Health Economics & Reimbursement

Health economics and reimbursement is not a condition or a treatment; it is the question of money. Whatever a psychedelic therapy does for a patient, a health system has to decide whether it is worth paying for, how it compares with existing treatments per pound or dollar spent, what it would cost to provide at scale, and whether insurers or governments will cover it. This page is about that economic case, and about a crucial distinction that is easy to lose: the difference between what the models project and what has actually been measured.

The headline findings here are unusually optimistic. Several studies conclude that MDMA-assisted therapy for trauma is not just cost-effective but cost-saving, and that scaling psychedelic therapy could save health systems billions. Those are striking claims, and they are also, almost without exception, the outputs of decision-analytic models, structured spreadsheets that take assumptions about how well a therapy works, how long the benefit lasts, how cheaply it can be delivered and what it will be priced at, and calculate a cost per unit of health gained. The results are only as good as those assumptions, and the assumptions are where the real uncertainty lives.

The single most important idea to carry through this page is that a favourable cost-effectiveness number is a projection, not a fact. Change the time horizon, the durability assumption, the comparator or the assumed price, and the same therapy can flip from "dominant" to "not worth it". The economics are inseparable from the practical questions of how the therapy is actually delivered and how it reaches a population, and they should be read as careful arguments about what might be true under stated conditions, not as evidence that psychedelic therapy will, in reality, save money.

Because there is no condition here, the relevant "evidence" is the small set of economic models, and they share a revealing feature: their conclusions swing on the assumptions, not the drug. A US model of psilocybin for treatment-resistant depression illustrates this perfectly: it reported a reasonable cost per QALY, but its cost-effectiveness went from a 1% probability at an assumed $10,000 therapy cost to 95% at $3,000[1]Transl Psychiatry (2025), psilocybin-TRD US model: ICER $117,517/QALY at an assumed $5,000 therapy cost; 1% cost-effective at $10,000 drug cost, 95% at $3,000 — the price drives the result. A UK model reached a similar verdict from the other direction, finding psilocybin cost-effective only if therapist costs were halved and the price held to a few hundred pounds[2]Psychol Med (2023), psilocybin UK model: cost-effective only when therapist support was cut 50% and the price reduced to £400-800 per person. In both cases the economics are a function of the assumed price and labour cost, not an intrinsic property of the treatment.

The same assumption-dependence runs through the most optimistic literature. MDMA-assisted therapy for PTSD is repeatedly modelled as cost-saving over a thirty-year horizon[3]PLoS ONE (2022), MDMA-PTSD phase-3 model: per-patient $11,537; projected net savings of $132.9M per 1,000 patients over a 30-year horizon (a projection), but the same kind of model shows that if the benefit is assumed to last only one year, the figure becomes a far more ordinary cost per QALY[4]PLoS ONE (2020), MDMA-PTSD phase-2 model: net savings over 30 years, but if benefits are assumed to cease after one year the ICER rises to $26,427/QALY, and a more conservative five-year analysis lands at a respectable but not spectacular cost-effectiveness ratio[5]PLoS ONE (2024), MDMA-AT US model: base-case ICER $83,845/QALY over a 5-year horizon (more conservative than the 30-year cost-saving models). The honest reading of the "standard" evidence is that the favourable numbers are real model outputs under stated assumptions, and that the assumptions, especially about how long the benefit lasts, are doing most of the work.

The decisive variable for the field’s economic future is durability, and it is also the least settled. Every cost-saving conclusion depends on the benefit of a single course of therapy lasting years, the Ukraine MDMA model, for instance, explicitly assumes three years of sustained efficacy[1]World Med Health Policy (2025), MDMA Ukraine model: ICER $1,537/QALY assuming THREE YEARS of efficacy; scaled projections of billions saved (assumption-driven). If benefits fade faster, or if patients need repeat or maintenance dosing (as ketamine patients typically do), the arithmetic that makes these therapies look cheap collapses. The most valuable economic research the field could do is therefore not another model but real long-term follow-up: actual data on how long remission lasts and how often treatment must be repeated.

The other frontier is reimbursement, and here the gap between ketamine and the classic psychedelics is stark. Esketamine is actually covered and priced, which generates real-world economic evidence, including the uncomfortable finding that the branded product is dominated by far cheaper generic ketamine from a health-system perspective[2]J Affect Disord (2022), branded esketamine was DOMINATED by cheaper IV ketamine from a healthcare perspective; only competitive when insurance shields the patient. Classic psychedelics, by contrast, are reimbursed essentially nowhere, and face a structural hurdle: health-technology assessors typically demand active-comparator trials that the psychedelic field has barely run[3]Eur Neuropsychopharmacol (2026), HTA bodies (NICE, IQWiG, HAS) often require active-comparator data, a particular hurdle for psychedelic therapies where such trials are almost entirely lacking. The honest outlook is that the economic case will be made or broken not by more optimistic modelling but by long-term durability data and by clearing reimbursement bars the field has not yet reached.

The economics of psychedelic therapy are shaped by who builds the models and who pays the bills. A notable share of the cost-effectiveness literature is sponsor-adjacent: the most optimistic MDMA analyses come from within the advocacy and development ecosystem, and the esketamine cost-per-remitter studies are framed from an industry perspective[1]J Comp Eff Res (2025), industry-framed analysis: esketamine cost-per-remitter $3,102 lower (commercial) than quetiapine, remission 50% vs 33%. That does not make them wrong, but it means the headline framing tends toward favourability, and independent analyses are few. The real decision-makers, the insurers, governments and health-technology bodies, apply a colder lens, and the recurring lesson is that a therapy can look cost-effective to society while being unaffordable to the payer who actually writes the cheque, as the esketamine cost-utility work shows when it flips from acceptable to far-above-threshold depending on perspective[2]PharmacoEconomics (2023), esketamine-TRD Italy: ICUR €16,314-22,133/QALY from a societal perspective but over €100,000/QALY from the healthcare-system perspective.

For an honest broker, health economics is where optimism is most easily dressed in the authority of numbers, and the responsible posture is to read the numbers as scenarios rather than conclusions. The favourable ICERs are genuine outputs of careful models, and the cost-saving potential, if the therapies are durable and can be delivered cheaply, is real in principle. But every figure rests on assumptions about durability, delivery cost, price and perspective that can be tuned to almost any answer, the same ketamine analysis is "dominant" or wildly expensive depending only on whether control-arm costs are counted[3]J Affect Disord (2025), KADS trial-based ketamine economics: dominant if control-arm costs are included, but $251,250/QALY if excluded — comparator choice drives the result. The most useful framing keeps three facts in view: these are models, not realised outcomes; durability is the assumption that decides everything; and patients themselves say they would pay far less than the models and the market assume[4]J Psychedelic Studies (2024), willingness-to-pay survey (N=1,221): users considered about $70-80 per hour reasonable, well below current market projections, a gap that the question of who can actually afford and access these therapies will ultimately have to close.