Stroke Improvement Group (SIG) Perspective
Document Date: 19 February 2026
Reviewer: SIG Leadership
Deck: LetsFIXtheNHS – Intelligence & Action Platform (Q1 2026, 15 pages)
—
1. KEY THEMES SUMMARY
T1: Complexity requires experimental, not linear, approaches
Cross-references: S2.6, S8.1, S13.2
The deck presents what are fundamentally complex problems (in Cynefin framework terms) but uses largely linear, plan-driven language. Complex problems require:
- Probe–sense–respond cycles, not design–implement–defend
- Minimum Viable Solutions (MVS) with rapid iteration (“MVP and pivot” / “suck it and see”)
- Explicit acknowledgment that we don’t yet fully understand either problems or solutions
The 6-phase lifecycle (Slide 8) appears linear despite “iterate” in step 6. The roadmap (Slide 13) phases sequentially over years 1–4+. Reality will be messier, more parallel, more experimental.
Recommendation: Frame the model explicitly as probe–sense–respond architecture for complex systems, not as a linear plan.
—
T2: Policy and cultural barriers are structural, not just informational
Cross-references: S2.1, S2.2, S2.3, S2.4, S7.2, S9.4, S12.3
The deck implies that better intelligence → better decisions → system change. But the real barriers are:
- Frozen policy frameworks: Good principles implemented inside outdated constraints; as context evolves, policy doesn’t. Need mechanisms to adjust policy/process while preserving principles.
- Professional worldview dominance: Clinicians assume they lead and frame problems clinically; alternative framings (lived experience, systems, community) are structurally excluded.
- Blame culture and risk aversion: Decision-makers hide behind “the system” because blame is applied with hindsight. No one is rewarded for running experiments that might fail on the way to breakthroughs.
- Headline-driven politics: Media and political pressure preserves imperfections; boldness is punished.
- Career fiefdoms and budget silos: Service integration across the stroke thriver’s journey cuts across promotion profiles, fiefdoms, and budget allocations – architecturally enshrined misalignment (S7.2).
Ingrained organizational self-interest (S9.4): Organizations like the Stroke Association may constrain progress to preserve donation revenue and salaries. The “proof of concept” is whether this platform can “put a bomb under” those constraints.
Recommendation: Address policy reform, cultural change, and power redistribution explicitly, not just intelligence infrastructure. Secretary of State-level intervention may be required (S7.2).
—
T3: No single source of truth in complex systems
Cross-references: S3.1, S3.3
The deck positions LetsFIXtheNHS as capturing “honest intelligence” vs “filtered” NHS data. This is valuable, but risks implying a single source of truth.
In complex systems, multiple perspectives can be simultaneously valid and correct within their contexts. Different observers (clinicians, patients, carers, managers, policymakers) see different truths.
Recommendation:
- Reference DIKW hierarchy (Data–Information–Knowledge–Wisdom): intelligence platforms help navigate from data to actionable knowledge, but don’t eliminate multiple valid interpretations.
- Reference bounded rationality (Herbert Simon): decision-makers can’t process all information; intelligence must respect cognitive limits and help people act despite complexity, not pretend to remove it.
- Frame as “multi-perspective synthesis” rather than “the truth.”
—
T4: Stroke Improvement Group (SIG) as the missing implementation layer
Cross-references: S4.1, S6.1, S13.1
The deck describes field teams, media, and AI but doesn’t clearly show how therapeutic interventions, digital peer support, and community-led recovery are integrated.
SIG is a coalface where:
- Therapeutic interventions happen
- Data-driven storytelling is captured
- Instrumented rehab devices generate data
- Recurrent neural networks and large language models process free-form text alongside device data
SIG could be the implementation of “field teams” for stroke – not vans gathering stories, but the actual digitally enabled peer support community network (DEPSCN) embedded in the stroke pathway.
SIG is public domain and global from day one (S13.1), not phased in over years.
Recommendation: Position SIG as the stroke-specific implementation model: the community platform where LetsFIXtheNHS intelligence infrastructure meets lived-experience-led recovery architecture.
—
T5: Rehab and long-term support need equal or greater emphasis than acute intervention
Cross-references: S9.1
The deck positions thrombectomy and BEFAST as “First Mission”, implying sequential focus. But:
- Rehab has many reasons why it needs just as much emphasis, if not more.
- We need parallel activity, not sequential.
