Competency intelligence
Training–Evidence Gap Index
The only quantified measure of the gap between what psychedelic-assisted therapy trials require of practitioners and what today's training actually teaches. Every number is computed live from the 663 competencies mapped on Blossom.
As of 8 July 2026.
Looking for something specific? Ask Blossom
550
Guideline-only (the gap)
Competencies evidenced in clinical-trial guidance but not yet taught on any indexed course page.
49
Matched
Competencies evidenced by both trial guidance and practitioner training pages.
60
Course-only
Competencies taught on course pages with no current clinical-guideline match.
Gap by protocol family
For each drug protocol family, the share of trial-evidenced competencies that training has not yet caught up on. "Evidenced" counts competencies backed by trial guidance; "taught" counts those on a course page; "gap %" is the evidenced set that remains untaught.
| Protocol family | Evidenced | Taught | Matched | Gap % |
|---|---|---|---|---|
| Psilocybin | 187 | 64 | 40 | 79% |
| MDMA | 184 | 46 | 35 | 81% |
| LSD | 115 | 25 | 24 | 79% |
| Ketamine | 172 | 48 | 34 | 80% |
| Ibogaine | 141 | 23 | 22 | 84% |
| DMT / Ayahuasca | 158 | 25 | 24 | 85% |
| 5-MeO-DMT | 138 | 22 | 21 | 85% |
| Mescaline | 36 | 12 | 11 | 69% |
Gap by competency category
The same measure across the twelve competency categories. Categories with a high gap are where trial protocols lean hardest on skills that training pages rarely name.
| Competency category | Evidenced | Taught | Matched | Gap % |
|---|---|---|---|---|
| Safety and Basic Knowledge | 101 | 27 | 16 | 84% |
| Pharmacology and Drug Interactions | 56 | 7 | 3 | 95% |
| Setting Management | 84 | 20 | 12 | 86% |
| Compliance and Risk Management | 97 | 19 | 4 | 96% |
| Ethics and Boundaries | 81 | 31 | 11 | 86% |
| Cultural Humility | 19 | 7 | 1 | 95% |
| Screening and Patient Assessment | 123 | 10 | 5 | 96% |
| Informed Consent | 43 | 6 | 5 | 88% |
| Psychological Support and Psychotherapy | 213 | 51 | 27 | 87% |
| Medical Monitoring and Crisis Intervention | 126 | 13 | 11 | 91% |
| Care Coordination | 54 | 19 | 3 | 94% |
| Data, Documentation, and Outcomes Measurement | 82 | 7 | 1 | 99% |
Where the gap is widest
The guideline-only competencies with the most trial-guidance backing. These are the skills protocols ask for most often and training pages address least.
Informed consent, decisional capacity, autonomy, and withdrawal rights
59 guidelines · 0 coursesConfidentiality, privacy, and research data protection
53 guidelines · 0 coursesAdverse event identification, documentation, and reporting
46 guidelines · 0 coursesManual fidelity, protocol adherence, and deviation management
39 guidelines · 0 coursesPost-dose follow-up, safety monitoring, and retention support
37 guidelines · 0 coursesDischarge readiness, escort safety, and post-session supervision
30 guidelines · 0 coursesBlinding, allocation concealment, and unblinding control
29 guidelines · 0 coursesMedication, substance-use, washout, and taper management
29 guidelines · 0 coursesStudy documentation, data integrity, and regulatory recordkeeping
27 guidelines · 0 coursesCo-therapist and multidisciplinary team coordination
24 guidelines · 0 courses
The full dataset
Every mapped competency with its categories, protocol families, source counts, and gap bucket, computed live.
Download the full list (CSV)How this is measured
A weekly pipeline reads clinical-trial guidance documents and practitioner course pages, then uses a large language model to extract the discrete competencies each one implies. Extractions are deduplicated against the existing map with a guarded matcher, so near-identical skills collapse onto one competency rather than inflating the count.
Each competency records how many trial-guidance sources and how many course pages evidence it. A competency is "evidenced" when at least one trial-linked guideline names it, and "taught" when at least one indexed course page does. The gap is the evidenced set that training has not yet caught up on.
Two honest limitations. Course extraction covers a subset of the courses we know exist, so the course-side counts understate what is taught in the wider field, which makes the measured gap a ceiling rather than a settled figure. And no human expert has reviewed the extraction yet, so treat the mapping as a structured starting point, not a verdict.