Prediction of sepsis within 24 hours at the triage stage in emergency departments using machine learning

Jingyuan Xie, Jiandong Gao, Mutian Yang, Ting Zhang, Yecheng Liu, Yutong Chen, Zetong Liu, Qimin Mei, Zhimao Li, Huadong Zhu, Ji Wu

Table 1. Comparison of sepsis patients and non-sepsis patients

Parameters Sepsis patients (n=8,219) Non-sepsis patients (n=8,400) P-value Age, years, median (IQR) 66 (55-78) 48 (30-63) <0.001 Gender, female, n (%) 3,724 (45.3) 4,662 (55.5) <0.001 Body temperature, ℉, median (IQR) 98.2 (97.5-99.0) 98.0 (97.5-98.6) <0.001 Heart rate, beats per minute, median (IQR) 95 (80-110) 83 (73-96) <0.001 Respiratory rate, breaths per minute, median (IQR) 18 (16-20) 18 (16-18) <0.001 SpO2, %, median (IQR) 97 (95-99) 99 (98-100) <0.001 SBP, mmHg, median (IQR) 118 (101-138) 133 (120-148) <0.001 DBP, mmHg, median (IQR) 67 (56-80) 78 (69-87) <0.001 MAP, mmHg, median (IQR) 84.67 (72.33-98.33) 96.67 (87.67-106.33) <0.001 Shock index, median (IQR) 0.795 (0.623-0.992) 0.628 (0.529-0.738) <0.001 Arrival transport, n (%) Walking in 1,787 (21.7) 5,431 (64.6) <0.001 Ambulance 5,606 (68.2) 2,702 (32.2) Helicopter & Other 826 (10.0) 267 (3.2) Acuity, n (%) 1 2,407 (29.3) 245 (2.9) <0.001 2 4,720 (57.4) 2,635 (31.4) 3 1,024 (12.4) 4,836 (57.6) ≥4 8 (0.1) 684 (8.1)