In the coming years, assessing the impact of efforts to reduce hospital readmissions will be important to policy makers and hospitals. To inform such assessments, we sought to define preexisting trends in readmission rates for patients by level of clinical severity using a difference-in-differences analysis of Medicare inpatient claims data from 1997, 2002, and 2007 for patients with acute myocardial infarction and congestive heart failure. We also examined trends in length of stay, in-hospital mortality, and postdischarge mortality by severity level to provide additional context for interpreting readmission rate trends. From 1997 to 2007, the difference in readmission rates between the highest and lowest severity quartiles increased. Length of stay and in-hospital mortality decreased for all patients; however, postdischarge mortality increased for the highest-severity patients and decreased for the lowest-severity patients. Assessments of recent policy reforms and quality improvement programs should account for underlying differential trends in readmission rates based on patient severity.
Keywords: hospital care; quality measurement; readmission; risk adjustment.