We empirically investigate the pattern of where heart attack patients are transferred between hospitals. Using 2011 Florida State Emergency Department and Inpatient Databases, we demonstrate the relative importance of three key factors in determining transfer destinations: (1) the distance between sending and receiving hospitals, (2) publicly-reported quality measures of receiving hospitals, and (3) the relationship between sending and receiving hospitals as identified by whether they are affiliated with the same multihospital system.
Our conditional logit analysis shows that hospital relationship plays a dominant role in the choice of transfer destinations, compared to distance and quality. This result is robust to three alternative specifications of choice sets using distance ranking, radius circles, and Hospital Referral Regions, and also robust to alternative measures of distance and quality. When using 30-day all-cause readmission rate to evaluate the health outcome of transferred patients, we find that relationship-based transfers are associated with a much worse outcome than distance-based and quality-based transfers. We also find that nonprofit hospitals are more likely to conduct quality-based transfers and less likely to conduct relationship-based transfers than their for-profit counterparts.
Our study calls for a reevaluation of current practice in interhospital transfer of heart attack patients — selecting transfer destinations based on quality or distance may potentially decrease hospital readmission rates.