SUD Benefit Limitations, One of the Most Overlooked Causes of Behavioral Health Claim Denials

Why This Matters

SAMHSA’s most recent National Survey on Drug Use and Health estimated that 52.6 million Americans needed treatment for a substance use disorder in the past year. Only about 10.2 million of them got it. That’s roughly four out of five people who needed care going without it. Cost, stigma, and provider shortages get most of the blame, and fairly so. But there’s a quieter contributor sitting inside that gap: care that was delivered, billed, and then denied anyway because of a benefit limitation nobody flagged at intake.

Most writing on this topic treats the denial as the starting point and jumps straight to the appeal, the forms, the phone calls. That’s useful information, but it skips the more interesting and more actionable question. Why did the denial happen in the first place, and what would it have taken to see it coming? For SUD benefit limitations specifically, the answer is almost always the same: the limitation was written into the plan design from day one. By the time the denial letter shows up, the moment where it could have been prevented has already passed.

For providers, this isn’t an abstract policy debate. It’s a recurring hit to cash flow that looks, on paper, like a billing error, when it’s actually a structural feature of how SUD benefits get built.

What Are SUD Benefit Limitations?

SUD benefit limitations are restrictions a health plan places specifically on substance use disorder treatment, separate from and often stricter than whatever applies to general behavioral health or medical care. In regulatory language these usually fall under what’s called a nonquantitative treatment limitation, or NQTL, a term that covers non-dollar, non-visit-count restrictions like prior authorization requirements, medical necessity criteria, and network adequacy standards, as opposed to a simple quantitative cap like “20 visits per year.”

In practice, they show up as a hard visit or day cap on outpatient SUD counseling regardless of clinical need. A step therapy requirement, where a lower level of care has to be tried and documented as unsuccessful before the plan will pay for something more intensive, an approach lifted straight from pharmacy benefit design and applied to addiction treatment. Or an outright exclusion of specific services, like recovery coaching or certain intensive outpatient formats, even when the same plan covers those same categories of service under its general behavioral health benefit.

Here’s what makes this genuinely hard to catch: a generic eligibility check often won’t surface, even though the underlying data sometimes exists and can be pulled with the right request. A patient can show as “active” with general behavioral health coverage while a separate SUD-specific rider underneath tells an entirely different story, one that a check run under the wrong service code will simply never ask about. Two people holding the exact same insurance card can have wildly different SUD benefits depending on their employer’s specific plan design or which state regulates that policy. More on exactly what a targeted check can and can’t tell you later in this piece.

Common Services Impacted

Some services take this hit harder than others, and the pattern maps closely to ASAM’s Criteria, the framework the American Society of Addiction Medicine publishes for determining medical necessity and level of care in SUD treatment. Several major commercial payers reference ASAM’s Criteria in some form during utilization review, though exactly which plans use it, and how strictly, varies by payer and is worth confirming against each payer’s own published medical necessity policy rather than assuming uniform adoption. Intensive outpatient programs and partial hospitalization programs draw frequent step-therapy requirements, since payers want documented failure at a lower ASAM level before they’ll authorize something more intensive. Residential treatment regularly runs into hard day limits or a concurrent review cycle that general behavioral health admissions never face. Medication-assisted treatment, despite decades of evidence behind it, still gets tangled in prior authorization or dosage caps at some plans, an irony that isn’t lost on anyone who treats opioid use disorder for a living. Even routine outpatient counseling for a SUD diagnosis can carry a lower visit cap than identical counseling billed under a general mental health diagnosis.

The thread running through all of it: these services get treated more restrictively because the primary diagnosis is substance use related, not because of anything about the clinical intensity of the care itself. That’s precisely the pattern the Mental Health Parity and Addiction Equity Act was written to prohibit.

Why These Claims Get Denied

A handful of things tend to be happening at once.

Benefit verification at intake usually checks general behavioral health coverage without drilling into SUD-specific riders, and that’s rarely a staffing failure so much as a time constraint. Those carve-outs live deep in plan documents that aren’t fast to pull, and thoroughly verifying every new patient’s SUD-specific benefit takes real time most intake teams don’t have.

