Fewer than 15% of Washington Medicaid emergency department visits for mental health conditions—and roughly 11% of visits for substance use disorders or alcohol use disorder—were followed by condition-concordant primary care within 30 days, according to a retrospective cohort study published in JAMA Network Open.
The analysis included 859,043 emergency department (ED) visit claims from 367,245 Medicaid beneficiaries aged 18 years or older who lived in Washington, were enrolled in Medicaid for at least 11 months in 2022, and had at least one ED visit that year.
Researchers assessed 30-day condition-concordant primary care follow-up following ED visits for mental health conditions and substance use disorders. Alcohol use disorder, a specific subtype, was examined separately. Follow-up was defined as a primary care visit in which the diagnosis matched the ED visit diagnosis.
Among 131,704 ED visits for mental health conditions, 18,722 were followed by condition-concordant primary care within 30 days, for a rate of 14%. Among 101,684 ED visits for substance use disorders, 11,353 had follow-up (11%). In the alcohol use disorder subgroup, 3,675 of 33,196 ED visits had follow-up, also approximately 11%.
Disparities in follow-up
Across all three models, non-Hispanic Black beneficiaries had the lowest predicted probability of follow-up. Compared with this group, non-Hispanic White beneficiaries had approximately 4 to 5 percentage points higher probability of follow-up, with similar advantages observed for Alaska Native or American Indian, Asian or Pacific Islander, and other racial groups.
Beneficiaries experiencing homelessness also had lower follow-up rates, with differences of approximately 3 percentage points following mental health–related ED visits and about 2 percentage points following substance use–related visits.
Differences by condition and patient characteristics
Associations with follow-up varied by condition.
Following ED visits for mental health conditions, female sex, rural residence, and greater comorbidity burden were associated with higher probability of follow-up, whereas increasing age was associated with lower probability
Following ED visits for substance use disorders, older age and greater comorbidity burden were associated with higher probability of follow-up
In the alcohol use disorder subgroup, older age, female sex, and greater comorbidity burden were associated with higher probability of follow-up, whereas rural residence was associated with lower probability
Higher comorbidity burden, measured using the modified Quan-Charlson Comorbidity Index, was consistently associated with greater probability of follow-up across all models, suggesting that patients with more complex medical needs may have more established primary care connections or care coordination.
Notably, rural residence showed different patterns depending on condition—associated with higher follow-up after mental health–related visits but lower follow-up after alcohol use disorder–related visits—highlighting potential differences in access to condition-specific services.
Interpretation and context
The findings are notable given that Washington implemented a statewide integrated managed care model by 2020 aimed at improving coordination between behavioral health and primary care services. Persistently low follow-up rates suggest ongoing challenges in achieving continuity of care even in systems designed to promote integration.
The study also comes amid policy changes that could affect access to care. The authors noted that recent federal legislation is projected to increase the number of uninsured individuals, which could further reduce access to primary care follow-up and increase reliance on emergency services.
Limitations
The authors emphasized several limitations. As a single-state analysis, the findings may not be generalizable to other Medicaid programs. The use of claims data may underestimate follow-up if clinicians addressed mental health or substance use concerns without coding the same primary diagnosis.
Follow-up was measured from the ED visit rather than hospital discharge, meaning patients who were hospitalized after the ED visit may have had less opportunity to complete a 30-day follow-up within the study window.
Importantly, the study captured only primary care follow-up. It did not assess follow-up in specialty behavioral health settings, addiction medicine programs, residential treatment facilities, or other care pathways. As a result, the findings reflect gaps in primary care follow-up specifically rather than absence of any follow-up care.
Takeaway
“Condition-concordant primary care follow-up after ED visits for MH conditions, SUDs, and AUD was infrequent,” wrote lead researcher Jonathan A. Staloff, MD, of the University of Washington, and colleagues. “Tailored care coordination and outreach may be necessary to improve continuity of and access to primary care services among these populations.”
No conflicts of interest were reported. The study was funded by the Washington Health Care Authority; two coauthors were affiliated with the funding organization at the time of the study.
Source: JAMA Network Open