Primary Care Follow-Up After Mental Health and Substance Use Emergency Department Visits in Medicaid

Medicaid beneficiaries-particularly those with mental health (MH) conditions, substance use disorder (SUD), and alcohol use disorder (AUD)-frequently use emergency departments (EDs). Timely primary care follow-up can reduce high 30-day ED revisit rates for these conditions; however, primary care follow-up after such ED visits remains understudied.

To examine rates of 30-day condition-concordant primary care follow-up for MH, SUD, and AUD after ED visits related to these conditions and identify characteristics associated with condition-concordant primary care follow-up.

This cohort study retrospectively analyzed 2022 Medicaid claims data from beneficiaries residing in Washington State who had at least 1 ED visit during the study period. Data were analyzed between March and May 2025.

ED visits for MH conditions, SUDs, or AUD.

The primary outcome was primary care follow-up for MH conditions or SUDs within 30 days following ED visits for those diagnoses (condition-concordant primary care follow-up), and the secondary outcome was condition-concordant primary care follow-up for AUD within the same time frame. Multivariable logistic regression with marginal effects was used to assess associations between beneficiary characteristics and probability of primary care follow-up.

This study included 859 043 Medicaid ED visit claims from 367 245 unique individuals in Washington (mean [SD] age, 41.7 [16.2] years); 496 775 claims (57.8%) were for women. Among the claims, 52 500 (6.1%) were for beneficiaries self-reporting as Alaska Native or American Indian, 48 103 (5.6%) as Asian or Pacific Islander, 89 548 (10.4%) as non-Hispanic Black, 476 968 (55.5%) as non-Hispanic White; and 173 996 (20.3%) as other race (17 928 [2.1%] did not provide or had missing information on race); 142 556 (16.6%) claims were for beneficiaries self-reporting as Hispanic and 678 961 (79.0%) as non-Hispanic (37 526 [4.4%] did not provide or had missing information on ethnicity). Condition-concordant primary care follow-up occurred after 18 722 of 131 704 MH condition- (14.2%), 11 353 of 101 684 SUD (11.2%)-, and 3675 of 33 196 AUD (11.1%)-related ED visits. All racial groups, and non-Hispanic White individuals in particular, had a higher probability of receiving condition-concordant primary care follow-up than non-Hispanic Black individuals (estimated marginal effects: MH, 4.47 [95% CI, 3.87-5.07] percentage points; SUD, 4.70 [95% CI, 4.12-5.27] percentage points; AUD, 4.00 [95% CI, 2.87-5.13] percentage points). Individuals experiencing homelessness had a significantly lower probability of receiving condition-concordant primary care follow-up (estimated marginal effects: MH, -2.74 [95% CI, -3.25 to -2.23] percentage points; SUD, -1.88 [95% CI, -2.33 to -1.44] percentage points; AUD, -1.86 [95% CI, 2.73 to -0.99] percentage points) compared with those not experiencing homelessness.

In this cohort study of Medicaid beneficiaries in Washington, condition-concordant primary care follow-up after ED visits for MH conditions, SUDs, and AUD was infrequent. Observed racial and social differences suggest potential barriers to care access and coordination following ED visits. Tailored care coordination and outreach may be necessary to improve continuity of and access to primary care services among these populations.