Plain X-ray is insufficient for correct diagnosis of tibial shaft spiral fractures: a prospective trial

Abstract

Abstract Purpose Tibial shaft spiral fractures and fractures of the distal third of the tibia (AO:42A/B/C and 43A) frequently occur with non-displaced posterior malleolus fractures (PM). This study investigated the hypothesis that plain X-ray is not sufficient for a reliable diagnosis of associated non-displaced PM fractures in tibial shaft spiral fractures. Methods 50 X-rays showing 42A/B/C and 43A fractures were evaluated by two groups of physicians, each group was comprised of a resident and a fellowship-trained traumatologist or radiologist. Each group was tasked to make a diagnosis and/or suggest if further imaging was needed. One group was primed with the incidence of PM fractures and asked to explicitly assess the PM. Results Overall, 9.13/25 (SD ± 5.77) PM fractures were diagnosed on X-ray. If the posterior malleolus fracture was named or a CT was requested, the fracture was considered “detected”. With this in mind, 14.8 ± 5.95 posterior malleolus fractures were detected. Significantly more fractures were diagnosed/detected (14 vs. 4.25/25; p < 0.001/14.8 vs. 10.5/25; p < 0.001) in the group with awareness. However, there were significantly more false positives in the awareness group (2.5 vs. 0.5; p = 0.024). Senior physicians recognized slightly more fractures than residents (residents: 13.0 ± 7.79; senior physicians: 16.5 ± 3.70; p = 0.040). No significant differences were demonstrated between radiologists and trauma surgeons. The inner-rater reliability was high with 91.2% agreement. Inter-rater reliability showed fair agreement (Fleiss-Kappa 0.274, p < 0.001) across all examiners and moderate agreement (Fleiss-Kappa 0.561, p < 0.001) in group 2. Conclusion Only 17% of PM fractures were identified on plain X-ray and awareness of PM only improved diagnosis by 39%. While experiencing improved accuracy, CT imaging should be included in a comprehensive examination of tibial shaft spiral fractures. Level of evidence II. Diagnostic prospective cohort study. Trail registration number DRKS00030075.