Interrater agreement for characterization of capitellar osteochondritis dissecans using photon-counting computed tomography technology

Background Imaging in children with capitellar osteochondritis dissecans (COCD) is crucial for surgical decision-making but remains challenging. Novel photon-counting computed tomography (PCCT) technology enables superior resolution while reducing radiation dose. This study aims to describe interrater agreement (IRA) for COCD characteristics assessed with PCCT and to make a comparison with conventional CT. Methods At a tertiary referral hospital, anonymized PCCT and conventional CT scans of COCD lesions were assessed systematically and independently by 2 experienced clinicians and 2 musculoskeletal radiologists, excluding postoperative scans. The intraclass correlation coefficient (ICC) with 95% confidence intervals was used to describe IRA among all raters. Results In the PCCT group (n = 29), ICCs for loose body count, lesion size in the sagittal plane, presence of an empty defect and presence of an osseous bridge were 0.75 (95% CI: 0.61–0.86), 0.68 (0.51–0.81), 0.64 (0.47–0.78) and 0.60 (0.43–0.76), respectively. ICCs of physeal status, lateral wall involvement, fragmentation, depth and tilting had a lower bound of the 95% confidence interval below 0.4. In the conventional CT group (n = 12), ICCs for loose body count and osseous bridging were statistically significantly lower than in the PCCT group after adjustment for multiple testing (p = 0.009 and p=0.025, respectively). Conclusion PCCT assessment of COCD demonstrates at least substantial IRA for loose body count and at least moderate IRA for lesion size and the presence of an empty defect or bony bridge. Moreover, PCCT may enable higher IRA than conventional CT.