Skip to content
Michael Kallfelz edited this page Oct 30, 2020 · 1 revision

CAP vocabulary

Overview

This vocabulary is based on a conversion of the CAP (College of American Pathologists) eCC protocols. It is implemented as a source vocabulary for improving ETL steps from concepts capturing relevant pathological report data.

Source

XML files provided by CAP were used to retrieve the data. The source items hierarchy was reflected in concept relationships and concept_synonym_names.

Concept characteristics

  1. Concept code A numeric value (C-key) originating from the source was used as a source code. The only exception - manually created CAP Protocol codes were created as modifications of a source file name.
  2. Names Descriptions attached to distinct codes were designated as their concept names. Alternative concept name was used to preserve a maximum of relevant source data. We keep parental relationships in a concept_synonym_name as a sequence separated by |-symbol, putting them in concept_synonym table. A composite name starts from the lowest hierarchy level concept and ends with the highest one reflecting the actual hierarchy.
  3. Domains and Concept Classes

Concepts in CAP vocabulary belong to one of the three Domains:

Domain Class Description
Observation CAP Protocol, CAP Header Concepts, describe items providing information from which distinct protocol or from which variables-values logic group it originates from.
Meas Value CAP Value Answers in CAP form, for example “3 o'clock position”
Measurement CAP Variable Question in CAP form, for example “tumor position clockwise”

Class and subsequently domain recognition was performed based on listed rules:

Class HTML-tag accessory Name restrictions
CAP Value LI No
CAP Header S Not equals 'Distance'
CAP Variable Q,S For Q-tag all items included, for S-tag name equality to 'Distance' was needed
CAP Protocol Not Applicable Not Applicable

Original DI-tag was considered as a comment, guide for a pathologist, not significant for Observational research

Relationships

  1. Internal relationships

CAP vocabulary includes a set of hierarchical and attributive relationships.

Relationship Reverse relationship Linked concepts
CAP Value of Has CAP value Cap Value ↔ Cap Variable
Has CAP parent item CAP parent item of Any concept_class ↔ Any concept_class
Has CAP protocol CAP protocol of Any concept_class ↔ Cap Protocol
  1. External relationships

Nebraska Lexicon as SNOMED extension was used as a primary mapping target. Also, other OMOP CDM standard vocabularies were used to represent clinically relevant CAP entities. We provide parallel relations to each Ckey mapped to Nebraska Lexicon where relationship 'CAP-Nebraska category' reflects more general mapping, and 'CAP-Nebraska equivalent' is used to preserve maximal possible granularity.

Relationship Reverse relationship Linked concepts
Maps to Mapped from Cap concept ↔ OMOP Standard concept (full match or mapping to more generic concept)
CAP to Nebraska Lexicon equivalent Nebraska Lexicon to CAP equivalent Cap concept ↔ Nebraska Lexicon/SNOMED concept - full match
CAP to Nebraska Lexicon category Nebraska Lexicon to CAP category Cap concept ↔ Nebraska Lexicon/SNOMED concept not a full match, some granularity is missing in Nebraska/SNOMED concept
Clone this wiki locally