Objectives: Individuals without health insurance in general receive fewer health services and are more likely than insured patients to experience poor outcomes. The main goal of this research was to study whether physicians' clinical recommendations vary for insured and uninsured patients, implying that physicians' choices of care may mediate insurance-related differences in health care use.
Methods: The authors designed clinical scenarios that describe routine decisions encountered by primary care physicians in ambulatory settings. Scenarios were designed to include discretionary, nondiscretionary, preventive, and diagnostic/therapeutic services. Insurance status of patients was indicated as either insured or uninsured for the service under consideration. Scenarios were presented to a nationally representative sample of primary care physicians (n = 1182) as part of the American Medical Association 1992 Socio-economic Monitoring System Survey. Physicians were assigned randomly to receive eight scenarios in which patients were either insured or uninsured. For each scenario, physicians were asked to indicate the percentage of patients for whom they would recommend a given service.
Results: After controlling for variables associated with nonresponse, we found that physicians who were presented scenarios with insured patients recommended service for 72% of patients, and physicians who were presented scenarios with uninsured patients recommended the same services for 67% of patients (P < 0.001). Physicians recommended both discretionary services (50% versus 42%; P < 0.001) and nondiscretionary services more often for insured than uninsured patients (93% versus 91%; P < 0.05).
Conclusions: In self-reports, physicians are more likely to recommend services for insured than for uninsured patients, and more so when services are discretionary. This provides evidence that physicians' recommendations may be important mediators of insurance-related variation in the use of health-care services. Higher rates of use among the insured may not always reflect higher quality of care, particularly when the service is discretionary in nature.