On paper, most health plans maintain an adequate behavioral health network. In practice, a U.S. Senate Finance Committee investigation found that members trying to book a mental health appointment through their insurance directory succeeded just 18% of the time.[1]
That's not a rounding error. It's the clearest signal that behavioral health network adequacy, as most plans currently manage it, is producing compliance documentation rather than actual access.
The regulatory and legal environment has shifted sharply in response. CMS expanded network adequacy requirements with explicit behavioral health provisions in its April 2024 final rule. The Department of Labor made network composition the centerpiece of federal MHPAEA enforcement and published a settlement agreement in January 2025 that sets a documentation benchmark the entire industry will now be held to. Federal courts are letting ERISA fiduciary breach claims tied to directory inaccuracies proceed to trial.
The problem itself isn't new. What's new is that the window for managing it quietly has closed.
The Behavioral Health Access Gap Is Bigger Than Most Plans Know
The numbers here are hard to ignore. SAMHSA's 2025 National Survey on Drug Use and Health found that 61.5 million U.S. adults — 23.4% of the adult population — experienced a mental illness in 2024, with nearly half receiving no treatment at all. Separately, 48.4 million Americans met the criteria for a substance use disorder that same year.[2] These figures aren't plateauing. They're still moving upwards.
What's happening on the supply side is more troubling. HRSA's 2025 behavioral health workforce analysis put the national mental health professional adequacy rate at just 27%,[3] meaning the country currently has roughly a quarter of the psychiatric providers needed to close shortage designations. The projected shortfall climbs to nearly 100,000 mental health counselors by 2038.[3] As of August 2024, more than 122 million Americans live in a federally designated Mental Health Health Professional Shortage Area,[3] which is over one-third of the entire population.
This behavioral health access gap isn't confined to rural counties or historically underserved markets. It runs through every geography and every health plan design. And while workforce shortages are part of the picture, a significant portion of this gap has nothing to do with provider supply. It comes down to whether plans know who in their network is available, practicing, and able to take a new patient. That's a network management problem, not a workforce problem.
Ghost Networks in Health Plans Are Now a Federal Enforcement Priority
The federal government has been quietly building a case, and the findings aren't flattering.
The HHS Office of Inspector General's October 2025 data brief looked at Medicare Advantage and Medicaid managed care plans across ten counties and found that plans inaccurately list 72% of their in-network behavioral health providers as available. In 15 of the MA plans studied, networks included fewer than 10% of the behavioral health workforce in those counties. Seven plans had no in-network behavioral health providers in certain counties whatsoever, and they were still filing as adequate.[4]
The OIG dug into why: close to three-quarters of the inactive providers in those directories should not have been listed at all,[4] whether because there was no one working at the listed location, the provider had retired, or a network change had never been reflected in the directory.
The GAO reached similar conclusions through a different route. In July 2024, investigators made covert calls to a random sample of behavioral health providers listed in TRICARE network directories and found most listings contained at least one inaccuracy: a wrong address, a disconnected number, a provider who had stopped participating. The TRICARE East contractor's own accuracy reporting showed 83% of listings were correct as of January 2024,[5] which sounds decent until you realize that figure still fell below the contractual threshold required.
On top of that, a U.S. Senate Finance Committee secret shopper investigation found that members trying to book a mental health appointment through their insurance directory succeeded just 18% of the time.[1] And when members do hit dead ends in directories, the financial consequences are real: research shows patients who encounter provider directory inaccuracies are four times more likely to end up with a surprise out-of-network bill.
Ghost networks in health plans aren't just a member experience problem. They're now a legal one. The 2024 federal lawsuit Hecht v. Cigna (in which a February 2025 ruling allowed ERISA fiduciary breach claims tied to directory inaccuracy to proceed) made that shift concrete. Plans that haven't taken directory accuracy seriously are operating with real legal exposure.
Network Adequacy Requirements CMS Is Enforcing Have Materially Changed
The regulatory floor for behavioral health network adequacy has been raised, and the pace of change isn't slowing.
The April 2024 CMS final rule for Contract Year 2025 Medicare Advantage introduced "Outpatient Behavioral Health" as a formal facility-specialty category subject to network adequacy evaluation. That single category now covers marriage and family therapists, mental health counselors, opioid treatment program providers, community mental health centers, and addiction medicine specialists, none of whom were previously part of the adequacy calculus in any structured way. Plans that filed clean networks while leaving these specialties unvalidated no longer have that option.
