The More Connected the Physicians, the Better the Referrals? Evidence from Patient-Sharing Networks

The demand for tertiary health care in the United States is largely determined by primary care physicians or medical specialists who make referrals but not by patients themselves. Thus, ideally, referring physicians should make a decision on behalf of their patients and ensure that the patients receive best available care given their conditions. Assuming that physicians in general care about their patients’ health, referring physicians would be able to send their patients to better tertiary care providers if they have better access to information on providers available in the market and quality thereof. Such information can often be obtained or updated through interactions with other doctors, so the shape of physician social networks can influence the referral decisions. This study examines this possibility using the sample of Medicare beneficiaries undergoing coronary artery bypass graft included in the 20% Medicare Carrier files (2008-13), and information on cardiac surgeons who performed the procedures and referring physicians (mostly primary care physicians and cardiologists). Building on the existing literature, the social network of each referring physician is defined by the patient sharing pattern in Medicare claims data, and the following three measures are employed as key explanatory variables that capture the network structure: 1) number of physicians connected (adjusted degree), 2) tightness of the network (clustering coefficient), and 3) influence of individual physicians in the network (eigenvector centrality). The outcome variables include average quality of the surgeon pool to which referring physicians send their patients each year and sensitivity with respect to a change in quality. The quality is measured by in-hospital mortality rates and the sensitivity is obtained by simulating the change in choice probabilities entailed by a one standard deviation increase in the mortality rate for each surgeon in the patient’s choice set. The results suggest that physicians respond more sensitively to the quality change of surgeons when making referral decisions, if they play a more central role in information sharing in their networks (i.e., eigenvector centrality is higher). This implies that, in addition to individual factors that have been extensively studied in literature, interpersonal factors of primary care providers can also be an important determinant of health care demand.