including deep venous thrombosis and pulmonary embolism (PE). The epidemiology of VTE in trauma patients showed that PE is the third major cause of death after trauma in patients who survive longer than 24 hours after onset of injury [1]. Besides, patients recovering from trauma have the highest rate of VTE among all subgroups of hospitalized patients. PE following development of DVT is one of the most preventable causes of death in hospitalized patients [2].

IVC filters have been suggested in some studies to decrease the risk of PE in various patient populations including the critically ill and trauma patients [2].

Methods: 32 trauma patients were admitted to ICU at National Care hospital in Riyadh,KSA with pelvic or femur fractures in the period from 4/2015 to 10/2016, all candidate patients were started on prophylactic anticoagulation with enoxaparin, were followed during their ICU stay and up to the hospital discharge, for development of pulmonary embolism and to compare those who had IVC filter inserted and those who didn’t have IVC filter, in relation to the development of pulmonary embolism events

Results: 32 adult trauma patients were enrolled in the study. 5 patients had no IVC filter inserted(15.6%). 2 patients of the no IVC filter group had PE (40%) (p = 0.000689). 27 patients had IVC filter(84.4%), none of them had positive PE. 2 patients died in the study but only one died of major PE and this patient had anticoagulation started but with no IVC filter inserted.

Conclusions: Trauma patients with lower limb long bone or pelvic fractures are at high risk of thromboembolic events including major PEs. Early insertion of IVC filter along with prophylactic anticoagulation prevented the development of PE. These results need to be confirmed in large scale randomized trial.


1. Sachdeva A, Dalton M, Amaragiri SV, Lees T. Elastic compression stockings for prevention of deep vein thrombosis. Cochrane Database Syst Rev. 2010 Jul 7;(7)

2. Torbicki A, Perrier A, Konstantinides S, Agnelli G, Galiè N, Pruszczyk P, et al. Guidelines on the diagnosis and management of acute pulmonary embolism: The Task Force for the Diagnosis and Management of Acute Pulmonary Embolism of the European Society of Cardiology (ESC) Eur Heart J. 2008;29:2276–315.

Fig. 22 (abstract P234).

See text for description

P235 Complications following resuscitative endovascular balloon occlusion of the aorta in patients with polytrauma

N Kono, H Okamoto, H Uchino, T Ikegami, T Fukuoka

Kurashiki Central Hospital, Kurashiki City, Japan

Introduction: Resuscitative endovascular balloon occlusion of the aorta (REBOA) is a temporary bleeding control method for patients who experience hemorrhagic shock. Since it is less invasive than resuscitative thoracotomy, REBOA is being increasingly used for polytrauma patients. However, despite the increased use of this procedure, its safety is yet to be thoroughly evaluated. The aim of our study is to review the complications we experienced related to REBOA in our emergency department.

Methods: This study is a retrospective cohort study conducted at a tertiary referral hospital in Japan from October 2014 to September 2016. All trauma patients who required REBOA for the control of bleeding were included. All procedures were performed by an emergency physician and any other combined treatments were managed at the discretion of the attending trauma surgeon. The primary outcome was complications related to REBOA. We collected data from our electronic medical records, including age, sex, site of injury, mechanism of injury, injury severity score (ISS), complications, and outcomes. We described these data as number (%) and median (interquartile range).

Results: Ten patients were included. The median age was 45 (24-83), and 6 patients (60%) were male. The mechanisms of injury were motor vehicle accident 8 (80%), fall 1 (10%), and entrapment by heavy machinery 1 (10%). The sites of injury were pelvic fractures 9 (90%), lower extremity fractures 7 (70%), and traumatic brain injuries 6 (60%). The median ISS was 46 (41-59), and 9 patients (90%) achieved systolic blood pressure > 90 mmHg after REBOA. Complications related to REBOA were observed in 3 patients (30%). Of these, 2 patients experienced a deterioration of traumatic intracranial hemorrhage, and the remaining patient experienced an intraoperative device problem (ruptured balloon during the procedure).

Conclusions: Our study revealed 3 REBOA related complications. Although REBOA may be less invasive, is potentially beneficial, and now it is widely indicated to control hemorrhage, our results suggest that the physician must pay more attention when undertaking REBOA due to its potential complications and safety issues.

