Development and validation of a new radiographic scoring system to evaluate bone and cartilage destruction and healing of large joints with rheumatoid arthritis: ARASHI (Assessment of rheumatoid arthritis by scoring of large joint destruction and healing in radiographic imaging) study

Objective: To assess the reliability and sensitivity of a novel scoring method to evaluate the radiographic appearance of and longitudinal changes including joint remodeling in large joints with early and established rheumatoid arthritis (RA). Methods: The ARASHI study group devised new radiographic scoring systems (Status score; range 0-16 points, and Change score; range -11 to 12 points) for evaluation of large joints with RA. Radiographs showing anterior/ posterior views of large joints (shoulder, elbow, hip, knee, and ankle joints) taken at two time points (mean interval 2.3 years) were collected from 25 patients with established RA (5 patients for each of the 5 joints, 50 films in total), and an additional 5 films of each joint with severe joint destruction were collected from 5 different sets of RA patients. After consensus on the definition of each component and reader training, images were evaluated using the Larsen's grading system and the ARASHI Status and Change score by 9 independent senior orthopedic surgeons. The reliability was estimated by intra-class correlation coefficients (ICCs) and measurement error by 95 % confidence intervals of minimum detectable change (MDC₉₅). Results: ARASHI Status score and Change score significantly correlated with Larsen's grade (r = 0.89, P <0.0001) and follow-up-baseline differences in Larsen's grade (r = 0.83, P <0.0001), respectively. Interreader ICCs were very high for both Status score (0.88, 95 % confidence interval [CI], 0.83-0.92, P <0.001) and Change score (0.92, 95 % CI, 0.87-0.96, P <0.001). Intrareader ICCs were also very high for both Status score (0.92, 95 % CI, 0.71-0.98, P <0.001) and Change score (0.97, 95 % CI, 0.91-0.99, P <0.001). The MDC95 for inter-reader agreement were 4.18 (25 % of maximum obtainable score, MOS) and 4.99 (21 % of MOS) for Status score and Change score, respectively. The MDC₉₅ for intra-reader agreement was acceptable with 2.82 (17 % of MOS) and 3.02 (13 % of MOS) for Status score and Change score, respectively. Conclusion: The ARASHI scoring method showed good inter-<intra-reader reliability with high ICCs and acceptable MDC₉₅ with respect to each large joint and the components of both Status and Change scores. The results suggest that the ARASHI scoring method might be useful for the assessment of status, as well as longitudinal monitoring of destruction and remodeling of large joints with RA.