Prior Authorization for Behavioral Health: Carelon, Optum, and the AI-Denial Problem

If you run revenue cycle for a treatment center, prior authorization is probably the line item that keeps you up at night. Not because it's intellectually hard. The rules are knowable. It's because it's the single biggest denial driver for the services that pay your bills. Intensive outpatient (IOP), partial hospitalization (PHP), and residential treatment all live or die on an authorization. That authorization has to be requested before admission, defended at every concurrent review, and renewed before it lapses. Miss any one of those touchpoints and a clinically appropriate stay becomes an unpaid one.

Prior-authorization issues account for roughly 20–30% of behavioral health denials in a typical book of business. That's more than coding, more than eligibility, more than documentation taken alone. And the ground is shifting under it. Carelon, Optum/UBH, and Magellan are the behavioral health benefit managers that sit between your center and most of your commercial and Medicaid volume. Increasingly, they route authorization decisions through automated systems. The reporting on payer AI is documented and specific. It matters for any center trying to keep an authorization pipeline staffed.

This guide is the operations-grade reference. It covers what prior authorization actually is and why behavioral health gets hit hardest, how the major benefit managers operate, and the step-by-step process including concurrent and continued-stay review for higher levels of care. It walks through the documented evidence on payer-side AI denials, the documentation that gets auths approved the first time, how to appeal and run a peer-to-peer, and where agentic systems on the provider side change the math. It's evidence-led, not conspiratorial. Automation isn't the villain. One-sided automation is the problem.

Sources: HHS OIG - Prior Authorization Denials in Medicaid Managed Care (2023) · KFF - Medicare Advantage Prior Authorization in 2024

What Prior Authorization Is - and Why Behavioral Health Gets Hit Hardest

Prior authorization (PA) is a payer's requirement that a provider get approval before delivering a service, or the payer won't cover it. The provider submits a request: clinical justification, diagnosis, requested level of care, expected duration. The payer's utilization-management (UM) team then decides whether the service is medically necessary under the member's plan. Only then does the clock on covered care start.

For most of medicine, PA touches a slice of services: an MRI, a brand-name drug, an elective surgery. For behavioral health, it touches the core of what a treatment center sells. Here's why behavioral health carries a disproportionate authorization burden.

Level-of-care intensity. Behavioral health is organized as a ladder: outpatient, intensive outpatient, partial hospitalization, residential, inpatient. Almost every rung above standard outpatient requires authorization to enter and continued authorization to stay. A residential admission isn't one PA. It's an initial auth plus a concurrent review every few days for the length of stay. The higher the level of care, the more authorization touchpoints, and the more revenue riding on each one.

Medical-necessity subjectivity. A payer reviewer assessing whether an MRI is warranted has objective imaging criteria. A reviewer assessing whether a patient needs PHP versus IOP is interpreting symptom severity, functional impairment, risk, and the patient's response to lower levels of care. Those are judgment calls, scored against criteria sets like ASAM (for substance use) or InterQual / MCG (for psychiatric levels of care). That subjectivity is exactly where denials live.

Parity enforcement gaps. The Mental Health Parity and Addiction Equity Act (MHPAEA) requires that payers apply no more restrictive authorization and review requirements to behavioral health than to comparable medical/surgical care. In practice, enforcement is uneven. Federal and state regulators have repeatedly found behavioral health subjected to tighter prior-auth and concurrent-review regimes than medical services. The 2024–2025 MHPAEA rulemaking tightened the comparative-analysis requirements precisely because the gap persisted.

Concurrent review is unique to BH intensity. Most medical PA is a one-time gate. Behavioral health at the IOP/PHP/residential level is gated continuously. The payer re-authorizes every few days, and any one of those reviews can downgrade the level of care or cut the stay short. That's a structural exposure most medical billing never sees.

The Major Behavioral Health Benefit Managers - How They Operate

Most commercial and Medicaid behavioral health benefits aren't managed by the parent insurer directly. They're carved out to a specialized behavioral health benefit manager, a "managed behavioral healthcare organization" (MBHO). That entity owns the provider network, the medical-necessity criteria, and the authorization decisions. Three names dominate.

Sources: Carelon Behavioral Health provider site · Optum Provider Express

A few operating realities worth internalizing, because they change how you staff PA.

