search

callahantiff / PheKnowVec

Translational Computational Phenotyping
2 stars 0 forks source link

Student 1 Verification: Hypothyroidism (drugs) #49

Closed callahantiff closed 5 years ago

callahantiff commented 5 years ago

Student (GitHub Username): @deringtonc Verification Number:** 1

Verification Assignments: Hypothyroidism

deringtonc commented 5 years ago

@callahantiff interesting observation while reviewing these: many of the products in this spreadsheet are agents that may cause hypothyroidism, but do not treat hypothyroidism.

I marked as "NOT clinically relevant," but it might be interesting to think about later, especially when developing algorithms to identify these phenotypes in EHR data.

callahantiff commented 5 years ago

@callahantiff interesting observation while reviewing these: many of the products in this spreadsheet are agents that may cause hypothyroidism, but do not treat hypothyroidism.

I marked as "NOT clinically relevant," but it might be interesting to think about later, especially when developing algorithms to identify these phenotypes in EHR data.

@deringtonc that is an interesting observation. Did you happen to investigate some of these within the scope of the phenotype definition? It looks like a lot of the drugs fall under the criteria for being included as a case patient. Does that make sense?

deringtonc commented 5 years ago

@callahantiff interesting observation while reviewing these: many of the products in this spreadsheet are agents that may cause hypothyroidism, but do not treat hypothyroidism. I marked as "NOT clinically relevant," but it might be interesting to think about later, especially when developing algorithms to identify these phenotypes in EHR data.

@deringtonc that is an interesting observation. Did you happen to investigate some of these within the scope of the phenotype definition? It looks like a lot of the drugs fall under the criteria for being included as a case patient. Does that make sense?

@callahantiff It does make sense. Correct me if I'm wrong: because these are thyroid altering drugs, we want to exclude those patients who have NO history of using these drugs in order to find cases with hypothyroidism.

So, we should mark these drugs as clinically relevant. I've changed it in the spreadsheet!

deringtonc commented 5 years ago

@trinklek I need your thoughts!

For this phenotype we are identifying iodine-containing products as a potential cause of hypothyroidism.

Through my literature search, topical iodine products have only been associated with hypothyroidism in children/neonates, not adults (Evidence: https://www.ncbi.nlm.nih.gov/pubmed/7489070 and https://www.ncbi.nlm.nih.gov/pubmed/9329422). So I'm counting them as clinically relevant. However, for topical iodine products that clearly wouldn't be used in children, should I count these as clinically relevant?

Example here with a douching product: https://docs.google.com/spreadsheets/d/1Va7cin600gumAUxQ-MT6InADK87iIwpqR0Qmca22dCc/edit#gid=1324889621&range=1537:1537

trinklek commented 5 years ago

Fantastic question! I would not include such examples as clinically relevant. As I know you will, commenting on this issue will be good. @callahantiff, I know you look at each of these, but want to make sure you see this conversation.

deringtonc commented 5 years ago

Completed 6/29/19

callahantiff commented 5 years ago

Fantastic question! I would not include such examples as clinically relevant. As I know you will, commenting on this issue will be good. @callahantiff, I know you look at each of these, but want to make sure you see this conversation.

@deringtonc @trinklek - Thanks so much for noting this!