My name is Sarah Coffey. I have been a BCBA for 15 years and currently serve as Vice President of Operations at Behavioral Innovations.
Behavioral Innovations provides center-based ABA to children with autism. We operate three centers in Virginia, in Chantilly, Fredericksburg and Manassas, with a fourth opening in Woodbridge in 2027. We also operate more than 120 centers across five other states. Most of the children we serve in Virginia are covered through Medicaid managed care, so we are invested in continuing to provide individualized services that produce meaningful outcomes and support children transitioning to less restrictive environments.
There is a great deal in this draft we support. Every ABA request being reviewed under EPSDT criteria before a denial is a stronger protection than exists in most states where we operate. Separating authorization requirements into initial, continued stay and EPSDT requests answers an existing problem. We also agree a provisional diagnosis is a legitimate basis for beginning treatment, given long wait times for a comprehensive evaluation.
There are some provisions in this draft that make a child's access or continued access to treatment, dependent on factors outside of what DMAS is trying to solve for. In some places it is documentation produced by a professional who is not treating the child; in others a caregiver's availability, a calendar interval, or a fixed threshold on service hours in advance of any individualized assessment. What these provisions have in common is access to care and the adequacy of the benefit, and where the benefit is delivered through managed care they also engage the Commonwealth's mental health parity obligations.
We are not asking the Department to remove a utilization management threshold, but rather to consider how DMAS reviews requests for services above 20 hours a week. Our comments go to how that review is defined, published and applied, because a standard not stated in advance cannot be met in advance.
This summary is limited by the comment field. Our full comment, twelve items and fifty-one recommendations with the supporting provision text, is being provided to the Department's Regulatory Coordinator and should be considered part of this submission.
The Four Items that Matter Most
1. The new weekly limit of 20 hours
The draft provides that services "may not exceed 20 hours per week, except as authorized under EPSDT review." The current Appendix D contains no limit on authorized hours; it refers to 20 hours only as the point above which a schedule of activities must be submitted. The effect is an ordinary authorization request above 20 hours carrying extra documentation becomes reachable only through EPSDT review.
We read the EPSDT clause as the Department intending 20 hours to operate as a utilization management threshold rather than a ceiling, and we think that intent is right. Our concern is the words. "May not exceed" is the language of a limitation, and the exception is written as an exception rather than the ordinary route for a child whose assessment supports more intensive care.
This reaches us principally through managed care. Each Virginia health plan builds its own criteria from the Department's and is bound to a definition of medically necessary services no more restrictive than the state's. A manual that states 20 hours as a limit gives every plan a limit to implement.
The code scope is unstated and it changes the number materially. The Department's 28 July 2026 bulletin describes a cumulative weekly limit. Read cumulatively, protocol modification and caregiver training consume part of the 20 hours and a child's ceiling for direct treatment is lower. Read as direct treatment only, the same child may receive supervision and caregiver training above it. A provider cannot build a treatment plan without knowing which applies.
The figure also sits below what CMS describes as appropriate for the children this benefit admits. CMS's August 2026 toolkit for state Medicaid agencies states there are no established metrics for what counts as a low or high intensity intervention and "no consensus on a set number of hours recommended for treatment." To formulate appropriate frequency and intensity, "providers consider the patient's level of ASD severity, which the DSM-5-TR defines using three levels from least (Level 1) to most (Level 3) severe." CMS illustrates: "Best practice is that the number of hours detailed in the ITP be aligned with these levels; for example, 10, 20, and 30 hours for children in levels 1, 2, and 3, respectively." Almost no child at level 1, and few at level 2, meet medical necessity for ABA. The children admitted to this benefit sit at the levels for which CMS illustrates 20 and 30 hours, and a single figure of 20 is set at the lower.
The same chapter reports that intensity is not correlated with improvement in outcomes, and we do not read the two passages as being in tension. CMS's point in both is that the number of hours belongs to the individual child rather than to a figure set in advance.
Texas operates the model CMS describes. Texas’ Medicaid manual states its frequency tiers "are guidelines," that "frequency and duration should be based upon the functional goals of treatment," and that its highest tier, "greater than 20 hours/week," may be delivered "without requiring additional submission of documentation or review by a physician for medical necessity."