Carelon Behavioral Health is the rebranded Beacon Health Options, now under Elevance. For a treatment center, Carelon often controls the authorization for patients whose card says Anthem or a Blue Cross plan. The medical benefit is Elevance, but the behavioral benefit routes to Carelon, with its own criteria and its own portal. The "carelon prior authorization" you submit is governed by Carelon's medical-necessity criteria, not the BCBS medical-side rules, and the two can diverge. Knowing which entity actually adjudicates the BH claim is half the battle.

Optum / UBH uses its Level of Care Guidelines and (where applicable) state-mandated criteria. Optum has been a central named party in the litigation and reporting on payer automation, which makes its UM process the most scrutinized of the three. For higher levels of care, Optum concurrent reviews are frequently conducted by phone, and the peer-to-peer pathway is well-trodden.

Magellan historically managed large public-sector and Medicaid behavioral health contracts and now sits under Centene. Its footprint varies sharply by state and plan. A center operating across multiple states will see Magellan governing one population and not another.

The practical takeaway: your "payer" for a behavioral health claim is often not the insurer on the member's card. Build your authorization workflow around the benefit manager that actually adjudicates: its portal, its criteria set, its concurrent-review cadence. Not the brand name in your patient's wallet.

The PA Process, Step by Step - Including Concurrent and Continued-Stay Review

The authorization lifecycle for a higher level of care isn't a single event. It's a sequence with several failure points. Here's the full path for an IOP/PHP/residential admission.

1. Verify benefits and PA requirements before admission. Before anything else, confirm the member is eligible, the level of care is a covered benefit, and whether PA is required (and by whom: Carelon, Optum, Magellan, or the plan directly). This is also where you capture session/day limits and any step-therapy requirements ("must fail IOP before PHP is authorized").

2. Submit the initial authorization request. Do this within the payer's required window, often before or within 24–72 hours of admission for urgent levels of care. The request carries the diagnosis (ICD-10), the requested level of care, the clinical justification mapped to the payer's criteria (ASAM for SUD, InterQual/MCG or the payer's own LOC guidelines for psychiatric), and the requested duration or number of days/sessions.

3. Initial determination. The UM reviewer approves, denies, or "pends" for more information. An initial approval typically authorizes a specific number of days or sessions, not the whole stay.

4. Concurrent / continued-stay review. This is the part that catches centers off guard. For residential and PHP, the payer re-reviews every few days. You submit updated clinical documentation showing the patient still meets criteria for the current level of care: ongoing symptoms, risk, functional impairment, and why a step-down isn't yet appropriate. The reviewer can authorize more days, downgrade the level of care, or deny continued stay.

5. Discharge / step-down review. When the patient is ready to move down a level, the transition itself often needs authorization for the next level of care. A center that nails the residential auth but forgets to authorize the step-down to PHP will get the PHP claims denied.

6. Peer-to-peer (if denied). When a concurrent review denies continued stay, the ordering clinician can request a peer-to-peer review: a phone conversation with the payer's physician reviewer, usually within a tight window (often 24 hours). More on this below.

Sources: AMA - Prior Authorization Practice Resources · ASAM Criteria overview

The structural insight: a higher-level-of-care stay isn't authorized once. It's re-authorized continuously, and the denial risk is highest at concurrent review, where a payer can quietly downgrade or cut a stay that's clinically justified. Staff your authorization function for the whole arc, not just admission day.

The AI-Denial Problem

Here's the change that's reshaping authorization economics. Payers are increasingly using AI and algorithmic systems to adjudicate claims and prior authorizations, reviewing and rejecting at machine speed and scale. This is documented, not speculative. It's worth walking through the evidence carefully, because the asymmetry it creates is the whole point.

Cigna's "PxDx." ProPublica's 2023 investigation documented an automated system at Cigna that let the insurer's doctors reject more than 300,000 claims over a two-month period, spending an average of about 1.2 seconds on each denials issued without opening the patient file. One Cigna physician described the workflow as clicking and submitting batches: "We literally click and submit… it takes all of 10 seconds to do 50 at a time." That's adjudication at a scale and speed no human review process can match, and no human appeals process can keep pace with.

nH Predict and UnitedHealth. UnitedHealth's use of an algorithm called nH Predict (from naviHealth, which it acquired in 2020) to guide post-acute care denials became the subject of a lawsuit (Estate of Gene B. Lokken v. UnitedHealth Group, filed November 2023). The suit alleged a roughly 90% error rate: that about nine in ten of the algorithm's denials were reversed when patients had the resources to appeal. That figure is an allegation in litigation, not an adjudicated finding, and it should be read that way. But the underlying concern is exactly the dynamic that makes one-sided automation costly for providers: an algorithm generating denials that overturn at high rates on appeal.

