First Real-World Study of twiist Automated Insulin Delivery System Shows Promising Glycemic Control

The twiist is an automated insulin delivery (AID) system designed to help adults with Type 1 diabetes manage their blood glucose more easily. Like other AID systems, it works by continuously sensing glucose levels and automatically adjusting insulin delivery based on those readings. This study was the first to evaluate how well twiist performs in real-world conditions with actual users.

Researchers followed 20 adults with Type 1 diabetes for 8 weeks while using twiist. The main measure of success was time in range (TIR)—the percentage of time glucose stayed between 70 and 180 mg/dL, a standard target range. On average, participants spent 70.3% of their time in this range, meaning glucose remained in a healthy zone for most of the day.

When researchers looked at other glucose metrics, participants met or maintained their targets for most measures. However, they spent more time above 250 mg/dL than the study's target goal, suggesting room for improvement in preventing very high glucose levels.

One of the most important safety features of any insulin pump is detecting when insulin delivery gets blocked. During the study, 15 participants experienced 67 occlusion events—times when something prevented insulin from flowing properly. Critically, the twiist system detected these blockages within 30 minutes, giving users a window to troubleshoot and prevent their glucose from rising dangerously.

This rapid detection is important because undetected occlusions can lead to missed insulin doses and blood sugar spikes. The system's ability to flag these events quickly supports user safety.

The study examined whether specific system settings influenced glucose control. Participants who set higher maximum basal insulin ratios—a measure of how much background insulin the pump can deliver—achieved higher time in range. Similarly, participants who adjusted their correction range (the insulin dose for bringing down high glucose) to 100 mg/dL or wider also saw better results. This suggests that taking time to personalize these settings may improve outcomes.

For people with higher baseline HbA1c levels (above 7%), the system helped bring that number down by an average of 0.4% over 8 weeks. HbA1c is a measure of average glucose control over 2–3 months, so even modest improvements can reflect meaningful progress.

This was a small, focused study: 20 people in one location over 8 weeks. While small studies have limits, they can offer valuable first-look insights into how a new device works in practice. The fact that 19 of 20 participants completed the full 8 weeks suggests the system was tolerable and usable for most people.

The results show that twiist can help users achieve solid glucose control and detect delivery problems quickly—two core goals of any automated insulin system. Larger and longer studies will help clarify how this system stacks up compared to others, and whether these early results hold up over months and years of real-world use.

Source: BMJ open diabetes research & care. Evidence type: PubMed indexed literature. Type1Cure is an information and intelligence hub, not a medical advice service. This article summarizes published research and does not provide diagnosis, treatment, or personal medical guidance. Always talk to your own care team before changing anything about your Type 1 diabetes management.