🏥 AI in the NHS Newsletter #1

May 31 - June 7, 2025 | Issue #1

Brought to you by Curistica - your healthtech innovation partner, 

helping builders and buyers work faster, safer, and smarter.
For help with clinical safety, data protection, and governance of Clinical AI, 

visit www.curistica.com or contact hello@curistica.com 

🎯 TL;DR - The Week That Was

A week of passionate debates on 24/7 online consultations, data governance battles over NHS Foresight, and the eternal question: "Is medicine an art or a craft?" Plus, AI scribes, clinical safety concerns, and why Blair's AI doctors might not be the solution we're looking for.

🔥 Hot Topic: The Great 24/7 Online Consultation Debate

This week's most heated discussion centered around whether GP practices should offer 24/7 online consultation access. The group was split into two camps:

Team "Always On"

  • The User Experience Advocates: Led by tech-minded voices arguing that patients should be able to submit non-urgent queries at any time (even if it's just a fungal nail infection at 2am!)

  • Key argument: "Anything that can alleviate the 8am rush is a good thing" - why restrict platforms when scarcity perception drives poor working conditions?

Team "Boundaries Matter"

  • The Clinical Safety Champions: GPs emphasizing that unmonitored overnight submissions create genuine risks

  • Dave Triska's reality check: "I've handled >200,000 OCs personally. I am never surprised at patients' ability to run towards danger quite merrily"

  • The evidence: Patients submit "wildly dangerous queries" overnight instead of calling 111/999

The Nuanced Middle Ground

Some suggested hybrid approaches:

  • AI triage to separate urgent from routine submissions

  • Automatic redirection of urgent cases to out-of-hours services

  • "Send at 8am" functionality like Microsoft's delayed email

Quote of the debate: "What core GP problems are critical to submit in the twilight hours? If we can't come up with a reasonable list, what problem are we solving?" - Dave Triska


🚨 Data Drama: NHS Foresight Project Paused

NHS England paused their Foresight AI project after the BMA and RCGP raised concerns about the use of GP data collected for Covid-19 research being used to train the AI model without their knowledge. The group's reaction revealed deep tensions around data governance:

The Facts

Foresight is a "groundbreaking AI initiative" trained on de-identified NHS data from 57 million people in England, designed to predict future health outcomes like hospital admissions and new diagnoses. The BMA and RCGP said they were "not aware that GP data, collected for Covid-19 research, was being used to train an AI model" and questioned whether "correct processes were followed to ensure data was shared in line with patients' expectations".

The Governance Gap

The Professional Advisory Group had approved the British Heart Foundation consortium's use of pandemic data specifically for Covid-19 research under emergency directive, but was never consulted about large language model training. The Joint GP IT Committee asked NHS England to refer itself to the Information Commissioner "so the full circumstances can be understood".

The Group's Divide

  • Innovation advocates saw this as "potentially groundbreaking research stopped because some people are too risk averse"

  • Data protection champions emphasized the importance of proper oversight, with one member noting previous controversies: Care.data (2013), GPDPR (2018), now Foresight (2025)

  • The middle ground recognized both the life-saving potential AND the need for robust safeguards

The Broader Context

This comes as former PM Tony Blair called it "absurd" that NHS data wasn't routinely available for innovation, while Tech Secretary Peter Kyle highlighted the UK's "extraordinarily powerful and rich data set".


Memorable quote from the group: "Data privacy costs lives... I think the balance is totally wrong! We use edge case scenarios to scare people and block progress."

🎨 Philosophy Corner: Is Medicine an Art or a Craft?

A thoughtful discussion emerged about whether medicine is an art form, sparked by Craig Nikolic's observation about a senior GP who could "tell a lot about a big sigh from a regular patient."

The Art vs. Craft Debate

  • Traditional view: Medicine as an art requiring intuition, experience, and human connection

  • Modern perspective: Medicine as a reproducible craft that can be standardized globally

  • AI's impact: Risk of turning senior "art" into "transactional grind"

Craig's Warning About AI in Medical Education

"The progression from junior to senior is often matched by progression from science to art... Too many AI products reduce the art to transactional. Junior folk will not learn the foundation to become that artful senior."

💻 Tech Talk: The Scribe Revolution (With Reservations)

AI scribes featured prominently, with mixed enthusiasm:

The Promise

  • Potential to free up time for genuine patient interaction

  • Could "rehumanize" healthcare encounters

  • Real productivity gains reported by early adopters

The Concerns

  • The happiness paradox: Time saved might just mean seeing more patients faster

  • Audio retention debates: Should recordings be kept for clinical safety?

