Development of the cervical myelopathy severity index: a new patient reported outcome measure to quantify impairments and functional limitations

Abstract

Background ContextExisting degenerative cervical myelopathy (DCM) severity scales have significant shortcomings, creating a strong impetus for the development of a practical measurement tool with sound psychometric properties. PurposeThis work reports the item generation and reduction of the Cervical Myelopathy Severity Index (CMSI), a new DCM patient-reported outcome measure of symptoms and functional limitations. DesignProspective observational study. Patient SampleAdult DCM patients belonging to one of three distinct treatment groups: (1) observation cohort, (2) preoperative surgical cohort, (3) 6 to 12 months postoperative cohort. Outcome MeasuresPatient-reported outcome measure of symptoms and functional limitations. MethodsItem generation was performed using semi-structured patient focus groups emphasizing symptoms experienced and functional limitations. Readability was assessed through think-aloud patient interviews. Item reduction involved surveys of DCM patients with a spectrum of disease severity and board-certified spine surgeons experienced in the treatment of DCM. A priori criteria for item removal included: patient median importance/severity <2 (of 4), 30% or more no severity (response of zero), item severity correlations ≤ 0.80 (Spearman), item severity reliability (weighted kappa <0.60) based on a 2-week interval and clinician median importance <2 with retention of items with very high clinical importance. ResultsThere were 42 items generated from a combination of specialist input and patient focus groups. Items captured sensorimotor symptoms and limitations related to upper and lower extremities as well as sphincter dysfunction. Ninety-eight patients (43, 30, 25 observation, pre- and postsurgery respectively) and 51 surgeons completed the assessment. Twenty-three items remained after application of median importance and severity thresholds and weighted kappa cutoffs. After elimination of highly correlated (>0.80) items and combining two similar items, the final CMSI questionnaire list included 14 items. ConclusionsThe CMSI is a new DCM patient-reported clinical measurement tool developed using patient and clinician input to inform item generation and reduction. Future work will evaluate the reliability, validity, and responsiveness of the CMSI in relation to existing myelopathy measurement indices.