Tenth International Symposium on Intensive Care and Emergency Medicine for Latin America

P01 KIR genes 2DS1, 3DS1, and 2DL5 are associated with lower organ dysfunction in a cohort of critical care patients without sepsis Luciana Mello de Oliveira, Jaqueline Beppler, Juliana Lindenau, Pamela Portela, Fernando Suparregui Dias, Clarice Sampaio Alho, Felipe DalPizzol, Luiz Fernando Jobim, Mariana Jobim, Rafael Roesler UFRGS – Universidade Federal do Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil; HCPA – Porto Alegre Clinical Hospital, Porto Alegre, Rio Grande do Sul, Brazil; HP – Pompéia Hospital, Caxias do Sul, Rio Grande do Sul, Brazil; PUCRS –Pontifical Catholic University of Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil; SJH – São José Hospital, Criciúma, Santa Catarina, Brazil Critical Care 2019, 23(Suppl 3):P01


Background
Natural killer cells are part of the innate immune system and their activity is regulated by activating and inhibitory signals transduced by killer cell immunoglobulin-like receptors (KIR). Seventeen KIR receptors have been characterized in humans: eight inhibitory types (2DL1-3, 2DL5A and B, and 3DL1-3), seven activating types (2DS1-5, 3DS1, and 2DL4), and two pseudogenes that do not encode a functional KIR receptor. We previously reported a reduced frequency of activating genes in patients with sepsis [1]. Objective We aimed to evaluate KIR genes in 60 critically ill patient without sepsis and their association with each system or organ dysfunction represented by the Sequential Organ Failure Assessment (SOFA) score. Methods Subjects were southern Brazilian residents, with a majority of subjects having European ethnicity and a minor number of individuals with African genetic traits, with an age range from 14 to 94 years admitted to the Intensive Care Unit, São Lucas Hospital, Pontifical Catholic University of Rio Grande do Sul, in Porto Alegre, Brazil, between 2004 and 2010. DNA samples of 60 critical care patients were genotyped using the polymerase chain reaction method with sequence-specific oligonucleotide (PCR-SSO kit, One Lambda® Inc, Woodland Hills, USA) for 16 KIR genes. SOFA scores were collected at admission and every day during the intensive care unit (ICU) stay. To evaluate the influence of individual KIR genes on each of the six system or organ dysfunctions as evaluated by the SOFA score, Poisson regression analysis was performed. A p value ≤0.05 after correction was considered to indicate a statistically significant difference.

Conclusions
Although the number of patients is small, this preliminary study shows that the expression of KIR genes 2DS1 and 3DS1 may protect against cardiac dysfunction, and KIR gene 2DL5 may protect against cardiac and renal dysfunction in critical care patients. Biological mechanisms explaining why presenting certain KIR genes or haplotypes influence organ dysfunction are not yet known; perhaps expressing a greater number of activating KIR genes results in a more effective response against critical disease. Further studies are needed to translate this genetic observation into clinical evidence.
The data from 2018 were excluded from the sample due to the possibility of updates.

Results
The total use of circulatory assist devices in the evaluated period was 346 with an increase from 2014; the annual average was of 57 ± 11.16. Of the total, 76% (n = 263) were requested as an emergency, with the remainder were installed as elective surgery. Only 13 of 27 federative units (48.14%) performed this procedure in the last decade, with no records of their use in the north of Brazil. Less than half (46.15%) of those federative units that used the resource used it five times or less. The southeast has the highest incidence (n = 217, 62.72%) followed by the midwest (n = 54, 15.61%) with an emphasis on São Paulo (n = 185, 53.47%) and the Federal District (n = 52, 15.03%). The total cost of hospitalization per procedure showed great variation, extending from R$266.18 to R$37,902.18, with an average of R$5,500.00. The mortality rate was 35 ± 14.23, which was higher in the south (58%) and lower in the midwest (25.9%). There are no recorded data that differentiate the types of circulatory assistance provided in the public health system, nor is the INTERMACS or equivalent score available.

Conclusions
The amount of mechanical circulatory support used in the last decade is a small significative amount, with none registered in more than half of the country. The available data do not differentiate the type of assistance provided, nor the profile of the patients in cardiogenic shock that would benefit from the therapy, as recommended by the Guidelines for Mechanical Circulatory Support of the Brazilian Society of Cardiology since 2016. Therefore, it reinforces the necessity to create a national registry to enable the analysis, comparison, and evolution of the use of devices in the country, such as the successful registries in Europe (EUROMACS), the United States (INTER-MACS), and internationally (IMACS). Background Acute myocardial infarction (AMI) is an unexpected event in individuals under the age of 40 years. It is important to analyze these patients in order to draw a profile of them and identify a possible mechanism in common for such an unexpected event.
Objective The present study aimed to describe the physiopathology of young people under 40 years of age who suffered an AMI by correlating the types of infarction, associated mechanism, presence of comorbidities, and mortality rates.

Methods
Retrospective analysis was performed on the myocardial infarction database of a private network of hospitals and, among 2783 cases of AMI attended in 15 hospitals in the state of São Paulo in the period between January 2014 and December 2018, 62 patients were <40 years old (2.15%). The cases were separated according to the ECG, whether or not there was cardiac catheterization, angiographic aspects and the type of infarction, and the probable mechanism of myocardial infarction in cases of AMI type 2.

Results
Of the patients analyzed, 53 (85.4%) had atherothrombotic disease (AMI type 1) as a mechanism while (14.5%) were patients with type 2 AMI. The mechanisms responsible for type 2 AMI were most commonly coronary embolism (44%), spontaneous coronary dissection (33%), and imbalanced supply × demand (22%). Diagnosis of STelevation myocardial infarction (STEMI) represented 47% of the type 1 AMI and 11% of the type 2 AMI. Among the AMI type 1 with atherothrombotic disease, 50% had lesions in the left anterior descending artery (LAD) as the most important mechanism revealed by catheterization and, of these, 35% had lesions present in the proximal third. It was also observed that 25% of the cases had lesions in the right coronary artery (RCA) as the most important mechanism and, among these cases, 50% presented with lesions in the middle third of the artery. About 20% of the patients who had AMI type 2 presented with hypertrophy of the left ventricle and this finding related to the event of imbalance of supply × demand). It was also found that, among the patients with AMI <40 years, 40% were smokers (41% of AMI type 1 and 37.5% of AMI type 2). Conclusions More than 85% of AMI cases in patients <40 years old is by atherothrombotic disease and of these in almost half shows STsegment elevation. The most frequente cause of infarction type 2 was the coronary embolism and smoking was very prevalent risk factor.

Background
Myocarditis is a generic term that encompasses a broad group of clinical situations that present with inflammation of the myocardium. A frequent but not very well explored group is myocarditis that mimics a picture of acute myocardial infarction. Objective We aimed to evaluate the clinical and laboratory pattern in the short and long term in a cohort of patients with acute myocarditis diagnosed after exclusion of coronary disease.

Methods
This was a multicenter observational study that included all consecutive patients presenting with chest pain and positive troponin without any significant obstruction on coronary angiography (invasive or noninvasive) in a 3-year period with a diagnosis of acute myocarditis based on characteristic late enhancement in the complementary investigation. The clinical and laboratory characteristics were analyzed at the baseline and the outcomes were evaluated every 6 months after discharge. Results A total of 61 cases of myocarditis were diagnosed in this period (54.1% presented a diagnosis of associated pericarditis) and all cases were considered to be of viral etiology, although 14.7% of the patients had no infectious disease 30× the limit of normality. The levels of these markers showed variations between measures >30%, but the serial curve already started in the descending phase in 81.9% of the cases. The clinical and laboratory characteristics are presented in Table 1. Only one patient died due to cardiogenic shock (in-hospital mortality of 1.6%, 95% confidence interval (CI) 0.01-9.5%). At a median of 2.5 years there was no death after hospital discharge and there was only one re-hospitalization (for heart failure).

