Predictors of Readmission in Heart Failure Patients Vary by Cause of Readmission: Hospital-Level Cause-Specific Readmission Rates Show No Correlation

Heart failure (HF) is a common condition with an associated high mortality and morbidity. Many patients are elderly and comorbidity is common. Readmission metrics typically combine all causes and look at 30 days, an arbitrary period. Less is known about other follow-up lengths, whether the predictors change over time or if combining all causes of readmission loses key information. We used national linked English data on 74,629 patients discharged alive from their first (index) HF admission during 2008-10. Following them up within the data set for a year, we found that several predictors for readmission for HF differ from those for other causes and that the effect of length of stay is complicated. This is reflected in the lack of correlation in hospital-level rates for HF and non-HF causes. This suggests that heart failure patients are not always managed holistically. Consideration should be given to stratifying by readmission cause.