By Steffie Woolhandler, M.D.
The Wall Street Journal, June 16, 2013
Paying doctors for better care — not just more of it — seems like a no-brainer. Yet rigorous studies of pay-for-performance bonuses have found no health benefits and some unintended harms.
An exhaustive analysis of pay-for-performance research by the Cochrane Collaborative, an international group that reviews medical evidence, unearthed “no evidence that financial incentives can improve patient outcomes.”
Consider these cases. In Britain’s massive pay-for-performance program, family doctors earned almost perfect scores (and big bonuses) for hypertension treatment, but population surveys found no decrease in blood pressure or its main complication, strokes. Meanwhile, aspects of quality that didn’t bring bonuses deteriorated.
The largest U.S. pay-for-performance experiment — Medicare’s Premier Demonstration — also flopped. The 200 hospitals that offered bonuses scored slightly worse on patient death rates than other hospitals.
Proponents argue that programs like these were flawed in one way or another, and that the next trial — or the one after — will certainly do better. They also claim successes with other programs. But none of these claims rest on rigorous science, and all those that have subsequently been subjected to rigorous tests have failed.
No Easy Measurement
Why do these programs consistently fall short? Measurement is distorted once you pay doctors based on the data they themselves create. High scores may reflect real excellence, but can just as easily reflect cherry-picking or gaming the measurement system.
One Boston-area hospital we observed improved its quality score 40 percent just by getting doctors to change the words they wrote in patients’ charts. Medicare gives hospitals more credit for saving patients with “acute respiratory decompensation” than those with “COPD exacerbations,” although these terms are synonyms. That kind of practice is neither illegal nor unusual.
Beyond that, it’s devilishly difficult to quantify doctors’ performance in the first place. Hospital death rates seem, at first glance, an ideal measure of medical quality. Yet, four widely used algorithms yield completely different mortality rankings; a hospital rated outstanding in one often looks downright dangerous in another.
Even if — as some proponents argue — we find performance measures that work for one group of doctors, it’s unlikely that they’ll work for all providers in all patient populations. Moreover, many providers interact in providing care, and influence each other and patients’ outcomes in complex ways. It’s hard to imagine that incentives could optimize this as a system.
There’s also psychology at work. Rewarding performance ignores the complexity of human drive, particularly the role of intrinsic motivation — the desire to perform an activity for its own inherent rewards. Offering your dinner-party host a $10 reward for cooking a wonderful meal isn’t likely to motivate future invitations.
Studies have found that financial incentives often crowd out intrinsic motivation. For instance, college students will spontaneously play with interesting puzzles, but once they’re paid to solve them, they lose interest in playing for nothing. When day-care centers in Israel imposed fines on parents for picking up children late, tardiness increased. Promptness transformed from a moral duty to a market transaction.
Pay for performance undermines the mindset required for good doctoring — the drive to do good work even when no one is looking. Moreover, it forces doctors to shift their attention from patients to computer screens — documenting trivial details useless for patient care but essential for compliance.
None can doubt medicine’s grave quality problems. As a remedy, pay for performance suggests manipulating greed. This can certainly change medicine, but not necessarily in the ways that we would plan, much less hope for.
Dr. Woolhandler is a physician and professor at the City University of New York School of Public Health. David U. Himmelstein, also a physician and professor at the School of Public Health, and Dan Ariely, the James B. Duke professor of behavioral economics at Duke University’s Fuqua School of Business, contributed to this article. They can be reached at firstname.lastname@example.org.