The clinical utility of computed tomography compared to conventional radiography in diagnosing sacroiliitis. A retrospective study on 910 patients and literature review.

OBJECTIVE: Ankylosing spondylitis (AS) is a progressive, debilitating disease with complex symptoms, unclear etiology and pathogenesis, and difficult diagnosis. Current imaging methods are useful in diagnosing AS and other spondyloarthropathies, and are frequently used in investigations of sacroiliitis. The radiographic diagnosis of sacroiliitis has large interobserver variations. Computed tomography (CT) has been used for evaluation of sacroiliitis since 1979, and has been evaluated in several studies, most of them with a limited number of patients. These studies have shown a large number of false-negative results from radiography.
METHODS: In a retrospective study of clinical data, we evaluated 910 patients with AS who were examined by radiography and CT within a 2-year period. The reported outcomes from radiography and CT were compared.
RESULTS: The agreement between radiography and CT data was only fair, with a kappa value of 0.2418. There were 35.0% false-positive radiography reports, 22.5% false-negative radiography reports, and 86.0% false-equivocal radiography reports. In total, 41.3% of all radiological reports gave a false answer. While the number of false negatives was similar to that previously reported, the number of false positives was much higher than previously reported, and is probably similar to everyday radiology reporting.
CONCLUSION: Our results indicate that the clinical utility of radiography for evaluation of sacroiliitis is limited. The high rate of inaccurate results should motivate the use of sectional imaging for its superior performance.