Conclusions
More than half of the cases of acute myocarditis presented altered electrocardiogram (EKG) with an associated diagnosis of pericarditis and more than 20% presented segmental alterations to the echocardiogram. The prognosis was good in the long term, although 10% remained with left ventricular dysfunction. The categorization of myocarditis in more homogenous groups will allow a better understanding of the natural history and the generation of better hypotheses of potentially effective treatments. Background Cardiac surgery is a standard procedure worldwide. Many of the patients have been followed in the intensive care unit (ICU) after surgery. Many concerns about intensive care treatment could be solved by publication in the literature. Research is an essential and unique field that intensivists all over the world deal with, in addition to their daily clinical practice. Published articles represent a central part of the research process. Objective We aimed to assess the current quality of evidence from published papers in Scopus Database of the management of cardiac surgical patients in the critical care unit. Methods An internet electronic search was made in the Scopus database looking for published articles between 2004 and 2018. The search strategy involved core terms related to CCM and the specific search strategy was as follows: TITLE-ABS-KEY ( "cardiac surgery" ) OR TITLE-ABS-KEY ( "heart surgery" ) AND TITLE-ABS-KEY ( "intensive care" ) OR TITLE-ABS-KEY ( "critical care" ) AND TITLE-ABS-KEY ( adult ) AND NOT TITLE-ABS-KEY ( animal ) AND ( LIMIT-TO ( SRCTYPE , "j" ) ). There were no restrictions on language but the abstract should be written in English.

Conclusions
Many cardiac surgical patients have been monitored in the ICU, although a search on research on this has found low numbers of good papers and systematic reviews within this group of patients. Improving the quality of studies will provide high-quality intensive care management after cardiac surgery.

Background
Although one-third or more of critically ill patients in the United States are obese, obesity is not incorporated as a contributing factor in any of the commonly used severity of illness scores.

Objective
We hypothesize that selected severity of illness scores would perform differently if body mass index categorization was incorporated and that the performance of these score models would improve after consideration of body mass index as an additional model feature.

Methods
This retrospective study was exempt from institutional review board approval due to lack of direct patient intervention, and the security schema for which the re-identification risk was certified as meeting safe harbor standards by Privacert (Cambridge, MA; HIPAA Certification no. 1031219-2). We performed a secondary analysis of electronic health records from a multicenter ICU database, which contains deidentified data for more than 200,000 intensive care unit (ICU) admissions from 208 distinct ICUs across the United States between 2014 and 2015. All patient in the ICU database were eligible for inclusion, but only the first ICU admission of patients with documented height and weight were included. The primary outcome was in-hospital mortality.

Results
Of 108,402 patients from 189 different ICUs across the United States included in the analyses, 4661 (4%) were classified as underweight, 32,134 (30%) as normal weight, 32,278 (30%) as overweight, 30,259 (28%) as obese, and 9070 (8%) as morbidly obese. Baseline characteristics are summarized in Table 1. To assess the effect of adding body mass index as a risk adjustment element to the Acute Physiology and Chronic Health Evaluation (APACHE) IV and Oxford Acute Severity of Illness scoring (OASIS) systems, we examined the impact of this addition on both discrimination and calibration. We performed three assessments based upon: 1) the original scoring systems; 2) a  (Fig. 1). We also performed a subgroup analysis in groups defined using World Health Organization guidelines for obesity. Incorporating body mass index into the models provided a minor improvement in both discrimination and calibration. In a subgroup analysis, model discrimination was higher in groups with higher body mass index, but calibration worsened.

Conclusions
The performance of ICU prognostic models utilizing body mass index category as a scoring element was inconsistent across body mass index categories. Overall, adding body mass index as a risk adjustment variable led only to a minor improvement in scoring system performance.

Background
The evolution of critical care has increased the number of patients who survive the acute insult but who remain with some organic dysfunction. Patients with this phenomenon are called chronic critically ill patients (CCIP). This condition is a serious and growing health problem in countries, resulting in high costs to the health systems. Background Recent data suggest that severity and the number of organ failures, management complexity, and performance status have higher impact on intensive care unit (ICU) and hospital mortality among cancer patients than the underlying neoplasm disease [1][2][3]. Objective We aimed to evaluate the risk factors for ICU mortality among cancer patients.

Methods
This was a retrospective study conducted at an 11-bed ICU of a public cancer hospital in São Luis, Maranhão. All cancer patients >18 years old requiring ICU stay >24 h from January 2016 to December 2017 are included. We evaluated demographic and clinical variables, and ICU support at admission and during ICU stay. The risk factors for ICU mortality were investigated through multiple logistic regression analysis.

Conclusions
Characteristics prior to and upon ICU admission were identified as risk factors for ICU mortality. The severity of acute illness at ICU admission suggests that prompt recognition of organ dysfunction, the possibility of early ICU referral, and the expansion of prevention strategies to reduce hospital-acquired infection could be important strategies for reducing ICU mortality rates.

Background
Heart failure (HF) is characterized by the inability of the heart to pump effectively as a result of structural or functional cardiac modifications. HF is considered severe and the 5-year survival rate is about 35%. It is estimated that it affects 23 million people in the world, and in Latin America is significant due to its economic, social, and cultural aspects.

Objective
We aimed to evaluate the epidemiological profile of patients hospitalized in the Brazilian public health system (SUS) with heart failure in the last decade, as well as the mortality rate and costs to the government.

Methods
Our study is a cross-sectional, retrospective, observational study of data collected through the hospital information system of the public health system (SIH/DataSUS) from January 2008 to December 2017. The data from 2018 were excluded from the sample due to the possibility of updates.

Results
The annual average number of hospitalized patients with heart failure was 241,500, representing 21.28 ± 2.11% of hospital admissions for diseases of the circulatory system. Comparing the years 2008 and 2017, a gradual decline in hospitalization was seen, with 27.3% less in 2017. The standard cost of hospitalization increased progressively, reaching R$1,582,00 in 2017, 67% more than 2008; moreover, there was an additional 1 day spent in the hospital. When we categorized by the age group, people who were 80 years or older had a higher prevalence, but patients from 60 to 79 years old represent 50% of the sample. Regarding the difference between genders, men between 30 and 69 years old were hospitalized about 5% more; however, women correspond to 56.5% of those patients aged 80 years or more. There were annual increases in mortality, reaching 10.7% in 2017. Excluding patients younger than 5 years old, mortality increases as the age group rises, with those aged 80 years or more representing 14.41%, followed by those aged 70 to 79 years at 9.96%, with no significant difference between genders. Concerning the age group in which mortality is high (60 to 80 years or older) there was an increase in the number of hospitalizations with morbidities in the same period analyzed, such as acute myocardial infarction (79.49%) atherosclerosis (94.03%), and sepsis (297.07%). In the context of region, the southeast represents 42%, followed by the northeast and south, which together correspond to 45.5%.

Conclusions
Despite a significant decrease in hospital admissions as a consequence of HF, mortality increases every year, as well as the financial cost. The most affected patient profile has a higher prevalence of morbidities, suggesting a more severe clinical scenario. In this study, we analyzed the patients admitted to a neurosurgical ICU, correlating pre-admission markers with the respective outcomes.

Methods
We analyzed data from 47 neurosurgical patients hospitalized at the Intensive Care Center of the General Hospital of Fortaleza from January to March 2019. In addition to the demographic data, we inferred severity through the Systemic Inflammatory Response Syndrome (SIRS) and the duration of surgical proceedings. The SPSS software program was used to process the statistical analysis.

