Evaluation of five scoring systems for cervical spondylogenic myelopathy

Comparison of measured clinical deficits and outcomes is vital for international discussion about the identification and treatment of cervical spondylotic myelopathy (CSM). There is currently little information comparing outcomes as assessed by different CSM scoring systems.

To qualitatively and quantitatively analyze five specific CSM outcome scores that are frequently used to assess the grade of severity and outcome after operative decompression.

This retrospective study evaluated the Nurick score, the Japanese Orthopedic Association score (JOA score), the Cooper myelopathy scale (CMS), the Prolo score, and the European myelopathy score (EMS).

The study included 43 patients with clinical and morphological signs of CSM, who underwent ventral decompression. Data were evaluated in sufficient detail to objectively assess the scores.

Clinical findings (funicular and radicular symptoms), recovery rate, symptom duration, age, economic situation, time away from employment, somatic-evoked potentials, and radiological findings were assessed.

Scores were assessed using both pre- and postoperative clinical data. Correlations between scores, score improvement, and how well the scores reflected the clinical, diagnostic, and anamnestic data were analyzed using nonparametric, descriptive statistical tests. The recovery rate, as a measure of cumulative outcome, was also assessed and compared for each scoring system.

All five scores were suitable for qualitatively assessing the clinical characteristics and progression of cervical myelopathy. All showed a statistically significant correlation (p<.05), and measured postoperative improvement (p<.001). All scores also reflected clinical deficits except for the Prolo score, which rates the severity of CSM with an emphasis on data related to the economic impact on the patient's situation rather than on clinical symptoms per se. Quantitative assessment of clinical symptom improvement varied greatly among the scores, for example, Nurick score (33%) versus JOA score (81%). The recovery rates, as a measure of cumulative improvement, showed less variation among most of the scores. The Nurick score and the EMS measured clinical deficit improvements in significantly fewer patients than did the JOA score (p<.05).

Evaluating the recovery rate is essential for comparing the results of the five CSM scores evaluated in this study. There was a large quantitative difference among the scores as the result of the different criteria used to produce each score. Qualitatively, all five scores allowed evaluation of cervical myelopathy, but only the recovery rate allowed for statistical comparison. Advancements in the treatment of CSM depend on the ability of clinicians to evaluate the therapeutic results of CSM studies. This study suggests that using the recovery rate to assess outcome is best for comparing studies that use different scores.