INTRODUCTION: Originally designed for surgery, the Zurich Claudication Questionnaire (ZCQ) is the most commonly used disorder-specific self-reported questionnaire for lumbar spinal stenosis (LSS). The psychometric properties of the symptom severity and physical function subscales, but not the satisfaction subscale, have been evaluated among patients with LSS undergoing nonsurgical treatments. The purposes of this study were to investigate the ability to distinguish patients who improved from those who did not, using the ZCQ satisfaction subscale as an external anchor in patients with LSS receiving nonsurgical treatment, and to estimate the minimal clinically important differences (MCIDs) of the ZCQ symptom severity and physical function subscales using this disorder-specific anchor.
METHODS: This study was a secondary analysis of a randomized controlled trial (RCT), in which patients with LSS received either supervised physical therapy or home exercise programs. The ZCQ satisfaction subscale was used by modifying the word “surgery” in the original scale in each question, to “rehabilitation.” To assess the usefulness of the ZCQ satisfaction subscale as an anchor, external responsiveness was assessed using Pearson’s correlation coefficient between the changes in outcomes and a score of the ZCQ satisfaction subscale, and receiver-operating characteristic (ROC) curve. According to the cutoff point of 2.5 for the ZCQ satisfaction subscales (range, 1–4), which is an average score, meaning the patient is at least “somewhat satisfied”, patients were divided into responders or nonresponders, and the area under the ROC curve (AUC) values were calculated. The MCIDs for the ZCQ symptom severity and physical function subscales were estimated using anchor and distribution approaches.
RESULTS: At 6 weeks, 84 (60 responders and 24 nonresponders) and at 1 year, 68 (53 responders and 15 nonresponders) of 86 patients from the original RCT were included after excluding patients who were lost to follow-up or received surgery (n = 10) during 1-year follow-up. Pearson’s correlation coefficient between the changes in outcomes and the ZCQ satisfaction subscale at 6 weeks and 1 year were 0.37–0.58 (symptom severity) and 0.40–0.45 (physical function subscales) (>0.30 is considered a good anchor). The AUC values at 6 weeks and 1 year were 0.66–0.72 and 0.63–0.71 for the symptom severity and physical function subscales, respectively (>0.7 is considered acceptable). In anchor-based approaches, the MCIDs at 6 weeks and 1 year estimated from the ROC curves ranged from –0.64 to –0.21 (symptom severity) and –0.1 to 0.1 (physical function) subscales, those from change differences between responder and nonresponder ranged from –0.29 to –0.44 and –0.39 to –0.37, and those from change differences between somewhat satisfied (ZCQ satisfaction: 2.0–2.5) and unsatisfied (³2.5) ranged from –0.18 to –0.13 and –0.24 to –0.1. In distribution-based approaches, the MCIDs at 6 weeks and 1 year estimated from 0.5 standard deviation of baseline scores ranged from –0.31 to –0.30 (symptom severity) and –0.29 to –0.27 (physical function) subscales.
DISCUSSION: These findings suggest that the ZCQ satisfaction subscale has limited ability to distinguish patients with LSS receiving nonsurgical treatment who improved from those who did not.