LOINC 3107-0: Urobilinogen [Mass/volume] in Urine (update annotation)

[callahantiff]

Update annotation for a low result on this lab from HP:0031890 to NOT(HP:0031890).

[kingmanzhang]

our annotation tool has a bug that prevent us from doing this. I am currently removing this annotation and add back in after fixing the bug.
monarch-initiative/loinc2hpo#94

[callahantiff]

If it's easier, I can just update it when I add the other codes for the next pull request (planned for tomorrow :D)? Or do you prefer to do it? Either way is fine by me!

[kingmanzhang]

However, I do have a concern here--we are overriding the medical interpretation. If a lab system thinks a value is too low (L), we should probably also think so. But by assigning the L to a normal phenotype, we are basically saying their interpretation is wrong. I think we are probably doing too much.

[callahantiff]

Currently, this annotation is incorrect. A low lab result does not imply "Increased urine urobilinogen", which is how it is currently annotated. I did some investigating before I made this issue and found that low urine urobilinogen is not clinically meaningful, which is why I was inclined to annotate it as NOT(Increased urine urobilinogen). See: https://www.labce.com/spg506382_clinical_significance_of_urobilinogen_in_urine.aspx

This is the pattern I follow for all the urine analysis results because for most urine labs only a high result is clinically meaningful. This is something I asked Peter about on Tuesday. If you disagree, I can hold off on annotating any low urine results.

[pnrobinson]

Actually, I think there may be a medically relevant too low phenotype for this, see
https://medlineplus.gov/lab-tests/urobilinogen-in-urine/

[callahantiff]

OK, fair enough! It's sometimes challenging (but still fun) to know what source to trust when reviewing these labs.

In that case I think a new terms are needed for:

• Abnormal urine urobilinogen
• Decreased urine urobilinogen

Should I make a new issue for these?

[pnrobinson]

Yes, let's make NTRs!

[callahantiff]

Woo hoo! I'm on it! :D

[kingmanzhang]

This is the pattern I follow for all the urine analysis results because for most urine labs only a high result is clinically meaningful. This is something I asked Peter about on Tuesday. If you disagree, I can hold off on annotating any low urine results.

I do have some hesitation. I know we talked about this during last meeting, but I think it is best to keep such results not annotated and tell user that we cannot map such results. Because if a clinical lab reports a lab value as abnormally low, we should not override it. Unless we agree that abnormally low value is medically meaningful (and thus map it to a value different than normal), we should just not map it.

[callahantiff]

OK, I will not add these types of annotations in the next pull request.