Interobserver and intraobserver reliability of the Walch classification in primary glenohumeral arthritis

In 1999, Walch et al introduced a novel classification scheme for glenoid morphology in patients with primary glenohumeral arthritis and reported substantial intraobserver and interobserver reliability. This classification system has been widely used by shoulder surgeons but a recent independent evaluation revealed considerable lower agreement. The goal of this study was to evaluate the reproducibility of the Walch classification.

Twenty-three consecutive patients (26 shoulders) undergoing total shoulder arthroplasty (TSA) or evaluated for TSA between March 2007 and November 2007 had shoulder CT scans performed and were included in this study. Three attending shoulder surgeons and 5 shoulder/sports medicine trained fellows independently and blindly evaluated CT scans of 26 consecutive patients with primary glenohumeral arthritis, and classified each patient according to the Walch classification to determine the interobserver reliability. The intraobserver reliability was assessed by comparison of the classification of each patient by the observers on 2 occasions separated by at least 6 weeks.

The overall interobserver agreement for all 8 observers was moderate (k=.508) for all Walch classes. The overall intraobserver reproducibility was substantial (k=.611).

We have shown that the interobserver reliability of the Walch classification is moderate while the intraobserver reliability is substantial. This is similar to or superior to the reliability of many commonly used orthopaedic classification systems. While the Walch classification system is not as reliable as initially suggested and improvement of this classification system would be of utility for future clinical studies, we have shown that this is an acceptable classification system and has good clinical and research applications.