Association between prehospital time and mortality in severe trauma: a restricted cubic spline analysis from a physician-staffed helicopter emergency medical service.

backgroundThe relationship between prehospital time and trauma mortality remains controversial. Most studies have employed linear models assuming a simple “shorter is better” relationship, potentially overlooking non-linear associations. We aimed to characterise non-linear associations between prehospital time intervals and mortality using restricted cubic spline (RCS) logistic regression in patients with severe trauma transported by physician-staffed helicopter emergency medical services (HEMS).

methodsThis retrospective cohort study included patients with severe trauma (Injury Severity Score ≥ 15) transported by physician-staffed HEMS to a Korean level I trauma centre between January 2019 and June 2025. The primary outcome was in-hospital mortality. Multivariable logistic regression was used to identify mortality predictors. RCS logistic regression models (df = 3) were used to examine non-linear associations between each prehospital time interval and in-hospital mortality (binary outcome). The HEMS operates under the Korean rendezvous model, in which patients are transferred from ground EMS at a designated landing zone. Generalised linear models with gamma distribution were used to analyse factors associated with HEMS on-scene time (HEMS OST).

resultsAmong 853 patients, 123 (14.4%) died during hospitalisation. RCS logistic regression models showed non-linear associations between prehospital time intervals and in-hospital mortality. Adjusted odds ratios (ORs) gradually decreased below unity at longer time intervals relative to reference values: dispatch time (reference 12 min) OR at 20 min = 0.88 (bootstrap 95% CI 0.58–1.50); HEMS OST (reference 4 min) OR at 15 min = 0.56 (95% CI 0.29–1.27); HEMS mission time (reference 45 min) OR at 60 min = 0.77 (95% CI 0.47–1.02). Bootstrap 95% CIs included unity at most time points, indicating exploratory findings. Tracheal intubation was the primary independent determinant of prolonged HEMS OST (Exp(β) = 2.76, p < 0.001), whereas i-gel® supraglottic airway insertion did not significantly prolong HEMS OST (p = 0.242).

conclusionsPrehospital time intervals showed non-linear associations with in-hospital mortality, challenging the conventional “shorter is better” paradigm. Tracheal intubation was the primary determinant of prolonged HEMS on-scene time. These exploratory findings warrant validation in prospective multicentre studies. CLINICAL TRIAL NUMBER: Not applicable.