Does Medicaid coverage of Medicare cost sharing affect physician care for dual-eligible Medicare beneficiaries?

Abstract

Objective: To assess changes in physicians' provision of care to duals (low-income individuals with Medicare and Medicaid) in response to a policy that required Medicaid to fully pay Medicare's cost sharing for office visits with these patients. This policy-a provision of the Affordable Care Act-effectively increased payments for office visits with duals by 0%-20%, depending on the state, in 2013 and 2014.

Data sources: Fee-for-service claims for a 5% random sample of Medicare beneficiaries in 2010-2016.

Study design: We conducted a difference-in-differences analysis to compare changes in office visits among Qualified Medicare Beneficiaries (QMBs)-the largest subpopulation of duals for whom payment rates were affected by this policy-to changes among other low-income Medicare beneficiaries for whom payment rates were unaffected (pooled across all states). Next, we conducted a triple-differences analysis that compared changes between QMBs and other low-income beneficiaries in 33 states with payment rate increases of approximately 20% to analogous changes in 14 states without payment increases.

Data collection: The study included administrative Medicare enrollment and claims data for QMBs and a comparison group of other low-income Medicare beneficiaries (1 914 073 beneficiary-years from 2010 to 2016).

Principal findings: Nationally, we did not find a differential increase in office visits among QMBs versus other low-income beneficiaries that coincided with this payment change. In the triple-differences analysis, we did not observe a greater increase in visits among QMBs vs other low-income beneficiaries in states where the policy resulted in large (approximately 20%) increases in payment rates vs states where payment rates were unaffected (triple-differences estimate: -0.12 annual visits, 95% CI: -0.28, 0.04; P = 0.15).

Conclusions: Physicians' provision of care to low-income Medicare beneficiaries may not be responsive to short-run payment changes.

Keywords: Affordable Care Act; dual eligibles; physician incentives.

FIGURE 1

Triple‐differences event‐study estimates. This figures display differential changes in outcomes among QMBs versus other low‐income Medicare beneficiaries in each year relative to 2012 between two groups of states: (1) 33 states in which the implementation of the ACA fee bump increased provider payments from < 97% to 100% of the Medicare allowed amount for physician office visits (plotted in green), and (2) 14 states in which the fee bump had no substantive impact on provider payments (payments from Medicare and Medicaid equaled ≥ 97% of the Medicare allowed amount for physician office visits in each year from 2010 to 2012; plotted in red). Analyses based on 1 914 073 beneficiary‐years from 2010 to 2016 in 47 states. The solid dots are regression estimates and solid vertical bars are 95% confidence intervals, which were estimated using heteroskedasticity‐robust standard errors clustered on state. To limit the influence of outliers, we Winsorized each dependent variable at the 99th percentile of the unadjusted person‐year distribution of that variable. We adjusted for state fixed effects, characteristics of Medicare beneficiaries, and attributes of beneficiaries’ ZIP code tabulations areas and counties as described in the notes to Table 2