Responsiveness and Minimal Important Change for Pain and Disability Outcome Measures in Pregnancy-Related Low Back and Pelvic Girdle Pain

Abstract

Abstract Background Pregnancy-related low back pain and pelvic girdle pain (LBP/PGP) are common and negatively impact the lives of many pregnant women. Several patient-based outcome instruments measure treatment effect, but there is no consensus about which measure to use with women who have these pain presentations. Objective The objective was to compare the responsiveness of 3 outcome measures in LBP/PGP: Oswestry Disability Index-version 2.0 (ODI), Pelvic Girdle Questionnaire (PGQ), and 0 to 10 numerical rating scale for pain severity (NRS); and to estimate a minimal important change (MIC) for these measures in pregnancy-related LBP/PGP. Design This was a methodology study using data from a pilot randomized controlled trial. Methods Women (N = 124) with pregnancy-related LBP/PGP were recruited to a pilot randomized controlled trial evaluating the benefit of adding acupuncture to standard care, of whom 90 completed an 8-week follow-up. Responsiveness was evaluated by examining correlation between change score and the external anchor (6-point global perceived effect scale) and by using receiver operating characteristic curve analysis. MIC was estimated using anchor-based methods. Results All measures showed good responsiveness, with areas under the receiver operating characteristic curve ranging from 0.77 to 0.90. The estimated MICs were 3.1, 11.0, 9.4, 13.3, and 1.3 for the ODI, PGQ-total, PGQ-activity, PGQ-symptoms, and NRS, respectively. All the measures, apart from ODI, had MICs larger than the measurement error. Limitations The lack of an optimal “gold standard” or external criterion for assessing responsiveness and MIC was a limitation of this study. Conclusion All 3 outcome measures demonstrated good responsiveness. MICs were derived for each instrument. The PGQ at 8 weeks postrandomization was identified as an appropriate outcome measure for pregnancy-related LBP/PGP because it is specific to these pain presentations and assesses both activity limitations and symptoms. The NRS is an efficient, shorter alternative.