Abstract Sat1304: A Predictive Score for In-Transit Cardiac Arrest in Trauma Patients: Development and Validation Using a National Registry

Abstract

Introduction: In-Transit cardiac arrest in trauma patients, although infrequent, is a critical event with high mortality. Accurate prediction of this event at emergency medical services (EMS) contact is essential for timely interventions. Objective: To develop and validate a simplified predictive score to estimate the risk of cardiac arrest during transportation in trauma patients, using data available at the time of EMS contact. Methods: We analyzed data from the Japan Trauma Data Bank (JTDB) between 2019 and 2023. The study population consisted of trauma patients with an Abbreviated Injury Scale (AIS) score ≥ 3 who did not have cardiac arrest on arrival of EMS. Patients with burns, AIS score of 6, or missing data were excluded. The primary endpoint was in-transit cardiac arrest. Patients were divided into a development cohort (2019-2022) and a validation cohort (2023). A multivariable logistic regression model was used to develop the score from the development set. The coefficients were simplified by rounding to generate an integer-based scoring system. Model performance was evaluated using 10-fold cross-validation in the development set and external validation in the validation set. Results: Among 176,054 trauma patients, 66,569 patients were analyzed (development: 53,537; validation: 13,032). In-transit cardiac arrest occurred in 0.30% (160/53,537) of the development cohort and 0.35% (45/13032) of the validation cohort. The scoring formula is shown in Figure A. The area under the receiver operating characteristic curve (AUC) of the score was 0.89 (95% CI: 0.85–0.92) in the development set and 0.88 (95% CI: 0.83–0.94) in the validation set (Figure B). In both the development and validation cohorts, the incidence of in-transit cardiac arrest increased with higher score categories: 0.1% (34/48,189) and 0.1% (9/11,758) for scores of ≤0; 0.6% (15/2,375) and 1.4% (8/568) for scores 1–2; 1.1% (8/725) and 1.1% (2/178) for scores 3–4; 2.6% (30/1,165) and 2.2% (6/276) for scores 5–6; and 6.7% (73/1,083) and 7.9% (20/252) for scores ≥7 (Figure C). Conclusions: We developed and validated a simplified predictive score to estimate the risk of cardiac arrest during transport in trauma patients, Predictable Trauma Death (PTD) Score. Patients with a score ≥7 had a risk exceeding 5%, suggesting that preemptive preparations for possible cardiac arrest should be considered in this group.