Development and validation of predictors of respiratory insufficiency and mortality scores: simple bedside additive scores for prediction of ventilation and in-hospital mortality in acute cervical spine injury.

BACKGROUND: Numerous studies have developed a "severity score" or "risk index" for mechanical ventilation and mortality, but there are few to predict outcomes for cervical spine injury (CSI) patients. Our objective in this study was to develop a simple bedside additive predictive score for requirement for ventilation and early in-hospital mortality for patients with CSI.

METHODS: Multivariate logistic regression analysis of the data obtained from 101 patients (development set) after surgical stabilization of traumatic CSI was performed to identify independent predictors of the need for mechanical ventilation and of early in-hospital mortality. Predictors of respiratory insufficiency and mortality (PRIM) scores were developed separately for ventilation and mortality by using the coefficients of the logistic regression model. The model was validated using the receiver operating characteristics curve to test its discriminatory ability and by comparing the predicted and observed outcomes. Validation was performed on an independent data set of 87 consecutive patients (validation set) with traumatic acute CSI.

RESULTS: Mechanical ventilation was required in 16.8% of the patients, and the in-hospital mortality rate was 17.8% in the development set. Independent risk factors for mechanical ventilation were severe injury (American Spinal Injury Association Impairment Scale Grades A and B), breath-holding time, pulmonary infection, hemodynamic instability, and progressive neurologic deterioration. Scores of 15, 20, 25, 25, and 15 were assigned to these variables, respectively. Independent predictors of death were severe injury (American Spinal Injury Association Impairment Scale Grades A and B), hemodynamic instability, progressive neurologic deterioration, and mechanical ventilation. The scores assigned for each of the variables were 20, 20, 40, and 20, respectively. The PRIM scores for mechanical ventilation and mortality had excellent discrimination (area under receiver operating characteristics curve >0.75). There was good correlation between predicted and observed outcomes in the development set and the validation set.

CONCLUSION: PRIM scores enable accurate prediction of individual patient risk of need for mechanical ventilation and in-hospital mortality in association with acute CSI.