Progression of rheumatoid arthritis on plain radiographs judged differently by expert radiologists and rheumatologists

Abstract

Objective. In a former study a panel of rheumatologists was used to assess which progression in radiological joint damage due to rheumatoid arthritis (RA) on hand and foot radiographs taken at one-year intervals was considered the minimally clinically important difference (MCID). We compare the judgments of the panel of rheumatologists with the judgments of 2 musculoskeletal radiologists. Methods. Two experienced musculoskeletal radiologists evaluated independently the same hand and foot radiographs as assessed by the panel of rheumatologists. Progression was defined as important if the radiologist would state it as substantial progression in their report. Two readers, different from the radiologists and rheumatologists, independently obtained the Sharp/van der Heijde scores. Receiver operating characteristic curve analyses were performed to quantify the minimally important progression defined by the radiologists expressed in Sharp/van der Heijde change-scores. The change-score with the highest accuracy represented the minimally important progression and was compared with the MCID defined by the panel of rheumatologists for 4 different settings (early versus advanced RA and mild versus high disease activity). Results. The minimally important progression defined by the radiologists was estimated at 6.5 Sharp/van der Heijde units. This was larger than the MCID defined by the panel of rheumatologists in 3 of the 4 clinical settings (3.0-4.5 units) and similar to the setting "advanced RA, mild disease activity." The panel of rheumatologists was inclined to change therapy in cases not reported as substantially progressive by the radiologists. The Sharp/van der Heijde progression scores of the radiographs on which the radiologists and rheumatologists disagreed related better with the rheumatologists' opinions. Conclusion. Changes that were riot regarded as substantial by the radiologists were judged clinically important by the rheumatologists in 3 of the 4 clinical settings. Thus, the radiologists appeared to be reserved in judging changes as important.