Decking Ralf, Brunner Alexander, Decking Jens, Puhl Wolfhart, Günther Klaus-Peter
Orthopaedic Department, RKU, University of Ulm, Ulm, Germany.
Skeletal Radiol. 2006 May;35(5):282-7. doi: 10.1007/s00256-005-0061-5. Epub 2006 Mar 14.
To assess the inter-observer and intra-observer reliability of two commonly used radiographic classification systems in the evaluation of hip dysplasia in skeletally mature adults.
Three observers with different levels of training independently classified 62 dysplastic hips on 51 standard anteriorposterior pelvis radiographs according to the criteria defined by Crowe and by Hartofilakidis. To assess intra-observer reliability, the same radiographs were reviewed 3 months later by the same observers.
At the time of the radiographic examination, the mean age of the 51 patients had been 54 years (range 18-82 years).
A high correlation concerning the inter- and intra-observer reliability of both systems was demonstrated. Inter-observer reliability displayed a weighted kappa coefficient of 0.82 for the Crowe and 0.75 for the Hartofilakidis classification. Intra-observer reliability showed a kappa coefficient of 0.86 and 0.79, respectively.
Both classification systems can be recommended to compare collectives of adult patients with congenital dysplasia of the hip. However, for future clinical practice, it would be advisable to agree on one universally accepted system as a standard in the literature.
评估两种常用的影像学分类系统在评估骨骼成熟成年人髋关节发育不良时观察者间和观察者内的可靠性。
三名接受不同程度培训的观察者根据Crowe和Hartofilakidis定义的标准，对51张标准骨盆前后位X线片上的62个发育不良髋关节进行独立分类。为评估观察者内可靠性，相同的观察者在3个月后对相同的X线片进行复查。
在进行影像学检查时，51例患者的平均年龄为54岁（范围18 - 82岁）。
两种系统在观察者间和观察者内的可靠性方面均显示出高度相关性。观察者间可靠性方面，Crowe分类的加权kappa系数为0.82，Hartofilakidis分类为0.75。观察者内可靠性方面，kappa系数分别为0.86和0.79。
两种分类系统均可用于比较成年先天性髋关节发育不良患者群体。然而，对于未来的临床实践，在文献中就一种普遍接受的系统作为标准达成共识是可取的。
