OHDSI Home | Forums | Wiki | Github

EHR data to OMOP CDM Work Group

Implementers
,
#79

Hello @TMS,

Probably convention note #10 from Condition Occurrence description in wiki might be helpful.

#80

@rookie_crewkie Thank you, that’s helpful !!:slightly_smiling_face:

#81

Am I correct, when following http://athena.ohdsi.org

Lab -> Domain=Measurement, Class=Lab Test
Vitals -> Domain=Measurement, Class=Observable Entity
Radiology -> Domain=Measurement, Class=Clinical Observations

Also what else we can map in measurement table?

#82

How can we bind the family history with observation table where relationship as Mother (ConceptID - 4248584, SNOMED - 72705000) having diabetes mellitus (ConceptID - 201820, SNOMED - 73211009)?

#83

@TMS

I would put Observation_concept_id as 4051255 (Family history with explicit context pertaining to mother) and value_as_concept_id as 201820 (Diabetes mellitus).

#84

Hi @TMS,

This question is best asked on the CDM Builders forum . This thread is about EHR WG related activities, so your question is likely to be overlooked.

But the short answer, you should map based on the vocabulary_id of your source data. If you have a LOINC code, join to the CONCEPT.concept_code WHERE vocabulary_id = ‘LOINC’.

I have a feeling I’m not fully answering your question. If I’m not, post it to the CDM Builders forum with a little more background. And we can figure it out.

Code V70.0
#85

Our next work group meeting is Friday, September 6th at 10am Eastern Time. We do not have a speaker or topic on the agenda. So, this will be an open discussion. The connection details:

Zoom meeting details:

Join from PC, Mac, Linux, iOS or Android: https://ucdenver.zoom.us/j/4984831362

Or iPhone one-tap :
US: +16468769923,4984831362# or +16699006833,4984831362#
Or Telephone:
Dial(for higher quality, dial a number based on your current location):
US: +1 646 876 9923 or +1 669 900 6833
Meeting ID: 498 483 1362
International numbers available: https://zoom.us/u/9hXjQ

#86

Hello,

I would love to join this work group. Can someone please add me and my colleague @marzbuzz. It would be great if we can join today’s meeting.

Thank you,
Shreya

#87

@ssanghani1,

The login information for today’s meeting is posted above. It is almost always the same login. Please give me your and @marzbuzz email addresses and I will add you to the calendar invite, so you can get it put on your calendar.

Cheers,
Melanie

#88

Hello Melanie,

These are the email addresses:
ssanghani1@northwell.edu
byang2@northwell.edu

Thank you very much. I appreciate your prompt reply.

Thanks,
Shreya

#89

I have forwarded both of you the invite! And corrected the spelling of your name, Shreya :slight_smile:

#90

Thank you so much! :slight_smile:

#91

I would like to ask members of the EHR group to comment on availability of the following events in their EHR data in the following issue here https://github.com/OHDSI/CommonDataModel/issues/287

  • provider sent a letter to patient
  • patient logged into PHR (e.g., if Mayo, patient logged into Mayo “tethered” PHR; one that is managed by Mayo; so not like Apple Health PHR )
#92

Hi Melanie
Can you add me to the calendar invites as well?
Thanks
Wilson Pace
Wilson.pace@gmail.com

1 Like
#93

For a transfusion project we have been asked to capture two datetime stamps for all laboratory related results.
Time specimen obtained and time reported out. Current idea is to add the specimen collection time as the datetime stamp for the result - as that was the patient’s situation at that time and add a row in the procedure table for the reported time with a link in the relationship table. Do others have better ideas of how to solve two datetime stamps for one result? Which datetime stamp goes where our research team can deal with but where to store the second datetime would be nice to have a set approach.

Thanks

#94

We put the collection time in the specimen table and the result time in the measurement table. Even though the result is not the exact time the sample was collected from the patient, it is the time the result is known and we don’t have any knowledge of the patients stated at the result time.

#95

I agree with the logic. We also use the collection time for Measurement data because it reflects the patients status at the time the specimen was taken.

Whether it is a Measurement/Specimen/Procedure depends on the domain_id for the concept.

As for

concept_id = 1175684, “date and time lab result reported” seems to be the best fit. It has a domain_id = Observation, which seems appropriate. And you can link the two using the Fact Relationship table.

I’ll add you to the invite!

#96

I misspoke earlier. We use the collection time both for the specimen table the measurement table. However, I think for CDM v6, there are new columns that you can use that will allow you to connect the specimen table and measurement table. I think this issue is the one that will link it and you can use the report time out in the measurement table. @rimma can explain a little more on how this new addition v6 can support this use case.

#97

Please include me ! Ella Young, ella.young@phsa.ca, PST

#98

Hello all!

I have canceled tomorrow’s EHR WG meeting. I am playing catch up from the Symposium. We will meet again on October 4th at 10am EST.

@ellayoung I’ll add you to the invite!

t