Assessing precision and accuracy of false-profile hip radiographs.

Abstract

The purpose of this study was to assess the accuracy and precision of pelvic rotation in existing false-profile (FP) radiographs and to devise a method to improve accuracy and precision of FP radiographs. An imaging protocol was developed to obtain FP radiographs. Pelvic rotation was calculated using the described method for FP images obtained in the 3 months prior to and after implementation of this protocol. Student's t-test and variance ratio tests were used to determine differences in mean and variance of pelvic rotation between the 2 cohorts. Pelvic rotation calculation methodology was validated by using fluoroscopic C-arm to obtain AP and rotated images of 10 osteologic pelvises. The ratio of the distance between hip centres of each rotated image and AP image (WP/W) was determined. Intraclass coefficient correlation (ICC) was used to verify the relationship between WP/W and pelvic rotation. Mean WP/W was 0.47 (95% CI, 0.45-0.49). There were significant differences in mean pelvic rotation of the pre-protocol group (47.6°; 95% CI, 45.6-49.5°) and the post-protocol group (60.0°; 95% CI, 58.7-61.3°, p < 0.0001). Additionally, there was a significantly wider distribution of measurements in the pre-protocol group (SD = 7.9°) compared to the post-protocol group (SD = 5.7°, p = 0.0035). The quality of FP radiographs obtained in the clinical setting may be inconsistent. Standardising FP imaging produces more accurate images. Appropriate FP radiographs should have a distance between hip centres that is approximately 0.5 times the same distance found on an anteroposterior (AP) radiograph.