An Optimal Physician Payment System

The researchers at Harvard have really done it this time!  Just when you thought it couldn’t get any more exciting, they release a study on…typology.

Ty-pol-o-gy: [tahy-pol-uh-jee] noun. a systematic classification or study of types.

Harvard researchers classified ‘types’ in this study; the ‘types’ studied were various compensation models for primary care physicians (PCPs) caring for Medicare patients.  In all the excitement, the researchers not only studied these payment models, they also studied how to study these various compensation models.

The interesting part, based on a follow up analysis, utilizes linear regression models to demonstrate which types of PCP compensation models tend to be associated with spending more or less money per patient.  Pulling from a survey of 2,000 PCPs across the country and the Center for Medicare and Medicaid Services claims data for these PCPs over the course of three years, this analysis looked at how the cost of care per patient was associated with individual physician compensation models.  Based on the previously-proven accurate methodology (a.k.a. typology), physicians were subdivided into seven major categories: (1) employed physicians paid salary, (2) employed physicians paid based on productivity, (3) employed physicians paid based on productivity and additional incentives in low capitation environment (<35 percent of revenue), (4) employed physicians paid based on productivity and additional incentives in high capitation environment, (5) practice owners without additional incentives, (6) owners with incentives in low capitation settings, and (7)  owners with incentives in high capitation settings. All care of the Medicare patients was reimbursed via the traditional fee-for-service manner.

Physicians who worked as employees, received productivity based compensation, and were in high capitation settings were found to spend the least of all other physicians. Total health care costs of patients cared for by physicians under this payment model were 3 to 13 percent less than PCPs in other compensation models based on total care costs per patient per year.  What’s more, the cost of care strictly for physician fees (as opposed to total care costs) per episode of care under this compensation model were 4 to 19 percent less than physician fees under other compensation models.

Physicians who were owners had higher costs of care than physician employees, although high capitation environments curbed this trend. Employed physicians paid based on productivity without other incentives had the highest costs of care.

Physicians compensated through fixed salary had significantly fewer episodes of patient care per year per patient than physicians under any other type of compensation model.  The number of episodes of care per year per patient was roughly the same among patients with PCPs in all other types of compensation models.

Commentary

As this analysis points out, only 21 percent of health care spending in our country is for physician fees, but using compensation to influence practice patterns of physicians, especially primary care physicians, affects utilization of other health care resources and therefore a much larger part of health care spending.  These data suggest that policy encouraging physician compensation through capitated contracts, whether through bundled payments or accountable care organizations, would reduce not only physician costs, but would also reduce health care costs classified under other sources.

This analysis also suggests that even if the majority of a physician’s services are reimbursed with fee-for-service, physicians who get a significant amount of their revenue from capitated contracts tend to use those same cost-conscious practice habits throughout their practice and do not change their practice habits based on the individual patient’s insurance coverage.

Lastly and inevitably, any conversation that discusses decreasing the cost of health care really should be focused on increasing the value of healthcare.  Maybe that should be the next study for the scientists at Harvard: valu-ology.

Landon BE, et al. The Relationship between Physician Compensation Strategies and the Intensity of Care Delivered to Medicare Beneficiaries. Health Serv Res. 2011 Jul 25. [Epub ahead of print]

by

Lisa Maurer, MD