Predictors of mortality and hospitalization in women with heart failure in the Digitalis Investigation Group trial.

We performed a retrospective follow-up study of 1926 women with heart failure who participated in the Digitalis Investigation Group trial. Adjusted hazard ratios and 95% confidence intervals for covariates that were significant independent predictors of all-cause mortality were age in years (1.02; 1.01-1.03; P<0.001), NYHA class IH-IV (1.56; 1.31-1.87), diabetes (1.63; 1.36-1.95), glomerular filtration rate in mL/min/1.73 m(2) (0.99; 0.98-4).996; P=0.001), pulmonary congestion by chest x-ray (1.57; 1.22-2.02), left ventricular ejection fraction (LVEF) (0.99; 0.98-0.993; P<0.001), and use of digoxin (1.20; 1.02-1.42). Covariates that were significant independent predictors of hospitalization due to worsening heart failure were: nonwhite race (1.28; 1.03-1.58), NYHA class III-IV (1.55; 1.30-1.84), diabetes (1.75; 1.47-2.09), glomerular filtration rate as mL/min/1.73 m(2) (0.99; 0.986-0.996; P<0.001), pulmonary congestion by chest x-ray (1.42; 1.12-1.81), and use of ACE inhibitors (0.67; 0.49-0.91). Longer duration of heart failure, higher NYHA classes, diabetes, chronic kidney disease, pulmonary congestion, and lower LVEF were significant independent predictors of all-cause mortality, all-cause hospitalization, and hospitalization for heart failure in women with heart failure.