callahantiff / PheKnowVec

Translational Computational Phenotyping
2 stars 0 forks source link

Clinical Source Code Verification: Student 2 #88

Closed callahantiff closed 4 years ago

callahantiff commented 5 years ago

Student (GitHub Username): @kandrews11 Verification Number: 2

Assigned Domain Expert: @jwyrwa

Verification Task Description: DropBox Paper Documentation

Phenotype Descriptions: project wiki

Verification Assignments:

callahantiff commented 5 years ago

@kandrews11 - Just a reminder, once you have finished verifying each file, can you please do the following:

Thanks and great work so far! 💪

callahantiff commented 5 years ago

@kandrews11 - Emily brought up a good point last night that I wanted to make sure that I informed you on. She asked me: "should I be determining whether the source string and code are a good match or should I be comparing the code to the phenotype (Crohn's disease)? For example, the data set includes the source code "felty syndrome" which is not a good match for Crohn's disease however is a good match when compared to the source string "felty"..."

I wanted to make sure that knew that this task will actually involve both of the things Emily described; thinking about the matches that are returned within the context of the the phenotype definition and the other codes/information we are provided. For the example you raised, the string "felty" is in reference to criteria that is used to exclude control patients. You can get to these specific details by clicking on the hyperlinks in the phenotype definition. So, you are right that it is not a good match to Crohn's disease, but in this case, it is a good match to Crohn's disease control patient exclusionary criteria.

Does this make sense?

callahantiff commented 4 years ago

Please feel free to re-open if more work is needed on this task.