Results
Of the total patients, 24 were men (51%), and the mean age was 50 ± 15 years, the length of stay was an average of 7 ± 10 days, and mortality was 5%. Surgeries were mainly (95%) elective. The main diagnoses were brain tumor (70.2%) and subarachnoid hemorrhage (19.1%). SIRS values ranged from 0 to 3 points, with <2 points in 72.3% of patients; 2 of 3 patients who died had only 1 point. In 70.2% of the patients the surgical procedure lasted around 3 and 6 hours. Using the Pearson chi-square test, the correlations had the following meanings: SIRS x duration of surgery, p: 0.153; SIRS x outcome, p: 0.470, and procedure duration x outcome, p: 0.829.

Conclusions
Given the low initial severity, inferred by the SIRS score, the low mortality and the absence of a significant correlation of surgical duration with the outcome, it seems rational to disregard an imperative admission to the ICU in the immediate postoperative window. In view of the high demand for ICU beds, costs, and ethical aspects, we suggest the creation of a unit specializing in post-neurosurgery.

P11
Early and late mortality following discharge from the intensive care unit: a multicenter prospective cohort study Background Long-term mortality is higher in intensive care unit (ICU) survivors than in the general population, yet few data exist on the incidence, causes, and predictors of early and late mortality after ICU discharge.
Objective We aimed to identify the incidence, causes, and predictors of early and late mortality among adult patients discharged alive from ICUs.

Methods
This was a multicenter prospective cohort study was performed in 10 tertiary hospitals in Brazil from May 2014 to December 2018. Adult ICU survivors with an ICU stay >72 h for medical and emergency surgical admissions or >120 h for elective surgical admissions were followed up for 12 months. The main outcomes were early (30 days) and late (31 to 365 days) mortality. Causes of death were independently extracted from death certificates and medical records by two critical care physicians and grouped according to the Tenth International Statistical Classification of Diseases and Related Health Problems.

Conclusions
In this multicenter prospective cohort study performed with adult ICU survivors, infection was the main cause of early and late deaths.
Older age, pre-ICU comorbidities and physical dependence, and severity of critical illness were associated with an increased risk of early and late mortality after ICU discharge, while ICU-acquired infections were associated with an increased risk of early mortality. These findings may guide future interventions aimed to prevent post-ICU deaths. The Simplified Acute Physiology Score III (SAPS 3) is the most accurate prognostic model for Brazil [1]. However, its performance in very elderly Brazilian patients (age ≥80 years) has not been adequately evaluated.

Objective
We aimed to compare the performance of SAPS 3 in patients older than 80 years with those younger than 80 years.

Methods
This was a retrospective cohort analysis of the Epimed® administrative database of adult patients (>18 years) on first admission to the intensive care unit (ICU) of Hospital Sírio Libanês (São Paulo, Brazil) from 2012 to 2016. Patients were categorized according to age (≥80 years or younger). We evaluated discrimination using the area under the receiver operating characteristic curve (AUROC) and the agreement between observed and expected mortality (calibration) using the calibration belt.

Results
We studied 7468 patients (age 65.9 ± 18.3 years, 53.3% male, SAPS 3 score 40.4 ± 13.4, 41.4% had cancer, 11.2% died during hospitalization), with 2002 patients (26.8%) older than 80 years. SAPS 3 discrimination was better for elderly patients below 80 years of age (AUROC = 0.81) than for very elderly patients over 80 years of age (AUROC = 0.70) (p < 0.001; Fig. 1). After we applied the calibration belt we observed that, in those aged 80 years and above, SAPS 3 underestimated mortality in those with low risk and overestimated it in those with middle risk . In those less than 80 years of age, SAPS 3 underestimated mortality only in those with low risk (Fig. 2).
Conclusions SAPS 3 has some limitations in those patients older than 80 years of age.   emergency physicians must be able to use strategies that revert the anticoagulant effect to deal with the urgent care.

Reference
Objective The aim of this study was to undertake a systematic review of the literature to display the conduct indications for the population in the use of NOACs during hemorrhagic events or urgent surgeries.

Methods
This article is a systematic review carried out from the descriptors "new anticoagulants" AND "emergency" on the Bireme data network where, from 335 scientific articles, 12 followed the inclusion criteria.
Were selected the full research in LILACS, MEDLINE, and IBECS databases, published between 2010 and 2018, in Portuguese or English and that had as the main subject "Anticoagulants" and "Emergency Hospital Service", excluding papers characterized as case reports.

Results
The surgical recommendation for the use of NOACs for a patient is when the plasma concentration of the drug is below 30 ng/ml. If the stipulated amount is exceeded, the indication is to postpone the surgery and to dose again after 12 h. In patients with compromised renal function, hemodialysis should be considered to reduce the level of the drug. When a postponement is not possible, it is prescribed to follow the procedure and, in case of abnormal bleeding, to administer an unspecific hemostatic factor (PCC or FEIBA), or the only specific hemostatic factor on the market, idarucizumab, for those who use dabigatran. In minor bleeding events, general actions for control and the delay of the anticoagulant are recommended. In major events, the following are considered: localization of the bleeding source; hydric control and maintenance of renal function; transfusion; use of activated charcoal; dialysis; and use of hemostatic agents. The immediate neutralization of the anticoagulant effect in intracranial hemorrhage is emphasized. In extended bleedings with plasma concentrations below 30 ng/ml, it is considered improbable that the drug provoked the occurrence, and reversion should not be carried out except when other actions cannot be performed. Finally, due to a lack of research, these recommendations cannot be taken as absolute guidelines but they can be a strategy against scarce scientific evidence.
Conclusions A large part of the literature within this area is based on the opinions of professionals, which focuses on evidence level 5. Thus, the designation of specific measures has become necessary, and an evolution of the guidelines is expected.

P14
Intensive approach to systemic lupus erythematosus associated with preeclampsia and its implications in a concept: a case report Background Preeclampsia is one of the most frequent complications of pregnancy in individuals with systemic lupus erythematosus (SLE), occurring in 16-30% in this group compared with 4.6% in the general obstetric population. HELLP syndrome is a possible outcome, presenting a challenging prognosis.

Objective
We aimed to report a case of uncontrolled SLE and its repercussion on pregnancy, which includes eclampsia, HELLP syndrome, fetal loss, and ischemic stroke. Case presentation A female patient, 28 years old, white, pregnant for the first time (22 weeks) and an SLE carrier attended our service reporting headache, abdominal pain, high blood pressure (HBP), and emesis; following the use of metoclopramide, the patient presented with lowering of sensory function. Although we used sulfate, she had a seizure (eclampsia). Evidence of ischemic stroke ( Fig. 1) was confirmed by a brain magnetic resonance imaging (MRI). The laboratory review showed a diagnosis of HELLP (Table 1). Pulse corticosteroid therapy with methylprednisolone, plasmapheresis with albumin, and doses of acetylsalicylic acid, simvastatin, and phenytoin were initiated. In addition, the induction of delivery began. At the time of writing, on the 29th day of hospitalization, the patient was conscious, despite presenting neurological deficits.

Discussion
High estrogen levels during pregnancy lead to an increase in body cytokines enhancing the action of SLE antibodies against cell membranes, which stimulates thrombogenesis and intensifies preeclampsia by damaging the endothelium and the platelets. Such damage causes HBP through vasoconstriction. Intensive treatment requires pregnancy interruption, immune system control, and neurological stabilization. Conclusions SLE implies potential risks for the pregnant woman and the fetus. Therefore, it is important to monitor these patients to investigate any clinical or laboratory changes and, thus, intercede for a better prognosis. Consent to publish Informed consent to publish has been obtained from the patient's family.  There is a lack of knowledge of how to use polymyxins, especially in the critically ill who often are given it empirically and have high rates of death [1].

