Observational Health Data Sciences and Informatics

The Anatomical Therapeutic Chemical (ATC) Classification System is a drug classification system that classifies the active ingredients of drugs according to the organ or system on which they act and their therapeutic, pharmacological, and chemical properties. It is controlled by the World Health Organization Collaborating Centre for Drug Statistics Methodology (WHOCC). Inasmuch as the ATC codes are semantic identifiers, the ATC vocabulary was taken as a basis for the OMOP Drug Hierarchy. As of 30 September 2021, there are 6497 valid codes in the ATC.

ATC source codes are obtained from WHOCC ATC/DDD Index 2021 and, in OMOP, stored in the table of class_drugs_scraper

The translation of source ATC codes to the OMOP Vocabulary is presented on the OHDSI GitHub.

1. ATC codes of 1st-3th Classes are written in UPPERCASE.
2. ATC codes of 4th Class are written in Initcap.
3. ATC codes of 5th Class are written in lowercase and contain the information about Active Ingredient(s) or Ingredient Group(s) together with the route of administration. The latter is separated from the Ingredient-related information by semicolon “;”. Different routes of one Drug Product are separated by a comma “,”.
4. OMOPized names of ATC Classes, containing information about the route of administration, are assigned by the vocabulary team through the concept_manual table. Details of ATC Adm.R-to-OMOP conversion are described below.
5. Some 5th ATC Classes, which have name duplicates in other hierarchical axes, can get additional characteristics inherited from the higher ATC class in brackets. Note, that brackets given by the source are placed before the semicolon, while OMOP-made brackets go after it.
6. Row ATC names as well as ATC names concatenated with DDD information are stored in class_drugs_scraper and concept_synonym_manual.
7. Either updated or deprecated ATC codes are marked with '[U]' or '[D]' at the beginning of the concept_name respectively.

ATC codes are alphanumeriс entities of different length, which reflects their Class:

All ATC codes are of the Drug domain, however, they may be hierarchically connected to the Device or Observation (concept_class_id = ‘Substance’) domains.

Valid ATC codes are Classifications, invalid ATC codes are Non-standard ones.

The source usually supplies replacement ATC codes for those which have been deprecated. However, there are several ATC Classes without replacement (they seem to be semantically wrong or being erroneously placed into the ATC hierarchy).

Relationships are defined within ATC and between ATC and RxNorm, RxNorm Extension, Multilex, SNOMED, NDFRT, and VA Class.

The ATC vocabulary is the hierarchical terminology, in which the 5th ATC Drug Classes, which reflect real Drug Products, are connected to the higher ATC Classes (4th-1st) via ancestry. In OMOP, the original ATC hierarchy is extended by Standard Drug Products of RxNorm and RxNorm Extension vocabularies, which are assigned to be descendants of the 5th ATC Classes.