What can go wrong with pay-for-performance incentives
From the March ACP Observer, copyright © 2006 by the American College of Physicians.
By Gina Shaw
Despite a lot of talk about performance measurement, the pay-for-performance movement may not be moving very far or very fast, at least not yet.
A recent report from the Center for Studying Health System Change (HSC) found that of 12 surveyed communities, only Orange County, Calif., and Boston had any significant pay-for-performance initiatives in place. The HSC report also exposed a wide swath of physician skepticism. "A concern of frontline physicians," the report's authors wrote, "is that [pay for performance] means a little more money and a lot more work."
Not getting the right mix of time and money is just one of many pitfalls pay-for- performance programs can fall prey to. Even physicians who have championed such initiatives within their own practice point to traps that can sabotage the best-intentioned performance plan: reporting problems; focusing only on the care being measured to the detriment of other key health issues; and not aligning physician incentives with broader systems change.
"I think physicians are rightly concerned that pay for performance could become just another form of managed-care 'pressure on you to keep costs down' scheme, wrapped in quality clothing," said Thomas Bodenheimer, FACP, professor of family and community medicine at the University of California, San Francisco, and an author of the HSC report. "Right now we pay for quantity and not for quality, and paying for quality is a good idea. But there are a lot of potential problems in making that a reality."
We asked physicians to flag several pay-for-performance pitfalls, and here are ones they singled out:
Pitfall # 1: 'Performing to the test.' One snare is that time-stretched physicians may focus on only one or two measures.
"Most physicians are up against the wall time-wise, and any additional energy that goes to one thing means something else has to give," said Dr. Bodenheimer. "If you are really, really busy and know you can make some extra money by doing very good work on diabetes, might you slough off people who don't have diabetes or who have more complex problems? We don't know, but it's possible."
Yogesh K. Patel, FACP, had the same thought during a recent quality conference at his hospital, where staff was discussing inpatient incentives being developed by the Centers for Medicaid and Medicare Services (CMS) for its hospital quality incentive demonstration project.
"We might strive for 90% on postoperative infections and the use of appropriate antibiotics, but then are we going to be at only 70% compliance in those areas that aren't among those the CMS has defined?" said Dr. Patel, a general internist at Copperfield Internal Medicine, one of almost 30 primary and specialty practices that make up NorthEast Medical Network in Concord, N.C.
When it comes to office-based performance measures, NorthEast has its own quality initiative focused on diabetes, setting 13 goals across eight measurable indicators derived from American Diabetes Association evidence-based guidelines.
Why so many measures? Because achieving only one objective measure, such as measuring a hemoglobin A1C level, may have little impact on actual population outcomes, Dr. Patel said.
"You need to have several measures in place to improve overall health care," he said, "and the indicators chosen should be representative of the population served. This is the problem when insurers start imposing their payment incentive measures onto a practice: Those measures might focus on a small portion of the practice's population, which isn't representative of their service."
What if physicians find themselves involved with a program that chooses only one or two measures? "Though not ideal, choosing one element that is narrowly focused is better than none," said Dr. Patel. "This is an important place to start to adopt a mind-set of population management that is data- and evidence-driven." His advice" Start somewhere—and then challenge the system, asking questions about whether [the measure] is important, practical, feasible and has impact."
Another way to address an imbalance: rotate the measures. "That's the strategy I use in my quality work," said Donald W. Robinson, MD, a solo practitioner in Hamburg, N.Y., who participates in three pay-for-performance programs run by area HMOs. "We'll concentrate on diabetes for three or four years until doctors have learned the process of quality care—it's not like it disappears." Once that level of care is where it needs to be, "we'll concentrate on asthma, then cholesterol, then stroke and heart attack. We have enough work to keep us busy for decades."
Pitfall #2: The reporting burden. Participating in pay for performance means spending administrative time gathering data—time not spent with patients. This can be a real problem, especially for small practices and those without electronic health records (EHRs).
Even in California, where pay-for-performance implementation is going relatively well, many practices struggle with data gathering, said Glenn D. Littenberg, FACP, a gastroenterologist in private practice in Pasadena, Calif., who chairs the ACP's Third-Party Relations, Coding and Reimbursement Subcommittee. "Certain measures are discovered by chart review, and there are many situations where the relevant information isn't in the chart or isn't easily available."
Even practices that have implemented some kind of EHR are "often stuck with systems that are still part paper and electronic," he added. "It's hard to pull the data out of the systems even if you have the time and expertise to do that."
According to NorthEast's Dr. Patel, vendors for many full-blown EHR systems are still working toward a more integrated and searchable record. "In the meantime," he said, "a simple program using an Excel or Access database can organize some core basic data for specific diseases." Creating a simple database is not the real pitfall, he said; the problem is changing the mindset of physicians to convince them that managing patient populations is as important as managing individual patients.
Pitfall #3: Poor time-money balance. Some measures may appear to be easy to achieve, but may not provide enough financial incentives to see them through.
