Incidence and prognosis of early hepatic dysfunction in critically ill patients--a prospective multicenter study.

OBJECTIVE In critically ill patients, hepatic dysfunction is regarded as a late organ failure associated with poor prognosis. We investigated the incidence and prognostic implications of early hepatic dysfunction (serum bilirubin >2 mg/dL within 48 hrs of admission). DESIGN Prospective, multicenter cohort study. SETTING Thirty-two medical, surgical, and mixed intensive care units. PATIENTS A total of 38,036 adult patients admitted consecutively over a period of 4 yrs. INTERVENTIONS None. MEASUREMENTS AND MAIN RESULTS Excluding patients with preexisting cirrhosis (n = 691; 1.8%) and acute or acute-on-chronic hepatic failure (n = 108, 0.3%), we identified 4,146 patients (10.9%) with early hepatic dysfunction. These patients had different baseline characteristics, longer median intensive care unit stays (5 vs. 3 days; p < .001) and increased hospital mortality (30.4% vs. 16.4%; p < .001). Hepatic dysfunction was also associated with higher observed-to-expected mortality ratios (1.02 vs. 0.91; p < .001). Multiple logistic regression analysis showed an independent mortality risk of hepatic dysfunction (odds ratio, 1.86; 95% confidence interval, 1.71-2.03; p < .001), which exceeded the impact of all other organ dysfunctions. A case-control study further confirmed these results: Patients with early hepatic dysfunction exhibited significantly increased raw and risk-adjusted mortality compared with control subjects. CONCLUSIONS Our results provide strong evidence that early hepatic dysfunction, occurring in 11% of critically ill patients, presents a specific and independent risk factor for poor prognosis.