ABA Denial Management: The Decision Guide for ABA Practices

ABA claim denials don’t get easier at scale. Multi-state payer rules, overlapping services, frequent funder changes, and high staff turnover create constant pressure on revenue and on the teams responsible for collecting it. Behavioral health organizations face 15-30% of claims denied or delayed due to errors [1], 20+ staff hours per month lost to rework, and 30-45 day average reimbursement cycles. For a growing ABA practice, the question isn’t whether you have denials. It’s whether your denial-management approach is keeping pace with your growth.

This guide walks through what ABA denial management actually is, why claims get denied, the codes that matter most, the in-house vs managed  and what “good” looks like in numbers.

What Makes ABA Denial Management Different

Denial management in ABA isn’t medical billing with a slightly different code set. The structural complexity that drives denials is specific to applied behavior analysis:

• BCBA-led prescription with RBT-delivered hours
• Multi-tier supervision documentation
• Multi-state Medicaid and commercial-payer mixes
• Authorization windows that gate every service hour
• Overlapping CPT codes (97151 assessment, 97153 protocol modification by technician, 97155 protocol modification by BCBA, 97156 family adaptive behavior treatment guidance) that get denied for reasons that don’t appear in generic medical-billing playbooks.

Many ABA claim denials originate before the billing process even begins. When clinical documentation, scheduling, and claims processing run in disconnected systems, data gets re-entered manually, and manual entry introduces errors before a claim ever reaches a payer. Multi-state ABA organizations face an added layer: payer policies vary by state and plan, and manual validation can’t keep pace with that volume. When one location’s workflow differs from another’s, denial patterns can develop and go undetected until they show up as a significant accounts receivable problem.

ABA denial management is therefore the operational discipline of preventing denials at the source (documentation, authorization, eligibility, modifier accuracy) and resolving the ones that still happen with the lowest possible days in AR impact. It is more than appeals work. It is a system.

The Cost of Letting Denials Pile Up

The financial pain of denials in ABA shows up in three places: the claims that take longer to pay, the staff hours spent on rework, and the revenue that walks out the door when claims aren’t recovered.

Initial denial rates. Behavioral health organizations face 15-30% of claims denied or delayed due to errors. The Change Healthcare data shows the average initial denial rate across healthcare specialties is roughly 12%, with rates ranging from 5% to 30% depending on specialty and payer mix [1]. For ABA specifically, more than 10% of all autism service claims are initially denied. A practice at 15% has roughly six weeks of revenue cycling through denials at any given time. A practice at 30% has serious operational debt accumulating in its AR.

Rework hours. Practices lose 20+ staff hours per month to rework, with rework costs estimated at approximately $25+ per claim. For a practice running 500 claims per month at a 15% denial rate, that is 75 denied claims, each requiring 10-30 minutes of biller time to identify, gather supporting documentation, and resubmit or appeal. Those hours translate into real labor costs before a single dollar of revenue is recovered.

Reimbursement cycle time. The industry average ABA reimbursement cycle is 30-45 days from claim submission to payment. BillAI by Rethink is engineered around a single KPI: days in AR, defined as the average number of days from when a claim is billed to when it is paid, tracked globally and by funder. Every day a denied claim spends in re-work is a day added to the practice’s cash conversion cycle.

Clinical staff cost. Underneath the billing-team cost sits a less-visible cost: the time BCBAs spend on billing-adjacent work that pulls them away from clinical care. Per RethinkBH’s own BCBA State of the Profession Survey (2025), 61% of BCBAs say administrative burden interferes with their ability to provide direct care, BCBAs spend 38% of their week on billable services that aren’t therapy, and only 25% on direct client care [2]. When denials require BCBA sign-off on appeals or additional documentation, that figure increases.

The combined picture: a 100-clinician practice at a 15% denial rate is leaving meaningful revenue on the table, dedicating significant billing resources to rework, and pulling clinical capacity into administrative work that takes time away from care.

The Most Common ABA Denial Codes

CMS-standard Claim Adjustment Reason Codes (CARCs) and Remittance Advice Remark Codes (RARCs) [3] appear on every Explanation of Benefits. Knowing which codes show up most often in ABA is the difference between treating each denial as a new problem and recognizing repeating patterns that can be prevented at the source. The codes below represent some of the most frequently encountered denial codes in ABA billing.

