OBJECTIVE: (i) To determine the efficacy of the Durban classification for children with juvenile idiopathic arthritis (JIA) where < 5 joints were involved at onset (with systemic arthritis excluded) by determining the proportion of the cohort that proved to be "unclassifiable"; (ii) to define reasons for cases being "unclassifiable," particularly regarding family history; and (iii) to compare the efficacy of a proposed hierarchical system (an unofficial modification of the Durban classification) with the Durban classification, where family history details are included as descriptors, rather than as classification criteria. METHODS: Charts were reviewed of 50 children with fewer than 5 joints involved at presentation for JIA, followed for at least 12 months, with systemic arthritis excluded. Cases were classified according to the EULAR criteria, the Durban criteria, and by a proposed "modified Durban" classification subject to hierarchy, with exclusions in the following order: systemic arthritis, rheumatoid factor (RF) positive arthritis, psoriasis or a combination of dactylitis and psoriatic nail changes (psoriatic arthritis), and HLA-B27 positive arthritis (enthesitis related arthritis), with the remainder of children being classified as having either RF negative polyarthritis or RF negative oligoarthritis, depending on number of joints involved, with additional information noted as descriptors. The "modified Durban" classification was proposed only to stimulate discussion among clinicians. RESULTS: Of 50 children, 56% were "unclassifiable" by the Durban classification, mainly because of inadequate family history despite appropriate questioning. Using the proposed "modified Durban" classification, 2% were "unclassifiable." Family history was classified as inadequate for the following reasons: The parents did not know family history; the child or parent was adopted; the father was unknown or parent died early; parents never attended; extended family had lost communication with parents; or a relative was considered to have psoriasis, but not confirmed by dermatologists. Other reasons for "unclassifiable" included: dermatologists unable to confirm psoriasis; family history of inflammatory bowel disease and sacroiliitis but B27 status unknown; proband B27 negative but family history of B27-related disease; family history of psoriasis, but patient had insufficient criteria for psoriatic arthritis and therefore excluded from oligoarthritis, psoriatic arthritis and other groups. CONCLUSION: (i) The Durban classification showed poor efficacy for JIA where < 5 joints were involved at onset, with more than half the cases being "unclassifiable". (ii) The most common reason was that appropriate family history was not available despite being sought by the clinician. (iii) A proposed hierarchical system, an unofficial modification of the Durban classification, showed good efficacy, with only one of 50 cases being "unclassifiable."
OBJECTIVE: (i) To determine the efficacy of the Durban classification for children with juvenile idiopathic arthritis (JIA) where < 5 joints were involved at onset (with systemic arthritis excluded) by determining the proportion of the cohort that proved to be "unclassifiable"; (ii) to define reasons for cases being "unclassifiable," particularly regarding family history; and (iii) to compare the efficacy of a proposed hierarchical system (an unofficial modification of the Durban classification) with the Durban classification, where family history details are included as descriptors, rather than as classification criteria. METHODS: Charts were reviewed of 50 children with fewer than 5 joints involved at presentation for JIA, followed for at least 12 months, with systemic arthritis excluded. Cases were classified according to the EULAR criteria, the Durban criteria, and by a proposed "modified Durban" classification subject to hierarchy, with exclusions in the following order: systemic arthritis, rheumatoid factor (RF) positive arthritis, psoriasis or a combination of dactylitis and psoriatic nail changes (psoriatic arthritis), and HLA-B27 positive arthritis (enthesitis related arthritis), with the remainder of children being classified as having either RF negative polyarthritis or RF negative oligoarthritis, depending on number of joints involved, with additional information noted as descriptors. The "modified Durban" classification was proposed only to stimulate discussion among clinicians. RESULTS: Of 50 children, 56% were "unclassifiable" by the Durban classification, mainly because of inadequate family history despite appropriate questioning. Using the proposed "modified Durban" classification, 2% were "unclassifiable." Family history was classified as inadequate for the following reasons: The parents did not know family history; the child or parent was adopted; the father was unknown or parent died early; parents never attended; extended family had lost communication with parents; or a relative was considered to have psoriasis, but not confirmed by dermatologists. Other reasons for "unclassifiable" included: dermatologists unable to confirm psoriasis; family history of inflammatory bowel disease and sacroiliitis but B27 status unknown; proband B27 negative but family history of B27-related disease; family history of psoriasis, but patient had insufficient criteria for psoriatic arthritis and therefore excluded from oligoarthritis, psoriatic arthritis and other groups. CONCLUSION: (i) The Durban classification showed poor efficacy for JIA where < 5 joints were involved at onset, with more than half the cases being "unclassifiable". (ii) The most common reason was that appropriate family history was not available despite being sought by the clinician. (iii) A proposed hierarchical system, an unofficial modification of the Durban classification, showed good efficacy, with only one of 50 cases being "unclassifiable."