A radiological measure of shoulder subluxation in hemiplegia: its reliability and validity

This study describes the reliability and validity of a new radiological method to measure shoulder subluxation that is based on taking a single x-ray of the affected shoulder in the plane of the scapula. A digitizer and a computer are used to quantify the measurements directly from the x-ray. One single radiological view of the shoulder taken in a standardized position provides four measures of subluxation including a measure of vertical subluxation, horizontal subluxation, a normalized distance sensitive to both vertical and horizontal subluxation, and an angular measure. A total of 72 volunteers with cerebrovascular accidents participated in the study. Thirty-six subjects had a clinical shoulder subluxation and 36 had no clinical subluxation. The construct validity analysis showed significant differences (p < .001) between the mean scores for subluxed and nonsubluxed groups on all four measures of subluxation. The concurrent validity of the x-ray measures in relation to external clinical measures (finger's breadth, measures in centimeters with calipers or a plexiglass jig) was moderate. Three measures were found to be reliable and valid whereas the fourth, a measure of horizontal subluxation has lower reliability. However, it does have clinical significance. The clinical and research implications of this study are presented.