Movement control exercise versus general exercise to reduce disability in patients with low back pain: randomized controlled multicentre study

Purpose: To assess the optimal cut-off for of a summary score of six movement control impairments tests. Methods: Subjects were 1180 patients with back pain and 102 healthy subjects without back pain. Design:A cross sectional study with cases and controls to find out the best cut off point score or the best test combination for discriminative validity between patients with back pain and healthy controls. Statistical analyses: Diagnostic odds ratios, sensitivity and specificity for the different cutoffs as well as area under the curve values were calculated with logistic regressions with corresponding 95% confidence intervals. Results: The optimal cut-off was with two or more positive movement control impairment tests out of six, resulting in a diagnostic odds ratio of 8.04 (95%CI 4.30 to 15.06 and a sensitivity of 68% and a specificity of 79%. If analysed individually, the test with the best discriminative validity was the prone knee bend (AUC0.67; OR 5.58, sensitivity 51%, specificity 84%), followed by four point kneeling (AUC 0.65, OR 3.7, sensitivity 52%, specificity 77%; and waiters bow (AUC 0.62, OR 4.6, sensitivity 33% and specificity 90%). Conclusion(s): The optimal cut-off for discriminating patients with back pain from healthy controls was with two or more positive movement control impairment tests out of six, resulting in a diagnostic odds ratio of 8.04. Implications: It is recommendable to use the whole test battery of six tests tpoevaluate the movement control ability of the subjects. If two or more tests are positive, the Odds are 6 to 1 that this person has back pain.