The two highest-leverage codes for ABA practices are usually CO-197 (authorization) and CO-16 (incomplete information), because both are preventable upstream. CO-50 (medical necessity) and CO-151 (MUE violations) are also common and require structural fixes, not one-off appeals.

The Five Drivers of ABA Denials

Five operational drivers account for the majority of preventable denials in ABA practices. Each one has a root cause and a preventable workflow gap.

1. Missing or incomplete documentation

Documentation issues are the single largest driver of claim denials. Session notes that don’t tie clearly to the goals in the authorized treatment plan, progress notes that don’t substantiate the units billed, or assessment documents missing required elements all surface as CO-16 denials. In ABA specifically, documentation problems compound because hours billed under 97153 (RBT-delivered) require both RBT session documentation and BCBA supervision documentation tied to the same authorization window. Either piece missing breaks the claim.

Prevention: Documentation captured at the point of care, with structured prompts that match payer requirements, tied to the same client record that generates the claim.

2. CPT coding and modifier errors

ABA uses a tight set of CPT codes that get denied frequently when modifiers are wrong. The most common errors include:

• 97153 (adaptive behavior treatment by protocol, technician) missing a 97-series modifier on multi-provider days
• 97155 (adaptive behavior treatment with protocol modification, BCBA) billed on the same day as 97153 without the correct modifier sequencing, triggering CO-97 bundling denials
• 97156 (family adaptive behavior treatment guidance) billed with the wrong place-of-service code
• 97151 (behavior identification assessment) billed without supporting assessment documentation, triggering CO-16

Prevention: Real-time validation against payer-specific modifier rules before the claim leaves the system. The pattern matters: rules must be tied to the payer, not to a generic medical-billing template.

3. Missing or expired prior authorization

CO-197 is the most common single denial code in ABA. Authorizations expire, units run out, plan years roll over, and patients switch plans. Any of those events without a parallel update to the practice’s authorization records produces a denial for services that were clinically legitimate.

Prevention: Authorization tracking integrated with scheduling and clinical documentation, so a session can’t be scheduled or rendered when the authorization is expired or out of units.

4. Missing or insufficient medical necessity documentation

CO-50 denials happen when the payer determines the documentation does not establish that ABA was medically necessary at the prescribed dosage. This is especially common when the treatment plan is older than the payer’s review window, when the diagnosis documentation doesn’t appear in the chart on the date of service, or when the prescribed hours don’t tie clearly to specific clinical goals.

Prevention: Treatment plans that document medical necessity in the format the payer expects, refreshed on the payer’s review cycle, with assessment data tied to the goals that justify the hours.

5. MUE (Medically Unlikely Edit) misuse

The Centers for Medicare and Medicaid Services maintains a Medically Unlikely Edits table that caps the units of a service one provider can bill for one patient on one day. MUE-driven denials in ABA (CO-151) are a published source of preventable revenue loss, particularly for high-frequency service codes [4].

Prevention: MUE thresholds embedded in pre-submission validation. When a claim’s units would exceed the MUE limit for the service code on a date of service, the system holds the claim for biller review rather than letting it submit and get denied.

The Decision Framework: In-House Billing, Managed Billing

The most consequential decision an ABA practice makes about denial management is who is responsible for it. There are two viable models, and the right choice depends on staffing capacity, payer-mix complexity, growth trajectory, and current denial-rate visibility.

In-house billing

The practice employs its own billing staff (one biller, a billing team, or a dedicated RCM department) and runs claims, denials, and appeals internally on practice-owned software.

This fits a practice when:

• There are at least one or two certified billing staff with ABA-specific denial-triage experience on payroll
• Payer mix is relatively contained (single state, fewer than five major payers)
• Leadership wants direct visibility and direct control over denial trends and recovery work
• Cost of an in-house billing team is less than the percent-of-collections fee a managed-billing service would charge

The risk: knowledge concentration in individual billers. When a biller leaves, the practice’s denial-management knowledge often walks out with them.

