Pay-for-performance programs that award bonuses to medical groups for reaching a fixed, common target appear more likely to maintain the status quo at high-performing groups rather than generate noticeable quality gains, according to a report in Wednesday's Journal of the American Medical Association. The study's principal lesson, the researchers said, is "incentive design matters." The researchers compared clinical data on doctors in a PacifiCare Health Systems pay-for-performance plan in California with data on similar physicians in Oregon and Washington who were not part of pay-for-performance plan. The California doctors showed a 5.3% improvement in use of cervical cancer tests from October 2001 to April 2004, compared with a 1.7% improvement in the comparison group. Both groups improved 2.1% on hemoglobin testing. In mammography -- the last clinical measure evaluated by the researchers -- the California doctors achieved a 1.9% improvement, compared with 0.2% in the comparison group. Some 75% of the bonus pool went to physician groups that were already above the threshold when the program began and had made limited gains.
The researchers said the findings could mean it takes more time to generate substantial quality gains or that the bonuses -- a maximum of $27 per patient -- weren't large enough to do so. Perceived fairness also matters in program design, said lead author Meredith Rosenthal, an assistant professor at Harvard School of Public Health. "Providers have to view these programs as fair and one can't understate the importance of that," Rosenthal said. Read the abstract. -- by Andis Robeznieks