Introduction: Cervical spondylotic myelopathy (CSM) is frequently treated surgically to relieve symptoms and to prevent further progression of the pathology. However, long-term outcomes have been seldom reported, especially considering different surgical accesses (anterior, posterior, combined anterior and posterior) in large patient cohorts. In particular, the length of the time period in which improvements can be expected after the treatment has never been extensively investigated.

Materials and methods: An institutional registry was searched for patients with CSM surgically treated between 2004 and 2018 and who completed follow-up assessments 3 months as well as 1,2, and 5 years after the surgery. Mixed ANOVA analysis followed (whenever appropriate) by post-hoc Student’s T-tests with Holm correction for repeated measures was performed to assess differences in the outcomes at the various time points (‘within-subjects’) with respect to the surgical access (‘between-subjects’). The outcome measures considered were: COMI score, Patient Acceptable Symptom State (PASS, 5-point Likert scale) and patient-reported Global Treatment Outcome (GTO, 5-point Likert scale).

Results: From a total of 628 patients, 421 returned complete COMI forms at all time points. ‘Within-subjects’ differences were significant for all outcomes (p<0.001 in all cases) whereas ‘between-subjects’ were not significant (COMI: p=0.85; PASS: p=0.76; GTO: p=0.83). Regarding COMI and PASS, post-hoc testing revealed significant differences (p<0.001) between all time points except for the comparison between 2 and 5 years (p=0.74). For GTO, significant differences were detected comparing 1 and 2 years (p<0.001) but not for the comparisons 3 months-1 year (p=0.55) and 2-5 years (p=0.71). Complications were reported for 28% of the patients at 12 months.

Discussion: No differences in outcomes were found regarding the various surgical accesses. All the examined outcome measures demonstrated significant differences up to 2 years after the treatment, indicating that the time scale in which improvements could be expected goes beyond those reported for other types of spine surgery.