Mapping rule
An external concept is called an exact map only when its meaning, scope, context, and intended use match the proposed concept. Broader, narrower, related, and safety-overlay concepts are retained as relationships, not substitutes.
The coding landscape, or why the ‘aha moment’ has no code.
To see the gap clearly, it helps to know the maps we already have. Here are the main code sets, what each is good for, and where the treatment experience falls through them.
Scope and method
What this covers, and what it does not.
The coding landscape is the set of standards used to turn one episode of interventional mental health care into a shareable record, from the molecule to the meaning. Most layers are codable today. The subjective treatment experience is the one layer with no broadly adopted code, and that gap is the subject of this page.
It covers ketamine and esketamine, transcranial magnetic stimulation, electroconvulsive therapy, and psychedelic-assisted therapy. Each layer is mapped to the standard that governs it and marked usable or not.
It does not create or modify any external standard. External identifiers remain governed by their respective standards-development organizations.
The maps we already have
The main code sets, in plain English.
- ICD-11. The international system for classifying diseases and diagnoses. Built to answer "what condition does this patient have," not "what did the patient experience in a session."
- SNOMED CT. A large, detailed clinical vocabulary for conditions, findings, and procedures. Rich, but organized around clinical findings, so a healing breakthrough has no natural home in it.
- MedDRA. The dictionary used to code adverse events and safety data, common in regulated research. It tends to read an unusual experience as a symptom to be flagged, which is exactly the wrong lens for a therapeutic breakthrough.
- RxNorm. The standard naming system for medications. This one works well for the drug side of interventional care.
- LOINC. The standard for lab results and clinical measurements, including many questionnaires. Useful for the measures a clinic tracks over time.
The stack
Each layer has a different steward and a different job.
Most of the stack is usable today. One layer is missing.
| Layer | Standard | Steward / jurisdiction | What it identifies | Maturity |
|---|---|---|---|---|
| Drug or molecule | RxNorm, with SNOMED CT | NLM (US) | Drug ingredient, clinical drug, and branded drug identity | Usable |
| Medical device | UDI / GUDID, with SNOMED CT | FDA (US) | Device identity and device classes | Usable |
| Units | UCUM | Regenstrief | Units of measure | Usable |
| Diagnosis (US) | ICD-10-CM, with SNOMED CT | NCHS and CMS (US) | The condition treated, for US records and billing | Mature |
| Diagnosis (international) | ICD-11 | WHO | The condition treated, as a global classification | Mature |
| Procedure reporting | CPT Category III | AMA (US) | Emerging-service reporting, not coverage or payment | Emerging |
| Instruments and observations | LOINC, with SNOMED CT | Regenstrief; SNOMED International | Questionnaire panels, items, calculated scores, and observations | Mixed |
| Safety | MedDRA, with CTCAE for severity | ICH; MedDRA MSSO | Adverse events and graded severity | Mature |
| Exchange | HL7 FHIR and implementation guides | HL7 International and profile sponsors | Exchange format and profiles | Emerging |
| US behavioral-health data | USCDI+ Behavioral Health | ASTP/ONC and SAMHSA | US behavioral-health data elements | Emerging |
| Experience | No broadly adopted, non-pathologizing terminology identified | Proposed below | The subjective treatment arc | Gap |
Jurisdiction matters. The United States uses ICD-10-CM for diagnoses; ICD-11 is the WHO global classification. A CPT Category III code set reports psychedelic-monitoring services, covering first physician, second physician, and clinical-staff monitoring in turn. Reporting a service is not the same as establishing coverage or payment for it.
Explore the landscape
The same landscape, interactive.
Filter, search, or isolate the gap. Select any marker to see its detail. The tables above stay the plain-text reference.
Uncoded Adjacent, not equivalent Instrument-anchored Officially coded size shows priority
Coding
Class
Tier
Showing 26 of 26, 18 in the gap
Coding
Priority
Anchor
Nearest code
Safety rule
List view: every concept and its coding status
Positions are illustrative, not measured percentages. Class, coding status, priority, anchor, and safety rule are real values from the review. No patient data, internal identifiers, or code numbers appear here.
The principle
We keep three kinds of record apart.
Treatment data needs three separations the field often collapses into one. Keeping them apart stops a meaningful experience from being filed as a symptom, and stops a mild physical effect from being filed as harm.
The experiential layer
Records what the patient experienced, in its own terms. Its default classification is an experiential observation, not a disorder, diagnosis, or adverse event.
Acute effects
Records a physical effect such as nausea during a session. An acute effect is not an adverse event unless it independently meets the criteria for one.
The safety layer
Uses MedDRA, SNOMED CT, and CTCAE only when symptoms, impairment, intervention, or harm actually occurred. A challenging experience is not an adverse event unless it became one.
The gap
The experiences the stack was not built to carry.
The medication and measurement layers are covered. The experiential layer is not. A profound shift a patient describes as healing can end up coded as a symptom of dissociation, because that is the closest label the safety dictionaries offer. Code it that way at scale and the research record will say the opposite of what happened. That is the gap the Standards Initiative is working to close.
These experiences are widely reported, measured by validated instruments or research methods, and can carry clinical weight. No broadly adopted code represents them without labeling them as disorders.
Explore each concept in the map above: its instrument anchor, nearest clinical code, and when safety coding applies. The map's list view is the full plain-text reference.
A label records what was described, not what is true. If a patient describes contact with a deceased person, the record keeps that description and notes who reported it. The label never turns the experience into a diagnosis.
Read why the weeks after dosing get lost, and what goes unmeasured while the vocabulary is missing.
Our answer
Document the ‘aha moment’, and more.
A practitioner running a session can mark the moment it happens. One tap records it against the session clock, with no form to complete and no attention taken from the room. The marker says a practitioner thought this point was worth returning to. The marker says nothing else.
Alongside it sits a governed catalog of experiential observations, chosen from a fixed list rather than typed as free text, so the same event carries the same meaning at every site. Acute physical effects sit in a separate class, so recording nausea never manufactures an adverse event. Where more than one source is permitted for a concept, the record also carries whether the patient reported it or the clinician observed it.
The record refuses to say what any of it meant. Interpretation stays with the clinician, and that restraint is the point. A code earns its keep only when it means the same thing to everyone reading it, and “the software determined this was a breakthrough” is not a code. That is an opinion wearing one.
Method
How concepts map, stated explicitly.
Every proposed concept carries its external relationships as typed links, not as a single overloaded code.
- Exact map. Meaning, scope, context, and intended use all match. Only then does an external code substitute for the concept.
- Broader than. The external concept covers more than the proposed concept.
- Narrower than. The external concept covers less than the proposed concept.
- Related to. Conceptually adjacent, kept as a relationship, never a substitute.
- Instrument anchor. The concept is measured by a named instrument domain or item.
- Safety overlay. A clinical or adverse-event code applied only when harm is present.
- No exact mapping identified. No equivalent found in this review. Stated as a finding, not a claim that none can exist.
Each concept is a governed node: a stable internal identifier, a preferred label, an operational definition, a default valence, and a provenance. External codes live in a separate mapping table, so one concept can carry several mappings, each with its own relationship, purpose, confidence, source, and verification date. Where the review found no equivalent, the record says so rather than forcing a near-match.
Read this before citing it
Provenance, and the open questions.
- Builds on prior work. The proposal draws on validated instruments, published phenomenology research, existing medical terminologies, and prior coding-gap analyses. It is not first.
- An open call to the field. What proves useful is offered to the relevant standards-development organizations as change requests.
- A dated snapshot. Codes change. Verify every identifier against its current authoritative release before relying on it.
- Open questions remain. Whether a given instrument's panel or total-score identity fits a specific implementation, instrument licensing for content and scoring, and the final local codes for several experiential concepts are all still under review.
Authorities as of this review. LOINC, Regenstrief, version 2.82; CPT Category III, AMA, with the CMS monitoring descriptors; ICD-10-CM FY2026, NCHS; USCDI+ Behavioral Health, ASTP/ONC and SAMHSA; RxNorm, NLM; SNOMED CT, SNOMED International; MedDRA, MSSO; UCUM, Regenstrief; HL7 FHIR, HL7 International. Last reviewed 2026-06-27.
Questions
Common questions about coding interventional mental health data.
What is the coding gap in interventional mental health data?
Most of the record is codable. Substances map to RxNorm, diagnoses to ICD, symptom scales to LOINC, and safety to MedDRA. The gap is the subjective treatment experience, such as insight, unity, or ego dissolution. Validated instruments measure these, but they have no broadly adopted clinical code, so the meaning does not travel between clinics.
Is there a standard code for psychedelic experiences like ego dissolution or the ‘aha moment’?
No broadly adopted, non-pathologizing code was identified for experiences like ego dissolution, unity, or a psychological insight event. Instruments such as the MEQ-30, CEQ, and EDI measure them, but those instruments often lack an official LOINC or SNOMED representation, so each clinic records them locally and comparison across sites breaks down.
How is a challenging experience coded versus an adverse event?
They are kept as two separate, linked records. A challenging or intense experience is recorded as an experience in its own terms. A safety code is added only when symptoms, impairment, intervention, or harm actually occurred. This keeps a meaningful experience from being filed as harm by default, and keeps the safety signal clean.
Which standards code ketamine and psychedelic therapy records today?
Substances map to RxNorm, diagnoses to ICD-10-CM in the United States and ICD-11 globally, symptom instruments to LOINC, safety to MedDRA, and records travel over HL7 FHIR. Psychedelic monitoring uses CPT Category III reporting codes, which report an emerging service rather than establishing coverage or payment. The subjective experience layer is the part without an adopted standard.