Association for Public Policy Analysis & Management

Medicaid programs increased payment to primary care physicians (PCPs) to parity with Medicare rates during 2013-14, so called the “Medicaid fee bump.” The size of the payment increase was over 50% nationwide and varied across states. Past literature found that physicians alter service pattern in response to a price change, so as to optimize income. Using the MarketScan Medicaid claims data for 2011-14, this study examined the effect of the Medicaid fee bump on primary care intensity, focusing on four office visit codes (CPT: 99212-99215) that accounted for over half of fee bump-eligible claims. Though high intensity codes (99214 and 99215), in contrast to low intensity ones (99212 and 99213), are for sicker patients that demand longer visit duration and more diagnostic tests, PCPs had a certain degree of latitude in choosing which code to claim. In most states the fee bump boosted the payment for high intensity codes disproportionally more than low intensity ones. Hence, we hypothesized that PCPs would claim more high intensity codes after the fee bump, and that the claimed high intensity codes might not be rendered if likelihood of receiving diagnostic tests and new medications following a high intensity visit dropped significantly. We stratified analyses by provider facility type (office vs. outpatient hospital [OP]) and patient type (fee-for-service [FFS] vs. Medicaid managed care [MMC] vs. dual Medicaid-Medicare eligible), as PCPs’ facility ownership, perception of benefit (per-visit bonus vs capitation rate increase), and other factors associated with provider/patient type might affect treatment and coding decisions.  

We found that office-based PCPs claimed more high intensity codes after the fee bump when treating FFS patients (2011-12: 44% of the four codes, 2013: 50%, 2014: 54%); this increase was absent for MMC patients, or FFS patients treated by OP-based PCPs. Among high intensity codes claimed by office-based PCPs on FFS patients, less involved lab tests (2011-12: 30%, 2013: 27%, 2014: 25%) or led to new drug prescriptions (2011-12: 17%, 2013: 16%, 2014: 15%).

Multivariable regression examined whether, at individual provider level, such shifts corresponded the size of fee bump, controlling for patient characteristics, year dummies, and provider fixed effects. The independent variable of interest was ‘payment difference-in-difference’ (PDiD) facing a provider, which measured post-fee-bump payment difference between high and low intensity codes subtracting its pre-fee-bump counterpart. PDiD captured the fee-bump-led incremental benefit of claiming high intensity code. We found that, for FFS patients treated by office-based PCPs, per dollar PDiD was associated with 0.4% increase in claiming high intensity codes in 2013 and 0.7% in 2014 relative to 2011-12. Conditional on a high intensity code rendered by office-based PCPs on FFS patients, PDiD was associated with lower likelihood of receiving new drugs, lab tests, or X-ray. Such correlations were much weaker or in the opposite direction for MMC and dual eligible patients.

This study suggests that office-based PCPs might respond to the fee bump by hiking the use of more expensive codes, and some of the code hiking might not be justified given the decline of associated quality indicators.