Medicaid enrollees face a disproportionate burden of severe mental illness yet have lower access to treatment.¹ Psychiatrists are less likely than other physicians to accept insurance, especially Medicaid.² This disparity is exacerbated by inaccurate health plan provider directories: a previous claims-based study of 2018 Medicaid managed care directories revealed that in some plans, over 90% of listed mental health care prescribers did not actively participate in the Medicaid program, but claims-based studies cannot assess wait times or enrollees鈥� direct experiences with accessing care.^3-5

We conducted a 鈥渟ecret shopper鈥� audit study to examine availability of and wait times for adult appointments with psychiatric prescribing clinicians across 4 of the most populous US cities.

We selected at random 80 psychiatric prescribing clinicians (psychiatrists, nurse practitioners, and physician assistants) listed as accepting new patients from the provider directories for the Medicaid managed care plans with the highest enrollment in New York City, Los Angeles, Chicago, and Phoenix, the 4 largest cities in states that have expanded Medicaid eligibility under the Patient Protection and Affordable Care Act. Using a standardized protocol and calling script, 5 researchers called sampled clinicians during normal business hours between May 12, 2023, and July 6, 2023, as Medicaid enrollees seeking the soonest available appointment. When an appointment could not be provided with the sampled clinician, callers requested consultation with an alternate clinician at the same practice (eAppendix in Supplement 1).

We examined appointment availability, wait times, and reasons an appointment could not be made with the sampled clinician. We made comparisons between cities for appointment availability with the 蠂² test with Bonferroni correction. We compared wait times with a 1-way analysis of variance test. Analyses were performed with Stata version 16 (StataCorp). We defined significance as 2-sided P鈥�<鈥�.05. The study was approved by the institutional review board at the Biomedical Research Alliance of New York, which provided a waiver of informed consent because the research involved minimal risk and could not be carried out without a waiver, and a waiver would not adversely affect individuals being studied.

Across the 320 clinician offices called, 87 (27.2%) had appointments available, including 57 (17.8%) with the sampled clinician and 30 (9.4%) with an alternate clinician at the same practice. The proportion of calls resulting in an appointment with any clinician was 36.3% in New York City, 30.0% in Phoenix, 27.5% in Chicago, and 15.0% in Los Angeles. The 2 cities with significantly different appointment availabilities were New York and Los Angeles (P鈥�=鈥�.002). Median wait times were 11 days (IQR, 6-20 days) in Phoenix, 23 days (IQR, 9-35 days) in Chicago, 28 days (IQR, 11-84 days) in New York, and 64 days (IQR, 24-126 days) in Los Angeles (Figure 1). The cities with significantly different wait times were Phoenix and Los Angeles (P鈥�=鈥�.001), Phoenix and New York City (P鈥�=鈥�.03), and Chicago and Los Angeles (P鈥�=鈥�.049).

Among the 263 sampled clinicians with whom appointments could not be made, 15.2% had a listed number that was incorrect or out of service and 35.0% did not answer the phone on either of 2 attempts (Figure 2).

In the largest Medicaid managed care plans across 4 of the largest US cities, only 17.8% of clinicians listed as in-network for Medicaid were reachable, accepted Medicaid, and could provide a new patient appointment. When access did exist, it often was not timely, with wait times up to 6 months.⁵

A limitation is the lack of a comparison group of enrollees with commercial health insurance, although previous studies have established that Medicaid enrollees have lower access to care than those with other forms of insurance. Calls were placed only to practices in urban areas and the cited diagnosis for requesting an appointment when prompted was depression; results may not generalize to rural areas or patients seeking treatment for other conditions.

A recent regulation from the Centers for Medicare & Medicaid Services requires states to conduct independent secret shopper surveys to directly assess network adequacy in Medicaid managed care.⁶ This study highlights the importance of greater enforcement of network adequacy standards through tools such as audit studies to safeguard access to care in the Medicaid program.

Accepted for Publication: June 14, 2024.

Published Online: July 31, 2024. doi:10.1001/jama.2024.13074

Corresponding Author: Diksha Brahmbhatt, MD, MPH, Department of Medicine, Brigham and Women鈥檚 Hospital, 75 Francis St, Boston, MA 02115 (dbrahmbhatt@bwh.harvard.edu).

Author Contributions: Drs Brahmbhatt and Schpero had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Both authors.

Acquisition, analysis, or interpretation of data: Both authors.

Drafting of the manuscript: Brahmbhatt.

Critical review of the manuscript for important intellectual content: Both authors.

Statistical analysis: Brahmbhatt.

Administrative, technical, or material support: Schpero.

Supervision: Schpero.

Conflict of Interest Disclosures: Dr Schpero reported receiving grants from the Agency for Healthcare Research and Quality, Commonwealth Fund, Flatiron, Laura and John Arnold Foundation, Milbank Memorial Fund, National Institute on Aging, National Institute on Minority Health and Health Disparities, New York State Department of Health, Patient-Centered Outcomes Research Institute, Physicians Foundation, Robert Wood Johnson Foundation, and United Hospital Fund outside the submitted work. No other disclosures were reported.

Data Sharing Statement: See Supplement 2.

Additional Contributions: We are grateful to Ella Blicker, BA, Tania Pe帽a-Ortiz, BS, Callista Clairmont, BA, and KT Tsang, BS (all at Weill Cornell Medical College) for their assistance in primary data collection. These individuals were not compensated for this work.