Observational Health Data Sciences & Informatics - Kristen Kotska
Kristen Kotska is an associate director at IQVIA, running the OMOP (Observational Medical Outcomes Partnership) data network and collaborating with OHDSI, whose mission is to improve health by empowering a community to collaboratively generate evidence that will promote better health decisions and ultimately better care.
OHDSI is a global, open source initiative, with over 2700 users in 18 countries, and 152 databases and 2.1B patient records. Community stakeholders are from biopharma, health systems, payers, patient engagement groups, academia, and government and represent multiple disciplines. One unique aspect of OHDSI is keeping data local, and only sharing results. Keeping the data in a common data model behind a firewall gives data owners complete control of their analytics and ensures proper governance, data provenance, and adherence to a standard framework.
Kristen concluded by describing OHDSI community research projects exploiting the mass of data at its disposal, to go, for example, from clinical characterization studies to patient level prediction. One case involved predicting symptomatic hemorrhagic transformation and acute ischemic stroke, where a team built a patient level prediction model and validated on six databases from three countries in one day. In a second, the huge LEGEND study showed that ACE inhibitors may not be the most effective first-line course of treatment.