Factors influencing hospital readmission penalties: Are they really under hospitals’ control?

The Hospital Readmissions Reduction Program (HRRP) established by the Centers for Medicare & Medicaid Services (CMS) penalizes hospitals that have excessive readmission rates. Research is needed to determine if there are factors that influence readmission penalties but may be out of the control of hospitals. In this study, we compare a set of geo-demographic, hospital, and quality of care characteristics of hospitals penalized by HRRP with those not penalized, and determine these characteristics’ association with the likelihood of penalization as well as the extent of penalties. We collected and integrated data from multiple sources such as the HRRP Supplemental Data File, the Hospital Compare database, CMS Impact files, the HCAHPS survey, and the U.S. census. We followed a two-step estimation procedure. First, we estimated a logistic regression model to find the relationship between various characteristics and whether a hospital was penalized. Second, we estimated a linear regression model to find the relationship between these characteristics and the extent of penalties. Results show that both the likelihood and the extent of penalties varied across several characteristics related to geo-demographics (such as location in the census division, urban vs. rural location, and racial make-up of the area), hospitals (such as ownership, bed capacity, teaching status, and case mix index), and quality of care (such as quality of communication by doctors and information provided at the time of discharge). Although quality of care characteristics are under hospitals’ control and can be improved, geo-demographic and hospital characteristics are likely consistent over time and largely out of the control of hospitals. The study supports the case for a comprehensive revision of HRRP's scoring methodology to calculate readmission penalties.