CH B AR THE LEONARD DAVIS INSTITUTE R ES E RI EF of HEALTH ECONOMICS Penn LDI 20 1 16 . - No Effect of Financial Incentives to Physicians, Patients, or Both on Lipid Levels: A Randomized Clinical Trial David A. Asch, Andrea B. Troxel, Walter F. Stewart, Thomas D. Sequist, James B. Jones, AnneMarie G. Hirsch, Karen Hoffer, Jingsan Zhu, Wenli Wang, Amanda Hodolfski, Antoinette B. Frasch, Mark G. Weiner, Darra D. Finnerty, Meredith B. Rosenthal, Kelsey Gangemi, Kevin G. Volpp Journal of the American Medical Association, Volume 314, Number 18 LDI RESEARCH BRIEF KEY FINDINGS: Can financial incentives be used to reduce cholesterol levels in high-risk patients? This randomized trial says modest reductions can be achieved only by targeting incentives to both patients and physicians, not to one or the other. THE QUESTION Each patient in the three intervention groups was assigned a quarterly goal to reduce lipoprotein cholesterol (LDL-C) To whom should financial incentives be targeted to achieve levels, which should have been achievable if the patient was a desired clinical or health outcome? Physician and patient fully adherent to his or her prescribed medication. incentives are becoming more common, but they are rarely combined, and effectiveness of these approaches is not PCPs in the physician incentives group accrued quarterly well-established. Using insight from behavioral economics, payments of $256 for each enrolled patient who met the a research team led by LDI Senior Fellows David Asch quarterly goal, paid semi-annually. PCPs in the patient and Kevin Volpp sought to determine whether physician incentives group received no payments; instead, their financial incentives, patient incentives, or shared physician patients participated in an automatic daily lottery with and patient incentives are more effective in promoting eligibility based on having taken the statin the day before. medication adherence and reducing cholesterol levels of PCPs and patients in the shared incentives group followed patients at high risk for cardiovascular disease. the same incentive structure as in the PCP or patient-specific groups but with payments of half the size. Total possible THE STUDY payouts were the same for all incentive groups. Physicians and patients in the control group received no goal-based The researchers randomly assigned 340 primary care incentives, but all participants received small participation physicians (PCPs) from three large primary care practices in payments. The interventions continued for 12 months, and the northeastern United States to one of four study groups: patients were followed up for an additional three months. control, physician incentives, patient incentives, and shared physician-patient incentives. More than 1,500 patients, all THE FINDINGS at high risk of cardiovascular disease, participated and were allocated to the same group as their PCP. Only patients in the shared physician-patient incentives group achieved reductions in LDL-C levels statistically Patients received their prescribed daily dose of statins in different from those in the control group. an electronic pill bottle, which, when opened, wirelessly transmitted a signal to a web platform. Research to Improve the Nation’s Health System. DATA DRIVEN. POLICY FOCUSED. ldi.upenn.edu of medication by PCPs and patient adherence to that medication. While the study points to the incentive structure that had the greatest relative impact, the improvements were modest and the authors stress that further information is needed to Source: JAMA. 2015;314(18):1926-1935. doi:10.1001/jama.2015.14850 understand whether the approach represents good value. Further, one limitation of the study was a lack of a true After 12 months of the intervention, 49% of patients in the “usual-care” control that did not receive electronic pill bottles. shared physician-patient incentives group had achieved their Patients in the control group received electronic pill bottles LDL-C goal compared with 40% in physician incentives, and may have been more adherent than is typical because they 40% in patient incentives, and 36% in control. were under observation. Other limitations include a relatively small number of enrolled patients per physician, which limited Although medication adherence was higher in the shared the potential rewards for physicians and may have reduced incentive and patient incentive groups, it was low across their motivation to go after these awards. all the groups. However, patients who were already LDI RESEARCH BRIEF taking statins before the start of the study (less than half of participants) showed large increases in adherence with Asch DA, Troxel AB, Stewart WF, et al. Effect of Financial Incentives to Physicians, Patients, or Both on Lipid Levels: A Randomized Clinical Trial. incentives. This suggests that the incentive was ineffective JAMA. 2015;314(18):1926-1935. doi:10.1001/jama.2015.14850. in promoting initiation of statin use in patients but effective in increasing adherence among those already taking statins. LEAD AUTHORS THE IMPLICATIONS This trial is the first of its kind to thoroughly test physician, David Asch is Executive Director of the Penn patient, and shared incentives of equivalent value, and is Medicine Center for Innovation. Dr. Asch’s notable for incorporating several insights from behavioral research aims to understand the clinical and economics: daily engagement, “regret” lotteries, a relatively economic decisions patients and providers make. high probability of a small reward and lower probability of a larger reward, and leveraging of loss aversion. Kevin Volpp is founding Director of the Center for Health Incentives and Behavioral Economics These findings are important for what they reveal about at LDI (LDI CHIBE), which explores how the what works and what does not work. Neither physician science of behavioral economics can improve nor patient incentives on their own lowered the LDL-C health. Dr. Volpp’s research program focuses on the impact level significantly more than the control. The lack of of financial and organizational incentives on health outcomes. effect of the physician-only intervention offers the first controlled evidence that adding these incentives to a fee- Andrea Troxel is Director of Biostatistics at LDI for-service payment model may not improve medication- CHIBE. Dr. Troxel is an expert on statistical related intermediate outcomes. The authors suggest that the methodology in the areas of missing data, effectiveness of the shared incentives model makes sense longitudinal studies, and clinical trials. with the LDL-C reduction likely driven by both provision Connect With Us: ldi.upenn.edu Since 1967, the Leonard Davis Institute of Health Economics (LDI) has been the leading university institute dedicated to data-driven, • ldi.upenn.edu/health-policysense policy-focused research that improves our nation’s health and health • @PennLDI care. Originally founded to bridge the gap between scholars in • www.youtube.com/user/LDIvideo business (Wharton) and medicine at the University of Pennsylvania, LDI now connects all of Penn’s schools and the Children’s Colonial Penn Center Hospital of Philadelphia through its more than 200 Senior Fellows. 3641 Locust Walk Philadelphia, PA 19104-6218 LDI Research Briefs are produced by LDI’s policy team. For more P: 215-898-5611 F: 215-898-0229 information please contact Janet Weiner at weinerja@mail.med.upenn.edu.