@inproceedings{10ff1cec85e04401aeedf3e3df71f7d3,
title = "Effects upon postoperative atrial fibrillation prediction of varied observation time windows",
abstract = "Post-operative atrial fibrillation (PoAF) occurs in 20-40% patients undergoing cardiac bypass graft surgery, but predicting its occurrence is a known challenge. Onset time of PoAF varies from one day to a few weeks after the cardiac surgery. We hypothesize that some of the difficulty in predicting PoAF and the inconsistency reported in previous studies stems from this variation, as patients with different time of onset after surgery also have different characteristics, and predictive models vary based on the window of observation after surgery. Here, we illustrate temporal dynamics in demographics and risk factors of AF incidence by developing models to predict PoAF onset day by day after surgery. In our results, age and prior AF remain associated with PoAF across different time observation windows (i.e. 0-7 days). The effect of gender, smoking, diabetes, and pre-operative beta blocker use differed with the length of post-surgery observation for AF.",
keywords = "EHR, Logistic regression, Postoperative atrial fibrillation, Predictive modeling",
author = "Yu Deng and Kevin Yu and Johnson, {Ethan M.I.} and Melnick, {David S.} and Sandhu, {Sukhveer S.} and Mozziyar Etemadi and Kho, {Abel N.}",
note = "Publisher Copyright: {\textcopyright} 2019 IEEE.; 7th IEEE International Conference on Healthcare Informatics, ICHI 2019 ; Conference date: 10-06-2019 Through 13-06-2019",
year = "2019",
month = jun,
doi = "10.1109/ICHI.2019.8904522",
language = "English (US)",
series = "2019 IEEE International Conference on Healthcare Informatics, ICHI 2019",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
booktitle = "2019 IEEE International Conference on Healthcare Informatics, ICHI 2019",
address = "United States",
}
