Diabetic ketoacidosis is getting worse, more frequent, and more deadly — and the medical devices meant to stop it are caught in a race against a crisis that’s quietly accelerating.
A 23-year British study tracking nearly 660,000 people has revealed a stark and troubling trend: DKA rates have more than tripled among adults with Type 1 diabetes and increased sixfold among those with Type 2. Recurrence — the dangerous pattern of patients cycling back through the same life-threatening emergency — is running at 31.5% in Type 1 and 12.1% in Type 2. This isn’t a blip. It’s a structural problem building over decades, and the healthcare system is only now beginning to confront it with tools that might actually matter.
A Crisis That Keeps Coming Back
“This long-term analysis shows that DKA is not only becoming more common, but also more likely to reoccur, and often in the same individuals,” said Pratik Choudhary, M.D., endocrinologist at Leicester Diabetes Centre and the University of Leicester. That last part — the same individuals — is what makes this so clinically maddening. It’s not random. There’s a cohort of high-risk patients cycling in and out of emergency wards, and traditional glucose monitoring hasn’t been enough to catch them before they crash.
How bad does it get? Separate UK data covering 2019 to 2022 puts 30-day mortality following a DKA episode at anywhere from 1.8% to 11%, depending on patient profile. In 2019 alone, recurrent DKA within 12 months was documented in 25.1% of Type 1 patients and 9.8% of those with Type 2 — figures that align grimly with the longer longitudinal trends.
Where Continuous Monitoring Comes In
The most compelling counterargument to that trajectory comes from the RELIEF study, which examined what happens when patients actually use Abbott’s FreeStyle Libre continuous glucose monitoring system consistently. The results were hard to dismiss: a 49% reduction in DKA hospitalizations for Type 1 patients and 48% for Type 2, measured over two years, with an 88% persistence rate among users. Those aren’t marginal gains. That’s nearly cutting the hospitalization rate in half.
Still, glucose monitoring only tells part of the story. DKA is fundamentally a ketone problem — the body’s acid load building to dangerous levels — and glucose readings alone can’t always catch it early enough. That’s why researchers have been investigating blood ketone testing patterns alongside CGM data, with preliminary case studies now exploring whether a dual glucose-ketone sensor could close that diagnostic gap for both Type 1 and Type 2 patients.
The Next Frontier — And Why It’s Not Here Yet
An international consensus framework has gone so far as to call continuous ketone monitoring transformational for DKA prevention. Abbott is actively developing dual glucose-ketone sensors. But here’s the catch: none of them are FDA-approved yet. The technology is tantalizingly close, the clinical rationale is well-established, and the regulatory finish line remains just out of reach.
That’s not the only complication hanging over Abbott’s monitoring ambitions. As of early 2026, the company had linked 860 serious injuries and seven deaths globally to inaccurate FreeStyle Libre 3 sensor readings — including cases involving DKA. It’s a sobering footnote to the otherwise compelling efficacy data, and it underscores a tension that runs through the entire CGM space: a device that’s supposed to prevent a medical crisis can, under certain failure conditions, contribute to one.
A Race With Real Stakes
None of this diminishes the urgency of the underlying problem. DKA is expensive, recurring, and killing people at rates that have been climbing for over two decades. The evidence that continuous monitoring reduces hospitalizations is real and substantial. The science behind ketone monitoring is advancing fast. But the path from promising technology to approved, widely deployed clinical tool is rarely as straight as the data would suggest — and in the meantime, patients are cycling back through emergency rooms at rates that should alarm anyone paying attention.
The sensors are getting smarter. The question is whether the system around them can keep up.