Authorization cycles move on their own schedule, too. A plan might require reauthorization every five or ten SUD sessions, on a timeline that has nothing to do with the practice’s standard authorization tracking. Miss that window by a few days and every session billed after it can deny, regardless of whether the treatment itself was clinically appropriate.

Nearly three-quarters of health plans require prior authorization to continue SUD treatment, and that’s one of the biggest points where claims get held up or denied. The good news is we can catch those prior auth requirements right during eligibility verification, just by selecting Behavioral Health as the service type.

Here’s what that looks like in practice. A patient starts an intensive outpatient program with a plan that covers ten IOP sessions before requiring reauthorization, a limit that sat quietly in a rider nobody pulled at intake because the general eligibility check came back active. Sessions one through ten go fine. Session eleven bills, and it denies, not because the treatment stopped being appropriate, but because nobody was tracking a ten-session clock that had nothing to do with the practice’s normal thirty-day authorization cycle. By the time billing catches it, sessions twelve and thirteen have already happened too, so now it’s three denied claims instead of one, and the reauthorization request that should have gone in before session eleven is going in after the fact instead, which some plans simply won’t backdate. That’s the whole mechanism in miniature: a limit that existed from day one, invisible to a standard check, surfacing only once the damage has already compounded.

Visit and day caps are still one of the most persistent, and least visible, reasons SUD claims stall or get denied. Providers end up stuck fighting the same authorization battles over and over, right in the middle of treatment.

Then there’s medical necessity documentation. ASAM’s Criteria, now in a Fourth Edition released in October 2023, assesses patients across six clinical dimensions, from withdrawal risk to recovery environment, to determine which level of care is justified. The six-dimension structure itself isn’t new to this edition, the Third Edition used six as well, but the Fourth reordered them and swapped out the old readiness-to-change dimension for a broader person-centered one. Adoption of the newer edition also varies by payer and by state, and plenty of reviews still run on the Third Edition’s framework rather than the Fourth. Either way, if a chart doesn’t map cleanly to what a payer’s ASAM-based review expects to see, the claim can be denied even when the clinical judgment behind it was entirely sound.

There’s also a confidentiality layer that’s unique to SUD claims and doesn’t touch general behavioral health at all. Substance use disorder treatment records carry extra protection under 42 CFR Part 2, which restricts how that information can be used or disclosed, including to a payer conducting utilization review. Historically, Part 2 required a separate, specifically scoped consent for each disclosure, so if a practice’s intake consent form wasn’t worded to cover the exact review a payer needed to do, the payer legally couldn’t get what it needed to process the claim, and the claim could stall or deny on a technicality that had nothing to do with the treatment itself. A 2024 final rule aligned Part 2 more closely with HIPAA and now allows a single consent to cover future treatment, payment, and operations disclosures, which should ease this over time, but full compliance wasn’t required until February 2026. Plenty of practices are still running on older consent language that predates the fix, which means this particular failure mode hasn’t gone away just because the rule changed.

And this is where I think the industry undersells how deep the problem runs. This isn’t just sloppy documentation on the provider side. It’s also a track record of payers writing their own internal medical necessity guidelines that diverge from generally accepted clinical standards, which is exactly the pattern MHPAEA’s nonquantitative treatment limitation rules were written to catch and prohibit.

Which brings up the regulatory backdrop providers are currently operating in, because it’s more unsettled than most billing teams realize. The requirement for plans to document and justify their NQTLs with formal comparative analyses actually predates the 2024 rule. It came from the 2021 Consolidated Appropriations Act, and the Departments of Labor, HHS, and Treasury have been reviewing and reporting on plan compliance since 2022. The September 2024 rule built on that existing requirement with more specific documentation standards. The track record under the original CAA requirement wasn’t encouraging: the Department of Labor’s early reports to Congress, covering more than 150 plans and over 200 distinct nonquantitative treatment limitations, found that not one comparative analysis reviewed was sufficient on first submission. Then, in May 2025, the Departments paused enforcement of the newer portions added by the 2024 rule while a legal challenge from the ERISA Industry Committee played out, and as of a March 2026 court filing, they’ve told the court they intend to issue a notice of proposed rulemaking, a formal step toward a revised rule rather than a finished one, by the end of 2026. The underlying statutory parity requirements from the 2013 rule and the 2021 Consolidated Appropriations Act amendments are still fully in force and still being enforced. But the newer, more specific NQTL documentation requirements that would have made SUD-specific limitations far more visible and accountable are, for now, in limbo.

