From pilots to patients: Why the UK must rethink NHS innovation adoption

How can the UK accelerate medical innovation at a time when health systems are under strain and investment is slipping away? That question dominated conversations at the recent PharmaTech Integrates conference in Glasgow, where speakers from government, academia, the NHS, and pharma sought to define the sector’s path forward.

Scotland was held up as proof that new models of adoption, AI-enabled diagnostics, and genomics can transform care at pace. But equally forceful were the warnings from industry executives: without stronger policy support and a more competitive business environment, the UK risks losing its edge in global life sciences.

Scotland as a “living laboratory”

Dame Anna Dominiczak, chief scientist (health) at the Scottish government and regius professor of medicine at the University of Glasgow, captured the optimism in her opening keynote. She described the Scottish approach as a “triple helix” of academia, industry, and NHS working in lockstep, which forms the engine behind accelerated adoption across the country.

She pointed to the Gemini AI breast cancer screening programme, which has increased detection rates by 12% without raising recall rates. “For the first time in my life, I have seen happy radiologists,” she joked, noting the relief among clinicians working in one of the most pressurised imaging specialities.

Dominiczak described the Accelerated National Innovation Adoption (ANIA) programme, which scans for promising technologies and moves them rapidly through strategic assessment, economic evaluation, and system-wide deployment. “To make these innovations their own, boards and executives need to want to adopt,” she stressed – a reminder that cultural buy-in is as vital as technical readiness.

That approach is already reshaping care. Within 10 months, a digital dermatology tool that allows GPs to attach images to referrals had been rolled out across Scotland. More than 7,000 patients have benefitted so far, many spared unnecessary in-person hospital visits. Once scaled, the system is expected to cover 130,000 patients each year, freeing up capacity for those who need urgent attention.

Genetic screening is also changing outcomes. Dominiczak explained that 28-30% of stroke patients carry a genetic variant that renders a widely used drug ineffective. Identifying those patients means treatment can be switched early, preventing repeat strokes. In newborns, one in 500 babies are at risk of profound deafness if treated with gentamicin for infections. Genetic testing now available in Glasgow can flag those infants in advance, allowing doctors to prescribe safer alternatives.

“Scotland is a living laboratory,” Dominiczak concluded – a place where prediction, prevention, and early therapy are already being delivered. But her optimism was soon tempered by harder truths.

The UK at risk

If Scotland was presented as a testbed for rapid adoption, the wider UK picture was more sobering. Mark Proctor CBE, vice president and site general manager of AstraZeneca UK Operations in Macclesfield, delivered a blunt assessment: “We’ve lost billions in investments. We cannot ignore this.”

Proctor argued that life sciences are a “vital national asset” underpinning national health, as well as economic strength – but one increasingly undermined by a deteriorating commercial environment. “The UK is a risk,” he warned, pointing to restrictive pricing, limited population approvals, and sluggish adoption pathways that make domestic manufacturing less attractive.

To prevent further fallout, he called for urgent reforms, including a reset of NHS innovation adoption, streamlined planning and permitting for new sites, better grid connections, and stronger resourcing for the MHRA. Trials, adoption, and manufacturing must be aligned, he argued, if the UK is to remain attractive for investment. “The health of the UK depends on the strength of its life sciences sector,” he said.

The implication was clear – without systemic reform and clean pathways to scale innovation, patients will not see the benefits of the breakthroughs Scotland is showcasing.

AI beyond the hype

Unsurprisingly, artificial intelligence was a throughline across panels. Dominiczak pointed to AI’s role in clinical trial recruitment, where algorithms can now read doctors’ notes and identify suitable patients “in hours, rather than weeks.” In an era when enrolment delays can stall promising studies, she argued, the ability to match the right patients quickly could accelerate access to lifesaving therapies.

She also suggested that embedding genetic information into electronic health records could prevent avoidable harm. “If every human being in Scotland had biogenetic testing in their EHR, you wouldn’t have adverse effects because you would know not to prescribe those medicines,” she said.

