<h3>BACKGROUND CONTEXT</h3> Spine trauma in the elderly is increasing in prevalence. Given their often extensive medical comorbidities and higher propensity to short-term mortality after spine trauma, surgical decision-making presents unique challenges when compared to younger patients with similar injury characteristics. The marked decrease in physiologic reserve, bone density and sometimes cognitive function in these patients contribute to a weakened response to injury—making traditional severity scores less reliable. <h3>PURPOSE</h3> Predictive calculators have proven useful in other areas of spine surgery; however, there have yet to be any meaningful clinical tools to predict mortality in elderly patients considered for surgical treatment of spine fractures. The goal of this study was to develop a predictive clinical calculator to predict mortality and better guide patient care and surgical decision-making in this complicated and rapidly growing population. <h3>STUDY DESIGN/SETTING</h3> Using data extracted from our clinical trauma registry, we conducted a single center retrospective cohort study. The outcome of interest was time to death within 3 years after trauma presentation. We prespecified and collected 12 baseline variables that are readily accessible and presumably relevant to mortality, including age, gender, marital status, hypotension on admission, max AIS, MFI-5, presence of high energy traumatic mechanism, injury level, spinal cord injury, closed head injury, WBC and HCT. A model-based predictive calculator was developed for patient's mortality at different time points after trauma presentation given his baseline characteristics. <h3>PATIENT SAMPLE</h3> A total of 1,746 elderly patients (65+) with spine trauma who presented to a single, level I trauma center from 2010-2019 were included. <h3>OUTCOME MEASURES</h3> Outcomes were identified and recorded including in-hospital mortality, mortality after discharge, medical and surgical complications. <h3>METHODS</h3> Retrospective chart review was completed to record comorbidities, presenting injury information, frailty scores, imaging evidence of sarcopenia and osteopenia, presence of spinal cord injury, as well as long-term outcomes for all patients who underwent operative treatment. Operative specifics including site, approach, number of levels, estimated blood loss were recorded as well as trends in lactate at the time of admission and time of surgery. Results from multivariate logistic regression analyses served as the basis for development of the model. Model calibration and discrimination were assessed graphically and by calculation of a concordance index, respectively. LASSO method was used to identify 12 variables which accounted for the majority of the variance in mortality. A subsequent model was developed with those 12 variables—including age, gender, marital status, hypotension on admission, max AIS, MFI-5, presence of high energy traumatic mechanism, injury level, spinal cord injury, closed head injury, WBC and HCT. A multivariable Cox proportional hazards model was fitted for time to death within 3 years after trauma presentation; the prespecified 12 predictors, surgical treatment as well as interactions between surgical treatment and other predictors are included in the model as independent variables. Internal validation was performed using bootstrap approach to evaluate its performance (discrimination and calibration) in future patients. <h3>RESULTS</h3> A total of 1,746 patients met inclusion criteria over the 9-year period. Patients were primarily male (n=927, 53%) with an average age of 76. All the predictors were significantly associated with mortality except presence of high energy traumatic mechanism. The most important predictors were age, MFI-5, closed head injury, HCT and hypotension. Older patients, males or unmarried patients suffered a significantly higher risk of death. Mortality hazard was also significantly higher in patients with higher MFI-5, presence of closed head injury, presence of hypotension, higher Max AIS, presence of spinal injury, higher level of WBC or injury site being cervical. A higher level of HCT decreased the risk. The effect of max AIS on mortality tended to be more noticeable in patients with surgical treatment, but injury site (cervical vs others) did not make a significant difference when patients had surgery. The model had an optimism-corrected C-index of 0.72. By visual inspection of calibration plot, the model predicted mortality risk well at a wide range of time points with a calibration slope around 0.88. The calculator was created to predict 30d, 90d, 180d and 365d mortality after injury for both operative and nonoperative management given the patient's baseline characteristics. Prospective patient information may be inputted to our online calculator to produce a personalized mortality risk. <h3>CONCLUSIONS</h3> This predictive calculator is the first comprehensive tool to be created to predict mortality in elderly victims of spine trauma. Prospective internal and external validation is needed to refine predictive power, but this tool offers valuable insight into surgical decision-making in this complex population. A better prediction of each patient's mortality risk will aid surgeons in patient and family counseling around these difficult clinical decisions. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.