37th International Symposium on Intensive Care and Emergency Medicine (part 1 of 3)

Introduction Imbalance in cellular energetics has been suggested to be an important mechanism for organ failure in sepsis and septic shock. We hypothesized that such energy imbalance would either be caused by metabolic changes leading to decreased energy production or by increased energy consumption. Thus, we set out to investigate if mitochondrial dysfunction or decreased energy consumption alters cellular metabolism in muscle tissue in experimental sepsis. Methods We submitted anesthetized piglets to sepsis (n = 12) or placebo (n = 4) and monitored them for 3 hours. Plasma lactate and markers of organ failure were measured hourly, as was muscle metabolism by microdialysis. Energy consumption was intervened locally by infusing ouabain through one microdialysis catheter to block major energy expenditure of the cells, by inhibiting the major energy consuming enzyme, N+/K + -ATPase. Similarly, energy production was blocked infusing sodium cyanide (NaCN), in a different region, to block the cytochrome oxidase in muscle tissue mitochondria. Results All animals submitted to sepsis fulfilled sepsis criteria as defined in Sepsis-3, whereas no animals in the placebo group did. Muscle glucose decreased during sepsis independently of N+/K + -ATPase or cytochrome oxidase blockade. Muscle lactate did not increase during sepsis in naïve metabolism. However, during cytochrome oxidase blockade, there was an increase in muscle lactate that was further accentuated during sepsis. Muscle pyruvate did not decrease during sepsis in naïve metabolism. During cytochrome oxidase blockade, there was a decrease in muscle pyruvate, independently of sepsis. Lactate to pyruvate ratio increased during sepsis and was further accentuated during cytochrome oxidase blockade. Muscle glycerol increased during sepsis and decreased slightly without sepsis regardless of N+/K + -ATPase or cytochrome oxidase blocking. There were no significant changes in muscle glutamate or urea during sepsis in absence/presence of N+/K + -ATPase or cytochrome oxidase blockade. Conclusions These results indicate increased metabolism of energy substrates in muscle tissue in experimental sepsis. Our results do not indicate presence of energy depletion or mitochondrial dysfunction in muscle and should similar physiologic situation be present in other tissues, other mechanisms of organ failure must be considered.

Introduction: Aberrant inflammation is a hallmark of acute respiratory distress syndrome (ARDS) pathophysiology. JAK/ STAT3 pathway is critical for macrophages and neutrophils activation and persistent inflammation. This study aims to investigate the therapeutic potential of inhibiting JAK1/3 activity using the small-molecule R548 inhibitor in LPS induce lung injury model. Methods: Lung injury was induced in adult male C57BL/6 mice, by intratracheal LPS administration followed by post subcutaneous injection of R548 inhibitor R548 inhibitor, prodrug for the active compound, R507 (Rigel Pharmaceuticals Inc.). Mice sacrificed at 6 h and 24 h after LPS administration. Lung inflammation was examined by protein content, number and type of inflammatory cells in bronchoalveolar lavage fluid (BALF). Protein expression levels of JAK1, p-STAT3, ERK1/2 were analysed by Western blotting. Results: LPS administration increased BALF cellularity, total protein content, and neutrophils cells number at both 6 h and 24 h. Elevated levels of JAK1, p-STAT3 and ERK1/2 protein expression were observed. In addition, post LPS treatment with the JAK 1/3 inhibitor significantly reduces BALF protein content (P < 0.05), total cells number (P < 0.01), neutrophils cells number (P < 0.01) as early as 6 h. Moreover, R548 treatment decreased JAK1 protein expression by 2 fold (P < 0.01) and p-STAT3 levels by 2.7 fold (P < 0.001) below the LPS group. Conclusions: These data suggest the JAK/STAT3 signaling pathway plays a critical role in ARDS mediated lung inflammation and injury. Additional studies are warranted to further investigate JAK/STAT3 inhibition as a therapeutic treatment for this serious and life threatening disease.
Introduction: In definition of ARDS PaO2 and PaO2/FiO2 (Horowitz) ratio are the most reliable parameters used. In this study it is aimed to investigate the correlation of Horowitz ratio with oxygenation saturation index (OSI) derived from SpO2 instead of PaO2. Methods: Demographical data of 307 patients treated with mechanical ventilation in our ICU between 01.01.2014-31.12.2015 due to respiratory failure were revised from iMDsoft (Metavision) system together with FiO2, SpO2, PaO2, Paw parameters. Oxygenation saturation index was calculated with the formula using [FiO2 x Paw]/ SpO2. ROC (receiver operator characteristic) analysis was used to search for the relation of Horowitz ratio and OSI. Results: The median age of the patients were 58,5 years (IQR 18-96). For the patients with Horowitz ratio < 100 cut-off value for OSI was found to be 7,0768 (sensitivity 94.4% and specificity 86.9%). OSI cut-off value was 3,7178 for Horowitz ratio < 200 (sensitivity 90.2% and specificity 79%). For those with Horowitz ratio < 300, OSI cut-off value was calculated as 2,561 (sensitivity 83.3% and specificity 72%).  Figure 2 The effect of CXCR7 on the migration of mMSCs. The concentration of VCAM-1 and CXCL12 in the supernatant of Transwell Chambers was examined using ELISA assay.  Conclusions: OSI was defined for ARDS diagnosis in pediatric population which includes the parameter of mean airway pressure in its formula rendering this index a more objective parameter for ventilatory support instead of PaO2/FiO2 ratio. It was a big controversy before Berlin definition that ARDS definition was relying on cirteria which do not directly consider the extent of ventiatory support on oxygenation, however there is still a debate that Berlin definition did not bring a solution in that aspect. Therefore, we think that this study, which we aim to reveal the OSI values that correlate with Horowitz values of mild, moderate and severe ARDS, should be supplemented with other studies and validated by further prospective studies, considering OSI as a reflection of positive pressure changes affecting oxygenation in ARDS.
Introduction: Tidal volume, pressure and flow are components of the energy load (EL). We investigated the relationship between intra-tidal lung inflation status variation and EL in mechanically ventilated ARDS patients.
Methods: Twenty-eight ARDS patients underwent end-inspiratory and end-expiratory low dose CT scans at PEEP 5 and 15 cmH2O, maintaining the same tidal volume 7 ± 1.8 mL/kg and respiratory rate 15 ± 4.2 breaths/min. Quantitative CT scan analysis was performed to obtain the amount of not, poorly, well and over inflated tissue (g). EL was computed as the area between the inspiratory limb of pressure-volume curve and the y axis, summed to the energy needed to inflate the PEEP volume. EL at PEEP 5 and 15 cmH2O was then normalized by the End-Expiratory Lung Volume at 5 cmH2O (EELV5).
Conclusions: Higher PEEP is associated to higher EL. Higher EL, increasing lung stress and intra-tidal opening and closing, could be unprotective. As shown, Pes is greater than either the total and the lung SP at the level of the esophageal balloon. Conclusions: If the increased total or lung SP is the main mechanism for lung collapse in ARDS patients and PEEP keeps the lung open mainly by counter-balancing the SP, Pes could not be used to select the best PEEP.
Introduction: The transforming growth factor β (TGF-β) pathway is activated in experimental ARDS and is associated with increased pulmonary edema (1) because it reduces transepithelial sodium transport by inducing endocytosis of ENaC (the epithelial sodium channel). In this way TGF-β decreases the osmotic gradient for re-absorption of water from the alveolar and interstitial spaces (2). We hypothesized that blocking TGF-β signaling with Galunisertib (LY2157299 monohydrate, a TGFβ R1 kinase inhibitor) would reduce edema and tested this in a porcine model of ARDS using aerosolized Galunisertib. Methods: Five piglets (25-30 kg) were randomized to receive (n = 3) or not (n = 2) Galunisertib 50 μM by aerosol. The animals were anesthetized, tracheotomized and ventilated in volume controlled mode (VT: 8 ml/kg, FiO2: 70%, RR:30, PEEP:5 cmH2O, I:E = 1:2). Cardiac output and extravascular lung water (EVLW) were continuously  measured (PiCCO, Pulsion Medical System). We measured pressure/volume curves and arterial pH, paO2, paCO2, HCO3-at baseline, immediately after induction of ARDS and then hourly for 6 h. ARDS was induced by repeated lung lavages and injurious ventilation. After the establishment of the ARDS model, Galunisertib was administered and the experiment was continued for 6 h, after which the animals were sacrificed. Histological samples from lungs, heart, kidney and liver were taken. Activation of the TGF-β pathway was immunohistochemically evaluated using an antibody against phosphorylated Smad2 (Small Mother Against Decapentaplegic) and activation of mechanical signaling in the lungs was evaluated using an antibody against phosphorylated FAK (Focal Adhesion Kinase). Moreover, we measured the wet/dry weight ratio of the lung samples.
Results: Two animals did not survive the experiment. The sample size is too small to draw any strong conclusion. However, we could not find any major difference in EVLW, wet/dry ratio, blood gas parameters or in lung mechanics between the two groups during the tested time period. However, the lung samples from the treated group showed a lower degree of TGF-β pathway activation, while the control samples from liver or heart showed no influence by Galunisertib.
Conclusions: We cannot prove or disprove that TGF-β blockade reduces pulmonary edema associated with experimental ARDS. However, the TGF-β blocker seemed to reach its target as indicated by reduced activation of the TGF-β pathway in the lungs, suggesting that aerosol administration is feasible.
Introduction: The low efficiency of homing to injured organ in mesenchymal stem cells (MSCs) has become the bottleneck for the treatment of acute respiratory distress syndrome (ARDS). In a previous in vitro study, the results showed that the CXCL12/ CXCR7 axis promoted the migration of MSCs. It is hypothesized that MSCs overexpressing CXCR7 could further benefits LPSinduced ARDS mice. Methods: mMSCs were transfected with CXCR7 by a lentiviral vector and were transplanted into mice following LPS-induced intratracheal lung injury. Histopathology with haematoxylin and eosin staining and lung injury scoring was performed to evaluate lung tissue injury.  Homing of mMSCs were assayed by NIR815 fluorescence imaging and immunofluorescent staining.
Results: The administration of MSC, MSC-GFP and MSC-CXCR7 all improved both lung histopathology (Fig. 9) and lung injury score. Moreover, the MSC-CXCR7 showed lower lung/body weight and lung injury score especially at 24 h compared with MSC-GFP. NIR fluorescence imaging and immunofluorescent staining showed that overexpressing CXCR7 significantly increased the accumulation of mMSCs in the lung especially at 24 h compared to MSC (Fig. 10).
Conclusion: It is demonstrated that overexpressing CXCR7 in mMSCs could further promote their homing to injured lung and improve the pathological damage in ARDS mice. Results: Total 8 patients underwent ILV during observed period (0,9% of all ventilated patients). All patients had paO2/FiO2 < 200, all were ventilated in pressure modes (P-AC, BiLEVEL), ventilators were not synchronized -PEEP, RR, FiO2 and Vt were set individually for each ventilator following blood gas results. Recruitment manuvers and NO were used at physician discretion. In 5 patients oxygenation improved significantly within 2 hours of ILV. 2 patients were upgraded to V-V ECMO (Wegener's, pneumonia), in 4 contraindication to ECMO was present. 3 patients died (lung carcinoma, emphysema, Wegener's). All tube placements were confirmed radiologically, 1 intubation with double lumen tube was performed via bougie and proper placement confirmed bronhcoscopically. In all patients left sided double lumen tubes were used only. In one patient carinal decubitus was observed due to double lumen tube's hook. Patients underwent ILV for median 4 days (range 2-10 days).
Conclusions: In severe single sided lung injury necessitating MV, ILV should be considered as it represents valuable adjunct in treating refractory hypoxemia. With lung separation both lungs can be ventilated according to protective ventilation strategy. Special care is needed to prevent tube dislocation, cuffs overinflation and trauma to larynx, trachea and bronchi. Since tubes' lumens are narrower special attention to possible tube blockage must be paid. Daily reevaluation and conventional tube placement is warranted as soon as possible.   Conclusions: In a significant percentage of patients with ARDS, targeted pH value can be achieved by using the lower threshold of recommended initial TV value. In patients with severe hyperlactatemic acidosis, higher initial TV values may be required. Introduction: Lung protective ventilation may improve patient outcome on ICU's and clear guidelines are available for ventilator settings. It is unknown whether nurses and doctors adhere to these guidelines. The purpose of this study was to determine the effect of predefined lung-protective ventilation regime in pressure support ventilation (PS) introduced via an educational programme on tidal volumes and airway occlusion pressure.

Methods:
The study was a case control single-center study. For the intervention group, an education programme was integrated consisting of eight clinical lessons of ±30 min, bed side teaching, four times written instruction by email and twice a dissemination of an educational presentation to the entire ICU team, in the period from May to June 2016. This program set guidelines for mechanical ventilation, with the specific instructions to ventilate patients with a tidal volume of 6-8 ml/kg/ideal body weight (IBW) and airway occlusion pressure (P0.1) of 2.8 to 6.0 cmH2O. Prior to ventilation instruction a control group was obtained from the general ICU population. Inclusion criteria were: mechanical ventilation (PS) within 7 days after presentation on the ICU, exclusion criteria were: any neurological diagnosis, re-intubation and delirium. Results: In total 14 patients were included in the intervention group versus 17 in the control group (baseline, Table 3). Thirteen parameters were monitored during the study, primary endpoints where tidal volume ml/kg/IBW and the P0.1. In the intervention group (Table 3) finale tidal volumes were 1.2 ml/kg/IBW lower (p = 0.04). The P0.1 shows no difference between the two groups (p =0.75). Conclusions: Conclusion: Adherence to lung-protective ventilation is increased with education and improves ventilation practice on ICU's. An educational program on specific guidelines of mechanical ventilation decreases tidal volumes to optimal values between 6-8-ml/kg/ IBW. P0.1 values could not be optimized by this regimen. It seems likely that higher pressure support levels lead to lower inspiratory work of breathing. Introduction: Intra-abdominal hypertension (IAH) is associated with increased morbidity and mortality. IAH reduces lung volumes and oxygenation and increases airway pressures. The optimal level of positive end-expiratory pressure (PEEP) to be applied in such patients remains unknown. Animal data suggests that high PEEP adjusted for intra-abdominal pressure (IAP) may counteract IAH-induced lung volume and arterial oxygen level reductions. In this pilot trial, our primary aim was to assess whether high and PEEP, adjusted for IAP, can be safely applied in patients with IAH requiring mechanical ventilation. Our secondary aim was to assess the effect of such PEEP levels on oxygenation.
Methods: Prior to enrolment, the patients next of kin were asked for written informed consent. Patients were excluded if deemed unsuitable or if predefined severe cardiovascular dysfunction or severe hypoxemia were present. Ventilation was standardized. Following a recruitment manoeuvre, the following PEEP levels in cmH2O were randomly applied: 5 (baseline), PEEP = 50% of IAP, and PEEP = 100% of IAP. At each measurement step, we allowed 30 min for equilibration before measuring arterial blood gases and respiratory parameters. For the safety of our patients, predefined stopping criteria were applied.
Results: 15 patients were enrolled. The protocol was stopped in one patient (excluded from analysis) due to an unexpected hypertensive episode (drug error unrelated to protocol). Three patients did not tolerate PEEP = 100% IAP due to hypoxemia, hypotension, or ETT cuff leak in one patient each. PaO2/FiO2 ratios (maen and SD) were 234 (68), 271 (99), and 329 (107) at each PEEP level respectively. The difference was significant (p = 0.009) only between baseline and PEEP = 100% IAP. Conclusions: PEEP = 100% was commonly not tolerated and only marginally improved oxygenation in ventilated patients with IAH. Such high PEEP pressures can therefore not be recommended for routine clinical practice. Legend Table 2 Baseline, Mean ± SD (%) Introduction: Neurally adjusted ventilatory assist (NAVA) is a complex intervention involving diaphragmatic monitoring and a proportional, neurally triggered ventilation mode [1]. Clinical effectiveness has not been demonstrated and feasibility data are required prior to larger definitive randomised controlled trials (RCTs). Although there is no agreement on minimum levels in clinical trials, low protocol compliance will impact on statistical power and interpretation of results in a definitive study [2]. The aim of this trial is to assess feasibility by measuring protocol compliance in the setting of a pragmatic RCT.
Methods: A feasibility RCT in 76 ventilated adult ICU patients, currently being conducted in a central London hospital. Patients at high risk of prolonged ventilation are randomly assigned to NAVA or Pressure Support (PSV) in the weaning phase. Feasibility end-points include protocol compliance (time in the NAVA mode), recruitment rate and acceptability. Secondary outcomes include ventilator free days and sedation load. The data from the first 48 patients (n = 25 and 23 in the NAVA and PSV groups respectively) are presented. Results: At 14/11/2016 480 patients had been screened and 55 recruited with a mean recruitment rate of 1.8 patients per month. 71 patients were approached for inclusion with 16 patients (22.5%) excluded due to personal or professional consultee refusal. 68.2% of patients in the intervention group were compliant [CI 47.3-83.6] with a median adherence to the NAVA mode of 87.5% [38. 9,99]. Ventilator free days to day 28 were 18 [8,24] days in the NAVA group vs. 19 [0,23] days in the PSV group. There were no adverse events related to the intervention. Conclusions: These data suggest an acceptable protocol compliance of > 65%. Further analysis of the reasons for non-compliance will be conducted; initial data suggest that a mix of technological, clinical and human factors underlie reduced adherence and protocol compliance. This is the first trial to compare NAVA to PSV in prolonged ventilation in the context of a pragmatic RCT. The results will provide guidance for the design of future trials.

P20
Effects of prone position and recruitment manoeuvres on gas exchange and regional respiratory mechanics G. Grasselli 1  Introduction: Prone position (PP) improves survival and delays the risk of ventilation-induced lung injury in severe ARDS [1]. PP may exploit its benefits in the presence of optimized alveolar recruitment. Thus, we performed prone positioning in combination with recruitment manoeuvres (RMs) to assess the effects on aeration and oxygenation. Methods: Eleven intubated, volume-controlled, ARDS patients (1 severe, 8 moderate, 2 mild) were studied in supine and PP before and after RMs (40 cmH2O airway pressure for 10 breaths). In each phase electrical impedance tomography (EIT) was recorded for 15 minutes to assess global and regional (dependent and non-dependent) respiratory system compliance (Crs), tidal volume (VT) distribution, end-expiratory lung volume changes (ΔEELV). Gas exchange and respiratory mechanics were monitored at the end of each phase. sub] improved after RMs (p = 0.054) and by interaction between PP and RMs (p = 0.046). ΔEELV significantly increased after RM performed in supine (p = 0.032; 40 ± 53 ml vs. pre-RM), in PP (p = 0.002; 95 ± 78 ml vs. pre-RM) but by similar magnitude (p = 0.089) post-RM supine vs. prone. PP didn't modify Crs, but dependent Crs significantly improved (p = 0.006) and non-dependent Crs (p = 0.009) decreased. VT distribution was more homogenous during PP and lung homogeneity index improved, albeit non-significantly. Neither regional Crs nor homogeneity were affected by RMs.
Conclusions: PP improves regional mechanics and homogeneity; RM performed after proning seems efficacious and might increase PP protective effects.
Introduction: We recently completed a large randomized clinical trial in critically ill patients to test the LTP vs. SRP in the prevention of ventilator-associated pneumonia (VAP). Here we focus on the feasibility of the study interventions. Methods: We randomized 194 and 201 critically ill patients into LTP and SRP group, respectively. Patients in LTP were placed in semilateral (60°) -Trendelenburg position and turned from one side to the other every 6 hours. LTP was encouraged specifically during the first days of mechanical ventilation, in compliance with the patient's wish. In the SRP group, patients were kept with the head of the bed elevated > = 30°. We assessed interventions feasibility, and nursing feasibility/workload Results: Patients in the LTP and SRP group were kept in the randomized position for 38% and 90% of the study time, respectively (p = 0.001). Median fraction of time in LTP reached up to 51.8% (interquartile range 20.7-79.2) during the first 2 days; yet, it progressively decreased in subsequent days and was inversely related to the level of sedation. Median lorazepam equivalents was 46.7 mg/day (interquartile range, 20.8-120.0) and 58.0 (28.0-124.2) in the SRP and LTP groups, respectively (p = 0.37); propofol dose was 1393 mg/day (627-2345) and 1661 (826-2700) (p = 0.13). Finally, morphine equivalents were 112 mg/day (30-217) in the SRP and 119 (31-234) in the LTP (p = 0.56). Nurses reported greater difficulties in positioning the patient in LTP and higher workload (p < 0.001) vs SRP. However, approximately 50% of the LTP patients were easily or very easily positioned. Finally, these nursing challenges slightly ameliorated through practice, as more patients were enrolled in each center.
Conclusions: The LTP was specifically applied during the first days of MV. Of note, higher level of sedation/analgesia was not needed to keep the patient in LTP. At times, the nurses encountered difficulties in executing the LTP.  Introduction: The semi-recumbent position (SRP) could increase risk of gravity-driven pulmonary aspiration and ventilator-associated pneumonia (VAP) (1)(2). We investigated whether the lateral-Trendelenburg position (LTP) vs. the SRP would prevent microbiologically confirmed VAP. Methods: 194 patients were randomized into the LTP group and 201 in SRP, and analyzed in an intention to treat approach. Patients in LTP were turned from one side to the other every 6 hours. Whereas, in the SRP group, patients were kept with the head of the bed elevated > = 30°. Primary outcome was VAP incidence rate, based on quantitative bronchoalveolar lavage fluid culture. Secondary outcomes were duration of mechanical ventilation, intensive care unit (ICU) and hospital stays, and ICU/hospital/28-day mortality. Results: The data safety monitoring board recommended stopping the study at the second interim analysis for low incidence of VAP in the control group, lack of benefits in any major secondary outcome and adverse events in the LTP group. Microbiologically confirmed VAP was 0.5% (1/194 patients) in patients positioned in LTP, and 4.0% (8/201 patients) in patients in SRP, risk ratio (RR) between groups 0.13, 95% confidence interval (CI) 0.02-1.03, p = 0.04. Microbiologically confirmed VAP per 1000 ventilator days was 7.19, 95%CI 3.60-14.37 and 0.88, 95%CI 0.12-6.25 in the LTP and SRP, respectively, RR 0.12, 95%CI 0.01-0.91, p = 0.02. Competing risk analysis, which accounted for the concomitant risk of death and discontinuation of MV on VAP, confirmed lower cumulative probability of VAP in the LTP (cause-specific hazard ratio 0.13, 95% CI 0.02-1.00, p = 0.05). No statistically significant between-group differences were observed in secondary outcomes. Conclusions: Critically ill patients positioned in the LTP had a statistically significant reduction in the incidence of VAP, in comparison with the SRP. Yet, further clinical examinations are mandatory to corroborate our findings, specifically in populations at high risk of VAP.     was most closely related to tidal volume, SAPS II score, and minute ventilation (Fig. 15). Conclusions: We present an approach to detect spontaneous breathing during mechanical ventilation. Our models may be useful to identify spontaneously breathing patient subsets in cohort studies, but predictive accuracy is insufficient to replace direct monitoring of individual patients.  Table 6. Cox proportional Hazard model was used to assess the risk of PMV.

