Determining a Clinically Applicable Cutoff in AI Algorithms for Predicting Clinical Deterioration: A Workload-Constrained, Alarm-Based Approach
Jaewon Jang, Yong Jun Choi, Taeyong Sim, Ki-Byung Lee, Ji-Hyun Kim, Eun Young Cho, Yuhyun Choi, Sungsoo Hong, Bo Mi Jung, Soo-Jeong Kim, Won Gi Hong, Jae Hwa Cho · 2026
Read Determining a Clinically Applicable Cutoff in AI Algorithms for Predicting Clinical Deterioration: A Workload-Constrained, Alarm-Based Approach by Jaewon Jang, Yong Jun Choi, Taeyong Sim, Ki-Byung Lee, Ji-Hyun Kim, Eun Young Cho, Yuhyun Choi, Sungsoo Hong, Bo Mi Jung, Soo-Jeong Kim, Won Gi Hong, Jae Hwa Cho (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13413000.txt.
Description
Jaewon Jang and co-authors argue in 2026 that hospital AI alarms should be tuned by workload, not by the usual statistical recipes. Their score, VC-MAES, predicts sudden deterioration such as unplanned intensive care transfer or cardiac arrest from ward data at a Korean hospital between 2020 and 2022. The team compared four ways of choosing the alert threshold, including textbook methods and one that caps alarms per hundred patient-days.
The AI score beat the standard NEWS and MEWS warning systems on accuracy in head-to-head tests on shared time points. Textbook thresholds produced alarm rates the authors judged unworkable on real wards. Their alarm-based cut-off kept sensitivity near double that of NEWS while holding alerts to about ten per hundred patient-days, a level staff could plausibly answer.
Catalog record
| CREATOR | Jaewon Jang, Yong Jun Choi, Taeyong Sim, Ki-Byung Lee, Ji-Hyun Kim, Eun Young Cho, Yuhyun Choi, Sungsoo Hong, Bo Mi Jung, Soo-Jeong Kim, Won Gi Hong, Jae Hwa Cho |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31955 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |