Effect of prehospital cervical spine immobilization on mortality and neurological outcome in trauma patients: an observational cohort study with propensity score matching based on the trauma registry of the German trauma society (TraumaRegister DGU®)

Abstract Background Prehospital cervical spine immobilization is standard trauma care intended to prevent secondary spinal cord injury, yet evidence for its effectiveness is limited, and adverse effects are increasingly recognized. Large registry-based comparisons accounting for injury severity are scarce. This study investigated the association between prehospital cervical spine immobilization and hospital mortality and neurological outcome in trauma patients. Methods Retrospective, multicenter, observational cohort analysis of the TraumaRegister DGU® (2020–2022). Patients aged ≥ 1 year who were primarily admitted in Germany, Austria, or Switzerland and had documented cervical spine immobilization, pupil status, and prehospital Glasgow Coma Scale (GCS) were included. A propensity score from logistic regression on injury severity, mechanism, and prehospital parameters was used for exact 1:1 matching, repeated in the subgroup with cervical spine injury (AIS ≥ 2). Primary outcome: hospital mortality; secondary outcome: Glasgow Outcome Scale (GOS) at discharge from the acute trauma admission. Results Out of 31,055 patients, the prevalence of cervical spine injury (AIS ≥ 3) was 3.5%, only 65.9% received cervical spine immobilization. Immobilized patients had higher injury severity and were more likely to have high-energy trauma or be treated by an emergency physician. In the matched cohort (n = 16,922), no clinically meaningful difference was observed in hospital mortality (12.0% with immobilization vs. 13.6% without; p = 0.001) or in crude neurological outcome at discharge as measured by the Glasgow Outcome Scale (GOS). In the subgroup with confirmed cervical spine injury (AIS ≥2; n = 1,362), mortality (17.9% non-immobilized vs. 18.2% immobilized, p = 0.888) and neurological recovery did not differ significantly between groups (vegetative state: 1.8% non-immobilized vs. 2.7% immobilized, good recovery 46.2% non-immobilized vs. 43.4% immobilized, p = 0.626). GCS change between scene and hospital arrival — a confounded, non-specific measure—showed marginally more improvement in immobilized patients (15.7% vs. 12.8%; p < 0.001). Conclusions In this registry-based analysis, prehospital cervical spine immobilization—recorded as a binary variable—was not associated with a measurable difference in hospital mortality or in crude neurological outcome at discharge (GOS), even in patients with confirmed cervical spine injury. These findings do not support routine, undifferentiated immobilization.