The reliability of the Low Back Outcome Score for back pain.

Abstract

A prospective test-retest study was conducted to investigate both new and follow-up patients with low back pain presenting to an orthopedic surgeon specializing in back pain. To further validate the internal consistency and test-retest reliability of the Low Back Outcome Score, and to compare these results with other condition-specific disability scales. To be useful in clinical practice, health-specific questionnaires must demonstrate reliability and validity. Several disease-specific questionnaires for low back pain have been validated to different extents. In this study, 102 new and 42 follow-up patients consecutively attending a consultant clinic completed the Low Back Outcome Score on their visit, then again after an interval of 1 week in postal form. This instrument also was completed by 230 patients presenting to a physiotherapist. A response rate of 90% was achieved for the postal questionnaire. A test of internal consistency conducted with the study sample achieved a Cronbach alpha coefficient of 0.85. Overall agreement for test-retest reliability was 84%, and the reliability coefficient (K) reached a range of 0.51 to 0.86 (P < 0.05). A Bland/Altman plot was calculated, demonstrating that only 5% of patient scores change by more than 11.6 scale points between test and retest, which is not sufficient to change outcome categories. The Low Back Outcome Score appears to have good internal consistency and test-retest reliability for use in clinical practice.