Add table PACK_CONTENT
This table already exists, we just never put it into the CDM documentation and haven't included it in the vocabulary zip file. This would be the chapter:
Table Description
The PACK_CONTENT table contains structured content of drug Packs. Packs are fixed combination of Drugs (Clinical Drugs, Branded Drug, Quant Clin Drug, Quant Branded Drug, Marketed Product without box size information). For example, contraceptives contain a certain amount of progesteron drug and estrogen preparations, sufficient for a month. This table is supplemental information to support standardized analysis of drug utilization.
User Guide
NA
ETL Conventions
The content of the table is part of the OHDSI Standardized Vocabularies and should be loaded with each release.
| CDM Field | User Guide | ETL Conventions | Datatype | Required | Primary Key | Foreign Key | FK Table | FK Domain |
|---|---|---|---|---|---|---|---|---|
| pack_concept_id | The Concept representing the Pack. | N/A | integer | Yes | No | Yes | CONCEPT | |
| drug_concept_id | The Concept representing the individual drug component inside the Pack. | N/A | integer | Yes | No | Yes | CONCEPT | |
| amount | The number of the above Drug components in the Pack. | N/A | integer | No | No | No | ||
| box_size | The number of Packs in a Drug Box. | N/A | integer | No | No | No |
To iterate on importance of this: currently, any and all HADES tools (e.g. those deriving DOSE_ERA) requiring drug dosage information ever only query the DRUG_STRENGTH table. At the same time, when source drug concepts are standardized to OMOP, the PACK_CONTENT table is populated with non-redundant data. That means that in current definition of OMOP CDM, actual dosage information for "pack-shaped" concepts is irretrievable.
https://github.com/search?q=repo%3AOHDSI%2FVocabulary-v5.0%20%22PACK_CONTENT%22&type=code
Although use of DOSE_ERA as well as dosage information in general in OMOP is rare, this needs to be fixed.
Potential concerns over drug_era calculations downstream. Based on @ekorchmar's comment above we should create a child issue detailing any updates to dose calculations that need to be implemented with the addition of the pack_content table
Potential concerns over drug_era calculations downstream
Considering that all available documentation suggests that drug era entries are defined by Ingredient content, it all depends on how the Ingredient is determined for this purpose.
Implementations relying on DRUG_STRENGTH.INGREDIENT_CONCEPT_ID will have to additionally consider PACK_CONTENT table, while those querying CONCEPT_ANCESTOR instead (example) will need no changes.