Reliability of three radiographic classification systems for knee osteoarthritis among observers of different experience levels

To assess the intra- and inter-observer reliability of three commonly referenced radiographic classification systems for knee osteoarthritis in a cohort of arthroplasty candidates.Pre-operative radiographs of 112 patients who subsequently underwent primary total knee arthroplasty were evaluated by four independent observers of varying experience. Each x-ray was de-identified, randomised, and classified according to the International Knee Documentation Committee, Kellgren-Lawrence, and Ahlbäck classifications. After a 2-week interval period, each x-ray was again randomised and re-classified.Regarding inter-observer reliability, the Ahlbäck and Kellgren-Lawrence classifications were shown to have 'substantial agreement' (AC 0.79 and 0.85 respectively), and the IKDC was shown to have 'almost perfect agreement' (AC 0.97). Regarding intra-observer reliability, the two more experienced observers demonstrated 'good' or 'excellent' reliability for all classification systems, and the two less experienced observers demonstrated 'moderate' intra-observer reliability for all classification systems.The International Knee Documentation Committee, Kellgren-Lawrence, and Ahlbäck radiographic classifications demonstrated adequate intra- and inter-observer reliability, supporting their potential implementation in surgical practice, or in epidemiological and clinical studies of knee osteoarthritis in a comparable cohort of patients. Clinical experience was positively correlated with intra-observer reliability. Whilst the International Knee Documentation Committee classification demonstrated the greatest reliability, this is likely due to its conservative definitions, and the Ahlbäck and Kellgren-Lawrence classifications are likely more reflective of the spectrum of disease severity encountered in an older patient cohort.