Leveraging patients' longitudinal data to improve the Hospital One-year Mortality Risk

Paper Title: Leveraging patients' longitudinal data to improve the Hospital One-year Mortality Risk Paper: https://doi.org/10.1007/s13755-024-00332-4 (full-text view-only version: https://rdcu.be/eccmN) GitHub Link: https://github.com/MEDomics-UdeS/POYM Description: This dataset accompanies Laribi et al. (2024) and contains synthetic data generated using the AVATAR method in partnership with Octopize. Files: dataset.csv: This file contains 248,485 rows and 248 columns, representing 248,485 synthetic visits from 123,646 synthetic patients. The first two and the last two columns are not used for prediction. Detailed descriptions of each column can be found in Laribi et al. (2024). To preserve patient's privacy, we did not save admission and discharge dates. Consequently, it is not possible to split the dataset temporally as done with the original dataset or to identify admissions with same-day discharge. Comparison of synthetic and original data: https://doi.org/10.21203/rs.3.rs-5363467/v1