Verhage S M, Rhemrev S J, Keizer S B, Quarles van Ufford H M E, Hoogendoorn J M
Department of Surgery, MC Haaglanden, The Hague, The Netherlands,
Skeletal Radiol. 2015 Oct;44(10):1435-9. doi: 10.1007/s00256-015-2179-4. Epub 2015 Jun 9.
Classification of malleolar fractures is a matter of debate. In the ideal situation, a classification system is easy to use, shows good inter- and intraobserver agreement, and has implications for treatment or research.
Interobserver study. Four observers distributed 100 X-rays to the Weber, AO and Lauge-Hansen classification. In case of a trimalleolar fracture, the size of the posterior fragment was measured. Interobserver agreement was calculated with Cohen's kappa. Agreement on the size of the posterior fragment was calculated with the intraclass correlation coefficient.
Moderate agreement was found with all classification systems: the Weber (K = 0.49), AO (K = 0.45) and Lauge-Hansen (K = 0.47). Interobserver agreement on the presence of a posterior fracture was substantial (K = 0.63). Estimation of the size of the fragment showed moderate agreement (ICC = 0.57).
Classification according to the classical systems showed moderate interobserver agreement, probably due to an unclear trauma mechanism or the difficult relation between the level of the fibular fracture and syndesmosis. Substantial agreement on posterior malleolar fractures is mostly due to small (<5 %) posterior fragments. A classification system that describes the presence and location of fibular fractures, presence of medial malleolar fractures or deep deltoid ligament injury, and presence of relevant and dislocated posterior malleolar fractures is more useful in the daily setting than the traditional systems. In case of a trimalleolar fracture, a CT scan is in our opinion very useful in the detection of small posterior fragments and preoperative planning.
踝关节骨折的分类存在争议。在理想情况下，分类系统应易于使用，具有良好的观察者间和观察者内一致性，并对治疗或研究有指导意义。
观察者间研究。四名观察者对100张X线片进行Weber、AO和Lauge-Hansen分类。对于三踝骨折，测量后踝骨折块的大小。采用Cohen's kappa计算观察者间一致性。采用组内相关系数计算后踝骨折块大小的一致性。
所有分类系统的一致性均为中等：Weber分类（K = 0.49）、AO分类（K = 0.45）和Lauge-Hansen分类（K = 0.47）。观察者间对后踝骨折存在情况的一致性较高（K = 0.63）。骨折块大小的估计显示一致性中等（ICC = 0.57）。
根据经典系统进行分类的观察者间一致性为中等，可能是由于创伤机制不明确或腓骨骨折水平与下胫腓联合之间的关系复杂。后踝骨折的高度一致性主要归因于较小（<5%）的后踝骨折块。在日常临床工作中，一种描述腓骨骨折的存在和位置、内踝骨折或深层三角韧带损伤的存在以及相关和移位的后踝骨折的存在的分类系统比传统系统更有用。对于三踝骨折，我们认为CT扫描对于检测小的后踝骨折块和术前规划非常有用。
