RESUMO
AIM: To estimate the effect of patients' axial rotation (AR) during pelvic radiograph acquisition, on the reliability and validity of sagittal pelvic parameters. MATERIALS AND METHODS: Lateral digitally reconstructed radiographs (LDRRs) were obtained from the pelvic computed tomography (CT) scans of eight children and nine adults. Then, the AR of the pelvis was simulated and the corresponding LDRRs were reconstructed at 5°, 10°, 15°, and 20° of the AR. Pelvic parameters were measured digitally on each radiograph. Intra- and interobserver variability were evaluated at each AR position (three operators repeated the measurements three times each). The bias on each clinical parameter, in each AR position, was calculated relatively to the 0° position. RESULTS: Interobserver variability increased similarly in children and adults with AR. It reached 4.4° for pelvic incidence and 4.7° for the sacral slope at 20° of AR. Biases on radiological parameters increased with AR and exceeded the acceptable threshold of errors when AR reached 10°. A linear regression was established (R2=0.834, p<0.0001) in order to estimate the AR of a patient on a lateral pelvic radiograph based on the measurement of the bifemoral distance normalized to the sagittal pelvic thickness. CONCLUSIONS: AR of patients during radiograph acquisition can be estimated in clinical practice, which would allow physicians to discard any radiographs where the calculated AR exceeded 10°.