Factors associated with carbon dioxide transfer in an experimental model of severe acute kidney injury and hypoventilation during high bicarbonate continuous renal replacement therapy and oxygenation membrane support.

OBJECTIVE: To investigate the factors influencing carbon dioxide transfer in a system that integrates an oxygenation membrane in series with high-bicarbonate continuous veno-venous hemodialysis in hypercapnic animals. METHODS: In an experimental setting, we induced severe acute kidney injury and hypercapnia in five female Landrace pigs. Subsequently, we initiated high (40mEq/L) bicarbonate continuous veno-venous hemodialysis with an oxygenation membrane in series to maintain a pH above 7.25. At intervals of 1 hour, 6 hours, and 12 hours following the initiation of continuous veno-venous hemodialysis, we performed standardized sweep gas flow titration to quantify carbon dioxide transfer. We evaluated factors associated with carbon dioxide transfer through the membrane lung with a mixed linear model. RESULTS: A total of 20 sweep gas flow titration procedures were conducted, yielding 84 measurements of carbon dioxide transfer. Multivariate analysis revealed associations among the following (coefficients ± standard errors): core temperature (+7.8 ± 1.6 °C, p < 0.001), premembrane partial pressure of carbon dioxide (+0.2 ± 0.1/mmHg, p < 0.001), hemoglobin level (+3.5 ± 0.6/g/dL, p < 0.001), sweep gas flow (+6.2 ± 0.2/L/minute, p < 0.001), and arterial oxygen saturation (-0.5 ± 0.2%, p = 0.019). Among these variables, and within the physiological ranges evaluated, sweep gas flow was the primary modifiable factor influencing the efficacy of low-blood-flow carbon dioxide removal. CONCLUSION: Sweep gas flow is the main carbon dioxide removal-related variable during continuous veno-venous hemodialysis with a high bicarbonate level coupled with an oxygenator. Other carbon dioxide transfer modulating variables included the hemoglobin level, arterial oxygen saturation, partial pressure of carbon dioxide and core temperature. These results should be interpreted as exploratory to inform other well-designed experimental or clinical studies.

Validation of a low-cost continuous renal replacement therapy dialysate fluid controller for experimental purposes.

BACKGROUND: Continuous renal replacement therapy (CRRT) support is crucial for critically ill patients and it is underexplored in specific situations. Experimental CRRT offers a means to gain insights into these scenarios, but the prohibitive cost of CRRT machines limits their accessibility. This study aimed to develop and validate a low-cost and precise dialysate controller for experimental CRRT. RESULTS: Our results demonstrate a commendable level of precision in affluent flow control, with a robust correlation (R2 = 0.99) for continuous flow and a strong correlation (R2 = 0.95) for intermittent flow. Additionally, we observed acceptable agreement with a bias = 3.4 mL (upper limit 95% = 43.9 mL and lower limit 95% = - 37 mL) for continuous flow and bias = - 20.9 mL (upper limit 95% = 54 mL and lower limit 95% = - 95.7 mL) for intermittent flow, in this way, offering a precise CRRT dose for the subjects. Furthermore, we achieved excellent precision in the cumulative ultrafiltration net (UFnet), with a bias = - 2.8 mL (upper limit 95% = 6.5 mL and lower limit 95% = - 12 mL). These results remained consistent even at low affluent flow rates of 8, 12, and 20 mL/min, which are compatible with CRRT doses of 25-30 mL/kg for medium-sized animals. Moreover, the acceptable precision of our findings persisted when the dialysate controller was subjected to high filter dialysate chamber pressure for an extended duration, up to 797 min. CONCLUSIONS: The low-cost dialysate controller developed and tested in this study offers a precise means of regulating CRRT in experimental settings. Its affordability and accuracy render it a valuable instrument for studying CRRT support in unconventional clinical scenarios, particularly in middle-income countries' experimental ICU laboratories.

Factors associated with carbon dioxide transfer in an experimental model of severe acute kidney injury and hypoventilation during high bicarbonate continuous renal replacement therapy and oxygenation membrane support / Fatores associados à transferência de dióxido de carbono em um modelo experimental de lesão renal aguda grave e hipoventilação durante terapia de substituição renal contínua com alto teor de bicarbonato e suporte com membrana de oxigenação

ABSTRACT Objective To investigate the factors influencing carbon dioxide transfer in a system that integrates an oxygenation membrane in series with high-bicarbonate continuous veno-venous hemodialysis in hypercapnic animals. Methods In an experimental setting, we induced severe acute kidney injury and hypercapnia in five female Landrace pigs. Subsequently, we initiated high (40mEq/L) bicarbonate continuous veno-venous hemodialysis with an oxygenation membrane in series to maintain a pH above 7.25. At intervals of 1 hour, 6 hours, and 12 hours following the initiation of continuous veno-venous hemodialysis, we performed standardized sweep gas flow titration to quantify carbon dioxide transfer. We evaluated factors associated with carbon dioxide transfer through the membrane lung with a mixed linear model. Results A total of 20 sweep gas flow titration procedures were conducted, yielding 84 measurements of carbon dioxide transfer. Multivariate analysis revealed associations among the following (coefficients ± standard errors): core temperature (+7.8 ± 1.6 °C, p < 0.001), premembrane partial pressure of carbon dioxide (+0.2 ± 0.1/mmHg, p < 0.001), hemoglobin level (+3.5 ± 0.6/g/dL, p < 0.001), sweep gas flow (+6.2 ± 0.2/L/minute, p < 0.001), and arterial oxygen saturation (-0.5 ± 0.2%, p = 0.019). Among these variables, and within the physiological ranges evaluated, sweep gas flow was the primary modifiable factor influencing the efficacy of low-blood-flow carbon dioxide removal. Conclusion Sweep gas flow is the main carbon dioxide removal-related variable during continuous veno-venous hemodialysis with a high bicarbonate level coupled with an oxygenator. Other carbon dioxide transfer modulating variables included the hemoglobin level, arterial oxygen saturation, partial pressure of carbon dioxide and core temperature. These results should be interpreted as exploratory to inform other well-designed experimental or clinical studies.

RESUMO Objetivo Investigar os fatores que influenciam a transferência de dióxido de carbono em um sistema que integra uma membrana de oxigenação em série com terapia de substituição renal contínua com alto teor de bicarbonato em animais hipercápnicos. Métodos Em um ambiente experimental, induzimos lesão renal aguda grave e hipercapnia em cinco porcos Landrace fêmeas. Em seguida, iniciamos terapia de substituição renal contínua com alto teor de bicarbonato (40mEq/L) com uma membrana de oxigenação em série para manter o pH acima de 7,25. Em intervalos de 1 hora, 6 horas e 12 horas após o início da terapia de substituição renal contínua, realizamos uma titulação padronizada do fluxo de gás de varredura para quantificar a transferência de dióxido de carbono. Avaliamos os fatores associados à transferência de dióxido de carbono através da membrana pulmonar com um modelo linear misto. Resultados Realizamos 20 procedimentos de titulação do fluxo de gás de varredura, produzindo 84 medições de transferência de dióxido de carbono. A análise multivariada revelou associações entre os seguintes itens (coeficientes ± erros padrão): temperatura central (+7,8 ± 1,6 °C, p < 0,001), pressão parcial pré-membrana de dióxido de carbono (+0,2 ± 0,1mmHg, p < 0,001), nível de hemoglobina (+3,5 ± 0,6g/dL, p < 0,001), fluxo de gás de varredura (+6,2 ± 0,2L/minuto, p < 0,001) e saturação de oxigênio (-0,5% ± 0,2%, p = 0,019). Entre essas variáveis, e dentro das faixas fisiológicas avaliadas, o fluxo do gás de varredura foi o principal fator modificável que influenciou a eficácia da remoção de dióxido de carbono de baixo fluxo sanguíneo. Conclusão O fluxo do gás de varredura é a principal variável relacionada à remoção de dióxido de carbono durante a terapia de substituição renal contínua com um alto nível de bicarbonato acoplado a um oxigenador. Outras variáveis moduladoras da transferência de dióxido de carbono incluíram o nível de hemoglobina, a saturação de oxigênio, a pressão parcial de dióxido de carbono e a temperatura central. Esses resultados devem ser interpretados como exploratórios para informar outros estudos experimentais ou clínicos bem planejados.

