Development of a Model for the Validation of Work Relative Value Units for the Medicare Physician Fee Schedule

Barbara O. Wynn ; Lane F. Burgette ; Andrew W. Mulcahy ; Edward N. Okeke ; Ian Brantley ; Neema Iyer ; Teague Ruder ; Ateev Mehrotra

RAND Corporation

RAND Corporation

关键词: Fee-for-Service for Health Care;    Health Care Costs;    Medicare;    United States;    Modeling and Simulation;    Medical Professionals;    Medicaid;   

DOI  :  10.7249/RR662
RP-ID  :  RR-662-CMS

学科分类：自然科学（综合）

美国|英语

来源: RAND Corporation Published Research

【 摘 要 】

The Centers for Medicare & Medicaid Services (CMS) uses the resource-based relative value scale to pay physicians and other practitioners for professional services. The work values measure the relative levels of professional time and intensity (physical effort, skills, and stress) associated with providing services. CMS asked RAND to develop a model to validate the work values using external data sources. RAND's goal was to test the feasibility of using external data and regression analysis to create prediction models to validate work values. Data availability limited the models to surgical procedures and selected medical procedures typically performed in an operating room. Key findings from the study include the following: RAND estimates of intra-service time using external data are typically shorter than the current CMS estimates. Model assumptions about how shorter intra-service times affect procedure intensity have implications for the work estimates. RAND estimates for work on average were similar to current work values if shorter intra-service time is assumed to increase procedure intensity and were on average up to 10 percent lower than current work values if shorter intra-service time is assumed to not impact on procedure intensity. The RAND estimates could be used for two key applications: CMS could flag codes as potentially misvalued if the RAND estimates are notably different from the current CMS values. CMS could also use the RAND estimates as an independent estimate of the work values. In some cases, further review will identify a clinical rationale for why a code is valued differently than the RAND model predictions.