OHDSI Home | Forums | Wiki | Github

How to remove the chief complaint from condition occurrence covariate data?

estimationmethods
patientprediction

#1

Hi.

I am doing research to predict frailty using all covariate data.
So I have to create all covariate data(condition occurrence, drug, …) using FeatureExtraction package.
And I want to remove chief complaint condition concept id.
I refered to the vignette ‘Creating custom covariate builders’ on FeatureExtraction package github to remove condition chief complaint concept id.
However I had to create all covariate data, but only condition occurrence covariate data.

And I added a conditional sentence ‘where condition_type_concept_id != 42894222(chief complaint concept id)’ to ‘FeatureExtraction/sql/sql_serverDomainConcept.sql’.
However ‘java.lang.nullpointerexception’ error arised.

How to remove the chief complaint from condition occurrence covariate data?
And can I apply the solution to procedure covariate data?

Thank you.


(Martijn Schuemie) #2

Looks like you are trying to create a custom covariate builder. You may want to take a look at the vignette on creating custom covariate builders or the vignette on creating custom covariates using cohort attributes.

If you are doing a one-off study, I recommend using the latter approach (using cohort attributes). However, if you’re looking to create a re-usable process that can be re-used over many studies you may want to dive into the former


t