Development and validation of the RACER (Readiness for Adult Care in Rheumatology) transition instrument in youth with juvenile idiopathic arthritis

Background: Current evidence suggests that many adolescents with juvenile idiopathic arthritis (JIA) do not successfully transfer to adult care, which can result in adverse health outcomes. Although a growing number of clinical programs have been designed to support healthcare transition, there is a lack of psychometrically sound instruments to evaluate their impact on development of transition-related knowledge and skills in youth with JIA. The purpose of this study was to develop and validate RACER (Readiness for Adult Care in Rheumatology), a self-administered instrument designed to measure stages of readiness for key transition-related skills in adolescents with JIA.

Methods: A phased approach was used to develop and evaluate the validity and reliability of RACER. Phase 1 A was a consensus conference with 19 key stakeholders to inform instrument domains and items. Phase 1B determined initial content validity using a sample of 30 adolescents with JIA and 15 clinical and research experts. Finally, Phase 2 was a prospective cohort study with repeated measures to evaluate the internal consistency, test-retest reliability, construct validity and responsiveness of the instrument within a sample of adolescents with JIA.

Results: In Phase 1 A, initial item generation yielded a total of 242 items across six domains from the consensus conference, which was subsequently reduced to a 32-item instrument. Phase 1B established the content validity of the instrument in adolescents with JIA. In the Phase 2 study, with a sample of 96 adolescents, the RACER instrument exhibited good internal consistency in five of its six subscales (Cronbach’s α > 0.7), and strong test-retest reliability between the first two administrations (ICC = 0.83). It also showed robust convergent validity by highly correlating with measures of self-management (SMSAG, rho = 0.73) and transition (TRANSITION-Q, rho = 0.76). The RACER was not correlated with unrelated measures (discriminant validity; PedsQL, rho = 0.14). The RACER scores increased significantly over time as expected, supporting measure responsiveness.

Conclusions: The RACER is a reliable and valid instrument which is sensitive to change for assessing transition readiness in adolescents with JIA.