Predicting in-hospital death for severe sepsis and septic shock: the role of troponin I

Autor: Capeci W, Falsetti L, Zaccone V, Tarquinio N, Martino M, Di Pentima C, Martini A, Fioranelli A, Nitti C, Viticchi G, Pellegrini F, Burattini M, Salvi A
Přispěvatelé: Capeci W, Falsetti L, Zaccone V, Tarquinio N, Martino M, Di Pentima C, Martini A, Fioranelli A, Nitti C, Viticchi G, Pellegrini F, Burattini M, Salvi A
Jazyk: angličtina
Rok vydání: 2018
Předmět:
Popis: Background: Severe sepsis and septic shock (SS) are often managed in internal and emergency medicine departments. Sequential organ failure assessment (SOFA) score is a validated prognostic tool. Troponin I (TnI) I has been proposed as a marker of worse prognosis for SS in some studies: we aimed to evaluate if TnI could predict in-hospital death independently of SOFA score. Methods:in the period 2015-2017 we enrolled all the consecutive patients admitted for SS in two internal medicine departments with specific expertise in critical care.We evaluated, at the admission: (1) SOFA score (2) TnI level (3) sex, age, PCR and procalcitonin (PCT), (4) length of in-hospital stay (5) comorbidities. The main outcome of the study was in-hospital death for SS. The best cutoff value for TnI and in-hospital death was evaluated with ROC curve analysis,adopting Youden index. We then prepared a Cox proportional Hazard model adopting (a) length of stay as time variable, (b) in-hospital death as main outcome, (c) SOFA score and TnI as predictors and (e) sex, age PCT and PCT as covariates. Results: 390 subjects (age:79,6±11,4;males:49,2%) with 144 (36,9%) deaths. Optimal cutoff for TnI was >0.315ng/ml. Cox proportional hazards model showed that (1) one-unit increase in SOFA score was associated to an increased hazard ratio of in-hospital death (HR:1.208;95%CI:1.134-1.287), (2) TnI predicted in-hospital death independently of SOFA score (HR:1.925;95%CI:1.278-2.902), even correcting for age, sex, PCR and PCT. Results TnI predicts in-hospital death in SS independently of SOFA score, age, sex, PCR and PCT.
Databáze: OpenAIRE