Summary: In adults with insulin-treated type 2 diabetes using continuous glucose monitoring, automated insulin delivery improved HbA1c and time in range more than continuation of the pretrial insulin-delivery method over 13 weeks. Hypoglycemia was uncommon in both groups, although one severe event occurred with automated delivery.
PICO Summary
| Element | Detail |
|---|---|
| Population | 319 adults with insulin-treated type 2 diabetes; 13-week multicentre randomized controlled trial in which all participants used CGM. |
| Intervention | Automated insulin delivery using a pump and control algorithm, allocated 2:1 (approximately 213 participants). |
| Comparison | Continuation of the participant's pretrial insulin-delivery method with CGM (approximately 106 participants). |
| Outcome | HbA1c fell by 0.9 points with AID versus 0.3 points with control; adjusted difference -0.6 percentage points (95% CI -0.8 to -0.4; P<0.001). Time in range increased by 14 percentage points (95% CI 11-17; P<0.001). |
Automated Insulin Delivery in T2D (2IQP)
Multicentre RCT - insulin-treated T2D - 13 weeks
Automated insulin delivery improved HbA1c and time in range over 13 weeks versus CGM with usual insulin delivery in adults with insulin-treated type 2 diabetes.
Expert Commentary
This trial moves automated insulin delivery for type 2 diabetes from plausible technology to randomized efficacy evidence. The adjusted HbA1c improvement of 0.6 percentage points and 14-point gain in time in range are clinically meaningful, particularly because both groups had CGM and the comparison therefore tested the added value of automation rather than glucose visibility alone. Hypoglycemia remained uncommon, although one severe event occurred in the automated-delivery group. The main limitation is duration: 13 weeks cannot establish long-term device persistence, skin and infusion-set burden, cost effectiveness, or durability across changing insulin requirements. The manufacturer funded the study, and trial support may exceed what is available in routine care. Can I use this with my patients? Yes, for adults with insulin-treated type 2 diabetes who remain above glycaemic targets, are willing and able to use a pump, and can receive structured onboarding and follow-up. It is especially relevant when complex insulin adjustment is limiting control. It is not a universal replacement for simpler regimens, and access, dexterity, cognition, health literacy, and affordability matter. Longer pragmatic studies should test sustained benefit, patient-reported outcomes, resource use, and performance in less supported settings. For now, clinicians can discuss AID as a legitimate intensification option rather than a technology reserved exclusively for type 1 diabetes.
References
Kudva YC, Raghinaru D, Lum JW, et al. A Randomized Trial of Automated Insulin Delivery in Type 2 Diabetes. N Engl J Med. 2025;392(18):1801-1812. doi:10.1056/NEJMoa2415948. PMID: 40105270.