Interobserver correlation in classification of bone loss in total knee arthroplasty

Objective: Considering the difficulty for classifying bone losses the present study was designed to analyze if the AORI classification based on pre-operative radiographies is consistent and reproducible between different orthopedic surgeons. Methods: Six orthopedists specialized in knee surgery were trained for the use of the classification based on radiographic evaluation. All the surgeons individually classified 26 pre operative knee radiographs. Results: There was a moderate (> 50%) matching of the classification in 24 of 26 cases in the femur and 22 of 26 in the tibia. A good matching (> 80%) was present in 12 of 26 cases in the femur and in 7 of 26 cases in the tibia. Conclusion: We observed that the AORI classification presented a moderate radiographic correlation between surgeons. Evidence of level III, Study of nonconsecutive patients; without consistently applied reference "gold" standard.