School screening for scoliosis: methodologic considerations. Part 1: External measurements.

In order to determine the applicability of school screening techniques for scoliosis, a methodologic survey was carried out within the framework of a school screening project. The accuracy of the measurements of rib hump height, angle of trunk rotation, and of moiré topography was investigated by assessing the intraobserver and interobserver variation. The validity of these techniques was tested by comparing their outcome to the Cobb angle. Intraobserver variation was measured over both short and long time intervals. The interobserver variation was determined among two orthopaedic surgeons and among a group of six doctors. Results are expressed in a Spearman correlation coefficient and a standard deviation. The Spearman correlation ranges from 0.46 (moiré) to 0.75 (rib hump height) in intraobserver variation, and from 0.60 (rib hump height) to 0.70 (angle of trunk rotation) in interobserver variation. The standard deviations illustrate the interobserver range of the measurements, for rib hump height, 3.7 mm, for rotation, 2.3 degrees, and for moiré, 0.7 lines. The validity of the three methods varied from 0.40 to 0.53 as correlated with the angle of Cobb. The conclusion is that these methods can be applied in school screening techniques, but that they do not allow a sharp distinction between normal and pathologic cases. Instead, it is preferable to define the borderline in terms of a danger zone rather than a strict single value. The danger zone for the rib hump height should be 5-10 mm, for the rotation 3-7 degrees, and for the moiré topography 1-3 lines. Recordings in these zones should be repeated within a few months.