- The “Total Stroke Pathway” language (Life in a Stroke Foundation) is correct, but the deck’s emphasis is heavily acute-focused.
Recommendation: Reframe as parallel missions across prevention–acute–rehab–long-term, not “first mission = acute, later = rehab.”
—
T6: Messaging is alienating for stroke survivors and doesn’t clarify benefit
Cross-references: S7.1, S12.1, S12.2
- London hub is alienating for most of the UK (S7.1).
- Sponsor tiers, gamification, prizes (Slide 12) are “pretty alien, unattractive messages for people who have had strokes” (S12.1).
- Benefit to stroke survivors and carers isn’t clear enough while costs (£5k/month sponsorship covering £2bn pathway investment?) don’t add up (S12.2).
- The deck looks very England-focused despite SIG being global (S7.1).
Recommendation:
- Decentralize language (not “London hub”).
- Reframe community engagement around intrinsic motivation (self-benefit, peer support, dignity) not extrinsic rewards (points, prizes).
- Clarify economic model and how sponsor revenue realistically funds pathway transformation.
—
T7: Key paradigm-shift elements are missing
Cross-references: S12.3
“Haven’t seen anything yet about the guidelines, standards, culture etc. that need to be created to affect a paradigm shift.” (S12.3)
The deck focuses on infrastructure (vans, AI, media) and business model (sponsors, SaaS). It does not address:
- Standards and governance for digitally enabled peer support communities (DEPSCN)
- Curriculum and professional competency changes (e.g. SSEF – Stroke-Specific Education Framework)
- Data ownership, consent, equity, safeguarding in peer-led digital communities
- Integration with formal pathways and role clarity (“professionals on tap, not on top”)
Recommendation: Incorporate the DEPSCN standards and governance work (SIG/ISRRA) as a foundational workstream, not an afterthought.
—
T8: NHS is a wicked problem, not a “whole system”
Cross-references: S13.2
Slide 13 describes “Whole System Scope: NHS + Social Care Day One.” But:
- The NHS is not a “whole system” – it’s a dynamic system of systems with emergent properties.
- This constitutes a wicked problem (Rittel & Webber, 1973) or even a super-wicked problem (characteristics: time is running out, no central authority, those seeking to solve it are causing it, policies discount future irrationally).
Recommendation: Use wicked/super-wicked framing explicitly. Reference Rittel & Webber (1973) and relevant Open University IBIS (Issue-Based Information System) work on navigating wicked problems.
—
T9: Critical factual error on thrombectomy penetration
Cross-references: S9.2
Slide 9 states “3 in 10 UK patients” have thrombectomy access (30%).
This is wrong by an order of magnitude. The real figure is 3% (possibly 5%), not 30%. Even in London, unlikely to be 30%.
Recommendation: Correct immediately. This undermines credibility with stroke clinical and policy audiences.
—
T10: Slides often function as presenter prompts, not standalone content
Cross-references: S5.2, S10.1, S11.1, S13.1
Multiple slides (5, 10, 11, 13) are described as “colourful but not standalone – might be a prompt for somebody who already knows the message, but on its own conveys nothing.”
Recommendation: Decide whether this is a leave-behind document or a presenter deck. If leave-behind, slides need to be more self-explanatory. If presenter deck only, that’s fine but clarify intended use.
—
T11: Real change comes from motivated individuals, not data alone
Cross-references: S4.4
The deck’s formula: “Coalface + Media + Data = Real Change” is overstated.
Reality: Those elements may deliver better inputs to the change process, but real change comes from motivated individuals who see benefit to themselves and act on it.
Recommendation: Reframe as “intelligence infrastructure enabling motivated actors” rather than “intelligence = change.”
—
2. DETAILED SLIDE-BY-SLIDE NOTES
Page 1 – Title Slide
S1.1: No comments at this time.
—
Page 2 – The Problem: Fragmented Intelligence
S2.1 – Frozen policy frameworks:
Policies are often based on good principles but implemented through the lens of current constraints. As context moves on, policy stays frozen. We need mechanisms to regularly adjust policy, processes, and procedures while preserving underlying principles, not just gathering more data about the problem.
S2.2 – Clinician-centric worldview:
Clinicians continue to assume they “lead” and frame problems within a clinical worldview. Alternative framings (system, community, lived experience) would reveal different solution spaces, but are structurally excluded.