The network adequacy requirements from CMS enforces CY2026 and CY2027 push further: county-level time-and-distance calculations, explicit specialty alignment across behavioral health sub-types, and telehealth integration for underserved markets. CMS has also put plans on notice that service area expansion applications can be denied on adequacy grounds, a consequence that had been applied inconsistently in earlier cycles but is now firmly in play.
On the parity side, the Department of Labor's January 2025 MHPAEA report to Congress made network composition and adequacy the centerpiece of enforcement. The inclusion of an unredacted 2024 settlement agreement with a multiemployer health plan, which detailed exactly what corrective actions were required for parity-related adequacy failures, effectively published a benchmark that other plans will now be measured against. Behavioral health compliance risk has moved out of the compliance department. It's a leadership-level conversation now.
Telehealth Has Changed the Access Equation, But Network Adequacy Hasn't Kept Up
Any honest discussion of behavioral health network adequacy in 2025 must account for how care delivery has fundamentally changed.
At the peak of the COVID-19 pandemic, telehealth accounted for roughly 40% of mental health and substance use outpatient visits. That number has since settled to around 36% post-pandemic, still far above pre-pandemic levels and not a trend that's reversing. Behavioral health now accounts for roughly two-thirds of all telehealth encounters across the healthcare system.[6] That isn't a niche access channel. It's how a substantial share of behavioral health care actually gets delivered.
The modality breakdown matters because it changes what network adequacy should be measuring. A geographic access analysis that counts only in-person providers captures an increasingly incomplete version of who's actually available to a member. For rural counties, elderly populations, and people managing substance use disorders, audio-only and video telehealth aren't backup options. They are the primary path to care.
CMS has recognized this in policy. Geographic and originating site restrictions for behavioral health, telehealth has been permanently removed for home-based delivery, freeing members from being constrained by provider location. Audio-only telehealth has been specifically and permanently permitted for behavioral health when a member can't or won't use video, which is a critical provision for populations without broadband or who face other digital barriers. CMS has also added new covered services to the behavioral health telehealth list for 2025, including caregiver training and additional substance use disorder services.
But here is where behavioral health network adequacy frameworks are failing to keep pace: most provider data systems still don't accurately capture which providers offer telehealth, in which modalities, or under which plan contracts. A provider seeing patients exclusively via video often sits in the same data field as one with a physical office. A provider technically in-network for telehealth but not enrolled to accept a specific members’ plan isn't reachable. The behavioral health access gap in telehealth is just as real as the in-person gap. It's just harder to detect without data that simultaneously tracks modality, enrollment status, and actual availability.
Provider Data Accuracy in Health Plans Is the Root Problem
Strip away the regulatory developments, the enforcement actions, and the litigation, and you keep arriving at the same place: provider data accuracy of the health plans is broken, and in behavioral health specifically, it's structurally broken.
Most behavioral health providers operate in solo or small-group practices. A psychiatrist who stops accepting new patients on Tuesday isn't sending a formal notification to eight health plans. A therapist who exits a network as reimbursement rates fall below viability doesn't always trigger an immediate directory update. A clinic that closes a satellite location and consolidates into one address may show up as two in-network sites for months. The pace of change in behavioral health provider availability is genuinely faster than any manual reconciliation process can handle.
On the plan side, provider information is being pulled from credentialing platforms, claims data, third-party verification feeds, and periodic outreach surveys, with each source reflecting a different version of reality from a different point in time. When these records conflict, as they routinely do, inaccuracies persist. By the time a member dials the number in the directory, that data may be a year out of date.
The downstream consequences for behavioral health network adequacy are direct. A plan may run its gap analysis and conclude it has 35 psychiatrists accessible within 30 miles of a given county. The actual count of providers who are active, in-network, and accepting new patients might be nine. Every adequacy report, every CMS filing, every parity compliance submission built on that foundation is telling a story the underlying data doesn't support. Poor provider data accuracy in health plans doesn't just create access barriers for members. It keeps plans from even seeing the behavioral health access gap they're responsible for closing.
True Network Adequacy: Moving From the Threshold to the Truth
A different standard has started to take hold across the industry. Rather than asking whether a network clears the regulatory threshold, plans are beginning to ask a harder question: can members actually get an appointment? The two questions sound similar. Operationally, the gap between them is enormous.
True network adequacy means having real-time visibility into which providers in the network are genuinely accessible to members, covering in-person visits, video telehealth, and audio-only care. It means being able to tell the difference between an active provider and a ghost listing at scale, not through a quarterly phone survey but through continuous, validated data. It means geospatial analysis that maps actual access at the sub-county level, not just the density of directory entries. And it means getting ahead of behavioral health compliance risk before a CMS audit or a state investigation forces the reckoning.
