Pay-for-Performance Plans a Major Component of Medicaid

NEW YORK, April 13 -- More than half the states have tied payment to performance for Medicaid services for some of their health care programs, according to a survey just published.

As of July 2006, 28 states had at least one pay-for-performance Medicaid option in place, and many of these payment schemes were at least five years old. By contrast, Medicare is tentatively approaching pay-for-performance as a new concept.

The nationwide survey of Medicaid programs, including interviews with state Medicaid directors, was published by the Commonwealth Fund. The report was prepared by researchers working for IPRO, a not-for-profit quality evaluation and improvement organization, and the Kuhmerker Group, a consulting firm founded by Kathryn Kuhmerker, a former director of New York's Medicaid program and co-author of the report.

There report found no one-size-fits-all Medicaid pay-for-performance plan. Instead, incentives and performance measures tended to be tailored to a state's specific Medicaid population and the state's ability to collect and review relevant data.

For example, Alabama is a state that ranks high in chronic diseases associated with obesity, so it offers a higher reimbursement rate to physicians who participate in a program that uses electronic records to monitor patients with diabetes and other chronic diseases.

By contrast Nevada, a state that has had rapid population growth in the past decade, takes a level-playing-field approach. It set up a bottom standard of performance beneath which no incentive payments are offered. And Massachusetts has a new plan on the drawing board that will offer incentives to reward both performance and improvement.

Likewise there was no standardization of measure. Some states rely on two performance measures while others use 10 or more.

The report pointed out that there are still no good measures of the effectiveness of pay-for-performance plans.

The study characterized incentives into six categories. There are bonuses, differential reimbursement rates or fees; penalties; auto-assignment of beneficiaries to specific plans or providers; withholds; and grants.

Not surprisingly, Medicaid directors said bonuses or differential reimbursement -- higher payment rates for providers who meet or exceed performance measures -- were the most effective incentives.

On the flip side, there was general agreement that penalties were not effective because they create "ill will between the medical community and the state, which may result in decreases in provider participation."

Auto-assignment is not technically a financial incentive but "it drives market share and, therefore, compensation," so it, too, was considered effective by the Medicaid directors.

Other findings of the survey:

• 70% of the Medicaid pay-for-performance plans operating in July 2006 are a feature of managed-care or primary-care case management health plans that provided coverage for women, children, and adolescents.
• New pay-for-performance plans -- including those recently implemented or in the planning stages -- are more likely to focus on quality and cost issues related to management of chronic diseases.
• Nine states -- Arizona, Kansas, Maine, Minnesota, New Hampshire, New York, Oregon, Vermont, and Washington -- are joining other payers, providers, employers, and consumers to form regional pay-for-performance initiatives.
• Some pay-for-performance plans are based not on clinical measures but on the willingness of physicians to adopt information technology such as electronic health records and electronic prescribing.

Kuhmerker and co-author Thomas Hartman, vice-president for health care quality improvement at IPRO, concluded the while pay-for-performance has secured more than a foothold in Medicaid, there are still no good measures of its effectiveness.

Establishing those efficacy measures, they wrote, is the "biggest challenge facing both state Medicaid [pay-for-performance] programs and those operated under other auspices."