Wæver Daniel, Madsen Mette Lund, Rölfing Jan Hendrik Duedal, Borris Lars Carl, Henriksen Mads, Nagel Lise Loft, Thorninger Rikke
Department of Orthopaedics, Regionshospitalet Randers, Randers, Denmark; Department of Orthopaedics, Aarhus University Hospital, Aarhus, Denmark.
Department of Orthopaedics, Regionshospitalet Randers, Randers, Denmark.
Injury. 2018 Jun;49 Suppl 1:S29-S32. doi: 10.1016/S0020-1383(18)30299-7.
Traditionally, distal radius fractures (DRFs) have been described using eponyms, e.g. Colles, Smith, Barton, Chauffeur. During the last half of the 20 century several classification systems for DRF have emerged. We evaluated the inter- and intra-observer agreement of the AO/OTA, Frykman and Older classification systems.
Four observers, an intern, an orthopaedic registrar, an orthopaedic consultant and a radiology consultant, independently evaluated DRF radiograms and classified the fractures according to the AO/OTA, Frykman and Older classification systems. After an interval of 6 months, radiograms of 30 randomly chosen patients were re-evaluated by the same observers.
Radiograms of 573 DRF patients were evaluated in the study. The inter-observer reliability of the AO/OTA fracture types (A, B and C) was 'weak' (kappa = 0.45). The agreement dropped to 'minimal' (kappa = 0.24) regarding the AO/OTA groups (A2, A3, B1, B2, B3, C1, C2 and C3). The reliability of the Frykman classification system was 'weak' (kappa = 0.41), and we observed the lowest inter-observer reliability for the Older classification system (kappa = 0.10). The kappa values for the intra-observer reproducibility of the AO/OTA fracture types (A, B and C) ranged from 0.58 to 0.87. For the AO/OTA groups (A2, A3, B1, B2, B3, C1, C2 and C3) the reproducibility was lower ranging from 'minimal' to 'weak'. The intra-observer reproducibility of the Frykman system was 'weak' to 'moderate' and even worse for the Older classification system.
Based on these findings the AO/OTA classification system seems to be most reliable for routine use, however, with lower kappa values concerning the agreement for the groups. The Frykman and Older classification systems cannot be recommended because of less convincing results.
传统上，桡骨远端骨折（DRF）一直使用人名来描述，例如科莱斯骨折、史密斯骨折、巴顿骨折、司机骨折。在20世纪后半叶，出现了几种针对DRF的分类系统。我们评估了AO/OTA、弗赖克曼和奥尔德分类系统在观察者间和观察者内的一致性。
四名观察者，一名实习生、一名骨科住院医师、一名骨科顾问和一名放射科顾问，独立评估DRF的X线片，并根据AO/OTA、弗赖克曼和奥尔德分类系统对骨折进行分类。间隔6个月后，相同的观察者对30名随机选择患者的X线片进行重新评估。
本研究共评估了573例DRF患者的X线片。AO/OTA骨折类型（A、B和C）在观察者间的可靠性为“弱”（kappa = 0.45）。对于AO/OTA组（A2、A3、B1、B2、B3、C1、C2和C3），一致性降至“极低”（kappa = 0.24）。弗赖克曼分类系统的可靠性为“弱”（kappa = 0.41），我们观察到奥尔德分类系统在观察者间的可靠性最低（kappa = 0.10）。AO/OTA骨折类型（A、B和C）在观察者内可重复性的kappa值范围为0.58至0.87。对于AO/OTA组（A2、A3、B1、B2、B3、C1、C2和C3），可重复性较低，范围从“极低”到“弱”。弗赖克曼系统在观察者内的可重复性为“弱”到“中等”，而奥尔德分类系统的情况更差。
基于这些发现，AO/OTA分类系统似乎最适合常规使用，然而，对于组间一致性的kappa值较低。由于结果缺乏说服力，不推荐使用弗赖克曼和奥尔德分类系统。
