Interobserver Reliability of the International Society of Arthroscopy, Knee Surgery and Orthopaedic Sports Medicine (ISAKOS) Classification of Meniscal Tears

Background: Consistency of arthroscopic evaluation and documentation in meniscal tears between investigators is essential to the validity of multicenter studies. A group of experts developed a classification of meniscal tears that may be used internationally. Hypothesis: The International Society of Arthroscopy, Knee Surgery and Orthopaedic Sports Medicine (ISAKOS) classification of meniscal tears provides sufficient interobserver reliability for pooling of data from international clinical trials designed to evaluate the outcomes of treatment for meniscal tears. Study Design: Cohort study (diagnosis); Level of evidence, 1. Methods: A pilot study was performed by having 8 members of the committee grade 10 arthroscopic videos for classification of tear depth, rim width, location, tear pattern, and quality of the tissue. The results of the pilot study were used to change the instruction sheet and evaluation form. International interobserver reliability was determined by having 8 orthopaedic surgeons who practice in different countries evaluate 37 arthroscopic videos selected to represent different meniscal tear characteristics. The Spearman rho correlation coefficient was used to compare the area of the meniscus excised, as drawn on the diagram, with the numeric percentage of meniscus excised. Results: There was an 87% agreement for anterior-posterior location of the tear (kappa=.65); 79% agreement for tear pattern (kappa=.72); 88% agreement for tear depth (kappa=.52); 68% agreement for anterior, middle, and posterior location of the tear (kappa=.46); and 72% agreement for tissue quality (kappa=.47). There was 54% agreement for the rim width (kappa=.25) and 67% agreement if the tear was central to the popliteal hiatus (kappa=.36). Based on the Landis and Koch criteria for kappa coefficients, there was substantial agreement for anterior-posterior location of the tear and tear pattern; moderate agreement for tear depth, anterior, middle, and posterior location of the tear, and tissue quality; and fair agreement for rim width and if the tear was central to the popliteal tear. Interobserver reliability based on the intraclass correlation coefficient (ICC) was good for tear length (ICC=.83) and moderate for percentage of meniscus that was excised (ICC=.65). The mean rho for all raters was .92 (95% confidence interval [CI],.89-.94) comparing the values for percentage of meniscus excised with the area on the diagrams. Conclusion: The ISAKOS classification of meniscal tears provides sufficient interobserver reliability for pooling of data from international clinical trials designed to evaluate the outcomes of treatment for meniscal tears.