Developing a simple field triage scheme compatible with regional trauma situations and medical systems using machine learning: A nationwide trauma database study DOI Creative Commons
Chie Tanaka, Takashi Tagami,

Shinnosuke Kitano

et al.

Hong Kong Journal of Emergency Medicine, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 27, 2024

Abstract Background and Objectives Trauma patients' characteristics prehospital medical systems differ between countries. We aimed to establish a simple field triage scheme for identifying severe injury patients that reflect the regional features of trauma demographics management systems. Methods This was retrospective cohort study using Japan Data Bank (JTDB). The main outcome all‐cause in‐hospital mortality. Blunt ≥18 years age were enrolled. Thirty‐five parameters (including demographic characteristics, mechanism injury, vital signs, abbreviated scale 98) assessed each patient machine learning method. After calculating feature importance variable from JTDB 2019–2021 dataset, we selected variables made schemes by including different combinations variables. analyzed receiver operating characteristic curve validated 2004–2018 dataset. Results 48,194 subjects in 2019–2021. Eight calculated evaluated 185,728 cases area under most accurate simplest 0.845. When cut‐off value set at 6, mortality rate positive group (i.e., score ≥6) 31.8%, undertriage 4.6%, overtriage 68.2%. Conclusions established an easy reliable suited Japanese system which comprised disturbance consciousness, respiratory failure, shock, head chest multiple fall height, age.

Language: Английский

Developing a simple field triage scheme compatible with regional trauma situations and medical systems using machine learning: A nationwide trauma database study DOI Creative Commons
Chie Tanaka, Takashi Tagami,

Shinnosuke Kitano

et al.

Hong Kong Journal of Emergency Medicine, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 27, 2024

Abstract Background and Objectives Trauma patients' characteristics prehospital medical systems differ between countries. We aimed to establish a simple field triage scheme for identifying severe injury patients that reflect the regional features of trauma demographics management systems. Methods This was retrospective cohort study using Japan Data Bank (JTDB). The main outcome all‐cause in‐hospital mortality. Blunt ≥18 years age were enrolled. Thirty‐five parameters (including demographic characteristics, mechanism injury, vital signs, abbreviated scale 98) assessed each patient machine learning method. After calculating feature importance variable from JTDB 2019–2021 dataset, we selected variables made schemes by including different combinations variables. analyzed receiver operating characteristic curve validated 2004–2018 dataset. Results 48,194 subjects in 2019–2021. Eight calculated evaluated 185,728 cases area under most accurate simplest 0.845. When cut‐off value set at 6, mortality rate positive group (i.e., score ≥6) 31.8%, undertriage 4.6%, overtriage 68.2%. Conclusions established an easy reliable suited Japanese system which comprised disturbance consciousness, respiratory failure, shock, head chest multiple fall height, age.

Language: Английский

Citations

0