Like everybody, health care professionals enjoy a pay raise for a job well done. But in some instances, financial incentives for health care performance may actually backfire.
A new UCLA study shows that patient-care performance ratings for 25 medical groups across California improved significantly following the launch of a statewide pay-for-performance program in 2004 - but not when incentives focused on doctors' productivity.
Reporting in the December edition of the Journal of General Internal Medicine, Hector P. Rodriguez, assistant professor in the department of health services at the UCLA School of Public Health, and colleagues found evidence that certain kinds of financial incentives for the purpose of improving patient care, in combination with public reporting of medical group performance ratings, have a positive effect on patient care experiences. However, they also found that some types of incentives may have a negative overall impact on how patients experienced their care.
The researchers analyzed how medical group performance ratings changed over time and found that ratings in specific measures representing three broad categories - physician communication, care coordination and office-staff interactions - improved substantially during the period after the start of the Integrated Healthcare Association's (IHA) pay-for-performance program. Incentives for addressing the quality of patient-clinician interaction and the overall experience of patient care tended to result in improved performance in those three areas, especially when the additional funds were used broadly by medical groups to positively reinforce a patient-centered work culture.
However, the greatest improvements were seen within those groups which placed less emphasis on physician productivity and greater emphasis on clinical quality and patient experience. And within groups where financial incentives were paid directly to physicians - rather than being used more broadly - the researchers found that placing too much emphasis on physician productivity actually had a negative impact on the experiences patients had when visiting their primary care doctor.
"As the Obama administration and Congress continue to grapple with health care reform, these findings provide timely information about the kinds of things medical groups can do - and can avoid doing - with financial incentives in order to improve the quality of patient health care experiences," said Rodriguez, the lead author of the study.
For the study, researchers looked at information collected from 124,021 patients of 1,444 primary care physicians at 25 California medical groups between 2003 and 2006 and conducted interviews with group medical directors to determine how financial incentives were used. All 25 groups, which represent six insurers, were awarded financial incentives for achievements in the broad categories of clinical care processes, patient care experiences and office-based information systems, in accordance with the IHA program, which was launched in 2004 with the goal of improving health care quality in California.
Medical groups were free to use the additional funds in various ways, with some groups paying incentives directly to physicians, and others using the incentives more broadly, with a focus on organizational priorities. The groups also participated in a public reporting program in which ratings in two of the three broad categories were released annually to the public in the form of a "health care report card" comparing the performance of the medical groups and insurers to one another.
"The current House bill being debated includes the establishment of a Center for Quality Improvement to identify and implement the best practices in the delivery of care," Rodriguez said. "Our study results suggest that the nature of financial incentives can affect the provision of patient-centered care. Therefore, quality improvement and reimbursement reform efforts should integrate patient-reported experiences of care as a central metric for evaluating reform effects."
Study finds health care evaluations of large language models lacking in real patient data and bias assessment