Network analysis of Medicare EHR incentive payments

In 2011 the Centers for Medicare and Medicaid Services (CMS) began providing incentive payments to eligible physicians who demonstrated meaningful use of a certified electronic health record (EHR). The goal of this and other federal EHR incentive programs is to foster adoption and use of health information technology. Since 2011, over $9 billion has been spent on Medicare EHR incentive payments. This paper investigates how physician characteristics and their professional networks relate to the acceptance of EHR incentive payments. Specifically, we model a single node network of physicians in the U.S. and explore common measures of social network structures. We examine measures such as centrality, density, and clustering among physicians participating in the incentive program, as well as those not currently participating. The resulting network structure variables are then used in predictive models of physician participation in incentive program to investigate the relationship between individual and network characteristics and the acceptance of incentive payments. The analyses employ two data sets recently made available through CMS’s efforts to increase data availability and program transparency: the eligible professional EHR incentive participation and payment data and the national provider file maintained by CMS for quality reporting purposes. Network characteristics and econometric results are investigated at a national level by specialty, geography, and types of institutional affiliations. The physician network structure overlaid with the incentive payment data provides a depiction of the distribution of incentive payments allowing for a more detailed understanding of the payments that have been allocated. Furthermore, assuming incentive payments are a reasonable proxy for meaningful use of EHRs, the results of the predicative models allow for inferences regarding the characteristics and relationships associated with physicians’ utilization of health information technology (IT). This paper contributes to both the health IT diffusion literature, and to a better understanding of the effect of Federal health IT policy. Understanding physicians’ relationships among each other as well as with incentives has significant implications for learning how to promote health IT and the development of a national health IT infrastructure. The paper also identifies avenues for future research of medical innovation diffusion and health care utilization and quality.