Spotting opportunities for national level service transformation
Diabetic Eye Screening
Context
Diabetic retinopathy screening (DES) is an essential NHS service offered free to people living with diabetes checking for early signs of a condition that can cause irreversible sight loss if not detected and treated early.
The programme operated across a complex landscape of multiple third-party suppliers, clinical teams, quality assurance providers and internal NHS stakeholders, each managing their piece of the service in isolation. Commercial contracts with key third-party providers were due to expire in 2026, creating both a business imperative and a design opportunity to rethink how the service worked before those decisions were forced.
I joined as part of a two-person user-centered design (UCD) team supported by a technical architect. brought in to map the end-to-end service, surface the systemic gaps, and work toward a desired to-be design that could inform the programme's next phase.
The Brief
Map the complete as-is service for diabetic retinopathy screening focussing on cohorting, understanding how patients, suppliers, clinical teams and programme managers interact with the service and identify opportunities for a to-be design that balanced user needs, clinical requirements and business constraints.
The Real Problem
The service wasn't broken because people were doing things wrong. It was broken because the system design made workarounds the only rational choice.
No single team could see the end-to-end service. The data used to identify who should be invited for screening was not entirely reliable as clinical coding practices varied significantly across GP practices. Cohort files were being managed across multiple suppliers and internal teams, creating duplication, inconsistency and risk that nobody could see from their own vantage point.
During the discovery and mapping of the as-is, we discovered something nobody had seen before: reporting was a repetitive process with no single team accountable. Multiple reports existed across the value chain but most were left unused or were duplicated by different providers. The result was wasted time, no clear reporting structure, and decisions being made on inconsistent data.
The problem wasn't a lack of data. It was a lack of shared visibility and shared ownership.
My Role
- Co-led service design and user research across the project
- Audited existing discovery artefacts and synthesised previous research
- Led stakeholder interviews across clinical, operational, supplier and programme teams
- Facilitated co-design workshops and reframing sessions using How Might We methodology
- Produced and iterated as-is and desired to-be service blueprints on Mural
- Led clinical feedback sessions to stress-test proposed design ideas with clinicians closest to the service
- Co-designed reporting structure directly with end users after discovering reporting had no single owner
- Presented findings and artefacts through a series of playbacks with programme and clinical stakeholders
Considerations
Working at the intersection of transformation and delivery
Our work sat at the intersection of transformation and delivery simultaneously. Our team brought an outside-in perspective, centring user needs and systemic insight while ensuring the research directly served the programme's goals and milestones. The two had to happen in parallel, not in sequence.
Commercial sensitivity
We were researching across third-party suppliers who had commercial relationships with NHS England. Getting honest insight without jeopardising those relationships required careful framing of research questions and deliberate stakeholder management throughout.
Designing for what was genuinely achievable
The to-be design couldn't be a blank slate, it had to work within the constraints of a live national programme. Research findings had to be grounded in what was genuinely achievable, not just desirable.
Multiple users with competing priorities
Our users weren't just patients, they were suppliers, GPs, clinical teams and programme managers, each with different priorities and pain points. Balancing these without losing sight of the end patient was a constant challenge.
Working under time pressure with real consequences
With contracts expiring in 2026, we didn't have the luxury of doing everything in the right order. Decisions about what to prioritise had real commercial and contractual consequences.
Our Approach — From Existing Knowledge to Co-Created Solutions
Phase 1 — Start with what we know
Rather than starting from scratch, we began by auditing everything that was already known. Previous teams had conducted discovery on diabetic eye screening before us. Rather than asking stakeholders to repeat themselves, we reviewed all existing artefacts, synthesised pain points that had already been identified, and built our understanding before going back with questions.
This respected stakeholder time, helped us build social capital quickly, and meant we could move faster because we weren't reinventing work that had already been done.
We synthesised everything visually on Mural, producing a high-level as-is and desired to-be end-to-end journey that gave every team a shared starting point, regardless of specialism. This artefact became the foundation that multiple outcome teams could zoom into for their specific area of the service.