Adverse events leading to intensive care unit admission in a low-and-middle-income-country: A prospective cohort study and a systematic review.

INTRODUCTION: Adverse events (AE) are frequent in critical care and could be even more prevalent in LMIC due to a shortage of ICU beds and Human resources. There is limited data on how relevant AE are among the reasons for ICU admission, being all of which published by High-Income-Countries services. Our main goal is to describe the rate of adverse events-related ICU admissions and their preventability in a LMIC scenario, comparing our results with previous data. METHODS: This was a prospective cohort study, during a one-year period, in two general ICUs from a tertiary public academic hospital. Our exposure of interest was ICU admission related to an AE in adult patients, we further characterized their preventability and clinical outcomes. We also performed a systematic review to identify and compare previous published data on ICU admissions due to AE. RESULTS: Among all ICU admissions, 12.1% were related to an AE (9.8% caused by an AE, 2.3% related but not directly caused by an AE). These ICU admissions were not associated with a higher risk of death, but most of them were potentially preventable (70.9% of preventability rate, representing 8.6% of all ICU admissions). The meta-analysis resulted in a proportion of ICU admissions due to AE of 11% (95% CI 6%-16%), with a preventability rate of 54% (95% CI 42%-66%). CONCLUSIONS: In this prospective cohort, adverse events were a relevant reason for ICU admission. This result is consistent with data retrieved from non-LMIC as shown in our meta-analysis. The high preventability rate described reinforces that quality and safety programs could work as a tool to optimize scarce resources.

Intensivist-led ultrasound-guided percutaneous tracheostomy: a phase IV cohort study / Traqueostomia percutânea guiada por ultrassom conduzida por intensivistas: um estudo de coorte de fase IV

ABSTRACT Objective: To describe, with a larger number of patients in a real-world scenario following routine implementation, intensivist-led ultrasound-guided percutaneous dilational tracheostomy and the possible risks and complications of the procedure not identified in clinical trials. Methods: This was a phase IV cohort study of patients admitted to three intensive care units of a quaternary academic hospital who underwent intensivist-led ultrasound-guided percutaneous tracheostomy in Brazil from September 2017 to December 2021. Results: There were 4,810 intensive care unit admissions during the study period; 2,084 patients received mechanical ventilation, and 287 underwent tracheostomy, 227 of which were performed at bedside by the intensive care team. The main reason for intensive care unit admission was trauma, and for perform a tracheostomy it was a neurological impairment or an inability to protect the airways. The median time from intubation to tracheostomy was 14 days. Intensive care residents performed 76% of the procedures. At least one complication occurred in 29.5% of the procedures, the most common being hemodynamic instability and extubation during the procedure, with only 3 serious complications. The intensive care unit mortality was 29.1%, and the hospital mortality was 43.6%. Conclusion: Intensivist-led ultrasound-guided percutaneous tracheostomy is feasible out of a clinical trial context with outcomes and complications comparable to those in the literature. Intensivists can acquire this competence during their training but should be aware of potential complications to enhance procedural safety.

