STUDY DESIGN: Predictive and concurrent validation study. STUDY PARTICIPANTS AND SETTING: One hundred fifteen Norwegian chiropractors collected prospective data on 875 patients with persistent LBP, defined as LBP for at least 2 weeks at baseline and a minimum of 30 days totaling within the preceding year.
METHODS: Data collection took place at first consultation, fourth visit, and 3 months using the BQ, the revised Oswestry questionnaire, and a 10-point pain box scale. Follow-up at 12 months included the BQ, Oswestry questionnaire, and additional questions on the number of days with LBP and the number of days off work in the past year. DATA ANALYSIS: Frequency of reporting of each 7 items in the BQ at baseline was identified as median value with 10th and 90th percentiles. Concurrent analyses of the 2 questionnaires were made at the 4 points in time with calculation of mean differences with limits of agreement together with Bland-Altman plots. Logistic regression was used to identify and compare the predictive values of the questionnaires and to test the relevance of each individual item in the BQ.
RESULTS: The median baseline values of the 7 items in the BQ ranged from 2 to 5. The 2 questionnaires did not agree on patients' status, and mean differences between the Oswestry questionnaire and the BQ were largest when patients reported higher scores. The predictive values for the 2 questionnaires were low, with no significant difference between the 2. The predictive value of the BQ could be improved by removing most of the 7 items. Certain items can predict specific outcomes.
CONCLUSIONS: The BQ is not a useful instrument to identify baseline status, monitor progress, or predict the 1-year progress in chiropractic patients having persistent LBP. However, certain individual items are useful to predict specific outcomes.
Click on the above link for the PubMed record for this article; full text by subscription. The abstract is reproduced here with the permission of the publisher.