What can physicians tell us about managed care tools?

Abstract

Over the past three decades, physician-directed managed care tools have been used extensively by health plans to influence physician practice and control health care costs. The success of these tools varies widely as do the specifics of how they are implemented. Organizations involved in implementation of managed care tools are evolving, as are the tools themselves. Gaining insight into physicians' views on how managed care tools impact quality of care could help in the development of more effective tools. The purpose of this thesis is to address two questions regarding physicians' attitudes towards managed care tools: Do physicians practicing in the same clinics have similar views of managed care tools? Do physicians' views of managed care tools differ across practice settings and organizational structures (e.g., types of clinics and health plans)? Using the literature on physician job attitudes and sociological and economic theory to guide the investigation, these questions are addressed by looking at the effects of physician, clinic and health plan characteristics on physicians' attitudes toward managed care tools. Data for this study come from the PEHP survey of physicians in 15 health plans and 5 major cities nationwide with supplementary health plan information from Interstudy (1997). The data was collected in 1998-99. The initial sample of 4,800 physicians was stratified to be half generalists and half specialists. The overall response rate was 68% and N = 3,459. The first part of the analysis consists of structural equation modeling (SEM) to test the new typology of managed care tools. The scales developed in the first part are then used as dependent variables in a hierarchical linear modeling (HLM) analysis to explore the role of physicians' work setting and health plan affiliation on attitudes toward managed care tools. Results of the SEM indicate that managed care tools can be described and classified in terms of the types of control they exert on physicians. Results of the HLM analysis suggest that physicians practicing in the same clinic and health plans share some similar attitudes toward managed care tools, but the majority of differences are still at the physician level. Physician characteristics included in this study do not explain these differences, leaving open the question of whether the majority of variation is due to real differences in how individual physicians experience managed care tools, or due to measurement error. Future analysis could clarify this issue if clinic associations can be identified more accurately and/or better physician level predictors are developed.