Pay-for-performance advocates acknowledge flaws: if not designed carefully, plans can warp physician behavior and fail to improve health care quality.

WASHINGTON -- If you're of the mind that the pay-for-performance plans instituted by federal as well as private payers are questionable at best and potentially dangerous at worst, don't worry: You're not alone. Many leaders of the pay-for-performance movement share your concerns.

Speaking at the fourth World Health Care Congress, advocates of pay-for-performance (P4P) acknowledged that if not designed carefully, these plans can create perverse incentives, warp physician behavior, and ultimately fail in their primary objective of improving health care quality.

P4P leaders admit that in many cases, they're not sure they're tracking the right measures. Even if they do get it right, there is little evidence that the measures are truly meaningful to ordinary people needing to make medical decisions.

This doesn't mean P4P is going away any time soon. In fact, P4P plans will only become more widespread in the coming years, spurred on by Medicare's embrace of the concept. But P4P advocates are rapidly finding out they need to assess the impact of their systems as closely as they monitor physician and hospital performance.

"Everything we do must be monitored for unintended consequences. P4P plans are no different. The movement is in its infancy," said Dr. Tom Valuck, director of value-based purchasing for the Centers for Medicare and Medicaid Services. He cited a recent Institute of Medicine report concluding that while P4P has potential to improve health care systems, experience is still very limited, close monitoring is essential, and plan developers need to build in provisions for rapid redesign and correction.

"P4P may lead to focus on wrong priorities. For example, we can end up focusing on individual accountability instead of system performance. This raises a lot of questions about rewards and incentives." Wrongly focused P4P could exacerbate health care disparities, leading to cherry-picking and cream-skimming, and detracting clinical attention from other priorities, he added. "We may end up teaching to the test, while ignoring the bigger picture."

Dr. Brent James is executive director of the Institute for Healthcare Delivery Research at Intermountain Healthcare, a health system with one of the most proactive quality improvement and performance measurement systems in the nation. An early advocate of P4P, Dr. James said he has learned some important lessons over several attempts at establishing P4P programs.

Where most P4P plans go awry is by being overly focused on arbitrarily chosen physician "accountability" measures and not focused enough on overall systems process measures that tie back to meaningful clinical outcomes, Dr. James said.

"You have to show end-of-day improvement in care. If everyone is doing 'perfect score' medicine but there's no improvement in outcomes, it means either people are gaming the system or the measures are irrelevant. If you build for system improvement, you'll get accountability data along the way. Build from the bottom up, so as not to damage care."

Dr. James defines systems transparency as meaning that "you have sufficient information to make a whole series of decisions, and this holds for patients and practitioners alike. It is not as if any one single piece of information tells the whole story or allows one to make a definitive decision. Transparency is a much broader, a much more profound concept than accountability."

Dr. James said that he is wary of plans that attach heavy financial rewards or penalties to individual physician measures. First, the measures may not be clinically important ones and may end up rewarding "performance" on tasks that do not really lead to better patient care. Second, financial incentives can skew care delivery "As you attach greater rewards or punishments to achieving a number, you get increasing propensity for suboptimization; you make one area look good at the expense of the others."

Finally, financial incentives create the wrong sort of motivations. "One of the worst things you can do to physicians is tell them that money is more important than their professional judgment. They will end up believing you," he said.

An effective P4P program motivates physicians by stressing improved patient care. "Extrinsic awards destroy intrinsic motivation for improvement. Get the professional incentives right and you get system improvement," Dr. James said.

Tom Sackville, chief executive of International Federation of Health Plans, and former Minister of Health in Britain, strongly agreed. "Doctors are highly trained, independent-minded, intelligent professionals. They know what they have to do. If they perceive distant bureaucrats throwing bits of fish, they'll start behaving like ... performing sea lions. Our doctors pride themselves on having a true vocation. We spoil that at our peril."

"The things that people measure in P4P are dictated by ivory tower thinkers. Their relevance to patients, or even to the administrative process, is very questionable," said Robert Burney, director of quality improvement for the U.S. Department of State.

Dr. James questioned the extent to which P4P data has any relevance to patients at all. "The truth is patients really do not use outcomes statistics to make their health care decisions. They rely on stories, based on relationships. They'll tell you they want data, but when we measure decision making, the data do not drive it. We have several good studies of this topic, where they gave patients carefully prepared statistics. Patients say the stats changed their decisions, but when we look closely, people do not change decisions based on data. Humans are more emotional than statistical."

If patients tend not to respond to data, physicians will ... eventually.

Dr. Varga said doctors tend to go through "a sort of 'Kubler-Ross acceptance process' when it comes to P4P, going from a denial attitude of, 'Your data stinks, its all BS,' through one of, 'Your data are meaningful but don't really apply to me,' through, 'The reasons my data are bad is because everyone's data are bad,' to finally accepting there's a need for improvement." But that's provided a P4P system is truly oriented toward system-wide care improvement and not simply punitive toward individuals.

Punitive ranking systems can have a very detrimental effect on health care, several experts said at the conference. On an individual level, P4P may favor older, more experienced practitioners at the expense of younger ones who may have less experience with a given procedure, and thus may get labeled early on in their careers as "lower quality." This can make it hard for younger doctors to build practices.

There's also a very real danger, Dr. Varga said, of putting smaller rural practices out of business if Medicare reimbursement is overly tied to rigid performance measures. "You can end up destroying health care delivery for small rural counties. A lot of smaller rural hospitals are working on very small margins. If you take away 5% of their Medicare revenue, they close their doors. They can't take that kind of hit."

At best, P4P is a tool set for improving health care outcomes, reducing iatrogenic illness and adverse events, and improving the overall return on each health care dollar spent. Advocates believe that with the right measures, P4P can achieve these goals.

"I think doctors are motivated to improve if they see objective data that they are not performing as well as their peers. It is not necessarily a financial incentive, but a patient care incentive that will motivate them," said Dr. Jack Lewin, CEO of the American College of Cardiology. ACC has developed a vigorous program of accountability guidelines aimed at improving the quality of cardiovascular care.

"Ultimately, we want to show individual cardiologists how they are doing in relation to their peers on real world indicators, and we want to give them tools for improvement." Given that cardiovascular disease consumes more than 43% of total health care dollars, a little improvement will go a long way, Dr. Lewin said.

ACC is currently studying "door to balloon" time at major centers, in an effort to reduce the interval from when a patient arrives at a hospital until he or she is in the angioplasty suite. "How fast do the best hospitals get you from the e-room door to the balloon angioplasty? You want this to happen within 90 minutes."

The National Cardiovascular Data Registries, which ACC supports, represent a major national project aimed at tracking hospital performance on a wide range of procedures including acute MI, balloon angioplasty, and defibrillator implantation. Data are being gathered in roughly 2,300 centers around the country.

"We can tell the medical staff how they are doing compared to their peers," Dr. Lewin said at the conference sponsored by the Wall Street Journal and CNBC. "We still need the patient outcomes side, but the program is underway, and some states mandate that hospitals participate if they want the states' Medicare and Medicaid data."

Dr. Peter Angood, codirector of the Joint Commission International's Center for Patient Safety, likened current quality improvement efforts, flawed though they may be, to the airline industry's efforts to improve safety performance.

"It took the aviation industry 40-45 years to improve performance quality and really get continuous quality improvement in place. In health care, we're just passing the stage where we acknowledge there's a problem. How to compress that 40-year curve down to just one generation?"

BY ERIK GOLDMAN

Contributing Writer