IPM Take
Diabetes care is often presented as one of the success stories of personalised, technology-enabled medicine. Insulin pumps, glucose meters, continuous glucose monitoring and automated delivery systems are changing how patients manage disease every day. But two recent FDA signals show the other side of that shift. When insulin delivery is interrupted or glucose information is unclear, the risk moves directly to the patient at home. The policy issue is not whether diabetes technology is useful. It is whether health systems have the safety, monitoring, communication and support infrastructure to manage technology-dependent care in the real world.
Executive Summary
The U.S. Food and Drug Administration has issued recent safety communications involving diabetes technologies used in routine care. On 4 June 2026, FDA issued an Early Alert for certain Insulet Omnipod Pods after identifying that some Pods from affected lots may have a small tear in the cannula, potentially causing insulin to leak outside the device instead of being delivered to the body. FDA warned that this may happen without an alert and that prolonged insulin under-delivery can lead to diabetic ketoacidosis. As of 20 May 2026, Insulet had reported 24 serious injuries and no deaths linked to the issue. Separately, FDA classified a TRUE METRIX blood glucose monitoring system recall as Class I, the most serious type, because the same E-5 error code can indicate either a very high glucose event or a test strip error. Trividia Health had reported 114 serious injuries and one death associated with that issue as of 16 January 2026. Together, the signals highlight a growing readiness challenge: personalised diabetes care is only as safe as the devices, instructions and support systems patients rely on every day.
Why it matters
- Regulators: Post-market surveillance, software design, device labelling and patient-facing safety communication are becoming central to diabetes technology oversight.
- Clinicians: Patients using insulin pumps or glucose meters may need clearer backup plans, device education and escalation instructions when readings or delivery appear abnormal.
- Patients: Technology can improve daily diabetes management, but device problems can create serious risks when they are hard to detect or interpret.
- Payers: Coverage decisions should not only consider device purchase costs, but also training, replacements, support services and safe transition pathways when recalls occur.
- Manufacturers: The safety bar is shifting from product performance alone to real-world usability, communication, rapid correction and patient trust.
Personalised diabetes care increasingly happens outside the clinic.
That is the promise.
It is also the risk.
For many people with diabetes, daily treatment now depends on a network of devices: insulin pumps, Pods, glucose meters, test strips, sensors, apps and automated delivery systems. These tools can improve control and reduce burden. But when something goes wrong, the patient may be the first person exposed to the harm.
Two recent FDA safety signals show why this matters.
On 4 June 2026, FDA issued an Early Alert for certain Insulet Omnipod Pods. The affected products include Omnipod 5, Omnipod DASH and Omnipod Eros Pods from specific lots. According to FDA, some Pods may have a small tear in the cannula, the tubing that delivers insulin under the skin.
If that tear occurs, insulin may leak outside the Pod instead of being delivered into the body.
That can lead to under-delivery of insulin.
In some cases, users may notice wetness on the skin, wetness on the adhesive or the smell of insulin. But FDA also warns that the problem may be difficult to detect and may go unnoticed. It may also happen without triggering a device alert.
For a person relying on insulin delivery, that is not a minor technical issue.
If insulin is not delivered properly, blood glucose can rise. If high glucose persists, diabetic ketoacidosis can develop. Diabetic ketoacidosis is a serious and potentially life-threatening condition requiring prompt medical treatment.
As of 20 May 2026, Insulet had reported 24 serious injuries and no deaths associated with the issue.
A second FDA signal points to a different kind of problem: not insulin delivery, but glucose interpretation.
FDA has classified a recall involving TRUE METRIX blood glucose monitoring systems as Class I, the most serious category. The issue concerns the E-5 error code. The same error code may indicate either a very high blood glucose event above 600 mg/dL or a test strip error.
That ambiguity matters.
If a user is having a very high glucose event but thinks the message is only a test strip error, treatment may be delayed. If a user assumes the error means very high glucose but blood glucose is actually low or normal, they may treat incorrectly. FDA warns that either delayed treatment or improper treatment can lead to serious consequences, including dehydration, altered mental status, loss of consciousness or death.
As of 16 January 2026, Trividia Health had reported 114 serious injuries and one death linked to the issue.
Together, these cases expose a bigger policy problem.
Personalised diabetes care is no longer only about prescribing the right medicine or setting the right insulin dose. It is also about whether technology-dependent care is safe, understandable and supported in everyday life.
That includes device design.
It includes software logic.
It includes instructions that patients can interpret under stress.
It includes pharmacists and clinicians knowing what to tell patients when a recall happens.
It includes replacement pathways so that patients are not left without essential monitoring or insulin delivery.
And it includes clear escalation plans when a device warning is ambiguous, absent or misunderstood.
The equity issue is also obvious.
Patients with strong specialist support, digital literacy and easy access to replacement devices may navigate recalls quickly. Patients with limited access to care, lower health literacy, language barriers or fragmented insurance coverage may face greater risk.
That is where digital diabetes care becomes a health-system readiness test.
The question is not whether these technologies should be used. They already are. The question is whether the safety infrastructure around them is strong enough.
For IPM, the signal is clear.
As cardiometabolic care becomes more personalised and more digital, regulators and health systems will need to treat devices as part of the care pathway, not as standalone products.
Because when the device fails, the pathway fails too.