References
Results: Of the preoperative risk factors (listed in Table 6), the indication pulmonary hypertension (PHT) was a statistical significant predictor, compared with COPD (P 0.04) and other pathology (P 0,0079) but not compared to pulmonary fibrosis (P 0,0927) when corrected for age, APACHE II >20, BMI < 20, dexmedetomidine or conventional sedation. After transplantation, the only significant predictor of PMV (P 0,0238) was extracorporeal membrane oxygenation (ECMO), corrected for age, APACHE II > 20, BMI <20, Dexdor or conventional sedation and reperfusion edema.
Limitations of this study are the retrospective and single-center design, and the low sample size. Conclusions: PMV after ssLTx is mainly explained by the indication for lung transplantation (PHT) and the postoperative need for ECMO. Introduction: Treatment with extracorporeal membrane oxygenation (ECMO) and/or hypoxaemia in patients with acute respiratory failure (ARF) has been suggested to cause short and long-term cognitive impairment. To explore the evidence for this, we performed a systematic review. Methods: We searched the databases Medline, PsycInfo, Cochrane Library, and Embase to identify publications, i.e. randomised controlled trials, nested case-control studies and cohort studies, reporting the possible association between hypoxaemia and cognitive impairment in patients treated for ARF with or without ECMO. Results: We identified 2606 citations. After eliminating duplicates, two reviewers screened 1825 publications for eligibility and read 30 full text papers. No study fulfilled the inclusion criteria. We identified six studies, mainly case series, which dealt with the study question. One case series reported a correlation between hypoxaemia (haemoglobin oxygen saturation < 90%) and cognitive dysfunction at discharge but not at 1 and 2-year follow-up. In another study which concerned cognitive impairment after ECMO treatment for all causes, no direct correlation was found between ECMO and cognitive impairment. Cognitive impairment was more common after veno-arterial ECMO, which is rarely used in ARF. Mikkelsen et al. reported that a lower PaO2 was significantly associated with cognitive impairment at 12 months follow-up.

Conclusions:
The evidence is sparse that that ECMO treatment for ARF or hypoxaemia during the course of ARF leads to long-term cognitive impairment. Therefore, any estimation of the actual risk of cognitive impairment as a result of using ECMO or employing hypoxaemia, permissively or inadvertently, in ARF cannot be made. More, high quality studies are needed to explore these clinically important questions.

P30
Patterns of morpho-functional pulmonary recovery after total structural consolidation G. Introduction: Extracorporeal membrane oxygenation (ECMO) can be used in patients with ARDS and life-threatening hypoxia, also to reduce VILI. We described 4 patients requiring ECMO treatment with massive lung consolidation and impaired lung ventilation. Methods: Patient 1, a 63 years old woman, was admitted for ARDS in S. pneumoniae pneumonia with massive bilateral consolidations requiring ECMO treatment; for 4 days she was minimally ventilated (2 mL/kg); after ECMO weaning, protective ventilation was applied for other 5 days. After 16 days she was discharged; CT scan showed residual minimal consolidation of the right lung. She had a good recovery after 6 months. Patient 2, a 50 years old woman, was admitted from another ICU for ARDS in bilateral P. aeuriginosa pneumonia. After 7 days of ECMO and 5 more days of protective ventilation, minimal bilateral consolidation with cystic areas were seen at CT scans; at discharge, after 18 days, she had a good recovery of respiratory function. Patient 3, a 49 years old woman, was admitted for A(H1N1) influenza virus pneumonia evolved in ARDS. Massive lung bleeding and complete pulmonary consolidation occurred; ECMO and minimal ventilation lasted 66 days, mechanical ventilation continued for 10 days after ECMO weaning. She was discharged after 81 days spontaneously breathing, with a reduction of bilateral consolidation at CT. She had a good recovery after 6 months. Patient 4, a 54 years old woman, arrived from another ICU for severe ARDS due to viral pneumonia and massive bilateral consolidations. ECMO treatment was started and for 30 days she was not ventilated for massive bilateral consolidations; weaning from ECMO was not possible for bacterial infection and evolution in pulmonary fibrosis. The patient died in ICU for multi organ failure after 300 days. Results: Massive pulmonary consolidation is not a terminal event in ECMO patients. Pulmonary consolidation can be related to an infection or may depend on massive intra-alveolar bleeding. In our series of 4 ECMO patients, pulmonary consolidation were caused in one case by bleeding, in the others by infectious diseases. Conclusions: We experienced that massive pulmonary consolidation can resolve, provided that patients is kept alive and VILI is minimized by ECMO support. In our series, outcome and long term sequelae seems mainly related to the underling disease, rather than to the extension of pulmonary consolidation. Periods of limited or absent ventilation do not seem to prevent recovery of alveolar function. Bleeding does not seem to exclude a full lung recovery, while infections may cause abscesses and pulmonary disruption with possible long term sequelae and residual lung dysfunction. The authors confirm that written consent for publication had been obtained. Introduction: Extracorporeal membrane oxygenation (ECMO) may be a lifesaving procedure in patients with severe respiratory and/or circulatory failure (1,2). We present the first 19 adult patients treated with ECMO at medical intensive care unit (MICU) of Phramongkutklao Hospital, a Thai tertiary referral hospital. Methods: A descriptive observational study was performed in 19 patients who received ECMO as a rescue therapy. Initial patient's characteristics, technical aspects, and ECMO-related complications were described. Results: Our adult ECMO program was started in August 2014. Since then we supported seventeen respiratory failure patients on venovenous (VV) ECMO, one electrical shock patient on veno-arterial (VA) ECMO, and one cardiac arrest patient on VA ECMO-assisted extracorporeal cardiopulmonary resuscitation (E-CPR). Ultrasound-guided femero-jugular cannulation was percutaneously done, by intensivist, in all VV ECMO patients, while femoro-femoral cannulation was done by cardiothoracic surgeon using open technique in both cases of VA ECMO. The most frequent indication was severe pneumonia with acute respiratory distress syndrome (n = 10). Mean duration on mechanical ventilation before ECMO was 6 days. Before ECMO was initiated, mean Murray score was 3.4, respiratory compliance was 18 mL/cmH2O, PaO2/FiO2 was 102.5, and blood lactate was 5.7 mmol/L. Mean duration of ECMO was 10 days, average length of MICU and hospital stay were 18 and 21 days, respectively. Eight out of nineteen (42%) were decannulated after improvements, while remaining eleven cases who developed multiorgan failure resulting in death while on ECMO. ECMO-related complications were found in eleven cases (58%), which the most complication was active bleeding at the cannulation site. Conclusions: Team organization and meticulous care during ECMO are crucial for minimizing ECMO-related complication. Even number of patients treated with ECMO in Thailand is increasing, a higher ECMO case volume may be required to increase the experience and improve quality of care. Introduction: Deep vein thrombosis (DVT) continues to be a serious complication in (intensive care unit) ICU patients despite the extensive use of pharmacological and mechanical prophylaxis. During extracorporeal membrane oxygenation (ECMO) treatment, the risk of cannula-related DVT is markedly increased, because the large calipter of cannulas causes endothelium damage and significantly decreases venous flow around the cannula. In this perspective it is possible to assume that during ECMO treatment cannula-related DVT could represent an additional risk of mortality and morbidity for ICU patients. We studied the incidence of cannula-related and not related DVTs in ECMO patients and their potential impact on the mortality and length of stay in this high risk subgroup of ICU patients. Methods: This is a retrospective study on 116 patients, 53 + 19 years old, with ARDS unresponsive to conventional treatments, who fulfilled the criteria for ECMO positioning set by the the Italian Ministry of Health. ECMO configuration included 78 (67.2%) patients with bicaval dual lumen mono-cannulation (J-ECMO), 21 (18.1%) with jugular-femoral cannulation (JF-ECMO) and 17 cases (14.6%) with femoro-femoral cannulation (FF-ECMO). Ultrasound exams for DVT diagnosis at the jugular-subclavian-axillary and the femoral-popliteal and sottopopliteal venous axes were performed within the first 48 hours after ECMO placement, once a week during ECMO treatment and within 48 hours after cannulas removal. During ECMO treatment the heparin infusion was adjusted to maintain aPTT target between 50 and 80 seconds. Results: We diagnosed 45 cannula-related DVT: 39 DVTs were located at the jugular vein and 8 at femoral veins. We found 36 (56.25%) not cannula-related DVTs; 9 cases were central venous catheter (CVC)-related DVTs. No DVT was symptomatic and all were diagnosed by scheduled ultrasound sourvellaince; asymptomatic pulmonary embolism was diagnosed in 5 cases, 3 of which in not cannula-related DVT. In our study, the overall incidence of cannula-related DVT was independent of vascular configuration of ECMO (J; FF; JF). The diagnosis of cannula related DVT was not associated with a higher ICU mortality or longer ICU length of stay; on the contrary Not cannula-related DVT significantly increased ICU mortality and ICU length of stay (p < 0.005).

P31
Conclusions: In our ECMO population only DVTs occurring in veins not involved in cannulation were significantly associated with increased mortality and ICU length of stay. These data strongly suggest that different pathophysiologic mechanisms are involved in the occurrence of cannula-related and not related DVT.

P33
Survival of septic patients with refractory ARDS requiring venovenous ECMO C. Lazzeri, G. Cianchi, M. Bonizzoli, G. Di Lascio, M. Cozzolino, A. Peris Careggi Teaching Hospital, Florence, Italy Critical Care 2017, 21(Suppl 1):P33 Introduction: The role of extracorporeal membrane oxygenation (ECMO) remains controversial in adult patients with septic shock and data are so far scare and not univocal. We investigated the outcome of 26 patients with refractory ARDS and septic shock, consecutively admitted to our Intensive Care Unit (which is an ECMO referral center) from January 2013 to December 2015 and treated with venous-venous ECMO (VV-ECMO). Methods: According to our center protocol, all patients were submitted to an echocardiography (transthoracic and/or transesofageal) before VV-ECMO implantation. Mortality during ICU stay was the primary outcome. Results: During the study period, 74 patients with refractory ARDS were consecutively treated with VV-ECMO at our center, among whom 16 patients had concomitant septic shock (35.1%). When compared to patients without septic shock, septic patients showed higher values of SOFA (sepsis: 13.1 ± 3 vs no sepsis: 10.2 ± 2, p < 0.001), lactate values (sepsis: 6.12 ± 5.90 vs no sepsis: 3.09 ± 5.65, p = 0.07) and procalcitonin (sepsis: 76.85 ± 95.37 vs no sepsis: 12.05 ± 30.15, p < 0.01). At echocardiography, no differences was observed between the two subgroups in the incidence of left ventricular dysfunction (chi-square: p = 0.657), right ventricle dilation (chi-square: p = 0.347) and right ventricle dysfunction (chi-square: p = 0.688). The overall mortality rate was 44% (33/74) but it did not differ between patients with and without septic shock (chi-square: p = 0.218). Conclusions: In patients with refractory ARDS requiring VV ECMO, septic shock is quite common. The use of VV-ECMO in these patients is associated with a comparable survival rate when compared to nonseptic patients. According to our data, VV ECMO is not contraindicated in patients with refractory ARDS and septic shock. Introduction: The H1N1 strain of Influenza A is associated with a number of complications including acute respiratory distress syndrome (ARDS). Invasive mechanical ventilation (IMV) does not always provide an optimal solution. Extracorporeal membrane oxygenation (ECMO) has been shown to be a rescue treatment of severe H1N1-induced ARDS. Methods: Here we describe first series of successful application of ECMO in these settings in Belarus. Results: From January to April 2016 three patients with positive serodiagnostics of H1N1 requiring IMV for severe ARDS were admitted in different urban hospitals. Despite ARDS-Net recommended approach to IMV, it was not possible to achieve PaO2 > 60 mmHg, SpO2 > 54%. Ad-hoc veno-veinous ECMO with femoral and internal jugular cannulation was initiated shortly. IMV parameters were adjusted as follow: tidal volume reduced to 3-4 ml/kg PBW; PEEP 12-16 cmH2O, FiO2 40-50%. Patients were further transferred to the ICU of our tertiary care center. No adverse effects of transportation were noted. Median time of transfer was less than 60 min. Ultraprotective IMV was applied using APRW mode with Pinsp 20-22 cmH2O, PEEP 12-16 cmH2O, and FiO2 35-45%. Target tidal volume was 3-4 ml/kg PBW, and inspiration/expiration ratio was 1.5:1-2.5:1. Two patients required myoplegia. After 1 to 4 days of IMV an APRW mode was changed to IntelliVent-ASV™. ECMO settings were as following: flow 3-4 l/min (60-80% of minute volume); fresh gas flow 1.5-2.5 (max 9.7) l/min depending on PaCO2; and oxygen fraction 60-100% to achieve SpO2 92-94%. The durations of ECMO were 9, 11 and 42 days. All patients received transcutaneous tracheostomy on 3-5 day of IMV, ventilator-associated pneumonia prevention bundle. We followed standard ECMO weaning procedure protocol after stabilization of respiratory function and achieving control of SIRS. Spontaneous breathing test was successful in two patients on day two after ECMO weaning. One patient with the longest duration of ECMO developed nosocomial septic shock with lung abscess complicated by hemopneumothorax, hemorrhagic shock and further left lung inferior lobectomy. He required additional 10 days of IMV following ECMO weaning, with successful return to natural airway breathing at day 67 after hospitalization. All patients survived hospitalization. Conclusions: We successfully used ECMO as a rescue therapy in three patients with severe H1N1 induced ARDS that allowed weaning of IMV and recovery. Introduction: Severe thrombocytopenia yields a high risk for bleeding thus representing a relative contraindication for anticoagulation and therefore ECMO. We herein report on a series of haematological patients with severe thrombocytopenia undergoing long-term ECMO without anticoagulation. Methods: Retrospective analysis of patient charts undergoing venovenous gas exchange between 2012 and 2016. Data were extracted from charts and the local ECMO registry. Six patients fulfilling the criteria of severe thrombocytopenia < 50 G/L without anticoagulation for more than 3 days were identified. Results: Patients suffered from acute myelogenous leukemia (n = 3), multiple myeloma (1) and acute lymphoblastic leukemia and consecutive stem cell transplantation (n = 2), respectively. ECMO was performed due to ARDS (n = 4) and graft versus host disease involving the lungs (n = 2). All ECMO systems used were heparin coated. Platelet count was 21 G/L (median, range 1-138), ECMO duration was 31 days (6-262), and ECMO was run without any anticoagulation for 17 days . Altogether, three clotting events were seen leading to oxygenator changes. Bleeding was common, leading to one fatal intracerebral bleeding, a minor subarachnoidal bleeding, a temporary gastric bleeding and persistent haematuria. Altogether, 27 packed red blood cells per patient (median, range 4-280) or 0,78 per day (0.57-1.25) and 22 platelet concentrates per patient (range 7-207) or 0.8 per day (0.57-1.26) were administered. Three of six patients could be weaned from ECMO, one patient survived ICU. Conclusions: In patients with severe thrombocytopenia, ECMO can be run without anticoagulation even for longer periods. Nevertheless, bleeding is common while clotting events seem to be rare. Given the high mortality rate in this population, however, the indication for ECMO should be scrutinized rigorously. Introduction: Patients who develop acute respiratory distress syndrome (ARDS) have high mortality and morbidity rates and survivors could have clinically significant physical and psychological disabilities. Several studies have reported that extracorporeal membran oxygenation(ECMO) may improve survival in severe ARDS but there have been only few studies evaluating long-term outcomes in ECMO-treated ARDS survivors (1). The purpose of the study was to assess intensive care unit (ICU) outcomes and long term outcome and quality of life of patients with ARDS receiving ECMO for refractory hypoxemia. Methods: We conducted a retrospective observational study in adult ARDS patients who had veno-venous-ECMO in a tertiary centre from January 2014 until April 2016. We collected demographic factors, Simplified acute physiology score (SAPS II) at admission, diagnosis, mechanical ventilation and ECMO duration, ICU outcome. We contacted telephonically all ICU-survivors and organized, when possible, a follow-up visits on average 6 months after discharge from ICU. The primary outcome variable was ICU survival and health related quality of life (HRQoL) measured with the Short-Form 36 (SF-36). Long-term quality of life assessment were assessed and compared to normative individual age-matched Italian population. Results: Sixty-four patients (mean age 50.4 ± 14.7 years, 62% males) were studied; 44 patients (69%) were retrieved from external intensive care units (ICUs) by a dedicated ECMO retrieval team. SAPS II at admission was 42.0 ± 15.1 (mean ± DS). Infectious disease were the leading causes of ARDS: 46% bacterial and 21% A(H1N1) influenza pneumonias. Mean duration of vv-ECMO support was 18.8 ± 27.5 days (median of 11 days). 38 patients (59%) survived to ICU discharge; 15 were discharged to other hospitals. Of the 38 ICU-survivors, 25 patients (65.7%) had a follow-up visit after 6 months from ICU discharge. 5 patient died within 6 months and 8 patients were lost/ unable to come to visit. HRQoL was evaluated for 21 patients: compared with age-matched controls, our ARDS survivors had significantly lower (p < 0.005) SF-36 physical function and physical role functioning; vitality, bodily pain, general health perceptions, emotional role functioning, social role functioning and mental health were comparable with those of general population. Conclusions: In this ARDS cohort treated with ECMO the ICU-survival and six months after ICU discharge survival rate was similar to other findings (1). Long term survivors had reduced physical health but their psychological domain scores were comparable with those of the general population. Introduction: Our group has recently developed an Advance Organ Support (ADVOS) system based on albumin dialysis to provide intensive care treatment for patients suffering multiple organ failure including liver, kidney and lung impairments. The system has already shown improved survival in two different animal models as well as safety and efficacy to eliminate water and protein-bound toxins in humans with liver failure [1,2]. In the present work, the ability of the ADVOS procedure to eliminate CO2 and stabilize blood pH together with the reduction of bilirubin and urea levels has been determined. Results were compared to a conventional renal dialysis machine (NIKKISO DBB-03). Methods: For this purpose an ex vivo model for respiratory acidosis was developed continuously infusing 110 ml/min CO2 into 5 liters swine blood. In addition, liver and kidney detoxification were simulated supplementing blood with bilirubin (275 mg/dl) and urea (30 mg/dl), respectively. Blood was subjected to hemodialysis in the ADVOS system for 4 hours through two dialyzers (2 × 1.9 m 2 ) using a blood flow (BF) of 400 ml/min and a dialysate pH of 10. The NIKKISO machine was run through a dialyzer (2.5 m 2 ) with a BF of 350 ml/min and a dialysate pH of 8. CO2, pH, bilirubin and urea levels were analyzed pre-and post-dialyzer. Blood was checked for hemolysis at the beginning and the end of the experiments. Results: During the whole hemodialysis time using the ADVOS procedure, an average CO2 removal of 108 ± 4 ml/min was achieved. The ADVOS system was able to maintain pH stable between 7.35 and 7.45 during the experiments, while with the NIKKISO machine pH decreased to 6.60 after one hour of treatment, being thereafter continuously out of the measuring range (hence no further calculations were possible) In the ADVOS system the main fraction was excreted as HCO3-(85%), while 15% was eliminated as dissolved CO2. In addition, post-dialyzer blood pH remained in both systems below 8. Urea was efficiently cleared with both machines (97% removal). Moreover, the ADVOS system reduced bilirubin levels about 3 times as much as conventional hemodialysis (59% vs. 21%). No signs for hemolysis were observed. Conclusions: The ADVOS system, in contrast to normal hemodialysis, was able to efficiently remove CO2, bilirubin and urea while maintaining pH in physiological levels in an ex vivo model for respiratory acidosis simulating additional kidney and liver failure. Introduction: Extracorporeal CO2 removal is used in COPD/moderate ARDS patients to avoid barotrauma and VILI. When renal failure coexists, a combined ECCO2 removal with RRT may be useful to improve CO2 clearance and kidney function in a single treatment. The aim of this study is to evaluate 1-the clinical feasibility of a CO2 removal device integrated in a CRRT platform.2 -The clinical safety of the system. 3 -The changes of main respiratory and metabolic indices Methods: Ten patients have been enrolled into the study. All patients had renal failure, were hypercapnic and mechanically ventilated, but one. The patients were connected to a low -flow CO2 removal device (Prismalung ®) integrated into a conventional RRT platform (Prismaflex) by a two single lumen catheters inserted in jugular and femoral vein or by a bilumen catheter in a femoral vein. Pump blood flow was started at 300 ml/min and progressively increased to 400 ml/ min. RRT was delivered in a CVHDF mode with an effluent flow of 45 ml/kg/hour. Heparin or citrate infusion (three cases) was used to anti coagulate the circuit. ACT and thromboelastography were used to monitoring the coagulation parameters. At basal time (T 0), 12 th (T 1), 24 th (T2) hours, the main respiratory parameters and the renal indices were evaluated. Bleeding complications, hemofilter clottings, catheter malfunctions were recorded. All data are expressed as Mean ± SD… ANOVA TEST one way with Bonferroni correction was used to compare the changes of the parameters in the time. P < 0.05 was considered statistically significant. Results: At Table 7 are reported the main results of this study. All patients, but one, survived to the treatment and 4/10 were weaned from the ventilator at the end of ECCO2R. The duration of the treatments was 60 ± 30 hours. There was a clotting of the circuit and two cases of catheter malfunction. No major bleeding episodes were observed. Conclusions: This study confirms previous study of Prismalung on animals.(1) ECCO2 removal with a membrane oxygenator integrated in a CRRT platform is clinically feasible and devoid of major complications. It allows clearance of CO2, with improvement of respiratory acidosis and control of renal failure. Introduction: The aim of this study was to assess safety and performance of the new ECCO2R device, i.e. Prismaflex® + Prismalung® (Baxter), with citric acid as regional anticoagulant by calcium chelating effect and acidification to displace the dissolved CO2 from HCO3and removing it through a membrane lung (Prismalung®, PL) with a blood flow (Qb) of 250 ml/min. Methods: This study was conducted on 6 male pigs (68 ± 4.5 kg) separated in two groups, 2 pigs in the control group (Group A) and 4 pigs in the intervention group (Group B). Both groups were submitted to Phase I consisting of 4 hours with citrate anticoagulation (3.3 mmol/L) no CO2 removal (circuit bypass). In Phase II, for 8 hours both groups were submitted to ECCO2R with a FiO2 100% sweep gas flow of 15 L/min; Group A was submitted to the same citrate anticoagulation system as in Phase I while Group B was switched to citric acid anticoagulation (2.7 mmol/L) and acidification (2.0 mEq/ min). All pigs were sedated and paralysed and connected to a ventilator (Servoi®, Maquet). The ventilator parameters were fixed mean tidal volume (7.11 ± 0.48 ml/kg) with PEEP 5cmH2O, FiO2 40%, V/min 8.1 ± 1.5 L/min (respiratory rate 18-20). Results: CO2 content pre and post PL and CO2 removal is depicted in Fig. 17. The systemic pH and pCO2 did not differ statistically between the two groups. The systemic ionized calcium was stable in both groups. During the study no major critical events were recorded (death, cardiac arrest, arrhythmias, etc.). PL CO2 removal in intervention group (51.8 ± 7.8 ml/min) was significantly higher than in control group (35.2 ± 6.9 ml/min). Conclusions: Using citric acid as a regional anticoagulant and acidification during ECCO2R is safe and more effective to remove the CO2 comparing standard ECCO2R.  Table 8. Patient comfort level was significantly higher during HFNC at lower flow and temperature (p = 0.015 and p < 0.001 respectively), but interaction was not significant. Peripheral saturation was significantly increased by higher flow rate (p = 0.026). Interestingly, in this convenience sample, higher HFNC support didn't modify the respiratory rate likely indicating clinical stability. Conclusions: Lower flow and temperature grant improved patient comfort. Titration of HFNC setting to obtain best comfort might be relevant to exploit clinical efficacy.

