The Five‐Year Effect of Medicaid Expansion on Community Health Centers: Coverage, Quality of Care, and Service Volume

Our objective was to assess the five‐year impact of Medicaid expansion on community health centers using nationally representative data on all US health centers, where 35% of the patient population was uninsured prior to expansion. We examined the impact of expansion on insurance coverage and type, quality of care, and utilization of services. Understanding longer term effects of expansion is critical given that some quality measures may take multiple years to be clinically affected while pent‐up demand may also result in short‐term effects on utilization. Using the 2011‐2018 Uniform Data System, we conducted a difference‐in‐differences (DID) analysis with inverse probability of treatment weights (IPTWs), based on propensity scores, to compare outcomes in centers located in expansion versus nonexpansion states. Outcomes included insurance coverage type (none, Medicaid, private), 47 utilization measures (number of patient visits) for select categories of service and chronic conditions based on CPT and ICD codes, and 8 primary care quality measures that may be sensitive to Medicaid expansion. Propensity scores included 23 baseline covariates (patient demographics, health center organizational features, county‐level characteristics). For each measure, using IPTWs, a difference‐in‐difference was calculated using generalized linear models. We included a treatment indicator, time in postperiod indicator, treatment*post‐time interaction, vector of time‐variant covariates, state and year fixed effects, and clustered errors at the center‐level. 100% sample of US health centers (N = 1061 centers/year, or 24.5 million patients/year, after exclusions). By 2018, compared to centers in nonexpansion states, centers in expansion states experienced a 12.0 percentage‐point decrease in the percent patients without health insurance ( P