The Sections Payers Read First in an ABA Assessment Report
A payer reviewer reads an ABA assessment report in a fixed order: diagnosis and date, functional impairment data, goals with baselines, requested hours tied to those goals, supervision structure, caregiver involvement, discharge criteria. The review stops at the first gap.
A payer reviewer is not evaluating your clinical judgment. They are checking your report against a list, in a set order, and they stop when they find the gap.
Your clinical adherence can be flawless and the report can still fail. Denials on assessment reports are usually a production problem. The thinking was sound. The document did not make it findable in the place the reviewer looked.
The reviewer works from a list, and the list has an order
You wrote the report front to back. That is not how it gets read.
Reports get reviewed the way they were written: diagnosis, history, assessment results, goals, hours, plan. A utilization reviewer with a queue of cases does something different. They move through the report hunting for specific evidence, in roughly this sequence:
- Diagnosis and the date it was made, checked against the payer’s recency window
- Functional impairment, scored, not described
- Goals, each with a baseline and a numeric mastery criterion
- Requested hours, traced to the goals that justify them
- Supervision structure and who is delivering which service code
- Caregiver involvement, with its own goals and its own data
- Discharge and generalization criteria
Anything the reviewer cannot find in the expected place reads as absent. It rarely triggers a phone call. It triggers a request for more information, which costs the family weeks, or a denial, which costs you an appeal.
The stakes are not hypothetical. In 2024, marketplace insurers denied 19% of in-network claims, and administrative reasons accounted for 18% of stated denial causes against 6% for medical necessity.¹ That data covers ACA marketplace plans, not ABA specifically. It still tells you where claims die: in the paperwork, before the clinical argument is ever weighed.
Two of these items are requirements in the payers’ own published criteria, not just field habits. Optum’s audit preparation guide states that ABA services are delivered “under the supervision of a Board Certified Behavior Analyst (BCBA)” or an equivalently trained licensed clinician.² Blue Cross Blue Shield of Michigan’s ABA supplemental clinical criteria requires “measurable criteria for completing treatment with projected plan for continued care” before discharge.³
Five sections that fail a structural read
Not one of them is about clinical judgment.
Goals without baselines or numeric mastery criteria. A goal written as “will improve requesting across settings” gives the reviewer nothing to measure continued need against. Every goal needs a starting number and a stated threshold that ends it.
Requested hours with no arithmetic behind them. Reviewers look for the connection between the hour request and the open goals, and it is usually absent from the page even when it exists in the clinician’s head. Write the hours against named goals so the math is visible.
Functional impairment described instead of scored. Narrative descriptions of skill deficits read as clinical impression rather than evidence. Standardized scores with dates and instrument names do the work that prose cannot.
Discharge and generalization planning left out. Its absence reads to a reviewer as open-ended service. Naming what discharge looks like is what establishes that the current request is bounded.
The section written last, under time pressure. Reviewers know which one it is. It is usually caregiver involvement, and it is usually what turns a clean approval into a request for more information.
Read your last three reports in payer order
Fifteen minutes. No software involved.
Take your three most recent assessment reports and read each one in the reviewer’s sequence above instead of your own. Note every point where you had to hunt for something a reviewer expects in the first two pages.
Then check one thing specifically: for each report, can you draw a line from the requested hour total to the goals that produce it? The hours are almost always clinically defensible. What fails is that the defense is not on the page.
If your three reports come back clean, your template is doing its job and there is nothing here to change. If they do not, look at the template before you look at the clinicians.
The same report, written by different BCBAs
Your reports are probably not weak. They are uneven.
Run that exercise across your team instead of your own caseload and a second pattern shows up. Report quality tracks how many payer responses a BCBA has personally read. The BCBA who has watched dozens of reports come back has internalized what draws a request for more information. The one two years into practice has not had the chance.
That gap comes from exposure. Peer review is the field’s answer to that gap, and it holds until the person doing the review is also carrying a caseload. A structural standard that lives in the report format, rather than in the reviewer’s head, is what makes a newer BCBA’s report read like your strongest one’s.
What any adequate report system has to do
Audit your own stack before you audit ours.
Whatever you use to produce assessment reports, it should:
- hold baselines and mastery criteria in structured fields rather than free text
- flag any goal with no supporting data before the report can be exported
- keep hour requests, tied to the specific goals they fund
- hold requirements per payer, since what Aetna flags and what a state Medicaid plan flags are not the same list
If your stack covers the first two and handles the rest with a senior BCBA reading everything before it goes out, that holds until the caseload grows. Assessment Builder is one system that covers all four, pulling structured data out of the EMR you already run and assembling it into the format the payer expects.
Confirm all four checklist behaviors ship today, specifically the pre-export flag on goals with no supporting data and per-payer requirement handling. Per-payer customization is the capability prospects have reacted to on demos, which makes it the first thing tested. If either does not ship, cut the line rather than softening it.
The report is the only part of your clinical work a payer ever sees. It should represent the work.
