Patient filtering recommendations while sharing institutional de-identified OMOP data

General
1

hello,
In light of the impending California Privacy Rights ACT and existing CCPA, we have been asked by our institution to redact the following types of patient data

  • Sexually Transmitted diseases (eg AIDS/HIV)
  • Alcohol drug services, substance abuse, diagnosis or treatment information
  • Genetic testing
  • Sexual and Reproductive Health
  • Gender affirming Care
  • Intimate partner violence
  • Developmental disabilities

Specifically calling out to other CA based institutions to understand if they are filtering out similar patient data. I would love to understand and collaborate on how to identify such data.

I would welcome your thoughts and feedback.
thanks so much!

2

@mik Any thoughts?

3

I’m interested and will follow. I think redacting the coded data won’t be hard. It’s the flowsheets, social history, free text lab results, etc. that will be difficult. I’m also curious how impactful the redaction will be on studies requiring SDoH data.

4

@PriyaDesai:

Question: What is the constellation this happens? Are you redacting an OMOP database, or some other data asset?

This is a superb list of the things that tend to have a social stigma attached to it. I totally understand that there is sensitivity. However, we will harm the ability to do research if we take out these important facts. This may or may not be research involving these particular items. For example, our predictive models we use for drug exposure or outcome occurrence might get it wrong if we withhold them, even though they are about something completely different.

Reason I am giving this sermon is that we have better ways to protect data. The OMOP CDM allows the application of analytics completely remotely without any access to the patient level data. Want to bring this up in your institution maybe as an alternative?