Psychometric characteristics and clinical usefulness of physical performance tests in patients with low back pain

Study Design. The psychometric properties and clinical use of a battery of physical performance measures were tested on 44 patients with low back pain and 48 healthy, pain-free control subjects. Objectives. Reliability, validity, and clinical use of nine physical performance measures were evaluated. Summary of Background Data. Although physical performance measures have potential use in evaluation, treatment planning, and determination of treatment outcome, there is sparse systematic investigation of their reliability, validity, and clinical use. Methods. Forty-four subjects with low back pain and 48 healthy pain-free subjects participated. The following physical performance measures were tested: distance walked in 5 minutes; 50- foot walk at fastest speed; 50-foot walk at preferred speed; 5 repetitions of a sit-to-stand task; 10 repetitions of a repeated trunk flexion task; timed up-and-go task; unloaded forward reach task; loaded forward reach task; and Sorensen fatigue test. Subjects were assessed twice on 2 days. Results. All measures had excellent intertester reliability (intraclass correlation coefficient [ICC]_1,1 >0.95). Test-retest (within session) reliability was adequate for all measures (ICC_1,1 >0.83) except repeated trunk flexion (ICC_1,1 >0.45) in the low back pain group. Test-retest (day-to-day) reliability ranged between 0.59 and 0.88 m the low back pain group and between 0.46 and 0.76 in the control group. Day-to-day reliability improved when the averages of two trials of repeated trunk flexion and sit-to-stand were used (0.76-0.91 low back pain group and 0.62-0.89 control group). Results of a multivariate analysis of variance showed a significant effect of group (F_10.65 = 3.52, P = 0.001). Results of univariate analyses showed significant group differences on all measures except the 50-foot walk at preferred speed and unloaded forward reach. Self-report of disability was moderately correlated with the performance tasks (r = 0.400 to -0.603). Conclusions. The results provide support for the use of these physical performance measures as a complement to patient self-report.