The Senate investigation. In October 2024, the U.S. Senate Permanent Subcommittee on Investigations (PSI) released a majority-staff report on major Medicare Advantage insurers: UnitedHealth, Humana, and CVS/Aetna. It found they used AI and algorithmic tools in ways that coincided with sharply higher prior-authorization denial rates for post-acute care. The report is congressional oversight documentation of payers deploying automation in the denial pipeline.

What the burden looks like to clinicians. The AMA's annual prior-authorization physician survey (n≈1,000 practicing physicians) quantifies the downstream cost. 95% report that PA causes care delays, 79% report it can lead patients to abandon treatment, and 26%, more than one in four, report that a PA delay led to a serious adverse event for a patient in their care.

Sources: ProPublica - How Cigna Saves Millions by Rejecting Claims Without Reading Them (2023) · STAT News - Denied by AI (2023) · KFF - Medicare Advantage Prior Authorization (2024)

Now the honest counter-framing, because it matters. Automation isn't inherently bad. The same systems that deny at machine speed can approve legitimate requests instantly. And not every denial is wrong. Some requests genuinely don't meet criteria, and a fast "no" beats a slow one. The problem isn't AI. The problem is the asymmetry. Payers adjudicate at machine scale while most treatment centers still appeal by hand, one fax and one phone call at a time. The most telling number above is KFF's: only about 11.7% of Medicare Advantage denials are appealed, yet 81.7% of those appeals succeed. Most denials stick not because they're correct, but because nobody has the bandwidth to fight them. The answer isn't to rail against automation. It's to give the provider side comparable leverage.

Documentation That Gets Auths Approved the First Time

The cheapest authorization is the one approved on first submission. Every pend, every request for more information, every denial-then-appeal is staff time and delayed cash. First-pass approval comes down to one thing: submitting clinical documentation that maps directly to the payer's medical-necessity criteria for the requested level of care.

What reviewers are looking for, and what to put in front of them:

• Diagnosis with specificity. The ICD-10 code and a clinical picture that justifies the requested level of care. Not just "F33.1, recurrent MDD," but the severity, risk, and functional impairment that make this level necessary now.
• Criteria-mapped justification. For SUD, write to the ASAM dimensions explicitly. For psychiatric levels of care, write to the payer's LOC guidelines (Optum's, Carelon's) or the InterQual/MCG criteria the payer uses. Don't make the reviewer translate your narrative into their checklist. Do it for them.
• Risk and acuity, objectively stated. Suicidal/homicidal ideation status, recent attempts, withdrawal risk, inability to maintain safety at a lower level. Vague "patient is struggling" language fails; "SI with plan, no means restriction possible in home environment" passes.
• Failed lower level of care, when relevant. Many criteria sets require evidence that a lower level was tried and was insufficient. Document the IOP that didn't hold before requesting PHP.
• Functional impairment with specifics. Work, school, relationships, ADLs. Make it concrete, current, measurable.
• Why not a step-down, at concurrent review. This is the single most missed element. Continued-stay reviews fail when the note documents progress but never explains why the patient still needs the current level rather than the next one down. Both can be true, improving and not yet ready to step down, but you have to write the second part.

A note that reads "patient doing better, continue current level of care" is a continued-stay denial waiting to happen. A note that reads "PHQ-9 down from 22 to 16, but persistent passive SI and inability to maintain structure independently; step-down to IOP premature pending two more weeks of stabilization" gets the days approved.

Appealing PA Denials - and the Peer-to-Peer

When a prior auth or continued-stay request is denied, you have two main tools: the formal appeal and the peer-to-peer review. Use both, and use them on the strongest denials first.

Peer-to-peer (P2P). When a concurrent review denies continued stay, the ordering clinician can request a phone review with the payer's physician reviewer, usually within a tight window, often 24 hours of the denial. This is your fastest path to reversing a continued-stay denial mid-treatment. It's underused because it requires getting a busy clinician on the phone, prepared, on short notice. Preparation is everything: have the clinician ready to speak directly to the specific criterion the denial cited, with the chart open and the risk factors at hand. Centers that run a standing P2P playbook (who calls, within what window, with what prep sheet) reverse far more denials than those that scramble each time.

