Reliability, Validity, and Responsiveness of the Norwegian Versions of the Maine-Seattle Back Questionnaire and the Sciatica Bothersomeness and Frequency Indices

Abstract

Cross-sectional and prospective study of patients with sciatica and disc herniation. To describe the process of translation of the Norwegian versions of the Maine Seattle Back Questionnaire (MSBQ), the Sciatica Bothersomeness Index (SBI) and the Sciatica Frequency Index (SFI), and to test the reliability, construct validity and responsiveness of these Norwegian versions. The instruments have been validated in English only. Reliability was assessed by internal consistency (Cronbach alpha), test-retest reliability (ICC) and measurement error (MDC and Bland and Altman's limits of agreement). Construct validity was assessed by comparing the 3 measures with subscales of SF-36, ratings of back and leg pain, and clinical findings. Responsiveness was assessed by standardized response means and Receiver Operating Characteristic curve analyses. Four hundred sixty six patients were included in the study, of whom 87 participated in the test-retest study. Completion time for the MSBQ was 1-2 minutes and 30 seconds for the SBI and the SFI, respectively. ICCs varied between 0.86 and 0.90. Cronbach alpha was 0.74, 0.70, and 0.65 for the MSBQ, SBI, and SFI, respectively. The measurement errors constituted 26% of the total MSBQ score range, 22% of the SBI and 27% of the SFI score range. Compared to the MSBQ the 2 Sciatica Indexes discriminated better between patients with normal versus abnormal clinical findings, but correlated less strongly with pain and physical functioning. Responsiveness was assessed by 2 external criteria at 3 months, a global change score (improved vs. unchanged) and surgical treatment status (operated vs. nonoperated). All standardized response means were >or=1.3 and all AUCs >or=0.75. Differences between the 3 measures were generally small. The Norwegian versions of the MSBQ and the Sciatica Frequency and Bothersomeness Indexes were rapidly administered, with acceptable internal consistency, test-retest reliability, measurement error, construct validity and responsiveness.