Minimum important change values for pain and disability: which is the best to identify a meaningful response in patients with chronic nonspecific low back pain?

Purpose: To examine the association between different minimum important change (MIC) values for pain and disability and a successful response in global perception of improvement in patients with chronic nonspecific low back pain (CNLBP). Methods: A prospective cohort study was conducted. At baseline, all participants completed a sociodemographic and clinical questionnaire, the Numeric Pain Rating Scale and the Quebec Back Pain Disability Scale (QBPDS). After a physiotherapy program, the Global Perceived Effect Scale (GPES) was completed together with pain and disability measures. The association of the different literature MIC values for pain and disability with a successful response on the GPES was analyzed using logistic regression models. The discrimination power, sensitivity, specificity and predictive values were computed. Results: A total of 183 patients with CNLBP participated in this study. A reduction of 30% on the QBPDS (OR = 7.8; area under the curve = 0.73; sensitivity = 0.72; specificity = 0.76) most accurately identified patients who perceived a global improvement on the GPES. Composite criteria using both pain and disability MIC values presented high odds ratios and specificity values, but failed to identify patients who perceived a meaningful improvement. Conclusion: A 30% reduction on the QBPDS is recommended to identify patients with CNLBP who achieve a clinical improvement with physiotherapy treatment.