A pay-for-performance initiative targeting sinusitis, for example, may require physicians to see patients before prescribing an antibiotic—in addition to prescribing the right antibiotic. "Changing behavior in antibiotic prescribing is pretty easy and doesn't require a lot of effort," said Howard B. Beckman, FACP, medical director of the Rochester Individual Practice Association in Rochester, N.Y. According to a paper he published in the October 2004 issue of American Journal of Managed Care, just such an initiative resulted in a 32% reduction in inappropriate antibiotic use.
"But if many people call for antibiotics, and you have to see them before prescribing, you'll either have to extend office hours into the evening or make the person wait until the next day," he added. "We found that the rate of seeing patients before prescribing didn't change because the small amount of money being offered didn't cover the cost of changing physicians' practice pattern."
"In order to catch a doctor's attention and motivate his behavior, there has to be real money involved—not 2%, but at least 5% and more realistically 15%," said Dr. Robinson. "The effort you put into gathering the data is considerable, and if the incentives are small, it's not worth a doctor's attention." According to Dr. Robinson, CMS officials have indicated that within 10 years, Medicare pay-for-performance programs may account for as much as 20% of physicians' reimbursement.
"At that point," he said, "pay for performance is not just for aficionados and idealists—it will be compelling."
What if your program isn't providing enough return to generate real change? Present that case to the payer, Dr. Beckman said. "For physicians to succeed in improving quality through pay for performance, we need to make informed choices, be explicit about our needs and represent ourselves."
Pitfall #4: Measures not based on evidence. Selected measures must be carefully chosen as well as evidence-based, Dr. Beckman added. "If you want physicians to focus on changing their behavior, you can ask them to do a fixed, relatively small number of tasks—and you have to pick those very carefully."
Measures designed or vetted by the National Quality Forum, for example, have a reliable evidence base behind them. "But doctors will satisfy themselves best by reviewing the evidence themselves," said Dr. Beckman. "If that is not possible or appropriate, it would also be helpful to have their specialty organization provide comments and participate in measure design."
Pitfall #5: Misaligned incentives. Some incentives target helping patients change their behavior—but insurers don't go far enough to help with that modification.
A physician may, for example, get points for counseling patients to stop smoking, but the insurer might not pay for smoking cessation programs. Or a measure may target diabetic lipid management and reward physicians for measuring diabetic patients' blood pressure, cholesterol and hemoglobin A1c levels. But even though obesity affects all three levels, health plans may not cover weight management programs, dietitians or nutritional counselors.
"The doctor may have an incentive to measure something, but not necessarily [the financial stimulus] to change behavior—nor does the patient," said Dr. Patel. "To truly impact real outcomes relevant to quality of life, morbidity and mortality, all participants must align incentives and work together."
Pitfall #6. Aiming at the wrong target. And many experts see a more global pitfall with pay for performance: Too many programs measure only individual physicians with their relatively small patient populations, not medical groups or even larger community systems.
Focusing primarily on incentives at the individual physician level misses the point, said L. Allen Dobson, MD, assistant secretary for health policy and medical assistance with North Carolina's Department of Health and Human Services. Dr. Dobson oversees Community Care of North Carolina, a state quality initiative to develop programs and systems to improve the care management of Medicaid patients.
"My fear with pay for performance is that by only measuring physicians, you won't get the needed coordination and system development to achieve lasting results," he said. Instead, improving health care quality should start at the level of community systems to better manage patients. "I think ultimately to get the right movement in health care, we have to reward communities and providers that develop their own systems to effectively manage patients and improve quality."
Rewarding medical groups, not individual doctors, is a key strength of one of the most ambitious plans to date, the pay-for-performance initiative rolled out in 2003 by the Integrated Healthcare Association (IHA), a nonprofit consortium of California medical groups, health plans and health systems. That program paid out $50 million in pay-for-performance bonuses to medical groups throughout the state in its first reporting year.
But not everyone agrees that group- or systems-based programs are the best way to go.
"To be a true incentive, it's best to have the reward linked directly to your own personal effort," said Dr. Robinson, who pointed out that HMOs have been following the impact of incentives paid out in western New York for years. "When you give money to an entire specialty, like cardiology, it doesn't work well. When you give money to a group, it works better. But it works best if you give the reward to an individual doctor."
However the reward system is structured, he said, what's important is the underlying philosophy. "What you're doing with pay for performance is giving doctors the incentive to reinvent the practice of medicine," he said. "Once those systems are moving and doctors see rewards—not just dollar rewards, but lives saved, better patient satisfaction and the feeling that we're doing what we really set out to do in medical school—they can't go back to their old ways."
California's Dr. Littenberg has no problems with rewarding individual physicians—but he sees a big role for health plans and independent practice associations in minimizing the pitfalls for physicians.
"Much of the administrative scut work needs to be out of sight and off the individual physician's shoulders," he said. "If I could get an accurate data report that included patients' phone numbers, my staff could do a much more efficient job getting them in for what they need."
Gina Shaw is a freelance health care writer based in Montclair, N.J.
The information included herein should never be used as a substitute of clinical judgment and does not represent an official position of ACP.
Internist Archives Quick Links
Have questions about the new ABIM MOC Program?
One Click to Confidence - Free to members
ACP Smart Medicine is a new, online clinical decision support tool specifically for internal medicine. Get rapid point-of-care access to evidence-based clinical recommendations and guidelines. Plus, users can easily earn CME credit. Learn more