P41
High flow nasal cannula prevent reintubation in post-extubated critically ill patients: a systematic review and meta-analysis T. Thamjamrassri, J. Watcharotayangul, P. Numthavaj, S. Kongsareepong Mahidol university, Bangkok, Thailand Critical Care 2017, 21(Suppl 1):P41 Introduction: Extubation failure and reintubation after planned extubation are associated with high mortality [1]. Multiple factors including impair secretion drainage, hypoxemia or hypercarbia are possible causes of extubation failure [2]. Therefore, giving appropriate oxygen therapy and respiratory care after extubation in critically ill patients should decrease incidence of reintubation. Recently, high flow nasal cannula (HFNC) has been used increasingly in critically ill patients because it has many advantages over conventional oxygen therapy (COT). HFNC provides higher flow rate and more humidity which facilitated secretion drainage. Moreover, PEEP causing from HFNC decreases work of breathing [3]. Previous study shown that HFNC decreased incidence of reintubation in post-extubated critically ill patients [4].
Although there is a meta-analysis compared HFNC with other methods of oxygen therapy, patients who failed extubation criteria were included in the study [5]. Therefore, we conduct a meta-analysis comparing HFNC with other methods of oxygen therapy in preventing reintubation in planned extubated critically ill patients. Introduction: High-flow nasal cannula (HFNC) is a non-invasive respiratory support associated to several physiologic advantages in acute hypoxemic respiratory failure (AHRF) patients [1]. Previous studies showed that these effects depend on set flow. Aim of the study was to identify the flow at which gas exchange, end-expiratory lung volume (ΔEELV), respiratory rate (RR), inspiratory effort and corrected minute ventilation ( The highest PaO2 value in 71% of patients was reached at HFNC 60 l/ min (p = 0.0047). The highest lung volume was obtained for 59% of patients at HFNC 60 l/min (p = 0.0094). The lowest respiratory rate (bpm) in 77% of patients was at HFNC 60 l/min (p = 0.0007) and the lowest value in ΔPes was obtained for 57% of patients at HFNC 60 l/ min (p = 0.0464). On the contrary, the percentage of patients with minimum value of MVcorr was equally distributed among the 3 HFNC flows (p = 0.662). Figure 19 shows flows distribution. Conclusions: Our results show that HFNC 60 l/min might be regarded as the optimal flow for most parameters. However, 30-40% of patients obtained best results at lower flows and with absence of "optimum" flow for corrected minute ventilation.  Introduction: Our goal is to analyze the impact and the results of the application of high flow oxygen therapy (HFNC) in hematological patients with respiratory failure Methods: We performed a retrospective study. All hematological patients who were treated with HFNC between May 2013 and April 2016 were included (n = 36) Results: Average age: 51. 67 ± 15. 77 (23-80); 50% male; mean SOFA: 9. 09 ± 5. 15 (2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19), and APACHE II: 22. 61 ± 8. 86 . The etiology of respiratory failure was hypoxemic in 91. 66% (n = 33) and mixed in 8. 33% (n = 3) of cases. HFNC avoided intubation and therefore mechanical ventilation (MV) in 38. 9% (n = 14) of the patients. 19 patients required MV after HFNC failure and in the other 3 HFNC was used after extubation. (Fig. 20) The mortality rate in the study was 44.4% (n = 16). In the group of patients who avoided MV was 14.28% (2/14). In the ones who required mechanical ventilation after HFNC failure the mortality rate was 68.42% (13/19). 3 Patients required HFNC after extubation, 1/3 died. (Fig. 21) The group of patients, who need mechanical ventilation, had higher mortality. (p <0.0001) However, the mortality rate was higher in this group due to delay of intubation. The average delay in the patients who died (n = 13) was 2.28 days vs. 1.83 days in the survivors group (n = 6) Conclusions: HFNC oxygen therapy appears to be a safety therapeutic option for the treatment of severe respiratory failure in patients with haematological malignancies. Intubation was avoided in a substantial proportion of patients. Also was related a lower mortality rate in this subgroup of patients. However, the delayed of intubation due to HFNC failure was related to higher mortality. Introduction: The study aims to compare endotracheal tube (ETT) and extraglottic devices (EGD) in the out-of-hospital setting, in terms of consequences and outcome. Secondary endpoints were evaluation of parameters that could correlate with survival and neurological outcome and evaluation of the influence of a mechanical ventilation (MV) vs bag valve mask ventilation (BVM) during the transport from the field to the hospital. Methods: This retrospective study examined 404 patients with a outof-hospital airway management between January 2013 and June 2016. 187 of them were transferred to the Intensive Care Unit (154 managed with ETT and 33 with EGD); they were analyzed for demographic, infective, radiologic and neurologic parameters. Patients with ETT were subsequently divided in 2 subgroups, the ones who received a MV during the transport to the hospital and the ones who were manually ventilated. Results: Our results show that there was no correlation between the use of ETT or EGD and mortality rates 48 hours after the admittance to ICU, but there was a significant difference in survival after 7 days. There was no significant difference between ETT and EGD in terms of neurological and pulmonary outcome when patients were dismissed from ICU. The ventilation technique during the transport revealed to be an accurate parameter of survival after 48 hours, as the MV reduced patients' mortality when compared to the BVM. Conclusions: The study confirmed the importance of advanced airway management in the out-of-hospital setting, regardless of the device used. Evidence shows that ETT is recommended whenever possible, and most of all the use of MV because this seems to reduce mortality rate, probably reducing pulmonary damage deriving from the impossible titration of volumes and pressures administered to the lung during BVM.

P48
Arterial oxygenation and hemodynamics during extrathoracic negative pressure in a porcine model of lung injury A. Nestorowicz 1 , J. Glapinski 2 , A. Fijalkowska-Nestorowicz 3   For all MV, MR100® is the BVM is one who offers the lower resistance. This is most likely due to the type of patient valve of the BVMs, because patient valve of AMBUs® is a mushroom valve conception, while patient valve of MR100® is a duckbill valve conception. This last type of valve seems to offer less resistance than mushroom valve. Introduction: The safe level of endotracheal tube (ETT) cuff pressure during anesthesia is between 20 to 30 cm H2O. The over-inflated cuffs may cause different complications, e.g. sore throat, cough, aspiration and trachea injuries, including rupture. The use of cuff pressure manometer was recommended in several studies (1), hovewer it is still not a routine during anesthesia in many hospitals allover the world (2). Often skilled anesthesiologists are sure that they could evaluate normal and high pressure in the cuff by palpating it after intubation.
The aim of this prospestive onservational study was to estimate whether anesthesiologists could evaluate a safe (20-30 cm H2O), high (30-60 cm H2O) and very high cuff pressure (more than 60 cm H2O) by palpation of the external cuff ballon on an ETT. Methods: The study was conducted in two Kyiv city hospitals at the department of anesthesiology and intensive care in 2015-2016. Anesthesiologists were asked to inflate an ETT cuff after intubation using their usual inflation technique and estimate by palpation if the cuff pressure is at the safe level, high level or very high level. After that the actual pressure in the cuff was validated with manometer measurements.
Results: ETT cuff pressure were measured in 146 patients by 24 anesthesiologists. The majority self-evaluate ETT cuff pressure as safe (86%), other as high (14%). The actual pressure in the cuff was safe in 26 (18%) of cases, high in 100 (68%) and very high in 20 (14%) of cases. There were no statistical difference between accuracy of ETT cuff pressure self-evaluation from one anesthesiologist to another.
Conclusions: Findings from this study shows that only in 18% of cases anesthesiologists could evaluate safe level of ETT cuff pressure by palpation of the ballon after intubation. Most of anesthesiologists underestimated the ETT cuff pressure. So, pressure manometer measurement of ETT cuff pressure should be implemented for safe medical practice. Introduction: Despite the use of endotracheal tubes (ETT) with high volume, low pressure cuffs, concerns about tracheal mucosal damage and long term tracheal stenosis following intubation in ICU remain.
Estimates of the incidence of laryngeal injury range from 15-94%. It is generally accepted that pressure exerted by the ETT cuff impairs mucosal blood flow and causes ischaemia. [1] We hypothesised that the lack of agreed 'safe' cuff pressure or relationship between cuff pressure and complications may be because cuff pressure is not a surrogate for mucosal pressure.
Methods: The study was performed using a cadaveric pig trachea. Sheridan endotracheal tubes in sizes 6,7,8 and 9 were used to intubate the trachea. The ETT cuff pressure was measured using a hand held manometer. A strain gauge was inserted under direct vision in a retrograde manor to align with the midpoint of the ETT cuff. The cuff was inflated in 1 ml increments and cuff pressure and applied pressure were recorded. Results: We found no correlation between ETT cuff pressure and applied tracheal mucosa pressure for a size 6 or 7 ETT. For size 8 and 9 ETT the graph of cuff pressure and applied pressure had an inflection point. Below the inflection point applied pressure remained low. Above the inflection point there was a linear relationship between applied pressure and cuff pressure.
Conclusions: We presume that for the smaller ETT, over the volumes tested, there is no circumferential contact between the cuff and trachea. The applied pressure remains independent of the cuff pressure. For size 8 and 9 ETT above the inflection point the cuff becomes in circumferential contact with the tracheal mucosa. The rise in cuff pressure is directly proportional to the applied mucosal pressure.
References Conclusions: Because Accidental decannulation is the most common serious complication associated tracheostomy, it is important to predict probability of tube displacement and use extended-length tube for risky patients. Pretracheal tissue depth difference as postural change measured by ultrasound was correlated well with risk factor of tubal displacement, pretracheal tissue depth and BMI. Pretracheal tissue depth difference as neck posture is helpful to predict risk of tracheostomy tube displacement. Introduction: Tracheostomy has been used traditionally to provide secure airway in patients undergoing radical resections for head and neck cancer. However, recent studies have questioned these practices, emphasizing endotracheal intubations to be equally safe. We planned a study to determine the safety and efficacy of overnight intubation followed by extubation the next morning (ETT group) as compared to tracheostomy (TT group) for postoperative airway management and to audit our airway management practices for major head and neck cancer surgeries. Methods: A prospective observational audit was conducted over one year at Tata Memorial Hospital from August 2015 to July2016. Adult patients undergoing intraoral surgeries were included. Patients undergoing emergency surgery and having a tracheostomy prior to enrolment were excluded. Extent of the disease, type of surgery, demographic details of the patient, anaesthesia details and airway management details were recorded. These patients were followed up in the post-operative period until discharge. Time to extubation was recorded in the ETT group. Time to oral intake, speech, complications until discharge and length of hospital stay were recorded in both groups. Results: We screened 4477 patients and included 720. There were 417 patients in ETT group and 303 patients in TT group. As compared to TT group, ETT group had shorter stay in hospital (7.2 ± 3.7 versus 11.5 ± 7.2 days, p = 0.00), less time to oral intake (5.1 ± 1.6 versus 7.2 ± 2.8 days, p = 0.00), and less time to speech (3.6 ± 1.6 versus 6.1 ± 2.7 days p = 0.00). The overall complications (4.3% versus 22.4%, p = 0.00) and airway related complications (1.68% versus 8.58%, p = 0.00) were lower in the ETT group compared to the TT group.
Conclusions: The airway in intra-oral major head and neck cancer surgery can be managed safely with overnight intubation with faster return to oral feeding and speech. Tracheostomy should be performed only in select patients. Introduction: Treatment of neurocritical patients is complex due to their different types of brain injury, however the realization of early tracheostomy shown to have lower incidence of pneumonia associated with mechanical ventilation (VAP), reduced hospital stay, lower mortality and lower costs hospital. Methods: Prospective observational study corresponding to patients with TCEs admitted to the Intensive Care Unit (ICU) of the General Hospital of Cd. Victoria "Norberto Trevino" over two years (2014-2015). Patients over 18 years were grouped into two groups; "Early" or "late" using as a cutoff 7 days. Tracheostomy We analyzed: age, sex, APACHE II, VM days, incidence of pneumonia associated with VM, days in ICU, ICU mortality. Continuous variables were expressed as mean and standard deviation, categorical variables as absolute value and percentage, test Students were used for independent samples to compare numerical variables and Chi-square test for categorical whether ordinal or nominal, measures of association with OR. SPSS v.23. Results: Early vs late tracheotomy tracheostomy: 62 patients (68% male) with severe head injury, divided into 2 groups they were included. Trach average age <7 days 27.7 ± 5.4 (22-35) vs Trach > 7 days 23.9 ± 10.1 (18-50) p = 0.672, VM days in Trach <7 days 8.5 ± 4.1 (4-13) vs Trach > 7 days 14.0 ± 3.0 (9-21) p <0.05, sedation days Trach <7 days vs 5.0 ± 2.3 Trach > 6.9 ± 3.0 7 days p = 0.21, APACHE in Trach <7-day 16.5 ± 7.4 (10-25) Trach vs > 7-day 15.0 ± 3.2 (10-22) p = 0.53, ICU days in Trach <7 days 10 ± 4.83 (4-14) vs Trach > 7 days 14.9 ± 4.00 (10-26) p <0.04, complications Trach <7-day 27.3% (12n) vs Trach > 7 days 72.7% (32n) p <0.001, mortality Trach <7 days 28.6% (4 N) vs Trach > 7 days 71.4% (10n) p = 0.34. Trach prevalence of VAP with <46% vs 7-day Trach > 7 days 84%. Pneumonias in Trach <7 days OR 0.160 (0.100-0.890). Conclusions: In this study in the Intensive Care Service General Hospital "Norberto Trevino" We conclude that the effectiveness of conducting early days of VM decreases tracheostomy, ICU stay and ventilator associated pneumonia (VAP). Results: There was no significant difference between two groups in terms of age, sex, timing of tracheostomy, duration of ICU stay, mortality and infectious status before the procedure. The incidence rate of stoma infection was significantly lower in PDT group vs ST group (16.7% vs 65.3%, p < 0.001). We performed the logistic regression adjusting for age, sex, the cause of hospitalization, tracheostomy strategy and timing of the procedure. The use of ST was associated with an increase in the risk of stoma infection (OR = 8.83 95% 4.40-17.71). Conclusions: The incidence rate of stoma infection was significantly higher in ST group compared to PDT group in NICU patients.

