Diabetes In Control. A free weekly diabetes newsletter for Medical Professionals. News and information for Medical Professionals.
For adults with type 1 diabetes choosing how to deliver insulin, the decision is increasingly more complicated than simply injections versus a pump. Smart insulin pens can track doses, calculate boluses, and integrate with continuous glucose monitoring (CGM), while automated insulin delivery (AID) systems can continually adjust insulin based on glucose readings.
A new randomized crossover trial provides a rare direct comparison of a smart insulin pen and automated insulin delivery. On average, automated insulin delivery produced considerably better glycemic outcomes. Yet nearly half of the participants evaluated during smart-pen treatment still achieved what researchers defined as near-optimal glycemic control.
For patients and clinicians weighing smart insulin pen vs AID options, the trial provides one of the clearest direct comparisons yet. The findings reinforce the effectiveness of AID while also raising an important clinical question: Does every person who can benefit from automation necessarily want—or need—the same level of technology?
Table of Contents
- What Did the EBIACE-1 Trial Compare?
- AID Produced Better Average Glycemic Outcomes
- The Smart Pen Still Worked Well for Some Participants
- Why AID Has an Advantage
- Why Some Patients May Still Prefer a Smart Insulin Pen
- Who May Benefit Most From AID?
- Treatment Burden Matters Too
- What the Study Cannot Tell Us
- Putting the Findings Into Clinical Practice
- The Bottom Line
- Frequently Asked Questions
What Did the EBIACE-1 Trial Compare?
The EBIACE-1 study was a randomized, open-label crossover trial involving 31 adults with type 1 diabetes. Participants used two different insulin-delivery approaches: the InPen smart insulin pen and the MiniMed 780G automated insulin delivery system.
Each treatment period lasted six months, after which participants crossed over to the other technology. This design allowed researchers to compare outcomes when individuals experienced both approaches rather than comparing two completely different groups of patients.
The co-primary outcomes were hemoglobin A1C and time in range, defined as the percentage of CGM readings between 70 and 180 mg/dL.
That direct comparison matters because smart pens and AID systems offer fundamentally different levels of automation. A smart pen assists with injection-based therapy, whereas an AID system uses CGM information and an algorithm to automatically modify insulin delivery throughout the day.
AID Produced Better Average Glycemic Outcomes
The clearest result was that automated insulin delivery produced better average glucose control.
During AID use, participants spent an average of 81% of the time between 70 and 180 mg/dL, compared with 66% during smart-pen treatment. That amounted to a 15-percentage-point improvement in time in range.
Average A1C was also lower with AID:
- 6.9% with automated insulin delivery
- 7.3% with the smart insulin pen
The difference in A1C was 0.4 percentage points, and both differences were statistically significant.
For clinicians, those results are substantial. An additional 15% of the day in range translates to roughly 3.6 more hours per day with glucose between 70 and 180 mg/dL.
The findings also align with the broader evidence supporting automated insulin delivery in type 1 diabetes. Current diabetes guidance increasingly favors AID for people who can use the technology safely and appropriately, while still emphasizing individualized device selection.
The Smart Pen Still Worked Well for Some Participants
The average results tell only part of the story.
Researchers also looked at whether participants could maintain near-optimal glycemic control while using the smart insulin pen. Among the 26 participants evaluated for this analysis, 12—or 46%—met that threshold.
That does not mean the two technologies performed equally. They did not. AID was clearly superior at the group level.
Instead, the finding suggests that there may be a subgroup of adults with type 1 diabetes who can achieve strong glycemic outcomes with injection therapy enhanced by connected technology.
This distinction matters when counseling patients. Population averages can help establish which therapy is generally more effective, but treatment decisions still occur one patient at a time.
Why AID Has an Advantage
The difference between the two approaches largely comes down to automation.
A smart insulin pen can record insulin doses, help calculate boluses, provide reminders, and combine insulin information with CGM data. These features can address several common problems with multiple daily injections, including forgotten doses and uncertainty about how much insulin remains active.
But the patient still has to administer the insulin.
AID goes further. These systems combine a CGM, insulin pump, and control algorithm that repeatedly evaluates glucose readings and adjusts insulin delivery. Depending on the system, it can increase, decrease, or pause insulin and may administer automated correction doses.
That continuous feedback loop gives AID something a connected pen cannot provide: the ability to respond automatically between treatment decisions.
Why Some Patients May Still Prefer a Smart Insulin Pen
Better glycemic averages do not eliminate the practical reasons someone may prefer injections.
Some people simply do not want to wear an insulin pump. Others may dislike having another device attached to their body or may find infusion-set management inconvenient.
Potential considerations include:
- Comfort with injections
- Desire to avoid wearing a pump
- Concerns about device visibility
- Alarm or notification burden
- Infusion-set management
- Cost and insurance coverage
- Access to training and technical support
- Comfort with changing existing diabetes routines
For patients who want to remain on injections, a smart pen can offer a useful middle ground between traditional multiple daily injections and pump-based automation.
Connected insulin pens may be particularly appealing to people who want better dose tracking and decision support without switching to a wearable pump.
Who May Benefit Most From AID?
The EBIACE-1 results strengthen the argument for discussing AID with adults whose current insulin regimen is not achieving desired glucose targets.
Because AID produced substantially greater time in range in the trial, patients struggling with persistent hyperglycemia, glucose variability, missed corrections, or the daily cognitive demands of insulin adjustment may have considerable room to benefit from greater automation.
