957-P: Feasibility Study of Use of the iLet Bionic Pancreas in Primary Care and Telehealth

Introduction & Objective: While about half of US adults with T1D are managed by primary care (PC), automated insulin delivery (AID) use is rare in PC compared to endocrinology (EN). The iLet Bionic Pancreas (BP) has a unique initiation process of simply entering the user’s weight, followed by perpetual autonomous adaptation to meet each user’s changing insulin needs; the BP may thus be more feasible for PC and/or telehealth (TH) deployment than AID systems with more intensive startup and follow-on care. This random order crossover trial (NCT05168657) evaluated feasibility of BP startup and management by PC versus EN and training via TH versus in person (IP).

Methods: Forty adults with T1D (20 at a PC site, 20 at an EN site) were assigned to complete two study arms (14 days of BP use, 14 days of their usual diabetes care [UC]) in random order. PC participants’ UC was required to be multiple daily injections with or without CGM; EN participants’ UC was CGM sensor augmented pump. Half of each group were trained and managed by IP visits and half by TH. Primary outcome: % of each group achieving average glucose (AG) <183 mg/dL during BP use.

Results: Both arms were completed by 39 participants. With BP use, 38 (97.4%) achieved AG <183 mg/dL, with no differences between PC and EN groups or the respective IP and TH subgroups (all p>0.05). Decreased AG was seen with BP use compared to UC in all four groups (all p<0.05). With BP use, PC and EN groups achieved similar results in AG, time <70, time between 70 and 180, time >180, and time >250 mg/dL. The PC group had slightly lower CV and time <54 mg/dL than EN with BP use (both p=0.03). There were no severe adverse events during BP or UC use.

Conclusion: Training and initiation of the BP by PC and EN subspecialists, and through IP versus TH visits, resulted in similar CGM outcomes with the BP, which were improved relative to UC. These findings suggest that the BP, with its uniquely simple initiation, use, and follow-on care, may expand AID access for people limited by geographic and/or subspecialty access.