 | The 131st Annual Meeting (November 15-19, 2003) of APHA |
|
| The 131st Annual Meeting (November 15-19, 2003) of APHA |
Yue Li, Department of Community and Preventive Medicine, University of Rochester, Medical Center, 601 Elmwood Avenue, Box 644, rochester, NY 14642, (585)273-2548, yue_li@urmc.rochester.edu, Dana B. Mukamel, PhD, Department of Community and Preventive Medicine, University of Rochester School of Medicine and Dentistry, 601 Elmwood Avenue, Box 644, Rochester, NY 14642, and David L. Weimer, University of Wisconsin Madison, 1255 Observatory Drive, Madison, WI 53706.
Managed care organizations (MCOs) typically contract only with a subset of the providers in their market. If the selection of providers for contracting is done unilaterally by MCOs - i.e. physicians always accept a MCO contract offer if made - the composition of provider panels is likely to be different from a bilateral contract process in which physicians may not accept a contract offer. The aim of this study is to empirically test the hypothesis that contracts between MCOs and specialists (cardiac surgeons) are made unilaterally by the MCO vs. the alternative that the choice is bilateral.
The main data source we used for analysis is the data set with information about contracts between New York State MCOs and cardiac surgeons offering coronary artery bypass graft (CABG) in 1998. We estimated models predicting the probability of a contract between a MCO and a cardiac surgeon conditional on MCO and surgeon characteristics. We compared three different model specifications: the binary probit and the panel probit specifications for the unilateral choice assumption, the bivariate probit specification for the bilateral assumption.
The Vuong test indicates that the bilateral model describes the contracting process better than both unilateral models (V=-53.9, P<0.01 comparing the bivariate probit to the binary probit and V=-49.5, P<0.01 comparing the bivariate probit to the panel probit). The same variables reached statistical significance at the 0.05 levels in all models but the effect size (the magnitude of the coefficients) was larger in the bivariate probit model. The probability of a contract increases with surgeon°¯s quality (low risk-adjusted mortality rate), surgeon years since graduation from medical school, and higher market share. Contracts were less likely when the MCO is a for-profit or if it is an HMO rather than an IPA. Based on the bilateral model we estimate the probability that the average MCO will offer a contract to the average surgeon at 0.85, and the probability that the surgeon will accept the contract at 0.78.
These findings suggest that the MCO/surgeon contracting process is likely a bilateral process in which MCOs select surgeons and make them an offer and then surgeons choose to accept or reject it. Since the composition of provider panels offered by MCOs depends on decisions made by both the MCO and the physician, policies designed to influence the composition of these panels, e.g. increase average quality, may need to address factors influencing both MCOs and physicians.
Learning Objectives:
Keywords: Managed Care, Quality of Care
Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.