JMIR Publications

ABSTRACT

Background:

People with diabetes and their support networks have developed open-source automated insulin delivery systems to help manage their diabetes therapy, as well as to improve their quality of life and glycemic outcomes. Under the hashtag #WeAreNotWaiting, a wealth of knowledge and real-world data has been generated by users of these systems but has been left largely untapped by research; the opportunity for multimodal studies investigating this group remains open.

Objective:

Developing a mixed-methods study to evaluate several aspects of open-source automated insulin delivery presents challenges relating to data management and security across multiple disparate online platforms, and implementation of follow-up studies with study participants. This research article reports on the feasibility of such a multimodal concept.

Methods:

A web portal was developed to manage both front-end participant interactions with study elements and back-end data management of survey responses and donated anonymized diabetes data. Participant data in REDCap and Open Humans was pseudonymously and securely linked and stored within a custom-built database using both open-source and commercial software. Participants—which included adults and children with diabetes, their partners and caregivers—were recruited through multiple online diabetes community groups. Participants were later given the option to include their healthcare providers in the study; database architecture was designed specifically with this kind of extensibility in mind.

Results:

Of 1052 visitors to the study landing page, 930 participated and completed at least one or multiple questionnaires. After the implementation of healthcare professional validation of self-reported clinical outcomes to the study, an additional 145 individuals visited the landing page, with 124 completing at least one or multiple questionnaires. Of the optional study elements, 7 participant-healthcare professional dyads participated in the survey, and 97 participants who completed the survey joined Open Humans to also donate their anonymized medical device data.

Conclusions:

The study design proved successful in being both accessible for participants and manageable for researchers while maintaining compliance with data regulations. The gateway proved scalable when tested with the later addition of validation of self-reported data. Custom software solutions like the gateway may become increasingly common in diabetes research, especially with medical device data donation and follow-up studies. The gateway portal code has been made available open-source and can also be leveraged by other research projects.