Formal internal appeal. If P2P fails or isn't available, file the internal appeal with the payer. Standard internal appeals typically resolve in about 30 days; urgent appeals in roughly 72 hours. Attach the documentation that addresses the exact denial reason. Not a resend of the original packet, but a targeted rebuttal to the criterion the payer cited.

External / independent review. If the internal appeal fails, most plans allow an external review by an independent organization, adding roughly 45–60 days.

The strategic point is the one KFF's data makes unavoidable: appeals work far more often than centers act on. About 82% of Medicare Advantage prior-auth appeals succeed, but only ~12% of denials get appealed at all. The binding constraint isn't whether you'd win. It's whether you have the operational capacity to file. Build the capacity, and prioritize: appeal the high-dollar continued-stay denials and the ones where your documentation clearly meets criteria, before chasing every low-value denial at a loss.

Sources: KFF - Medicare Advantage Prior Authorization (2024) · AMA - Prior Authorization

AI / Agentic Systems on the Provider Side

AI in revenue cycle has moved past "better OCR." What's emerging now is agentic systems that handle entire workflows end-to-end, with human review at the edges. For prior authorization specifically, that shift is the provider's answer to the asymmetry above. It meets machine-speed adjudication with machine-speed preparation, so a center isn't bringing a stopwatch to an algorithm fight. The two agents that map directly to the authorization workflow are the benefits verification agent and the benefits accuracy agent.

Here's what that looks like in practice. Before a patient is admitted, the benefits verification agent pulls benefits across 5,000+ payers. It logs into payer portals, and when a portal can't surface what's needed, it places an actual voice call to the payer and completes the conversation with the rep to retrieve accurate benefits. For prior authorization, that means surfacing which services require PA, who adjudicates them (Carelon, Optum, Magellan, or the plan directly), and what the session or day limits are. Before the first visit, not after the first denial. The benefits accuracy agent then validates that raw data and surfaces the non-obvious details that drive auth denials: prior-auth requirements by level of care, session limits, deductible and copay structure for IOP versus outpatient versus testing, and step-therapy rules. When the agent learns a payer caps psychotherapy at, say, 20 sessions a year, it flags that patient's submissions as they approach visit 18. A foreseeable, avoidable denial gets caught before it happens rather than appealed after. The same logic applies to a concurrent-review cadence: knowing a residential auth covers seven days means flagging day five for the next review, not discovering the lapse on an EOB.

Because these agents are part of a continuously learning system, what one learns feeds the others. When a denial reveals that a payer is rejecting a level-of-care request for a missing element, that signal can feed back into documentation guardrails in Supanote. The next clinical note then prompts for exactly what the reviewer wanted to see, turning a one-time denial into a standing prevention.

Honest limits. Agents can't override a payer's policy or make a payer authorize care that doesn't meet criteria. They can't make a clinically thin justification billable, and they won't fix a workflow that's broken upstream. What they can do is the high-volume, rules-based, time-sensitive work that humans get wrong simply because there's too much of it: tracking every auth window, every session limit, every concurrent-review date across every payer and every patient, without dropping one. That's the work that decides whether your authorization pipeline holds.

If you're interested, book a demo here to learn more.

Quick Wins

Things a treatment center can act on this week.

1. Map every active payer to its real benefit manager. Build a one-page reference: which payers route BH auths to Carelon, which to Optum/UBH, which to Magellan, which adjudicate directly, with each one's portal and concurrent-review cadence. Your team is probably guessing on some of these.
2. Capture session/day limits and PA requirements at the benefits check, not at denial. Add prior-auth-required, session-limit, and step-therapy fields to your pre-admission verification so nothing surfaces for the first time on an EOB.
3. Build a concurrent-review calendar. For every active higher-level-of-care patient, track the date the current authorization expires and submit the continued-stay review before it lapses. The lapse is where the revenue leaks.
4. Stand up a peer-to-peer playbook. Define who requests the P2P, within what window (often 24 hours), and with what prep sheet open. Then a denial mid-treatment triggers a fast, prepared call instead of a scramble.
5. Appeal the high-dollar continued-stay denials first. Given ~82% appeal-success rates on prior auth, the constraint is capacity, not merit. Prioritize the denials where you clearly meet criteria and the dollars are biggest.
6. Audit your last 25 PA denials by reason. Bucket them: wrong benefit manager assumed, criteria not addressed, window missed, concurrent review failed. The shape tells you where to invest first.

