Guide
A repeat supply is a clinical decision, not a rebill
Retention is the whole economics of an online clinic. It is also the part of the service most likely to fail an inspection, because the second, sixth and twelfth supply are usually handled by the billing system rather than by a prescriber. The subscription renews. The order goes out. Nobody made a decision.
This guide covers what a repeat supply has to involve, what the record has to show afterwards, and how to handle the two events that get confused constantly: a patient stopping treatment, and a card being declined.
The reassessment gate
Every repeat supply is a fresh clinical decision. That is not a Sync-RX opinion, it is the basis on which prescribing is judged.
The reference points prescribers are measured against come from the NICE appraisals, which cover NHS funding. They are not private-sector rules, but they are the published clinical thresholds, and a private prescriber departing from them should be able to say why.
- Semaglutide. NICE recommends considering stopping if less than 5% of initial weight has been lost after six months of the maintenance dose. It also recommends a maximum of two years of treatment, within a specialist weight management service providing multidisciplinary management. NICE TA875.
- Tirzepatide. If less than 5% of initial weight has been lost after six months on the highest tolerated dose, decide whether to continue, weighing benefits and risks for that person. NICE TA1026.
- Eligibility at the start. BMI thresholds, with a reduction of usually 2.5 kg/m² for people from South Asian, Chinese, other Asian, Middle Eastern, Black African or African-Caribbean backgrounds.
What that means operationally: a six-month point exists in every course where somebody has to look at the numbers and make a call. If your system cannot tell you which patients are at that point this week, you do not have a reassessment process. You have a billing schedule.
What ongoing follow-up has to produce as a record
The GPhC’s April 2026 themed review analysed 77 inspection reports and the concerns raised between January 2024 and December 2025. The weaknesses it names are almost all record weaknesses, not clinical ones:
- risk assessments and clinical governance arrangements
- incomplete or insufficient patient consultation records
- inconsistent processes for independently verifying weight and BMI
- poor communication with patients
- limited arrangements for ongoing clinical follow-up and customer support
The single sharpest line for anyone running repeats: prescribers must independently verify weight and BMI, not accept what the patient typed into a questionnaire. The review is published by the regulator as Weight management medicines and services: a review of GPhC inspections and concerns.
A repeat supply record that survives an inspection shows, for each cycle:
- When it was reassessed and by whom, with a name, not a system user ID
- An independently verified weight, and how it was verified
- Current BMI, calculated rather than declared
- Side effects and tolerance asked about and answered
- The dose decision: continue, increase, hold or reduce, with the reason
- Any flag that fired, and what the prescriber did about it. A prescriber may absolutely continue supply despite a flag. The record has to say why they did
- What the patient was told, and when
If a supply went out with no reassessment record, the honest reading is that the sale happened and the clinical decision did not.
Stopping, pausing and tapering
Stopping a GLP-1 is not dangerous in the way stopping some medicines is. There is no withdrawal syndrome to manage. The real issues are different:
- Appetite returns and weight typically rebounds. Patients should know this before they start, not when they cancel.
- Other medicines may need review. Blood pressure medicines, diabetes medicines and others adjusted downward during weight loss may need revisiting when the weight loss stops or reverses. This is the risk most cancellation flows miss entirely.
- Tapering is not established as better than a planned stop. Either is defensible. Neither is defensible without a plan.
So the requirement is not “keep them on it”. It is that a stop should be a planned clinical event with a record, the same as a start.
The distinction that matters: cancellation is commercial, stopping is clinical
These are two different events and most systems treat them as one.
| Event | What it is | What should happen |
|---|---|---|
| Patient cancels their subscription | Commercial | Cancel immediately, no hurdles. Then offer a stopping conversation, and flag any other medicines that need review |
| Patient stops treatment on clinical advice | Clinical | Documented stop plan, and the billing stops as a consequence |
| Card payment fails | Billing | Hold the order. Retry. Contact the patient. Do not treat it as a decision to stop treatment |
| Patient goes quiet at reassessment | Clinical | Supply pauses. Chase. Do not ship on the old decision |
The failed payment case is the one nobody writes into their terms. A declined card is a bank event. It says nothing about whether the patient should continue treatment, and it should never silently end a course.
