Public reporting and pay for performance in hospital quality improvement

A Medicare pay-for-performance pilot project that rewards hospitals based on quality measures contributes to "modest" improvements in quality of care, according to a study published online Friday in the New England Journal of Medicine, the Wall Street Journal reports (Tomsho, Wall Street Journal, 1/29).

The project, which launched in October 2003, includes about 260 hospitals in 38 states. Under the program, hospitals can earn bonuses if they rank among the top 20% in providing specified treatments in at least one of five areas of patient care: joint replacement, coronary artery bypass graft, heart attack, heart failure and pneumonia. CMS officials on Friday announced that the three-year pay-for-performance pilot program has improved patient care. CMS officials on Friday also announced that performance bonuses of $8.7 million will be awarded to 115 hospitals that were the top performers based on 30 quality measures in the second year of the project. Premier, a not-for-profit hospital alliance, is managing the program (Kaiser Daily Health Policy Report, 1/25). For the NEJM study, researchers compared results over a two-year period from 207 hospitals taking part in the CMS project with 406 hospitals that were not offered quality-care reimbursements. Both groups of hospitals were participating in efforts to publicize quality-of-care data. The study found that hospitals in the pay-for-performance program had greater improvement overall in all combined measures of quality. Researchers found that hospitals involved in the pay-for-performance program had a 2.6% improvement in treatment for heart attack patients and a 3.4% improvement in care for pneumonia patients compared with hospitals not in the program (Wall Street Journal, 1/29). The bonuses were relatively small, accounting for 1% to 2% of the cost of providing care for patients with one of the evaluated conditions.

According to researchers, a possible problem with the quality reimbursements was that they were awarded to hospitals relative to the performance of other hospitals. As a result, "most of the payments went to hospitals that showed the least improvement but had already begun the program with higher quality scores," HealthDay/Washington Post reports. In addition, lead study author Peter Lindenauer -- from the Division of Healthcare Quality at Baystate Medical Center and the Tufts University School of Medicine -- said that it is not clear whether the increased costs associated with the program compared with standard public reporting programs are justified by the small improvements in patient care. Lindenauer also said that paying hospitals for improvement in care for a limited number of conditions could be problematic, adding, "We don't know whether care for other conditions could have suffered as a result of hospitals paying less attention to those conditions, to excel in the areas under study" (Reinberg, HealthDay/Washington Post, 1/27). Arnold Epstein, a professor of health policy and management at the Harvard School of Public Health, in an accompanying editorial wrote that "the findings still leave us with many uncertainties concerning the level of financial incentives needed and the optimal formula for payment that might be used for attaining high levels of performance." Epstein added that "rather than adopt a single new payment system for all of Medicare, a series of regional models could accelerate learning and allow Medicare officials to find out more about the effect of differing levels of incentives and formulas for payments" (CQ HealthBeat, 1/26).

An abstract on the study is available online. An abstract of the editorial is available online.