Objective
We aimed to compare patients who received guided and empirical therapy with polymyxins during intensive care unit (ICU) stay.

Methods
We looked at a historical cohort of all consecutive patients who were given polymyxins to treat ventilator-associated pneumonia from 1 January 2017 to 31 January 2018 during hospitalization in an ICU from a university hospital endemic for carbapenem-resistant bacteria. The aim of this study was to identify risk factors for 28-day mortality in patients with CR-KP infection.

Methods
We conducted a retrospective study of CR-KP infections in critically ill patients between January 2015 and October 2017 at the Hospital São Rafael. Continuous variables were compared by the Student t test and categorical variables by the χ 2 test. Multivariate logistic regression analysis was performed to determine the independent risk factors for mortality. The study was approved by Hospital São Rafael's Ethics Board, approval number 2890446. Results A total of 90 patients with CR-KP infections were identified. Twentyeight-day mortality was 40.0% (n = 36). Variables associated with 28day mortality in the univariate analysis are shown in Table 1. Charlson Comorbidity Index (CCI; p = 0.001; odds ratio (OR) 1.38, 95% confidence interval (CI) 1.14-1.68) and Sequential Organ Failure Assessment (SOFA; p = 0.014; OR 1.33, 95% CI 1.06-1.66) were independently associated with mortality in the multivariate analysis.

Conclusions
In critically ill patients infected with CR-KP, 28-day mortality was independently associated with SOFA and CCI. Background Intradialytic hypotension (IDH) is one of the most common reasons for modification of the dialysis prescription from the intermittent to the continuous method. It is a barrier for effective treatment in critically ill patients and can also contribute to a delay in recovery of renal function and improvement in other organ failures. However, fluid status remains poorly defined by clinical examination findings, and sonography is a noninvasive method that has been proposed to enhance the bedside assessment of intravascular fluid volume.

P17
Objective The aim of this study was to determine the prevalence and risk factors of IDH in critically ill patients undergoing intermittent dialysis. We also sought to evaluate different profiles based ion sonographic findings just before the initiation of hemodialysis and their relation to IDH.

Methods
From January 2016 to April 2018, a prospective study was conducted in a single center with 248 patients with acute renal injury with an indication for intermittent dialysis.

Results
Hypotension was seen in 31.9% of patients and 28-day overall mortality occurred in 50 patients. Among the factors significantly associated with hypotension were high levels of lactate and low mean arterial pressure. Of the patients, 65.8% had a diagnosis of sepsis, 40.5% were using noradrenaline, and 25.3% required mechanical ventilation.

Conclusions
We identified risk factors related to fluid status and vasomotor tone that may be implicated in the occurrence of IDH. However, further studies should be considered to validate these results and design an appropriate specific treatment.

P20
The

Conclusions
The role of renal ultrasonography in critically ill patients with acute renal injury still needs to be determined; in patients with a history suggestive of urinary tract obstruction, renal ultrasound for evaluation of AKI should be indicated. The use of renal RI could be a tool to reflect vasomotor tone in patients with AKI related to sepsis.  The use of renal replacement therapy (RRT) in acute kidney injury (AKI) patients in the intensive care unit (ICU) is associated with high hemodynamic instability, leading to an in-hospital mortality of about 50%. AKI is a frequent complication in the ICU and is independently associated with the development of end-stage renal disease and higher mortality rates [1,2]. Despite the significant advances in medical treatments, AKI requiring RRT in the ICU is associated an increased in-hospital mortality of about 50% [3].

Objective
The aim of this study was to compare hemodynamic parameters among continuous, intermittent, and hybrid RRT in AKI in ICU patients.

Methods
This was a systematic review conducted in accordance with PRISMA and registered at the PROSPERO database (CRD42018086504). Randomized clinical trials (RCTs) involving patients with AKI in the ICU submitted to continuous, intermittent, or hybrid RRT were included. We investigated the PubMed, Embase and Cochrane databases.

Results
Out of 3442 citations retrieved, 12 RCTs were included, representing 1419 patients. Most of the studies did not find differences in hemodynamic parameters across different RTT modalities, except a continuous venovenous hemofiltration (CVVH) group heart rate decrease after 1 and 4 h in comparison with an intermittent hemodialysis (IHD) group, an increase in systolic blood pressure after 0.5 and 2 h of CVVH in contrast to IHD, and significantly higher doses of dobutamine in patients from the continuous venovenous hemodiafiltration (CVVHDF) group when compared with IHD. Lower baseline mean arterial pressure (MAP), greater MAP variation on dialysis, higher number of pressors at baseline, and an increase in pressor dose during dialysis were associated with shorter survival time; greater MAP variation on dialysis was negatively correlated with renal recovery.

Conclusions
Most of the studies did not find differences between the groups in relation to heart rate, blood pressure, hypotensive episodes, or use of catecholamines. We believe that this study can provide important insights for further clinical trials designed specifically to evaluate hemodynamic parameters across different types of RRT.

Background
Aneurysmal subarachnoid hemorrhage (SAH) is an acute, and often catastrophic, cerebrovascular event, with high mortality and morbidity. Data on predictors of mortality in low-and middle-income countries are scarce.

Objective
The purpose of this study was to investigate clinical predictors of hospital mortality in patients admitted with subarachnoid hemorrhage in a large sample of Brazilian intensive care units (ICUs).

Methods
We performed a retrospective cohort study of patients admitted with spontaneous SAH to 57 hospitals in Brazil during 2014 and 2015. We retrieved the clinical and outcome data of patients from an electronic ICU quality registry. Simplified Acute Physiology III (SAPS 3) non-Neuro and Sequential Organ Failure Assessment (SOFA) non-Neuro scores were calculated, subtracting the Glasgow Coma Scale (GCS) contribution from the original score values. We used mixed multivariable logistic regression analysis to identify factors associated with hospital mortality. Results A total of 1114 patients were included. Fifty-five percent (n = 610) of patients were female and 71% were admitted from the emergency room while 18% were transferred from another hospital. Median age was 57 years (interquartile range (IQR) 45-72) and median ICU length of stay was 5 days (IQR 2-10). Median SAPS 3 non-Neuro score was 46 (38-55) and SOFA non-Neuro score was 2 (0-5). A total of 446 (30%) patients presented with poor grade SAH (World Federation of Neurosurgeons grading scale IV and V) and hospital mortality was 35%. In univariate comparisons, nonsurvivors were older and had higher SAPS 3 non-Neuro and SOFA non-Neuro scores (all (all P<0.001). Poor grade SAH, use of vasopressors, mechanical ventilation, intracranial pressure monitoring, external ventricular drainage, blood transfusions and renal replacement therapy were all more frequent among nonsurvivors (all P<0.001). Mortality was also higher with initial lactate above 2 mmol/L, in those admitted to public hospitals and when admission to ICU was delayed more than 24 hours after ictus. After adjusting for common predictors (age, gender and WFNS) SAPS 3 non-Neuro, SOFA non-Neuro, early vasopressor use and admission to a public hospital were independently associated with hospital mortality. Moreover, the area under the curve for prediction of mortality with SAPS3, SOFA and WFNS was 0.86.

Conclusions
Mortality is elevated and highly variable in this large sample of SAH patients. Age and severity of clinical presentation, both systemic and neurological, as well as the presence of organ dysfunction were associated with mortality.