S2.3 – Blame culture and decision-maker filibuster:
Decision-makers can hide behind risk because blame is applied only with hindsight, not credited for foresight. They avoid bold decisions, filibuster, and blame “the system” or policy for problems because they’re not bold enough to say: “We’ll try something, we know it will fail in some ways, but the learning will allow us to succeed.” Culture is built on newspaper headlines, which preserves imperfections rather than allowing improvements.
S2.4 – Media and political headline culture:
Newspaper and political headline pressure rewards blame-avoidance and punishes experimental, learning-oriented improvement.
S2.5 – Over-design in advance:
The system is predisposed to designing solutions in advance, but real design is only possible when we understand both the problem and potential solutions – and we often don’t. We need experimental processes focused on implementing change, not just providing cautious incremental tweaks to care.
S2.6 – Cynefin and complexity:
This is largely a complex domain (Cynefin framework: simple/complicated/complex/chaotic/unclear). Need explicit reference to probe–sense–respond approach for complex contexts, not “design–implement–defend” which is appropriate only for complicated or simple domains.
—
Page 3 – The Intelligence Advantage
S3.1 – No single source of truth:
The slide risks implying there is a single “source of truth.” In complex systems, multiple perspectives can be valid and correct within their contexts simultaneously. Different observers (clinicians, patients, carers, managers) see different truths.
S3.2 – Positive framing needed:
The “Current NHS Data” negatives are reasonable, but framing should also positively describe what is needed (A, B, C capabilities), not just what’s missing.
S3.3 – DIKW and bounded rationality:
This needs explicit reference to:
- DIKW hierarchy (Data–Information–Knowledge–Wisdom): intelligence platforms help people move from raw data to actionable knowledge, but don’t eliminate complexity or interpretation.
- Bounded rationality (Herbert Simon): decision-makers have limits on what they can process and know. Intelligence must respect those limits and help people act despite complexity, not pretend to remove it.
—
Page 4 – The Solution: Distributed Intelligence Network
S4.1 – SIG as coalface implementation:
The Stroke Improvement Group (SIG) is itself a coalface where both therapeutic interventions and data-driven storytelling happen, and where instrumented rehab devices generate data. SIG could be the implementation of “field teams” for stroke – the actual digitally enabled peer support community network embedded in the pathway, not external vans gathering stories.
S4.2 – Storytelling as sticky message:
The storytelling engine concept aligns with Malcolm Gladwell’s “tipping point” notion of sticky messages – messages that people remember and repeat.
S4.3 – AI layer needs specificity:
The “AI & Data” block on the slide feels generic. In the SIG model, there is something concrete: recurrent neural networks, large language models processing free-form text, plus data from instrumented rehab aids – a real, specific AI implementation.
S4.4 – Change requires motivated individuals:
“Coalface + media + data = real change” is overstated. Those may deliver better inputs to the change process, but real change comes from motivated individuals who see benefit for themselves and act on it.
—
Page 5 – How the Intelligence Network Operates (6-step loop)
S5.1 – London hub is alienating:
A “London hub” is alienating for almost everybody outside London. Language should be decentralized.
S5.2 – Icons add nothing:
The icons around the diagram are pretty but add up to nothing substantive on the slide – possibly prompts for the presenter rather than meaningful content for the audience.
S5.3 – Slide lacks substance:
The whole slide feels like it doesn’t say much of substance.
S5.4 – England-only appearance:
Despite mentioning 16 regions earlier, this looks very England-focused, whereas SIG is global, not just UK.
—
Page 6 – The Three Pillars in Depth
S6.1 – Field teams feel like PR exercise:
The field teams column looks like data gathering wrapped in a PR exercise. It was a surprise at this point in the deck, so the concept probably wasn’t explained clearly earlier. Assumed field teams gather data from the community, so for stroke within SIG architecture, this would be in addition to centralized community platform data gathering (not a replacement).
S6.2 – Redundancy:
Obviously you need to put data through an intelligence engine, but not sure this slide says much more than already said on previous slides.
—
Page 7 – National Rollout (16 Regional Hubs map)
S7.1 – National differences and alienation:
National (UK-wide) approach will involve administrative wrangles due to differences between England, Scotland, Wales, and Northern Ireland. Reinforces concern that this looks England-centric when SIG is global.
