Reed Wellbeing compressed the time between diagnosis and first step

Reed Wellbeing delivers the NHS Diabetes Prevention Program (Healthier You) across seven regions of England, supporting people identified by their general practitioner as being at risk of developing Type 2 diabetes. The clinical case for speed of first contact had long been established by the NHS: the window immediately after a GP diagnosis is the moment a patient is most open to behavior change. By 2023, with referral volumes rising and the active backlog standing at approximately 15,000 people, Reed Wellbeing recognized that the next step in serving more people would call for a different operating model. In June 2024, following a successful three-region pilot, Reed deployed NiCE Proactive AI across all seven regions, autonomously orchestrating outreach across SMS, email, and voice from the moment a GP referral arrives. The wait between GP diagnosis and first patient engagement fell by 80 percent. Time to program start fell from 28 days to five. Reed Wellbeing is now the highest-performing provider on the program in the UK.

Before

A national-scale public health mission running into the limits of a manual outbound model

Reed in Partnership is a trusted provider of public services, working with governments and local authorities to deliver solutions to some of the most important challenges society faces. Reed Wellbeing is the division that carries that mission into preventative health, with a purpose grounded in helping people make changes that last a lifetime. Since 2016, Reed Wellbeing has been the provider of Healthier You, the NHS Diabetes Prevention Program, across seven regions of England.

The scale of the challenge is significant. Diabetes UK estimates that 6.3 million people in the UK are at increased risk of developing Type 2 diabetes. The condition is a leading cause of preventable sight loss in working-age adults and a major contributor to kidney failure, heart attack, and stroke. Treating diabetes costs the National Health Service approximately £10.7 billion a year, around six percent of its entire budget. The wider cost to the UK economy, accounting for absence, ill health, and economic inactivity, runs to an estimated £74 billion a year.

The nine-month Healthier You program supports people identified by their general practitioner as being at risk of developing Type 2 diabetes, guiding them through behavior change, nutrition, and physical activity with the goal of reducing that risk by around 20 percent. More than 350,000 people have been referred to Reed’s delivery since launch. More than 170,000 have been onboarded by the team in Reed Wellbeing’s Liverpool contact center.

The clinical case for speed had long been understood. The NHS recognized that the window immediately after a GP diagnosis is the moment a patient is most open to change. Motivation is highest. The doctor’s words are still ringing. Reach someone in that window and they are far more likely to act. Let it close, and urgency fades.

Reed Wellbeing’s contact center had been built to do exactly that. From 2016, the team made up to three outbound call attempts within thirty days of every GP referral, working through a continuous queue across seven regions and 200 to 400 daily referrals. The model worked at the scale it was designed for. By 2023, with referral volumes continuing to rise, the average wait between GP diagnosis and first patient contact had grown to 28 days. The active referral backlog stood at approximately 15,000 people. The structural ceiling of a manual outbound model had become clear, and the next step in serving more people would call for a different approach.

Opportunity to change

A pilot in three regions, a hypothesis ready to test

In 2023, Reed Wellbeing commissioned a review of its contact center operation to identify where improvements could be made. The review surfaced an introduction to Contact Engine, now NiCE Proactive AI, an intelligent, proactive conversational AI capability that had not yet been deployed at this scale in NHS-commissioned preventative healthcare.

Nick Hughes, Divisional Director for Reed Wellbeing, met with the NiCE team in Liverpool. The framing that came out of that meeting was decisive: AI not as a replacement for human contact, but as the bridge that made human contact possible at the right moment, with the right person, ready to act. Reed Wellbeing’s mission had always been human-centered, and the coaches delivering the nine-month program along with the agents guiding participants into it depended on human connection. What was described worked alongside that connection, absorbing the outbound contact burden so that people could focus on the conversations that actually required them.

There was candid acknowledgment on both sides that the approach was unproven. The application of this technology to NHS-commissioned preventative healthcare at this scale was genuinely new ground. What tipped the decision was the logic of the NHS hypothesis itself: the faster a patient is reached after a GP diagnosis, the more likely they are to act. If the technology could compress that gap, more people would join the program. Reed Wellbeing ran a three-month pilot from January 2024 across three deliberately selected regions, mixing urban and rural geographies to test how participants in different settings would respond.

By the end of three months, early signs of greater program uptake were visible. Reed entered a full contract. In June 2024, the solution was live across all seven regions.

The solution

An orchestrated journey, built to get smarter

The transformation Reed Wellbeing implemented was not a swap of outbound calls for automated messages. It was a redesign of the entire early-stage participant journey, with every touchpoint from GP referral to first session attendance connected, coordinated, and continuously improved based on what participants and contact center advisors experienced in practice.

