Minimum Clinically Important Difference in SF-36 Scores for Use in Degenerative Cervical Myelopathy

Post-hoc analysis of 606 patients enrolled in the AOSpine CSM-NA or CSM-I prospective, multicenter cohort studies.

The aim of this study was to determine the minimum clinically important difference (MCID) in SF-36v2 Physical Component Summary (PCS) and Mental Component Summary (MCS) scores in patients undergoing surgery for degenerative cervical myelopathy (DCM).

There has been a shift toward focus on patient-reported outcomes (PROs) in spine surgery. However, the numerical scores of PROs lack immediate clinical meaning. The MCID adds a dimension of interpretability to PRO scales; by defining the smallest change, a patient would consider meaningful.

The MCID of the SF-36v2 PCS and MCS were determined by distribution- and anchor-based methods comparing preoperative to 12-month scores. Distribution-based approaches included calculation of the half standard deviation and standard error of measurement (SEM). Change in Neck Disability Index (NDI) served as the anchor: "worse" (ΔNDI>7.5); "unchanged" (7.5≥ΔNDI>-7.5); "slightly improved" (-7.5≥ΔNDI>-15); and "markedly improved" (ΔNDI ≤-15). Receiver operating characteristic (ROC) analysis was performed to determine the change score for the MCID with even sensitivity and specificity to distinguish patients who were "slightly improved" versus "unchanged" on the NDI.

The MCID for the SF-36v2 PCS and MCS were 4.6 and 6.8 by half standard deviation and 2.9 and 4.3 by SEM, respectively. By ROC analysis, the MCID was 3.9 for the SF-36v2 PCS score and 3.2 for the SF-36v2 MCS score. Using a cutoff of 4 points, the SF-36v2 PCS had a sensitivity of 72.2% and specificity of 68.1%, and MCS 61.9% and 64.6%, respectively, in separating patients who were "markedly improved" or "slightly improved" from those who were "unchanged" or "worse."

We found the MCID of the SF-36v2 PCS and MCS to be 4 points. This will facilitate use of the SF-36v2 as an outcome in future studies of DCM.

3.