Poster Paper: Medicaid Today, Uninsured Tomorrow? Medicaid Beneficiary Disenrollment Rates and Physician Participation in Medicaid

Saturday, November 10, 2018
Exhibit Hall C - Exhibit Level (Marriott Wardman Park)

*Names in bold indicate Presenter

Adam S. Wilk, Xu Ji, Cathy Lally and Benjamin G. Druss, Emory University

Background: Only about two-thirds of physicians accept Medicaid patients, harming Medicaid beneficiaries’ access to care. Most of the literature seeking to explain this practice decision focuses on relatively low fees in Medicaid relative to what is paid in Medicare or through private insurance, but other factors may play a key role. In particular, physician practices may be less willing to accept Medicaid patients if they have a high likelihood of disenrolling and becoming uninsured and, thus, posing a greater risk of bad debt for the practice. For beneficiaries with chronic physical or mental health conditions requiring regular care, this factor may be especially important in practices’ decision-making, and the health consequences of poor access to care are exacerbated.

Objective: To examine the effects of local Medicaid beneficiaries’ average disenrollment rates on physician practices’ Medicaid participation decisions and whether these effects vary by specialty: primary care physicians treating adults versus pediatricians versus behavioral health specialists.

Design: We conduct a cross-sectional, observational study, controlling for other established drivers of physician participation in Medicaid, including state Medicaid fee levels and practice size. We account for endogeneity in our analysis by using an instrumental variables approach exploiting differences in states’ Medicaid eligibility redetermination regulations.

Participants: We examine this relationship within a nationally representative sample of U.S. physician practices and specific specialty subgroups of interest.

Main Measures: Our outcomes include dichotomous indicators for whether practices were accepting new Medicaid patients and whether at least 2, 5, or 10 percent of their practice revenues came from Medicaid. The main exposures are county-level rates of disenrollment from Medicaid among the chronically ill patient groups commonly treated by each physician specialty type.

Key Results: For every 10 percentage-point increase in the county’s disenrollment rate among chronically ill adult beneficiaries (due to restrictive eligibility redetermination rules), primary care physicians treating adults were 8.5 percentage points less likely to accept any new Medicaid patients (p=0.04). Similarly, for every 10 percentage-point increase in the county’s disenrollment rate among chronically ill youth beneficiaries, pediatricians were 9.8 percentage points less likely to accept new Medicaid patients (p=0.04). Estimated effects on behavioral health specialists’ Medicaid participation were muted, as were overall effects estimated across specialties. Findings were similar in analyses of physician practice revenues from Medicaid.

Conclusions: States concerned about access to care for Medicaid recipients should consider retaining relaxed program eligibility redetermination rules and other approaches to address physicians’ concerns about Medicaid patients’ disenrollment risk.