Inter- and intra-observer reliability of a novel Lisfranc injury classification in comparison to the Myerson classification

Aim: We investigated the inter- and intra-observer reliability of the novel Column Involvement Severity System (CISS) in comparison to the commonly used Myerson classification for Lisfranc injuries. Methods: Sixteen observers (orthopedic or trauma surgeons with experience in foot and ankle trauma) rated radiographs and computed tomography videos of 31 cases of patients treated at our level 1 trauma center between 2010 and 2020. Four weeks after completing the first rating session, eight observers (50%) rated the cases again. The inter- and intra-observer reliability for both classifications were assessed using intraclass correlation coefficients (ICC). Results: ICC for Myerson classification was 0,277 (0,056–0,944) for interobserver and 0,505 (0,390–0,604) for intra-observer reliability. However, the Myerson classification was found to be non-applicable by at least 25% of observers in 48,4% of cases (range 0–63%). For the CISS classification, overall ICC was 0,455 (moderate) and 0,676 (substantial) for interobserver and intra-observer reliability respectively. Conclusion: The Myerson classification was found to be non-applicable by a large number of observers. The CISS classification demonstrated better interobserver and intra-observer reliability. It may thus provide better communication about the injury severity and involvement of the three midfoot columns in Lisfranc injuries.