January 14, 2015
Changing health care pay-for-performance programs to account for differences among providers in their patient characteristics can make the incentive schemes more equitable and avoid a redistribution of resources away from providers who care for socioeconomically disadvantaged patients, according to a new RAND Corporation study.
Researchers developed a new pay-for-performance method that adjusts payments to health providers based upon patient characteristics so as to avoid potential unintended consequences that may discourage health providers from caring for disadvantaged patients.
Simulating use of the approach, the research team found that it nearly doubled quality-based incentive payments to provider organizations that serve disadvantaged patients as compared with the existing incentive scheme. Importantly, the approach retained financial incentives to encourage all providers to improve quality. The findings are published in the January edition of the journal Health Affairs.
“Incentive programs run the risk of encouraging providers to avoid patients who may have greater challenges in meeting quality targets that providers are now being held accountable for,” said Cheryl Damberg, the study's lead author and Distinguished Chair in Health Care Payment Policy at RAND, a nonprofit research organization.
Because providers who care for disadvantaged patients tend to perform less well on average, the impact for these providers is fewer resources to help more-challenging patients achieve quality targets, researchers say.
“Our pay-for-performance approach strengthened quality-based incentive payments to providers who care for disadvantaged patients, while only slightly weakening incentives for other provider organizations and eliminating the redistribution of money to well-resourced providers,” said Marc Elliott, an author of the study and the RAND Distinguished Chair in Statistics.
The new approach eliminated the redistribution of financial resources away from providers who care for disadvantaged patients. Furthermore, within each group of providers the approach still holds the providers accountable for better performance.
Pay-for-performance programs in health care have grown in recent years as a way to improve the quality of care delivered by doctors, hospitals and other health care providers. Under the schemes, providers receive bonuses for meeting certain quality measures such as making sure patients with high blood pressure have their blood pressure under control.
Many providers say it's more challenging to meet quality measures when they have large numbers of disadvantaged patients who, for example, may lack the economic and social support needed to get to appointments with doctors or take medications on schedule.
Case-mix adjustment, a statistical method to control for patient factors when computing quality scores, is one approach that can address this issue. However, evidence suggests that even after such adjustment disadvantaged patients receive care from poorer providers. Thus case-mix adjustment is not sufficient to address what may be a downward spiral of resources.
RAND researchers developed a method that allows for existing pay-for-performance payments to be adjusted using factors that categorize providers, such as the proportion of the patient population with low socioeconomic status or the monthly base capitation payment received by the provider, which is a marker of the resources a provider has to care for patients.
RAND researchers simulated the effect of the new payment scheme by analyzing 153 provider organizations in California eligible for incentive payments in 2009 as a part of the Integrated Healthcare Association's pay-for-performance program.
Using census block data and other statistical measures, researchers sorted provider organizations into pre-defined groups based on the income of their patients and their monthly capitation rate. The mean incentive payment was held constant across the provider groups, while within group provider organizations earned higher incentive payments with higher quality performance.
The results showed that the RAND approach fully eliminated differentials across provider organizations based on categories defined by patient income. The approach also eliminated about half of the differentials observed based on patient race/ethnicity and geographic location.
Meanwhile, the approach continued to provide larger rewards to provider organizations that performed better on quality measures, thus preserving incentives to improve health care.
Support for the study was provided by the Robert Wood Johnson Foundation. Brett Ewing was the other author of the study.
RAND Health is the nation's largest independent non-profit health policy research program, with a broad research portfolio that focuses on health care costs, quality, payment reform, and public health preparedness, among other topics.