Interventional psychiatry has become the standard value creation lever in outpatient behavioral health: an existing patient panel, a defined protocol, and contribution margin per square foot that talk therapy cannot match. The demand case is usually right. The reimbursement case is usually wrong in four specific ways.

TMS and Spravato/esketamine programs are attractive for a reason. They monetize patients a platform already treats, they carry a documented clinical rationale for treatment-resistant depression, and at steady-state utilization the unit economics are strong. The problem is that the models built around them describe steady state and skip the two years it takes to get there.

1. Authorization dependence, not contracted rate

Diligence tends to focus on the contracted rate per session because that is the number in the fee schedule. The rate is rarely the binding constraint. Almost all of this revenue sits behind prior authorization, and a meaningful share of course completions require re-authorization partway through.

Three numbers matter more than the rate: the first-pass authorization approval rate, the average days from referral to first treatment, and the share of approved courses that are completed. A program with an excellent contracted rate, a 60% first-pass approval rate, and a 21-day authorization lag has a revenue profile closer to a program with mediocre rates and clean authorization. Those two look identical in a rate-based model and nothing alike in cash.

Authorization performance is also the most fragile thing in the operation, because it usually depends on one or two people who know each payer's documentation preferences. That is a personnel risk masquerading as a payer risk.

2. Medical policy variance by payer and state

Coverage for interventional psychiatry is governed by medical policy, and medical policy is set by payer, plan type, and state. Criteria for the number and class of prior antidepressant trials, the treatment-resistance definition, the allowed course length, the coverage of maintenance or re-treatment, and the handling of accelerated protocols all vary. A platform that performs well in one state can be repriced in another without changing anything about how it operates.

A roll-up thesis compounds this. Acquiring five practices across four states means inheriting four distinct policy environments, and the model that annualizes the best-performing site across the whole footprint is overstating the opportunity. The same applies to Medicaid managed care participation, where coverage of these service lines is inconsistent and where a favorable state posture is a policy decision rather than a contract.

3. Utilization ramp, staffing floor, and the gap between them

Device and chair economics are driven by utilization, and utilization ramps slowly. Patients present over a course of weeks with daily or near-daily sessions, which means schedule density, no-show behavior, and dropout rates determine realized revenue far more than the number of devices installed.

Staffing works in the opposite direction. Monitoring requirements for esketamine set a floor on clinical coverage per treatment session regardless of how many patients are on the schedule, and prescriber availability gates both intake and ongoing management. The result is a cost base that is close to fixed against a revenue line that ramps over quarters, and an economic profile that is negative for longer than most models assume.

The diligence questions are concrete: what is the actual ramp curve of the most recently opened site, what share of started courses are completed, what is the no-show rate against a daily protocol, and what does site-level contribution margin look like in month six versus month eighteen.

4. Referral dependence inside the panel

The stated advantage of adding interventional psychiatry to an existing platform is internal referral: the panel is already there. In practice, internal referral concentrates in a small number of prescribers who are comfortable with the modality and consistent about offering it.

If two prescribers generate most of the interventional referrals, the service line has the same single-point dependency as any concentrated revenue base, and it is exposed to exactly the provider attrition risk that follows a change of control. Referral-source concentration inside the platform should be measured with the same discipline applied to external referral concentration.

The demand case for interventional psychiatry is usually sound. The sequencing case is where the money is lost.

What a defensible underwrite looks like

  • Model by state and payer, not blended. Build the revenue line from medical policy and authorization performance in each state the platform operates in.
  • Underwrite realized revenue per started course rather than rate per session, incorporating completion and dropout behavior.
  • Hold staffing at its protocol floor through the ramp, and fund the resulting negative contribution period explicitly.
  • Test authorization operations as a dependency. Who does it, what documentation standard they apply, and what happens to approval rates if they leave.
  • Sequence de novo openings against credentialing and enrollment timelines, which govern when a site can bill at all.
  • Stress-test one adverse policy change in the largest payer in the largest state, and look at what it does to the equity case.

None of this argues against interventional psychiatry as a value creation strategy. It argues for underwriting it as a reimbursement and operations problem, which is what it is, rather than as a demand problem, which it is not.

About the authorEdisa Shirley, PhD, LMHC, is Founder and Managing Partner of Aluria Advisory, which advises private equity firms, family offices, and behavioral health operators on diligence, integration, and platform growth. Start a confidential conversation.