Venue: The Fuqua School of Business, Duke University, 1 Towerview Drive, Durham, NC 27708-0120

 

Presentation

Overlap in HMO Physician Networks and Convergence in HMO Performance

Authors: Daniel D Maeng (Pennsylvania State University); Michael E. Chernew (Harvard University); Dennis P. Scanlon (Pennsylvania State University); Tim Gronniger (Congressional Budget Office); Walter P. Wodchis (University of Toronto); Catherine G. McLaughlin (University of Michigan)

Presenter: Daniel D. Maeng (Pennsylvania State University)

Discussant: Gilbert Gimm (Mathematica Policy Research)

Session: Physician Groups

Room: Seminar E

When: Monday 1 p.m. - 2:30 p.m.

One approach to reforming the health care system involves instilling more competition in health care markets. Some proposals focus on competition among health care providers while others, such as managed competition, focus on competition among health plans. The assumption behind the models of competition at the plan level is that well-informed consumers can choose their plan based on costs, benefit structures, and quality and thereby reap the benefits commonly attributed to competitive markets. In reality, however, competition among health plans may be limited and choice of consumers illusory because of the non-exclusive relationships between plans and providers.

To the extent that health plans attempt to establish their optimal-size networks within each market while providers contract with multiple health plans to ensure patient base and revenue, purchasers and consumers are faced with a certain degree of 'overlap' in the provider networks that the available health plans offer. As the provider network overlap grows, the plans would presumably become less able to distinguish themselves from one another based on the provider networks they offer to consumers, resulting in reduced competitions among plans in terms of provider quality. With the rising popularity of Preferred Provider Organizations (PPO) and the corresponding decline in the staff-model HMOs in recent years, these provider network overlaps are likely to have increased significantly over the years.

This lack of plan competition at the provider level would be of particular concern if plan quality measures are largely determined by the plans' networks of providers rather than by the plans themselves. Existing measures of plan performances commonly include process measures - e.g. rates of prescription of beta blockers following a heart attack, retinopathy exam rates for diabetics, preventive screening test utilization rates such as mammography or preventive immunizations, etc. - that are likely to reflect physician practice styles and behaviors as well as the characteristics of the plan, patients, and the markets in which they operate. As such, comparisons across plans based on such measures may not, in practice, necessarily convey the information that purchasers, consumers and policy makers seek, limiting the usefulness of the measures.

In this paper, we explore the extent to which a provider network overlap is correlated with the convergence of plan performance measures using a unique data set that allows us to directly examine this relationship; that is, how much of the variation in plan performance measures can be explained by the provider network overlap? If plan performance measures do indeed largely capture physician characteristics rather than those of the plans, one may then expect to see a convergence of the measures in markets in which there is a large provider network overlap, suggesting a limited scope of plan competition based on provider choice and quality. Our results suggest that indeed a larger physician network overlap is significantly associated with a greater degree of convergence in HMO performance measures.