Writing Defensible ABA Treatment Plans: Aligning Clinical Goals with Medical Necessity for Autism Services

What Medical Necessity Means When a Payer Reads Your Plan

Medical necessity (the test payers use for funding) decides what gets paid for in ABA (therapy built on behavior science). Dr. Jessica Osos defined it as a health care standard. The question is whether a service is needed to diagnose, treat, or relieve a health condition. The service must fit generally accepted medical standards. It cannot rest on convenience. That includes convenience for the client, the family, or the provider.

She used Botox to make the idea concrete. A doctor may prescribe Botox to treat migraines. Insurance often covers that. Botox for fine lines is a choice, so it sits outside coverage. Same drug. Different reason. Different answer.

Here is the catch. Health care has no single definition of medical necessity. Each plan writes its own rules and its own benefit language. So access shifts from plan to plan. That gap stings in ABA, where plans often call for many hours over many months.

Osos tied the pressure to the Triple Aim (a plan for improving health systems) from the Institute for Healthcare Improvement. Its three goals are a better care experience, better health for whole groups, and lower cost per person. The cost goal creates real pressure. Insurers get stricter about services that take a lot of time and money. ABA sits right in that group.

medical necessity is a healthcare standard that's used to determine whether a service is essential to diagnose, treat, or to relieve a health condition

From the talk — Dr. Jessica Osos

From Lab Research to Insurance Funded Autism Services

ABA did not start in clinics or insurance plans. It started in basic lab research (studies of how behavior works). Researchers soon saw that these rules applied to behavior that mattered in daily life. In 1968, Baer, Wolf, and Risley published the seven dimensions of ABA (the seven features defining our field). The field was named as analytic, applied, technological, and effective. It was also conceptually systematic, able to produce generality, and focused on socially important behavior.

Before insurance, most full programs ran through universities or grants. The best known is the UCLA Young Autism Project, led by Lovaas. In 1987, Lovaas published landmark findings. These children received 40 hours a week of intensive ABA. Results showed that 47% reached intellectual and educational outcomes like their neurotypical peers. Those data spread fast. They drove the rise of early intensive behavioral intervention (many hours of early ABA), or EIBI. The case for high hours rested on the catch up hypothesis (learning faster to close gaps).

Parents saw the gains and pushed for coverage. State autism mandates followed. By 2019, every state and the District of Columbia had some form of autism insurance mandate. The job changed with the money. Clinicians now had to argue in payer terms. They had to show less functional impairment (real trouble with daily tasks), clinical fit, cost, and social validity (does the change matter to families).

Growth also brought new players. ABA became a multi-billion-dollar part of health care within a few years. Between 2015 and 2024, private equity firms (outside firms that buy and combine companies) completed well over 100 acquisitions across hundreds of sites. Payers can trim a plan through utilization controls (limits payers set on services). They can also apply what Osos called medically unlikely unit limits, caps on billable units. The field built tools in response. CASP, the Council of Autism Service Providers, publishes practice guidelines. The Patient Outcome Planning Calculator, or POPC, links assessment data to deficits and goals, mostly for dosage (how many hours you request).

ABA was no longer grant-funded or only university-based services.

From the talk — Dr. Jessica Osos

Why Two Identical Clients Get Different Authorization Answers

Each insurer defines medical necessity on its own. So two clients with the same clinical picture can get different answers. The differences show up in several places. Which goal types count as allowable. How many hours get approved. How often you must reassess. What documents the plan requires.

That means your job is bigger than good clinical design. You also translate the plan into each payer's format. A plan that fits one insurer's rules can be cut or denied by the next. Osos described this as a constant balance between client need and payer rules.

Peer review (a payer clinician reviewing your plan) is where those gaps surface. Denials often come from paperwork gaps, not clinical disagreement. Osos pointed to work by Papatula and colleagues, which she placed around 2016. Common misses include three things. The assessment was not clearly tied to the goals. The reason for intensity was never spelled out. Sometimes the data were not graphed the way the payer expects.

Going in with thin paperwork costs you. You may face delays or service gaps. You may also get a read only determination (a denial the reviewer will not reconsider). That means the reviewer takes no new information, and the denial stands. Families notice those stops in care. Preventable denials shape how they see ABA.

many of these denials are coming from just missing links in the justification chain

From the talk — Dr. Jessica Osos

Why Payers Keep Coming Back to Autism Core Deficits

Most payers fall back on autism core deficits (the symptoms used to diagnose autism). So your plan needs one clear chain. Assessment findings point to named impairments. Impairments drive the goals. The goals explain why ABA is needed.

The DSM-5 (the manual used for diagnosis) names two core areas. First, lasting deficits in social communication and social interaction. Second, restricted and repetitive patterns of behavior, interests, or activities. Socially significant (behavior that matters in real life) is not enough by itself. The target also has to touch one of those two areas.

This is where plans often stall. Goals that look custodial or academic get flagged as not medically necessary. Toileting and academic targets are the usual examples. Osos was careful here. There is no blanket rule that those goals fail. It depends on how you tie each one back to a core deficit for that person.

She shared an example from a BCBA (board certified behavior analyst) group online. Someone asked how to write a goal for teaching sight words. Her advice was simple. Do not bury the goal inside a safety goal. Say what you are teaching. Then name the barrier. A client with high prompt dependency (leaning on adult cues to act) may need ABA teaching to learn the skill. That is the story a reviewer can approve.

We're losing trust in funders when we're trying to be sneaky and hide goals in other goals.

From the talk — Dr. Jessica Osos

The Three Ingredients of a Defensible Treatment Plan

Osos framed medical necessity as a recipe with three main ingredients. Tie the plan to autism core deficits. Pick an assessment that supports your argument. Then show that ABA works for this client.

