Assessing Treatment Efficacy in Diabetes Clinical Trials

Assessing Treatment Efficacy in Diabetes Clinical Trials

A Paediatric Diabetes Clinical Trial

04/10/2026

Lily Antell

This research project explores a clinical trial assessing the safety and efficacy of a hybrid closed-loop insulin delivery system in children aged 2 to 6 years with type 1 diabetes (Wadwa et al., 2023). The primary aim was to estimate the effect of closed-loop control on 13-week HbA1c and investigate the robustness of this estimate to covariate adjustment, missing-data handling and Bayesian modelling. The analysis began with a complete-case comparison, followed by adjusted regression models, different frequentist methods of imputation and Bayesian approaches. The results consistently favoured closed-loop control over standard care, with adjusted treatment-effect estimates ranging from approximately 0.36 to 0.44 percentage points across the frequentist analyses. The Bayesian results were broadly consistent with these findings. A further exploratory aim was to model 26-week HbA1c outcomes during the trial's extension period using Bayesian longitudinal models. However, these results were associated with greater uncertainty, particularly because participants originally assigned to standard care switched to closed-loop treatment after 13 weeks. Overall, the findings provide reasonably consistent evidence that closed-loop control reduced 13-week HbA1c compared with standard care, while conclusions about the 26-week treatment effect remain more uncertain.

Artificial Pancreas, Type 1 Diabetes, Paediatrics, Clinical Trial, Clinical Trial Data, Biostatistics, Clinical Trial Analysis, Longitudinal Data Analysis, Statistical Modelling, Data Analysis, Digital Health, R Project, R Programming, RStudio, Bayesian Statistics, Frequentist Statistics, Mixed-Effects Models, Markov Chain Monte Carlo (MCMC)