Interrater and intrarater reliability of arthroscopic measurements of articular cartilage defects in the knee

Background: Cartilage lesions of the knee are difficult to treat. Lesion size is a critical factor in treatment algorithms, and the accurate, reproducible sizing of lesions is important. In this study, we evaluated the interrater and intrarater reliability of, and correlations in relation to, various arthroscopic sizing techniques. Methods: Five lesions were created in each of 10 cadaveric knees (International Cartilage Repair Society grade 3C). Three orthopaedic surgeons used 4 techniques (visualization and use of a 3-mm probe, a simple metal ruler, and a sliding metallic ruler tool) to estimate lesion size. Repeated-measures data were analyzed using a mixed-effect linear model. The differences between observed and gold-standard (plastic mold) values were used as the response. Intraclass and interclass correlation coefficient (ICC) values for intrarater and interrater reliability were computed, as were overall correlation coefficients between measurements and gold standards. Results: The mean lesion size was 2.37 cm2 (range, 0.36 to 6.02 cm2). Rater, lesion location and size, and measurement method all affected the cartilage defect measurements. Surgeons underestimated lesion size, and measurements of larger lesions had a higher percentage of error compared with those of smaller lesions. When compared with plastic molds of lesions, 60.5% of surgeon measurements underestimated lesion size. Overall, the correlation between measurements and gold standards was strongest for the simple metal ruler method and weakest for the visualization method. Conclusions: Several factors may influence arthroscopic estimation of cartilage lesion size: the lesion location, measurement tool, surgeon, and defect size itself. The intrarater and interrater reliability was moderate to good using a 3-mm probe, sliding metallic ruler tool, or simple metal ruler and was fair to moderate using visualization only. Clinical Relevance: There is a need for more accurate methods of determining the size of articular cartilage lesions.