P60
Occurrence of ventilator associated pneumonia using tracheostomy tubes with subglottic secretion drainage P. Terragni 1  Introduction: Ventilator-Associated Pneumonia (VAP) represents a cause of morbidity and mortality in critically ill patients who require mechanical ventilation. Subglottic secretions above the endotracheal cuff are associated with bacteria colonization of lower respiratory tract, causing VAP. A preventive strategy to avoid subglottic secretions progression is to remove it by drainage with the use of special tracheal tubes effective in preventing both early onset and late onset VAP. [1] The purpose of this study is to measure VAP incidence in tracheostomized patients with suction above the cuff. Methods: Study design: matched cohort study with historical control in 3 academic italian ICUs. Procedures and measurements: upon admission to ICU, patients requiring mechanical ventilation were submitted to tracheostomy with a tracheal tube allowing drainage of subglottic secretions (treatment group) to reduce the incidence of VAP. A control group of tracheostomized patients without the ability of suctioning above the cuff was created applying the propensity score matching technique on a dataset including patients enrolled in the previous ELT study [2]. Primary endpoint: occurrence of post-tracheostomy VAP incidence (as determined by clinical pulmonary infection score) at 28-days from intubation. Secondary endpoints: 28-days mortality after tracheostomy; number of ventilator-free days and ICU-free days at 28-days from intubation; time of decannulation, total number of days of inpatient hospital stay; assessment of SOFA score. Results: 125 patients were enrolled in the treatment group from July 2014 to April 2016; 232 patients without suctioning were selected as a control group. Overall incidence of VAP was 10 patients (8%) in treatment group and 45 patients (19.4%) in the control group (p value = 0.004) with OR: 0.361 and CI (0.175; 0.745). In order to balance the two groups for timing of tracheostomy, gender, age, SAPS and SOFA covariates, a propensity score matching was performed: VAP incidence (120 patients) was 8.3% and 21.7% in treatment and control groups respectively (p value = 0.0408) with OR: 0.329 and CI (0.109; 0.990). Conclusions: Subglottic secretions drainage reduces incidence of VAP in critically ill patients requiring ongoing mechanical ventilation via tracheostomy. Introduction: Spontaneous breathing trial (SBT), a routine procedure during ventilator weaning, entails cardiopulmonary stress for the ventilated patient, which is higher in the patients failing the trial. SBT has been shown to activate an intense inflammatory response, with an increase of cytokines in plasma such as TNF-a, IL-6, IL-1b, and IL-18, a pro-inflammatory cytokine which is being considered as a potential sepsis biomarker. Sepsis has been associated with severe muscle wasting and/or ICU-aquired weakness. Aim of this study was to investigate the effect of sepsis on weaning outcome in critically ill patients. Methods: A total of 55 intubated and mechanically ventilated patients were included in the study (29 with a septic condition, 26 non septic, intubated for other reasons). Patients were assessed during a 30 minute SBT trial SBT. Blood samples were collected at baseline and at the end of the 30 minute trial. IL-18, caspase-3, TLR-2, IL-6, and TNF-á were measured by ELISA and were correlated with markers of inflammation and the weaning outcome. Results: In these preliminary data we included 29 mechanically ventilated patients (53%) with a septic condition and 27 (47%) patients with other than sepsis conditions. Septic patients had a higher percentage of weaning failure compared to non septic ones (38% vs 27%), and this was associated with statistically significantly lower levels of serum albumin (p = 0.026) and CRP (p = 0.05). No statistically significant differences were observed in the cytokine levels between the failure/success groups in septic and non-septic patients, as well as, at the two measurement time points. However, IL-18 levels were higher in septic patients that failed SBT at 30 mins compared to the success group [192 (75-583) vs 147 (12.5-581), median (range)]. In both septic and nonseptic patients that failed SBT the stress during the trial led to an increase of IL-18 levels (although, non statistically significant). Finally, IL-18 was correlated at both time points with serum albumine (p = 0.012 and p = 0.011, respectively, and r = −0.344 and r = −0.347, respectively), PCT (p < 0.001, r = 989), and APACHE II score (p = 0.05, r = 0.383) Conclusions: Although the number of patients per group is small in these preliminary data, our results indicate that SBT failure in septic patients was associated with a higher inflammatory profile of the patients. Among all the measured cytokines, IL-18 seems to reflect the severity of sepsis associated with weaning failure. Introduction: Dysphagia is a common problem in the intensive care unit (ICU), and has been associated with prolonged ICU length of stay and increased risk of pneumonia, reintubation and death. However, no national guidelines on dysphagia prevention, screening and management exist for ICU patients. Therefore, we performed a national survey to learn what strategies are being used in Dutch ICUs. Methods: A survey was developed based on current literature and experts' opinions. It comprised both open and multiple choice questions regarding hospital and intensive care characteristics, perceived prevalence and importance of dysphagia, screening strategies including the use of certain diagnostic tests and specialist consultation, modalities used to prevent complications such as aspiration, and interventions used to improve swallowing function and follow-up. It was sent to all 90 nonpediatric ICUs in The Netherlands with two reminders by telephone and e-mail. Results: 67 of 90 addressed ICUs (74%) replied to our survey. A median relevance score of 4 (IQR 4-5) out of 5 was given to the topic of dysphagia. In 22% of ICUs, patients were always screened for dysphagia after extubation, in 45% of ICUs screening was always performed after tracheotomy. The water swallow test was always part of the work-up in 88% of ICUs. Fiberoptic endoscopic evaluation of swallowing (FEES) was used as the gold standard in 60% of ICUs, versus videofluoroscopic swallowing study (VFSS) in 25%. In 49% of ICUs no standardized active rehabilitation protocol for dysphagia existed. In the remaining 51% swallowing exercises supervised by the speech language pathologist were part of standard rehabilitation, occasionally supplemented by electrical stimulation or sEMG biofeedback training in 6% and 10%, respectively. Conclusions: Despite the known possible consequences of dysphagia and it being considered relevant by the majority of respondents, most Dutch ICUs do not regularly screen for dysphagia after extubation or tracheotomy and almost half do not have a treatment or rehabilitation protocol. The diagnostic tests and therapies used vary between hospitals, including some of unproven validity. A well-defined screening and treatment algorithm based on an evidence-based guideline could provide more standardized approach to this problem, thus improving quality of care. Introduction: Intra-abdominal hypertension can increase airway pressures and impair ventilation. Little is known about the relationship between increasing intra-abdominal volume (IAV) and airway pressures. We asseseds the effect of increasing IAV on airway and intra-abdominal pressures (IAP) in a porcine model Methods: Seven pigs (41.4 +/−8.5 kg) received standardized anesthesia and mech. ventilation. A latex balloon in the abdomen was inflated in one liter steps until IAP exceeded 40 cmH2O. At each step, we measured peak airway pressure (pPAW) and IAP. Results: Raising IAV led to an exponential increase of pPAW and IAP while decreasing both abdominal and respiratory system compliance. pPAW increased by approx. 40% of the increase of IAP.
Conclusions: The exponential nature of IAV on pPAW and IAP implies that small reductions in IAV may lead to significant reductions in airway and abdominal pressures in the presence of high grades of intra-abdominal hypertension. Treatment to reduce IAP may be more effective at high grades of IAH. Introduction: Lung donation occurs in a limited number of potential donors due to the deterioration of lung function after brain death and the strict eligibility criteria. "Protective" ventilator strategy markedly increased lung donation rate [1]. However, to our knowledge no studies have investigated individualization of PEEP and recruitment maneuvers in this setting. Thus, we explored the use of electrical impedance tomography (EIT) in a hypothesisgenerating study. Methods: Single center observational study. Potential lung donors underwent a 10 minutes evaluation by EIT (PulmoVista 500 Dräger Medical) after neurologic determination of death. Potential donors were ventilated with a "protective" protocol. Gas analysis and respiratory system compliance (Crs) were assessed. Two regions of interest were defined: non-dependent lung zones (ROIndep) and dependent lung zones (ROIdep). We measure the Vt distending each region (VtROIn-dep, VtROIdep); the heterogeneity of Vt distribution (VtH) and regional values of compliance (CrsROInon-dep, CrsROIdep Conclusions: EIT monitoring showed that ventilation is preferentially distributed in ventral lung zones when the protective ventilator strategy is employed in potential organ donors. Next step will be to verify whether PEEP titration and recruitment maneuver based on EIT findings contribute to enhance respiratory performance and suitability for lung transplantation. desaturation, or pneumothorax. Different methods have been proposed to identify patients who can benefit from a RM but conclusive recommendations are still lacking. We hypothesized that the shape of a pixel-level pressure-volume curve (PVpx) could predict recruitability defined as Recruited Volume (RV) after a RM at rising PEEP levels.
Methods: 12 ARF patients (P/F < 300 mmHg) were mechanically ventilated with TV of 6-8 ml/kg. A 5-step PEEP trial -increasing (5,10,15 cmH[sub]2[/sub]O) and decreasing (10,5 cmH[sub]2[/sub]O) limbwas performed with a RM (inspiratory hold,40 cmH2O CPAP for 40s) between PEEP10 and PEEP15. At each step, lung mechanics and EIT data were recorded during a quasi-static pressure-volume curve (PV) maneuver. The PV before the RM was adapted to an isogravitational pixel level plotting the variation of impedance in each pixel row and the variation of pressure in the respiratory system ( fig. 28); 19 pixellevel PV curves (PVpx) were obtained [1] and fitted in the equation V = a + b•Pao + c•Pao 2 [2]. The "c" factor derived from the fitting (C) indicates the shape of the curve: positive C is related to a compliance increase during the inflation, while negative C to its reduction. We correlated the RV after the RM with the C before the RM at each pixel level, from non-dependent (pixel 1) to dependent lung (pixel 19). Results: The C had a significant positive correlation with RV (ml/kg/ PBW) for pixel levels 13-17,19 (dependent lung); positive PVpx values predicted recruitability of the dependent lung. In the central lung, no correlation was founded. A negative (non-significant) correlation was founded in non-dependent lung indicating that a preexisting nondependent overinflation could be inversely correlated with the effects of the RM.
Conclusions: The C factor from PVpx predicts lung recruitability at the bedside and could help to identify patients who might benefit from lung RM.  Lung recruitment decreases CL but may induce regional OD. PEEP may be optimal when the balance between CL and OD is optimal. The distribution of this balance and changes in local ventilation distribution at varying PEEP levels throughout different lung fields in ARDS is unknown. Regional OD, CL and centre of ventilation (COV, a measure of homogeneity of ventilation distribution) can be visualized using Electrical Impedance Tomography (EIT). Methods: Ten ARDS patients (P/F-ratio 170, SD 53) were studied. Ventral and dorsal OD and CL (% overdistended or collapsed alveoli) and COV were measured by EIT during an incremental and decremental PEEP trial (8-20 cmH2O). COV was expressed as dorsal to ventral ventilation distribution. A COV of 50% indicates an equal distribution between ventral and dorsal lung fields. COV >50% indicates a shift of ventilation distribution towards ventral. We estimated the balance between overdistended and collapsed alveoli by subtracting CL from OD (ODCL). An ODCL of 0% indicates an optimal balance between CL and OD. Data are expressed as mean ± SEM. 1 or 2-way ANOVA was used as appropriate.
Results: At 8 cmH2O, dorsal ODCL (−30% ±6) was significantly lower than ventral ODCL (−6% ±3) (p < 0.001). During the increment in PEEP, ODCL significantly increased to +20% ±2 dorsally and +49% ±2 ventrally (p < 0.001), the magnitude of this increase was similar in both lung fields (p = 0.6, Fig. 29a). The COV shifted dorsally from 57% ±1 to 50% ±1 (p < 0.001, Fig. 29b). Conclusions: The balance between CL and OD at varying PEEP levels differs between ventral and dorsal lung fields. Lung recruitment reduces CL and increases ventilation in dorsal lung fields and a dorsal shift of the centre of ventilation. However simultaneously ventral OD occurs. When determining optimal PEEP settings, variations in the balance between CL and OD between lung fields should be taken into account. Visualization of local ventilation by EIT appears to be a useful tool for this. Introduction: Electrical Impedance Tomography (EIT) is a bedside monitoring technique of regional ventilation that measures the changes in the impedance within the thorax. The tight correlation between variation of impedance (ΔZ) and variation of lung volumes (Vt) is known. Unless the Vt is measured by an external reference (e.g. spirometry) its absolute value (in ml) cannot be determined, however measurement of Vt would be useful in non intubated subjects. Methods: We performed a prospective study on thirteen healthy volunteers (>18 years) without obesity, pregnancy, chest circumference <75 cm and history of pulmonary or medical disorders. The subjects were connected to the EIT monitor. Calibration phase: the subjects breathed 10 times, from functional residual capacity, fully distending a non-elastic bag with a volume of 1680 ml. In this way the conversion factor between Vt and ΔZ (K[sub]Vt/ΔZ[/sub]) was found for each subject. Validation phase: tidal volumes were estimated (estVt = ΔZ* K[sub]Vt/ ΔZ[/sub]) from impedance variations; the accuracy was assessed measuring tidal volumes (Vt) with a ventilator through which the subjects breathed by a mouthpiece. Four different ventilator settings were used changing pressure support ventilation (PSV) and positive end expiratory pressure (PEEP). Results: The correlation between Vt and estVt was tight (r 2 = 0.89) with a within-subject mean r 2 of 0.91 ± 0.07. The fit equation was estVt = 0.9 * Vt + 10.1 (Fig. 30). The highest correlation was found at PEEP 0 (mean: estVt = 0.93 * Vt) vs PEEP 8 (mean: estVt = 0.8 * Vt), p = 0.01. No differences were found between PSV 0 (mean: estVt = 0.97) and PSV 8 (mean: estVt = 0.93 * Vt), p = 0.50. The Bland-Altman plot showed a systematic bias of −95,5 ml (−10%) and 95% CI of -396 ml (−40%); 205 ml (20%).  Introduction: The recruitment maneuver (RM) is a part of clinical strategy of respiration supporting, The subject of the research was to monitor the phenomena occurring during RM when the artificial ventilation is applied at the mouth of animal ARDS model, using modified low frequency forced oscillation technique (FOT). Methods: The phenomenon of the opening alveoli was modelled according to the estimations made in the complex mathematical model of respiratory system. The complex model structure was reduced into its simpler, identifiable analog of nonlinear resistance and capacitance of the lung. A Large White pig weighing 50 ± 5 kg were included into the study. Under general anaesthesia the animal was tracheostomised, intubated and ventilated in a volume-controlled mode (Puritan Bennett 840) at RR-18/min, with VT-8-10 ml/kg, I:E ratio −1:2 and FIO2-1.0. Lungs injury was induced by repetitive lavages with 1.5-1.8 l warm 0.9% NaCl. The low frequency (0,5Hz) sinusoidal flow oscillations were induced at the intubation tube by means of mechanically controlled syringe. The flow and pressure changes were monitored by pneumotachograph (Hans Rudolph) while the external negative pressure (eNP) in the whole body size-box was changing. For the estimation of timevarying mechanical parameters (Rest and Cest) of the lung the Kalman filter method of data analysis was used. Results: Application of the proposed method of the identification of the parameters of lung during the vacuum changes in the chamber results as in the example figure (Fig. 31). There were significant differences in Rest and in Cest at the body size-box negative pressure 0, −4, −8, −12, and −16 cmH2O. Conclusions: The modified low frequency forced oscillation technique allows observation of changes in the parameters typical for recruited lungs and assess the possibility of monitoring of changes of physiological parameter values when the RM is applied in artificially ventilated lungs. Introduction: There is limited data on safe tidal volume and pressure setting in pressure support ventilation (PSV). Methods: We included 59 mechanically ventilated patients in SICU with PSV mode. We used monitoring of airway pressure, tracheal pressure, and esophageal pressure, plotted dynamic «tracheal pressure-volume» (Ptr-Vt) and «esophageal pressure-volume» (Pes-Vt) loops. We measured tidal volume (Vt), patient's work of breathing (WOBp), delta esophageal pressure (ΔPes), plateau pressure (Pplat), transpulmonary plateau pressure (Ptp plat), transpulmonary pressure at PEEP (Ptp PEEP), dynamic compliance of respiratory system (Cdyn), dynamic lung compliance (Clung dyn), dynamic chest wall compliance (Ccw dyn), delta transpulmonary pressure (ΔPtp = Ptp plat -Ptp PEEP), tracheal pressure at PEEP level (PEEPtr), minimal tracheal pressure during triggering (Ptrig tr), and calculated difference between PEEP tr and Ptrig tr. After that we plotted loops and estimated its shape. We collected data at 6 steps: (1) at baseline PS and PEEP level, (2) at PS + 4 mbar level and baseline PEEP, (3) at PS-4 mbar level and baseline PEEP, (4) at PEEP + 4 mbar level and baseline PS, (5) at PEEP set by end-expiratory transpulmonary pressure level (PEEPtp0) and baseline PS, and then (6) at PEEP-4 mbar level and baseline PS. Results: We discovered 3 typical shapes of inspiratory part of «tracheal pressure-time» (Ptr-t) curve: triangle, square and Sshape. We found significant differences between different shapes of inspiratory part of «Ptr-t» curve in WOBp (p < 0.0001), ΔPes (p < 0.0001), PEEPtr-Ptrig tr (p < 0.0001), and PS-ΔPtp (p = 0.002). We discovered 4 types of dynamic «Ptr/Vt» loops: inverted (n = 41), classical (n = 7), linear (n = 7), and S-shaped (n = 4). We found significant differences between different shapes of dynamic «Ptr/Vt» loops in ΔPtp (p = 0.05) и Clung dyn (p = 0.020), which allowed to estimate delta transpulmonary pressure without esophageal pressure monitoring. We found significant differences between different shapes (v-,u-,w-and v + −shaped) of inspiratory part of «Pes-t» curve in WOBp (p = 0.002), PEEPtr-Ptrig tr (p = 0.05), PS-ΔPtp (p = 0.034), and Ccw dyn (p = 0.014). Vt was more than 6 ml/kg of ideal body weight (IBW)(7.8 (6.9-9.1)ml/kg) in 88.1% of patients at baseline while ΔPtp was 16.3 (12.1-19.7) mbar, and «safe» Vt (in which ΔPtp <15 mbar) had 46.2% of patients. Conclusions: Monitoring of «Ptr-t» curve and dynamic «Ptr/Vt» loop allow for estimation of the lung injury and WOBp without Pes an Ptp monitoring.  Introduction: Evaqua technology allows the humidity in the expiratory limb to diffuse through the permeable membrane of the expiratory limb wall before it has an opportunity to condense into liquid water within the circuit limb. As the membrane is not permeable to pathogens it seems to reduce the risk of pathogen transfer to the patient. With conventional circuits, condensate can form inside the expiratory limb when humidity in the expiratory gas flow touches cooler surfaces along the gas path. In this study, we aimed to evaluate the effect of respiratory circuit systems on the development of ventilator associated pneumonia (VAP). Methods: We evaluated 41 patients retrospectively who were admitted to intensive care unit (ICU) and mechanically ventilated using 2 different circuit systems between May, 2013-August, 2015 (Group 1: Adult dual heated wire ventilator circuit, Group 2: Adult ventilator circuit dual heated with Evaqua technology). Results: Patients' characteristics are showed in Fig. 32. There was statistically difference only in gender status between two groups. Respiratory failure (28.6 vs 30%), Postresuscitation Syndrome (PRS) (23.8 vs 30%) and multiple traumas (28.6 vs 20%) were the main reasons for admission to ICU. Mean duration of hospitalization in ICU and duration of mechanical ventilation were similar in two groups (53.5 (9-174) vs 37 (11-102) days, 45 (9-152) vs 32.5 (4-102) days, respectively). Positive result in tracheal culture was obtained 100% and 85% in Group 1 and 2. Acinetobacter Baumannii, Pseudomonas Aeruginosa and Klebsiella Pneumoniae were the most detected agents in the weekly tracheal culture results. Mean duration time to positive tracheal culture was similar in two groups (10.1 ± 5.9 vs 13.6 ± 9.1 day). Conclusions: We could not find any significant advantage on development of VAP in favor of Evaqua technology. Well-designed prospective studies with more patients are necessary for better understanding of the effect of respiratory circuit system on development of VAP. Introduction: Opioids are commonly used for postoperative pain management but often decrease respiratory drive and can cause opioid-induced respiratory depression. Current respiratory monitoring in non-intubated patients relies on late indicators of respiratory depression, such as pulse oximetry or point of care measurements of RR [1,2]. Here, we assess the effectiveness of respiratory rate (RR) alone to detect respiratory depression by using a non-invasive respiratory volume monitor (RVM) which accurately measures minute ventilation (MV), tidal volume (TV), and RR [3]. Methods: Impedance-based RVM (ExSpiron, Respiratory Motion, Waltham MA) was used to non-invasively collect MV, TV and RR measurements from 104 patients (55 males, BMI: 27.2 ± 5.0 kg/ m2) recovering from elective major abdominal surgery. MV, TV and RR were calculated from 30-second respiratory segments for up to 48 hours following surgery. Predicted MV (MVPRED) was calculated for each patient based on body surface area. LowMV was defined as MV < 40% MVPRED and LowRR was defined as RR < 6 breaths/min. RR rate values were compared to MV measurements and sensitivity and specificity of LowRR as a predictor of LowMV were calculated. Results: Patients were monitored for an average of 34.5 ± 14.7 hours in the PACU and general hospital ward. Analysis of all 417,850 paired MV and RR measurements revealed that although MV is a function of RR (MV = TV*RR), there was poor correlation between a given MV measurements and its corresponding RR measurements (r = 0.35). 56.4% of MV recordings were below 80% MVPRED and 19.1% were below 40% MVPRED while only 4.0% were below 6 breaths/min. A variety of RR alarm conditions (4-8 breaths/min (bpm)) were explored which showed that a substantial fraction of low MV measurements remain undetected. Specifically, with a RR cutoff of 8 bpm, 74% of all MV measurements < 40% MVPRED would be missed. Decreasing the RR cutoff to 4 bpm misses 93% of MV measurements <40% MVPRED. Overall, LowRR was a poor predictor of LowMV with a sensitivity of 15.0% and specificity of 98.6%. Furthermore, 29% of all low RR events were associated with adequate MV, indicating that patients were adequately ventilated even in the presence of LowRR. Conclusions: Our data suggest that LowRR alone does not accurately reflect episodes of LowMV, and is not sufficient for accurate assessment of respiratory status. The tidal volume of each breath is at least equally critical to ensure respiratory sufficiency. Introduction: Weaning failure of cardiac origin can be diagnosed by the elevation of the left ventricular (LV) filling pressure. This hydrostatic pulmonary edema is associated with hemoconcentration du to hypooncotic fluid movement from the vascular compartiment to the interstitium. The aim of this study was to search for a correlation between the protidemia and the hemoglobin elevation during the weaning test and the presence of LV filling pressure elevation.
Methods: This prospective observational study was conducted during two years (between January 2014 and January 2016). Every patient with weaning failure was included. The variation of the biological and ultrasound criteria between the periods before and after the ventilator weaning was analysed. Results: We included 56 patients which failed during the first weaning test. 20 patients (35,7%) had an elevation of the LV filling pressure during the second weaning test with E/A > 0.95 and E/Ea > 8.5 at the end of the test period and within these 20 patients, 12 (60%) failed this second weaning test. When compared to the 36 patients who didn't present a pulmonary edema, these 12 patients required additional delay for the weaning: 2.5 + 3.7 days versus 0.75 + 2.4 days (P = 0.023). the unique predictive factor associated with the occurrence of pulmonary edema was positive body weight difference between the admission in the ICU and the inclusion in the study (4.6 + 5.6 versus 3.5 + 7.3, p = 0.004). the was no significant variation of the hemoglobin and protidemia during the weaning for the patients who presented a pulmonary edema. Conclusions: During this study we didn't find a correlation between the hemoglobin or the protidemia variation during the weaning period and the ultrasound criteria of th LV filling pressure elevation. The difference of body weight between the admission in the ICU and the inclusion, which reflect a positive fluid balance, was the unique factor associated with weaning failure of cardiac origin. Its control could allow a more frequent weaning success and the diminution of the morbidity and mortality due to the diminution of the mechanical ventilation duration Introduction: The prevalence of diaphragmatic dysfunction (DD) in patients admitted to Intensive Care Units (ICUs) has often been underestimated, even though it occurs quite frequently.1,2 The first aim of our study is to estimate its prevalence in patients with acute respiratory failure admitted to our ICU; the second aim is to show if Non Invasive Ventilation (NIV) can be helpful in these court. Methods: We enrolled patients with acute respiratory failure admitted to our ICU, scheduled to perform NIV. To measure the diaphragm excursion, a trans-thoracic ultrasound (US) examination was performed before (T0) and after 30 minutes of NIV (T1).3 NIV was performed through a full-face mask with a pressure support ventilation of 6 to 10 cmH2O and a Positive End Expiratory Pressure (PEEP) of 5 cmH2O. Inspired oxygen fraction was set to obtain an SpO2 > = 92-93%. Results: In our general ICU, the overall prevalence of DD was 55% (11 over 20 pts). A subgroup analysis revealed that 90% (9 over 10 pts) of the post-operative patients had DD, compared to only 20% (2 over 10 pts) of medical origin. All patients with normal diaphragmatic functionality were NIV responder. The sensitivity and specificity of DD in predicting NIV failure was, respectively, 100% and 60%; negative predictive value was 100% while positive predictive value was 46%. Conclusions: The main finding of this study is that DD showed a high prevalence: 55% of our mixed population (surgical and medical). The presence of a normal diaphragmatic functionality seems to be an indicator of good response to NIV. Introduction: Mechanical ventilation unloads the respiratory muscles in order to prevent development of muscle injury and patient discomfort. On the other hand, over-assist may lead to diaphragm inactivity associated with disuse atrophy and patient-ventilator asynchrony. Clinicians are often unable to recognize diaphragm inactivity based on pressure and flow signals on the ventilator. In recent studies an index was described, the neuromuscular efficiency (NME), in which the electrical activity of the diaphragm (EAdi) was used to determine the respiratory muscle effort. However, conversion of EAdi (in microvolt) into pressure has not been very well validated today. The aim of the current study was to assess the repeatability coefficient (RC) of the NME (dynamic and during occlusion). Methods: We included 31 mechanically ventilated adult ICU patients with a dedicated naso-gastric feeding tube for assessing diaphragmatic EMG activity. NMEoccl was calculated by measuring the change in airway pressure (delta Paw) divided by the EAdi during an end expiratory occlusion; repeated 5 times with a 1-minute interval at inclusion and 12, 24, 72 hours respectively. NMEdyn was calculated by dividing the peak inspiratory effort of the diaphragm (Pmus = chestwall compliance + esophageal pressure) by the corresponding EAdi during regular tidal ventilation. Results: The repeatability coefficient of NMEoccl was 81.7%, which means that the difference between two repeated measurements lies between this ratio with a 95% confidence interval. For example, with a calculated NME of 0.93cmH2O/μV and a repeatability coefficient of 81.7% it is expected that 95% of the subsequent measurements will be between 0.17-1.69 cmH2O/μV. Both the Eadi peak as the Paw peak showed a wide variation in which just a moderate correlation was found (p = 0.44). Also after additional analysis, the RC did not improve. The repeatability coefficient of NMEdyn was 125,7%. Conclusions: The measurements of the NME is feasible in daily practice, but the repeatability coefficient of both NMEoccl and NMEdyn seems too high for clinical use. Introduction: A preserved respiratory muscle function is essential for the liberation of patients from mechanical ventilation. Evidence suggests that a prolonged time of ventilation leads to diaphragm atrophy and dysfunction. Bedside diaphragmatic function monitoring may allow for a more complete evaluation of the patient's respiratory effort. Surface Electromyography (sEMG) is a non-invasive technique for the assessment of diaphragmatic electrical activity. [1] Diaphragm Ultrasound (US) provides a reliable evaluation of its motion. [2] Aim of the present study was to assess, through US, the motility corresponding to sEMG-derived electrical activity signals of the healthy diaphragm. Methods: 12 healthy volunteers underwent a standard breathing protocol including three different respiratory patterns: Quiet Breathing (QB), Deep Breathing and Voluntary Sniffing (VS). Analysis of sEMG recordings was performed to assess baseline and maximum electrical activity (sEMG max, sEMG delta), and the Area Under the Curve (sEMG AUC). Simultaneously we assessed baseline (Tmin) and maximum muscle thickening (Tmax) with a linear 10 Mhz probe positioned at the 9th intercostal space in the mid-axillary line. We calculated Thickening Fraction (TF) as: 100(Tmax-Tmin)/Tmin. We then assessed Diaphragm Excursion (DE) with a convex 7.5 Mhz probe placed below the right costal margin. Results: In healthy volunteers the simulation of different respiratory patterns led to statistically significant differences in terms of both sEMG and US parameters. TF differed among QB (24,8 ± 11,4%), DB (154 ± 40,8%), and VS (96,9 ± 33,3%, p < 0.0001). As well, differences in sEMG delta could be measured (QB: 0.99 ± 0.5 μV; DB: 16.8 ± 9.43 μV; VS: 13.03 ± 5.87 μV, p < 0.0001). A significant correlation could be identified between several parameters obtained by the two techniques. Particularly, sEMG max vs. Tmax  Introduction: In ICU the dysfunction of respiratory muscles, during mechanical ventilation is a relevant problem which may lead to prolonged ventilation and difficult weaning. Some data suggest that ventilator induced diaphragmatic dysfunction (VIDD) can be aggravated by neuromuscular blockers and high doses corticosteroids. Aims of this study are to evaluate the coherence between ultrasound (US) and computed tomography (CT) measurements of diaphragmatic thickness in mechanically ventilated patients, and their relationship with clinical features and outcomes. Methods: We enrolled intubated or tracheotomized patients undergoing mechanical ventilation for at least 48 hours undergoing chest CT scan for clinical reasons. Patients under the age of 18 years, with neuromuscular diseases, phrenic nerve lesions or air leakage were excluded. In enrolled patients thickness of the right hemidiaphragm was evaluated with US as described by Goligher [1] (within 12 hours prior to or after CT scan), while thickness of right and left diaphragm (anterior, posterior pillars, domes) was evaluated with CT scan. At the time of enrollment clinical data were collected. Results: We enrolled 24 patients in whom 35 CT and US were performed. US evaluation of the right hemidiaphragm was feasible in all patients, and with CT scan in all patient except one. CT measurements of thickness in different portions were tightly related (r = 0,631 ± 0,142). Consistency between US and CT is supported by significant correlations between measurements, in particular posterior diaphragmatic pillars evaluated with CT scan showed the highest correlation coefficient with US, r = 0,439;p-value = 0,009 and r = 0,573;p-value < 0,001 respectively for right and left. Finally we found significant correlations between thickness of posterior pillars and duration of controlled mechanical ventilation (Fig. 36 -  Introduction: Weaning failure (WF) from mechanical ventilation (MV) may be due to lung derecruitment, cardiac dysfunction and respiratory muscles weakness. Transthoracic echocardiography (TTE) [1], lung ultrasound (LUS) [2] and diaphragm ultrasound (DUS) [3] have shown their value in early identification of the failing patients separately. A combination of TTE, LUS and DUS could improve the identification of failing patients and of WF etiology [4]. We aimed to estimate the value of a combined ultrasound assessment (heart-lung-diaphragm) to early identify patients at high risk of WF from MV. Methods: Prospective observational multicenter international study including all patients undergoing a 30' spontaneous breathing trial (SBT) before extubation. Patients with neuromuscular diseases and MV <48 hours were excluded. TTE and LUS were performed before SBT and at its end; DUS at beginning and end of SBT. TTE included: MAPSE (mitral annulus plane systolic excursion), EF% (ejection fraction), E/A, E/Ea. We computed LUS score (0-36) [2]. DUS assessed diaphragm excursion. Extubation was considered as failed in case of reintubation, non-invasive ventilation or death within 48 hours. We used classical criteria for SBT failure. Extubation was decided by an independent operator. Results: We enrolled 18 patients (age 77.7 ± 11.4 yrs, BMI 29.9 ± 7, SAPSII 54.4 ± 17.4, MV length 9.5 ± 7.3 days). 6 patients (33%) failed the SBT and were not extubated. 12 patients (67%) were extubated; 5 failed. Two populations were identified: WF (failed SBT or extubation) and weaning success (WS). LUS: SBT-LUS was higher than MV-LUS in the whole population (17 ± 4 vs. 12 ± 3; p = 0.019); this was more remarkable in WF (20 ± 3 vs. 11 ± 3; p = 0.0004). SBT-LUS was higher in WF vs. WS (20 ± 3 vs. 13 ± 4; p = 0.03). SBT-LUS predicted SBT and weaning failure (respectively AUC 0,877 and AUC 0,883). TTE: E/A and E/Ea (both MW and SBT) did not predict WF, extubation failure nor SBT-failure. MV-MAPSE predicted WF (AUC 0,833); if < =10 mm it predicted WF with sensitivity 0.67 and specificity 1. SBT-MAPSE predicted extubation failure (AUC 0.833). DUS: No correlation between diaphragm excursion and WF was identified. Conclusions: LUS and MAPSE seem to be the most useful parameters to predict weaning failure. Introduction: Tidal volume and respiratory rate are the usual variables to follow the weaning from mechanical ventilation. Although different studies show controversies; because these parameters are influenced by several factors on critically ill patients, and could do not reflect with accurately the work of breathing. Our objective was to study the relationship between respiratory pattern and variables related with the work of breathing. Methods: We studied 11 patients ready to maintain spontaneous breathing after recovering form acute respiratory failure from different causes. Respiratory flow (V'), airways and esophageal pressure (Paw, Pes) were registered for ulterior analysis of these signals. Respiratory rate (RR), delta Pes and pressure time product (PTP) were measured from Pes. The respiratory mechanics during control mechanical ventilation, dynamic elastance (Ers) and total resistances (Rrs), were calculated by multiple linear regression techniques. Results: Age 44 ± 18 y, ventilated since 1 to 70 days (23 ± 22). Ers 26 ± 9 cmH2O/l, Ecw 10 ± 6 cmH2O/l, Rrs13 ± 7 cmH2O/l/s, PEEPi 4 ± 6 cmH2O. On CPAP (5 ± 4 cmH2O), tidal volume 0.350 ± 0.185 l, Inspiratory V' 0.68 ± 0.16 l/s, RR 24 ± 14 bpm. Variables related with respiratory effort were: PTP 7 ± 4 cmH2O/l•s, PTPm 189 ± 108 cmH2O/l•s/min, delta Pes 10 ± 5 cmH2O. These variables were related with tidal volume (r 0.8;0.5;0.7), however respiratory rate was unrelated (0.05;0.5;0.3).
Conclusions: The respiratory effort may not be reflected in the global ventilatory pattern during trial of spontaneous breathing, with risk of respiratory muscle injury and fatigue. Therefore it is necessary to analyze other breathe components by additional monitoring. Introduction: Lung ultrasound (LUS) is useful to assess lung edema in many clinical situations. We hypothesized that LUS might be a good tool to assess gas exchange defect (GED) in patients with the Acute Respiratory Distress Syndrome(ARDS). Methods: This was a prospective observational study at an academic intensive care unit. We Included adult septic patients with ARDS. A simplified lung edema scoring system (SLESS) was used, and 6 thoracic regions were evaluated. Four LUS patterns were considered, from normal aeration to consolidation. To evaluate the GED, the SLESS was compared with the PaO2/fraction of inspired oxygen ratio (PaO2/FiO2) and the partial pressure of carbon dioxide (PCO2). Results: Fifty-seven patients were enrolled. Clinical characteristics are presented in Table 11. Figure 37.a shows correlation between SLESS and PaO2/FiO2; Figure 37.b shows the correlation between the SLESS and PCO2 levels. Figure 38 shows a stepwise increase in SLESS among ARDS stages (mean SLESS = 12.21; 14.32 and 16.22, respectively, for mild, moderate and severe ARDS; p < .05 for * and **). Conclusions: The SLESS seems to be useful to assess the gas exchange deffect in patients with ARDS. It also seems to correlate with ARDS severity stages. Introduction: In critically ill patients ultrasonography has gained widespread acceptance for lung monitoring of ventilated patients [1]. Lung Ultrasound Score (LUS) has been demonstrated to be a valid tool to monitor semi-quantitatively lung aeration using four patterns corresponding to increasing loss of air content [2]. Computed tomography (CT) scan remains the reference method, nonetheless X-ray exposure is a relevant concern, particularly in children. We developed a modified pediatric lung ultrasound score (pLUS) to assess aeration in infants. Aim of this study was to compare pLUS with CT scan. Methods: Mechanically ventilated children < = 1 years old who underwent CT-scan for clinical purposes received a lung ultrasound examination. Each hemithorax was divided in six regions according a systematic protocol examination [1]. For each explored region, the worst finding was scored as follows: normal: 1, single B-lines: 2; multiple non-coalescent B-lines: 3; coalescent Blines: 4; white lung: 5; consolidation: 6. A cumulative pLUS was calculated as the sum of each examined region in order to obtain a comprehensive picture of the lung. CT scans were analyzed determining, for each lung region the median density and percentage of aerated lung (fraction of voxels with a density < −500 Hounsfield Units). Results: Nine infants (median age 86 days [36-242]) were enrolled between September 2015 and September 2016 (18 lungs, 108 regions). Regional pLUS and cumulative pLUS of each lung showed a good correlation with median CT-density (R = 0.83; p < 0.05 and R = 0.71; p < 0.01, respectively). Moreover, regional pLUS had a closed correlation with fraction of aerated lung (R = 0.99; p < 0.01) (Fig. 39).   Conclusions: According to these preliminary data, pLUS seems to be a reliable method to assess lung aeration in infants showing a close correlation with aeration lung determined by CT scan. Introduction: Lung ultrasound (LUS) is a useful tool for lung diseases' assessment and monitoring [1]. A lung aeration score can be computed on the basis of type and number of visualized artefacts per scan [2]; this allows semiquantification of lung aeration. LUS score is affected by the length of visualized pleura, significantly different in different scans (longitudinal (Long) vs. transversal (Transv)) [3]. The choice of the probe may also affect the length of visualized pleura, and therefore type and number of artefacts useful for lung assessment and LUS score computation. No clear indications on the most appropriate probe are reported in recommendations [4] and different kinds of probe (microconvex-Mi, linear-Li, phased-array-Pa, convex-Co) can be found in literature [5][6][7][8], limiting the possibility to standardize the exam. We prospectively compared 4 probes. Methods: Prospective observational monocenter study. In each patient, we scanned 6 standard areas per lung (anterior, lateral, posterior, each divided in superior and inferior) with 4 probes (Pa 2.5 MHz, Co 4 MHz, Li 10 MHz, Mi 10 MHz), in Long (craniocaudal orientation) and Transv (aligned with intercostal space) scan. The length of visualized pleura was measured by a caliper. Results: We enrolled 6 patients (2 males, age 63 ± 23 yrs, BMI 24.7 ± 2.0 kg/m2), corresponding to 72 areas and 288 scans. In all probes, Long has poorer performances than Transv scan: it visualizes significantly shorter pleura with higher pleural length variance (Table 12). In Transv, the length of visualized pleura shows significant differences among different probes. Co visualizes the longest pleura, but also has the highest pleural length variance.

