Cervical Spine Injury Recovery Prediction Scale: a means of predicting neurological recovery in patients with acute subaxial cervical spine injury.

Abstract

To develop a comprehensive Cervical Spine Injury Recovery Prediction Scale (CSIRPS) to predict outcomes for patients with acute subaxial cervical spine injury (CSI). 42 men and 18 women (mean age, 41 years) with acute subaxial (C3-C7) CSI and a neurological deficit were evaluated clinically and radiologically after crutch-field skeletal traction. The comprehensive CSIRPS comprised 5 predictor variables: American Spinal Injury Association (ASIA) impairment scale, maximum cord compression, maximum canal compromise, signal intensity pattern in the cord, and Cervical Spine Injury Severity Score (CSISS). Point values of each predictor were weighted based on consensus and experience, and the total effect was computed using the CSIRPS. A receiver operating characteristic (ROC) curve was plotted with the CSIRPS data to obtain the critical value that can predict walkers from non-walkers. The Chi squared value for CSIRPS was 55.771 and the contingency coefficient was 0.694 (p<0.0001), both of which were highest among each predictor's corresponding values. A ROC curve was plotted with the CSIRPS data. With 100% area under the ROC curve, the critical value of 50 was identified with a sensitivity and specificity of 100%. All the patients with CSIRPS above this level regained the useful functions on follow-up, whereas patients with lower scores did not. The CSIRPS is a practical and accurate means of predicting outcomes in patients with acute subaxial CSI.