Our own authorizations point the same way. Of 27 children in active treatment in Virginia today, roughly 80 percent are authorized at or above 20 hours a week of direct treatment under 97153 alone, before protocol modification or caregiver training. Those are hours the Department and its plans have already approved as medically necessary. We offer it as one provider's snapshot, and the Department can test the question against its own authorization data. If the figure operates as a limit, exception review becomes the ordinary route for this population, a volume of review neither the Department nor providers are set up to absorb.
We ask the Department to:
2. The provisional diagnosis pathway is newly limited to children five and under
The current criteria carry no age limit. The draft admits a child on a provisional diagnosis only where "the youth is age 5 or under." A child of six or seven with a provisional diagnosis, waiting for a comprehensive evaluation, can be admitted today and could not be under the draft.
The children affected are not a marginal group. Diagnosis after age five is common, particularly for girls and for children whose presentation is less overt. We understand the concern that a provisional diagnosis should not become an indefinite basis for treatment. The draft's own six and twelve month checkpoints already address it. The age limit does not add a safeguard. It closes a door that is open today.
We ask the Department to remove the age limit from admission criterion 2(b); to apply the same continued stay and EPSDT review mechanisms to any child admitted on a provisional diagnosis; and, if an age limit is retained, to extend the documentation route in item 2(c) to children above five and accept evidence of a scheduled evaluation or active waitlist placement.
3. The functional impairment criteria for admission
Admission requires a second showing beyond the diagnosis. Four of the manual's five criteria correspond to the diagnostic domains for autism. The fifth, frequent intense outbursts that are self-injurious or aggressive, is not a diagnostic criterion. It is an associated feature, present in some children and absent in many.
The criteria are also written to a severity the diagnosis does not require. The diagnostic criteria ask whether social deficits are persistent; the manual asks whether impairment is severe. The diagnostic criteria ask whether repetitive behaviors are present; the manual asks whether they are disruptive. A verbal child with a confirmed diagnosis, whose deficits are persistent but not severe and whose routines are rigid but not disruptive, can fail two of five. The Department accepts that child's diagnosis and declines to allow the child to access care.
The criteria also point a reviewer toward problem behavior as the marker of medical necessity, when the treatment that follows is largely skill acquisition.
We ask the Department to align the functional impairment criteria with diagnostic domains; to confirm that "severe" and "disruptive" describe the presentation rather than set a threshold above the diagnosis, or to remove those qualifiers; and to confirm a child who meets the diagnostic criteria at the lowest severity level is admissible where the assessment supports medical necessity.
4. The new diagnostic documentation requirement
The requirement itself is reasonable. The open questions are what happens to children already in treatment and whether a school-based evaluation counts.
We ask the Department to confirm that an authorization in effect on the effective date remains valid for its approved period and to state the first request at which the new requirement applies; to apply it prospectively, deferring it for children already in treatment to the first continued stay request at least six months after the effective date, or until changes from the Department's ABA workgroup are reflected in the manual; to add licensed school psychologists to the qualifying professionals; to communicate the requirement directly to evaluators through the Department's own bulletins, so the obligation does not reach them first through the ABA provider; to confirm the provider may supply the evaluator's Type 1 NPI; to accept a comprehensive evaluation completed by telehealth using a validated instrument; and to accept any validated autism screening instrument under item 2(b), including one completed by the child's primary care provider.
Comments on remaining items
5. Evidence obtained directly from the evaluating entity: The clause carries two readings. If it means the document must originate with the evaluator, say so and we will comply. If it means the evaluator must transmit it to us, a child's treatment becomes contingent on an office outside the treatment relationship that neither family nor provider can compel. We ask that the evidence originate with the evaluating entity but may be submitted by the provider, the family or the evaluator; that where the evaluator does not respond within a stated period, the family's documentation plus the provider's documented efforts suffice and services may not be denied on that basis alone; and that a documented waitlist placement carry forward on the family's confirmation that it remains active.
6. The twelve-month diagnostic checkpoint: Retain the twelve months measured from admission. Provide that EPSDT review may not result in denial or reduction where the only outstanding item is an appointment controlled by a third party and the child continues to meet medical necessity and show documented benefit, and that a documented active waitlist placement satisfies the requirement while it remains active. Clarify that the comprehensive evaluation in continued stay item 1 is submitted once completed and that until then items 2 and 3 govern; as drafted it carries no timing and could be read to apply ahead of the checkpoints the draft sets for the same children. Carry the age distinction the draft makes at initial authorization into the continued stay criteria, or state the basis for a single interval.
