LAUSR

INTRODUCTION The AO/OTA classification for diagnosing femoral trochanteric fractures (31A fractures) was revised in 2018. No studies have investigated whether the addition of CT to radiographic diagnosis improves the inter-rater reliability of classifying 31A fractures with the current AO/OTA criteria. The study aimed to test the hypothesis that the addition of three-dimensional CT (3D-CT) to radiographic diagnosis would improve diagnostic reliability. METHODS A retrospective review was conducted to assess the diagnostic reliability of classification of 31A fractures with current AO/OTA criteria. Radiographs and 3D-CT images from 89 cases were assessed. Major fracture types (A1, A2, and A3) and subgroups were diagnosed by nine orthopedic surgeons who were classified into three groups (high-, intermediate-, and low-experience) according to their clinical experience. Anterior-posterior and lateral radiographs were provided to diagnose fracture type (first assessment). After a 6-week interval, radiographs and 3D-CT images of all cases were evaluated by each rater (second assessment). Fleiss's Kappa was used to determine inter-rater reliability. RESULTS In the first assessment, the Kappa value indicated fair inter-rater reliability in all groups (high-experience group: κ = 0.296, 95% confidence interval [CI] 0.239-0.352; intermediate-experience group: κ = 0.367, 95% CI 0.305-0.428; low-experience group: κ = 0.304, 95% CI 0.246-0.362). With the addition of 3D-CT (second assessment), reliability improved from fair to moderate in the high- and intermediate-experience groups (κ = 0.483, 95% CI 0.428-0.539 and κ = 0.409, 95% CI 0.352-0.466, respectively). By contrast, reliability remained fair in the low-experience group (κ = 0.322, 95% CI 0.322-0.431). The inter-rater reliability of diagnosing subgroup fracture types improved for A2.3 and A3.1 in all three groups and for A3.2 and A3.3 in the intermediate- and low-experience groups. CONCLUSION The current AO/OTA classification revised in 2018 provided fair reliability in diagnosing femoral trochanteric fractures in all three surgeon groups. The addition of 3D-CT to radiographic image evaluation improved reliability in high- and middle-expertise groups. The addition of 3D-CT to radiographic evaluation often improved the diagnostic reliability for unstable fractures, although there was some variation among fracture subgroups.