185. A clinical model to predict postoperative improvement in subdomains of the modified Japanese Orthopedic Association Score for degenerative cervical myelopathy

Abstract

<h3>BACKGROUND CONTEXT</h3> Longstanding and progressive compression of the cervical spinal cord can lead to irreversible loss of neurologic function due to demyelination and apoptosis of oligodendrocytes. A commonly used metric to quantify the severity of cervical myelopathy is the modified Japanese Orthopedic Association (mJOA) score. The mJOA score comprises six items to assess the impact of spinal cord compression on: (1) ability to feed oneself, (2) ability to walk, (3) loss of feeling or numbness in arms, (4) loss of feeling or numbness in legs, (5) loss of feeling or numbness in body and (6) ability to urinate. <h3>PURPOSE</h3> In the present study, the primary objective was to construct a clinical prediction model for improvement of mJOA subdomains at 12 months following surgery utilizing data from a longitudinal, multicenter clinical spine registry. <h3>STUDY DESIGN/SETTING</h3> This study was conducted using data from the cervical module of the Quality Outcomes Database (QOD), a longitudinal, multicenter, prospective spine outcomes registry. <h3>PATIENT SAMPLE</h3> A total of 5,000 patients who underwent elective surgery for cervical myelopathy were enrolled into the registry and had complete 12-month follow-up. <h3>OUTCOME MEASURES</h3> mJOA subdomain scores. <h3>METHODS</h3> This study was conducted using data from the cervical module of the Quality Outcomes Database (QOD). The outcomes of interest were the subdomains or items of the mJOA at 12 months following surgery. A multivariate multivariable proportional odds ordinal regression model was developed for patients with cervical myelopathy using a latent variable approach. The latent variables were assumed to follow a multivariate logistic distribution which was constructed by univariate logistic margins and a t copula. Patient demographic, clinical, and surgery covariates as well as baseline subdomain scores were included as the independent variables. The model was internally validated using bootstrap resampling to estimate the likely performance on a new sample of patients. <h3>RESULTS</h3> A total of 5,000 patients who underwent elective surgery for cervical myelopathy were enrolled into the registry and had complete 12-month follow-up. The mean age for the patients was 60.9 (±11.4) years and comprised of 47% (n=2,339) females. Patients had statistically significant improvement from baseline to 12-months postsurgery on all the mJOA subdomains(p<0.001). The multivariable analysis identified that the baseline subdomains of the mJOA were the strongest predictors of 12-month scores, with numbness in legs and ability to walk predicting 5 of the 6 mJOA items. Additional covariates that significantly predicted 3 or more of the subdomain mJOA scores at 12 months included age, preoperative anxiety or depression, gender, race, employment status, duration of symptoms, smoking status, and presence of listhesis on radiology. Surgical approach, the presence of motor deficits, number of surgical levels involved, history of diabetes mellitus, worker's compensation claim and a patient's insurance had no impact on the 12-month scores of the mJOA. The discriminative ability of the model regarding joint probabilities measured by unweighted generalized C-index was 0.753. <h3>CONCLUSIONS</h3> In conclusion, our study has developed and validated a clinical prediction model for improvement in mJOA scores at 12-months following surgery. Results highlight the importance of assessing preoperative numbness symptoms and walking ability as well as the modifiable variables of anxiety/depression and smoking status prior to surgery. Additional patient demographic variables to consider are age, gender, race, employment status, duration of symptoms, and presence of listhesis when counseling patients prior to surgery. This prediction model has the potential to assist surgeons, patients and families when considering surgery for cervical myelopathy and provides clinically useful information in the preoperative setting. Future steps include prospective, external validation of the predictive model in order to assess the reproducibility and clinical utility of this work. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.