A clinical prediction model to determine outcomes in patients with cervical spondylotic myelopathy undergoing surgical treatment: data from the prospective, multi-center AOSpine North America study.

BACKGROUND Cervical spondylotic myelopathy is a progressive spine disease and the most common cause of spinal cord dysfunction worldwide. The objective of this study was to develop a prediction model, based on data from a prospective multi-center study, relating a combination of clinical and imaging variables to surgical outcome in patients with cervical spondylotic myelopathy. METHODS Two hundred and seventy-eight patients diagnosed with cervical spondylotic myelopathy treated surgically were enrolled at twelve different sites in the multi-center AOSpine North America study. Univariate analyses were performed to evaluate the relationship between outcome, assessed with the modified Japanese Orthopaedic Association (mJOA) score, and various clinical and imaging predictors. A set of important candidate variables for the final model was selected on the basis of author consensus, literature support, and statistical findings. Logistic regression was used to formulate the final model. RESULTS Univariate analyses demonstrated that the odds of a successful outcome decreased with a longer duration of symptoms (odds ratio [OR] = 0.80, 95% confidence interval [CI] = 0.65 to 0.98, p = 0.030); a lower baseline mJOA score (OR = 0.74, 95% CI = 0.65 to 0.84, p < 0.0001); the presence of psychological comorbidities (OR = 0.51, 95% CI = 0.29 to 0.92, p = 0.024); the presence of broad-based, unstable gait (OR = 2.72, 95% CI = 1.47 to 5.06, p = 0.0018) or other gait impairment (OR = 3.56, 95% CI = 1.75 to 7.22, p = 0.0005); and older age (OR = 0.96, 95% CI = 0.93 to 0.98, p = 0.0004). The dependent variable, the mJOA score at one year, was dichotomized for logistic regression: a "successful" outcome was defined as a final score of ≥16 and a "failed" outcome was a score of <16. The final model included age (OR = 0.97, 95% CI = 0.94 to 0.99, p = 0.0017), duration of symptoms (OR = 0.78, 95% CI = 0.61 to 0.997, p = 0.048), smoking status (OR = 0.46, 95% CI = 0.21 to 0.98, p = 0.043), impairment of gait (OR = 2.66, 95% CI = 1.17 to 6.06, p = 0.020), psychological comorbidities (OR = 0.33, 95% CI = 0.15 to 0.69, p = 0.0035), baseline mJOA score (OR = 1.22, 95% CI = 1.05 to 1.41, p = 0.0084), and baseline transverse area of the cord on magnetic resonance imaging (OR = 1.02, 95% CI = 0.99 to 1.05, p = 0.19). The area under the receiver operator characteristic curve was 0.79, indicating good model prediction. CONCLUSIONS On the basis of the results of the AOSpine North America study, we identified a list of the most important predictors of surgical outcome for cervical spondylotic myelopathy.