ABA leans on two broad types of assessment. Criterion referenced tools (tests scored against a set skill list) pinpoint specific skill gaps. Norm referenced tools (tests that compare a person to same age peers) show functional impairment relative to peers. Common examples of criterion tools are the VB-MAPP and the ABLLS-R. The Vineland-3 (a standardized developmental scale) is the norm referenced tool most ABA providers use. The two work together. Criterion tools shape your goals. Norm tools show impairment next to peers. Payers often want to see both.

Decades of research show ABA improves cognitive, language, adaptive, and social outcomes. Early intensive programs at roughly 30 or more hours a week tend to show the strongest gains in young children. You do not have to prove that again. You have to prove benefit for this person.

Use more than one kind of proof. Show assessment scores moving up across time points. Show graphed session data, weekly trends, and rates of skill acquisition. Show the percent of goals mastered inside the authorization period (the span a payer approves). Show movement to harder skills. When progress stalls, name the barrier and what you did about it. Write down the protocol change, the added caregiver support, or the call to another provider. That record is part of your argument.

I like to think of medical necessity as a recipe.

From the talk — Dr. Jessica Osos

Framework Steps 1 to 3: Ask the Payer, Assess, Then Pick Targets

Step one is context. Before you write a goal, learn what this plan counts as medically necessary. Contact the insurer's utilization management team (staff who review service requests) or its ABA benefits department. Ask for the current medical necessity criteria, coverage guidelines, and documentation expectations. A mismatch up front is the most common reason plans get cut.

Step two is assessment. Ground the plan in a full assessment that names skill deficits tied to core deficits. Those results become the baseline for every goal. Osos recommends both assessment types when time allows. In her experience, norm referenced testing is often expected yearly rather than every authorization. Some funders name the exact tool. She has seen plans ask for the Vineland and others accept any standardized measure. Check first.

Step three is target selection. Turn diagnostic and assessment data into specific behavior change needs. Each target must link to a documented deficit. It must also be treatable with behavior analytic methods. The need should show up where the client lives, learns, and takes part. Lean on all three legs of evidence based practice (research, clinical skill, family values). Then check your targets against the plan's own guidelines. If a target falls outside them, make the case in writing.

we really should treat insurers as collaborators in that authorization process rather than our adversaries

From the talk — Dr. Jessica Osos

Step 4: Goals, Dosage, and the Exit Plan

Step four starts with social significance. Goals should matter to the client and the family, not just look interesting on paper. Family centered care gives you the method: shared decisions, real teamwork, and respect for culture. Caregiver involvement also strengthens generalization and maintenance (using skills elsewhere and keeping them), since caregivers run the natural setting. One caution matters. Current guidance is clear that access to medically necessary ABA cannot depend on caregiver participation. Work schedules, travel, and child care are real barriers.

Next comes scope and intensity. Comprehensive and focused ABA (how many skill areas you treat) sit on a continuum. Intensity refers to hours. There is dose response evidence (more hours linked to more gain), mostly in early intervention. Newer work shows some children gain at lower intensity, based on baseline skills, target choice, and how care is delivered. So guidelines now push individual dosage over fixed hour blocks. Recommend what fits this client's needs and capacity to learn.

Then write the goals. Each one should be measurable, observable, and criterion based, judged against a clear bar for mastery. Osos pointed to the ABA Coding Coalition (a group that sets ABA billing standards) for the required parts. Each goal needs an operational definition (a plain description anyone can use). It needs the client's baseline performance. It needs the measurement system, such as frequency or percentage. It needs mastery criteria (the bar that counts as learned). It needs a timeframe and the conditions under which the skill should occur. Finally, it needs a plan for generalization and maintenance. She added one habit from her own supervision. Pick the teaching protocol and the data method before the goal goes into the plan. Also match the number and complexity of your goals to the hours you request.

Step four closes with the exit. A transition is the point where a client can move to a less structured setting with less support. Set those performance criteria with the family at the start. Discharge planning (deciding when services should end) starts on day one too. Some clients meet their goals and hold gains as hours fade. Others stop benefiting, and your data and protocol changes support that call. Some families end services while ABA is still needed. Write down your recommendation and the criteria for restarting later.

Establishing these expectations early creates transparency and also helps ensure that fading or discontinuing service is data-driven rather than arbitrary.

From the talk — Dr. Jessica Osos

Steps 5 and 6: Running the Plan and Earning Reauthorization

Step five is about doing the work well. Use teaching and intervention methods with a research base, and skip ones without support. Then support the people running them. Technicians need materials, real familiarity with the plan, and protocols with clear prompting, error correction, and mastery criteria. That is what keeps treatment integrity (running plans the way they are written) high.

The second half of step five is data review. Your measures must be sensitive to the target behavior and produce reliable data. Look at graphs often and change course when progress lags. Osos noted that some research points to as few as 10 data points. That number only works if the data are collected reliably. Record every protocol change. Use session notes and phase change lines (marks showing when you changed something). Many payers now ask for graphs in the updated plan.

Step six is reauthorization (asking the payer to keep funding). Bring objective, individual proof. Show measurable gains on goals and better assessment scores. Show that procedures were run correctly. Show movement toward functional outcomes that matter. The ABA Coding Coalition goes further. It calls for graphed direct observation data, documented protocol changes, and generalization probes (checks that a skill shows up in new settings). It also wants a status line on every goal. Reviewers want to know if each goal was met, not met, or modified.

If progress stalled or skills slipped, say so. Name the barrier. Then describe your plan to fix it, or what you already changed. Osos framed the whole framework as one story a reviewer can follow from assessment to hours. It takes work. It also builds trust with funders and strengthens how the public sees ABA as health care.

reauthorization is not simply like copying and forwarding an existing treatment plan

From the talk — Dr. Jessica Osos