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

Abstract

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.

Editorial: Equitable digital medicine and home health care

“…Thanks to advanced sensors and devices such as smartwatches or smart bracelets, patients can measure their vital parameters, monitor physical activity, check glucose levels or track other important indicators of their health. This data is sent in real time to HWs who can carry out constant monitoring and intervene promptly in the event of anomalies ( 25 – 29 ). This allows personalized treatment and more accurate monitoring of the patient's health condition as well as increasing patient empowerment.…”