New Study Challenges Conventional Wisdom on Pay-for-Performance Incentive Programs

RWJF Investigator tests impact of incentives for clinicians, and draws conclusions with implications for health reform.

    • September 19, 2013

One of the many approaches the Affordable Care Act takes to making health care more efficient is pay-for-performance incentives for clinicians. The underlying idea is that if doctors, nurses, physician assistants, and others are judged and rewarded based on the quality of the care they provide and actual patient outcomes, more of the necessary interventions will occur while fewer unnecessary tests and procedures will be performed, saving money and improving the quality of care.

Various studies have tested the concept, reaching mixed results. But two new studies by Robert Wood Johnson Foundation (RWJF) Investigator Award in Health Policy Research recipient R. Adams Dudley, MD, MBA, and colleagues suggest that for incentives to be effective, they must be carefully targeted and designed.

Dudley's two studies were published this month in the Journal of the American Medical Association. The first tests the effectiveness of incentives for small medical practices that have implemented electronic health records (EHRs). The second focuses on how incentives should be distributed to achieve maximum effect—to individual clinicians, to their practices, or to both.

Small Practices, Big Reach

"Most pay-for-performance research has been on large group practices, but prior research done on small practices suggests that their overall performance is worse, leading some to conclude that they might not be able to improve," Dudley says. "Our study was different because it's based in small practices that had all implemented EHRs that featured clinical decision support and other features necessary to identify patients with certain needs," thus creating some of the information infrastructure such practices have lacked in the past.

"In the days before EHRs," Dudley explains, “practices that wanted to focus on controlling blood pressure for patients with diabetes, or on asthmatics whose condition had landed them in the emergency room, would assign a staff member to go through patient charts by hand.” For small practices, such a task is a considerable burden, but as Dudley points out, if the data is in a computer, "it doesn't take a large-scale practice to find all the diabetics with blood pressure problems. By making it easier to find information, we make it easier for clinicians to improve their performance, at least in theory."

Dudley and colleagues focused on New York City primary care clinics with fewer than 10 clinicians, taking advantage of a local initiative that had provided EHR programs and training to such practices. Half of the 84 participating clinics received incentives; half did not. Incentives were focused on treatment of cardiovascular risk, gauging whether clinicians provided recommended care for patients, as well as achieving intermediate outcomes for those patients—control of blood pressure and cholesterol, and smoking cessation.

Significantly, the study included patients who were "harder to treat because of comorbidities or insurance status." Rather than dropping such patients from the study, as some incentive programs do, the study design provided extra rewards for clinicians who made progress with them. "We were trying to build an incentive system that's consistent with how doctors think," Dudley says, by paying more for what's harder to accomplish. "Each time a patient walked into the room, we wanted doctors to think, 'If I do it right, if I get the right outcome, I get an incentive.'" Dudley describes such features as "very, very different than any incentive program that's been implemented elsewhere."

In their test, Dudley and his colleagues write, "we found that EHR-enabled small practices were able to respond to incentives to improve cardiovascular care processes and intermediate outcomes," contradicting assumptions to the contrary. As the authors note, 82 percent of U.S. physicians practice in groups of fewer than 10 clinicians, and since EHRs are almost certainly the way of the future, the results underscore the value of incentive programs. But Dudley points out that the study also demonstrates the importance of what exactly is measured and whether that aligns with how clinicians think about treating their patients.

Who Gets the Incentive?

In the second study, Dudley and colleagues—led by Laura A. Petersen, MD, MPH, a National Advisory Committee member and Veterans Affairs (VA) Liaison for the RWJF Generalist Physician Faculty Scholars Program—tested whether incentives are most effective when aimed at individual physicians, their entire practices, or both.

This time, the study focused on treatment of hypertension in 12 hospital-based clinics in VA health networks. Clinics were assigned randomly to one of four study groups. One group featured incentives for individual physicians; one offered incentives to the practice, which then distributed money equally among all of its participating providers, including non-physicians; one provided incentives to both the physician and the practice; and one provided no incentives at all. The study ran for 20 months, and researchers gathered data for 16 months afterward to see if improvements outlived the incentives.

Over the course of the study, researchers provided all clinicians with reports on their performances. Clinicians earned incentives when patients achieved guideline-recommended blood pressure levels or, that failing, if clinicians had responded appropriately to uncontrolled blood pressure. The incentives averaged about 1.6 percent of physicians’ salaries, a dollar figure chosen by VA administrators to approximate "real-world" budget constraints.

The results suggested that incentives are most effective when they go directly to clinicians. Dudley and colleagues write, "We found that physicians who were randomized to the individual incentive group were more likely than [those in the control group] to improve their treatment of hypertension as measured by achievement of blood pressure control or appropriate response to uncontrolled blood pressure."

Taken together, Dudley says, the two studies undercut "the thought that doctors don't respond to incentives. These are two large, well-conducted trials that make clear that individual providers are pretty unequivocally improving their outcomes in response to financial incentives, and that even small practices can improve…. But it's also very important as we design incentives, to know what the organization is going to do with it, and to be sure that the incentives make sense in terms of how doctors think about their patients."

 

Related Websites

Learn more about the RWJF Investigator Awards in Health Policy Research.
Learn more about the RWJF Generalist Physician Faculty Scholars Program.
For an overview of RWJF scholar and fellow opportunities, visit www.RWJFLeaders.org.