Abstract The Durban classification system for juvenile idiopathic arthritis (JIA) is reviewed in a historical context and in association with a review of problems that have become apparent. These include: (i) a family history of psoriasis as an exclusion factor (oligoarthritis and enthesitis-related arthritis) and as in inclusive factor (psoriatic arthritis); (ii) a family history of HLA-B27-associated disease as an exclusion factor (oligoarthritis) and an inclusive factor (enthesitis-related arthritis); (iii) the requirement of a dermatological opinion for psoriasis; (iv) the absence of HLA-B27 antigen in proband, with the presence of antigen in the family history; (v) the definition of time of onset; (vi) the presence of rheumatoid factor (RF) with oligoarthritis; (vii) HLA-B27-positive males with an onset of arthritis after 8 years of age. Modifications are suggested to maintain the homogeneity of groupings for research, whilst providing a practical scheme for clinicians. Three main modifications are suggested. (A) That the family history be included in descriptors rather than as inclusive or exclusion criteria. (B) Further development of the hierarchical system, which is partly used in the Durban classification. (C) That the following changes be made: rheumatoid-factor-positive oligoarthritis and polyarthritis be classified together; extended oligoarthritis and polyarthritis be classified together; HLA-B27-positive disease be classified with fewer inclusive and exclusion criteria; the criteria for psoriatic arthritis be modified; the classification of the disease in a particular child be changed in the event of relevant changes in the child's disease or laboratory profile. These suggestions are made to stimulate discussion.
