Objective: There is no agreement on how to classify and diagnose reactive arthritis (ReA) and it is also unclear what kind of specific clinical and laboratory investigations are appropriate. We define relevant points of agreement and identify points of disagreement among an international group of experts in the field.

Methods: Prior to the 4th International Workshop on Reactive Arthritis, Berlin, July 1999, we sent questionnaires to 42 experts identified by personal knowledge and recent publications.

Results: The response rate was 81% (n = 34). There was agreement on the nomenclature and recommendation to use the term "reactive arthritis" only if the clinical picture and the microbes involved are HLA-B27 and spondyloarthropathy (SpA) associated, whereas the term "infection related arthritis" is used for all other arthritides related to or associated with infections. A differentiation between acute and chronic ReA with a cutoff of 6 months is recommended. The history of a preceding symptomatic infection is thought to be most relevant for a diagnosis of ReA. The minimal interval between preceding symptoms and arthritis is proposed to be 1-7 days, maximally 4 weeks. The joint pattern in ReA is asymmetrical, with predominance of the lower limbs. SpA related symptoms may contribute to the diagnosis. A search for chlamydia in urine/urethra/cervix is recommended, while in the case of diarrhea enterobacteria should be searched for in stool and antibodies against them in serum. There were also areas of disagreement, such as: Is arthritis essential for the diagnosis of ReA?, Is it oligoarthritis or any arthritis?, What are the role and value of polymerase chain reaction investigation?, The role and value of serology?, Is the diagnostic sensitivity of microbiological tests for ReA increased by HLA-B27 determination?

Conclusion: The points of agreement will support better communication in this area, and clarification of the disagreements will lead to further studies and discussion.