FAQ

Q: Who actually decides my behavioral health prior authorizations - the insurer or a benefit manager?

A: Often a benefit manager, not the insurer on the member's card. Many Anthem/Elevance and BCBS plans carve behavioral health out to Carelon Behavioral Health (formerly Beacon). UnitedHealthcare routes most BH to Optum / United Behavioral Health. Magellan (now under Centene) manages select commercial and a lot of public-sector/Medicaid BH. The benefit manager owns the criteria, the portal, and the decision. Build your workflow around it, not the brand on the card.

Q: What's the difference between an initial authorization and a concurrent review?

A: The initial authorization gates entry into a level of care and usually approves a set number of days or sessions, not the whole stay. Concurrent (continued-stay) review re-authorizes the stay every few days, requiring updated clinical documentation that the patient still meets criteria for the current level of care. For residential and PHP, concurrent review is where most denials happen, because it recurs continuously and each cycle can downgrade or cut the stay.

Q: Why does behavioral health get hit with so much more prior authorization than medical care?

A: Three reasons. Behavioral health is organized by level of care, and almost every level above outpatient requires authorization to enter and to stay. Medical necessity for psychiatric and SUD care is more subjective than, say, imaging, which gives reviewers more room to deny. And parity enforcement is uneven: despite MHPAEA, regulators keep finding behavioral health subjected to tighter review than comparable medical care.

Q: What is "carelon prior authorization" and how is it different from BCBS prior auth?

A: When a patient's medical benefit is Elevance/Anthem or a Blue Cross plan but their behavioral benefit is carved out to Carelon, the BH authorization is governed by Carelon's medical-necessity criteria and submitted through Carelon's portal, not the medical-side BCBS rules. The two can diverge. Confirm which entity adjudicates the BH claim before you submit, or you'll write to the wrong criteria.

Q: How fast can payers deny an authorization now, and does AI really play a role?

A: It's documented. ProPublica reported Cigna's PxDx system let doctors reject 300,000+ claims in two months at about 1.2 seconds each, without opening files. UnitedHealth's nH Predict algorithm for post-acute denials drew a lawsuit alleging a ~90% error rate (an allegation in litigation, not a finding). A 2024 Senate PSI report documented major MA insurers using algorithmic tools alongside rising prior-auth denials. The point isn't that AI is evil. It's that payers adjudicate at machine speed while most centers still appeal by hand.

Q: If denials are so often overturned, why don't more centers appeal?

A: Capacity, not merit. KFF found only about 11.7% of Medicare Advantage prior-auth denials are appealed, but 81.7% of those appeals succeed. Most denials stick simply because no one has the bandwidth to fight them. The economic implication is stark: a lot of denied-but-winnable revenue is being written off for lack of staff time, which is exactly the gap automation on the provider side is meant to close.

Q: What's a peer-to-peer review and when should I use one?

A: A peer-to-peer (P2P) is a phone conversation between your ordering clinician and the payer's physician reviewer, used to challenge a denial, most commonly a continued-stay denial mid-treatment. The window is tight, often 24 hours from the denial. It's your fastest reversal path, but it requires getting a prepared clinician on the phone fast. Centers with a standing P2P playbook reverse far more denials than those that improvise each time.

Q: How do I get an authorization approved on the first submission?

A: Map your clinical justification directly to the payer's criteria set: ASAM dimensions for SUD, the payer's LOC guidelines or InterQual/MCG for psychiatric levels. State risk and acuity objectively, document failed lower levels of care when criteria require it, and at concurrent review, explicitly explain why a step-down isn't yet appropriate. The single most common continued-stay failure is a note that shows progress but never says why the patient still needs the current level.

Q: How long do appeals take?

A: Internal appeals typically resolve in about 30 days standard, 72 hours for urgent. If the internal appeal fails, an external/independent review adds roughly 45–60 days. For a mid-treatment continued-stay denial, the peer-to-peer (often within 24 hours) is faster than any formal appeal and usually the right first move.

Q: Does prior authorization apply to standard outpatient therapy, or just higher levels of care?