  • The authenticity question: Multiple "versions" of truth from clinician, patient, and AI scribe

🏥 System Struggles: IT Chaos and Resource Realities

The week included some sobering reminders of NHS realities:

IT Horror Stories

  • Entire GP practice unable to work due to simultaneous computer updates during patient appointments

  • The eternal EMIS complaints (a group bonding experience!)

  • Windows 11 rollouts causing widespread disruption

Resource Reality Check

From Dr. Amar Ahmed: "Primary care delivers over 90% of all NHS appointments on less than 6% of the budget... Around half the population is seen in primary care each month."

🎪 Lighter Moments & Group Dynamics

The EMIS Solidarity

When someone suggested getting rid of EMIS, the group found rare unanimous agreement!

Tech Bro Teasing

Good-natured ribbing of the "tech bros" in the group, with reminders that profit isn't inherently evil when it drives innovation.

The Musk-Trump Tangent

A brief but entertaining discussion about the Musk-Trump relationship breakdown and its implications for AI development.

📊 By the Numbers

Weekly Activity Stats

  • Total Messages: 400+ messages over 7 days

  • Peak Day: Friday, June 6th (80+ messages)

  • Busiest Time: Morning discussions (8-10am) and evening winds-down (6-8pm)

  • Most Active Day for Debate: Saturday, June 7th (24/7 OC discussion)

🏆 Top Contributors (Message Count)

  1. Dave Triska - 45+ messages (The Clinical Reality Check Champion)

  2. Dr. Amar Ahmed - 35+ messages (The Productivity & Economics Voice)

  3. "Me" (PT) - 30+ messages (The Tech-Clinical Bridge Builder)

  4. Gus Kennedy (447473596596) - 25+ messages (The Pragmatic Problem Solver)

  5. Farhan Amin - 20+ messages (The Systems Thinking Advocate)

🔥 Hottest Debate Topics (by message volume)

  1. 24/7 Online Consultations - 60+ messages (spanning 2 days)

  2. NHS Foresight Data Project - 25+ messages

  3. AI Scribes & Clinical Documentation - 20+ messages

  4. Art vs. Craft in Medicine - 15+ messages

  5. IT System Failures & EMIS Complaints - 12+ messages

📈 Engagement Patterns

  • Longest Single Thread: 24/7 OC debate (lasted 8 hours across Saturday morning)

  • Most Controversial Take: "Data privacy costs lives" discussion

  • Most Unified Response: Universal EMIS hatred

  • Most Philosophical: Craig Nikolic's thoughts on medical education progression

  • Most Practical: Auto-filing results and triage system implementations

🎯 Discussion Quality Metrics

  • Evidence-Based Arguments: 85% of posts included personal experience or data

  • Cross-Specialty Perspectives: GPs, tech developers, policy experts, and clinical informaticians

  • Constructive Disagreement: High - passionate but respectful debate throughout

  • Action Items Generated: Multiple (conference attendance, demo requests, system implementations)

🔮 Looking Ahead

Upcoming Events Mentioned

  • June 5-6: "25 years of engaging communities to improve health and their records" conference in Glossop

  • July 14: Farhan's implementation of new systems

Ongoing Themes to Watch

  • The evolution of Crown indemnity for AI tools

  • Development of Class 2a medical device AI solutions

  • The future of GP partnership models and data controllership

💬 Quote Wall

"You get hungry if you start picking" - Wisdom from the group

"FATPOA, innit?" - Dr. Amar Ahmed on NHS economics

"Bloody tech bros" - The affectionate group term for innovation advocates

"Humans will quickly find the online consultation that asks the fewest questions" - Universal truth about user behavior

🎭 Group Personality Snapshot

This week revealed the group's unique character: a blend of clinical pragmatism, technological optimism, and healthy skepticism. The discussions showed deep expertise across clinical practice, health tech, and policy, with members challenging each other while maintaining mutual respect. The group clearly values both innovation and patient safety, even when those priorities create tension.

Whether debating the philosophy of medical practice or the practicalities of overnight triage, this community continues to grapple with the fundamental question: How do we use AI to enhance healthcare without losing the human elements that make medicine meaningful?

Next Issue: Will cover June 8-14, including any developments on the Foresight project and continued evolution of the great online consultation debate.

Want to contribute to next week's discussion? The group is always welcome to passionate, evidence-based perspectives on AI in healthcare - tech bros and skeptics alike!


Previous
Previous

🏥 AI in the NHS Newsletter #5

Next
Next

🏥 AI in the NHS Newsletter #2