On the industry side, Proctor tied AI directly to resilience: “We are investing in autonomous manufacturing,” he said, describing systems designed to cut lead times from development to supply. Professor Yvonne Perrie MBE, chair in drug discovery at the Strathclyde Institute for Pharmacy and Biomedical Sciences, University of Strathclyde, highlighted AI’s role in predicting whether drugs will hit their targets, reducing wasted time and failed studies. Matt Winterman, executive vice president, product supply, operations strategy & transformation officer for Perrigo, argued that the real prize lies in integration: “When we can use multiple unintegrated data sets […] that creates the velocity to get to outcomes.”

But panellists were equally clear that AI brings environmental and ethical costs. Training large models consumes vast amounts of energy. Daniel Markl (University of Strathclyde) noted that “open access to LLMs is the gamechanger,” but acknowledged their heavy carbon footprint compared with narrower machine learning models. The consensus was that the industry should be selective, using smaller, purpose-built models wherever possible – a pragmatic step to avoid adding emissions while chasing insights.

Data ownership was another sticking point. IntelligenceQ CEO and co-founder Swapnil Khandke warned: “If OpenAI owns the data […] you won’t be able to get the IP, and you won’t be able to get the medicine approved.” For companies operating in highly regulated spaces, keeping control of data and models is as important as the insights AI delivers.

And then there was the human factor. Syneos Health’s Janet Ferihough reframed AI’s purpose as supporting, not replacing, decision-making: “The sharp, pointy end of AI is in improving work for people. It’s not just up to the tech part of the business to throw AI at the wall and hope for the best. It’s not just automating decisions. It’s reducing unnecessary labour to give people more time to dedicate to the tasks humans excel at.”

Manufacturing: Resilience meets responsibility

If AI is reshaping decisions, according to speakers across sessions, manufacturing is being challenged to keep pace.

Proctor emphasised that fixed-dose products have long languished in slow pipelines and argued for greater autonomy in plants to reduce lag time between development and commercial supply.

Jagjit Singh Srai, director of research and head of the Centre for International Manufacturing at the University of Cambridge, highlighted the shift from batch to continuous manufacturing, describing micro-factories as a way to meet modern expectations for immediacy. “We have next-day delivery in e-commerce. This is the world we are working in – the expectation of immediacy,” he said.

For Lisa Farrell of the National Robotarium, robotics are about removing the “dull, dirty, and dangerous” aspects of work. Peter Deegan, business development manager at PACE, added that automation helps reduce the inevitability of human error: “The human condition means that at some point we will fail.”

But sustainability loomed large. GSK’s Cammy Mitchell pointed to the tension between flexibility and sustainability: modular equipment and cleaner changeovers will be essential, but he warned that pharma’s inherent conservatism slows progress. He also highlighted the carbon footprint of inhalers as an urgent case for change: “One puff from a Ventolin inhaler emits more CO₂ than a car trip.” It was a striking reminder that innovation is not just about speed, but also responsibility to patients and the planet.

A sector at a crossroads

Perhaps the biggest takeaway to emerge from PharmaTech Integrates 2025 was not a single vision, but a tension: Scotland proving that innovations can be deployed quickly, while the UK overall risks being left behind.

On one side, Scotland continues to validate what many have long preached: cross-sector collaboration, fast pipelines, AI-enabled diagnosis, and therapeutic innovation can shape patient care at scale. Dominiczak’s case studies – AI cancer screening, dermatology triage apps, and genetic testing that spares newborns from deafness – illustrated how this concept is already benefitting patients in the real world. On the other, industry veterans’, including Proctor’s, warnings highlighted what happens when systemic inertia prevails: investment drains away, and patients wait longer for proven treatments.

For all the talk of AI, continuous manufacturing, and personalised dosing, the underlying question remained the same: can the UK move beyond pilots and fragmented progress to system-wide transformation?

Ultimately, the most telling reframing came from Matt Winterman: “Everybody learns. But, smart people learn from other people.” In short, the next stage won’t come from isolated success stories alone – it will come from Dominiczak’s idea of the “triple helix” – systems, governance, and leadership that allow the entire UK to scale.