Usefulness of the Direct and/or Total Bilirubin to Predict Adverse Outcomes in Patients With Acute Decompensated Heart Failure

Abnormal liver function test results are often observed in acute decompensated heart failure (ADHF). However, the prognostic value of bilirubin fractionation has not been elucidated. The prognostic value of direct bilirubin (DB), in comparison with total bilirubin (TB), was examined in 556 consecutive patients with ADHF. Patients with elevated DB showed mostly similar patient characteristics including signs of elevated right-sided pressure (frequent hepatomegaly, jugular venous distention, dilated inferior vena cava, and elevated gamma-glutamyltransferase) and decreased cardiac output (cold extremities, decreased pulse pressure, and lower blood pressure) and other parameters of heart failure (HF) severity (increased plasma renin activity, decreased sodium, total cholesterol, and ejection fraction) to elevated TB; however, only patients with elevated DB showed a significant difference in the frequency of HF history and alkaline phosphatase value. Kaplan-Meier analysis showed that patients with elevated DB had a significantly higher rate of the composite end point of all-cause mortality or HF readmission (p = 0.021) compared with those with normal DB, whereas patients with elevated TB did not show a statistically significant difference compared with those with normal TB (NS). A multivariate Cox hazards model showed that DB was an independent predictor of adverse events (adjusted hazard ratio 1.052, 95% confidence interval 1.001 to 1.099, p = 0.034), whereas TB was not (adjusted hazard ratio 1.017, 95% confidence interval 0.985 to 1.046, p = 0.27). Adding DB to existing prognostic variables resulted in higher C-statistics than adding TB (C-statistics: 0.670 to 0.675, 0.670 to 0.674, respectively). In conclusion, elevated DB in ADHF was an independent prognostic predictor that was superior to TB. DB may be useful for further risk stratification in ADHF.