Background: The results of surgical treatment for femoroacetabular impingement have been increasingly reported, and more advanced intra-articular disease has been identified as an important predictive factor of outcome. Yet, the reliability of arthroscopic hip disease classification has not been well defined. Purpose/
Hypothesis: To determine the intraobserver and interobserver reliability of the Beck classification of labral and articular cartilage disease (anterior-superior acetabular rim) encountered in hip arthroscopy. Secondly, we identified the sources of poor reliability that may be improved with future disease classification schemes. Our hypothesis was that the Beck classification of labral and chondral lesions would demonstrate substantial reliability, while the differentiation of early forms of disease would be a common source of disagreement.
Study design: Cohort study; Level of evidence, 3.
Methods: Four experienced hip arthroscopists reviewed standardized arthroscopic videos of 40 cases. Arthroscopic findings at the anterior-superior acetabular rim were classified using the Beck classification of labral and articular cartilage disease. Repeat classification of videos was performed at least 2 weeks later. The reliability of arthroscopic classification was defined using the average weighted Cohen κ values and agreement rates.
Results: Arthroscopic classification of labral disease using the Beck classification demonstrated moderate to substantial interobserver reliability (average κ = .62; range, .48-.78) and an overall agreement rate of 81.7%. Intraobserver reliability showed a similar level of reliability (average κ = .65; agreement rate, 80.6%). The differentiation between labral degeneration and labral detachment was a common source of disagreement. Similarly, the Beck classification of articular cartilage disease had moderate to substantial interobserver reliability (average κ = .65; range, .49-.78) and overall agreement rate of 57.5%. Intraobserver reliability showed a slightly better level of reliability (average κ = .80; agreement rate, 77.5%). The differentiation between articular cartilage malacia and debonding was a common source of disagreement.
Conclusion: The arthroscopic classification of acetabular rim disease with the Beck classification has substantial interobserver reliability. This level of reliability is similar to previously reported arthroscopic disease classifications in the knee and shoulder and seems appropriate for future outcome reporting. Future classifications that eliminate common sources of disagreement may further improve the reliability.