S7.2 – Integration across silos is key (government-level issue):
Major topic: The transformation of stroke needs to be integrated across silos, but this cuts across:
- Career-based promotion profiles
- Fiefdoms
- Budget allocations
Need service integration aligned along the journey that a stroke thriver takes (the patient journey). We have a systematically and architecturally enshrined service delivery model that does not match service user patterns. This is key and needs to be addressed at government and Secretary of State level.
—
Page 8 – From Intelligence to Action (lifecycle phases)
S8.1 – Lifecycle looks linear despite iteration:
Concerned that the lifecycle looks linear even though box 6 says “iterate”. Given the Cynefin framework (simple/complicated/complex/chaotic/unclear domains), if we’re to actually describe what will happen, we need:
- Minimum Viable Solution (MVS)
- Probe–sense–respond (Cynefin complex domain approach)
- Other vocabulary: “suck it and see” or MVP and pivot (Minimum Viable Product, then pivot based on learning)
The linear presentation doesn’t reflect the experimental, iterative reality needed for complex problems.
—
Page 9 – First Mission: Universal Thrombectomy & BEFAST
S9.1 – “First” implies sequential, need parallel emphasis:
Immediately reacting to the idea of “first” implying there is sequential activity. We need to be doing things in parallel. Rehab has many reasons why it needs just as much emphasis, if not more, than thrombectomy and BEFAST.
S9.2 – Critical factual error on penetration rate:
The slide says thrombectomy is available to “3 in 10 UK patients” (30%). This is wrong by an order of magnitude. The real figure is 3% (possibly 5%), not 3 in 10. Even in London, doubt it’s 30%. This error undermines credibility.
S9.3 – BEFAST vs FAST is correct:
True that BEFAST (adding Balance and Eyes to Face, Arms, Speech, Time) is an improvement over FAST alone.
S9.4 – Organizational self-interest as barrier:
Real challenge with BEFAST: it shows there are ingrained self-interests in organizations like the Stroke Association that stand in the way of progress. Proof of concept here is that we have to put a bomb under those who would constrain progress so they can keep asking for donations to pay their salaries.
—
Page 10 – Intelligence Revenue Model (Enterprise SaaS)
S10.1 – Not standalone:
Colourful, but not standalone – might be a prompt for somebody who already knows the message to speak to, but on its own it conveys nothing.
—
Page 11 – Business Model (Revenue Stack & Strategic Logic)
S11.1 – Not standalone:
As per previous slide: colourful, not standalone, conveys nothing on its own without presenter.
—
Page 12 – Sponsor Partnership Model
S12.1 – Alienating messaging for stroke survivors:
What’s this trying to say? That you’re going to get funding from the private sector and invite people to participate because they can win prizes? Those are pretty alien, unattractive messages for people who have had strokes.
S12.2 – Benefit unclear, costs don’t add up:
Don’t think sponsor/gamification model will appeal strongly enough to those motivated by self-interest – the benefit to them isn’t clear enough. The costs are also unclear: £5k/month sponsorship needs to cover an awful lot of people if it’s going to fund £2 billion (mentioned in other materials for Total Stroke Pathway investment).
S12.3 – Missing paradigm-shift elements:
Haven’t seen anything yet about the guidelines, standards, culture etc. that need to be created to affect a paradigm shift. Infrastructure and business model are necessary but not sufficient. Need explicit work on:
- DEPSCN (Digitally Enabled Peer Support Community Networks) standards and governance
- SSEF (Stroke-Specific Education Framework) curriculum changes
- Data ownership, consent, equity, safeguarding in peer-led communities
- Role clarity for professionals (“on tap, not on top”)
—
Page 13 – Transforming Health & Social Care (roadmap)
S13.1 – Presenter prompt, not standalone:
Back to: it provides a prompt to talk to, but not substantive on its own. Also, SIG is public domain and global from day one, not phased in over years 1–4+.
S13.2 – NHS is a wicked problem, not a “whole system”:
NHS is not a “whole system” as stated on slide. It’s a dynamic system of systems with emergent properties, meaning it constitutes a wicked problem (Rittel & Webber, 1973) or even a super-wicked problem.
