Roentgenologic assessment of spondylolisthesis. I. A study of measurement variations.

Variations when measuring slip and lumbar index were studied in three lumbar specimens with different degrees of artificially created L5-S1 spondylolisthesis. Lateral radiographs were obtained with the specimens tilted and/or rotated mimicking an every-day examination situation. Slip and lumbar index were measured; slip was estimated according to a measuring method advocated by BOXALL et coll. (3) but slightly modified by us. Six radiologists evaluated each film independently. The magnitude of interobserver and intraobserver errors and variations due to positioning was found to be equally large--about 15 per cent each. Although the total measurement variation was not equal to the sum of errors involved, radiologic evaluation of degree of spondylolisthesis was found to be highly unreliable.