Measurement of vertebral rotation: Perdriolle versus Raimondi

SummaryThe measurement of vertebral rotation according to Perdriolle is widely used in the French-speaking and Anglo-American countries. Even in this measurement technique there may be a relatively high estimation error because of the not very accurate grading in steps of 5°. The measurement according to Raimondi seems to be easier to use and is more accurate, with 2° steps. The purpose of our study was to determine the technical error of both measuring methods. The apex vertebra of 40 curves on 20 anteroposterior (AP) radiographs were measured by using the Perdriolle torsion meter and the Regolo Raimondi. Interrater and intrarater reliability were computed. The thoracic Cobb angle was 43°, the lumbar Cobb angle 36°. The average rotation according to Perdriolle was 19.1° thoracic (SD 11.14), 12.7° lumbar (11.21). Measurement of vertebral rotation according to Raimondi showed an average rotation of 20.25° in the thoracic region (11.40) and 13.4° lumbar (10.92). The intrarater reliability was r=0.991 (Perdriolle) and r=0.997 (Raimondi). The average intrarater error was 1.025° in the Perdriolle measurement and 0.4° in the Raimondi measurement. Interrater error was on average 3.112° for the Perdriolle measurement and 3.630° for the Raimondi measurement. This shows that both methods are useful tools for the follow-up of vertebral rotation as projected on standard X-rays for the experienced clinicial. The Raimondi ruler is easier to use and is slightly more reliable.