It is worth reading your own terms against this. A pattern common across UK weight-loss providers is a clause saying orders cannot be cancelled or refunded once payment is taken, because they go straight to the dispensing pharmacy, with no separate provision at all for a failed payment. That clause is already awkward. It becomes considerably harder to hold once the DMCCA’s proportionate-refund rules for services are in force.
The cancellation mechanism itself
The DMCC Act’s subscription regime was brought forward on 10 August 2026 and is now expected to apply from January 2027, having previously been set for spring 2027. The secondary legislation and guidance are not yet published, so the detail below is the direction of travel rather than final law.
What is clear enough to build for now:
- Cancellation must be straightforward, with no unnecessary steps. In practice, a patient who signed up in two clicks online should be able to cancel in two clicks online. Not a phone call, not an email to support, not a retention script.
- Cooling-off rights on renewals. Two windows: an initial one when the contract starts, and a renewal one after a free or discounted trial ends or after a contract of twelve months or more auto-renews.
- Notices before auto-renewal, on a durable medium, with the purpose of the message immediately apparent to the reader. Email counts. A line buried in a monthly statement does not.
- Refunds. The government response confirms an approach to refunds for services; the mechanics are in the legislation still to come.
Two things you can do now without waiting for the statutory instrument: make cancellation genuinely one click, and make sure every renewal notice would survive being read out loud to a regulator.
Where Sync-RX fits
- Reassessment gates tied to the clinical record, not the billing cycle. Supply does not renew because a card is on file. It renews because a prescriber made a decision and it is recorded.
- Independently verified weight and BMI. BMI is calculated by the platform from verified figures, not accepted from a free-text box, with the Asian-background threshold adjustment applied automatically.
- A failed-payment path that holds the order. A decline pauses the supply and starts a retry and contact sequence. It does not lapse treatment and it does not quietly end a course.
- One-click cancellation, logged. However the patient signed up, they can cancel the same way, and the cancellation is timestamped in the audit trail.
- A stop is a clinical event. Cancelling triggers a stopping prompt: rebound counselling, and a review flag for any other medicines adjusted during treatment.
- An audit trail covering both sides. Commercial events and clinical decisions sit on the same patient timeline, so an inspector reading it sees one story, not a billing export next to a clinical export.
- Due-for-review reporting. Who is at six months, who has a flag outstanding, who has not submitted a verified weight, who is approaching the two-year point.
FAQ
Is the 5% rule a legal requirement for private prescribing?
No. It comes from NICE technology appraisals covering NHS funding. It is the published clinical benchmark, so a private prescriber who continues supply below that threshold should be able to show they considered it and why they continued. That reasoning belongs in the record.
Does NICE cap treatment at two years?
For semaglutide, TA875 recommends a maximum of two years within a specialist weight management service. The tirzepatide appraisal contains no equivalent cap. Do not apply a blanket two-year rule across every product.
What counts as independently verifying weight?
Not a number typed into a form. Options in practice include a video or timed photo of the patient on scales with the reading visible, connected scales, or a measurement taken by a healthcare professional. The GPhC's concern is with services that accept self-reported figures at every supply.
A patient's card failed. Can we just cancel them?
You can stop supplying, but do not record it as the patient choosing to stop, and do not let the treatment lapse without contact. A decline is a billing failure. Hold the order, retry, tell the patient, and if they do choose to stop, handle that as a stop.
Do we have to let people cancel online?
The DMCCA regime, expected from January 2027, requires cancellation to be straightforward without unnecessary steps. If sign-up was online, plan for cancellation to be online. Building it now costs less than retrofitting it in December.
Should we taper patients off?
There is no strong evidence tapering beats a planned stop. What matters is that a plan exists, the patient understands weight and appetite are likely to return, and any other medicines adjusted during treatment are reviewed.
Does stopping cause withdrawal?
No withdrawal syndrome as such. The risks are appetite and weight rebound, and medicines that were reduced while the patient was losing weight now needing review.
Book a walkthrough
Want to see reassessment gates working on real patient records?