Temporal Generalizability of Machine Learning Models for Predicting Postoperative Delirium Using Electronic Health Record Data: Model Development and Validation Study

Koutarou Matsumoto   ¹ , MPH, PhD ;   Yasunobu Nohara   ² , PhD ;   Mikako Sakaguchi   ³ , BSN ;   Yohei Takayama   ³ , MS ;   Syota Fukushige   ⁴ , MS ;   Hidehisa Soejima   ⁵ , MD, PhD ;   Naoki Nakashima   ⁶ , MD, PhD ;   Masahiro Kamouchi   ^{7, 8} , MD, PhD

¹ Biostatistics Center, Kurume University, Kurume, Japan

² Big Data Science and Technology, Faculty of Advanced Science and Technology, Kumamoto University, Kumamoto, Japan

³ Department of Nursing, Saiseikai Kumamoto Hospital, Kumamoto, Japan

⁴ Department of Inspection, Saiseikai Kumamoto Hospital, Kumamoto, Japan

⁵ Institute for Medical Information Research and Analysis, Saiseikai Kumamoto Hospital, Kumamoto, Japan

⁶ Medical Information Center, Kyushu University Hospital, Fukuoka, Japan

⁷ Department of Health Care Administration and Management, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan

⁸ Center for Cohort Studies, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan