Pekka Rantanen
^A1 Orthopedic Hospital of The Invalid Foundation, Helsinki, Finland

DOI: 10.1080/165019701300006515

The objective of this study was to assess the diagnostic value of common questionnaires and measures of physical performance in low back pain (LBP) syndrome. One hundred and fourteen patients with LBP classified according to the Quebec Task Force were compared with 50 patients with different pain syndromes but without apparent LBP. The discriminating value of each variable was estimated by calculating the area under the receiver operating characteristics (ROC) curve. The diagnostic value of the Million and Oswestry disability questionnaires was evident, with the area under the ROC curve varying between 0. 73 and 0. 88. The isometric trunk extension-flexion strength test with concomitant reaction-time test could not distinguish between patients (area under ROC curve 0. 50-0. 68). Sensitivity of pain drawing was excellent but specificity was low: 47% for men and 39% for women. In conclusion, disability questionnaires have discriminating power. The trunk muscle strength test does not perform well as a diagnostic tool. The area under the ROC curve and the use of other patients as controls make it easier to assess the diagnostic specificity of a particular method.