For plans still relying on manual processes and periodic audits, making this shift is a real undertaking. But staying with the old approach isn't the safe choice it might appear to be. Filing networks whose accuracy can't be verified is the posture that's drawing federal investigations, state penalties, and federal lawsuits right now.
AI Network Management in Healthcare Makes True Network Adequacy Operational
Behavioral health networks are particularly hard to manage. High provider churn, fragmented data sources, dynamic availability, multiple sub-specialties, and now multiple care modalities: the data problem is genuinely complex, and it compounds over time if not actively managed.
AI network management in healthcare is the practical answer to that complexity. Rather than validating provider data through manual outreach on a quarterly or annual schedule, AI-driven platforms reconcile records continuously, across multiple sources, at a scale no human process can replicate. Inaccuracies get flagged before they become a complaint from a member or a compliance finding. Access gaps become visible while there's still time to address them.
HiLabs' MCheck ® NetworkIQ is an innovative network intelligence solution with a purpose-built R3 engine at its core, assessing every provider attribute for Reliability, Recency, and Relevancy across internal credentialing systems, claims and utilization data, and thousands of real-time external sources. It identifies ghost providers, applies confidence scores to directory listings, and runs geospatial access analysis that accounts for how care is being delivered, both in person and across telehealth modalities. The results are concrete: over 97% of ghost providers identified and cleansed with no impact to plan-level adequacy, with operational cost savings by 40% compared to fragmented manual workflows.
What AI network management in healthcare makes possible isn't just faster compliance. It's a fundamentally different relationship with network data, one where what's in the system exactly reflects what's in the field.
A National Behavioral Health Plan Bets on True Network Adequacy
In April 2026, HiLabs announced that a leading national managed behavioral health organization serving millions of members had deployed MCheck® NetworkIQ.,the timing was deliberate.
Federal regulators had made their expectations clear. State regulators were pursuing multi-million-dollar penalties against plans whose networks looked fine on paper but failed in practice. The OIG had formally recommended that CMS investigate a national provider directory as a response to systemic accuracy failures, a signal that the window for voluntary remediation is narrowing. The plan's existing approach, built on periodic outreach, manual reconciliation, and compliance-threshold monitoring, couldn't provide what the regulatory environment now demands: real-time confidence in what the network looks like.
The deployment shifted the plan from reactive compliance management to verified access. NetworkIQ brings together provider data from multiple internal and external sources, validates it continuously, surfaces access gaps by geography and specialty, and supports proactive provider recruitment to close those gaps before they register with regulators. As HiLabs highlighted in the announcement: "The plans choosing to deploy it today are the ones that will define what behavioral health network adequacy looks like in the next era of managed care."
Behavioral Health Compliance Risk Is Now an Enterprise Problem
Behavioral health network adequacy has become an enterprise risk issue, full stop. The OIG is publishing data briefs that identify specific patterns of failure across plans. State attorneys general have been filing suits. The DOL is using unredacted settlement agreements as public benchmarks. Members are better informed about their parity rights than at any point before, and they are using them.
Staying ahead of behavioral health compliance risk at this scale takes more than a well-maintained spreadsheet and an annual audit cycle. It requires infrastructure capable of validating provider data in real time, detecting access gaps before they widen, and producing the kind of documented evidence that holds up under regulatory scrutiny.
Plans that build that infrastructure now isn't just reducing compliance exposure. They're changing what behavioral health network adequacy means for their members, shifting it from a figure in a filing to something a person in crisis can rely on.
Source:
[1] U.S. Senate Committee on Finance (Majority Staff), 'Medicare Advantage Plan Directories Haunted by Ghost Networks' report (May 2023)
[2] American Hospital Association, 'SAMHSA Releases Results from Annual Survey on Drug Use and Mental Health' (July 2025)
[3] Health Resources and Services Administration (HRSA), National Center for Health Workforce Analysis, 'State of the Behavioral Health Workforce, 2025' report (December 2025)
[4] HHS Office of Inspector General, 'Many Medicare Advantage and Medicaid Managed Care Plans Have Limited Behavioral Health Provider Networks and Inactive Providers' report (October 2025)
[5] U.S. Government Accountability Office, 'Defense Health Care: DOD Should Improve Accuracy of Behavioral Health Provider Information in TRICARE Directories' report (July 2024)
[6] Lo, J., Rae, M., Amin, K., Cox, C., Panchal, N., & Miller, B.F., 'Telehealth Has Played an Outsized Role Meeting Mental Health Needs During the COVID-19 Pandemic' KFF (March 2022)