P236 Evaluation of trauma patients admitted to ICU with trauma and injury severity score (TRISS)

M Simoes, E Trigo, P Coutinho, J Pimentel

Centro Hospitalar e Universitario de Coimbra, Coimbra, Portugal

Introduction: Trauma is one of the leading causes of mortality and morbidity in ICU. Injury severity and prognosis assessing is a complex process. The Trauma and Injury Severity Score (TRISS), introduced in 1981, was developed combining patient age with an anatomical score, the Injury Severity Score (ISS), and a physiological score, the Revised Trauma Score (RTS). Despite some limitations, it still remains the most used scoring system in trauma patients.

Methods: Retrospective analysis of clinical records of all adult severe trauma patients admitted to a trauma center in a university hospital’s ICU in one year (2015). Gender, age, mechanism of injury, type of injury, length of ICU and hospital stay and mortality were studied. RTS, ISS and TRISS scores (using 1995 and 2009 revisions) were determined for every patient based on clinical records. Statistical analysis of TRISS’s performance as predictor for survival.

Results: In 2015 were admitted to our ICU 124 adult trauma patients, the majority were male (83.1%). Age ranged from 18-91 years, with mean of 52.4 years. Most patients were admitted directly from Emergency Room (82.3%), 40.3% of all the patients came from other hospitals. Regarding mechanism of injury, road traffic collisions were responsible for 46.0% of cases, followed by falls (39.5%). Road traffic collision patients were younger than those with injuries caused by falls (45.72 years and 62,14 years, respectively). Trauma brain injury was the most frequent type of lesion admitted to ICU (67.7%), followed by thoracic injury (45.2%). The average length of ICU stay was 13.8 days. The average hospital length of stay was 32.6 days, although this value was probably higher, as 56 patients were transferred back to smaller hospitals. The hospital mortality rate was 26.6% (25 patients died at the ICU and 8 died after ICU discharge), but no information regarding the outcome of transferred patients is known to the authors, therefore the mortality ratio might have been higher. Mortality ratio was higher in older patients (47.2% in patients aged 65 and over and 18,2% in patients under 65 years). Excluding those transferred patients, TRISS showed a statistical difference between the survivors and non survivors (87.7% vs 78.0%, p = 0.0249).

Conclusions: TRISS was a valuable tool for assessing the clinical outcome, as significant difference was found between survivors and deceased patients. Injury evaluation and outcome prediction is relevant issue in critical care trauma patients.

P237 Management of major bleeding trauma in ICU: preliminary data on the use of thromboelastometry in a tertiary care hospital

A Franci, D Basagni, M Boddi, M Cozzolino, V Anichini, A Cecchi, A Peris

Careggi Teaching Hospital, Florence, Italy

Introduction: Diffusion of viscoelastic methods, combined with the availability of coagulation factors’ concentrates, allowed the introduction of goal-directed treatment’s protocols (GDT), aimed at the correction of specific haemostatic alterations.

The aim of the study was to evaluate if thromboelastometry could determine a better outcome and a reduction in blood products’ consumption and in treatment’s costs in bleeding trauma patients.

Methods: We conducted an observational retrospective study, including patients with an Injury Severity Score (ISS) > =15 and transfused with at least 3 units of Red Blood Cells within the first 24 H. The patients have been divided into 2 groups, according to the hemostatic treatment received: Group A (GDT) and Group B (conventional therapy).

Results: The 2 groups didn’t shown differences in outcome, though the use of blood products and coagulation factors’ concentrates was significantly greater in Group A. This might depend on the trend of using thromboelastometry in the most critical bleedings: Group A had on average a greater ISS and a lower mean arterial pressure and received a more aggressive fluid therapy in prehospital setting. We can also hypothesize a tendency to adopt a traditional approach for the treatment of these patients (high consumption of plasma and low of concentrates). However, it may partially depend on the earlier detection of the coagulopathy by the thromboelastometry.

Conclusions: Introduction of thromboelastometry wasn’t associated with an improvement in outcome and a reduction in blood products’ consumption and costs. Therefore, we decided to broadcast staff training programs and to implement goal-directed treatment’s protocols for bleeding trauma.

Table 13 (abstract P237). See text for description

P238 How inflammatory markers are correlated with the onset of fever in ICU trauma patients; preliminary results

D Markopoulou, K Venetsanou, I Papanikolaou, T Barkouri, D Chroni, I Alamanos

Kat Hospital Athens, Kifisia, Greece

Introduction: The acute multiple trauma state is often followed fever. It’s known that traumatic injuries stimulate an inflammatory reaction and cytokines cascade. The aim of this study is to investigate the correlation of fever with inflammatory markers and circulating endotoxin, in patients admitted in ICU.

