Dear OHDSI community,
I am a graduating oncology fellow at Stanford, soon to be a GI Oncologist at Northwell Health specializing in pancreatic cancer. I aim to build a large curated database of patients with pancreatic cancer, and then explore whether differences in survival exist among centers, and whether those differences relate to several clinical and molecular covariates. To facilitate analyses across centers I was thinking of building the database with a OMOP CDM. Much of the data input data I will be using lie in the hospital’s EMR, but these data need to be curated (to be accurate/reliable), added to new data that arise when I see patients, and then stored somewhere (to be aggregated/analyzed) to be analyzed with other institutions.
I would appreciate advice from the community for how to accomplish the above, i.e. to build an OMOP-compatible database from an existing EMR to analyze data across centers.
I did a PhD in genomics and comfortable in perl/R, am a novice in mysql. I learned about OMOP last week at Stanford’s clinical informatics conferences, thank you to those who presented there.
Dan King MD PhD