Phase 2 — Reframe and co-create
From there we zoomed in, conducting thematic analysis of pain points across the cohorting phase and categorising them into five themes. We then reframed them as How Might We statements, transforming problems into design opportunities and opening up creative possibility.
Internal democratic voting sessions prioritised ideas before we tested them. We then ran a series of co-creation playbacks with programme stakeholders; not presentations, but genuine stress-tests, with the people closest to the service challenging, refining and building on our thinking in real time.
Phase 3 — Translate ideas into journey design
We translated the strongest ideas into a first draft of the future cohorting journey. Using Mural, we developed several rough iterations, deliberately avoiding premature solutionising and keeping the design system-agnostic, so it could stand independently of existing technology rather than being retrofitted to current products.
Through iteration, we merged the strongest ideas and discarded others after assessing feasibility. The outcome was a first-pass journey map capturing the experience of key users: participants, clinicians, diabetic eye screening providers and QA teams.
We engaged technical teams, product leads and business analysts in parallel to validate both the user-centred design and technical feasibility simultaneously.
Phase 4 — Validate with clinicians
Before finalising the to-be design, we ran a series of clinical feedback sessions to stress-test our proposals with the people closest to the service.
The sessions validated our core proposals, particularly around improving cohort accuracy using additional data sources, and the appetite for a centralised repository that could eventually support broader diabetes care. They also surfaced a critical constraint that desk research alone couldn't have revealed: that even well-designed solutions would only succeed if the wider system supported adoption. Technically sound ideas and real-world adoption are not the same thing.
This fundamentally refined our thinking, not just about what to design, but about what would actually work in practice.
Phase 5 — Co-design reporting with end users of reports
Discovery on one of the first providers in the value chain revealed a problem nobody had explicitly named: reporting was a repetitive process with no single team accountable.
Multiple reports existed across the system but most went unused. End users didn't know which to trust, so they pulled their own, creating inconsistency, wasted effort and decisions made on unreliable data.
We co-designed the reporting structure directly with end users. We structured what providers could do, evaluated system capabilities, and built a shared understanding of reporting ownership that hadn't existed before.
What We Proposed
Three core proposals emerged for transforming the DES cohorting service:
Improve data confidence
Validate who should be invited for screening using multiple data sources rather than relying on a single, variable coding system — reducing the risk of eligible people being missed or incorrectly included.
Close the feedback loop
Enable data captured during screening appointments to flow back to the teams responsible for cohort management — so the system learns from what happens at the point of care rather than operating in a one-way direction.
Create shared ownership
A centralised repository and clearer reporting structure so no team is working from their own version of the truth — and so the service can be understood and managed as a whole, not just in parts.
What Changed
A culture of working together
Before this work, teams were doing their best within their own area of expertise. After it, co-design became the way of working, not an exception to it. The project shifted how teams related to each other, not just what they designed.
Trust between transformation and programme
The relationship between the transformation team and the programme team shifted fundamentally; from a team presenting a vision to stakeholders approving or rejecting it, to genuine consultation and shared ownership. That trust became the foundation for everything that followed.
Reporting that people actually used
Co-designing the reporting structure with end users meant that for the first time, providers had a clear, accountable reporting process. Unused reports became used ones. Wasted effort became structured insight.
A shared artefact that outlived the project
The as-is and to-be journey map became the starting point for multiple outcome teams across the programme, something teams could zoom into for their specific area rather than starting from scratch.
Reflection
"Just because you don't provide the whole service, doesn't mean you're not responsible for the outcome." — Lou Downe
This project taught me that the most valuable thing a design team can sometimes do is not the artefact they produce-but the conversations they create, the silos they break, and the shared language they leave behind.
The to-be design didn't immediately become the next phase of delivery. But the relationships, the shared understanding and the culture of co-design it built did. That's an outcome worth measuring too.