A sensor that watches for the danger a normal glucose reading can hide

The gap this is meant to close

There is a form of the condition called euglycaemic DKA, in which ketones rise dangerously while blood glucose stays within a range that looks unremarkable. Someone checking a continuous glucose monitor would see nothing wrong. Abbott says the device is aimed particularly at people taking SGLT2 inhibitors, a class of medication where this presentation is a known risk.^1,2

That is the entire argument for a second analyte. A monitor that shows a reassuring number during the hours when a person is becoming seriously unwell is not merely unhelpful. It is actively misleading, and the reassurance is the harm.^1,3

Abbott cites published research reporting that hospital admissions for DKA rose 55 percent over the past decade. That figure is Abbott's citation of a 2024 study rather than an Abbott measurement, and we have not examined the underlying paper.¹

What was actually authorised

The device is worn for up to ten days by people with diabetes aged two and over who take insulin or other glucose-lowering medication. It measures both analytes in interstitial fluid every minute, and Abbott says it is the first system to meet the FDA's integrated continuous glucose monitoring standard and a new integrated glucose and ketone standard created for this category.^1,3

The wear-time claim comes with its own numbers, and Abbott publishes them rather than rounding them away. In its data, 84.1 percent of sensors in adults and 68.8 percent in children lasted the full ten days. A device advertised as ten days will, for roughly a third of paediatric users, not reach ten days.¹

Abbott received Europe's CE mark in May, ahead of the American authorisation, and says the system is designed for compatibility with leading automated insulin delivery systems.^1,3

What continuous means, and what it costs in attention

Every minute is a high sampling rate, and it produces a second stream of numbers for someone who is already managing one. That is not a trivial addition. People living with diabetes describe the mental load of monitoring as one of the harder parts of the condition, and a device that adds a measurement also adds a thing to interpret at three in the morning.^1,2

The design question is therefore how quiet the second stream stays. A ketone reading that only speaks when it matters is a safety net. One that produces frequent borderline alerts becomes something people learn to dismiss, which is worse than not having it, because the dismissal habit forms before the real warning arrives.^1,3

Abbott's announcement does not describe alert thresholds or how the two measurements are presented together, so this cannot be assessed yet. It is the first thing to look at when the product reaches actual users, and the first thing a clinician will be asked about.¹

The questions the announcement leaves open

Abbott says the product will launch in the United States later this year and gives no price. For a wearable replaced every ten days, price and insurance coverage are not details. They determine whether a cleared device becomes an available one, and the announcement is silent on both.^1,3

• There is no published price, and no statement about insurance or Medicare coverage.
• Later this year is the only stated timing for United States availability.
• Ketone thresholds and alert behaviour are not described in the announcement, and how a device warns matters as much as what it measures.
• The paediatric wear-time figure of 68.8 percent is worth raising with a clinician if the device is being considered for a child.
• Nothing here changes existing guidance. A person on SGLT2 inhibitors should still know the symptoms of DKA rather than delegating that to a sensor.

Why this one is worth noticing

Most announcements in consumer health hardware add a metric that a healthy person might find interesting. This adds a measurement that a specific group of people has a clinical reason to want, aimed at a failure mode that existing devices cannot see. That is a narrower claim and a more serious one.^1,2

It also arrives through the regulatory route reserved for devices with no predicate, which means the FDA had to define what an acceptable version of this category looks like. The standard created here is likely to outlast this particular product.^1,3

The caution is the familiar one for any monitor. A device that watches for a dangerous state does not treat it, and the value depends entirely on what the person and their clinician do when it raises a warning. An authorisation is the beginning of that question, not the answer to it.^1,2