Rate of adjudication of radiological progression in rheumatoid arthritis randomized controlled trials depending on preset limits of agreement: a pooled analysis from 15 randomized trials

Objective The aim of this study is to provide data on the adjudication rate for a predetermined threshold of difference in change score between two readers in randomized controlled trials (RCTs). Methods Fifteen datasets from RCTs in RA were scored by 13 experienced readers as pairs according to the modified Sharp-van der Heijde method. The theoretical adjudication rates for thresholds of between 3 and 20 units were calculated. We investigated the influence of the number of time points within the same session, the length of the interval and disease duration on the adjudication rates. Results A total of 21,295 time points from 7643 patients from 15 databases were included in the analysis. The adjudication rate was inversely related to the threshold. Higher adjudication rates were observed with a higher number of time points, longer time intervals and in early versus established RA. The adjudication rates ranged from 0% to 22% depending on the scenario. Conclusion With trained and experienced readers, the adjudication rate in RA RCTs is low even with very conservative adjudication thresholds.