OHDSI Home | Forums | Wiki | Github

Question Mapping

(JD Liddil) #1

So we have the question:
“Is the Stem Cell Transplant status date recorded?”

Looking for advice how to map. Visit_Detail? with yes/no concepts?

(Christian Reich) #2


Not sure what that question even means. Stem cell transplant status date? What’s a status date? What has VISIT_DETAIL to do with it?

(JD Liddil) #3

I don’t write the questions :slight_smile:

I just need to figure out the best way to map the question and the yes/no response since this is not a question in SNOMED or other vocabularies

(Christian Reich) #4

But if you don’t know what it means how are you going to do analytics on it? OMOP is an analytical model. Data that have no use case don’t belong there. It’s not an atic where you put things that you don’t need just in case it becomes useful one day.

(Michael Kahn) #5

Christian: Not all ODHSI folks use their CDM just for generating analytic data sets. As you know, others use it as a core data warehouse model that supports a wide landscape of (unanticipated) analytics that are not pre-specified. So having “an attic where you put things until it become useful one day” is a core use case. Supports unexpected use cases such as FDA suddenly wanting data on blood products from our attic…

(Davidcarnahan) #6

Good point @mgkahn but I’m with @Christian_Reich on this one. What does that mean? A status of a transplant would no doubt be determined by a clinician who would then write a note stating the patient was stable and then code it accordingly. I personally would like a use case of what they mean by ‘status of transplant’.

I’m not sure but it sounds like they want an easy button:

SELECT transplant cohort
FROM my_cdm
WHERE status = ‘good’

(Christian Reich) #7


Hm. Not sure. Look. if there is an analytical use case - we add it! No problem. And we definitely want to be comprehensive, no question about it.

But why do you want to store something you don’t know what it is?

(JD Liddil) #8

So I am seeking clarification for our clinical folks. Again I did not write the question. I agree it is vague at best.

(Dave Barman) #9

There are good information architecture for storing unused data in its raw format. Converting data that is not of use into OMOP CDM seems like a lot of effort for no gain.

(JD Liddil) #10

So this derives from this:
Stem cells transplant status

It is in OMOP but I could not find it since I was using Stem “Cell” not “Cells”

(Christian Reich) #11


Ah! Understood. So, the notion of “status” in the context of a diagnosis means “the patient has had it” or “the patient didn’t have it”. In this case, the stem cell transplantation. The date of that “status” is the date of the actual procedure. This is ugly bad jargon, but hey, not your fault. :slight_smile:

So, map it just like that ICD10 code, to an observation of past history (dated at the time stamp of the observation) and the procedure stem cell transplantation (dated at that “status date”):

Makes sense?

(JD Liddil) #12

Now it does. Thanks. I’m getting the hang of this. Yes I agree the wording is vague/bad. I’m running into a few of these where we pulled questions directly from various organizations.

(Christian Reich) #13

@jliddil1: Very well. And keep it coming! :slight_smile: