Availability of OHDSI tools in Secure Data Environments

AlexKnightHDR · December 5, 2023, 3:04pm

Hi! As a relative newcomer to the OHDSI world, I am hoping that you can help me answer a question that has come up in my work, but might be of general interest.

I’m Project Manager for Data Standards at Health Data Research UK, and I’m working with several teams who are currently engaged in mapping a variety of large data sets to OMOP.

Some teams, particularly those working with National Health Service data, have to work in “locked down” Secure Data Environments which strictly restrict what software tools can be run. For example, in the national NHS SDE, only R packages from CRAN can be installed/run. Unfortunately this means that some parts of the OHDSI stack are unavailable - specifically the Data Quality Dashboard.

Have other parts of the community encountered this kind of problem, and how did you work round it? Are there solutions to making the tools more easily available?

Thanks in advance for your thoughts

Alex

solmaz.eradat · December 5, 2023, 4:23pm

Hi @AlexKnightHDR ,

The best way we found in getting round the problem of bringing in OHDSI tools such as WhiteRabbit, Usagi, DQD which are hosted on GitHub into a secure environment was by downloading the source code directly from GitHub, copying the zip file into the secure environment and installing them inside the secure environment. This is a manual step, but we found it was not to time consuming, manageable and maintainable.

Hope this helps

Many thanks,
Solmaz

Patrick_Ryan · December 5, 2023, 4:33pm

@AlexKnightHDR , there are numerous examples in our community of folks working within secure environments, But a very timely one was highlighted on the OHDSI community call today , @Chungsoo_Kim presented about their recent paper around creating an infrastructure to use OMOP CDM and deploy OHDSI tools for South Korean national claims data.

You should definitely check out the presentation, will be posted here: Community Calls – OHDSI . And read the team’s paper, published here: https://www.nature.com/articles/s41597-023-02580-7.