Reduction of emergency department use in people with disabilities.

OBJECTIVES: To examine emergency department (ED) use by individuals with disabilities in safety-net clinics that have adopted the patient-centered medical home (PCMH) model.

STUDY DESIGN: This is a retrospective matched cohort study. Prior to matching, we identified 2269 nonelderly Medicaid beneficiaries with disabilities from a Los Angeles Medicaid managed care plan in PCMH clinics and 21,897 in non-PCMH clinics.

METHODS: To minimize self-selection bias from clinics and individuals, we created 3 comparison groups through a series of propensity score matching schemes that included matching clinics with similar health service utilization per patient and matching individuals with similar demographic characteristics and underlying health conditions. Rates of having at least 1 ED visit per year and excess ED use (defined as ≥2 ED visits per year) were compared across beneficiaries who received care from PCMH clinics and matched comparisons using logistic regression analyses.

RESULTS: After matching on clinic- and individual-level characteristics, the adjusted odds ratio (OR) of excess ED use was 25% to 33% lower (P <.05) in the PCMH group compared with the non-PCMH group. When limiting the study population to patients with at least 1 office visit, the OR of having at least 1 ED visit decreased by 21% (P <.05) for the PCMH group. Similarly, the OR of having excess ED use decreased by 38% (P <.05) for the PCMH group.

CONCLUSIONS: Our study highlights that the adoption of the PCMH model in safety-net clinics was associated with reduced ED use in Medicaid beneficiaries with disabilities.