A sizable bonus or sharp pay cut riding on the quality of patient care did little to change doctors' performance in a study.
The research, published in the latest issue of Health Affairs, found doctors at Fairview Health Services in Minneapolis did no better on quality than doctors elsewhere in the state even after Fairview said 40% of its paychecks would be tied to quality performance. The result is notable as hospitals, medical groups and health insurers test how best to tailor incentives to change the quality and cost of healthcare. Money alone may not be enough to get results.
“Putting more money on the table for financial incentives doesn't seem to be the answer to get financial incentives to be more effective,” said Jessica Greene, one of three researchers behind the study and a professor of nursing at George Washington University. “This was an instance where a lot of money was on the table.”
But the results also come with a critical caveat. The study measured quality improvement among Fairview doctors for the first two years that the health system put 40% of physician pay at risk for quality performance. Researchers compared Fairview's improvement to that of five other Minnesota health systems—including the Mayo Clinic—for the same period, using quality data reported to the state.
Researchers do not know whether the Mayo Clinic or other Minnesota health systems also made changes to how they paid their doctors for quality during the same period.
However, Greene said Fairview's quality incentive was unusual and others were unlikely to adopt similarly significant changes.
If that's the case, the result could suggest limited gain from increasing the size of incentives, said Dr. Mark Friedberg, senior natural scientist for the RAND Corp. and co-author of a recent RAND report on physician incentives for the American Medical Association.
But without knowing how other Minnesota systems paid physicians for quality during the period, the results' should be viewed with some caution, he said.
Teasing out how an incentive may influence behavior is a challenge at a time when so much work is underway to achieve the same result, Friedberg said. The significance of the study is limited without more information, he said.
In recent months, Medicare and commercial insurance companies have announced plans to aggressively increase their use of financial incentives tied to quality and greater control over the cost of medical care. Those incentives are largely weak, often the promise of the bonus without the risk of financial losses. Indeed, Medicare recently said it plans to allow hospitals and doctors to avoid potential financial losses under the program's largest test of accountable care.
Critics, meanwhile, have called for greater use of steeper incentives, without which doctors will have little incentive to change and will continue to respond to other incentives that reward greater use of healthcare services, regardless of possible harm and waste.
The study of Fairview's incentives also found that Fairview physicians who began with the weakest quality performance improved the most—three to six times more than other Fairview doctors—during the two-year period. The gap in quality performance narrowed between doctors with poorer quality and top-performing doctors as a result. Doctors with poorer quality also had a disproportionate share of lower-income patients and quality gains helped narrow socioeconomic disparities, the paper said.
The results suggest “any one patient was going to get higher quality care and more consistent care,” said Greene. “That's a really good thing.”
The study also found that performance improvement on vascular and diabetes quality among Fairview doctors varied (10 percentage-point gain versus 4 percentage-point gain) even though both measures accounted for 12% of doctors' compensation. Meanwhile, cancer screening accounted for 6% of doctors' pay but saw an 18% improvement in quality scores.
That suggests that factors other than financial incentives are at work, Greene said. For example, cancer screenings may be easier to change than quality measures for diabetes care, one of which is going tobacco free.
The lack of correlation between the size of the incentives and the quality improvement also suggest the size of the financial incentive may not be as powerful as believed, she said.
But those dynamics are unclear without data for comparison from other systems that lacked large financial incentives, Friedberg said.
Fairview, however, appears convinced. The system abandoned the incentive after roughly three years for one that reduced the amount of pay at risk for quality performance to 25%, the paper said. (Valerie Overton, vice president for quality and innovation for the Fairview Medical Group and one of the study's authors, was unavailable, a spokeswoman said.)
“This study adds to the accumulating evidence that payment reform is complex and may not be a “magic bullet,” the study concluded.