An estimate of the number of people with clinical depression eligible for psilocybin-assisted therapy in the United States

Psilocybin-assisted therapy (PSIL-AT) has been designated by the FDA as a breakthrough therapy for major depressive disorder (MDD) and treatment-resistant depression (TRD). TRD is defined in the paper as failure to have significant symptom relief after at least two adequate antidepressant trials during the current episode. With potential FDA approval and medical legalisation under consideration, understanding the prospective public health impact of PSIL-AT in the United States requires an estimate of demand among people with clinical depression. This study sets out to produce conservative lower-bound, mid-range and upper-bound estimates of the number of patients with MDD and TRD who would be clinically eligible for PSIL-AT. The authors frame these bounds by applying exclusion criteria drawn from the largest PSIL-AT trials, then relaxing or adjusting those criteria to reflect likely real-world practice and to account for comorbidity-driven double counting. The intent is to provide policymakers, payers and providers with a range of plausible demand scenarios to inform resource planning and regulation.

The analysis begins with national estimates of MDD prevalence based on the 2021 National Survey on Drug Use and Health and then restricts the population to those who reported receiving treatment in the past year, reasoning that PSIL-AT uptake is most likely among people actively engaged in care. Patients meeting an operational definition of TRD were identified as a subgroup of those treated for MDD. Three eligibility scenarios were constructed. Scenario 1 (lower-bound) applies clinical trial exclusion criteria from the largest PSIL-AT trials to produce a conservative estimate of who would qualify. Scenario 2 (mid-range) removes exclusion criteria judged unlikely to be enforced in routine clinical practice, drawing on pharmacological mechanisms and clinical judgement to decide which conditions remain exclusionary. Scenario 3 (upper-bound) retains the mid-range exclusions but further adjusts for co-occurrence of disqualifying conditions to avoid double counting individuals with multiple comorbidities. The study team excluded from consideration transient conditions (for example pregnancy), conditions modifiable by a provider (for example tapering an SSRI), rare or poorly documented exclusions, study-design constructs (for example prior psychedelic use), and experimental interventions. Prevalence inputs came from published sources where available; when prevalence among people with depression was not reported, general population estimates were used. The investigators explicitly considered neurobiological reasons for exclusions — potential destabilisation of mania or psychosis, seizure risk, serotonergic cardiovascular effects, and severe hepatic dysfunction — when forming the real-world exclusion set. To characterise uncertainty, the authors performed a probabilistic sensitivity analysis. They assigned beta distributions to comorbidity prevalence parameters with ranges of ±50% around baseline values and ran simulations using @RISK software (Palisade Corporation, version 8.1.1) to generate 95% confidence intervals for the final eligible population estimates.

The baseline burden estimated in the paper is 14.8 million people with MDD in the United States, of whom 9.0 million were receiving treatment and 2.7 million met the study's criteria for TRD. Applying the three eligibility scenarios to these treated populations produced markedly different pools of potentially eligible patients. Under the lower-bound (trial exclusion) scenario, 24% of treated patients with depression would be eligible for PSIL-AT. This equates to approximately 2.2 million people among those treated for MDD and about 0.6 million when limited to TRD. The mid-range (real-world) scenario increases the eligible proportion to 56%, corresponding to roughly 5.1 million treated MDD patients and 1.5 million with TRD. The increase from lower-bound to mid-range is largely driven by removing alcohol and substance use disorders as exclusion criteria; the authors attribute 32% of the difference between these scenarios to that change. The upper-bound estimate adjusts the mid-range figures to account for comorbidity overlap and thereby avoid double counting, yielding 62% eligibility overall. In absolute terms this represents about 5.6 million treated patients with MDD and 1.7 million with TRD. The adjusted rise is chiefly due to accounting for co-occurrence among cardiovascular and psychiatric comorbidities, each contributing an estimated 3%–4% increase in eligible patients after correction for overlap. The probabilistic sensitivity analysis produced 95% confidence intervals of 4.7 million to 6.6 million eligible individuals with treated MDD and 1.4 million to 1.9 million eligible individuals with TRD, reflecting uncertainty in the comorbidity prevalence inputs.

Rab and colleagues present a range of plausible estimates for the number of people with treated MDD or TRD who might be clinically eligible for PSIL-AT in the United States, and they position these figures within broader health-system and policy contexts. The authors note that prescribing practice often extends beyond trial-eligible populations — a cited retrospective analysis found that most people currently prescribed antidepressants would not meet typical trial eligibility — implying that real-world use of PSIL-AT could diverge from the trial-based lower-bound estimate if off-label or expanded indications occur. At the same time, several practical constraints could limit effective demand below the study's projections. Key limiting factors highlighted include the availability of trained providers, geographic access (with rural areas at greater risk of limited services), state-level regulatory variation, and patient preferences related to cost, treatment duration and cultural acceptability. The authors discuss how states that have already moved to legalise psychedelic therapies, such as Oregon and Colorado, may see disproportionate early uptake, and they point out that state rules (for example Colorado's facilitator restrictions) could either reduce or allow exceptions for patients with comorbidities. Insurance coverage decisions, especially by Medicaid, are emphasised as critical determinants of realised demand. Medicaid covered 85 million beneficiaries in 2023 and is estimated to have 18%–20% prevalence of clinical depression among its population; thus, whether and how Medicaid reimburses PSIL-AT will strongly influence access. The researchers also note that if FDA approval were limited to monotherapy, uptake could be constrained because about 70% of people with MDD use antidepressants and many find tapering difficult. The authors acknowledge important limitations. They state there are no reliable data to predict future treatment-seeking behaviour, which complicates demand projections; their approach intentionally focuses on currently treated patients to keep estimates conservative. They also recognise uncertainty introduced by variation in comorbidity prevalence estimates and by potential future approvals of other psychedelics that could redistribute demand. Finally, the paper calls for additional research to refine projections: studies examining regional and demographic variation, the impact of state regulations, cultural attitudes, and longitudinal real-world monitoring of PSIL-AT implementation would help align early estimates with eventual uptake and inform policy and resource allocation decisions.