Study Design and Data Analysis of Artificial Pancreas Device Systems with Closed-Loop Glucose-Sensing Insulin Delivery

OBJECTIVE: The objective of this article is to provide a high-profile review and discussion on the study design and statistical analysis of pivotal clinical trials conducted to demonstrate the safety and effectiveness of closed-loop investigational artificial pancreas device systems (APDSs) in premarket approval applications. METHODS: The United States Food and Drug Administration (FDA) guidance on the content of investigational device exemption and premarket approval applications for APDSs is reviewed with special emphasis on study design and statistical analysis of the pivotal clinical trials. The two pivotal studies for the MiniMed 670G hybrid closed-loop system by Medtronic in their premarket approval application are summarized and discussed. RESULTS: The United States FDA established detailed recommendations on the study design and statistical analysis of pivotal clinical trials for the industry that seek market investigational APDSs and for FDA scientific reviewers that regulate the device applications. The recommendations cover specifics regarding patient population, clinical endpoints, and strategies for data analysis. However, the two pivotal studies that demonstrated the effectiveness of the FDA-approved MiniMed 670G hybrid closed-loop system were not typical randomized controlled trials as per FDA recommendations. CONCLUSION: The development and regulation of investigational APDSs require careful and sophisticated clinical study designs and data analysis in premarket approval applications. The regulatory evaluation process of the APDSs is rather complicated since the devices consist of multiple components that collaboratively function to mimic human pancreases.