Methods: Eighteen trauma patients were admitted in ICU and 18 healthy volunteers enrolled in this study as control group; 20 ml of peripheral blood was sampled from the trauma patients (10 ml during admission and 10 ml during the onset of fever and 10 ml from the control group. Clinical and demographic data were recorded on admission. Serum/plasma samples were isolated with centrifugation and stored at -70 ° C Interleukin-6 (IL-6),Lipopolysacharide binding protein (LBP), Procalcitonin (PCT), C-reactive protein (CRP) measured with ELISA and endotoxin with chromatometric assay.

Results: On admission, IL-6 (P < 0.001), CRP (P < 0.001) and PCT (P < 0.01) release, were significantly higher compared to control group, while LBP and endotoxin had no significant difference. The onset of fever was accompanied by abundant LBP release (P < 0.001), parallel reduction of circulating endotoxin (P < 0.01) and further significant increase of PCT and CRP (P < 0.05), compared to admission values. No significant difference found for IL-6 (P > 0.05).

The onset of fever was positively correlated with LBP (P < 0.001, Spearman coefficient (Spc) 0,866) and PCT (P < 0.05, Spc 0.281) and negatively with LAL (P < 0.01, Spc -0.436), but no significant results were found for IL-6.

Conclusions: LBP, the specific protein binding for circulating endotoxin, is more reliable and patent marker than PCT, CRP and IL-6, as far as fever concerned trauma patients.


1. Saxena, Manoj, et al. " Intensive care medicine 41.5 (2015): 823–832.

2. Markopoulou D et al. Int Care Med Exp 2016; 4 SUPPL: A306.

P239 The ECS algorithm application in trauma patients: a pre-post analysis

E Cingolani1, MG Bocci2, L Pisapia2, A Tersali2, SL Cutuli2, V Fiore3, A Palma1, G Nardi3, M Antonelli2

1Azienda Ospedaliera San Camillo Forlanini, Roma, Italy; 2Fondazione A. Gemelli, Rome, Italy,3Infermi, Rimini, Italy

Introduction: This study, performed by Policlinico Gemelli, aims to evaluate blood components consumption, mortality and morbidity associated with trauma, hemorrhage and blood transfusion pre and post the introduction of the Early Coagulation Support (ECS)[1]. The ECS is an algorithm developed by the Trauma Centers Network (TUN) (Fig. 23), aiming to improve and homogenize the treatment of trauma patients with significant bleeding and at high risk of massive transfusion.

Methods: ECS algorithm was applied to all severely injured patients with a ISS > 15, admitted to the hospital for trauma occurred within six hours, excluding patients who suffered of cardiac arrest prior to admission. A Propensity Score Analysis (PSA) was performed. A prospective Study Period (SP) started from 01 Jan. 2014 to 31 Dec. 2014 and a retrospective Control Period (CP) lasted from 01 Jan 2012 to 31 Dec. 2012.

Results: The PSA was performed on 64 patients, 32 in each group (CP and SP), matched on the estimated probability to receive a massive transfusion. Variables considered for PSA were: age (47 ± 20 vs 51 ± 19), SBP (91 mmHg in both), ISS (41[IQR 29-66] vs 41[IQR 27-57]), pH (7.28 ± 0.1 vs 7.25 ± 0.2), lactates (4.6 ± 2.9 vs 5 ± 2.9 mmol/l), platelets (196 ± 101 vs 200 ± 94 109/l), hemoglobin (10.7 ± 2.5 vs 11.4 ± 2.4 g/dl), fibrinogen (169 ± 78 vs 196 ± 59 mg/dl). Statistically significant reductions were observed in the SP group for blood components consumption, volume of infused crystalloids (Fig. 24). Furthermore, there was a significant reduction of surgical procedures performed in the first 24 hours of hospitalization in the SP but no relevant differences were observed neither for damage control surgery nor for each single type of trauma surgery. The 30-days survival analysis showed no difference between the two groups. It was observed a decrease of morbidity in SP group, although it was not statistically significant.

Conclusions: The ECS must be considered as a part of a comprehensive Damage Control Resuscitation. This algorithm allows to avoid plasma use in patients who need PRBCs massive transfusions, reducing related complications. Furthermore, the early restoration of fibrinogen blood concentration improves coagulation support. This study shows the saving in numbers of blood products and the carrying out of pertinent therapeutic strategies without an increasing of morbidity and mortality.


1. Nardi G, Agostini V, Rondinelli BM, et al. Prevention and treatment of trauma induced coagulopathy (TIC). An intended protocol from the Italian trauma update research group. J Anesthesiol Clin Sci. 2013;2(1):22.

Fig. 23 (abstract P239).

See text for description

Fig. 24 (abstract P239).

See text for description