Low serum albumin predicts early mortality in patients with severe hypoxic hepatitis

AIM: To evaluate the incidence, etiology, and predictors of mortality of severe hypoxic hepatitis. METHODS: We used computerized patient records to identify consecutive cases of severe hypoxic hepatitis admitted to a tertiary hospital in Singapore over a one-year period. We defined severe hypoxic hepatitis as elevation of serum transaminases more than 100 times upper limit of normal in the clinical setting of cardiac, circulatory or respiratory failure after exclusion of other causes of hepatitis. We used multivariable regression analysis to determine predictors for mortality. RESULTS: We identified 75 cases of severe hypoxic hepatitis out of 71380 hospital admissions over one year, providing an incidence of 1.05 cases per 1000 admissions. Median age was 65 years (range 19-88); 57.3% males. The most common etiologies of severe hypoxic hepatitis were acute myocardial infarction and sepsis. Fifty-three patients (71%) died during the hospitalization. The sole independent predictive factor for mortality was serum albumin measured at the onset of severe hypoxic hepatitis. Patients with low serum albumin of less than 28 g/L have more than five-fold increase risk of death (OR = 5.39, 95%CI: 1.85-15.71). CONCLUSION: Severe hypoxic hepatitis is uncommon but has a high mortality rate. Patients with low serum albumin are at highest risk of death.