Redundant radiography before CT in axial spondyloarthritis: high radiation burden, limited diagnostic value

The objectives of this study were to quantify the number and cumulative radiation dose of pelvic and lumbar radiographs obtained before sacroiliac CT confirmed diagnosis of radiographic axial SpA (r-axSpA), and to assess potential dose reduction if CT had been used instead of repeated radiographs.

Pelvic and lumbar radiographs performed prior to CT were identified from institutional and national databases. Literature-derived effective dose estimates were applied to calculate minimum- and maximum-bound cumulative radiographic doses per patient. Hypothetical scenarios in which diagnosis was assumed based on a single low-dose CT (0.5 mSv) or conventional CT (4-10 mSv) were compared. Paired Wilcoxon signed-rank tests were used to compare cumulative X-ray doses with CT benchmarks.

A total of 235 radiographs were performed before CT (mean 2.47 per patient, range 0-8). Most patients (77.9%) had ≥2 combined radiographs, often clustered into multiple imaging sessions. The minimum-bound cumulative X-ray dose ranged from 0.00 to 1.60 mSv (median 0.40 mSv), and the maximum dose ranged from 0.20 to 11.20 mSv (median 2.80 mSv). Under minimum assumptions, radiographic exposure was similar to low-dose CT (P = 0.307), with 10.5% ≥0.5 mSv. Under maximum assumptions, exposure was markedly higher (median +2.30 mSv, P < 0.001; 82.1% ≥0.5 mSv). Compared with a fixed 4.0 mSv conventional CT benchmark, no patients in the minimum scenario and 7 (7.4%) in the maximum scenario exceeded threshold.

In CT-confirmed r-axSpA, multiple radiographs are frequently obtained before definitive imaging, adding substantial radiation without improving diagnostic yield, whereas a single appropriately indicated low-dose CT could achieve structural assessment with less redundant exposure.