P24
Clinical predictors of early and late mortality after cardiac arrest: a retrospective multicenter cohort study Pedro Kurtz 1 , Christian Storm 2 , Marcio Soares 3

Methods
We performed a retrospective cohort study of survivors from cardiac arrest in 57 hospitals in Brazil during 2014 and 2015. We retrieved clinical and outcome data of patients from an electronic ICU quality registry. We used multivariable logistic regression analysis to identify factors associated with hospital mortality.

Results
A total of 2295 patients were included. Both patients with primary admission diagnosis of cardiac arrest (n = 800, 35%) and those that arrested on admission to the ICU (65%) were included. Median age was 67 years (interquartile range (IQR) 54-79) and 53% were male. Median Simplified Acute Physiology III score (SAPS 3) was 70 (57-83), Sequential Organ Failure Assessment (SOFA) score was 10 (7-13) and hospital mortality was 83%. Among nonsurvivors, 47% died in the first 48 h of ICU admission (early mortality). Only 1% received therapeutic hypothermia and 6% underwent withholding/withdrawal of life support. After adjusting for SAPS 3 and SOFA, early mortality was associated with hemodynamic compromise (systolic blood pressure 4 mmol/l and use of vasopressors) and hypo-or hyperthermia. Late mortality was independently associated with delayed admission to the ICU (>48 h) and palliative care.

Conclusions
This large cohort of post-cardiac arrest survivors demonstrates extremely high mortality rates and negligible implementation of hypothermia in Brazil, with half of nonsurvivors dying within 48 h of admission with severe hemodynamic compromise and organ dysfunction. Late mortality was related to delayed admission to ICU and withdrawal/withholding measures.

Methods
This was a multicenter prospective cohort study performed in 10 tertiary hospitals in Brazil from May 2014 to December 2018. Adult ICU survivors with an ICU stay >72 h for medical and emergency surgical admissions or >120 h for elective surgical admissions were followed up for 6 months. The main outcomes were anxiety and depression symptoms assessed by the Hospital Anxiety and Depression Scale (HADS) , PTSD symptoms assessed by the Impact Event Scale-6 (IES-6), and quality of life assessed by the short-form (SF)-12 questionnaire.

Results
A total of 1616 post-ICU patients were enrolled. Of these, 579 were assessed for mental health symptoms at 6 months. The prevalence of any mental health disability was 36.3% (210 patients). The prevalence of anxiety, depression, and PTSD were 24.2% (140 patients), 20.9% (121 patients), and 15.9% (89 patients), respectively. Patients with post-ICU mental health disabilities had lower scores of quality of life in both physical and mental domains compared with patients without post-ICU mental health disabilities. Age <65 years (p<0.001), female sex (p=0.01), history of depression (p<0.001), need of renal replacement therapy (p=0.02) and decrease in functional status (p<0.001) were independently associated with anxiety. History of depression (p<0.001), increased risk of death at ICU admission (p=0.04) and decrease in functional status (p=0.003) were independently associated with depression. Age <65 years (p=0.01) and decrease in functional status (p=0.001) were independently associated with PTSD.

Conclusions
The high burden of mental health disability among survivors of critical care is a matter of concern for public health due to its negative impact on quality of life. The network of potential risk factors for anxiety, depression, and PTSD among post-ICU patients is complex and involves factors of multiple domains such as sociodemographic, premorbid health, and ICU-related variables and sequela of critical illness. Anemia is frequent among patients with traumatic brain injury (TBI) and is associated with an increased risk of poor outcome. The optimal hemoglobin concentration to trigger red blood cell (RBC) transfusion in TBI patients is not clearly defined. Objective We aimed to evaluate the feasibility and safety of two different hemoglobin thresholds for blood transfusion in patients with moderate or severe TBI. Methods All eligible consecutive adult patients admitted to the intensive care unit (ICU) with moderate or severe TBI were randomized to a 'restrictive' (hemoglobin transfusion threshold of 7 g/dl) or a 'liberal' (threshold 9 g/dl) transfusion strategy. The transfusion strategy was continued for up to 14 days or until ICU discharge. The primary outcome was the mean hemoglobin difference between groups. Secondary outcomes included transfusion requirements, intracranial pressure management, cerebral hemodynamics, length of stay, mortality, and 6-month neurological outcome.

Results
A total of 44 patients were randomized, 21 patients to the liberal group and 23 to the restrictive group. There were no baseline differences between the groups. The mean hemoglobin concentrations over the 14-day period were 8.4 ± 1.0 and 9.3 ± 1.3 (p<0.01) in the restrictive and liberal groups, respectively. Fewer RBC units were administered in the restrictive than in the liberal group (35 vs. 66, p=0.02). There was a negative correlation (r=-0.265, p<0.01) between hemoglobin concentration and middle cerebral artery flow velocity as evaluated by transcranial Doppler and the incidence of posttraumatic vasospasm was significantly lower in the liberal strategy group (4/21, 3% vs. 15/23, 65%; p<0.01). Hospital mortality was higher in the restrictive than in the liberal group (7/23 vs. 1/21; p=0.048) and the liberal group tended to have a better neurological status at 6 months (p = 0.06).

Conclusions
The trial reached feasibility criteria. The restrictive group had lower hemoglobin concentrations and received fewer RBC transfusions. Hospital mortality was lower and neurological status at 6 months favored the liberal group. The prevalence of malnutrition in hospitalized patients is high and may even increase during hospital stay. Adequate nutritional management can improve this scenario, and, currently, the Multiprofessional Nutrition Therapy Team is the most important instrument in this context for obtaining improvement. However, a Multiprofessional Nutrition Therapy Team is not always available in hospitals and the alternative would be to use the telemedicine service to increase the access and reach of specialized professionals. Objective This study aims to compare the care assistance quality of nutritional therapy in critical patients, analyzing indicators before and after the implementation of the telemedicine service in a tertiary hospital. Methods This is a case-control study comparing the indicators of the proportion of the prescribed volume that was infused (volume adequacy), the proportion of the protein target reached, and the diarrhea rate prior to the implementation of telemedicine first period (October 2016 to February 2017) and after the telemedicine second period (from June 2017 to October 2017). All patients who received enteral nutritional therapy during the period participated in the study.

Results
A total of 150 patients (65 patients from first period and 85 from the second period) participated in the study. The median age was 57.5 (interquartile range (IQR) 44-64) years, 51.8% were male and 48.2% were female, and most of the patients were receiving treatment for pulmonary focus sepsis. In general, the median percentage of diet volume infused was 88.8% (75.7-98.1%), the patients received a median of 1.13 (0.9-1.33) g/kg of protein, and 14.7% presented with diarrhea. The crude diet infused volume increased from 83.3% (IQR 71.3-95.7%) in the first period to 91.7% (IRQ 80.1-99.2%) in the second period (p = 0.01) and the same has occurred for the protein received, but this was not statistically significant (1.08 ± 0.3 g/kg in the first period and 1.16 ± 0.3 g/kg in the second period; p = 0.10). After adjusting for age, patients enrolled in the first period (19.6%, 95% confidence interval (CI) 9.4-36.1) compared with those enrolled in second period (0.0%, 95% CI 0.0-7.2) exhibited a lower risk of diarrhea (odds ratio (OR) 0.04, 95% CI 0.02-0.7; p = 0.03).

Conclusions
The telemedicine seems to achieve good results in the follow-up of patients who received enteral nutrition when compared with no specialized professional follow-up; this study demonstrated that it improved the care quality indicators. Background Dysphagia has already been reported to increase the length of hospital stay and the risk of mortality in all age groups [1]. A simple swallowing screen can be used to identify patients at risk for bronchopulmonary penetration/aspiration, the need for further swallowing assessment, and the safety of patient oral intake [2]. Currently, no consensus exists on a standard method of clinical swallowing assessment [3].

