Utah Doesn't Pay for ABA in Schools, the Drive There, or the Charting After
The Utah Medicaid ASD manual, updated January 2026, bars fee-for-service ABA in school-based settings apart from a care plan meeting, and lists travel time and post-session charting among non-covered services. Its discontinuation test turns on protocol modifications rather than a calendar.
Florida's ABA Task Force opened its first meeting by asking how many children receive 30 or more hours of ABA a week in school settings. Utah has already answered that question for itself, in a manual updated this January, and the answer is that it does not pay for it.
No fee-for-service ABA in schools
Section 8-4.1 of the Utah Medicaid Provider Manual for Autism Spectrum Disorder Services is one sentence long where it matters:
“Apart from the psychologist's or behavior analyst's participation in the evaluation of a child's school-based care plan, the Medicaid agency shall not reimburse fee-for-service ABA services in school-based settings.”
ASD-related services required under IDEA may still be delivered in schools, but they run through a different door: the Medicaid School-Based Skills Development Services benefit, with its own provider manual. The scope note matters too — school-based settings “also include charter schools but not privately funded schools.”
One carve-out survives. If the psychologist or behavior analyst, working with the family and school staff, judges it medically necessary to take part in the child's care plan development meeting, that specific participation may be billed fee-for-service. The meeting, not the treatment.
The paperwork and the drive are not billable either
Section 9-1 lists what Utah will not pay for, and two entries are the kind that quietly account for a chunk of a schedule:
- Charting and data collection done separately from the session. Non-covered is “time spent by the behavior analyst, behavior analyst in training, or technician charting or collecting data that is occurring separate from the time spent documenting direct observations that occur when the provider is working directly with the member.” Documentation made while with the member is inside the service. Writing it up afterwards is not.
- Travel. “Provider's time traveling and other associated costs to get to the member.” For a home-based caseload spread across a county, that is a real number.
This is the same line the HHS Office of Inspector General has been drawing in its state audits, where failure to distinguish billable from non-billable time appears in report after report. Utah has written the distinction into the manual rather than leaving it to be discovered in a post-payment review.
A plateau test, not a clock
Most payers now define when ABA stops by counting time — six months at Health Net, two successive six-month authorization periods at CareSource in Georgia. Utah defines it by what the clinician did next. Non-covered:
“ABA services rendered when measurable functional improvement is not expected or progress has plateaued with no further protocol modifications made to address deficits.”
Read the second half. A plateau alone is not the trigger — a plateau with no further protocol modifications is. The manual is describing a clinical record in which someone noticed and changed something. That is a different test from a calendar, and a different documentation burden: it asks what you modified and when, not how long the child has been in service.
The rest of the non-covered list
- ABA “not primarily provided to ameliorate a behavioral or maladaptive behavior.”
- “Any service that does not follow the established treatment plan.”
- Custodial care — assistance with bathing, dressing, eating, hygiene and safety.
- Respite care, defined as care given to relieve or cover for the usual caregiver.
- Setting costs, with the manual naming resorts, spas, therapeutic programs and camps as examples.
Requests above the quantity limits go to secondary medical review or a medical review committee, and the qualified health professional must file justification with each prior authorization request. Non-covered services can be sought case by case through the Medicaid exception process.
The limits
This is a standing manual, not a change: it carries an “Updated January 2026” stamp, and we have not established which provisions are new in that revision as against the prior edition. It governs fee-for-service Utah Medicaid; members enrolled with a health plan follow that plan's rules, which may differ. The school-based exclusion redirects services rather than eliminating them — the School-Based Skills Development Services benefit is a separate manual we have not read here, and its terms decide what a school-based program can actually bill.
What you must know or do
- Utah providers billing fee-for-service ABA in a school: stop and re-route. The manual bars it outright, with one exception for the analyst's participation in a care plan development meeting. Read the School-Based Skills Development Services manual before your next authorization, not after a denial.
- Check whether any of your school placements are privately funded. Charter schools count as school-based settings. Private schools do not, which changes which rule applies.
- Find out what your team charts after the session ends. If notes and data entry happen away from the member, that time is non-covered in Utah. This is worth measuring rather than estimating — it is also the exact defect OIG has cited in four state audits.
- Take travel out of your billable model. Time and associated costs to reach the member are non-covered. A home-based caseload built on the assumption that drive time is recoverable needs re-pricing.
- If a client has plateaued, the record must show what you changed. Utah's discontinuation test turns on the absence of protocol modifications, so a note documenting a plateau without a modification is the finding, not the defense.
- Multi-state operators: this is now a fourth distinct discontinuation standard — six months at Health Net, twelve at CareSource, a plateau-plus-no-modification test in Utah, and documented attributable progress at Fidelis. They are not interchangeable, and a single template satisfies none of them well.