Juan Pablo Montoya is Founder and CEO of Solum Health, building AI automation for healthcare operations.

getty

The same question has landed in my inbox a dozen times this month, and almost none of it comes from practices that run in Indiana; they run applied behavior analysis (ABA) practices in other states, and they have been watching the news out of Indianapolis. They all want to know the same thing: does this reach us?

The answer is that it already has. It just has not shown up on their balance sheet yet.​

On April 1, 2026, Indiana's Health Coverage Programs put Bulletin BT202627 into effect and reset the economics of ABA in a single move. It cut individual service rates 6%, with another 4% due in April 2027. It capped comprehensive therapy at 4,000 lifetime hours, and it set a floor of one hour of BCBA supervision per eight technician hours. This pulled the synchronous-telehealth modifier from five core assessment and treatment codes.​

I do not read that as an Indiana story; I read it as a template. State Medicaid agencies watch one another, and when one finds a bundle of levers that bends the spend curve without an obvious access scandal, the others borrow the design. Indiana has already gone further, freezing new ABA agency enrollments and ownership transfers under a federally approved moratorium. But the real lesson is not regulatory. Each of these levers is a test of whether your systems can produce a number on demand, and most cannot.​

So here is what I would pressure-test inside any multi-state ABA organization this quarter. Each item is a technology problem wearing an operations costume.​

Run the supervision math before a regulator runs it for you.​

A 1:8 supervision ratio reads like a clinical-quality rule. It functions as a billing rule: every eight hours of technician time must sit inside a documented, authorized hour of BCBA oversight, or the units underneath are exposed in an audit. Most multi-location practices cannot tell me whether their billed units fall inside a valid supervision span, because the supervision calendar and the billing system live in different places and nobody reconciles them until a denial forces the question.​

The fix is integration, not headcount. Until those two systems write to one record, no rule can flag out-of-ratio units before the claim goes out, and the 1:8 floor stays a manual cleanup instead of a real-time control. Model what that floor does to your BCBA coverage per site.​

Treat a lifetime hour cap as a tracking problem, not a clinical one.​

A 4,000-hour lifetime cap quietly changes what an authorization is. It stops being a six-month window you renew and becomes a running balance against a ceiling that follows the patient for life, across every provider they have seen. The danger is delivering care that is clinically sound against hours a previous provider already burned.​

This is a data-model problem before it is a clinical one. Systems built around six-month episodes cannot hold a lifetime balance that travels with the patient. What's needed is a longitudinal record that counts used hours as a live number, reconciled against the payer's count, with alerts as the ceiling nears. Ask your team a blunt question this week: can our systems even produce that number today?​

Re-engineer your telehealth-dependent workflows before the codes disappear.​

Indiana pulled the synchronous-telehealth modifier from five codes: 97151, 97152, 97153, 97154 and 0373T. If any piece of your assessment or supervision flow quietly depends on delivering those services remotely, it is one bulletin away from breaking. Operators treat this as a scheduling nuisance when it is really a capacity constraint: in-person assessment slots are scarcer, and in Indiana, treatment cannot be authorized until a completed assessment is on file.​

Workflows hard-wired to a single billing modifier are a design liability. Model service delivery so that codes and modifiers are configuration, not foundation, and a rule change becomes a settings update your systems absorb rather than a rebuild.​

Build accreditation readiness and audit readiness off one dataset.​

Indiana is also steering ABA providers toward third-party accreditation to keep their Medicaid eligibility, joining Massachusetts, which already requires it outright.

Most operators handle accreditation, audit defense and private-equity diligence as three separate fire drills.​ They are one dataset. Build a single intake-to-claim record where every delivered session links its assessment, authorization, supervision and documentation, and accreditation, an audit lookback and a buyer's data room become three queries against one source instead of three reconstructions. That is an enterprise data-architecture decision, and the real technology story inside this overhaul. Build it once, not three times under deadline.​

Model the rate cut as a margin redesign, not a line-item trim.​

A 6% cut now and a 4% cut in 2027 do not subtract politely from the bottom line. They compound against the administrative leakage most ABA practices already tolerate. When the rate drops, every preventable denial, every lapsed authorization and every hour a BCBA spends on paperwork costs proportionally more than it did the week before.​

You cannot recover margin you cannot see leaking. Denial rate by payer, authorization-lapse rate and BCBA administrative hours belong on a live dashboard, instrumented as operational data, not assembled by hand each quarter. That visibility is what turns "wait and hope volume covers it" into a margin you actively manage, and the recovered revenue absorbs the cut.​

The Bottom Line​

Indiana did not invent a new way to squeeze ABA; it gathered levers other states have been testing one at a time and pulled them together, with effective dates attached. What it really exposes is infrastructure: whether your systems can reconcile supervision to billing, track a lifetime balance, survive a code change and produce one clean dataset on demand.

This gives multi-state operators a narrow window to build that spine while it is still a planning exercise, not the week a bulletin lands in their state. Studying Indiana can help buy you the one thing the next state's rule never will, which is time.​​​

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?