How a Cardiff dietitian got 88% of a Type 2 diabetes group to the finish line

A Type 2 diabetes diagnosis is most reversible in the first months, when daily food and activity changes do the most work. The landmark DiRECT trial, run for Diabetes UK, found that nearly half of participants on an intensive weight-management programme were in remission at one year. Rhian Morgan, a registered dietitian in Cardiff, built a 12-week group around the same principle — and watched it leak members on email and a spreadsheet, finishing barely half the people who started. One Bitir cohort took completion from 52% to 88%.

Where she started

Rhian spent nine years in NHS dietetics across South Wales before setting up Tŷ Maeth in Cardiff, where much of her work is with people in the months after a Type 2 diabetes diagnosis — the window where structured change matters most. She runs the programme as a group, not one-to-one, partly on cost and partly on conviction: people changing how they eat do better alongside others doing the same.

The clinical design was sound — twelve weekly sessions, each with a focus, each with home practice on food, movement, and glucose monitoring. The problem was everything between the sessions.

Her coordination ran on three things:

• A weekly group email with the session recap and the week's home tasks, which a third of members never opened.
• A shared spreadsheet for logging daily readings that only the most organised members ever filled in.
• A WhatsApp group that quickly became a place for general questions and went quiet on the actual logging.

She could send the right guidance. She had no idea who was acting on it until someone arrived at a session with worse numbers and admitted they had not logged a meal in three weeks.

The breaking point

One cohort brought it to a head. A member in her early fifties, diagnosed three months earlier and genuinely motivated, drifted out at week five. She had stopped opening the emails, never managed the spreadsheet, and quietly decided she was failing — when in fact her early readings had been improving.

By the time Rhian realised she had gone, the window to catch her had closed. The member did not come back, and her diabetes care returned to twice-yearly GP reviews.

"I was giving people the right plan and then losing them in the gap between sessions," Rhian says. "The woman who left at week five wasn't unmotivated — she was invisible to me. I couldn't see she was drifting until she'd already gone, and by then a clinical opportunity that only really exists in those first months had passed."

What she set up in Bitir

Rhian rebuilt the programme as a single Bitir cohort, with one aim: make the daily between-session work visible without turning herself into someone chasing twelve people for spreadsheet entries. The structure follows the logic we set out in our guide to progress tracking — measure the daily behaviour, not only the blood result that lags it.

1. One private group per cohort. All 12 members in one closed space, replacing the email list and the WhatsApp group, with each week's session material pinned rather than buried.
2. A personal goal per member. Each person writes their own outcome for the 12 weeks — a target weight, a steps figure, a fasting glucose range their GP agreed — visible to that member and to Rhian, never to the rest of the group.
3. A daily food-and-glucose check-in. A short recurring check-in where members log the day's readings and whether they hit the day's eating plan. Rhian sees a per-member view, so a member who stops logging shows up within a day.
4. Weekly assignments tied to the session. Each week's home practice goes out as a structured assignment with a due date, so it lands as a clear task rather than a paragraph in an email.
5. Private check-ins on the members who slip. When the daily log goes quiet for someone, Rhian sends a short private message — never a public nudge — the approach we argue for in how to stop members drifting out of a group.

Member privacy mattered more here than in most groups. A diabetes diagnosis is sensitive, and members log personal health numbers daily, so individual data stays between each member and Rhian — the cohort only ever sees aggregate encouragement, never another person's readings.

What changed in the numbers

Rhian compared the last email-and-spreadsheet cohort with the first full Bitir cohort, each starting with 12 members.

The compliance figure is the one Rhian watches, because the daily logging is what the clinical result rides on. The HbA1c drop matters most to members and their GPs, but she is careful about it: those bloods were taken by each member's own GP, the sample is the nine completers who chose to share, and the change reflects consistency rather than any new clinical advice from her.

What she puts the change down to, in order:

1. The daily check-in made the logging a 30-second habit instead of a spreadsheet chore, so people actually did it — and a member who logs daily is a member Rhian can coach.
2. The per-member view meant a quiet logger surfaced in a day, so the week-five drifter became a private message on a Tuesday, not a no-show in a month.
3. Self-set goals gave the daily effort a point a member owned, rather than a target handed down in a clinic.

Re-enrolment onto her follow-on 12-week maintenance group reached 71%, against a trickle before. A member who finished feeling seen signs up for the next stage; one who quietly faded does not.

What she would tell another practitioner

Rhian's advice to other dietitians and health coaches running group programmes is direct.

• The plan is rarely your weak point. Delivery is. A good programme that members can't stay inside finishes no one.
• Make the daily behaviour a two-tap log, not a spreadsheet. People will tap a check-in every day where they will abandon a workbook by week two.
• Watch the logging, not the bloods. The blood result tells you what happened months ago; the daily log tells you who needs you this week.
• Keep individual health data private by default. Members log honestly only when they know their numbers aren't visible to the rest of the group.

For the end-of-programme summary she now shares with members' GPs, Rhian leans on the approach in our guide to reporting coaching results — lead with what changed, keep individual data named only to the person it belongs to.

What's next

Rhian is piloting the same structure for a prediabetes prevention group, where the goal is to stop progression rather than manage a diagnosis, and is building a simple shared dashboard so a member's own GP can, with consent, see the aggregate programme outcome. She is also training a second dietitian to run a parallel cohort, with the per-member view making it straightforward to hand a group over without losing the thread.

Related reading: progress tracking for coaching groups: what to measure and how, and how to show group coaching results to clients and sponsors.