References
Conclusions: Transv visualizes longer and more constant pleura with any probe; it's confirmed to be a better approach to assess the lung. The length of visualized pleura and its variance differ among probes; linear probe seems to offer the best combination of high visualized pleura and low pleural length variance. Introduction: Chest ultrasonography is commonly performed in the ICU to identify pleural fluid and to guide thoracic drainage. Aim of this study was to compare the accuracy of two ultrasound methods of estimating the volume of pleural effusions(PEV) with sonography in intubated ICU patients Methods: The first method used to measure PEV was V = 20xSep [1]. Sep was the maximal distance between parietal and visceral pleura, recorded in end-expiration at the posterior axillary line in supine position with trunk elevation at 15°. The compared method multiplied PE paravertebral length (LUS), assessed between the apical and the caudal limits in supine patients, by its cross-sectional area at mid length (AUS) [2]. Moreover the estimated PEV was compared to the volume drained. Patients with incomplete drainage, suspect of empyema, high bleeding risk or interpleural distance less than 10 mm were excluded. T-test, Bland-Altman were used to compare the results. Results: 10 patients (average age 63,2 ± 12,5) were recruited. Bland-Altman analysis showed no differences between the two methods, which can therefore be regarded as interchangeable (fig. 40). Both ultrasound approaches resulted highly correlated with drained PEV (respectively R0,75 R20,56 P0,012, and R0,92 R20,86 P0,001), and so accurate in estimating it. Coefficients of correlation were totally similar to those reported in previous publications. Ultrasound was performed after a 5-h professional training. These results underlined the simplicity of execution and learning of the ultrasound methods.  Conclusions: Based on the results in the current study we conclude that: 1) It is possible to accurately detect arterial oxygen saturation changes with reflection PPG-recordings based on sternal blood flow on an individual level.
2) Arterial oxygen saturation can be calculated using only the DCcomponent of the PPG-signal from infrared and red lights.
3) Arterial oxygen saturation changes are detected markedly faster on the sternum compared to more distal sites of measurement.   Introduction: Marathon race transiently elevates the probability of sudden death. Also during long-distance run may occur various gastrointestinal symptoms with range from mild nausea to hemorrhagic stool. However microcirculatory nature of this disturbances is not clear. Microcirculation of sublingual mucosa is part of interest, because it is easy and noninvasively accessible, changes have relation with mortality and it is part of the upper digestive tract. Here, we evaluate changes in sublingual microcirculation induced by a marathon race. Methods: Thirteen healthy male controls and 13 male marathon runners volunteered for the study. We performed sublingual microcirculation, using a Cytocam-IDF device (Braedius Medical, Huizen, The Netherlands), and systemic hemodynamic measurements four times on the marathon runners: 24 hours prior to their participation in the Kaunas Marathon (distance: 41.2 km), directly after finishing the marathon, 24 hours after the marathon and one week after the marathon. Results: The marathon runners exhibited a higher functional capillary density (FCD) and total vascular density of small vessels at the first visit compared with the controls. Overall, we did not find any changes in sublingual microcirculation in the marathon runners at any of the visits. However, in a subgroup of marathon runners with a decreased FCD after finishing the marathon race compared to increased FCD had shorter running time (190.37 ± 30. Introduction: We hypothesized that that sublingual (SL) microcirculatory parameters 1. Can predict microvascular response to resuscitation during sepsis; and 2. Are associated with tissue level lactate, lactate/pyruvate ratio (L/P) and tissue to blood lactate gradient (T-BLac). Methods: Lipopolysaccharide (LPS) was administered to 23 anesthetized Yorkshire-Durock pigs for 45 minutes. Thirteen animals received late (90 min after LPS) and 10 early (immediately after LPS) resuscitation. Five animals per group had available data for this opportunistic study. Sublingual microcirculatory parameters (microvascular flow index (MFI) and perfused vessel density (PVD)) were collected. Tissue level lactate and pyruvate were measured using microdialysis catheters inserted in the liver, kidney and tongue at baseline and pre-/post-resuscitation (PreR, PostR). Resuscitation was driven by MAP and SvO2 as targets and using SVV for fluid responsiveness. Data are shown in median (interquartile range). Non parametric statistics were used whenever appropriate. Results: Pre-R MFI correlated with Post-R MFI (r2 = 0.562, p = 0.008). Microvascular 'responsive' animals (i.e. increase in MFI > 50% or any increase in PVD) had lower Pre-R MFI and PVD (Fig. 42A). Pre-R PVD correlated with Pre-R SL lactate (r2 = 0.4, p = 0.08). A higher Pre-R MFI (>2.5) was associated with lower SLT-BLac (−0.42 (0.74) vs. 3.49 (0.34), p = 0.008). At the Post-R time point, PVD was associated with liver L/P (Fig. 41B), and MFI with renal L/P (r2 = 0.31, p = 0.09). Animals with increased Post-R renal T-BLac had lower Pre-R MFI (1.25 (0.25) vs. 1.75 (0.75), p = 0.07). Conclusions: Pre-resuscitation MFI and PVD may predict postresuscitation microvascular responsiveness. Importantly, pre-and post-resuscitation MFI and PVD were associated with local metabolic changes in the tongue as well as in vital organs, particularly the kidney and the liver.

