A whole new ball game: a retrospective cohort study using healthcare administrative data to uncover predictors of timely transition from paediatric to adult type 1 diabetes care in British Columbia, Canada.

Abstract

To assess predictors of timely transition to adult diabetes care among individuals diagnosed with type 1 diabetes during childhood and adolescence. We hypothesised that older age at the last paediatric visit and urban residency would be predictors of timely transition. Retrospective cohort study using healthcare administrative data in a jurisdiction with a universal healthcare system. 2045 adolescents and young adults diagnosed with type 1 diabetes between the ages of 0.5 and 18 years. We ascertained age at the last paediatric diabetes visit (LPDV), age at the first adult diabetes visit (FADV) and transition duration, defined as the time between LPDV and FADV. Timely transition was defined as a transition duration of <1 year. Logistic regression models were fitted to assess predictors of timely transition. Only 31.3% of individuals saw an adult provider within 1 year of their LPDV. Each 1-year increase in the age at LPDV was associated with increased odds of timely transition (adjusted OR 1.82, 95% CI 1.71 to 1.93, p<0.001). Urban residency was also associated with increased odds of timely transition (adjusted OR 1.93, 95% CI 1.40 to 2.71, p<0.001). Sex and older age at diagnosis were not associated with timely transition in the multivariable regression models (p>0.05). Older age at the LPDV and urban residency are associated with increased odds of timely transition. Interventions should be developed to help keep adolescents engaged in paediatric care until an older age before referring them to adult diabetes care. Limitations of this study include unmeasured confounding and limited generalisability to non-universal healthcare systems.