OHDSI / OncologyWG

Oncology Working Group Repository
https://ohdsi.github.io/OncologyWG
Apache License 2.0
56 stars 24 forks source link

Open up for group discussion: Metastasis #476

Closed rtmill closed 1 year ago

rtmill commented 1 year ago

Problem Space

Standard representation for metastasis.

Proposed Solutions

Findings of metastasis when attributed to a cancer would be represented as modifiers (measurements) with FK back to that condition. Those modifiers would consist of a standard discrete set using the "Metastasis" class within the Cancer Modifier Vocabulary.

One point of contention is whether or not to represent all concepts that contain evidence of metastasis solely as measurements, or also to persist them as conditions. For instance, there are many ICD codes that imply metastasis and are in the condition domain, and this is helpful when leveraging the condition hierarchies to find all forms of neoplastic disease. The proposals seem to boil down to:

Option 1. map all of those conditions to measurements and consequentially break the condition hierarchy of neoplastic disease Option 2. map those condition concepts to BOTH measurements (within Cancer Modifier vocab) AND to standard condition concepts