By Amy Norton
Reuters
June 29, 2010

The findings, researchers say, raise the possibility that the so-called "pay-for-performance" initiative could inadvertently worsen existing healthcare disparities.

Pay-for-performance reimbursement plans essentially reward hospitals and doctors for meeting certain treatment goals established in medical guidelines. For example, guidelines state that heart attack patients should be given aspirin and drugs called beta-blockers when they are admitted to and discharged from the hospital; centers that better meet that goal would get greater reimbursements.

Pay-for-performance systems are in place in some countries, like the UK, Australia and Taiwan, and have been operating on a limited scale in the U.S. for several years. Now the federal government is poised to implement a nationwide pay-for-performance policy within Medicare, the health insurance program for Americans age 65 and older, and people with disabilities.

But concerns have been raised about the fairness of such a system, since it assumes that all hospitals have the resources they need to meet clinical performance measures.

So for the new study, published in the journal PLoS Medicine, researchers looked at how U.S. hospitals' estimated "performance scores" would vary according to local economic conditions.

Using data reported by 2,700 hospitals to Medicare between 2004 and 2007, the researchers assessed how well the centers were meeting goals for treating heart attack and heart failure patients. They also used government data to look at the economic health of each hospital's country -- including the level of poverty, unemployment and the availability of health professionals and college graduates in the local workforce.

In general, the study found, hospitals improved their performance scores over time. However, those in the most disadvantaged geographic areas started off with lower scores than their better-off counterparts and still lagged behind by the end of the study period.

The researchers calculated hospitals' scores using the same methodology that has been proposed as the basis for Medicare's new pay-for-performance reimbursement plan.

The details of the program, which is set to begin widespread implementation in 2013, are still being hammered out, noted lead researcher Dr. Jan Blustein, of New York University Medical School.

Her team's findings highlight an important issue for policymakers to think about as they proceed, Blustein told Reuters Health in an interview.

One approach they could take, she said, is to "change the way they define 'improvement.'"

As it stands, the proposed performance-assessment system gives hospitals that start at a low baseline less credit for performance improvements than centers that start at a higher baseline. So "low-attainers" must have a greater increase in their absolute performance scores in order to "improve" as much as hospitals that start off with relatively high scores.

An alternative, Blustein and her colleagues suggest, would be to credit improvement regardless of a hospital's starting point. That could help hospitals in economically disadvantaged areas, where many of the low-attaining hospitals exist.

"We need to be mindful of whether hospitals' have a level playing field to begin with," Blustein said.

The Centers for Medicare and Medicaid Services, the federal agency that administers the Medicare program, has acknowledged that pay-for-performance has the potential to worsen the situation of hospitals that have long faced problems in meeting performance goals. And it has said that it will monitor the distribution of funds as the new payment program is implemented to see if some hospitals are being put at a disadvantage.

But Blustein and her colleagues say their findings argue for a more "proactive" approach -- one that, from the outset, tries to limit the chances of worsening healthcare inequalities.