Wicked problem characteristics (Rittel & Webber):
- No definitive formulation
- No stopping rule
- Solutions are not true/false but better/worse
- No immediate/ultimate test
- Every solution is a “one-shot operation” (consequences matter)
- No enumerable set of potential solutions
- Every wicked problem is essentially unique
- Every wicked problem is a symptom of another problem
- Discrepancies can be explained in numerous ways
- Planner has no right to be wrong
Super-wicked additions:
- Time is running out
- No central authority
- Those seeking to solve the problem are also causing it
- Policies irrationally discount the future
Reference note: There’s an Open University software product that implements the problem language (likely IBIS – Issue-Based Information System, based on Rittel’s work, but exact name needs confirmation).
—
Page 14 – Thank You / Contact
S14.1: Final slide is contact information (founder, email, phone). No substantive comments.
—
Page 15 – (If present, final back page)
Note: If deck has a 15th page beyond “Thank You,” add here. Otherwise, deck ends at page 14.
—
Education Framework)
Context: Educational curriculum reform to prepare healthcare professionals to work within and support digitally enabled peer communities, not just deliver traditional clinic-based care.
Application to LetsFIXtheNHS: Part of the paradigm shift needed (currently missing from deck). Professionals need new competencies: facilitation, community partnership, digital literacy, shared decision-making in peer contexts.
—
4. RECOMMENDATIONS SUMMARY
Immediate corrections:
- Fix thrombectomy penetration rate (Page 9): Change “3 in 10” to “3%” (correct by order of magnitude).
- Clarify slide numbering/page references for external audiences.
Strategic framing changes:
- Adopt Cynefin complexity framing explicitly: probe–sense–respond, MVP/pivot, experimental cycles – not linear planning.
- Reframe NHS as wicked/super-wicked problem (not “whole system”), with explicit acknowledgment of emergent properties and need for adaptive approaches.
- Add DIKW and bounded rationality to intelligence advantage framing (Page 3): intelligence helps navigate complexity, doesn’t eliminate it.
- Replace “single source of truth” language with “multi-perspective synthesis.”
Content additions:
- Add paradigm-shift section: Standards, governance, DEPSCN, SSEF, culture change (currently missing, noted at S12.3).
- Position SIG as stroke implementation model: The coalface where LetsFIXtheNHS infrastructure meets DEPSCN architecture.
- Parallel missions, not sequential: Rehab and long-term support need equal emphasis to acute intervention (thrombectomy/BEFAST).
- Address structural barriers explicitly: Policy reform, career structures, budget silos, organizational self-interest (Pages 2, 7, 9) – these require power redistribution, not just better data.
Messaging changes:
- Decentralize language: Remove “London hub,” ensure UK-wide (and acknowledge SIG global) framing.
- Reframe community engagement: Focus on intrinsic motivation (self-benefit, dignity, peer support), not extrinsic gamification (points/prizes) which alienates stroke survivors.
- Clarify economic model: How does £5k/month sponsorship realistically scale to fund £2bn+ pathway transformation?
- Make slides more standalone or clarify deck is presenter-support only (currently many slides function only as prompts, noted at S5.2, S10.1, S11.1, S13.1).
Realism adjustments:
- Reframe “coalface + media + data = real change” as “intelligence infrastructure enabling motivated actors” – change requires motivated individuals who see self-benefit, not just data.
- Secretary of State-level intervention likely required for service integration across silos (S7.2) – make this explicit.
—
5. NEXT STEPS
For LetsFIXtheNHS / Steve Podmore:
- Review factual correction (thrombectomy rate)
- Consider strategic framing shifts (complexity, wicked problems, parallel not sequential)
- Engage with SIG on positioning DEPSCN as stroke implementation model
- Clarify economic model and community value proposition for stroke survivors
For SIG:
- Prepare positioning paper: “How SIG DEPSCN plugs into LetsFIXtheNHS and Life in a Stroke”
- Offer to contribute paradigm-shift section (standards, governance, SSEF) to future deck versions
- Identify opportunities for pilot integration (e.g. Hybrid Hubs BHF bid)
For joint discussion:
- Role of SIG as global, public-domain stroke coalface vs LetsFIXtheNHS UK-centric, commercial platform – how these complement
- Data governance and beneficial ownership models (avoid PatientsLikeMe-type exploitation)
- Timeline: SIG public domain from day one vs LetsFIXtheNHS phased Years 1-4+ approach
—
Document prepared: 19 February 2026
For: SIG leadership review and onward discussion with LetsFIXtheNHS / Life in a Stroke Foundation
Author: Simon via Perplexity