Abstract
Spinal cord injury (SCI) following traumatic spinal fractures substantially elevates the risk of morbidity and mortality; however, reliable early prediction is difficult, particularly in resource-limited clinical environments. This study aimed to identify early admission predictors of SCI in patients with spinal fractures and to explore factors associated with in-hospital mortality within this cohort. In this prospective cohort study, consecutive adult (≥ 18 years) polytrauma patients with clinical signs of spinal injury were enrolled at two tertiary trauma centers in Isfahan, Iran (2023-2024). Among 1000 screened patients, 422 had radiologically confirmed spinal fractures and formed the primary cohort. Admission demographics, injury characteristics, vital signs, and Injury Severity Score (ISS) were recorded. Multivariable logistic regression was used to identify independent predictors of SCI. Of the 422 patients with spinal fractures, 85 (20.1%) sustained SCI. The model identified higher ISS (OR 1.16, 95% CI 1.12-1.20, p < 0.001) as an independent admission predictor of SCI, while ICU admission (OR 3.33, 95% CI 1.94-5.70, p < 0.001) and surgical fixation (OR 3.67, 95% CI 2.10-6.50, p < 0.001) were clinical factors significantly associated with SCI. Cervical fracture level was protective (OR 0.64, 95% CI 0.47-0.87, p = 0.005). In an exploratory univariable analysis, non-survivors (n = 18) had significantly lower Glasgow Coma Scale scores, higher ISS, higher heart rate, lower oxygen saturation, and lower educational attainment (all p < 0.05). Higher ISS independently predicts SCI at admission, while ICU admission and surgical fixation are strongly associated with SCI, offering a practical early risk-stratification framework. Exploratory mortality analyses identified associated physiological and injury-severity factors, highlighting avenues for future prognostic research. This prospective study from a low-middle income country underscores the clinical utility of simple admission data in trauma decision-making.