Consider a woman in her early fifties who tells her GP about a fluttering in her chest. It comes and goes. A ten-minute ECG in the surgery shows nothing, and she is advised to return if it happens again. Months pass, the cardiology queue lengthens behind her, and the arrhythmia that might explain everything remains unrecorded. A £184 million framework agreement between NHS Supply Chain and Cambridge Heartwear is aimed at precisely this gap.

The framework allows integrated care boards to buy a wireless, medical-grade ECG wearable with adaptive AI analysis, which sends encrypted readings to the patient's smartphone for clinician review. It holds UKCA approval, is made in Britain, and emerged from the University of Cambridge. Its design addresses the long-documented underdiagnosis of heart disease in women. The case for continuous home monitoring of intermittent rhythm problems is credible, and the government's stated ambition to move care out of hospitals gives the purchase a clear policy home.

Yet a framework is a route to purchase and carries no obligation to spend. The £184 million figure is a ceiling, and it says little about how many devices will reach patients. Integrated care boards are meanwhile cutting running costs and merging functions, which leaves fewer people to design the clinical pathways that a new diagnostic tool requires. A simpler procurement route removes one barrier. The others remain in place.

The most pressing of these is the demand that detection creates. Continuous monitoring finds arrhythmias that episodic testing misses, including asymptomatic and low-risk findings. Each flagged trace needs a clinician to review it, and a share of them will lead to echocardiography, cardiology referral or decisions about anticoagulation. If primary care has no protected time to interpret the output, and community diagnostic capacity is already stretched, the device relocates the queue without shortening it. A patient wearing a monitor at home may simply wait for a reviewer instead of a clinic slot.

The AI component sharpens the question. Its value rests on specificity as much as sensitivity, because every false positive consumes scarce clinical time and may cause real anxiety. Boards should expect independent evaluation in NHS populations, including the older and more diverse patients who present with palpitations, before committing large volumes. Because the algorithm is described as adaptive, the MHRA and local clinical safety officers will also need clarity on how changes to its behaviour are controlled once devices are in the field.

Data governance deserves equal attention. Readings pass through patients' own phones, then to clinicians, and presumably into GP and hospital records. Public confidence in NHS data sharing has been damaged more than once, most visibly by earlier national programmes that were announced before the public was properly informed. Patients will want to know who holds their cardiac traces, for how long, and whether the data may train future algorithms. Boards that answer those questions plainly at the outset will find adoption easier than those that treat them as a compliance task to be completed later.

For industry the agreement carries a wider lesson. The government wants a stronger domestic life sciences sector, and a UK manufacturer with a large patent portfolio and a pending FDA application is the kind of company that policy hopes to produce. A framework of this size gives it a reference customer of real weight. The NHS has a long record of welcoming such innovators and then failing to adopt their products at scale, because pilots rarely become funded, staffed and commissioned services. Health-tech suppliers will watch whether this contract breaks that pattern.

The sensible measure of success is therefore not the value of the purchase order. It is the median time from a patient's first report of palpitations to a confirmed diagnosis, and the proportion of monitored patients who go on to need specialist care. Leaders who track those figures from the first deployment will learn quickly whether the technology is shortening the path to treatment. Without them, £184 million could buy a great deal of data and very little improvement in access.

‍