Background/Purpose: It is unclear which tool should measure functional ability in adolescents with JIA. The proxy-completed Childhood Health Assessment Questionnaire (P-CHAQ) is completed on the adolescent’s behalf and an adolescent version (A-CHAQ) has not been validated. Since adolescence parallels transfer to adult care, the adult HAQ may be preferable to capture functional ability throughout transition. However, it is unclear how the HAQ compares with the two CHAQ tools. Agreement between the P-CHAQ, A-CHAQ and HAQ was assessed in adolescents with JIA during the three years following presentation to rheumatology.

Methods: Adolescents aged 11 to 17 years recruited before 1^st January 2013 to the Childhood Arthritis Prospective Study (CAPS), a UK multicentre JIA inception cohort, were selected. Adolescents had complete data on proxy-completed P-CHAQ and adolescent-completed A-CHAQ and HAQ at presentation. These assessments were continued annually. At baseline, Wilcoxen signed-rank tests compared median scores, Spearman’s correlations assessed pairwise correlations and percent agreement (defined as cores within 0.25 points) was assessed. Univariate and age and sex-adjusted associations between scores were assessed via zero-inflated negative binomial models. Multiple imputation accounted for missing data for longitudinal models, which had robust clustering at the patient level. 

Results: Of 94 adolescents included, median age at diagnosis was 13 years (IQR 12 to 15) and 61% were female. Median disease duration at diagnosis was seven months (IQR 5 to 14) and the most common subtype was oligoarticular JIA (40%). Median baseline HAQ (0.5) was marginally lower than both CHAQ scores (both 0.6), although this difference was not clinically significant. In accordance, the highest agreement was between the two CHAQ tools (78%) and lowest between the HAQ and P-CHAQ (71%). Where discordant, the majority of HAQ scores fell below those from either CHAQ. Discordance between CHAQ scores was more evenly distributed (Table 1). Despite marginally different medians, the strongest correlation was between the HAQ and the A-CHAQ (0.91), with the lowest between the two CHAQ tools (0.83). After adjustment for age and sex, there was around 11% difference in scores at baseline and 8% over the course of three years (Table 2).   

Conclusion: There was strong correlation, good concordance and similar associations between the P-CHAQ, A-CHAQ and HAQ in adolescents with JIA. The strong relationship between the HAQ and either CHAQ tool indicate the utility of HAQ in adolescents with JIA.