For providers, the practical takeaway is blunt: the rules meant to make SUD benefit limitations more transparent are exactly the ones currently on pause. That makes catching these limitations at intake more important, not less, because you can’t count on regulatory pressure to surface them for you right now.

The Financial Impact

One denied SUD claim rarely sinks a practice on its own. Volume and timing are what turn it into a real problem. Because if a plan applies a visit cap or reauthorization schedule to one patient, it’s applying that same rule to every patient it covers. A practice that missed the limitation for one person almost certainly missed it for a dozen others on the same plan.

The numbers around denials generally are already sobering enough. Experian Health’s 2025 survey of 250 revenue cycle leaders found that 41 percent reported at least one in ten of their claims getting denied. Premier Inc.’s 2024 national survey of over 500 hospitals put the cost of adjudicating and reworking claims at $25.7 billion industry-wide for 2023 alone, up from $19.7 billion the year before, driven by both a rising denial rate and a rising cost per claim, which Premier calculated at $57.23 in 2023. Reworking a single denied claim runs around $25 for a physician practice, according to the Journal of AHIMA, considerably less than the $181 Journal of AHIMA figure often cited for hospital settings, since the two face very different claim volumes and complexity. Providers overturn roughly 54 percent of the denials they actually bother to appeal, per Premier’s data. The catch, per the same AHIMA analysis, is that as many as 60 percent of denied claims are never resubmitted at all, a ceiling estimate rather than a typical rate, and other researchers have put the figure even higher. Whatever the precise number, a meaningful share of denied claims simply get written off. That’s not because the care wasn’t warranted. It’s because appealing takes staff time a lean behavioral health practice often doesn’t have lying around.

Behavioral health sits inside that baseline unevenly. The American Society of Health Economists’ review of denial patterns found that commercial claims were initially denied at an average rate of 14 percent in 2022 and 2023, and specifically flagged mental health services, alongside lab tests and procedures with complex prior-authorization requirements, as among the categories most frequently subject to denial in the first place. SUD-specific limitations layer directly onto that already elevated baseline.

And here’s the distinction that matters most, the one that gets lost when denials get treated as one undifferentiated category. A denial caused by a documentation gap or a coding error is fixable after the fact. A denial because the patient’s plan genuinely doesn’t cover session eleven of outpatient SUD counseling is not, no matter how well the appeal letter is written. I haven’t found data that isolates appeal success rates specifically for benefit-limitation denials versus provider-error denials, so I won’t claim a number here, but the underlying logic holds regardless: an appeal can only fix what was actually wrong, and a plan enforcing its own written terms isn’t wrong by its own definition. The service wasn’t covered. The plan is often correct according to its own terms, terms nobody surfaced before treatment started.

Prevention Starts Before the First Appointment

The single highest-leverage moment to catch an SUD benefit limitation is intake, before a first session ever happens. That means going past a general eligibility check to specifically verify whether the plan carries a separate SUD benefit distinct from general behavioral health, what visit or day limits apply and over what benefit period, whether step therapy or fail-first requirements exist for the ASAM level being considered, what the reauthorization schedule looks like and how many sessions it actually covers, and whether specific services like recovery coaching, certain IOP formats, or particular MAT protocols are excluded outright.

That’s more upfront work, no argument there. But it’s a fraction of the effort an appeal demands after the fact, and it means the practice can have an honest coverage conversation with the patient before treatment starts instead of after a surprise bill arrives. It also means building SUD-specific reauthorization dates into whatever tracking system the practice runs, kept separate from general behavioral health authorization tracking, because the timelines frequently don’t line up. And intake is the start of this, not the end of it. Visit counts and remaining sessions shift as treatment progresses, so the check that matters most for a patient in week nine is the one run before that week’s appointment, not the one run in week one.

How Should a Good Biller Handle It

Good billers earn their keep on three things, and they all happen before or during the visit, not after the claim gets denied.