However, eligibility for a technology and willingness to use it are not the same thing.
A person who strongly dislikes pump therapy may use it inconsistently or discontinue it. A technically less automated option that fits comfortably into someone's daily life may sometimes prove more sustainable.
A patient who wants more automated insulin adjustment may find an AID system especially appealing. Someone who strongly prefers injections and is already meeting individualized glycemic goals may have less reason to switch.
Treatment Burden Matters Too
One particularly interesting part of EBIACE-1 involved diabetes-related life interference.
In an exploratory analysis, participants reporting greater diabetes-related interference with daily life at baseline were less likely to achieve near-optimal control with the smart pen.
Because this was an exploratory finding from a small trial, it should not be used as a clinical prediction rule.
Still, it raises an intriguing possibility.
Patients already experiencing substantial burden from diabetes self-management may be precisely the people for whom automation is most helpful. Multiple daily injection therapy can require repeated calculations and treatment decisions throughout the day. AID cannot eliminate diabetes-management work, but it can automate part of it.
Conversely, patients who already have highly consistent injection routines and feel little interference from those routines may be more capable of maintaining strong control with a connected pen.
Further research will be necessary to determine whether patient-reported measures can reliably help identify who is most likely to succeed with each approach.
What the Study Cannot Tell Us
EBIACE-1 provides valuable evidence, but it was a small study.
Only 31 adults were enrolled, and fewer participants ultimately contributed paired data to the main crossover analysis. That limits how confidently the results can be generalized to the broader type 1 diabetes population.
The study also compared two specific products—InPen and MiniMed 780G. Results should therefore not automatically be assumed to apply identically to every smart insulin pen or every AID system.
The investigators reported no episodes of severe hypoglycemia or diabetic ketoacidosis during follow-up, but the trial was not large enough to establish meaningful differences in uncommon safety events.
Most importantly, the finding that some participants achieved excellent control with a smart pen should not be interpreted as evidence that smart pens are generally equivalent to AID. The primary comparison pointed clearly toward better average glycemic outcomes with automated insulin delivery.
Putting the Findings Into Clinical Practice
The clinical takeaway from this comparison of smart insulin pens and AID systems is not that both technologies perform the same.
AID produced better average glucose outcomes.
The more nuanced message is that the best technology decision involves more than asking which device produces the strongest average result in a trial.
A useful discussion may include:
- Current A1C and CGM metrics
- Frequency and severity of hypoglycemia
- Glucose variability
- Missed or delayed insulin doses
- Comfort with infusion sets and wearable devices
- Diabetes-related distress and treatment burden
- Ability to manage device troubleshooting
- Insurance coverage and affordability
- Lifestyle considerations
- The patient's own priorities
A patient who wants more automated insulin adjustment may find an AID system especially appealing. Someone who strongly prefers injections and is already meeting individualized glycemic goals may have less reason to switch.
Technology selection should ultimately reflect the person's clinical needs, preferences, circumstances, and ability to use the device consistently.
The Bottom Line
In the EBIACE-1 trial, automated insulin delivery clearly outperformed a smart insulin pen on average. Participants using AID had substantially greater time in range and lower A1C.
But the study also showed why diabetes technology should not become a simple hierarchy in which every person is expected to move toward maximum automation.
Nearly half of the evaluable participants achieved near-optimal glycemic control with the smart pen. For patients who prefer multiple daily injections and can achieve their individualized treatment goals safely, a connected pen may remain a meaningful option.
For clinicians, the goal is therefore not simply to identify the most sophisticated device. It is to match the capabilities of the technology with the clinical needs, daily burden, resources, and preferences of the person who will actually use it.
Frequently Asked Questions
Is an automated insulin delivery system better than a smart insulin pen?
In the EBIACE-1 randomized crossover trial, AID produced better average glycemic outcomes. Participants had 81% time in range with AID versus 66% with the smart insulin pen, and average A1C was 6.9% versus 7.3%.
Can someone with type 1 diabetes still achieve good control with a smart insulin pen?
Yes. In this small trial, 12 of 26 participants evaluated during smart-pen therapy achieved the researchers' definition of near-optimal glycemic control. That result does not establish equivalence with AID, but it indicates that some patients can achieve strong outcomes while remaining on injection therapy.
What does a smart insulin pen do?
Connected or smart insulin pens can record insulin doses and timing, assist with dose calculations, provide reminders, and integrate insulin information with other diabetes data such as CGM readings. Unlike an AID system, however, they do not continuously deliver or automatically adjust insulin.
Does automated insulin delivery completely automate type 1 diabetes management?
No. Current AID systems automate important parts of insulin delivery, but they do not eliminate all diabetes-management tasks. Users still require appropriate training, troubleshooting skills, and ongoing management and support.
Should everyone using a smart insulin pen switch to AID?
Not necessarily. AID offers important glycemic advantages on average, but treatment decisions should also account for current glucose control, hypoglycemia risk, treatment burden, device preferences, cost, access, and willingness to use pump technology. Patients should make technology changes in consultation with their diabetes care team.
Who might be a good candidate for a smart insulin pen?
A smart insulin pen may be appropriate for someone who prefers multiple daily injections but wants better dose tracking, reminders, insulin-on-board information, or integration with glucose data. Clinical suitability depends on the person's individual treatment needs and goals.
This article is intended for educational purposes and does not substitute for individualized medical advice. Treatment and diabetes-technology decisions should be made with a qualified healthcare professional.