RESUMO Objetivo: Descrever, com um número maior de pacientes em um cenário do mundo real após a implementação rotineira, a traqueostomia percutânea guiada por ultrassom conduzida por intensivistas e os possíveis riscos e complicações do procedimento não identificados em estudos clínicos. Métodos: Trata-se de estudo de coorte de fase IV de pacientes internados em três unidades de terapia intensiva de um hospital acadêmico quaternário que foram submetidos a traqueostomia percutânea guiada por ultrassom conduzida por intensivistas no Brasil de setembro de 2017 a dezembro de 2021. Resultados: Entre as 4.810 admissões na unidade de terapia intensiva durante o período do estudo, 2.084 pacientes receberam ventilação mecânica, e 287 foram submetidos a traqueostomia, 227 das quais foram realizadas à beira do leito pela equipe de terapia intensiva. O principal motivo para a admissão na unidade de terapia intensiva foi trauma, e para a realização de uma traqueostomia foi comprometimento neurológico ou incapacidade de proteger as vias aéreas. O tempo médio entre a intubação e a traqueostomia foi de 14 dias. Residentes de terapia intensiva realizaram 76% dos procedimentos. Ao menos uma complicação ocorreu em 29,5% dos procedimentos, sendo instabilidade hemodinâmica e extubação durante o procedimento as complicações mais comuns, com apenas três complicações graves. A mortalidade na unidade de terapia intensiva foi de 29,1%, e a mortalidade hospitalar foi de 43,6%. Conclusão: A traqueostomia percutânea guiada por ultrassom conduzida por intensivistas é viável fora do contexto de um estudo clínico com resultados e complicações comparáveis aos da literatura. Os intensivistas podem adquirir essa competência durante seu treinamento, mas devem estar cientes das possíveis complicações para aumentar a segurança do procedimento.

Association of fasting in the first 72 h of intensive care unit stay with outcomes of critically ill patients.

BACKGROUND: Whether fasting early in critical illness course is acceptable is not clear and high-quality data on this topic are lacking. To generate equipoise for future clinical trials and bring additional data to current literature, we compared outcomes of patients fasted during the first 72 h of intensive care unit (ICU) stay to patients receiving any nutrition support during this period. METHODS: Retrospective cohort study of a medical ICU from a tertiary academic center in Brazil. Adult patients treated between November 2017 and February 2022 with an ICU length of stay of ≥5 days were included. Baseline and daily data were retrieved from the prospectively collected administrative database. We did 1:1 propensity score matching to compare patients fasting for at least 72 h with controls. Primary outcome was hospital mortality and secondary outcomes were other resources' use. RESULTS: During the study period, 1591 patients were cared for in this ICU, of which 998 stayed ≥5 days. After excluding readmissions and propensity score matching, 93 patients in the fasting group were matched to 93 controls. Hospital mortality was similar between fasting and matched control groups (odds ratio = 1.04; 95% CI = 0.56-1.94; P > 0.99). Secondary outcomes were not different between groups, including length of stay, days on mechanical ventilation, and incidence of new infections. CONCLUSION: Withholding nutrition support in the first 72 h of ICU stay was not associated with worse outcomes in this cohort of severe critically ill patients.

Intensivist-led ultrasound-guided percutaneous tracheostomy: a phase IV cohort study.

OBJECTIVE: To describe, with a larger number of patients in a real-world scenario following routine implementation, intensivist-led ultrasound-guided percutaneous dilational tracheostomy and the possible risks and complications of the procedure not identified in clinical trials. METHODS: This was a phase IV cohort study of patients admitted to three intensive care units of a quaternary academic hospital who underwent intensivist-led ultrasound-guided percutaneous tracheostomy in Brazil from September 2017 to December 2021. RESULTS: There were 4,810 intensive care unit admissions during the study period; 2,084 patients received mechanical ventilation, and 287 underwent tracheostomy, 227 of which were performed at bedside by the intensive care team. The main reason for intensive care unit admission was trauma, and for perform a tracheostomy it was a neurological impairment or an inability to protect the airways. The median time from intubation to tracheostomy was 14 days. Intensive care residents performed 76% of the procedures. At least one complication occurred in 29.5% of the procedures, the most common being hemodynamic instability and extubation during the procedure, with only 3 serious complications. The intensive care unit mortality was 29.1%, and the hospital mortality was 43.6%. CONCLUSION: Intensivist-led ultrasound-guided percutaneous tracheostomy is feasible out of a clinical trial context with outcomes and complications comparable to those in the literature. Intensivists can acquire this competence during their training but should be aware of potential complications to enhance procedural safety.

