Novel computational methods on electronic health record yields new estimates of transfusion‐associated circulatory overload in populations enriched with high‐risk patients

• DOI: 10.1111/trf.17447
• Publication Date: 2023-05-30T05:41:31Z
• Authors: Michelle Wang; Gregory M. Goldgof; Ayan Patel; Barbee Whitaker; Artur Belov; Brian Chan; Evan Phelps; Benjamin Rubin; Steven Anderson; Atul J. Butte
• Keywords: Odds; Health records

Abstract Background Transfusion‐associated circulatory overload (TACO) is a severe adverse reaction (AR) contributing to the leading cause of mortality associated with transfusions. As strategies mitigate TACO have been increasingly adopted, an update prevalence rates and risk factors using growing sources electronic health record (EHR) data can help understand transfusion safety. Study Design Methods This retrospective study aimed provide timely reproducible assessment TACO. Novel natural language processing methods, now made publicly available on GitHub, were developed extract ARs from 3178 reports. Other patient‐level extracted computationally UCSF EHR between 2012 2022. The odds ratio estimates calculated multivariate logistic regression analysis case‐to‐control matched sex age at 1:5. Results A total 56,208 patients received transfusions (total 573,533 units) during period 102 was estimated be 0.2% per patient (102/total 56,208). Patients history coagulopathy (OR, 1.36; 95% CI, 1.04–1.79) transplant 1.99; 1.48–2.68) increased Discussion While serious AR, events remained rare, even in populations enriched high‐risk patients. computational methods used find continually surveil for ARs. suggest that or presence organ should carefully monitored potential risks