May 13 (Bloomberg) -- Pay-for-performance bonuses for U.S. hospitals may backfire, penalizing institutions that serve the poor and lack money to improve quality ratings, a study found.

Hospitals that treat more patients who are uninsured or are covered by state Medicaid programs for the poor made the smallest improvements in government performance scores, according to the study in tomorrow's Journal of the American Medical Association. Proposed pay-for-performance rules would widen the disparity with richer hospitals, researchers said.

Health plans increasingly are adopting ratings-based financial rewards and penalties to encourage hospitals to make fewer mistakes and follow sound practices. Raising scores costs money that hospitals serving the poor and racial minorities don't have, said Rachel M. Werner, a physician and health economist at the Veterans Affairs Department in Philadelphia and lead author of the study.

``Safety-net hospitals need a lot more help, and I don't think we should leave it up to the whims of the market to see if they can compete,'' said Werner in a telephone interview. ``We should not back away from pay-for-performance measures that have been adopted, but they may need to be applied differently to safety-net hospitals.''

Such hospitals have been given too little funding by Medicaid for decades and have too few insured patients who can make up the difference by paying higher prices, Werner said. Their strained finances mean they have less money to invest in equipment, training, staffing and other initiatives that bolster performance, she said.

Medicare Ratings

The study examined changes from 2004 to 2006 in ratings of care reported by 4,000 hospitals to Medicare, the U.S. health insurance program for the elderly.

Hospital Compare, Medicare's rating system, scores the treatment of heart attacks, heart failure and pneumonia, based on dozens of performance measures reported by hospitals. The scores cover tasks that include whether patients are promptly given appropriate drugs upon admission.

Hospital Compare has expanded its scope and now also tracks the care of patients with chronic lung disease, adult diabetes, chest pain, heart and blood vessel surgery and a range of abdominal and orthopedic operations.

Werner ran a simulation of payments for all U.S. hospitals based on performance bonuses that are being used in a Medicare demonstration program with 250 facilities. In the experimental program, each hospital in the top 10 percent of performers gets a bonus of 2 percent of Medicare payments, and those ranked in the next 10 percent get a 1 percent reward. Hospitals in the bottom 10 percent are penalized.

Fewer, Smaller Bonuses

Safety-net hospitals consistently performed worse than those with fewer low-income patients in the simulation, resulting in fewer and smaller bonuses to the neediest hospitals, she said.

Payments to the hospitals with the fewest poor patients grew 32 percent in 2006 compared with 2004, while rewards for patients with the highest share of poor patients fell 66 percent, Werner found.

``It puts them ever further behind the eight-ball in their efforts to improve care,'' Werner said. ``We're just penalizing the patients who get care at those hospitals.''

A study published yesterday in the Annals of Family Medicine found that U.K. doctors who get performance bonuses said fulfilling a checklist of tasks detracts from giving patients personal attention, although it improved care and doctors' earnings.

*t For related news:

*t

To contact the reporter on this story: Avram Goldstein in Washington at agoldstein1@bloomberg.net.

Last Updated: May 13, 2008 16:00 EDT