Sensitivity to change of the Roland-Morris Back Pain Questionnaire: part 2
Riddle, D. L.; Stratford, P. W.; Binkley, J. M.
BACKGROUND AND PURPOSE: One purpose of this study was to determine whether the Roland-Morris Back Pain Questionnaire (RMQ) could be used to detect clinically meaningful change in individual patients. The construct that served as the basis for this study was that RMQ change scores should be greater for patients meeting their treatment goals than for patients who did not meet their goals. The second purpose of the study was to determine whether sensitivity to change (STC) varies depending on the magnitude of the initial RMQ score. Subjects and METHODS: Of the 143 patients with low back pain who completed the study, 104 patients achieved their goals and 39 patients did not achieve their goals. Receiver operating characteristic (ROC) curve analysis and likelihood ratios were used to determine the RMQ change scores that best classify patients as having met or not met their goals. RESULTS: The area under the ROC curve for the entire RMQ scale was 0.68, while the curve areas for smaller RMQ intervals varied from 0.80 to 0.97. CONCLUSION AND DISCUSSION: The STC for the entire RMQ scale was poor for the construct examined in this study. The likelihood ratios for smaller RMQ intervals support the construct validity of the RMQ for assessing change in disability. Initial RMQ score magnitudes must be taken into account to improve the rate of making correct predictions about whether meaningful change in disability will occur following treatment.