Punished for Success? A Natural Experiment of Displaying Clinical Hospital Quality on Crowdsourced Review Platforms

Information Systems and Operations Management

Speaker : Paul PAVLOU  (University of Houston)

Room Bernard Ramanantsoa

Abstract

The healthcare market faces severe information asymmetry — patients struggle to evaluate the clinical quality of hospitals and make informed decisions. To inform patients about hospital selection, crowdsourced platforms (e.g., Yelp) have begun to display hospitals’ clinical quality (alongside hospital consumer reviews).  In 2017 and 2019, Yelp added a feature displaying maternity care clinical scores (e.g., C-section rates) of hospitals that deliver babies in select markets. We examine how clinical quality measures on Yelp – especially for high-quality hospitals (i.e., those that score high on clinical quality measures) – influence the subsequent consumer reviews of hospitals. Surprisingly, our difference-in-difference estimation shows that when quality scores are displayed, high-quality hospitals are actually punished in terms of subsequent lower consumer ratings and a lower review sentiment on Yelp, especially for those hospitals with low staffing capacity. This novel finding is important for hospitals as patient dissatisfaction has a significant effect on the federal funding of hospitals (CMS.gov). To tease out the underlying mechanism, we employed transfer deep learning and hospital demand proxies to show the observed phenomenon is likely due to a patient surge to hospitals with high maternity care clinical scores that results in staffing shortages and patient dissatisfaction. We contribute to the theory and practice by shedding light on the role of clinical quality measures in affecting patient satisfaction with hospitals.