Dear all,
I am quite new to the OHDSI community and hope that someone can help me with my problem.
I currently have some trouble finding suitable vocabulary concepts. I am trying to identify standard measurement_concept_ids and unit_concept_ids for the following parameters, but unfortunately have not found satisfying matches:
Step Count : We measured the number of steps patients took. The data cannot be converted from 24 hours or 1 week to a different time interval. Do you have suggestion for an appropriate measurement_concept_id?
Flights Climbed : The closest concept I could find is Athena but again, conversion from 24 hrs values to ours is not possible. Our data only includes the number of flights climbed (e.g. 1). Do you have any suggestions for a standard measurement_concept_id?
Stair Ascent and Descent Speed: Also, I am looking for a standard measurement_concept_id for the parameter 3a) Stair Ascent Speed and 3b) Stair Descent Speed. The parameter describes the velocity (m/s) when the patient is climbing the stairs (up or down). I only found Athena but I am not sure if it is the best fit. Do you have a better alternative?
Boolean value: What is the correct unit_concept_id for a Boolean value (simple true/false)
I would greatly appreciate any guidance you can provide!
Thank you so much in advance!
Thank you so much for your help! I think your suggestions will work well – I doubt there are any concepts that would fit better, so I’ll go with your ideas.
Would you also happen to have any thoughts on the others (Boolean, Stair Ascent/Descent Speed)? I’d really appreciate any input!
Thanks again!
Your line of questioning makes me wonder, are you converting survey data to the OMOP CDM? If yes, we have a Survey WG tackling these types of questions. You should link up with them to find and implement best practices.
Hi Melanie,
Thank you for your reply! No, I’m not converting survey data; the data was measured. Do you have any suggestions on how to approach this? Thank you!
I think you are going about this the correct way, @Fred_B
These free-text strings are difficult to map because most are not covered by any vocabulary. The best you can do is find the closest matching standard concept. I highly suggest someone with a medical background does the mapping. There’s a lot of nuance in the semantics of medical / healthcare vocabularies.
There isn’t a unit for a Boolean value. This field is nullable, so leave it NULL.
Hi! I’m currently working with a parameter that simply indicates whether the patient stood during an hour or not. The parameter answers the question: “Did the patient stand or not?” (Yes/No).
Now, I want to find the correct unit_concept_id for this parameter?
Hm. This is difficult. First, it is not clear how this is a health-related fact, unless it is part of a diagnosis (e.g. orthostatic hypotension) or a Measurement (blood pressure while standing). And second, this is a massive violation of the Closed World assumption. Which means, you cannot rely on this fact being recorded (patient is standing, sitting, lying down, talking, whistling…). It is almost 100% likely not to be recorded except in your survey-like dataset for some special situation and use case.
You have several options:
Represent it through an Expansion, i.e. fields and tables that are outside OMOP. There is likely nobody else in the world looking for this fact and having data on it.
Combine it with a relevant fact (e.g. diagnosis or measurement).
Shove it into the Observation table, filling the value_as_concept_id field with some made up Boolean concept (you have all the concept_id’s above 2 Billion to yourself).
Thank you very much for your answer! It helps a lot - I’ll take some time to consider which of the four options will fit best for my use case. Thanks again!