Investigação da associação de dois oxigenadores em paralelo ou em série durante o suporte respiratório com oxigenação por membrana extracorpórea / Exploring the association of two oxygenators in parallel or in series during respiratory support using extracorporeal membrane oxygenation

RESUMO Objetivo: Caracterizar as pressões, as resistências, a oxigenação e a eficácia da descarboxilação de dois oxigenadores associados em série ou em paralelo durante o suporte com oxigenação veno-venosa por membrana extracorpórea. Métodos: Usando os resultados de insuficiência respiratória grave em suínos associada à disfunção de múltiplos órgãos, ao modelo de suporte com oxigenação por membrana extracorpórea veno-venosa e à modelagem matemática, exploramos os efeitos na oxigenação, descarboxilação e pressões do circuito de associações de oxigenadores em paralelo e em série. Resultados: Testaram-se cinco animais com peso mediano de 80kg. Ambas as configurações aumentaram a pressão parcial de oxigênio após os oxigenadores. O teor de oxigênio da cânula de retorno também foi ligeiramente maior, mas o efeito na oxigenação sistêmica foi mínimo, usando oxigenadores com alto fluxo nominal (~ 7L/minuto). Ambas as configurações reduziram significativamente a pressão parcial de dióxido de carbono sistêmico. Como o fluxo sanguíneo na oxigenação por membrana extracorpórea aumentou, a resistência do oxigenador diminuiu inicialmente, com aumento posterior, com fluxos sanguíneos mais altos, mas pouco efeito clínico. Conclusão: A associação de oxigenadores em paralelo ou em série durante o suporte com oxigenação veno-venosa por membrana extracorpórea proporciona um modesto aumento na depuração da pressão parcial de dióxido de carbono, com leve melhora na oxigenação. O efeito das associações de oxigenadores nas pressões de circuitos extracorpóreos é mínimo.

ABSTRACT Objective: To characterize the pressures, resistances, oxygenation, and decarboxylation efficacy of two oxygenators associated in series or in parallel during venous-venous extracorporeal membrane oxygenation support. Methods: Using the results of a swine severe respiratory failure associated with multiple organ dysfunction venous-venous extracorporeal membrane oxygenation support model and mathematical modeling, we explored the effects on oxygenation, decarboxylation and circuit pressures of in-parallel and in-series associations of oxygenators. Results: Five animals with a median weight of 80kg were tested. Both configurations increased the oxygen partial pressure after the oxygenators. The return cannula oxygen content was also slightly higher, but the impact on systemic oxygenation was minimal using oxygenators with a high rated flow (~ 7L/minute). Both configurations significantly reduced the systemic carbon dioxide partial pressure. As the extracorporeal membrane oxygenation blood flow increased, the oxygenator resistance decreased initially with a further increase with higher blood flows but with a small clinical impact. Conclusion: Association of oxygenators in parallel or in series during venous-venous extracorporeal membrane oxygenation support provides a modest increase in carbon dioxide partial pressure removal with a slight improvement in oxygenation. The effect of oxygenator associations on extracorporeal circuit pressures is minimal.

Evaluation of eleven immunochromatographic assays for SARS-CoV-2 detection: investigating the dengue cross-reaction.

COVID-19 disease is spread worldwide and diagnostic techniques have been studied in order to contain the pandemic. Immunochromatographic (IC) assays are feasible and a low-cost alternative especially in low and middle-income countries, which lack structure to perform certain diagnostic techniques. Here we evaluate the sensitivity and specificity of eleven different IC tests in 145 serum samples from confirmed cases of COVID-19 using RT-PCR and 100 negative serum samples from blood donors collected in February 2019. We also evaluated the cross-reactivity with dengue using 20 serum samples from patients with confirmed diagnosis for dengue collected in early 2019 through four different tests. We found high sensitivity (92%), specificity (100%) and an almost perfect agreement (Kappa 0.92) of IC assay, especially when we evaluated IgG and IgM combined after 10 days from the onset of symptoms with RT-PCR. However, we detected cross-reactivity between dengue and COVID-19 mainly with IgM antibodies (5 to 20% of cross-reaction) and demonstrated the need for better studies about diagnostic techniques for these diseases.

Updates on aneurysmal subarachnoid hemorrhage: is there anything really new?

BACKGROUND: Aneurysmal subarachnoid hemorrhage (aSAH) is a severe disease, with systemic involvement and complex diagnosis and treatment. Since the current guidelines were published by the AHA/ASA, Neurocritical Care Society and the European Stroke Organization in 2012-2013,there has been an evolution in the comprehension of SAH-associated brain injury and its multiple underlying mechanisms. As a result, several clinical and translational trials were developed or are underway. Objective: The aim of this article is to review some updates in the diagnosis and treatment of neurological complications of SAH. Methods: A review of PubMed (May, 2010 to February, 2022) was performed. Data was summarized. Results: Content of five meta-analyses, nine review articles and 23 new clinical trials, including pilots, were summarized. Conclusions:Advances in the comprehension of pathophysiology and improvements in critical care have been reflected in the reduction of mortality in SAH. However, despite the number of publications, the only treatments shown to be effective in adequate, well-controlled clinical trials are nimodipine and repair of the ruptured aneurysm. Thus, doubts about the optimal management of SAH still persist.

Comparison of the accuracy of residents, senior physicians and surrogate decision-makers for predicting hospital mortality of critically ill patients. / Comparação da acurácia de residentes, médicos seniores e decisores substitutos na previsão da mortalidade hospitalar de pacientes críticos.

OBJECTIVE: To compare the predictive performance of residents, senior intensive care unit physicians and surrogates early during intensive care unit stays and to evaluate whether different presentations of prognostic data (probability of survival versus probability of death) influenced their performance. METHODS: We questioned surrogates and physicians in charge of critically ill patients during the first 48 hours of intensive care unit admission on the patient's probability of hospital outcome. The question framing (i.e., probability of survival versus probability of death during hospitalization) was randomized. To evaluate the predictive performance, we compared the areas under the ROC curves (AUCs) for hospital outcome between surrogates and physicians' categories. We also stratified the results according to randomized question framing. RESULTS: We interviewed surrogates and physicians on the hospital outcomes of 118 patients. The predictive performance of surrogate decisionmakers was significantly lower than that of physicians (AUC of 0.63 for surrogates, 0.82 for residents, 0.80 for intensive care unit fellows and 0.81 for intensive care unit senior physicians). There was no increase in predictive performance related to physicians' experience (i.e., senior physicians did not predict outcomes better than junior physicians). Surrogate decisionmakers worsened their prediction performance when they were asked about probability of death instead of probability of survival, but there was no difference for physicians. CONCLUSION: Different predictive performance was observed when comparing surrogate decision-makers and physicians, with no effect of experience on health care professionals' prediction. Question framing affected the predictive performance of surrogates but not of physicians.

OBJETIVO: Comparar o desempenho preditivo de residentes, médicos seniores de unidades de terapia intensiva e decisores substitutos dos pacientes logo no início da internação na unidade de terapia intensiva e avaliar se diferentes apresentações de prognóstico (probabilidade de sobrevida versus probabilidade de óbito) influenciaram seus desempenhos. MÉTODOS: Os decisores substitutos e os médicos responsáveis pelos pacientes críticos foram questionados durante as primeiras 48 horas de internação na unidade de terapia intensiva sobre a probabilidade do desfecho hospitalar do paciente. O enquadramento da pergunta (isto é, a probabilidade de sobrevida versus a probabilidade de óbito durante a internação) foi randomizado. Para avaliar o desempenho preditivo, comparou-se a área sob a curva ROC para desfecho hospitalar entre as categorias decisores substitutos e médicos. Também estratificaram-se os resultados de acordo com o enquadramento da pergunta randomizado. RESULTADOS: Entrevistaram-se decisores substitutos e médicos sobre os desfechos hospitalares de 118 pacientes. O desempenho preditivo dos decisores substitutos foi significativamente inferior ao dos médicos (área sob a curva de 0,63 para decisores substitutos, 0,82 para residentes, 0,80 para residentes de medicina intensiva e 0,81 para médicos seniores de unidade de terapia intensiva). Não houve aumento no desempenho preditivo quanto à experiência dos médicos (ou seja, médicos seniores não previram desfechos melhor que médicos juniores). Os decisores substitutos pioraram seu desempenho de previsão quando perguntados sobre a probabilidade de óbito ao invés da probabilidade de sobrevida, mas não houve diferença entre os médicos. CONCLUSÃO: Observou-se desempenho preditivo diferente ao comparar decisores substitutos e médicos, sem qualquer efeito da experiência no prognóstico dos profissionais de saúde. O enquadramento da pergunta afetou o desempenho preditivo dos substitutos, mas não o dos médicos.

