A Clinical Prediction Rule for Functional Outcomes in Patients Undergoing Surgery for Degenerative Cervical Myelopathy: Analysis of an International Prospective Multicenter Data Set of 757 Subjects

• Published in: Journal of bone and joint surgery. American volume Vol. 97; no. 24; p. 2038
• Main Authors: Tetreault, Lindsay, Kopjar, Branko, Côté, Pierre, Arnold, Paul, Fehlings, Michael G
• Format: Journal Article
• Language: English
• Published: United States 16.12.2015
• Subjects: Adult; Aged; Aged, 80 and over; Cervical Vertebrae - surgery; Decision Support Techniques; Decompression, Surgical - methods; Disability Evaluation; Female; Follow-Up Studies; Humans; Laminectomy; Laminoplasty; Logistic Models; Male; Middle Aged; Multivariate Analysis; Poisson Distribution; Prospective Studies; Recovery of Function; ROC Curve; Severity of Illness Index; Spinal Cord Diseases - etiology; Spinal Cord Diseases - physiopathology; Spinal Cord Diseases - surgery; Spinal Fusion; Spondylosis - complications; Spondylosis - physiopathology; Spondylosis - surgery; Treatment Outcome
• DOI: 10.2106/JBJS.O.00189

Cervical spondylotic myelopathy (CSM) is a progressive spinal condition that is often managed surgically. Knowledge of important predictors of surgical outcome can provide decision support to surgeons and enable them to effectively manage their patients' expectations. The purpose of this study was to identify the most important clinical predictors of surgical outcome in patients with CSM using data from two multinational prospective studies. A total of 757 patients treated surgically for CSM participated in either the CSM-North America or the CSM-International study. The model was designed to distinguish between patients who achieved a modified Japanese Orthopaedic Association (mJOA) score of ≥16 at the one-year follow-up and those who did not (mJOA < 16). A score of 16 was chosen as the cutoff as an mJOA of ≥16 translates to minimal impairment. Univariate analyses evaluated the relationship between outcome and various clinical predictors. Multivariate Poisson regression was used to create the final prediction rule and estimate relative risks. Based on univariate analyses, the probability of achieving a score of ≥16 decreased with the presence of certain symptoms, including gait dysfunction, the presence of certain signs such as lower limb spasticity, positive smoking status, higher comorbidity score, more severe preoperative myelopathy, and older age. The final model consisted of six significant and clinically relevant predictors: baseline severity score (relative risk [RR], 1.11; 95% confidence interval [CI], 1.07 to 1.15), impaired gait (RR, 0.76 [ref. = absence]; 95% CI, 0.66 to 0.88), age (RR, 0.91 per decade; 95% CI, 0.85 to 0.96), comorbidity score (RR, 0.93; 95% CI, 0.88 to 0.98), smoking status (RR, 0.78 [ref. = non-smoking]; 95% CI, 0.65 to 0.93), and duration of symptoms (RR, 0.95; 95% CI, 0.90 to 0.99). Patients were more likely to achieve a score of ≥16 (indicating minimal impairment) if they were younger, had milder preoperative myelopathy, did not smoke, had fewer and less severe comorbidities, did not present with impaired gait, and had shorter symptom duration.