Callers Got a Mental Health Appointment on 24 Percent of Answered Calls. Across All 5,822 Calls, It Was 9 Percent

Provider directories are the lists insurers give members to show which clinicians are in network, and their accuracy has been a documented problem across specialties for years. A study published October 16 in JAMA Health Forum measured it for commercial mental health coverage. Callers posing as patients phoned 5,822 listed psychiatrists and psychologists, and 532 of those calls, 9.1 percent, ended in an appointment with the clinician on the list.

How the calls were made

The team, led by Simon Haeder and Wendy Xu of the Ohio State University College of Public Health, called from October 2025 through March 2026. Each caller asked the closest listed psychiatrist or psychologist for an appointment on behalf of a simulated patient with a standard PPO plan and a nonemergency condition such as anxiety or grief. The plans came from the largest commercial carrier in each of Delaware, Montana, New Jersey, Rhode Island and Virginia, with CareFirst added for Virginia because Anthem does not cover the whole state.

Callers made one attempt per listing and did not leave messages.

Two denominators, two answers

Of the 5,822 calls, 3,570 (61.3 percent) could not be evaluated, mostly because the caller reached only an answering machine or an automated system. That left 2,252 calls that reached a person. Among those, 23.6 percent produced an appointment with the listed clinician (95 percent confidence interval, 21.9 to 25.4), and 32.1 percent did if the caller would see someone else. The Ohio State release rounds these to 24 and 31 percent.

The gap between 9.1 and 23.6 percent is not evidence that 91 percent of listings were fake. A clinician who lets a call go to voicemail may be fully available. The authors say as much, writing that the yield is "a lower bound on what a persistent patient might achieve." What the number does show is how often one phone call, the way a directory invites a patient to start, goes nowhere.

What went wrong when a person answered

Among the calls that connected, 55.6 percent hit a directory inaccuracy. Wrong contact information accounted for 31.0 percent of calls and a clinician with the wrong specialty for 22.2 percent. Another 21.6 percent failed on capacity, such as a practice not taking new patients. Haeder said providers "move, change phone numbers, leave practices or otherwise become unavailable," and that insurers can list every clinician in a contracted group even when an individual is not seeing patients.

Patients who did get an appointment waited an average of 20.5 days. Psychiatry waits averaged 22.5 days against 17.3 for psychology, though appointment success was similar, 24.5 and 22.4 percent.

The pattern varied by carrier and by geography

Horizon Blue Cross Blue Shield of New Jersey had the highest adjusted appointment rate at 32.9 percent. Blue Cross Blue Shield of Montana came in at 19.0 percent and CareFirst in Virginia at 18.0 percent. Rural clinicians were more likely to offer an appointment than nonrural ones, 33.2 against 21.9 percent, and their listings had fewer errors, 44.3 against 57.9 percent. The cost showed up in travel: rural residents averaged 30.8 minutes to a clinician against 12.7 minutes for others, and Montana averaged 41.8 minutes.

What the study cannot say

The test covered five states and one carrier or two per state, so the authors caution that the findings may not generalize to smaller carriers or other plan types. The work was funded by Inseparable, a national mental health advocacy organization.

A comparison point exists, with a caveat. A 2024 JAMA study from Weill Cornell Medicine, led by then-student Diksha Brahmbhatt and senior author William Schpero, found that 18 percent of listed psychiatrists accepting Medicaid in New York City, Los Angeles, Chicago and Phoenix were reachable, took Medicaid and offered an appointment. Schpero said that "actual access to care is apparently quite low" even among listed participants. Insurance type, cities and call protocol differ, so the two figures are not a trend.

The new paper's own reading is that barriers documented in public insurance also reach privately insured patients. The No Surprises Act requires insurers to verify directory information at least every 90 days, and the authors note that more than three years have passed since those provisions took effect.

The bottom line

The evidence is solid that these five-state directories frequently listed clinicians who could not be booked. It is thin on how the result carries to other states, smaller insurers or Medicare Advantage plans, and silent on how long a persistent patient would need.

Haeder said he would like a centralized directory that clinicians maintain and insurers draw from, rather than each insurer building its own. No study cited here has tested that idea. What the paper does supply is a baseline: a measured appointment rate, taken after the 90-day verification rule had been in force for over three years.