7. Discharge where a caregiver is unable to participate: A caregiver's work schedule is not a clinical finding, and the draft treats it as one by placing caregiver participation among discharge criteria that are otherwise clinical determinations. This is the provision most likely to end treatment for a child who still needs it, and it will fall hardest on families with inflexible hours, single-caregiver households and families without reliable transportation. We ask that caregiver participation be removed from the discharge criteria and addressed through the ISP; that providers instead establish an individualized participation plan, document efforts and barriers, and modify training methods, format or scheduling where clinically appropriate; that services not be terminated or reduced solely for barriers to participation where the provider has made and documented reasonable efforts and the child continues to benefit; or in the alternative that the criterion be limited to a caregiver who declines to participate after documented efforts, removing "or is unable." We also ask what direct family involvement "at a minimum of weekly" requires, including whether it is satisfied by documented contact or a specified duration, whether it is separately billable, and whether telehealth satisfies it. A requirement carrying a discharge consequence should be stated precisely enough to be met.
8. Discharge after a 31-day lapse: Allow an exception where the provider documents the cause of the interruption and the clinical basis for resuming the existing episode of care, permit resumption on an addendum rather than a new assessment and plan, and confirm that an absence documented in advance in the ISP does not start the period.
9. The ten-day transition period: Allow up to 30 days where clinically necessary and transition efforts are documented, and allow extension where the receiving provider's authorization is pending, so a child is not discharged into a gap created by a second review.
10. Schedules required above 20 hours: Confirm the schedule describes the structure of the treatment day and the therapeutic purpose of each portion, and does not fix the specific intervention delivered in each timeframe, which remains clinical judgment during the session. State the requirement to distinguish therapeutic from recreational and non-therapeutic time as a general one applying to every setting in which ABA is delivered, or remove it; as drafted it reaches clinic-based services alone and the draft gives no reason for the difference. Confirm that play, meals, toileting and transitions count as therapeutic time where they target plan goals and progress is measured. For young children a substantial part of teaching happens in those moments, and a standard reading them as non-therapeutic would misdescribe the intervention.
11. Care coordination with the primary care provider: As drafted the requirement describes an outcome rather than an effort, so a record can fall short for a reason the provider does not control. We ask that providers make and document reasonable efforts to coordinate with the primary care provider and other treating professionals, consistent with consent and privacy requirements, and that a family's decision not to authorize an exchange of information be a documented barrier rather than a provider deficiency.
12. Supervision and direction are used interchangeably: The manual requires observation and supervision monthly and clinical supervision twice monthly, and uses the two terms in overlapping ways. They are different activities. Clinical direction means working alongside the technician and adjusting the protocol; supervision required by the health regulatory board is a separate obligation; the administrative supervision tied to the RBT credential is a third. We ask that they be distinguished and each described consistently with the scope of practice of the professional performing it.
Implementation
We ask that DMAS issue implementation guidance to contracted plans covering the effective date, authorizations already in place, the 20-hour threshold and the EPSDT pathway, and each new documentation requirement. State a single effective date for this appendix and confirm how it relates to any corresponding state plan amendment. Confirm, for the manual as a whole, that a plan's criteria and documentation requirements may be no more restrictive than the Department's. Allow a defined period between publication of the final manual and enforcement.
Recommended Corrections
The CPT code set should be updated. The manual references Category III codes 0362T and 0373T, which the AMA deletes effective 1 January 2027 and replaces with 97148 and add-on 97149, and 97159 and add-on 97160.
The diagnostic instrument listed as "Autism Diagnostic Review (ADI-R)" should read Autism Diagnostic Interview-Revised, in both locations where it appears.
The age boundary for the provisional diagnosis pathway is stated four ways across admission criterion 2(b), item 2(c) of the Initial Service Authorization Request, and the Additional Continued Stay Criteria. We ask that one formulation be used throughout, and that the Department confirm whether a child who has turned five but not six is included.
Summary
As highlighted above, four items matter most, and all four determine whether a child receives the service at all rather than how it is documented. The weekly limit should be written as the threshold the Department appears to intend, because every plan will build its criteria from whichever word the manual uses. The age limit closes to children over five a route open to them today. The functional impairment criteria ask for a second showing beyond the diagnosis the Department has already accepted. Lastly, the caregiver criterion makes a family's schedule a basis for discharge.
Each can be addressed without weakening the Department's utilization management, and our full comment proposes language that does so. Thank you for the opportunity to comment, and we welcome the chance to discuss our comments with the Department!