By Sarah Klein and Martha Hostetter

On Making Performance Measures Meaningful
IMPORTED: www_commonwealthfund_org____media_images_newsletters_tnfkaqm_2015_oct_navathe_amol_115x100.png

Amol Navathe, M.D., Ph.D., assistant professor of health policy and medicine at the University of Pennsylvania Perelman School of Medicine

Q: When it comes to designing workplace incentive programs, are there unique things about the way physicians work that necessitate a different approach?
A: Absolutely. In economic terms, we think of the principal–agent relationship, meaning physicians act as agents first and foremost for the patients—making decisions and recommendations on their behalf. This is unlike other employees who have an almost singular duty to their employers or customers. However with health care reform, physicians are increasingly being asked to serve as agents for not just patients and society, but also for health insurers, for accountable care organizations, and as agents of value-based care. As a result, employers must adapt incentive programs so they align the multiple agency relationships that physicians carry.

Q: What about transparency in reporting performance at the individual physician level. How important is that?
A: When using unblinded reports, physicians have to feel it’s a fair comparison, that it’s people they perceive or know to be their peers. If not—say in a large-scale reporting program in which physicians are being compared to someone in another market and it’s unclear what resources they have—comparisons may actually be harmful rather than helpful. It’s important to remember that as a group, physicians are quite motivated to meet professional standards and not lag [relative to] colleagues, so peer effects can be very powerful.

Q: What are some of the biggest unanswered questions in this field?
A: We don’t yet understand how the design of incentives influences behavior and performance. It is very much an open space still. The many open questions include the what, when, and how of information feedback. We know that we probably need to simplify the hundreds of quality measures the National Quality Forum has endorsed down to a practical set that we can make salient, actionable, and tied to meaningful amounts of both financial and nonfinancial incentives. We just don't know which measures and how many to use.

Q: Your work focuses on using behavioral economics to change physician behavior. What are some good examples of how health systems used them—perhaps unwittingly?
A: Most, if not all, health systems that use behavioral economics were trying to solve their business needs through trial and error, rather than with an explicit agenda to use psychology and economics to motivate behaviors. There are several good examples, including group-based incentives, dashboards with peer comparisons, and public award ceremonies.

On Communicating Your Approach
IMPORTED: www_commonwealthfund_org____media_images_newsletters_tnfkaqm_2015_oct_patel_pankaj_115x110.png

Pankaj Patel, M.D., M.Sc., senior medical director, Advocate Physician Partners

Q: How does Advocate communicate its strategies for increasing the value of the care it provides?
A: There are a couple of things. One is having an entire team of people who meet monthly with all of our clinicians. They explain the program, ask if there are problems or issues, and liaise back and forth with the local physician–hospital organization and Advocate Physician Partners’ management team. Then we have a variety of other communications—from 15-minute videos to flash reports that go out via email to say this is a big issue, or we are behind in this, or we are doing well here. There is also a requirement that physicians participate in training sessions and attend meetings that go over how the program works, how it applies to them, and what are the measures relevant to them. We also have developed an interactive continuing medical education (CME) program that physicians are required to participate in. For instance, the immediate past president of the American Academy of Allergy, Asthma, and Immunology is one of our physicians and he helped us develop a CME program on what we think are our best practices for asthma care—it condenses the NIH guidelines into 45 minutes.

Q: You’ve said Advocate didn’t set out to employ the principles of behavioral economics in its physician incentive program, but in essence did. What’s an example?
A: We have these annual end-of-year meetings where the independent doctors get their incentive checks. So you can imagine they are well attended. We take that opportunity to highlight big things that are happening in health care, and to explain to physicians what we are doing to try to stay ahead of the curve. And then we do some nonfinancial things. So the top 25 percent of primary care physicians, specialists, they get a framed certificate. I am always amazed—some physicians seem more interested in their plaques and certificates on their wall than any kind of financial incentives. Of course, that’s not everybody. About a year ago a physician wrote me this lengthy letter saying “I am a top performer and I got this excellent physician recognition. Please do not use precious health care dollars on printing out a certificate.”

On Creating Feedback Loops
IMPORTED: www_commonwealthfund_org____media_images_newsletters_tnfkaqm_2015_oct_debrantes_francois_115x110.png

François de Brantes, M.S., M.B.A., executive director, Health Care Incentives Improvement Institute

Q: Years ago you helped create Bridges to Excellence, an incentive program that allowed employers and payers to reward providers for among other things reducing avoidable complications in patients with chronic conditions. You’ve said one of the keys to its success was relying on data in physicians’ medical records—rather than claims data—to assess performance. Why was this so important?
A: It’s not a scorecard that’s calculated using algorithms off of claims data, which often have really significant errors and are the reason why, when physicians say “This is garbage,” it’s not hyperbole. We said this is your information, not ours: if you are not documenting stuff, which many of them didn’t, if you are not correctly assessing your patients’ vitals, that is your problem, not my problem. After that one of the most striking comments that came up, again and again and again, was, “I am just not doing as well as I thought I did.” That was a very powerful incentive.

Q: What other mistakes do incentive programs make?
A: I think it’s retrospective review. Any time you create distance between performance measurement and reporting, you are removing one of the more impactful aspects of a feedback mechanism, which is its timeliness. Ultimately what you are you doing with pay-for-performance and quality improvement is providing feedback. It is how well have I done, how well I am doing, and can you tell how well I am doing in relation to other peers I respect? So it has to be timely and it has to be accurate. And of course if you are judging it in a quantitative way, the methods must be fair and transparent.

On Traditional Pay-for-Performance Programs
IMPORTED: www_commonwealthfund_org____media_images_newsletters_tnfkaqm_2015_oct_bishop_tara_115x110.png

Tara Bishop, M.D., assistant professor of health care policy and research at Weill Cornell Medical College

Q: When it comes to traditional pay-for-performance programs, what have we learned about what works and what doesn’t?
A: There have been very mixed results. There have been studies that have shown an effect particularly around very specific types of services. There was for example a study looking at blood pressure management and management around cardiovascular disease—done in conjunction with New York’s Department of Public Health—which did show an effect of paying providers for those quality metrics, but then there have also been studies that show no effect. Most of these programs are built upon a fee-for-service model of payment where physicians continue to get paid for the services they provide and then get additional incentives for meeting and reaching quality metrics.

Q: Is the fact that they are building on top of fee-for-service payments sending a mixed message?
A: I think it has to do with the relative incentive. If you can make more money by seeing more patients a lower incentive payment might not change behavior so much. We are still in a phase of trying to figure out what is the clearest incentive in the fee-for-service environment: is it the amount? Or how you pay? We found in a national survey nearly 17 percent of doctors don’t even know if they are getting incentive payments. There may not be enough transparency about how much you are getting in bonuses and how that affects your salary or take-home pay.