Causal Impact of the Hospital Readmissions Reduction Program on Hospital Readmissions and Mortality
Estimating causal effects of the Hospital Readmissions Reduction Program (HRRP), part of the Affordable Care Act, has been very controversial. Associational studies have demonstrated decreases in hospital readmissions, consistent with the intent of the program, although analyses with different data sources and methods have differed in estimating effects on patient mortality. To address these issues, we define the estimands of interest in the context of potential outcomes, we formalize a Bayesian structural time-series model for causal inference, and discuss the necessary assumptions for estimation of effects using observed data. The method is used to estimate the effect of the passage of HRRP on both the 30-day readmissions and 30-day mortality. We show that for acute myocardial infarction and congestive heart failure, HRRP caused reduction in readmissions while it had no statistically significant effect on mortality. However, for pneumonia, HRRP had no statistically significant effect on readmissions but caused an increase in mortality.
READ FULL TEXT