Objective
The purpose of the present study was to assess the validity of a simple screening swallowing instrument used in a large public-school hospital in Brazil.

Methods
The study was approved by the Ethics Committee for the Analysis of Research Projects of Hospital das Clínicas School of Medicine University of São Paulo, approval number 1.781.177. The Dysphagia Risk Evaluation Protocol (DREP) screening version contains four items (altered cervical auscultation, altered vocal quality, coughing, and choking after the swallow) that were previously indicated as independent risk factors associated with the presence of dysphagia. Trained speech-language pathologists administered and scored the DREP screening version to consecutive patients referred by the hospital's medical team to perform a videofluoroscopy of swallowing (VFSS) Twenty percent of all enrolled patients were then randomly allocated to VFSS 24 h after being submitted to the DREP screening version.

Results
Two hundred and eleven patients received the swallowing screen: 99 failed and 112 passed. One in every five patients were randomized to receive a VFSS (Fig. 1). The DREP screening version demonstrated excellent validity with sensitivity of 92.9%, specificity 75.0%, negative predictive values of 95.5%, and an accuracy of 80.9% (Table 1).

Conclusions
The DREP screening version is a simple and accurate tool to identify bronchopulmonary penetration/aspiration in patients who are not tube fed, who have good level of alertness, no history of repetitive pneumonia or current respiratory compromise, and are not using a tracheostomy tube. Background Worldwide, patients are subject to medication errors due to poorly written or misinterpreted prescriptions, which can lead to adverse events ranging from those that are mild to even death. To minimize this type of error, it is necessary to use strategies to avoid their occurrence such as simple guidelines to prescribing physicians and/or the use of electronic prescriptions. Objective We aimed to evaluate the quality of intravenous antimicrobial prescriptions in an intensive care unit (ICU) of a quaternary hospital in Fortaleza, Ceara, Brazil, during a period of 1 month.

Methods
This is a descriptive, observational, cross-sectional and retrospective study. We identified all the antimicrobials for intravenous use prescribed during the month of November 2018 and analyzed the presence of the following items: 1) generic name; 2) concentration; 3) pharmaceutical form; 4) dose; 5) volume; 6) diluent; 7) administration route; 8) infusion rate; and 9) dosage. We considered any prescription that contained all nine items as safe.

Results
Three hundred and ninety-four antimicrobial prescriptions were identified in the period. Of these, we found adhesion rates of 94.9% for generic name, 26.7% for concentration, 25.9% for pharmaceutical form, 100% for dose, 90.9% for volume, 91.4% for diluent, 99.2 % for administration route, 68.3% for infusion rate, and 98.2% for dosage. Of the total, only 19.5% contained all the recommended items.

Conclusions
Our data suggest that we did not prescribe antimicrobials safely in the unit since only one in five prescriptions were complete. It is necessary to develop counseling strategies for physicians so that low adherence to the full prescription, an indirect indicator of low quality, does not become an error and lead to adverse events in patients. Background With the evolution of the therapeutic arsenal, as well as a growing concern on patient safety, lots of information has become necessary on a medical prescription to ensure its correct understanding. This information, when absent, incomplete or illegible, adds a high probability of occurrence of errors, sometimes fatal errors. The investigation of prescribing habits allows us to detect such failures, which makes possible the implementation of appropriate and timely interventions.

Reference
Objective We aimed to analyze the results of intervention on the intravenous antimicrobial prescriptions in an intensive care unit (ICU) of a quaternary hospital in Fortaleza, Ceará, Brazil, during a period of 1 month.

Methods
This is an analytical, interventional, before-and-after, longitudinal and retrospective study. We identified all the antimicrobials for intravenous use prescribed during the month of November 2018 and analyzed the presence of the following items: 1) generic name; 2) concentration; 3) pharmaceutical form; 4) dose; 5) volume; 6) diluent; 7) administration route; 8) infusion rate; and 9) dosage. During the month of December, through cell phone messages and printed information, we disclosed to ICU physicians the correct way to prescribe antimicrobials. After that, we collected data from the prescriptions of January 2019 and analyzed the efficiency of the intervention.

Results
Three hundred and ninety-four antimicrobial prescriptions were analyzed in November 2018. Of these, 26.7% contained the concentration, 25.9% had the pharmaceutical form, and only 19.5% were complete. In January 2019, 428 antimicrobial prescriptions were analyzed. We observed an increase in the concentration prescription (46.3%) and in the pharmaceutical form (45.1%), with a statistically significant difference (p <0.05). In addition, 42.8% of the prescriptions had all the recommended items, which represented an increase of 119.5% compared to the previous period, also with statistical significance (p <0.05). Despite this, adherence to the infusion rate prescription remained low [68.3% (November) vs 63.1% (January)]. The remaining items were present in more than 90% of the prescriptions in the two analyzed months.

Conclusions
Our study suggests that the intervention can be considered an efficient measure, but with results still lower than those expected of a safe antimicrobial prescription. It is therefore necessary that measures are taken to ensure the presence of essential information for a quality prescription, mainly through permanent orientation with the physicians. The meropenem recommended dose usually cannot achieve its target in critically ill septic patients in the intensive care unit (ICU) against the most common MIC >2 mg/l strains that can impact the desired outcome [1].

Objective
The aim of the study was to investigate meropenem 3 g daily by extended infusion based on the pharmacokinetic-pharmacodynamic (PK/PD) approach in burn patients receiving the initial dose regimen against nosocomial-susceptible strains. Methods Ethical approval was obtained (no. 069/09-2015). Thirteen burn patients were investigated. They had a median age of 27 years, weight of 70 kg, with 34% median total burn surface area, and Simplified Acute Physiology III score (SAPS 3) of 58; 8/13 had inhalation injury, with preserved renal function undergoing meropenem treatment at the earlier period of septic shock receiving 1 g every 8 h as a 3-h infusion. A series of two blood samples were collected (1.5 ml each) at the steady state for drug serum measurement by liquid chromatography [2]. PK results from burn patients were compared with data previously described in healthy volunteers [3]. The PK/PD approach was performed to estimate the probability of target attainment (PTA) based on the predictive index of drug effectiveness (100% fT > MIC) [4].

Results
Effective free serum trough levels occurred at the recommended dose regimen (1 g q8h 3-h infusion; Fig. 1A and B). Total isolates of bacterial infection were stratified in Enterobacteriaceae and Pseudomonas aeruginosa at the earlier period of septic shock (Fig. 1C-E). Target was attained for all patients by eradication of nosocomial pathogens up to MIC 4 mg/l (Fig. 1F). Pharmacokinetics are illustrated in Fig. 2

. Conclusions
The meropenem serum level is altered in the earlier period of septic shock in burn patients with impacts on PK and drug effectiveness. The desired outcome was reached during the therapy and the clinical cure occurred in all patients against MIC 4 mg/l.  The vancomycin empiric dose regimen usually recommended cannot reach the target in critically ill septic patients in the intensive care unit (ICU) against the most common MIC >1 mg/l strains and this can impact the desired outcome. Objective The aim of the study was to compare vancomycin 2 g versus 3 g daily based on the pharmacokinetic/pharmacodynamic (PK/PD) approach in ICU burn patients.

Methods
Ethical approval was obtained (no. 069/09-2015). Ten burn patients (1 female and nine male) had a median age of 25 years, weight of 78 kg, with 29% total burn surface area, and Simplified Acute Physiology III score (SAPS 3) of 63; 8/10 had an inhalation injury, with preserved renal function undergoing vancomycin therapy (1-h infusion) at the earlier period of septic shock receiving 1 g every 12 h (set 1) or 1 g every 8 h (set 2). A series of two blood samples were collected (1.5 ml each) at the steady state for drug serum measurement by immunoassay. PK results obtained from burn patients were compared with data previously described in healthy volunteers [1]. The PK/PD approach was performed to estimate the probability of target attainment (PTA) based on the predictive index of drug effectiveness (AUC/MIC >400) [2].

