Predicting Poor Outcomes in Heart Failure

摘要: Background: Health plans must prioritize disease management efforts to reduce hospitalization and mortality rates in heart failure patients. Methods Results: We developed a risk model predict the 5-year of or for among patients at large health maintenance organization. identified 4696 who had an echocardiogram diagnosis from 1999 2004. We observed 56% five-year death (95% confidence interval, 54% 58%). The hazard ratios data contributed statistically significantly model, but findings did not improve our ability accurately once we accounted demographic characteristics clinical findings. A more complex demonstrated modest capacity risk. Our discriminated highest- lowest-risk with limited success–the was 3 times higher highest quintile, compared quintile. Conclusions: Using available electronic records, series risk-prediction models poor outcomes failure. found that relatively simple is as effective all only accuracy. Until better prediction variables are patients, may be valuable prioritizing centralized program by stratifying according their absolute outcomes.