P90
Retrospective study over a 12-month period looking at the national early warning score as a screening tool for patients with sepsis admitted to intensive care H. Lyons, A. Trimmings East Sussex Healthcare, East Sussex, UK Critical Care 2017, 21(Suppl 1):P90 Introduction: The definition of sepsis has changed recently, pushing us to redefine and examine how we screen for sepsis [1] [2]. The bedside clinical score termed quickSOFA (qSOFA) has been introduced as an additional tool to identify patients with suspected sepsis who are likely to have a prolonged ICU stay or to die in hospital. The trigger score for qSOFA correlates with a NEWS of either 4 or 5 depending on which qSOFA parameters are abnormal. We have recently updated the sepsis screening tool at our institution, amending that published by the UK Sepsis Trust with their permission. We decided to use a NEWS score of 5 as a trigger by local 'sepsis steering group' consensus. Whilst reaching this consensus decision, however, we also wanted to look at NEWS for our local population admitted to critical care with sepsis. Methods: Our Trust is a District General Hospital comprising 19 medical and surgical critical care beds across two separate sites. All patients admitted to critical care with a diagnosis of sepsis as per ICNARC coding method from April 2015 until March 2016 were identified using 'Wardwatcher software' (Critical Care Audit Ltd, West Yorkshire, UK). This period was before the introduction of the updated sepsis screening tool. We then looked retrospectively at highest NEWS prior to admission to critical care using early warning score electronic records (VitalPac, The Learning clinic, London, UK). Admissions to critical care were from the emergency department, general medical and surgical wards. Results: 200 patients were admitted to critical care with a diagnosis of sepsis during the study period. 47 patients were excluded as NEWS data was not available for these patients. Overall mortality was 34%. 17.6% had a NEWS between 0 and 4. This cohort consisted mainly of young patients with single organ dysfunction who had the lowest mortality (6% of all deaths). The majority of patients fell into the middle category -47.8% scored NEWs 5-9 The remaining score a NEWS of greater or equal to 10, this accounted for 34.6% of selected patients. The mortality in this group was high as was the length and complexity of the stay. Conclusions: In our local population NEWS score of 5 is a reasonably sensitive screening trigger for identifying the septic patient at risk of significant deterioration. The patients with a NEWS less than 5 and admitted to ICU with sepsis had a lower morbidity and mortality than those with a NEWS score over 5. Introduction: Microvascular alteration is associated with organ dysfunction and adverse outcome in several subsets of critically ill patients. We performed a daily monitoring of sublingual microcirculation to evaluate the association between microcirculatory alterations and the development of organ dysfunction in polytraumatic patients. Methods: This is a subgroup analysis of a perspective observational study on a mixed Intensive Care Unit (ICU) population. 36 polytraumatic patients. Sublingual microcirculation was monitored with Sidestream Dark Field (SDF) imaging daily from admission to discharge/death. Parameters of vessel density and flow quality were calculated with a dedicated software. Organ function was observed and SOFA score was recorded daily. Results: Patients who had a SOFA ad day 4 < 6.5 showed a higher Total Vessel Density (TVD) at day 1 and 2 (Fig. 43). TVD at day one was also a good predictor of SOFA > = 6.5 at day 4 (Area Under the Receiver Operating Characteristic Curve 0.826; p = 0.002; 95%CI = 0.681-0.970) (Fig. 44). Analogue results were obtained for the Perfused Vessel Density (AUC 0.p99773; p = 0.009; 95%CI = 0.604-0.942). Microvascular Flow Index did not show any predictive value towards development of organ dysfunction. Conclusions: The presence of altered microcirculatory vessel density and perfusion in the first 48 hours from the ICU admission resulted to be associated with the development of organ dysfunction in polytraumatic patients. Introduction: Skin mottling is extreme manifestation of microcirculatory changes in septic shock. Skin discolouration is thought to be due to heterogenous changes in oxygen saturation and volume of microcirculatory blood which results in alterations in the amount of reflected light in various parts of the spectrum. These alterations can be captured using hyperspectral imaging. We aim to describe mottled knee skin relative total hemoglobin concentration and microcirculatory saturation values in survivors and non survivors of septic shock and assess their prognostic information. Methods: Hyperspectral imaging was performed in 62 intensive care patients with septic shock enrolled within 24 hours of admission in single centre observational study to obtain relative oxy/deoxyhemoglobin concentration maps and numerical values from mottled area around the knee. Images were processed to identify areas with purple blue discolouration. Nonlinear fitting of optical density spectra was used to calculate relative oxy/deoxy haemoglobin concentration and obtain total hemoglobin concentration (a.u.) and microcirculatory oxygen saturation values (%). In addition, demographic (age, sex, primary site of infection), hemodynamic (mean arterial pressure, dose of vasopressor agents) and disease severity (SAPS II, SOFA score) data were collected.   values were significantly lower (29.7 ± 36.1% vs 76.3 ± 30.7%;P < 0.001). Skin microcirculatory oxygen saturation of more than 22% provided 72% specificity and 69% sensitivity for predicting survival. Increase of relative skin total haemoglobin concentration (RR 1.11 (1.05-1.20)) and decrease of skin oxygen saturation by 5 units (RR 1.19 (1.09-1.30)) was associated with increase in hospital mortality. Conclusions: We used a non contact hyperspectral imaging system to quantify total haemoglobin and oxygen saturation in dermal microcirculation of septic shock patients. Increase in total dermal hemoglobin and decrease in oxygen saturation was strongly predictive of 28 day mortality. Introduction: There is a growing interest in the state of the microcirculation during critical illness. Sidestream Dark Field Imaging (SDFI) is widely used to study microcirculation in vivo. One such SDFI device is the Glycocheck®, which estimates the thickness of the endothelial glycocalyx by measuring the thickness of the Perfused Boundary Region (PBR). Though single measurements are current clinical practice, intra observer variability is unknown. Methods: Microcirculation was assessed in a pilot study of 49 mechanically ventilated mixed ICU patients using SDFI (Glycocheck®). Each patient was assessed three times consecutively within 30 minutes by the same experienced operator. The operator was unaware of the results of the measurements until all three measurements were completed. The Glycocheck® calculated a PBR for each of the measurements. The intraobserver agreement was calculated using an intra-class correlation coefficient (ICC) using SPSS. Results: The mean (SD) PBR was 2,02 (0,26) μm. The ICC for single measures was 0,362, indicating a poor intraobserver agreement for single measures. ICC for average measures was 0,630 indicating reasonable reliability of the averaged value of 3 consecutive measurements. Conclusions: Intraobserver variability in this pilot study was higher than expected. Averaging the results of three consecutive measurements improved data quality. Whether intraobserver agreement was limited by patient factors such as hemodynamic instability or whether it is inherent to the method itself remains unclear. Reliability of measurement of PBR by SDFI should be increased by averaging (at least three) multiple consecutive measurements. A single measurement is insufficient to yield reliable data on PBR diameter assessed by glycocheck. Introduction: To compare the microcirculatory reserve at two levels of mean arterial pressure (MAP) in septic shock patients. Methods: In 22 septic shock patients receiving norepinephrine with a MAP > 75 mmHg (so-called "high MAP") within the first six hours of resuscitation, we decreased the norepinephrine dose in order to target MAP to 65-70 mmHg (so-called "low MAP"). We measured muscle tissue oxygen saturation (StO2) at the thenar eminence by using nearinfrared spectroscopy and the StO2 recovery slope after a vascular occlusion test at "high MAP" and "low MAP". Transpulmonary thermodilution cardiac index was also measured at the both MAP levels. Results: On average, MAP was decreased from 81 ± 3 to 67 ± 3 mmHg. Cardiac index did not change (3.26 ± 1.09 vs. 3.12 ± 1.05 L/min/m 2 , respectively). The decrease in MAP was associated with a decrease in mean StO2 recovery slope (3.00 ± 1.40 vs. 2.61 ± 1.46 units/sec, respectively, p < 0.05). StO2 recovery slope decreased by more than 5% in 16 patients and increased by more than 5% in 5 patients. Decreasing MAP from "high" to "low" values was not associated with any changes in mean StO2 (81 ± 8 vs. 81 ± 9%, respectively). StO2 decreased by more than 1% in 11 patients and increased by more than 1% in 8 patients. There was no difference in baseline cardiac index between patients with a decrease or an increase in StO2 or in StO2 recovery slope. Changes in StO2 or in StO2 recovery slope were not different between patients with (n = 17) and without (n = 5) a past medical history of arterial hypertension. There was no correlation between changes in StO2 and changes in StO2 recovery slope. Conclusions: Decreasing MAP from "high" to "low" values in septic shock patients was associated with a decrease in the mean StO2 recovery slope without any changes in mean StO2. Nevertheless, there was a large interindividual variability with two opposite behaviours: the StO2 recovery slope seems to be greater at the high MAP level in the majority of patients but greater at the low MAP level in a smaller group of patients. Thus, it could make sense to monitor the microvascular reserve in septic shock patients in order to individualize the MAP level to target above 65 mmHg. Introduction: We investigated the relationship between the ultrasonic B profiles and Spectral tissue Doppler echocardiography (E/E' ratio), a non-invasive surrogate for left ventricular diastolic pressures, in patients presenting with suspicion of acute pulmonary edema Methods: This is a prospective observational study of 61 consecutive patients presenting with acute pulmonary edema and B -profile detected by echocardiography with a 5 MHz curvilinear probe. The Filling pressure of the left ventricle considered high when E/E' is equal or > 15 or when value between 9 and 14 with ultrasound chest B pattern. The filling pressure is considered normal if E/E' is equal or below 8 or the value between 9 and 14 with A-line pattern (1). Results: Sixty-one participants were included (49.2% male, with a mean age 66.8). The mean E/E' level in the patients with B-profile was (20.8), compared with the mean level in the patients with an A-profile of (8.2) (p = 0.003). Based on the value of E/E' , the sensitivity and specificity (including the 95% confidence interval) were determined and are shown in Table 13. The systolic function in the subjects with a B-profile was below 50% in 74.3% of the subjects. All the subjects with B profile and systolic function > 50% had elevated NT-proBNP and E/E' > 15. Conclusions: Detecting the B-profile in lung ultrasound is highly sensitive and specific for elevated left ventricular diastolic pressures in patients with acute pulmonary oedema.  Introduction: Haemodynamic parameters are measured as unindexed "raw values" and in part adjusted to unspecific biometric data. We hypothesized that 1.) Indexation to unspecific biometrics is inappropriate. 2.) Haemodynamic parameters might be associated not only to biometrics, but also to "contexts" such as mechanical ventilation MV, position of the CVC used for transpulmonary thermodilution TPTD and (patho)physiological contexts (e.g. heart rate HR, heart rhythm Rh). Methods: Database analysis (10936 TPTDs; 608 patients). Multivariate regression regarding independent association of GEDV, stroke volume SV, SVV, EVLW and CVP to age A, gender G, weight W, height H, HR, Rh, MV and CVC. Results: 1.) All haemodynamic parameters were independently associated to biometrics and contexts (Table 14)  Introduction: We aimed to run a training day focused on practical use of haemodynamic monitoring (HM). Methods of HM have rapidly expanded, trainees' knowledge of, and exposure to, all of them is likely to be variable. HM is of increasing importance which has been mirrored with an increasing array of devices. Trainees undergo assessment of their knowledge of these devices during examinations. Practical application however is learnt by apprenticeship. To encourage a wider and more detailed practical understanding we undertook a HM training day. Methods: We conducted a training day consisting of didactic lectures followed by small group practical sessions. We conducted pre and post course surveys assessing subjective confidence of setting up and interpreting data, and objective assessment via multiple choice questions. Results: There was an increase in confidence across all four monitoring techniques in both setting up and interpreting ( fig. 45 and 46). The MCQ showed a pre and post course mean of 74% and 78% respectively.
Conclusions: The survey has demonstrated inconsistent knowledge of and exposure to the many haemodynamic monitors available. The course has improved both subjective and objective measures. There was a proportionately greater increase in confidence with ECHO and PAC, which may be indicative of the lower overall prior exposure. The MCQ showed minimal increase, and this may reflect the aims of the course were to focus on hands on experience and understanding rather the exam based knowledge.  Introduction: Tachycardia events observed in the Intensive Care Unit (ICU) could lead to cardiorespiratory instabilities and significant morbidity and mortality [1]. We apply machine learning to predict whether and how soon tachycardia will occur based on continuous hemodynamic monitoring data.
Methods: An episode of tachycardia is defined as heart rate over 130 for at least 10% of a 5 minute interval. Based on heart rate, respiratory rate, blood pressure and plethysmography from data in MIMIC-II [2], we extracted 42 statistical features over 30-minute time windows right before the onset of tachycardia. We trained two random forest classification models. The first used the first occurrences of tachycardia during patient stay as cases, and randomly selected controls from patients without tachycardia. The second model used the same cases as above but controls were sampled from tachycardia patients data way ahead of its onset. Results: We evaluated the models using 10-fold cross validation, and tracked predictions made over the entire stays of test patients. Fig. 47 summarizes predicted risk scores and 95% confidence intervals for cases (solid) and controls (dashed) over 3.5 hours leading to tachycardia. Fig. 48 depicts the ratio of risk scores (lift) for cases during 3.5 hours leading to the onset (solid) vs. the risk computed for the same patient during the first few hours of their ICU stay (dashed). Conclusions: Estimated risks scores for cases and controls differ significantly throughout the test period suggesting potential utility of the first proposed model in triage as it helps identify who is going to develop tachycardia with high confidence, and could inform monitoring and care resource allocation in the ICUs. The second model applied to patients at risk reliably signals the upcoming episode more than 1 hour ahead of its onset, allowing preemptive treatment.
The results suggest improvements of quality of care and patient outcomes, and mitigation of costs. Introduction: Several predictive models and scores were developed to predict critical care disease severity and prognosis. QT dispersion, one of the proposed measures of heterogeneity of ventricular repolarization. It was studied extensively in cardiology and there was a trend toward increased cardiac adverse outcome as well as mortality in normal subjects with a prolonged QTd. We aimed to clarify the relationship between QTd     Introduction: It is recognised that both a prolonged tachycardia and a high daily mean heart rate are associated with an increase in Intensive Care patient mortality [1]. This study aimed to look at the highest recorded heart rates during Intensive Care Unit (ICU) admission and the associated rate of survival to discharge from hospital. Methods: A retrospective analysis looking at a total patient population of 11,685 which included consecutive patients admitted to a mixed medical and surgical ICU over a 22-year period. We assessed the highest heart rate recorded during the first 24 hours of ICU admission. Patients were grouped in incremental sets depending on highest heart rate. The defined outcome was survival to discharge from hospital. Fischer's exact test was applied to determine two-tailed p values for the dataset.
Results: Of those patients with a highest recorded heart rate of 90 beats per minute (bpm), or less, 77.4% survived to discharge from hospital. This decreased to 70.6% in patients with a high heart rate of between 101 and 110 bpm (p value < 0.0001) and to 50% or less in patients with a high heart rate of more than 170 bpm (p value < 0.0001) (see Fig. 49). The difference in mortality between those with a maximum heart rate of 90 bpm (77.4%) and those with a maximum heart rate of 100 bpm (75.5%) was not statistically significant (p = 0.172).
Conclusions: The occurrence of tachycardia of greater than 100 bpm during an ICU admission is associated with a reduced rate of survival to discharge when compared with those patients with a high heart rate of 90 bpm or less.  Table 15 present data collected globally and from our centre.
Populations are similar in preoperative risk and comorbidities. Significant differences concern the type of surgery, mean extracorporeal circulation (ECC) and cross clamping (CC) time.
All centres rely on less invasive monitoring, our centre showing remarkable preference for minimally invasive monitoring. While no patient had invasive intraoperative CO monitoring, this was used for 2% of patients in ICU. 36% of patients were monitored intraoperatively by transesophageal echocardiography (TEE), with all patients being monitored by TEE or transthoracic echocardiography (TTE) in the ICU. All centres used preferentially norepinephrine and dobutamine, while milrinone and levosimendan remain peripheral in clinical practice. Mortality is low, with no significant differences in postoperative complications or ICU stay.
Conclusions: Data shows that in our centre low rates of mortality and morbidity can be maintained using minimally invasive monitoring and relying on echocardiography. If this can be a change of paradigm or an isolated report is to be determined.    Introduction: Mean systemic filling pressure (MSFP) is the elastic recoil pressure of the systemic circulation at zero blood flow and it reflects the stressed volume [1,2]. Clinical estimation models have been developed [1] and values reported from dead ICU patients [2]. The time course of vascular pressure equilibration at the end of life has not been reported. Methods: In eight pigs, HES was infused to expand estimated blood volume (6% body weight) by 15% [1]. Pressures were measured in the aortic arch and both venae cavae (PSVC and PIVC). MSFPRAO was estimated by occlusion of the right atrium with an inflatable balloon and averaged from 9 to 12 seconds into the venous pressure plateau. The animals were then killed with potassium chloride in deep anesthesia without relaxation [1]. Circulatory standstill was defined as loss of pulmonary flow (ultrasonic flow probe From initial standstill, it increases to an early venous plateau, then further increases to arterio-venous equilibrium, after which all pressures decline together. The equilibrium pressure is not stable. This should be taken into account when MSFP is estimated after cardiac arrest [2]. Introduction: The aim of this study was to determine the change in critical closing pressure (CrCP) during experimental endotoxemia (Lipopolysaccharide, LPS) and sepsis. In addition, we determined the effect of vasopressors on the vasomotor tone during experimental endotoxemia.
Methods: We performed a prospective observational study, at the intensive care department (ICU) of a tertiary care university hospital the Netherlands, in 40 healthy subjects during experimental human endotoxemia and in 10 patients with severe sepsis or septic shock. Subjects in the LPS trial were randomized to receive either a 5 hour infusion of 0,05 μg/kg/min noradrenaline (n = 10, "LPSnor"), 0,5 μg/kg/min phenylephrine (n = 10, "LPS-phenyl"), 0,04 IU/ min vasopressin (n = 10, "LPS-vaso") or placebo (n = 10, "LPS-placebo"). In patients with sepsis, fluid resuscitation and vasopressor use was performed at the discretion of the medical team, aiming at normovolemia and a mean arterial pressure (MAP) >65 mmHg, using noradrenalin. The mean flow velocity in the middle cerebral artery (MFVMCA) was measured by transcranial Doppler (TCD) with simultaneously recordeding of heart rate, arterial blood pressure, respiratory rate and oxygen saturation. CrCP was determinded by a cerebrovascular impedance model. Introduction: Right ventricular function (RVF) has been recently reviewed as having a major role in the development of hemodynamic alteration in the critically ill 1 . Measuring RVF is not easy because of RV geometry and limitations in monitoring techniques, nevertheless, single-beat (SB) approaches to measure RV efficiency and its relationship with the pulmonary vasculature has been described using invasive approaches 2 or MRI 3 Methods: Using max pressure extrapolation from tricuspid regurgitation Doppler flow we designed a completely noninvasive approach to measure RV coupling (RVAC). We aimed to test it in a critically ill ventilated patient. To design end-systolic pressure volume relationship (ESPVR), we defined maximal isovolumic RV pressure (RVPmax) as (dp/dt*tau)/π + CVP as previously described 4 . We used tricuspid regurgitation flow to assess dp/dt noninvasively. We then calculated mPAP, an approximation of end systolic pressure 2 , by adding mean tricuspid regurgitation gradient to CVP, and SV using continuity equation as commonly used. Effective pulmonary arterial elastance (RVEa) has been defined as mPAP/SV, right ventricular elastance (RVEes) is the result of the equation (RVPmax-mPAP)/SV 2 . Their ratio is RVAC. We compared SB invasive technique using a pulmonary artery catheter and completely noninvasive echocardiographic method in a ventilated septic shock patient using 10 seriated measurements modulating afterload condition by changing 4 levels of PEEP (Zeep, 5 cmH20, 10 cmH20, 15cmH20) and 10 others by varying preload conditions using passive leg raising (PLR) Results: New SB noninvasive technique performed well under every load condition except with levels of PEEP above 10 Conclusions: In this preliminary first case we describe a new method of measuring RVAC in a completely noninvasive way. We plan further investigations to validate such approach in larger series before assessing the impact of this technique in clinical practice Introduction: Poor right ventricular ejection fraction (RVEF) is an important predictor of mortality in the patient with ischemic cardiomyopathy 1 . Accurate measurement of RVEF foresees precise RV volume assessment as this can be obtained from cardiac MRI. Moreover, RVEF can be estimated using thermodilution (TD) using PAC, but this method can underestimate the correct value under several condition 2 Methods: Using max pressure extrapolation from tricuspid regurgitation Doppler flow, we measured right ventricular elastance (RVEes) and pulmonary artery elastance (RVEa) so designing a new method to measure RVEF.
To calculate RVEF, we used the previously validated formula 3 RVEF = RVEes/(RVEes + RVEa). We then designed end-systolic pressure volume relationship (ESPVR) using maximal isovolumic RV pressure (RVPmax) defined by sine extrapolation of systolic and diastolic portions of RV pressure curve as described 4 , but the RV pressure curve was assumed from tricuspid regurgitation Doppler flow using the Bernoulli's. Effective pulmonary artery elastance (RVEa) was then calculated as mPAP/SV and right ventricular elastance (RVEes) as the result of the equation (RVPmax-mPAP)/SV. mPAP can be calculated by adding mean tricuspid regurgitation gradient to CVP, and SV can be measured via continuity equation. We then tested this method against TD and surrogates of RV function as tissue doppler S' curve and TAPSE at several preload condition using passive leg raising and fluid adminisitration.
Results: New single-beat technique was effective in measuring RVEF in every load condition (Delta increase of EF was 11% after PLR, 5% after subsequent fluid challenge). RVEF absolute values were higher than the ones measured by TD, as expected 2 , but their respective trends correlates well as they are concordant with surrogates measurements. Conclusions: A new method of measuring RVEF independent of RV volumes is here described. Further investigation is needed to assess the real impact of this technique in clinical practice. Introduction: Mobile right heart thrombi are rare finding in patients with pulmonary embolism, described in up to 4% of cases, but are associated with increased mortality. The optimal treatment strategy is unknown. Methods: We performed a retrospective review of patients charts who were admitted to intensive care unit at University medical center Ljubljana with intermediate or high risk pulmonary embolism between January 1, 2008 and September 20, 2016. 315 patients were screened and in 31 cases right heart thrombus was reported on echocardiographic examination. The primary end point was survival to out of hospital discharge and comparison between different treatment strategies. Data were analyzed using one-way ANOVA statistical test. P of 0,05 was considered to be statistically significant.
Overall survival at hospital discharge was 65% (20 patients). All the 31 patients received unfractionated heparin (UFH). 9 patients (29%) received only UFH, 17 patients (55%) received thrombolytic therapy in addition to UFH, 2 patients (6%) had thromboaspiration in addition to UFH and 3 patients (10%) had surgical embolectomy in addition to UFH, survival to hospital discharge was 4 (out of 9; 44%), 12 (out of 17; 71%), 1 (out of 2; 50%) and 3 (out of 3; 100%) respectively. There was no statistically significant difference in survival between groups as determined by one-way ANOVA (F = 1,217, p = 0,323). In three cases the initial treatment strategy was inefficient and they had additional intervention. In two cases thrombolysis did not achieve haemodynamic stabilization so one patient had additional thromboaspiration and the second one had surgical embolectomy. One patient after surgical embolectomy had an additional thromboaspiration. All three patients who had additional interventions survived to hospital discharge. Conclusions: Finding of right heart thrombus in patient with pulmonary embolism is associated with increased mortality. No treatment strategy, UFH alone, UFH + thrombolytic therapy, UHF + thromboaspiration or UHF + surgical embolectomy, was associated with statistically significant survival benefit. It is possible to combine two of treatment strategies if the one attempted first is unsuccessful. Introduction: Autoimmune haemolytic anaemia (AIHA) is a rare disorder in which red blood cells are destroyed by haemolysis by either IgG or IgM auto-antibodies against red cell membrane antigens. We report a case of a 49-year-old woman who had been diagnosed with AIHA since 2 years. The therapy she received didn't include any thromboprophylaxis. During her last admission for exacerbation of haemolysis she developed a massive pulmonary embolism (PE) with cardiovascular collapse. In this case the patient died despite intensive care treatment, appropriate cardiopulmonary resuscitation and fibrinolytic therapy (alteplase). Thereafter we performed a literature search to assess the prevalence of venous thromboembolism (VTE) in patients with AIHA. Methods: PubMed was searched for studies that focus on the prevalence of thromboembolic complications in patients with AIHA. The following key words were used: 'autoimmune haemolytic anaemia' , 'pulmonary embolism' and 'venous thromboembolism'. A total of 32 studies were found. After screening the titles and abstracts only 6 articles were retained. The full text of these articles was analysed.

References
Results: The review of literature showed 3 case reports similar to our case of pulmonary embolism associated with AIHA. Although our search showed a lack of studies concerning this topic, one retrospective review reported an incidence up to 20% of thromboembolic events in patients with AIHA [1]. Two other reviews did not determine an exact percentage but showed that AIHA is an independent risk factor for thromboembolic complications and these patients may need to be considered for thromboprophylaxis [2,3].
Conclusions: By reporting this case we want to raise awareness for thromboembolic events in patients with AIHA. VTE and PE are frequent, serious and potential lethal complications in this patient group. Early recognition and treatment of VTE and venous thromboprophylaxis are advised to prevent disastrous outcome. Further studies are needed in order to determine more precisely the prevalence of these complications in the selected patient group. Introduction: Hemodynamic monitoring plays a pivotal role in the management of acutely ill patients. Many monitoring systems are available but they differ from the degree of invasiveness [1]. Aim of this study was to compare a non invasive hemodynamic monitoring system (ClearSight-Edwards Lifesciences) and a minimally invasive system (Vigileo FloTrac-Edwards Lifesciences) in a population of patients undergone coronary revascularization surgery.
Methods: This prospective study was conducted in a 12 beds Cardiothoracic Surgical ICU in patients with preoperative normal left ventricular function undergone coronary bypass surgery. Exclusion criteria were age <18 and >79 years, arrhythmias, hemodynamic instability requiring vasopressors or mechanical assistance, heart valves disease. In the immediate postoperative period patients were monitored with Vigileo FloTrac System and ClearSight System in the contralateral arm. The following parameters were monitored every hour till patient discharge and after fluid challenges if needed: Systolic, Diastolic and Mean Arterial Pressures (SP, DP, MAP), Stroke Volume (SV), Stroke Volume Variation (SVV), Cardiac Index (CI). Fluid challenges were performer by the administration of crystalloid solution 500 ml in 20 minutes. Analysis of data was performed with Med-Cal Software. The agreement between the devices was described by the Bland-Altmann plot. Results: 15 patients with age ranging from 57 to 76 years were enrolled from June 1st and June 30th 2016. Analysis of the data are reported in Table 17. All the parameters showed a significant difference between the two devices. In fig. 53 and 54: Bland-Altmann plots regarding CI and SVV are reported. All the parameters analyzed and their response to fluid challenge showed agreement even if absolute values were different. Conclusions: The two monitoring systems showed significant agreement. However the mean value of the parameters were significantly different between the two devices.

P114
Echocardiographic measurement of cardiac output through modified subcostal window: consistency analysis L. Colinas 1 , G. Hernández 1       Methods: We included 100 hemodynamically stable patients (age 67 ± 16 y.o., SAPSII 52 ± 19, 54% mechanically ventilated, 16% with atrial fibrillation). Three successive examinations were performed by two different board certified operators, the first and the third examinations by one operator and the second examination by the other operator.
Within each examination, we performed three successive measurements for each variable at end-expiration, without moving the probe. For every echocardiographic variable, we calculated the LSC, either from the series of measurements performed within one examination (for the within-examination precision) or from the series of repeated echocardiographic examinations, each one considering the average of three measurements (for the between-examination precision).
Results: Regarding within-examination reliability, LSC of VTI was 13 ± 11% when examination included one single measurement. It dropped to 8 ± 6% if three measurements were averaged. Regarding between-examination precision, if two different operators performed the two successive examinations, the LSC of VTI, LVEF, E, e' and LVEDA was 14 ± 13%, 10 ± 8%, 12 ± 12%, 24 ± 23% and 15 ± 11%, respectively. If the same operator performed the two successive examinations, the LSC for VTI, LVEF, E, e' and LVEDA was 19 ± 17%, 11 ± 10%, 14 ± 12%, 28 ± 27% and 18 ± 15%, respectively. Conclusions: Within each echocardiographic examination, averaging three measurements is enough for obtaining an acceptable LSC for the VTI measurement. When examinations are repeated, the LSC of echocardiographic measurements is slightly better when the two examinations are performed by the same than by two different operators.