How the journey now runs

From the moment a GP referral arrives, the system initiates the participant journey autonomously. Without waiting for an advisor to trigger it, the system determines which channel to try first, sequences its contact attempts, interprets how a participant responds, and decides the appropriate next step. Where complexity arises that the system cannot resolve, the participant is routed to a human advisor with full context preserved. The contact center steps in when its expertise is needed. The system handles everything else.

Outreach goes out across multiple channels: SMS, email, and automated voice. Participants are presented with course options, including face-to-face sessions near home and fully digital alternatives through the Second Nature app. Each contact attempt gives the participant the option to enroll, gather more information, or opt out. Participants respond in natural language. The system interprets intent and guides them through the next step.

Eligibility is not assumed. Each participant is screened against the details submitted by their GP, with inconsistencies queried in real time before a booking is confirmed. Once enrollment is complete, automated reminders go out ahead of the first session, including venue details and the name of the assigned coach.

What the people taught the system

The solution that exists today is meaningfully different from what launched in January 2024. Reed Wellbeing and NiCE met twice weekly throughout the early phase of deployment, reviewing the participant journey together and building improvements from what the data and the people revealed.

Six months in, a formal evaluation gathered feedback from participants, contact center advisors, and program coaches simultaneously. One finding stood out: some participants, particularly older ones, missed speaking to a person in the early stages of their journey. A one-click option was added within the outreach message, connecting a participant directly to the contact center as a priority inbound call, bypassing the standard queue with full context preserved. Automation and human support became a choice, not a trade-off.

Pilot data from Manchester showed that younger participants needed to fit the program around working hours, not just home locations. A workplace postcode option was added alongside the home postcode, giving participants course options near where they work. In more rural areas, route planning details were built into the outreach flow. Language adaptations were introduced for communities where English was not the first language. Each change came from evidence, not assumption.

What Reed Wellbeing and NiCE built over eighteen months was not a fixed deployment. It was a living system. Every piece of evidence made it better. Every person who used it taught it something new.

Results

The NHS hypothesis, proven at scale

The results across Reed Wellbeing’s seven regions since full deployment in June 2024 confirm what the NHS had long hypothesized: speed of first contact drives program uptake. They also exceed what Reed anticipated when the pilot began.

The wait time between GP referral and Reed’s first patient engagement has fallen by 80 percent. A participant referred by their GP today can start a face-to-face program within five days, where the same journey previously took an average of 28 days. The time between referral and first session attended has reduced by almost 30 percent. The active referral backlog has fallen from approximately 15,000 people to around 3,500, shifting Reed Wellbeing’s contact center from four to five months of work in hand to approximately one.

The number of participants joining the service has increased by 16 percent. Referral-to-start conversion has risen from the 50 to 55 percent range to 62 to 64 percent across the seven regions, reaching 70 percent in Manchester. Reed Wellbeing now converts referrals to program starts at a rate ten percentage points above any competitor in the market and is the highest-performing provider on the program. Digital program take-up has grown by 22 percent, reflecting both the expanded options the system surfaces for participants and the ease of the digital onboarding pathway.

The inbound calls they now handle come from participants who want support, have questions, and are ready to engage. Conversations are longer, more meaningful, and squarely within the expertise of the people having them.

Advisor satisfaction scores on Reed’s quarterly staff survey have moved from a range of 7.8 to 8.0 out of 10 to a range of 8.5 to 10. No staff were made redundant. As outbound call volume fell, natural attrition reduced further and the team stabilized. The work became the work it was always meant to be.

The Future

Diabetes was a starting point, not the finish line

The next phase extends the technology beyond the enrollment journey and into program retention. Participants who fall off the program will receive proactive re-engagement, with the system identifying dropoff early and prompting return before absence becomes withdrawal. The full nine-month behavior change journey, not just the first session, becomes the scope of the technology’s reach.

Three additional NHS workstreams are being brought into the model. The Behavioural Support for Obesity Prescribing program, introduced alongside weight-loss injection prescriptions in England, is already using the technology to reach and inform eligible patients. A longterm conditions pilot is applying the same engagement model to people with high blood pressure across two initial areas. The Type 2 Diabetes Remission program, a more intensive twelve-month pathway for newly diagnosed patients, is being scoped for integration, with the scheduling complexity of one-to-one delivery being addressed as part of the design.

What began as a solution to a contact center backlog has become the operating architecture for a different kind of healthcare engagement: faster, more personalized, more continuous, and built for the moments when people are most ready to change.