Back

Kineret - The largest Israeli chain of hospitals data available now in OMOP common data model format enabling broad collaborative research

Rappoport, N.; Livne, G.; Perry Cohen, N.; Makover, N.; Eshel-Geva, H.; Kapach, H.; Hadad, T.; Alon, Y.; Rubin, R.; Chai, S.; Da Villa, S.; Hochman, O.

2025-07-14 health informatics
10.1101/2025.07.04.25330663 medRxiv
Show abstract

BackgroundIn 2021, the Israeli Ministry of Health established the Kineret initiative for standardizing clinical data across its medical centers and making it accessible for secondary use, research, and development. The main objectives were to reduce burdens and bottlenecks in data extraction, data cleaning, and data sharing in order to enable the reuse of patient data for translational research. The Directorate of Government Medical Centers is the governing body for the network of government healthcare centers, which includes 25 medical centers spread across Israel with 11 general medical centers, 9 mental health centers, and 5 geriatric healthcare centers. MethodsAfter considering several alternatives, the Observational Medical Outcomes Partnership Common Data Model (OMOP CDM) was chosen as the data standard for harmonizing the terminologies across institutions. An instance of ATLAS, an open-source software for scientific analyses, was made available for researchers on a secure cloud for researchers within the medical centers. This enables self-feasibility tests and exploratory analysis in an efficient way. Approved studies are conducted in a secure virtual environment on a cloud-based platform, which provides flexibility for a variety of computational resource needs. ResultsSince the initiation of the project, six medical centers have joined Kineret and their data has been fully integrated into the data lake. The seventh medical center is expected to be fully integrated during 2025. ConclusionKineret enables not only internal multi-site studies but also international studies with ease. We see that Kineret enables studies for improving local and worldwide health and healthcare. Data description is available at https://kineret.health.gov.il/

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.