MIT-LCP / mimic-omop

Mapping the MIMIC-III database to the OMOP schema
MIT License
123 stars 47 forks source link

Various ETL issues - not ok 1 - number of unique patients who die in the database #72

Open MattMuffin opened 2 years ago

MattMuffin commented 2 years ago

The query to build the populate the DEATH from MIMIC-III to OMOP looks for patients in two ways: 1) from the admission table, WHERE deathtime IS NOT NULL, joining then on patients 2) from the patient table, WHERE dod_ssn IS NOT NULL , and for which the patient has not been selected from the previous subquerry (admission table)

The check calls for 2 counts:

SELECT count(dod) FROM mimiciii.patients; Result: 15759 SELECT count(death_date) FROM omop.death; Result: 14849

This gives an error because of the mismatch in result. The issue here is that the count is on the dod field in the table mimiciii.patients , while the ETL is based on either admissions.deathtime or patients.dod_ssn.

If I take patient with subject_id = 263, it shows a dod but dod_ssn is null, resulting in the patient not being included in the DEATH table.

I think there is an issue: 1) either the check counts the wrong field (dod but not dod_ssn) 2) either the etl queries on the wrong field (dod_ssn and not dod)

In any case, I think we are comparing apples to oranges as dod does not imply that dod_ssn will be filled.

Please advise.

I'll dig the other errors I received from the etl_checks.