Along those lines, my colleagues at Georgia Tech created a CT.gov scraper a while back and re-ran yesterday which they sent my way. I will add to our resources as well. It does a bit of data structuring. Results can be found here:
My pleasure, Jon – I was wondering whether I had reinvented the wheel here. Let me know what directions, if any, I can help to expand, or if we are out of the automated phase and into the manual curation phase. Thanks!
In my lab, we have created (and are improving) a similar ClinicalTrials.gov code to arrive at a set of relevant COVID19/sars-cov-2 trials. The level of structuring may be even deeper than file posted by Jon. It is based on this schema here:
Folks - I may be ghost-writing a commentary article for publication this week in a peer-reviewed journal. One fact I am trying to track down is # of studies. Can I write something like “The open science consortium Observational Health Data Science & Informatics (OHDSI) COVID-19 working group has found 134 studies …” I am saying 134 because that is what is in the Jon Duke google sheet, for all I know it’s been updated. Anyway, question stands… if anyone knows of a larger list, please let me know.
Thanks
Shawn
consultant to Gates Foundation
202-460-4660
Hi - I will need counts of everything eventually. But your document(s) as you define them might make for an excellent statistic and give me a chance to give a shout out to OHDSI.
Hi, I was the principal manager of the OMOP project when it started at PhRMA back in 2005. One of the things that occurred to me yesterday was whether we could get chemoprotective data for the anti-malarial drugs based on usage from lupus or rheumatoid arthritis patients. It’s a shot in the dark and because it’s still difficult to link pharmacy and clinical data across diverse datasets one would probably have to go into an HMO setting where they are linked. Because of the large case load of SARS-CoV-2 in Seattle, Kaiser-Permanente might be in a position to look at their patient database (I don’t know the prevalence of the patient population on therapy and the numbers may be too small). This was one of the goals of the original business plan for the project.