Lessons from an international survey of paper cases of 10 real patients from an early arthritis clinic

Objective. To determine how experts would classify 10 early-arthritis cases (7 atypical) and to study discrepancies in diagnoses relative to ACR crite ria for rheumatoid arthritis (RA) or ESSG criteria for spondyloarthropathy (SpA). Methods. Ten real cases (5 met ACR criteria for RA, 6 ESSG criteria for SpA , 3 both and 2 neither) followed for 28.5 +/- 4.8 months were sent as paper cases to 20 international and 12 French experts. Each expert selected a di agnosis among 8 possible choices and rated it on a 0-10 confidence scale. F or each case, 3 analog scales (0-100 mm) were used to indicate the probabil ity of RA, SpA or undifferentiated arthritis (UA). Results. Experts often disagreed about diagnoses (up to 5 different diagnos es for a given case, with a mean of 3.9 per case). Similarly, expert opinio ns on probabilities for RA and SpA differed widely, with great overlap betw een confidence for RA, SpA and UA. Fulfilment of ACR or ESSG criteria was p oorly related to the experts' diagnosis and evaluation of probabilities for RA and SpA. However, UA was a relatively infrequent choice (19%). Conclusions. There was no general consensus about the nosology of early RA and SpA. Classification of atypical early arthritis was not resolved by cur rently available criteria far RA and SpA. This may have implications for th erapy in early disease.