Comparação da acurácia de residentes, médicos seniores e decisores substitutos na previsão da mortalidade hospitalar de pacientes críticos / Comparison of the accuracy of residents, senior physicians and surrogate decision-makers for predicting hospital mortality of critically ill patients

RESUMO Objetivo: Comparar o desempenho preditivo de residentes, médicos seniores de unidades de terapia intensiva e decisores substitutos dos pacientes logo no início da internação na unidade de terapia intensiva e avaliar se diferentes apresentações de prognóstico (probabilidade de sobrevida versus probabilidade de óbito) influenciaram seus desempenhos. Métodos: Os decisores substitutos e os médicos responsáveis pelos pacientes críticos foram questionados durante as primeiras 48 horas de internação na unidade de terapia intensiva sobre a probabilidade do desfecho hospitalar do paciente. O enquadramento da pergunta (isto é, a probabilidade de sobrevida versus a probabilidade de óbito durante a internação) foi randomizado. Para avaliar o desempenho preditivo, comparou-se a área sob a curva ROC para desfecho hospitalar entre as categorias decisores substitutos e médicos. Também estratificaram-se os resultados de acordo com o enquadramento da pergunta randomizado. Resultados: Entrevistaram-se decisores substitutos e médicos sobre os desfechos hospitalares de 118 pacientes. O desempenho preditivo dos decisores substitutos foi significativamente inferior ao dos médicos (área sob a curva de 0,63 para decisores substitutos, 0,82 para residentes, 0,80 para residentes de medicina intensiva e 0,81 para médicos seniores de unidade de terapia intensiva). Não houve aumento no desempenho preditivo quanto à experiência dos médicos (ou seja, médicos seniores não previram desfechos melhor que médicos juniores). Os decisores substitutos pioraram seu desempenho de previsão quando perguntados sobre a probabilidade de óbito ao invés da probabilidade de sobrevida, mas não houve diferença entre os médicos. Conclusão: Observou-se desempenho preditivo diferente ao comparar decisores substitutos e médicos, sem qualquer efeito da experiência no prognóstico dos profissionais de saúde. O enquadramento da pergunta afetou o desempenho preditivo dos substitutos, mas não o dos médicos.

ABSTRACT Objective: To compare the predictive performance of residents, senior intensive care unit physicians and surrogates early during intensive care unit stays and to evaluate whether different presentations of prognostic data (probability of survival versus probability of death) influenced their performance. Methods: We questioned surrogates and physicians in charge of critically ill patients during the first 48 hours of intensive care unit admission on the patient's probability of hospital outcome. The question framing (i.e., probability of survival versus probability of death during hospitalization) was randomized. To evaluate the predictive performance, we compared the areas under the ROC curves (AUCs) for hospital outcome between surrogates and physicians' categories. We also stratified the results according to randomized question framing. Results: We interviewed surrogates and physicians on the hospital outcomes of 118 patients. The predictive performance of surrogate decisionmakers was significantly lower than that of physicians (AUC of 0.63 for surrogates, 0.82 for residents, 0.80 for intensive care unit fellows and 0.81 for intensive care unit senior physicians). There was no increase in predictive performance related to physicians' experience (i.e., senior physicians did not predict outcomes better than junior physicians). Surrogate decisionmakers worsened their prediction performance when they were asked about probability of death instead of probability of survival, but there was no difference for physicians. Conclusion: Different predictive performance was observed when comparing surrogate decision-makers and physicians, with no effect of experience on health care professionals' prediction. Question framing affected the predictive performance of surrogates but not of physicians.