1. Checking the correct benefits at the time of eligibility verification. This isn’t just confirming the patient has coverage. It means pulling the actual SUD (substance use disorder) benefit details, including visit limits, prior authorization requirements, and any carve outs for the specific service being billed. A biller who just checks “active coverage” and moves on is setting up a denial down the line.
2. Routing flagged cases back the moment a visit cap is reached. If a client has used 20 of 20 allowed outpatient visits, that needs to surface immediately, not get discovered when the claim bounces back weeks later. A good biller treats the cap as a trigger point, not a line item to reconcile after the fact.
3. Tracking visit counts closely enough to work hand in hand with front desk staff and providers. This is the piece that actually prevents the problem instead of just catching it. When billing keeps a running count and flags it early, front desk can flag the client at check-in and providers can plan the treatment course around what’s actually left, rather than continuing to schedule visits that won’t get reimbursed.

How Technology Can Help

I want to be specific here rather than wave at “the right software” the way a lot of writing on this topic does, because the honest answer is that standard eligibility checks, the ones every practice already runs before a patient’s first visit, can surface a meaningful chunk of this. Just not automatically, and not if the check is run the same generic way every time.

Insurance systems generally distinguish between different categories of behavioral health service when they respond to a coverage check, and substance use treatment often has its own distinct category, separate from general mental health. A generic check that only asks about broad coverage won’t necessarily return the details tied to that more specific category. A practice has to specifically ask about SUD coverage, not just mental health coverage in general, to have a chance of getting SUD-specific answers back. Payer support for this varies, so what works for one insurer won’t necessarily work for another, and testing each payer directly is the only reliable way to know what actually comes back.

When an insurer does return the right level of detail, visit and day limits often come back in a genuinely structured way rather than buried in a policy document somewhere. A response can state something as specific as a set number of allowed outpatient visits per year, along with how many of those have already been used. That’s a real, demonstrated capability, though whether every insurer returns that level of detail specifically for SUD coverage, as opposed to general behavioral health coverage, isn’t something a practice can assume without checking. When an insurer does return it this cleanly, a system can flag a patient approaching that limit automatically, before the limit gets billed past and denied. Prior authorization requirements can often be checked the same way, telling a practice outright whether a specific service needs sign-off before it’s delivered.

Carve-outs are where this gets more honest about its limits. When a payer routes behavioral health or SUD benefits to a separate administrator, the standard coverage check usually won’t hand over that administrator’s actual limitations. What it often will do is name the administrator and hand over the patient’s identifying information with them. That’s not nothing. It tells a practice a second, targeted check against that specific administrator is what’s actually needed, rather than treating the first response as the full picture. But it does mean one check often isn’t enough for a carved-out SUD benefit, and a workflow built assuming one check equals one answer will still miss things.

There are real gaps technology doesn’t close. Step therapy and fail-first requirements, the specific clinical documentation a payer’s medical necessity review expects, and a plan’s exact criteria for authorizing a higher level of care generally don’t show up in a routine coverage check at all. Those sit in payer policy manuals and utilization review guidelines that still require someone to go find and read.

So the honest version of “how technology can help” isn’t that software makes SUD benefit limitations disappear. It’s that the data to catch a real chunk of them, the visit caps, the prior auth flags, the existence of a carve-out, is often already available through the checks practices are running anyway. Most just aren’t set up to ask the more specific questions or look for the specific details where that information lives, and they’re checking once at intake instead of before every visit when the numbers that matter most are the ones still changing.

Final Thoughts

Go back to that SAMHSA number from the top of this piece: roughly four out of five people who needed SUD treatment in the past year didn’t get it. Most of the reasons behind that gap are bigger than any one practice can fix. Stigma, cost, provider shortages, none of that changes because a billing team gets sharper at intake.

But some of that gap is quieter and more solvable than it looks. It’s the patient who did start treatment, whose care was appropriate and delivered, and who ends up back at square one anyway because a session-eleven limit nobody surfaced turned into a bill nobody expected. That version of the gap isn’t a stigma problem or an access problem in the usual sense. It’s a visibility problem, and visibility problems are the ones a practice can actually do something about before the first appointment instead of after the tenth.