Reliability and discriminatory capacity of a clinical scale for assessing abdominal muscle coordination

This study evaluated the reliability and discriminatory capacity of a novel clinical scale for assessing abdominal muscle coordination. We investigated the interrater reliability of this tool in patients with chronic low back pain (LBP) (reliability section); the ability of this tool to discriminate healthy and LBP subjects (discriminatory section); and the association between the score and measures of pain, disability, and kinesiophobia (correlation section).

For the reliability section of this study, 14 patients with chronic LBP were included. For the discriminatory section, 10 patients with chronic LBP and 10 pain-free controls were recruited. In the correlation study, data from the 10 chronic LBP patients in the discriminatory section were used. The clinical test was conducted by a blinded examiner while the subjects attempted to independently activate transversus abdominis separate from the rest of the abdominal muscles (hollowing or draw-in maneuver). The intraclass correlation coefficients, receiver operating characteristic curve, and Pearson r correlation coefficients were calculated to assess reliability and validity.

An intraclass correlation coefficient((2,1)) of 0.72 (95% confidence interval, 0.33-0.90) was recorded for interrater reliability. The tool correctly identified the subject condition in 97% of the cases. The score did not correlate substantially with any clinical measures, with Pearson r ranging from 0.122 (P = .737) to 0.493 (P = .148).

This study showed that the proposed scale is a reliable tool and may be useful in discriminating patients with chronic LBP from pain-free controls. The poor correlation between the score and clinical measures may be due to the multidimensional nature of chronic LBP.