P117
The contribution of (pvco2-paco2)/(cao2-cvo2) ratio level as an Introduction: Blood lactate level in patient with septic shock is dependent on various factors that their interpretation can be difficult. CTA: (PvCO2-PaCO2)/(CaO2-CvO2) superior to 1.4 was described as parameter of anaerobic metabolism's evaluation. The aim of this study is to describe the correlation between fast hemodynamic status 's variation and CTA and if this parameter is related to blood lactate level in patients with septic shock. Methods: We included all patients with septic shock who are intubated, ventilated, sedated, with stable hemodynamic status requiring noradrenalin and with cardiac catheterization. Noradrenalin infusion was reduced in patients who had PAM superior to 85 mmhg under noradrenalin and this to develop a PAM at 65 mmhg. Cardic output was measured with heart ultrasound. Hemoglobin, arterial, venous blood gases were measured before and after noradrenalin modifications.
Results: 30 patients with documented septic shock were included. The mean age of patients was 60 yaers olds [20-82]. Blood lactate levels was 3.5 mmol/l(2-8.5) .23 patients had an elevated CTA at 1.4. The falling of PAM resulted in the rise of CTA from 6% to 130% in 20 cases and the falling of CTA between −40% et −10% in 10 cases. In 26 patients, blood lactate level remained unchanged while 23 patients had a CTA variation from 6% to 130%. The absence of CTA variation is always associated with an absence of lactate level variation. There was no correlation between variation of CTA and variation of cardiac output.
Conclusions: Hemodynamic status's modification in patient with septic shock put under noradrenalin can result in modification of anaerobic metabolism detected with CTA variation and still indetected with lactate level variations. Introduction: Venous-to-arterial carbon dioxide differences (PCO2 V-A Gap) were reported during shock states. An inverse relationship between PCO2 V-A Gap and cardiac output was described, highlighting the importance of blood flow on venous CO2 accumulation. We performed the study to assess the role of veno-arterial CO2 gap (Pco2 V-A) in septic patients and it's correlation with other markers of perfusion and hemodynamics. Methods: We studied 50 patients diagnosed with sepsis and/or septic shock.. Arterial and venous blood gas samples were analyzed at different time intervals (time 0,6 hours,12 hours and 24 hours). Venoarterial Co2 gap (PCo2 V-A) was calculated. Routine labs including lactate and base deficited were reported at the same time intervals. Hemodynamic monitoring was done using non-invasive bioimpedence(ICON). Introduction: The objective of this study was to better characterize the relationship between initial emergency department (ED) lactate and need for ICU-Specific intervention amongst patients with suspected infection. Elevated serum lactate has been shown to be a predictor of mortality in septic populations. To date, however, the relationship between serum lactate and requirement for Intensive Care Unit (ICU)-Specific interventions has not been explored. As need for ICU-Specific intervention is a more proximal and actionable endpoint than all-cause mortality, understanding its relationship to lactate may be useful in assigning patient disposition from the ED. Methods: We performed a retrospective chart review of all patients with a suspected infection presenting to the ED of a single urban tertiary care center between January 2010 to December 2014. Suspected infection was defined as receiving antibiotics and having a culture drawn within 24 hours of ED presentation, consistent with criteria used in the derivation of the Sepsis III definition. ICU interventions were defined as needing intravenous insulin, a central venous catheter, arterial catheter, urgent hemodialysis, pulmonary artery catheterization, intubation, non-invasive ventilation, and/or vasopressors. We excluded those with missing lactate values. The sensitivity and specificity of lactate > 4 for predicting need for ICU-intervention were calculated.
Conclusions: In a population of patients with suspected infection, a lactate > 4 was highly specific in predicting need for ICU level care, therefore suggesting the majority of these patients should be admitted to the ICU. This is the first study to use 'ICU-Specific intervention' as an endpoint for prediction, which we believe to be a more practical endpoint as compared to the traditionally used'all-cause mortality.' Introduction: Increase in lactate levels (hyperlactatemia) is a sensitive marker in early identification of patients who are critically ill. Capillary lactate measurement by handheld lactate devices may allow for rapid determination of test results and extend the possible use in the pre-hospital arena. Thus, the aim of this paper is to study the agreement of handheld lactate analyzers for the measurement of capillary lactate as compared with reference venous blood lactate level assessed using a blood gas analyzer in the Emergency Department (ED). Methods: Two hundred and forty patients triaged as 'urgent' (Category 3 of the five category triage scale), aged 18 or above, who presented to the ED in 2016 were recruited. Venous and capillary blood samples were collected for lactate analysis. Venous lactate levels were measured by blood gas analyzer were used as reference (VL-Ref). Capillary lactate level were measured using two handheld analyzers (Nova StatStrip Xpress Lactate Meter and Lactate Scout + Analyzer) (CL-Nova and CL-Scout+). Venous lactate measurements were also performed using two handheld analyzer (VL-Nova and VL-Scout+). Agreement of handheld lactate analyzers with blood gas analyzer will be determined by using Bland-Altman agreement analysis. Introduction: Assessing the sensitivity of the peripheral chemoreflex (SPCR), we can predict the likelihood of developing respiratory and cardiovascular disorders during the treatment of these patients, during surgery and general anesthesia, to predict the course of the disease and its outcome. At present, there is sufficient evidence that SPCR is often increaseв in chronic heart failure and it is one of the markers of disease progression and even a prognostic marker [1]. However, the existing methods for SPCR assessing are difficult to use in routine clinical practice due to their complexity. Breath-holding test performed well in this regard, in healthy people, and the result of this test is inversely correlated with peripheral receptor sensitivity to carbon dioxide [2]. The aim of the study was to compare the breath-holding test to single-breath carbon dioxide test in the evaluation of the sensitivity of the peripheral chemoreflex in subjects with chronic heart failure.

Methods:
The study involved 42 patients with stable chronic heart failure due to left ventricular systolic dysfunction (56 ± 12 years). In all participants, breath-holding test was performed in the morning before breakfast: voluntary breath-holding duration was assessed three times, with 10 min intervals. After inspiration of a volume equal to 2/3 of the vital lung capacity, the participant was asked to hold their breath and the duration of voluntary apnea was measured from the beginning of the voluntary inspiration until reflex contractions of the diaphragm were noted by palpation. A mean value of the duration of the three samples was calculated. The single-breath carbon dioxide test [3] was performed the next day. The study was approved by the local ethics committee. All subjects provided signed informed consent to both tests. The reported study was funded by RFBR, research project No. 16-34-60147 mol_a_dk.
Results: The average sensitivity of peripheral chemoreflex measured with single-breath carbon dioxide test was 0.31 ± 0.13 L/min/mm Hg., the average breath-holding duration was 41 ± 12 seconds. During the correlation analysis a significant negative correlation between the results of two tests was noted (−0.83, p <0.05).
Conclusions: A breath-holding test reflects the sensitivity of the peripheral chemoreflex defined by the single-breath carbon dioxide test in patients with chronic heart failure.
Introduction: Patients with ischemic heart failure (iHF) have a high risk of neurological complications such as cognitive impairment and stroke. We hypothesized that iHF patients have a higher incidence of impaired dynamic cerebral autoregulation (dCA). Methods: Adult patients with iHF and healthy volunteers were included. Cerebral blood flow velocity (CBFV, transcranial Doppler, middle cerebral artery), end-tidal CO2 (capnography), and arterial blood pressure (Finometer) were continuously recorded supine for five minutes at rest. Autoregulation index (ARI) was estimated from the CBFV step response derived by transfer function analysis using standard template curves. Results: Fifty-two iHF patients and 54 age-, gender-, and BP-matched healthy volunteers were studied. Echocardiogram ejection fraction was 40 (20-45) % in iHF group. iHF patients compared to control subjects had reduced EtCO2 (34.1 ± 3.7 vs. 38.3 ± 4.0 mmHg, p < 0.001) and lower ARI values (5.1 ± 1.6 vs. 5.9 ± 1.0, p = 0.012). ARI < 4, suggestive of impaired CA, was more common in iHF patients (28.8% vs. 7.4%, p = 0.004).
Conclusions: IHF patients are more likely to have impaired dCA in comparison with age-matched controls. The relationship between impaired dCA and neurological complications in iHF patients deserves further investigation. Introduction: To assess the effects of intra-aortic balloon pump (IABP) on cerebral hemodynamics in high-risk patients undergoing cardiac surgery with cardio-pulmonary bypass (CPB). Methods: A prospective, randomized, single-center, pilot study was performed in a surgical ICU of a university teaching hospital. Sixtyseven high-risk patients undergoing coronary artery bypass surgery was included, and were randomized in a 1:1 ratio to surgery with or without IABP. Cerebral blood flow velocity (CBFV, transcranial Doppler) and blood pressure (BP, Finometer or intra-arterial line) were continuously recorded over 5 minutes preoperatively (T1), after 24 h (T2) and 7 days after surgery (T3). Autoregulation index (ARI) was estimated from the CBFV response to a step change in BP derived by transfer function analysis. Diagnosis of delirium was based on the Confusion Assessment Method for ICU. Two cognitive scales were applied before and 6 months after surgery. Results: No significant differences were found in the IABP group in comparison with controls for CBFV or ARI. The incidence of delirium or cognitive decline was similar for both groups.
Conclusions: IABP does not affect cerebral hemodynamic or rates of post-surgical delirium and cognitive decline after cardiac surgery with CPB. These results suggest that IABP per se is unlikely to contribute to the occurrence of early or late neurological complications of cardiac surgery. Introduction: Intra-aortic balloon pump (IABP) is the most common used device in the setting of cardiac surgery. It improves cardiac output, reduces systemic vascular resistance and decreases afterload. In high-risk patients, the prophylactic use in cardiac surgery might reduce postoperative complications, such as cardiogenic shock and myocardial ischemia. However, in the last years, the evidence regarding the prophylactic use of IABP is controversial, mainly based in retrospective data. This study aims to evaluate the role of prophylactic IABP in reducing complications in high-risk patients undergoing cardiac surgery. Methods: A prospective randomized controlled trial that evaluated 181 patients undergoing coronary artery bypass at the Heart Institute of the University of Sao Paulo from 2014 April to 2016 June to receive or not intraaortic balloon pump after anesthesia induction just before skin incision. Inclusion criteria were left ventricular ejection fraction (LVEF) < = 40% and/or EuroSCORE > = 6. Eligible patients were randomly assigned, in a 1:1 ratio, to IABP group (n = 90) or control group (n = 91). The primary outcome was the composite endpoint of mortality and major morbidity in 30 days after cardiac surgery (cardiogenic shock, stroke, acute renal failure, mediastinitis, prolonged mechanical ventilation and need for reoperation. Introduction: Critically ill cardiac patients consist a heterogeneous population, but usually are congestive. Among the variety of methods used to evaluate volemic status in Intensive Cardiac Care Unit (ICCU) none of them has been checked in a prospective study.
Results: During mean 6 ± 3 days of hospitalization 7 (32%) patients died. Admission CVP was 12.1 ± 5.9 mmHg and not significantly lower values in the group of deaths as compared to survivors were found (11.0 ± 4.8 vs. 12.6 ± 6.5 mmHg, p = 0.6). Among patients who died significantly lower central venous oxygen saturation, higher high-sensitivity troponin I, higher TAPSE and higher iv fluids infusion with significantly negative fluid balance within the first 24 hours were found (p < 0. Patients who died instead of higher iv fluid therapy within the first 24 hours had negative fluid balance which, together with higher TAPSE, might suggest paradoxical dehydratation. At the same time false interpretation of CVP encouraged clinicians to the incorrect use of high doses diuretics. In this aspect, among other assessed parameters, TAPSE seems to be a better tool to guide the balance between the dose of diuretics and iv fluid therapy in critically ill non-septic cardiac patients. Results: In 10 patients ("responders"), CI increased > =15% (22 ± 8%) during volume expansion. The IAP at baseline was 21 ± 4 mmHg in responders and 17 ± 1 mmHg in non-responders (p = 0.10). The PLR test increased CI in responders (11 ± 16%) but not in non-responders. The PLR test was positive (PLR-induced increase in CI > =10%) in 3 responders (true positives) and negative on 7 responders (false negatives). The PLR test was negative in 2 non-responders (true negatives) and positive in 1 non-responders (false positives). Thus, the sensitivity of the PLR test was 30% (95% confidence interval: 7-65%) and the specificity was 67(9-99)%. Introduction: For the moment, a reliable assessment of the hemodynamic effects of volume expansion and passive leg raising (PLR) requires a direct measurement of cardiac output, which is often invasive. However, it has been suggested that changes in common carotid and common femoral artery blood flows and in their velocities may reflect the variations in cardiac output. We tested whether these changes measured by Doppler could accurately follow the changes in cardiac output during PLR and volume expansion. Methods: In 34 ICU patients, before and during PLR and before and after volume expansion, we measured cardiac index (pulse contour analysis, PiCCO2) and carotid and femoral blood flows, which were calculated from arterial diameter and velocity time integrals (for femoral arteries) or time average velocities (for carotid arteries). Peak systolic velocity was also recorded. We performed 53 PLR tests. Volume expansion was then performed in 23 cases where cardiac index increased > = 10% during PLR. Results: Taking into consideration measurements performed during PLR and volume expansions (n = 76), cardiac index increased by 14,6 ± 13,7%, carotid blood flow by 15,3 ± 35,4% and femoral blood flow by 21,1 ± 36%. The correlation between changes in cardiac index and in carotid blood flow and between changes in cardiac index and in femoral blood flow were not significant (p = 0.60 and p = 0.69, respectively). In 27 patients, PLR was positive (increase in cardiac index > = 10%). During PLR in these patients, cardiac index increased by 19 ± 11%, carotid blood flow increased by 13 ± 38% and femoral flow increased by 3 ± 12%. Neither the changes in carotid blood flow (area under the ROC curve: 0.55 ± 0.10) nor the changes in femoral flow (area under the receiver operating characteristic (ROC) curve: 0.57 ± 0.16) could detect a positive PLR test. The changes in peak systolic velocity during PLR also failed to detect a positive PLR test, for carotid as for femoral blood flows.

P128
Conclusions: The assessment of carotid and femoral artery blood flow as well as the measurements of their peak systolic velocity are not reliable methods to assess a PLR test in ICU patients.

P133
Could bioreactance assess the effects of passive leg raising in critically ill patients?  Introduction: The effects of end-expiratory occlusion on cardiac output were shown to reliably predict fluid responsiveness in patients under mechanical ventilation. Nevertheless the threshold value of 5% which was found requires that the method of measurement of cardiac output is of high precision, what is doubtful when echocardiography is used. We aimed at assessing whether 1) fluid responsiveness can be predicted by the effects of endexpiratory occlusion on the velocity-time integral (VTI) of the left ventricular outflow measured by echocardiography and 2) adding the effects of end-inspiratory to those of end-expiratory occlusions on VTI can predict fluid responsiveness with the same reliability as end-expiratory occlusion alone but with a higher threshold value, more compatible with the precision of echocardiography, which is assumed to be close to 5%. Methods: In 30 patients, we measured pulse contour analysis-derived cardiac index (PiCCO2) and VTI during the 5 last seconds of 15second end-inspiratory and end-expiratory occlusions, separated by 1 minute, and after infusing 500-mL saline. Patients where volume expansion increased cardiac index (transpulmonary thermodilution) > = 15% were defined as "responders". Results: End-expiratory occlusion increased VTI more in responders than in non-responders (11 ± 5% vs. 3 ± 1%, respectively, p < 0.0001) and end-inspiratory occlusion decreased VTI more in responders than in non-responders (12 ± 5% vs. 5 ± 2%, respectively, p = 0.0002 Results: Two hundred and ten ARDS patients were included in our study. The majority of their ARDS severities were moderate to severe. Overall, mortality at 28th day was 67%, which was associated with age, APACHE II, PaCO2, cumulative fluid balance in the first 3 days and the first 7 days. After multivariate analysis, only APACHE II and the first 7-day cumulative fluid balance were the independent predictors for 28-day mortality. Moreover, the risk of death increased with increasing 7-day fluid balance (Table 19).

Conclusions:
The study demonstrated the more positive cumulative fluid balance during the first week of ARDS, the higher 28day mortality. The first 7-day cumulative fluid balance threshold for increased risk of 28-day mortality in ARDS patients was −1,000 ml. However, further studies are needed to confirm these results.   Introduction: It is well known that fluids are vital to patient management after cardiac surgery. The question is, how much do you give? Some of the latest research suggests that that the previously held belief of "the more fluid, the better" might not hold true. 1,2 .
Methods: To help answer our question of how much, we carried out a retrospective observational analysis of patients admitted to an intensive care unit at Papworth Hospital, a cardiothoracic centre, between 01/01/2014 and 31/12/15. After extracting data from an electronic database, patients were divided into four quartiles, determined by daily IV fluid administration (FA). Parameters analysed included mortality, CAM-ICU status, alkaline phosphatase score and bilirubin score. These two scores were defined by the number of times they were higher than the upper limit of normal for the parameter, using the highest value during the patient's stay. Each quartile was compared to Quartile 1 (Q1) for a variety of different adverse outcomes, using relative risk (RR) and 95% confidence intervals (CI).
Results: Although 5519 patients were found at first, only 4222 qualified for the study as 945 patient records contained missing data. Each quartile was therefore composed of 1055 or 1056 individuals. A statistically significant relationship was found between higher FA and unfavourable patient outcomes in patients in Q4 compared to Q1 for mortality (RR = 109.1 (CI 6.7-1764.3)), CAM-ICU status (RR = 7.4 (CI 5.6-9.6)), an alkaline phosphatase score >3 (11.1 (CI 2.6-46.7)) and a bilirubin score >3 (RR = 6.0 (CI 3.0-12.1)). The results found here are similar to those found in other studies 1,2 .
Conclusions: The correlations found here between increased FA and mortality and morbidity are significant, suggesting fluid overload can have serious negative consequences. However, further statistical analysis is needed to determine whether these are independent associations, and if so, this may then support a more restrictive approach.. Introduction: Fluid resuscitation is central to the management of septic shock [1]. The topic has attracted much attention over the last decade with a raft of contradictory results. In 2011 a study published in the NEJM was stopped early because aggressive fluid resuscitation in children with sepsis had significantly worse outcomes [2]. This led many to question whether fluid is actually a good thing and focus has shifted to restrictive fluid regimes. At our institution we are considering the use of peripheral noradrenaline to enable earlier vasopressor therapy in septic shock. We first wanted to quantify how much fluid was being administered to those with septic shock before noradrenaline was started. This was the period after "adequate fluid resuscitation" in areas outside the ICU and before noradrenaline was started. We hypothesised that delay in placement of central venous catheter (CVC) may delay vasopressor administration and potentially increase unnecessary fluid administration.
Methods: Our Trust is a DGH consisting of 19 mixed surgical and medical critical care beds across two separate sites. We retrospectively analysed all ICU admissions with ICNARC coding of "sepsis" between April 2015 and April 2016. Diagnosis of septic shock was also confirmed by interrogation of the notes and requirement for vasopressor support after adequate fluid resuscitation outside the ICU. Patients taken directly from theatre and re-patriated were excluded. We looked at the time taken to place CVC, start noradrenaline and the amount of fluid administered on ICU before noradrenaline was started.
Results: 79 patient were identified as having septic shock and were included in the analysis. The mean time taken to place a CVC was 2 hours 48 minutes and mean time to commencing noradrenaline was 4 hours 53 minutes. The mean volume of fluid administered prior to starting vasopressors was 1174 ml. This did not include fluid resuscitation on the ward. Conclusions: Our results show that large amounts of fluid are being administered to patients with septic shock on admission to ICU before vasopressor initiation. It appears from the data that placement of a CVC represents a barrier to the timely initiation of vasopressors. Peripheral administration may represent a means of overcoming this delay.

