Inaccurate mental health directories block care long before a first visit is booked
A secret shopper study of large commercial insurance plans found a stark disconnect between what online provider directories promise and what patients encounter when they try to schedule mental health care. Fewer than one in four evaluable calls to listed psychiatrists and psychologists ended in a scheduled appointment, even though the callers were seeking nonemergency care and were directed to clinicians appearing to be in network.
The problem was not simply long wait times. In many cases, the path collapsed before scheduling even became possible.
Investigators from Ohio State University and Stanford University placed more than 5,800 calls to listed psychiatrists and psychologists between October 2025 and March 2026. The callers simulated patients in five states and contacted the geographically nearest clinician listed in each insurer’s online directory. Those plans were the state’s predominant commercial carriers: Blue Cross Blue Shield of Rhode Island, Blue Cross and Blue Shield of Montana, CareFirst BlueCross BlueShield in Virginia, Highmark Blue Cross Blue Shield Delaware and Horizon Blue Cross Blue Shield of New Jersey.
Among the 2,252 calls that reached a human and could be evaluated, 23.6% ended with an appointment on the books. When callers said they were willing to see another available clinician, the scheduling rate rose to 31.1%. Even that improvement left most callers without access to a visit. For patients already struggling with anxiety, depression, grief or another mental health concern, the first barrier was often the phone tree, the wrong number or a provider who no longer participated in the plan’s network.
Directory errors, capacity limits and dead ends shaped the encounter
A majority of the evaluated calls, 55.6%, involved inaccurate information in the insurer’s directory. Incorrect contact details were the most common problem, and 11.4% of calls reached a disconnected or out-of-service number. Nearly 22% failed because the provider’s practice had no capacity. Those capacity constraints were more common in psychology than psychiatry.
The findings fit the familiar complaint about so-called ghost networks: directories that suggest broad access but do not reflect real-world availability. In practice, a patient may spend time searching for an in-network clinician, only to discover that the listing is obsolete, the practice has closed to new patients, or the phone number belongs to a dead line. The mismatch is especially consequential in behavioral health, where delays can push care farther from the moment when a patient first reaches out.
Wait times were long even when scheduling succeeded. The average wait to an appointment was 20.5 days, and psychiatry visits took longer to secure than psychology visits. Travel time averaged 17.1 minutes overall, but rural patients faced a much longer commute, averaging 30.8 minutes to reach the appointment they did obtain. Rural clinicians were more likely to offer appointments than nonrural clinicians, a pattern that may reflect different local supply conditions, practice structures or willingness to accept new patients.
The gap between network status and usable access is measurable
The study’s most important contribution is not just the low scheduling rate. It quantifies the breakdown between insurer directories and actual care delivery. A directory entry is supposed to function as an access tool, yet the data show that the tool often fails at the point of first contact. For patients, especially those with limited time, limited transportation or psychological distress, a failed call can become a failed treatment pathway.
Across carriers, the success rate varied from 18% in Virginia to 32.9% in New Jersey. Psychiatry and psychology were similarly difficult to access overall, which suggests the problem is not confined to one type of clinician. The study’s structure also captures an important reality in mental health access: even when a plan formally lists many providers, the number of usable providers may be far smaller once contact errors and capacity limits are counted.
Directly speaking, a network can look broad on paper and still function as narrow at the bedside. Patients do not receive care from directory listings; they receive care from offices that answer, verify coverage and have an opening.
Regulatory action has not yet closed the gap
These findings arrive three years after directory accuracy provisions of the No Surprises Act took effect. The investigators interpreted the results as evidence that current requirements are not enough or are not being enforced in a way that materially improves accuracy. That conclusion carries weight because the study examines not theoretical coverage promises but actual attempts to book care.
The issue has drawn policy attention beyond the commercial market. CMS has already moved toward a national provider directory for Medicare patients, including a preview release that aggregates provider data from multiple sources. On the same day this study appeared, CMS also proposed a rule for Medicare Advantage plans that would require directory verification and updates at least every 90 days. The proposal would also shield patients from higher cost sharing if they received care from a clinician incorrectly listed as in network.
Those measures speak to the same structural flaw exposed by the secret shopper calls: administrative data age quickly, while patients act on them immediately. When a clinician leaves a practice, changes a phone number or stops taking a given plan, the directory can lag behind the actual market for months. In mental health care, that lag can mean a missed evaluation, a worsening symptom burden and a longer path to treatment.
Centralized maintenance may matter more than repeated listings
Simon Haeder, the Ohio State associate professor who helped lead the work, argued for a centralized provider directory maintained by clinicians and accessible to insurers rather than each payer trying to maintain a separate database. The logic is straightforward. Multiple payers, each with their own update cycles and data standards, create multiple points of failure. A shared directory would not solve capacity problems on its own, but it could reduce the volume of outdated contact information and misclassified network status that currently frustrates patients before a visit is ever scheduled.
The current model asks patients to navigate a system that often looks organized from the insurer’s perspective but fragmented from the patient’s. A person seeking help is not trying to audit network files. They are trying to find a clinician who answers the phone, accepts the plan and can see them within a reasonable time.
This study shows how often that basic sequence breaks. The directory may say “in network,” but the practice may have moved, the line may be dead, or no appointment may be available at all. The result is not just administrative inconvenience; it is delayed care in a field where delay can deepen illness and make the next call harder to place.