Changes in utilization of health care services following procedure code bundling and consequent Medicare fee adjustments

Background: Policymakers are concerned that the Medicare Physician Fee Schedule may overstate the relative costs of certain imaging procedures. To distribute payments across services fairly and to lower the overall spending, the Centers for Medicare & Medicaid Services (CMS) recently cooperated with the American Medical Association (AMA) to identify services with distinct Current Procedural Terminology^® codes often performed and billed together. To correct for improper coding leading to inappropriate payments, the AMA continuously has introduced new bundled codes for the identified services, while the CMS assigned these codes updated Relative Value Units to “better” reflect the actual performed work.

Research objective: The main goal is to assess the extent to which the exogenous changes in Medicare fees affect provision of health care services in both Medicare and privately insured populations. The two primary outcomes in our study, which we compare, are the volume of health care services provided to Medicare beneficiaries and the volume of health care services provided to privately insured individuals.

Data: Our major data sources are the Medicare Standard Analytical Files for years 2010-2014 and the Truven Health Analytics MarketScan^® Commercial Claims and Encounters database for the same period. The Medicare database includes health care claims of a random 5% sample of all Medicare beneficiaries. The MarketScan^® database contains health care claims of more than 50 million privately insured individuals across the United States. In addition, information about health care supply variation across the United States is obtained from the 2010-2013 Area Health Resources Files.

Methods: The theoretical framework for our research is based on the so-called stepwise demand model that describes a multiple market situation in which the physician is a price-setter in the private market and a price-taker in the Medicare market. The volume of provided services then results from altering the payer mix depending on the relative profitability in the two markets. We use the administrative changes in the Medicare Physician Fee Schedule resulting from the recent procedure code bundling of various radiology services as the exogenous price shocks in the Medicare market. In the first stage, we estimate the effect of Medicare price shocks on prices in the private market. In the second stage, we substitute the actual prices in the private market for the predicted prices resulting from the first stage, and estimate the effect of the altered profitability of certain procedures on the volume of provided health care services using the simultaneous equations system approach.

Results: We have established initial estimates using 2010-2013 data. The preliminary results support the predictions of the stepwise demand model. However, the most recent 2014 data will become available in January 2016, and we will present the final five year analysis results at the conference. Our previous study (Horný et al. 2015), which focused on utilization of the combined computed tomography scan of the abdomen and pelvis, documented an increased utilization rate (+11.4% per month; p = 0.034) in the private market after corresponding code bundling and Medicare fee reduction occurred in 2011.