Results
We demonstrated a significant difference (p<0.0002) between the empiric and adjusted daily dose, figure 1A. Total isolates of gram-positive strains were stratified in Staphylococus spp, Streptococcus spp and Enterococcus faecalis isolated from fluids at the earlier period of septic shock, figure 1B.Target was attained after dose adjustment in 70% (7/10) patients by eradication of nosocomial pathogens up to MIC 2 mg/L, figure 1C. Pharmacokinetics was altered at the earlier period of septic shock, figure 2.

Conclusions
Vancomycin serum levels were reduced during septic shock as a consequence of increases in total body clearance and reduction in the biological half-life in burn patients with an impact on drug effectiveness. The desired outcome was reached during the therapy once the clinical cure occurred in 7/10 patients against MIC 2 mg/l. In septic pediatric burn patients, it is a challenge to maintain effective vancomycin plasma levels, once changes occur in the pharmacokinetics; the control of infection is therefore a challenge in these critically ill pediatric patients [1]. Objective The aim of the study was to investigate vancomycin effectiveness against infections mainly caused by Staphylococcus spp. MIC 0.5-2 mg/l in pediatric septic burn patients after the empiric dose (set 1) against the individualized dose regimen (set 2) by a pharmacokinetics/pharmacodynamics (PK/PD) approach.

Methods
The protocol was approved (no. 0069/15). Twenty patients (14 male and six female) were investigated, with the following characteristics: creatinine clearance (CrCl) median 241 (quartiles 202-295) ml/min, age 5.0 (3.0-9.5) years, weight 22 (16-38) kg, total burn surface area 30 (20-39) %. Fire was the reason for the burns, inhalation injury occurred in 12 patients, and vasoactive drugs and mechanical ventilation were required in 13/20. Blood was sampled at 3 h and 5 h (1.5 mL each) after the beginning of 1-h infusion. Patients were investigated in two sets after the recommended empiric daily dose (set 1) and after individualized drug therapy (set 2). Drug serum measurements were performed by liquid chromatography. A PK/PD approach was performed based on the predictive index AUC/MIC >400, where MIC is the minimum inhibitory concentration value for each isolated pathogen.

Results
We showed a significant difference between the recommended dose regimen and the adjusted dose ( Fig. 1A and B). Total isolates related to the infection caused by Gram-positive strains at the earlier period of septic shock are shown in Fig. 1C and D. The percentage of target attainment (PTA) is shown after both therapies in Fig. 1E.

Conclusions
The vancomycin serum level is altered at the earlier period of septic shock in pediatric burn patients with an impact on drug effectiveness. After dose adjustment, the desired outcome was obtained for all patients against S. aureus MIC 0.5-1 mg/l while the PTA was reduced to 65% (13/20) of patients against MIC 2 mg/l strains. Background A vancomycin initial dose regimen is recommended for critically ill pediatric patients with bloodstream infection caused by Grampositive strains. Drug effectiveness must be guaranteed by the area under the curve/MIC ratio (AUC/MIC) [1,2]. Objective The aim of the study was to evaluate if dose adjustment at the earlier period of septic shock must be performed for target attainment against Gram-positive MIC 1 mg/l strains based on a pharmacokinetics/pharmacodynamics (PK/PD) approach in burn and nonburn patients.

Methods
Thirty-four septic pediatric burn and nonburn patients (12 female and 22 male), with median creatinine clearance (CrCl) >240 ml/min, aged 5-10 years and weighing 16-22 kg (quartiles) were included in the study. Therapy started with 40-60 mg/kg daily equiv. 10-15 mg/ kg every 6 h, and the dose was adjusted if required based on the PK/PD target (AUC/MIC >400). Blood was sampled (1.5 ml each) at hours 3 and 5 after starting infusion; serum levels were obtained by immunoassay.

Results
We showed significant differences between the recommended dose regimen and individualized therapy (Fig. 1A and B). The percentage of target attainment (PTA) is illustrated after both therapies in those with burns and those without in Fig. 1C and D. Changes in pharmacokinetics are described for both groups by comparison with data previously reported in healthy volunteers (Fig. 2) [3]. After dose adjustment, the target was attained up to MIC 1 mg/l strains for all patients. The target was reached in 65% (13/20) of burn patients and in 43% (6/14) of nonburn individuals against MIC 2 mg/l strains.

Conclusions
Since vancomycin pharmacokinetics were altered in septic pediatric patients, when nosocomial Gram-positive strains MIC >0.5 mg/l were isolated, the dose must be adjusted to 80 mg/kg daily equiv. 20 mg/kg every 6 h to eradicate S. aureus MIC 1-2 mg/l. A vancomycin PK/PD approach performed in real time permits an earlier clinical intervention and desired clinical outcome with a cure of the infection. Background Piperacillin/tazobactam is largely prescribed in the intensive care burn unit (ICBU) for patients with infections caused by Gramnegative strains. The recommended dose cannot achieve the target once serum levels are below those required for effectiveness [1].

Objective
The aim of the study was to investigate if the empiric dose of 4.5 every 6 h by 2-h infusion can provide drug effectiveness by a pharmacokinetics/pharmacodynamics (PK/PD) approach in septic burn patients against susceptible MIC 8 mg/l strains. Methods Ethical approval was obtained from the hospital Ethics Committee (no. 069/09-2015). We studies nine burn patients (three female and six male) with preserved renal function undergoing treatment for septic shock; they had a median age of 32 years, weight of 72 kg, and 29% total burn surface area. Mechanical ventilation plus vasopressors were used in 8/9 patients, and inhalation injury was present in 6/9 of them. Only two blood samples were collected (1.5 ml each) for drug serum measurement by liquid chromatography. Pharmacokinetics (PK) data obtained from patients were compared with results from healthy volunteers [2]. A PK/PD approach was performed to estimate the probability of target attainment (PTA) based on the new predictive index of drug effectiveness (100% fT > MIC) [3]. Results A piperacillin serum trough after a 2-h infusion is illustrated in Fig.  1A and B. Target was attained for all patients by eradication of Gramnegative strains Pseudomonas aeruginosa and Enterobacteriaceae against MIC 8 mg/l (Fig. 1C). Changes in pharmacokinetics occurred as shown in Fig. 2.

Conclusions
Piperacillin pharmacokinetics are altered in septic burn patients improving drug effectiveness. A clinical cure occurred in all patients against isolated susceptible strains after the recommended dose was administered by extended infusion.

P36
Accuracy of troponin I as a marker in the differentiation of primary or secondary infarction due to sepsis: the RESCA  Background Serum troponin is a specific marker for myocardial necrosis. However, it is not specific for a coronary etiology (primary infarction) because it may change in systemic situations of reduced circulatory supply or increased demand. In these situations of instability, physicians may judge significant increases in troponin as a primary infarction, triggering inappropriate invasive coronary procedures. Objective We aimed to assess the accuracy of a troponin elevation magnitude in differentiating primary infarction or sepsis in critically ill patients who were troponin positive. Methods From the Acute Coronary Syndrome Registry (beginning in 2010) and Sepsis Registry (beginning in 2013) in our hospital, patients with non-ST elevation myocardial infarction and sepsis, respectively, were consecutively selected, both groups consisting of individuals admitted to intensive care and whose troponin I was positive (detection ≥0.012 μg/l; VITROS, Johnson & Johnson). Patients with creatinine on admission ≥1.5 mg/dL were excluded. Peak troponin values were tested for the differentiation of these two groups.