A: It varies by payer, but the heaviest authorization burden is on IOP, PHP, residential, and inpatient. Standard outpatient psychotherapy is more often governed by session limits (e.g., 20 visits/year) than by upfront PA, though some payers do require auth after a threshold number of visits. Capture both the PA requirement and the session limit at the benefits check.

Q: We operate across multiple states. Why do our authorization rules differ so much?

A: Because the benefit manager and the criteria vary by state and plan. Magellan might govern one state's Medicaid BH and not another's; a Carelon contract in one market may use different criteria than another. Medicaid managed-care BH rules are especially state-specific. A multi-state center needs its payer-to-benefit-manager map maintained per state, not nationally.

Q: Can software actually prevent prior-auth denials, or just process them faster?

A: Both, but the prevention is the bigger lever. Surfacing PA requirements, the right benefit manager, and session/day limits before admission prevents the foreseeable denials. Tracking concurrent-review dates prevents lapse denials. What software can't do is override a payer's policy or make care that doesn't meet criteria billable. It can't make a payer say yes to a request that genuinely fails the criteria. It removes the avoidable losses, not the legitimate ones.

References

• American Medical Association. (2024). 2024 AMA Prior Authorization Physician Survey. https://www.ama-assn.org/system/files/prior-authorization-survey.pdf
• American Medical Association. (2025). Prior Authorization Practice Resources. https://www.ama-assn.org/practice-management/prior-authorization
• American Society of Addiction Medicine. (2025). The ASAM Criteria. https://www.asam.org/asam-criteria
• Carelon Behavioral Health. (2025). Provider Resources. https://www.carelonbehavioralhealth.com/providers
• Centers for Medicare & Medicaid Services. (2024). Mental Health Parity and Addiction Equity Act (MHPAEA) Final Rule. https://www.cms.gov/marketplace/private-health-insurance/mental-health-parity-addiction-equity
• HHS Office of Inspector General. (2023). High Rates of Prior Authorization Denials by Some Plans and Limited State Oversight Raise Concerns About Access to Care in Medicaid Managed Care (OEI-09-19-00350). https://oig.hhs.gov/reports/all/2023/high-rates-of-prior-authorization-denials-by-some-plans-and-limited-state-oversight-raise-concerns-about-access-to-care-in-medicaid-managed-care/
• ICANotes. (2025). Behavioral Health Billing Metrics & KPIs 2025. https://www.icanotes.com/2025/09/24/behavioral-health-billing-metrics-kpis/
• KFF. (2024). Medicare Advantage Insurers Made Nearly 53 Million Prior Authorization Determinations in 2024. https://www.kff.org/medicare/medicare-advantage-insurers-made-nearly-53-million-prior-authorization-determinations-in-2024/
• KFF. (2025). Claims Denials and Appeals in ACA Marketplace Plans in 2024. https://www.kff.org/patient-consumer-protections/claims-denials-and-appeals-in-aca-marketplace-plans-in-2024/
• Optum. (2025). Provider Express - Behavioral Health Provider Resources. https://www.providerexpress.com/
• ProPublica. (2023). How Cigna Saves Millions by Having Its Doctors Reject Claims Without Reading Them. https://www.propublica.org/article/cigna-pxdx-medical-health-insurance-rejection-claims
• STAT News. (2023). Denied by AI: How Medicare Advantage Plans Use Algorithms to Cut Off Care. https://www.statnews.com/2023/03/13/medicare-advantage-plans-denial-artificial-intelligence/
• U.S. Senate Permanent Subcommittee on Investigations. (2024). Refusal of Recovery: How Medicare Advantage Insurers Have Denied Patients Access to Post-Acute Care. https://www.hsgac.senate.gov/subcommittees/investigations/
• HFMA. (2025). Navigating Medical Necessity Denials: Strategies for Successful Resolution. https://www.hfma.org/revenue-cycle/navigating-medical-necessity-denials-strategies-for-successful-resolution/
• Global AHS. (2025). Behavioral Health Billing Denial Rate Benchmarks. https://www.globalahs.com/behavioral-health-billing-denial-rate-benchmarks/
• Supahealth. (2026). Aggregate billing, authorization, and denials data from 200+ behavioral health practices. Internal dataset.

For more on behavioral health revenue operations, see our guides on Why Behavioral Health Claims Get Denied and Behavioral Health Billing for Treatment Centers.