Managed billing

The practice contracts billing operations to an external team that runs claims, denials, and appeals on the practice’s behalf, typically billed as a percent of collections.

This fits a practice when:

• The practice cannot recruit or retain certified billing staff at acceptable cost
• Payer mix is complex (multi-state, mixed Medicaid and commercial)
• Leadership prefers a predictable per-claim cost over fixed billing-staff payroll
• Internal team capacity is needed for clinical operations or location expansion rather than for billing

The risk: loss of visibility. Practices that outsource billing often lose direct insight into denial patterns by funder, by service line, or by location, and find out about systemic issues only when month-end reports surface them.

RethinkBH supports both models within the same platform. Practices can manage billing in-house with the BillAI software or leverage RethinkBH’s Managed Billing services.

The single most important question to answer before making that decision: do you know your current denial rate by funder and by service line right now? If yes, you have the visibility to run denial management in-house and judge whether you are improving. If no, a managed service or a connected platform’s reporting layer typically delivers that visibility faster than building it internally.

How Connected Billing Reduces Denials at the Source

The structural reason most ABA practices have higher denial rates than they want is fragmentation. Clinical documentation lives in one system. Scheduling lives in another. Claims processing lives in a third. Each handoff is an opportunity for data to be re-entered, transposed, or lost. Each manual transfer is an opportunity for a claim to leave the practice with an error already baked in.

A connected platform changes the math. When session notes, authorization tracking, scheduling, and claims processing share one source of truth, practices are less likely to see situations where data is correct in the clinical record but incorrect on the claim.

Pre-service stage. Eligibility verification runs via 270/271 transactions through the clearinghouse before the session is scheduled. Prior authorization is captured via 278 transactions for participating payers. Co-payment and deductible verification happens up-front, not after the claim comes back partially adjudicated.

Visit and charge-capture stage. The session is rendered, documented, and validated against the rules engine in real time. Charges are captured at the point of care, so the claim is built from clinical data, not transcribed from it after the fact.

Claim review and submission. The Business Rules Engine validates the claim against both system-level rules (maintained by RethinkBH across all clients and payer types) and practice-specific custom rules (configured to the practice’s payer mix and historical error patterns). Claims that fail validation are held for correction with an issue log and task assignment rather than submitted with errors. Claims that pass go through automated submission via Robotic Process Automation.

Insurance-response and follow-up. The clearinghouse returns 997/999 functional acknowledgments and 277CA payer confirmations. The 276/277 transactions request and receive claim status updates. AI Identify & Track monitors every claim against expected milestones from submission to resolution, flagging deviations the moment they occur and triggering automated corrections through RPA. Denied or rejected claims surface on an outstanding worklist with mapped rejection reason codes, ready for biller action.

The result is a touchless claims process that catches issues earlier, resolves them faster, and keeps cash flowing without adding headcount. The pattern is that automation handles volume, and the billing team focuses on exceptions that need human judgment.

The Business Rules Engine plays an important role in this process. It works in two ways:

1. System Rules are maintained by RethinkBH across all clients and payer types, so every client benefits from collective intelligence across the network.
2. Custom Rules are configured at the practice level, learning the practice’s specific payers, error patterns, and resubmission requirements. Every corrective action updates the engine.

The practice owns the knowledge, retained in the system, not dependent on any individual biller. When a biller leaves, the institutional knowledge stays.

What “Good” Looks Like at the Practice Level

Industry averages provide a baseline. Practices running BillAI by Rethink report results that exceed those benchmarks.

The BillAI figures come from the BillAI launch announcement, April 2026 [5]; the industry-average 30-45 day reimbursement cycle is documented as an ABA industry benchmark in the RethinkBH Reimagined Enterprise Messaging.

The KPI to anchor on is days in AR, defined as the average number of days from when a claim is billed to when it is paid, tracked globally and by funder. Lower days in AR means faster cash flow, less administrative burden, and a healthier practice. Every day a stalled claim is caught sooner is a day removed from days in AR, compounded across every claim the practice submits.

For practices running this at scale, BillAI Reporting and Dashboards surface the operational and financial views required to manage it.