Epidemiological State, Predictors of Early Mortality, and Predictive Models for Traumatic Spinal Cord Injury A Multicenter Nationwide Cohort Study

Study Design. Multi-center, retrospective cohort study. Objective. To determine the epidemiology, identify predictors of early mortality, and develop predictive models for traumatic spinal cord injury (SCI). Summary of Background Data. Despite improved initial care and management strategies, traumatic SCI remains a devastating event. Knowledge of the epidemiological state and predictive factors for mortality is important for developing strategies and counseling; however, they have not been adequately investigated, and predictive modeling regarding outcomes remains an underused modality for patients with traumatic SCI. Methods. Using a nationwide trauma registry-the Japan Trauma Data Bank-we identified adult (18 yrs) patients with SCI between 2004 and 2015. The endpoint was in-hospital mortality. Characteristics of each patient were described. Multivariate logistic regression analyses were performed to identify factors significantly associated with in-hospital mortality and develop a predictive model. Results. In total, 236,698 patients were registered in the database. Of the 215,835 adult patients, 8069 (3.7%) had SCI. The majority had SCI at the cervical level with falls at ground level being the primary etiology. Over the study period, median age, the proportion of cervical SCI, and the etiology of falls at ground level increased. The mortality rate was 5.6%. The following eight factors, age, sex, Glasgow Coma Scale on arrival (GCS), hypotension on arrival, bradycardia on arrival, severe head injury, Injury Severity Score (ISS), and neurological severity of SCI, were independently associated with mortality. A predictive model consisting of these variables predicted mortality with area under the receiver operating characteristic curve of 0.88 (95% confidence interval [CI], 0.86-0.90). Conclusion. Over the 12-year period, patient characteristics, etiology, and post-SCI outcomes significantly changed. We identified eight prognostic factors of early mortality. A predictive model including these factors showed excellent performance and may improve treatment strategies, healthcare resource allocation, and counseling.