Histopathology grading systems for characterisation of human knee osteoarthritis – reproducibility, variability, reliability, correlation, and validity

SummaryObjectiveTo determine the reliability, reproducibility, variability and validity of the Osteoarthritis Cartilage Histopathology (OACH) assessment system and Mankin Histological–Histochemical Grading System (HHGS) when applied to the characterisation of the osteoarthritic human knee.MethodOsteoarthritic knees of 10 patients undergoing unilateral knee arthroplasty were assessed, and assigned Kellgren–Lawrence and Line Drawing Atlas (LDA) radiology scores. The tibial plateaux were scored using the Modified Collins (MC) and Société Française d’Arthroscopie (SFA). Three observers twice scored both the OACH and HHGS across a single complete medial and lateral tibial plateau transect taken to include the region with the most severe osteoarthritis (OA) lesion. Intra- and inter-observer reliability, reproducibility, variability and validity were assessed, and the correlation between the two histopathology scoring systems was calculated.ResultBoth histopathology scoring systems were determined to be reliable and reproducible exhibiting similar variability, when applied to characterise OA specimens sampled from a well defined patient group with knee OA. A strong correlation between the mean OACH and mean HHGS scores was identified (Spearman’s ρ 0.980, P