Results
A total of 337 patients with infarction (age 64 ± 13 years, 60% men, 92% in Killip I, 0% Killip IV) were compared with 41 patients with sepsis (age 69 ± 16 years, 59% men, systolic blood pressure 121 ± 33 mmHg, 27% admitted with vasoactive drugs). The median peak of troponin I in infarction was 0.33 μg/l (interquartile range (IQR) 0.06-3.2), a value three times higher than the median found in septic patients (0.11 μg/l, IQR 0.02-0.37). A 10-fold increase over the 99th percentile value was prevalent in both groups and more frequent in the infarction group (50% versus 29%; p = 0.01). The 90th percentile of the peak troponin I was 15 μg/l in the infarction group and 7.8 μg/l in the sepsis group. The area under the receiver operating characteristic curve for troponin of 0.65 was nonsignificant in differentiating the two conditions (95% confidence interval 0.56-0.74; p = 0,002).

Conclusions
Patients with non-ST elevation myocardial infarction present higher troponin elevations when compared with septic patients. However, the accuracy of troponin I to differentiate the two conditions is questionable and a significant proportion of septic patients present with substantial elevations in this marker.  Background Sepsis is currently the leading cause of death in burn patients. It is possible that the incidence of infection, increased total body surface area (TBSA), and inhalation injury require a longer hospital stay. There are few studies on sepsis in burn patients on the intensive care unit (ICU). Objective We aimed to determine the demographic profile, clinical presentation, evolution, procedures, and treatments used in burn patients affected by sepsis attending a university hospital. Methods This retrospective study was performed in a university hospital ICU by analyzing medical records of adult burn patients who developed sepsis from November 2015 to May 2018. We analyzed patient information, internment, and sepsis. The Sepsis-3 criteria were used for confirmation of a sepsis diagnosis.

Results
We studied 44 patients with an average age of 42.1 ± 16.88 years, with 75% of them being male. The median TBSA burned was 50% and was significantly associated with mortality (p = 0.013). Mortality was 50%. The median length of stay at hospital was 35 days and on the ICU was 21.5 days. Orotracheal intubation and tracheostomy were the most prevalent aggravations during the internment, occurring in 77.2% and 56.8% of patients, respectively. The median time to first sepsis episode was 7 days and the average total time in sepsis was 13.2 days. The median length of hospital stay in patients with septic shock who died was significantly lower than those who lived (p = 0.031). Bacterial culture was positive in 79.5% of cases and the majority of these were typical ICU bacteria. Broad-spectrum antibiotics were administered, the most used being meropenem and vancomycin.

Conclusions
This study shows that sepsis occurs more frequently in men, with higher TBSA burned, large internment time, and with aggravating complications. The infections were caused by typical ICU bacteria causing patient mortality of 50%, mostly by septic shock. The emergency department is a dynamic environment in which a large volume of patients with different levels of urgency should be evaluated and treated in a timely manner, especially by the nurses who provide first care. Patients with trauma may have different degrees of severity related to the mechanism of trauma (blunt, penetrating, or both) and each level of severity may determine more or less dependency on care.

Surgery/Trauma
Objective The aim of this study was to measure blunt and penetrating trauma patient dependency during a stay in the emergency room (ER). Methods A cross-sectional study was carried out on patients with severe trauma who were assisted in an ER in the city of Rio de Janeiro. Adult patients who died with blunt or penetrating trauma were evaluated for their dependency level during the first-aid stay using the Brazilian version of the Jones Dependency Tool (JDT) [1]. Data were collected in May and June 2018. To estimate the severity of trauma, the Revised Trauma Score (RTS), Injury Severity Score (ISS), New Injury Severity Score (NISS), and Trauma and Injury Severity Score (TRISS) indices were used. The severity was compared according to the type of trauma. The Cronbach's alpha was used to verify the reliability of the JDT.

Results
There was predominance of males (83.3%), blunt trauma (66.7%), and red screening category (79.2%). As for the mechanism of trauma, falls and penetrating injuries (firearms and knives) accounted for 33%. The mean age was higher among patients with blunt trauma (50.5 ± 23.6 years). There was an equivalence in the severity of trauma when the mean RTS, ISS, and NISS indices were analyzed. Mean TRISS was higher in patients with penetrating trauma (85.1 ± 24.8) ( Table 1). The reliability of the JDT was 0.91. In patients with blunt trauma, there was a predominance of high and moderate dependency of 40%. In patients with penetrating trauma, only high and total dependency levels were observed of 50% (Fig. 1).

Conclusions
In patients with blunt and penetrating trauma, dependency levels represented the demand for care required by patients in the ER. There was an indication of greater dependency in patients with greater trauma severity. The JDT proved to be valid and sensitive in the measurement of patient states. Background Patients who undergo high-risk surgical procedures represent a large proportion of intensive care unit (ICU) admissions. The accurate identification of patients who are at high risk of complications and death on ICU admission is important for the management of these patients [1][2][3]. Objective The aim of this study was to assess the prognostic factors in the first 24 h of admission for ICU mortality in a surgical ICU.

Methods
This was a prospective cohort study conducted on patients admitted to the surgery ICU of the Hospital Santa Luzia Rede D'Or, Brasília, DF, Brazil, over 2 years. Patients transferred to another ICU were excluded.

Results
Of 1185 patients included, the mean age was 57 ± 17 years, the Acute Physiology and Chronic Health Evaluation (APACHE) II score was 8 ± 5, the Simplified Acute Physiology II score (SAPS 2) was 23 ± 12, and 91% underwent elective surgery. ICU mortality was 3.5%. The nonsurvivors were older (68±20 vs 56±16, p<0. Background Early mobilization is a fundamental aspect of the recovery of critical patients, both clinical and surgical [1], and the functional evaluation using a metric instrument is of great importance.

Objective
The objective of this study was to compare mobility between surgical and clinical patients at intensive care unit (ICU) admission versus discharge. The secondary objective was to stratify the evolution of mobility barriers between admission and discharge from the ICU.

Methods
This was a retrospective study using a physiotherapy database from the Department of Critical Care of a private hospital in São Paulo, from July 2017 to January 2018. The mobility was evaluated through the Perme Intensive Care Unit Mobility Score instrument at admission and discharge from the ICU [2]. Patients were older than 18 years and hospitalized for at least 24 h in intensive care; patients with previous functional dependence, those who died during intensive care hospitalization, or those with Perme Score instrument filled out incorrectly were excluded. The study was approved by the hospital's Ethics Committee. There was no consent form as data were extracted from a database where patients were not identified, making it impossible to obtain consent.

Results
The study included 628 patients, the majority of whom were males (n = 337, 53.6%), with a mean age of 70 ± 17.8 years, and were mostly clinical patients (n = 464, 73.8%). The main barriers were mental status, the use of invasive and noninvasive mechanical ventilation, and pain. There was an improvement in mobility at discharge from intensive care compared with admission (17.3 ± 8.8 versus 9.1 ± 7.2, p (17.3 ± 8.8 x 9.1 ± 7.2, p <0.001). Surgical patients had a higher mobility than the clinical patients (19 ± 8.3 x 16.7 ± 8.9; p = 0.001) on discharge day ( Figure 1). All barriers decreased until discharge from the intensive care unit. Factors associated with improved mobility: surgical patients, decreased level of consciousness and no ventilatory support at admission.

Conclusions
Patients presented improvement in their mobility status upon discharge from intensive care compared with admission; surgical patients had higher levels of mobility than patients, and all items considered as barriers decreased during hospitalization.