Predictors of fifty days in-hospital mortality in decompensated cirrhosis patients with spontaneous bacterial peritonitis

AIM: To determine the predictors of 50 d in-hospital mortality in decompensated cirrhosis patients with spontaneous bacterial peritonitis (SBP). METHODS: Two hundred and eighteen patients admitted to an intensive care unit in a tertiary care hospital between June 2013 and June 2014 with the diagnosis of SBP (during hospitalization) and cirrhosis were retrospectively analysed. SBP was diagnosed by abdominal paracentesis in the presence of polymorphonuclear cell count ≥ 250 cells/mm(3) in the peritoneal fluid. Student's t test, multivariate logistic regression, cox proportional hazard ratio (HR), receiver operating characteristics (ROC) curves and Kaplan-Meier survival analysis were utilized for statistical analysis. Predictive abilities of several variables identified by multivariate analysis were compared using the area under ROC curve. P < 0.05 were considered statistical significant. RESULTS: The 50 d in-hospital mortality rate attributable to SBP is 43.11% (n = 94). Median survival duration for those who died was 9 d. In univariate analysis acute kidney injury (AKI), hepatic encephalopathy, septic shock, serum bilirubin, international normalized ratio, aspartate transaminase, and model for end-stage liver disease - sodium (MELD-Na) were significantly associated with in - hospital mortality in patients with SBP (P ≤ 0.001). Multivariate cox proportional regression analysis showed AKI (HR = 2.16, 95%CI: 1.36-3.42, P = 0.001) septic shock (HR = 1.73, 95%CI: 1.05-2.83, P = 0.029) MELD-Na (HR = 1.06, 95%CI: 1.02-1.09, P ≤ 0.001) was significantly associated with 50 d in-hospital mortality. The prognostic accuracy for AKI, MELD-Na and septic shock was 77%, 74% and 71% respectively associated with 50 d in-hospital mortality in SBP patients. CONCLUSION: AKI, MELD-Na and septic shock were predictors of 50 d in-hospital mortality in decompensated cirrhosis patients with SBP.