To compare mortalities in patients with sepsis due to biliary tract infections (BTIs) and due to infections from other sources, and to identify independent predictors of mortality in these patients.METHODS
This study was part of a community-acquired bloodstream infection (BSI) study that prospectively collected comprehensive clinical, laboratory, and outcome data from 937 consecutive patients with microbiologically documented BSI in the emergency department. BTI was the confirmed source of 145 of the 937 BSIs. We determined the independent prognostic factors by evaluating the correlation between 30-day mortality and various factors, for example, comorbidity, clinical severity, related hepatobiliary complication, and decompressive procedures.RESULTS
Patients with biliary sepsis had a high percentage of Gram-negative (88.3%), polymicrobial (26.9%), and anaerobic infections (6.9%). The 30-day overall mortality was 11.7%. Cox proportional hazard regression analysis disclosed five significant independent predictors: acute renal failure (hazard ratio, 95% confidence interval: 6.86, 6.02–25.5), septic shock (5.83, 4.36–15.64), malignant obstruction (4.35, 1.89–12.96), direct type hyperbilirubinemia (1.26, 1.1–1.42), and Charlson score ≥6 (1.57, 1.12–2.22). Compared with the remaining 792 patients in the source population, patients with bacteremic BTI had significantly better prognosis (log-rank test, P= 0.007). Adjusting for age, comorbidity, and clinical severity, BTI was still independently associated with better 30-day survival (0.25–0.76).CONCLUSIONS
Though the mortality rate in patients with bacteremic BTI is substantial, survival is better than in those with bacteremia from other sources. The main prognostic factors identified in this study may help clinicians recognize patients at high risk for early mortality so that they can give prompt, appropriate treatment.