The cost reduction induced by the incentives was concentrated in outpatient procedures and referrals to non-primary care physicians.
Americans spend more than $1 trillion dollars annually on health care. Because physicians play a central role in the allocation of these vast resources, financial incentives that influence physician decisionmaking have been the subject of high stakes litigation and intense public controversy.
Much of the attention surrounding physician incentives has focused on the contracts that managed care organizations write with the physicians in their network. Managed care organizations have become the dominant form of health insurance in the United States, based largely on the promise of controlling the cost of care through cleverly designed systems of financial and non-financial incentives.
Critics of managed care are concerned that financial incentives cause physicians to alter their treatment at the expense of quality. Proponents of managed care counter that incentives are necessary tools to control costs and that cost containment need not compromise care. Until now, no one has taken a close and detailed look into a managed care organizations' incentive system to determine who is right. This is just what NBER Research Associates Martin Gaynor and James Rebitzer, and their co-author Lowell Taylor have done in their paper Incentives in HMOs (NBER Working Paper No. 8522).
The authors present a case study of a health maintenance organization (HMO) that wrote incentive contracts with the roughly 1,000 primary care physicians in its network. During the period of study, 1994-7, the HMO utilized the common "gatekeeper" model, in which primary care physicians are held responsible for the medical utilization costs incurred by their patients. The incentive? If groups of primary care gatekeepers kept costs below actuarially determined target levels, they received a sizeable bonus. Physicians' strategies for reducing utilization costs included: 1) teaching patients to better manage chronic diseases to avoid hospital visits; 2) discouraging non-essential specialty referrals and "unnecessary" testing; and 3) reducing emergency room visits by offering patients extended office hours and more responsive answering services.
The HMO limited the effect of its cost-control bonus by imposing a "stop-loss" provision for seriously ill patients. For example, a very sick patient who incurred expenses of $100,000 in a year would only be factored in for a maximum of $15,000 of a physician's annual utilization costs. Thus, the provision removes incentives to withhold care from the sickest patients. Equally important, in 1997, the final year of the study, the HMO also introduced financial incentives linking bonus payments to certain quality targets.
The authors estimate that a typical physician in the network gained ten cents in income for every one dollar reduction in medical utilization costs. This incentive appears overall to have caused a 5 percent reduction in utilization costs, leading to total annual savings of more than $3 million. Consistent with the effect of the HMO's stop-loss provisions, financial incentives had little or no effect on in-hospital costs. Instead, the cost reduction induced by the incentives was concentrated in outpatient procedures and referrals to non-primary care physicians.
Analysis of the "quality" incentives suggests that physicians respond to financial incentives linked to measures of quality just as they do to incentives linked to cost-containment. Indeed, physician groups that were best at keeping costs below "target" levels were also best at hitting their quality targets. However, this finding does not mean that low-cost physicians necessarily provided higher quality care. The quality measures used by the HMO, while standard in the industry, focus on preventive care practices, such as the percentage of immunized children and women with recent mammograms. The authors caution that although their findings lead them to believe that incentives have an impact on achieving the target quality measures, the overall impact on the quality of care is uncertain. It is possible that physicians are responding to the HMO's incentives by producing higher measured quality and shirking on aspects of quality that aren't measured. Nonetheless, these findings raise the possibility that properly designed incentive systems needn't produce a reduction in quality.
Overall, the results of the study offer some comfort to both critics and supporters of HMOs. Physicians' clinical decisions do indeed differ from those they would choose in the absence of incentives. At the same time, these same incentives have a measurable effect in controlling costs. Perhaps more importantly, it seems that devices such as "stop-loss" provisions and "incentives for quality" may prove helpful in assuring the quality of care. The authors conclude that the incentive contracts HMOs write with their physicians are powerful management tools, whose impact likely depends on the details of their implementation.
-- Tracy Certo