Analysis of dedicated sacroiliac views to improve reliability of conventional pelvic radiographs

Abstract

Objective. To compare the antero-posterior (AP) pelvis view with the Ferguson view of the SI joint in order to resolve whether one modality has a clear advantage for grading of sacroiliitis. Methods. One hundred and nine patients fulfilling Assessment of SpondyloArthritis international Society (ASAS) criteria for axial spondyloarthritis who had AP pelvis and Ferguson views on the same day were identified from an axial spondyloarthritis clinic registry. Two rheumatologists independently scored the AP pelvis and Ferguson views according to modified New York (NY) criteria. Intra- and inter-reader agreements were obtained for both evaluations by using the kappa statistic and intraclass correlation coefficient (ICC). Any change in diagnostic category dictated by the Ferguson vs the AP pelvis views was also evaluated. Results. A total of 266 radiographs were read from 109 patients. Intra-observer reliability of the observers showed similar ICC scores; this was also reflected in the kappa for diagnosis of AS fulfilling modified NY criteria between the observers. The inter-rater agreement showed similar kappa values between the two modalities. When separately evaluating SI joints with score grading of 0-2, grade 2 showed the lowest kappa, reaching a low of 0.1 and 0.19 for the right SI joint for Ferguson and AP pelvis views, respectively. Both modalities were concordant diagnostically; reclassification from AS to non-AS and vice versa was in the range 5-11%. Conclusion. There was general agreement between the Ferguson and AP pelvis X-ray ICC and kappa scores. Either modality can be employed to evaluate the SI joint for sacroiliitis with the Ferguson view showing no clear superiority over the standard AP pelvis view.