P145
The Introduction: Fasting, anaesthesia and surgery affect the body's physiological capacity not only to control its external fluid and electrolyte balance but also the internal balance between the various body fluid compartments. Conversely, abnormalities of fluid and electrolyte balance may adversely affect organ function and surgical outcome. The aim of the study -to determine association between fluid administration during the perioperative period and complication development after major abdominal operations. Methods: A retrospective study of the perioperative period after major abdominal operations in 560 patients was performed. The physical condition of patients corresponded to 3 class of ASA. The median age was 46.0 (38,0-62,0) years. The duration of the operations was more than 180 minutes. All patients received standart fluid management according to the ICU rules and were divided into 2 groups according to complication development during postoperative period: 1 -with complicated postoperative period (n = 169), without complications (n = 391 . Emotional triggers were more common, than physical in tacotsubo patients (50% vs 35%, p < 0.05), in 15% no evident triggers were found. In 28% patients, tacotsubo cardiomopathy caused acute heart failure in early postoperative period, which was significantly higher, then in MI group. In 50% of patients with tacotsubo psychiatric abnormalities (severe dementia, delirium or schizophrenia) were found, which was significantly higher when compared with patients with MI (13%). Introduction: Upper abdominal operations demand higher doses of opioids. The aim of the study was to investigate whether the ulrashort-acting opioid remifentanil provide better cardiovascular and humoral response control than fentanyl during and after total gastrectomy in elderly parient.
Methods: A total of 60 patients aged over 65 years (ASA II calssification) were randomised in two equal groups to receive remifentanil (0.1 μ g/kg continuous iv infusion) or fentanyl (1.5 μ g/kg iv bolus), as analgesic part of general balanced anaesthesia and continued postoperatively. The values of cardiovascular parametersblood pressure and heart rate and level of cortisol, glicemia was measured at six points in time: baseline value-before inducion in anaesthesia (0), after intduction (1), two minutes after intubation (2), after manipulation with abdominal organs (3), after extubation (4), after 12 h and 24 h postoperatively (5,6). Results: After induction of anaesthesia, a greater decrease of systolic pressure was found in the remifentanil group (32% ± 5 vs. 25 ± 4%). However, in the same group after the intubation of trachea a smaller increase of systolic pressure was registered (16% vs. 33%). Systolic blood pressure was less incressed during the phase of extubation in fentanyl group (24%). Diastolic pressure was significantly higher during the phase of extubation in fentanyl group. After induction in anaesthesia, a significant fall in diastolic blood pressure was found in remifentanil group Levels of cortisol were increased after extubation, but significantly more in the remifentanil group. Level was significantly increased after 12 h of the operation in the both groups. After 24 h was increased but not significantly. Glycaemia was without changes after intubation. After first incision and during procedure. Glucose level was significantly increased in phase of tracheal extubation only in remifentanil group. During the postoperative period was within the normal values in both groups. Conclusions: Fentanyl assured better cardiovascular and humoral stability than remifentanil in elderly gastrectomised patients. Introduction: The purpose of this study was to compare clevidipine (CLEV) vs sodium nitroprusside (SNP) as adjunct agents to esmolol (ESM) for blood pressure (BP) management in aortic dissection. Intravenous (IV) vasodilators are commonly added to beta-blocking agents to reach BP goals in aortic dissection [1]. Our institution has transitioned to adding CLEV instead of SNP to standard therapy ESM for initial BP management. CLEV has been proven efficacious in lowering BP in both pre-and post-operative cardiac procedures and hypertensive emergencies [2,3]. To our knowledge, this is the first study to evaluate CLEV use in acute aortic dissection.
Methods: A single-center retrospective chart review evaluated all patients diagnosed with aortic dissection from September 2010 through September 2016. Included patients were over 18 years old with new diagnosis of aortic dissection at presentation, were initiated on ESM as their primary beta blocker, received CLEV or SNP adjunct therapy and had complete hemodynamic data. Excluded patients were initiated on IV beta-blockers other than ESM, received other concomitant IV anti-hypertensive therapies, did not reach primary endpoint prior to surgical management, or were pregnant or breastfeeding. The primary endpoint was defined as time to reach patient specific systolic blood pressure (SBP) goals after ESM initiation. The secondary endpoint was defined as efficacy of CLEV and SNP for maintaining BP within patient specific goals using area under the curve (AUC) analysis of both positive and negative excursions until 24 hours post agent initiation or time of surgical management, whichever was less. Statistical analyses were conducted using R statistical software. Introduction: Lung ischemia-reperfusion injury after thoracoabdominal aortic occlusion represents a major complication, which increases morbidity and mortality. In the present study we hypothesized that lazaroid U-74389G intravenous administration protects from lung ischemia-reperfusion injury through lipid peroxidation inhibition. Methods: A total of 24 pigs were randomized in three groups. Group I (n = 8) underwent sham operation, group II (n = 8) underwent thoracoabdominal aortic occlusion for 45 min and received placebo and group III (n = 8) received 3 doses of lazaroid (3 mg/kg) 60 and 30 min before thoracoabdominal aortic occlusion and at 30 min of thoracoabdominal aortic occlusion (duration 45 min). Aortic occlusion was performed with aortic balloon-catheters under fluoroscopic guidance. All animals were sacrificed at the 7th postoperative day and lung specimens were received for molecular analysis. Results: mRNA levels of leukotrienes LB4, LC4 and nitric oxide synthase isoforms including eNOS, nNOS and iNOS were determined with real-time RT-PCR. Nitric oxide can either induce (iNOS) or inhibit (eNOS and iNOS) lipid peroxidation based on its specific isoform origin. Group III showed significantly reduced levels of both LB4 (−63.7%) and LC4 (−35.9%) when compared with group II (P < 0.05). Isoform nNOS was not detected in lung specimens of all three groups. iNOS was significantly reduced (−60.2%) in Group III when compared with group II (P < 0.05). Finally, eNOS was slightly increased (+2.1%) in group III when compared with group II (P = 0.467). Conclusions: Lazaroid U-74389G may represent an effective pharmacologic intervention in reducing lung ischemia-reperfusion injury following thoracoabdominal aortic occlusion. Introduction: Nitric oxide may be used in many medical conditions, such as primary pulmonary hypertension and acute pulmonary embolism. In the setting of cardiac surgery, its role is unknown. The aim of this study was to to investigate the efficacy and safety of perioperative administration of nitric oxide in cardiac surgery. Methods: The authors conducted a systematic review of randomized, controlled, parallel-group trials in accordance with a previously registered protocol (International Prospective Register of Systematic Reviews registration no. PROSPERO 2016:CRD42016032702), from inception to March 2016. Primary outcome was intensive care unit (ICU) stay, secondary outcomes were mortality, duration of mechanical ventilation, reduction of mean pulmonary artery pressure. Results: The study included 17 RCTs comprising 760 patients. We calculated the pooled odds ratio (OR) and the mean difference (MD) with Random-Effects model. Quantitative synthesis of data demonstrated a significant reduction in the length of ICU stay (mean difference −0.33 days, CI [−0.59, −0.07] p = 0.01) and of mechanical ventilation duration (MD −4.88 hours, CI [−8.07, −1.69] p = 0.003) when compared to controls with no differences in mortality. Conclusions: Nitric oxide might be beneficial in patients with pulmonary hypertension undergoing cardiac surgery. Large randomized trials are needed in order to further assess its effect on survival.

P155
Preoperative beta-blocker & vasodilator combined with neuraxial block increase risk of perioperative cardiac complication in patients admitting to the general surgical ICU S Kongsayreepong, R Sungsiri, P Wongsripunetit Siriraj Hospital, Mahidol University, Bangkok, Thailand Critical Care 2017, 21(Suppl 1):P155 Introduction: Increase serious cardiac complications in patient receiving preoperative high dose beta-blocker and neuraxial block1. Anyhow with the perioperative benefit of neuraxial block especially the combined thoracic epidural and general anesthesia that could reduce stress & decrease ventilator hours including ICU length of stay. So questions were asked which could be the causes of these serious cardiac complication associated with neuraxial block. So the aim of this study was to study factors associated with perioperative cardiac morbidity in sick surgical patient receiving neuraxial block admitting to the general surgical ICU.
Methods: This was a part of prospective ICU database of general surgical ICU of Siriraj Hospital, Mahidol Univeristy. The data from 80 surgical patients receiving neuraxial block admitting to this general surgical ICU postoperative & stay in ICU > 24 hours. Studied information included: age, sex, BW, BMI, co-morbidity (stroke, TIA, coronary artery disease, NewYork Heart Classification, arterial vascular disease, DM), ASA classification, preoperative medication (beta-adrenergic blocking agent, vasodilator, ACEI), neuraxial block technique, preoperative hydration, type & duration of anesthesia/surgery, ventilator hours & ICU length of stay. Perioperative cardiac complications included intraoperative hypotension (MAP < 65 mmHg > 25 mins that needed fluid resuscitation and/or vasopressor), early perioperative myocardial ischemia/infarction, early perioperative arrhythmias, cardiac death. Neuraxial block in this study included: spinal, combined lumbar or thoracic epidural anesthesia. Results: The result of this study showed that perioperative cardiac complications that associated with neuraxial block significant associated with preoperative beta-blocking agent, vasodilator, hypovolemia, high dose local anesthetic agent, high level thoracic epidural combined with general anesthesia, not using continuous technique, poor cardiac function (EF < 40%) (p < 0.05). As the sample size was not enough to do the multivariate analysis, so this part of the result was not be reported. Conclusions: Despite the benefit of perioperative use of neuraxial block. Caution should be taken into consideration when using this neuraxial block in combind with perioperative beta-adrenergic blocking agent, vasodilator or in dehydration condition to prevent serious cardiac complications associated with this neuraxial block.
Conclusions: Levamisole exerts direct effects on vascular tone in the carotid artery, enhancing the contractile response to ET-1 through the activation of ETA receptors. This effect is independent of cocaine actions, suggesting that levamisole could potentiate the deleterious effects of cocaine on vascular tone.

P157
Effect of volume vs. vasoconstriction on regional oxygen saturation during hemodynamic management UB Borg, AM Neitenbach Medtronic, Boulder, CO, United States Critical Care 2017, 21(Suppl 1):P157 Introduction: The objective of this animal study was to determine effects of volume infusion versus vasoconstriction on regional oxygen saturation. Near infrared technology (NIRS) is used in cardiac surgery to monitor cerebral oxygen saturation (CrSO2) as an indicator of adequate perfusion. The most common intervention in reaction to decreasing CrSO2 is to increase mean arterial blood pressure (MAP). The question is whether volume or vasoactive drugs is the correct path. The chosen path may influence the incidence of organ insufficiency such as acute kidney injury. Organ perfusion, other than brain is not easily assessed with INVOS in adults, and may be compromised Methods: In a porcine study, approved by the local animal use committee, we investigated the effect on CrSO2 and perirenal rSO2 (RrSO2) (INVOS™ 5100C, Medtronic, Boulder, USA) of restoring MAP with volume or vasoconstriction. Seventeen animals were put under general anesthesia, intubated and ventilated to normocapnia. MAP, heart rate, cardiac output CrSO2 and RrSO2 were continuously monitored. Hypovolemic state was achieved by removing 50% of the animals calculated blood volume. The shed blood or vasoconstriction using norepinephrine (NE) was used to return CrSO2 to baseline values. The shed blood was returned to the animal in steps of 200 ml/ step and NE infusion was stepwise increased until CrSO2 was returned to baseline. Results: Table 21 show mean values and SD for baseline, hypovolemia, effect of blood infusion or NE. Both blood infusion and NE returned the CrSO2 to baseline. Only blood infusion returned RrSO2 to baseline and there was a significant difference (p < 0.0002) between RrSO2 blood vs. NE even though a significantly (p < 0.0001) higher MAP was restored. A significantly (p < 0.0001) higher MAP was required to reach baseline CrSO2 with NE. Conclusions: NIRS monitoring is often used in an attempt to monitor and prevent oxygen desaturation of the brain. In this study we demonstrated that restoring CrSO2 may result in unintended consequences for other organs. When CrSO2 was restored using NE the RrSO2 showed no improvement thus possibly indicating that the renal perfusion may have been compromised despite "normal" CrSO2 and MAP. The clinical implications of these results should be elucidated by clinical studies in patients.

P158
Dynamic arterial elastance reflects pressure-flow uncoupling in an experimental endotoxic septic shock M García 1 , P Guijo González 2 , M Gracia Romero 2 , P Saludes Orduña 2 , A Gil Cano 2 , A Rhodes 1 , RM Grounds 1 , M Cecconi 1 1 St. George's Healthcare NHS Trust and St George's University of London, London, United Kingdom, 2 Hospital SAS de Jerez, Jerez de la Frontera, Spain Critical Care 2017, 21(Suppl 1):P158 Introduction: Dynamic arterial elastance (Eadyn), the ratio between pulse pressure variation (PPV) and stroke volume variation (SVV), has been suggested as a functional parameter of arterial load. We aimed to determine the effects of an endotoxic septic shock on Eadyn and its components (PPV and SVV), and the impact of hemodynamic resuscitation, considering both macro (arterial and cardiac factors) and microcirculatory influences, and the relationship with cardiac energetics. Methods: 18 New Zealand rabbits. Animals received placebo (SHAM, n = 6) or lipopolysaccharide (LPS) with or without (EDX-R, n = 6; EDX-NR, n = 6) hemodynamic resuscitation (fluids and norepinephrine). Animals were monitored with an indwelling arterial catheter and an esophageal Doppler. The arterial load was evaluated by a 3-element Windkessel model. Cardiac influence on Eadyn was assessed by heart rate and the profile of the aortic blood flow. External work generated by the heart is the sum of the pressure energy (PE) and kinetic energy (KE), and represents the driving force for the circulation. Tissue oxygen saturation (StO2) was assessed using near-infrared spectroscopy Results: LPS infusion resulted in a hyperdynamic profile with an increase in CO and a reduction in blood pressure and arterial load. LPS administration led to a sustained decrease in StO2, while resuscitation produced a modest increase. Eadyn, PPV and SVV increased in both EDX and EDX-NR groups, but not in SHAM animals (Fig. 64). Even if PPV and SVV were related to arterial and cardiac factors, only blood flow pattern (velocity and ejection time) were associated with Eadyn. Cardiac energetics were associated with Eadyn. So, the higher the percentage of work performed by the heart to accelerate blood flow (KE), or the lower the ability to produce PE, the higher the Eadyn. Conclusions: In this experimental model, Eadyn reflected the pressure-flow impairment during endotoxic septic shock. Hemodynamic resuscitation restored this relationship but a higher cardiac energetic cost. Eadyn could be a potential index of ventriculo-arterial coupling. Introduction: Patients in surgical settings are often at risk of developing hypotension, which can lead to poor outcomes such as acute kidney and myocardial injury [1]. We have developed a hypotension  probability indicator (P(↓BP)) for real-time prediction of intraoperative hypotension using features of the arterial pressure waveform. We assessed the performance of P(↓BP) on independent data being collected at UCI Medical Center (UCIMC). Methods: We studied patients undergoing surgery for a wide range of procedures, including spinal, abdominal, and cardiac. 993 patients consented to this study, and 155 patients with arterial lines were analyzed to assess the accuracy of P(↓BP). A hypotensive event was defined as any time period where MAP < 65 mmHg for > = 1 minute. An ROC analysis was performed to assess AUC of P(↓BP) to identify and predict an event. Time to event, P(↓BP) and MAP values when P(↓BP) > 51 prior to event were also assessed. Clinical records were reviewed for any vasopressor, inotrope, crystalloid, colloid, or blood product interventions within 10 minutes before or after the start of an event.
Results: 123 of the 155 patients had at least one hypotensive event, totaling 12,212 hypotensive events. P(↓BP) accurately detected an event up to 15 minutes prior to event start (Fig. 65). In addition, using a threshold of 51, P (↓BP) warned of hypotension~20 minutes prior ( Introduction: Hypotension often represents the first sign of an acute decompensated cardiovascular system that precedes organ hypoperfusion. Thus, predicting arterial hypotension could be a relevant aspect in hemodynamic management of critically-ill patients. In this regard, a new hemodynamic parameter based on the analysis of the arterial pressure waveform features has been proposed to predict hypotensive episodes: hypotension probability indicator (HPI). The aim of this study was to report the preliminary results of the application in real clinical practice of the performance of the HPI™ for predicting arterial hypotension. Methods: Patients had an indwelling radial arterial catheter connected to a FloTracIQ™ sensor (Edwards Lifesciences, Irvine, CA). A hypotensive episode was defined as MAP < 65 mmHg for at least 1 minute. ROC curve analysis was performed to determine sensitivity/specificity and the area under the ROC curve of the HPI to predict a hypotensive event 5 minutes prior to the start of event.
Conclusions: Although further confirmation with a larger sample size is required, our preliminary results showed that HPI predicted hypotension with high reliability in real clinical practice and could be a valuable information during hemodynamic management of critically-ill patients.   Introduction: Hypotension occurs frequently in intraoperative and critical care settings and is associated with an increased incidence of complications. We have recently developed a hypotension probability indicator (HPI™) to predict hypotension based on machine learning techniques. The objective here is to test the accuracy of HPI in predicting artificially induced hypotension in a porcine model. Methods: With IRB approval, hypotension is induced in animals both by hemorrhage and by vasodilation using the protocol described in fig. 68. Radial arterial pressure waveforms were monitored invasively by FloTrac (Edwards Lifesciences) to calculate HPI. ROC analysis was performed to assess the performance of HPI.
Results: 5 pigs were studied. 12 hypotensive episodes occurred: 6 by hemorrhage and 6 by vasodilation. Typical examples of animal's hemodynamic variables are shown in fig. 69: HPI increases with the introduction of hypotension, indicating there is higher probability for hypotension to occur. ROC analysis shows HPI can predict hypotension with a sensitivity and specificity of 100% and 100%, respectively, at 5 minutes prior to hypotensive episodes.
The area under the curve is 1. Changes can be found in stroke volume variation as a preload measure, dynamic arterial elastance as a functional measure of afterload, and dP/dt as a measure of LV contractility. Conclusions: Our study suggests that HPI can predict both hemorrhage and vasodilation induced hypotension 5 minutes before its occurrence with high sensitivity and specificity. Information about preload, contractility or afterload may help to take appropriate therapeutic measures.  70).
Conclusions: Those data demonstrate that hypotension occurs quite often in ICU patients, and HPI can predict hypotension with high sensitivity and specificity. HPI may serve as a useful addition in the care of critically ill patients with the potential to reduce the occurrence of hypotension.  Introduction: The radial artery is often catheterized in the intensive care unit (ICU), allowing for direct measurement of local blood pressure. Along with ultrasonic estimates of arterial displacements, such measurements can be used to estimate local tissue stiffness and its changes in different conditions. However, because of the presence of the catheter, ultrasound and pressure measurements cannot be made on the same side.
Methods: Image sequences of left and right radial arteries of 5 normotensive subjects (38.5 ± 3.10 yo) were acquired at~30Hz using B-mode ultrasound and a linear array transducer. Arterial diameters were derived from displacement waveforms produced from image sequences using adaptive block matching. Diastolic and systolic blood pressures were measured at the brachial artery using sphygmomanometry. Tissue properties were quantified through radial strain, distensibility and compliance coefficients and Peterson's elastic modulus (Table 23). Results: Figure 71 shows an ultrasound image of the radial artery and its diameter waveform. Table 23 shows average ± std values for the estimated parameters. Tissue stiffness of the left side was similar to that of the right, based on a Wilcoxon rank sum test (p-value > 0.05). The observed large variabilities were attributed to intersubject anatomical differences reflected in different diameters and blood pressures.
Conclusions: These preliminary findings suggest that blood pressure measurements in one radial artery can be used to reliably derive stiffness indices in the contralateral side. Introduction: Non-invasive measurement of hemodynamic parameters and microvascular reactivity are increasingly reported and recommended for assessment of critically ill patients [1][2][3][4]. In this proof of concept study, objectives were 1) assess non-invasive measurement of peripheral arterial flow velocity in the Dorsalis Pedis artery using a novel ultrasound sensor [5], 2) analyze velocity waveforms and 3) assess response under a variety of control and pathological conditions including 1) resting control state, 2) Valsalva maneuver, 3) hyperemic response and 4) in active sepsis with norepinephrine treatment. Methods: In age matched male subjects in the basal position (135 degrees), the sensor was positioned above the left Dorsalis Pedis artery. Results: Figure 72 shows velocity waveforms from control and septic patient. Waveforms depict forward flow velocity (max peak), reverse flow velocity (min peak) and additional reflected waves, originating from "blind ends" or arterioles and capillaries of microvascular beds. Note the increased variability in max velocity in sepsis (range from 32.4 to 54.7 cm/s) compared to control (range from 46.9 to 50.78 cm/s)) and prolonged stasis (40% of time) between 2nd and 3rd max peaks. Figure 73 reveals norepinephrine had a rapid effect on arterial flow velocity. As norepinephrine increased from 25 to 35 mcg/min there was a rapid decrease in max velocity over 15 minutes. Increasing the dose to 45 mcg/min resulted in a further decrease in arterial velocity over 10 minutes. This was associated with increased MAP. In control experiments (see poster) transient responses in max flow velocity in the Dorsalis Pedis artery were detected in response to 20s Valsalva maneuver, 50s vascular occlusion and to passive leg raising.
Conclusions: Non-invasive peripheral arterial flow velocity, measured in the Dorsalis Pedis artery using an ultrasound sensor, is sensitive to a variety of physiological transient responses and to the effects of norepinephrine in sepsis. Taken together, peripheral arterial flow velocity is an additional hemodynamic parameter which may have broader application in terms of patient monitoring, fluid resuscitation and vasopressor evaluation, inter alia, though further validation is required.   Introduction: Vasopressin (VP) has an important role in restoring blood pressure in septic shock patients. The aim of this study was to determine the outcome of patients with septic shock needing vasopressin to achive a MAP of 65 mm˙Hg despite NOR use. Methods: Retrospective, observational, cohort study. The variables collected were: age, gender, SAPS 3, SOFA at admission, days on MV, need of RRT, LOS and, ICU mortality. Data were presented as percentages for cathegorical variables and means for continuous variables. Comparisons were made with the use of Student's T test, Mann-Whitney U test, or Pearson Chi-square test as appropriate. To estimate the effects of vasopressin use on mortality, we used binary logistic regression with SAPS3, SOFA, and need for RRT as covariates. All statistical analysis were made using SPSS vs 20.0. Results: There were 163 consecutive patients treated for septic shock, between January 2015 and August 2016, divided in group 1 (NOR plus VP; n = 54) and group 2 (NOR only; n = 109). The mean age (yr), male gender (%), days on MV, RRT (%) and LOS ICU was 58.8 and 62.00(NS), 61.1 and 52.3 (NS), 8.1 and 9.3 (NS), 24.1 and 11.9 (p < 0.001), 9.6 and 11.9 (NS), respectively. Severity and organ dysfunction score and, ICU mortality are in Table 24. After stepwise binary logistic regression modeling, vasopressin use was still significantly associated with ICU mortality, with an OR of 5,34 (2,41-11,87). Conclusions: Patients needing vasopressin to sustain a MAP at least 65 mmHg had worst severity scores at admission, more severe organ dysfunction and needed more RRT during ICU stay. VP rescue therapy during septic shock had a high mortality in comparison to those who sustain MAP only with NOR even when adjusting for severity of disease.  Introduction: High-dose insulin (HDI) has hemodynamic and metabolic properties in several myocardial dysfunction scenarios. This study was carried out to evaluate if HDI infusion would improve hemodynamic, metabolic and inflammatory parameters in a clinically relevant septic shock model treated with fluid resuscitation and antibiotics.
Methods: Sixteen pigs (35-45 kg) were anesthetized, monitored with pressure-conductance catheters and submitted to peritonitis by fecal inoculation (1 g/kg). In the insulin group, dextrose (50%) was infused to maintain glucose in the 60-150 mg/dL range, and potassium to maintain a level greater than 2.8 mmol/L. After persistent hypotension, animals received antibiotics, fluids and vasopressors according to a predefined algorithm and were randomized to an insulin (HDI) (n = 9) or control (n = 7) group. The HDI group received an insulin infusion of 3 units/kg/hour during six hours. Inflammatory response was evaluated by plasma, myocardial and pulmonary concentration of cytokines (IL-6 and IL-8) by ELISA. Outcomes from the repeated-measures analysis were modeled using a mixed-effects linear model. Submit your next manuscript to BioMed Central and we will help you at every step: