Code of Conduct

1. Purpose and principles

This document sets out the Coalition for Responsible Digital Health’s (CoRDH) proposed proportionate, evidence-based safety standards for the distance prescribing of GLP-1 and GLP-1/GIP receptor agonists for weight management.

Patient safety is our first and overriding priority. CoRDH’s ambition is responsible digital health: ensuring patients can access effective, regulated care through modern digital services, supported by robust clinical governance, appropriate safeguards and clear accountability.

Our concern is not whether the remote prescribing of GLP-1s should be regulated - it should be - but whether current regulatory expectations are smart, consistent, proportionate to real clinical risk, and whether they support or inadvertently undermine patient safety in a rapidly evolving healthcare landscape.

Our aims are to:

• Reframe the safety profile of GLP-1 medicines, based on current evidence and real-world data.

• Describe the modern patient and service context in which these medicines are prescribed.

• Propose a clear, consistent and proportionate safety framework that enables prescribers to make safe, well-documented decisions using appropriate digital tools. This should be something the GPhC can accept and inspectors should apply consistently.

• Highlight areas where current inspection practice appears misaligned with existing GPhC guidance and out of step with national digital health policy.

2. About the Coalition for Responsible Digital Health (CoRDH)

The Coalition for Responsible Digital Health (CoRDH) is a newly formed coalition of regulated, digital-first healthcare providers operating in the UK, including Numan, Manual/Voy, Juniper, SheMed and Simple Online Pharmacy.

CoRDH was established to ensure that responsible digital health providers can constructively support discussions around innovation in care delivery, patient-centered regulation, and how strong governance and high clinical standards can be delivered digitally.

The coalition is underpinned by three core principles:

• Safe prescribing – robust, evidence-based clinical processes that protect patients and uphold the highest standards of care.
• Ongoing programme support – holistic, end-to-end care models that provide continuous monitoring, education, and escalation pathways, rather than transactional prescribing.
• Responsible communication – transparent, accurate information that supports informed patient choice and reduces the risk of harm from unregulated or misleading sources.

CoRDH members already operate within comprehensive clinical governance frameworks, including structured risk assessments, audit cycles, incident reporting, and clinician accountability. This document reflects those existing practices and sets out how they can be recognised and applied consistently by regulators.

2.1 CoRDH Asks

As a coalition committed to responsible digital health and patient safety, CoRDH seeks to work constructively with the General Pharmaceutical Council to support clear, consistent and proportionate regulation of digital prescribing.

At this stage, CoRDH is inviting the GPhC to:

1. Recognise proportionate, evidence-based safety frameworks.
   Formally recognise that - when implemented as described - the safety standards in Section 6 constitute a reasonable, proportionate and compliant framework for distance prescribing of GLP-1 medications. Specifically, CoRDH suggests that mandating video-call BMI verification on weighing scales is inappropriate, not in patients’ best interests, and may unnecessarily increase stigma and reduce access for people living with disabilities.

2. Provide clarity on acceptable digital verification methods.
   Confirm in writing that independent digital verification (photos, documents, NHS App evidence) is acceptable where access to the Summary Care Record (SCR) is not available.

3. Clarify expectations on GP involvement.
   Confirm that sharing information with a patient’s GP should be strongly encouraged but not mandated, where a competent adult has made an informed decision to withhold consent and the associated risks are actively managed within the service’s clinical governance framework.

4. Support consistent and predictable inspection practices.
   Issue clear guidance to inspectors to prevent local variation or “goalpost-shifting”, including illustrative case examples of acceptable digital models that meet GPhC standards.

5. Align regulatory expectations with national policy and clinical evidence.
   Engage with CoRDH and other stakeholders to align GPhC expectations with:

   • The NHS Long Term Plan and 10-Year Health Plan.
   • NICE and MHRA positions on the safety and use of GLP-1 medicines.
   • The UK Life Sciences and Industrial strategies, including the Growth Duty on regulators.

6. Enable safe transition when expectations evolve.
   Commit to a transition period when new expectations are introduced, allowing distance-selling pharmacies to adapt without fear of sudden enforcement action where there is no imminent patient-safety risk.

2.2 Why these asks

CoRDH recognises the GPhC’s commitment to patient safety and the effective regulation of distance-selling pharmacies. However, evidence from inspections suggests that the application of current guidance has, in some cases, led to unintended and inconsistent expectations.

We are concerned, for example, that current inspection practice:

• Over-emphasises analogue proxies (e.g. in-person BMI measurement, GP pre-verification, universal SCR access) and under-values robust digital verification and risk-management processes.

• Results in variation between inspections and between providers, creating uncertainty, undermining trust, and reducing predictability for regulated services.

• Creates expectations that are difficult to reconcile with wider national policy that promotes digital-first, patient-centred care and the use of regulated digital providers to reduce reliance on unsafe informal markets (Healthcare Improvement Scotland+4NHS England+4NHS England Digital+4).

CoRDH does not believe this reflects the intent of the updated guidance, but is a consequence of how that guidance is interpreted and applied in practice. (Pharmacy Regulation Authority+2gppharmacy.club+2).

Clarifying acceptable approaches, supported by clear examples of compliant digital models, would support inspectors and providers alike. Moreover it would improve consistency and better align regulatory practice with patient safety objectives and the larger national goals for a healthcare system fit for the future.

The proposals set out at the start of this document are intended to provide that clarity, by setting out a proportionate, evidence-based framework for safe digital prescribing that can be applied consistently in practice.

3. Context

3.1 Growth of GLP-1 prescribing and public concern

GLP-1 receptor agonists (e.g. semaglutide) and dual GLP-1/GIP agonists (e.g. tirzepatide) are now central to evidence-based management of obesity and type 2 diabetes. Large trials and reviews describe a favourable overall safety profile, with well-characterised class effects such as gastrointestinal symptoms, gallbladder disease and pancreatitis, and emerging but still unproven signals around mental health and pregnancy. Healthcare Improvement Scotland

At the same time, regulators have legitimate concerns about:

• Unlicensed / counterfeit supply of GLP-1 medicines.
• Ineligible patients gaining access, including:
  • Misrepresentation of medical history.
  • Misrepresentation of weight.
• Supply from providers operating outside robust clinical governance frameworks. Healthcare Improvement Scotland

CoRDH fully shares these patient safety concerns. Our argument is not that GLP-1s are risk-free, but that:

1. The risks are well understood and can be managed safely through smart, proportionate controls within an appropriately governed digital model.

2. The GPhC’s current stance often overstates medicine-intrinsic risk and under-recognises service-level mitigations, leading to disproportionate expectations for distance-selling pharmacies.

3.3 Digital-first healthcare and the changing role of the GP

Over the last decade, national policy has explicitly committed to a digital-first NHS, where patients can access primary care and outpatient services remotely and manage their health via digital channels.

The NHS Long Term Plan, the Digital Clinical Safety Strategy, the What Good Looks Like Framework and subsequent digital requirements commit that every patient should be able to access “digital-first” primary care - including online and video consultations and full online access to their record and that digital pathways -as the default for many prescribing processes, long term condition reviews, follow-up appointments and population health interventions.

The Government’s new 10-Year Health Plan for England and the continued expansion of the NHS App reinforce this direction, placing digital transformation,self management, remote monitoring, and neighbourhood-based care at the heart of a modern, sustainable health service.

In this context:

• The GP is no longer the exclusive gatekeeper to all care. Patients routinely use multiple providers (NHS and independent, online and face-to-face) often without needing a GP referral or in-person verification.
• Patients are expected and encouraged to manage their own health information, including via the NHS App and other digital tools.
• Clinical governance is evolving from a focus on physical attendance to a focus on whether appropriate controls are in place to support safe decision-making.
• Digital care is not inherently riskier; its risks are different and require different mitigations.

Regulation designed for an analogue, GP-centric system must therefore evolve to support safe digital care rather than constrain it.

However, current GPhC inspection practice for distance-selling pharmacies often appears to assume a traditional model in which:

• In-person measurements and GP-mediated verification are treated as the default or gold standard.
• Digital models are presumed to be higher risk and therefore subject to additional requirements rather than being assessed against equivalent safety outcomes
• Additional barriers e.g. rigid in-person requirements, GP confirmation, restrictions on remote prescribing, are applied selectively to digital providers, even when equivalent risks exist in walk-in, out-of-hours, or traditional community pharmacy services where healthcare providers also do not have full GP record access.

This approach is increasingly misaligned with wider NHS strategy and wider national policy, as well as the realities of how patients now access care.

4. Reframing the safety profile of GLP-1 medicines

4.1 Evidence summary

Key points from the literature and regulatory reviews:

• GLP-1 receptor agonists have a favourable overall safety profile when used within licence, titrated appropriately and with appropriate monitoring, particularly compared with the morbidity and mortality associated with untreated obesity and type 2 diabetes. PMC+1

• Known class risks include:

  • Gastrointestinal adverse effects, usually mild–moderate and dose-dependent.
  • Uncommon but serious pancreatitis and gallbladder disease.
  • Pregnancy risk (insufficient data; generally contraindicated).
  • Potential interactions with drugs of narrow therapeutic index due to delayed gastric emptying.
  • Emerging signals around mental health, under active investigation but not yet causally established (CKS+3Healthcare Improvement Scotland+3BMJ+3).

These risks arise from the pharmacology of GLP-1 RAs and patient-level factors, and are not unique to the mode of care delivery. They can be effectively mitigated through proportionate, evidence-based controls, including:

• Structured eligibility and exclusion criteria.
• Exclusion of high-risk cohorts.
• Safe prescribing and titration.
• Clear patient information, counselling and contraception advice.
• Close monitoring of symptoms and weight.
• Robust pathways for safety netting, side-effect management and escalation.
• Audit and clinical governance.

4.1.1 Real-world safety and benefit-risk of GLP-1 agonists in a regulated digital service

We are able to evidence

Real-world evidence from large, regulated digital weight-management services demonstrates that GLP-1 receptor agonists can be prescribed safely at scale within appropriately governed digital models.

A peer-reviewed analysis of over 100,000 adults treated with semaglutide and tirzepatide in a national digital service shows:

• Low overall clinical incident rates.
• No excess safety signal associated with greater engagement in digital support.
• Prescribing error rates comparable to, or lower than, those reported in other digital and traditional care settings.

These findings are consistent with major clinical trials and support the conclusion that patient safety is driven by clinical governance and risk controls, rather than by the mode of care delivery.

Full methodology and results are provided in Appendix A.

one of the large-scale, peer-reviewed real-world evidence research on outcomes and safety from Voy, a national digital weight-management service (DWLS) prescribing GLP-1 RAs. In a cohort of almost 107,000 adults treated with tirzeptatide and semaglutide, patients engaged in the digital Voy platform (medications with coaching access; app usage and educational content; and weight tracking) achieved at month 12: 21.5% versus 17.0% weight loss for those in the service that used primarily medication alone. Both weight loss outcomes are either in-line with or superior to clinical trials outcomes at the 12-month follow-up period, which may lead to beneficial clinical and psychosocial outcomes.

The study across 290,050 patient-months of exposure showed equitable and favourable safety profiles:

• Overall clinical incident rate was 1.57 events per 1,000 patient-months.
• With no excess risk for those that utilised the full digital service provided by Voy (incidence-rate ratio 0.83, 95% CI 0.60–1.15) compared to users not engaged fully in the digital platform.
• Most adverse events were no harm or minor harm (88%).
• With a low prescribing error rate of 1.71 per 1000 patients (95% CI, 1.47–1.97) compared to other digital health providers.

The real word picture is consistent with major clinical trials. SELECT, STEP and SURMOUNT show that GLP-1 RAs deliver large, sustained weight loss and reduce cardiovascular risk, with adverse effects dominated by dose-dependent gastrointestinal (GI) symptoms. Serious events are rare relative to the burden of untreated obesity.

When benchmarked against widely accepted long-term medications prescribed over-the-counter or online, the Voy DWLS safety signal sits well within, and often below routine regulatory tolerability thresholds:

• Statins: widely used for primary and secondary prevention, but associated with muscle, liver and renal adverse events and small diabetes risk, however safety events for muscle related events are 190 cases per 100,000 patient years; these are accepted in view of cardiovascular benefit.
• Combined hormonal contraceptives: often prescribed online or in an emergency has a risk as high as 12 per 10 000 women per year of use for developing venous thromboembolism (VTE), yet remain standard of care because benefits outweigh this manageable risk.
• Anti-hypertensives: such as Ramipril can lead to acute kidney injuries (AKI) which can have serious sequelae, with AKI event rates up to 2.54 per 1000 person-years (95% CI, 2.34–2.76), yet is prescribed due to the blood pressure control efficacy and impact on preventing further co-morbidity development.

NSAIDs: Ibuprofen is one of the commonest consumed over the counter (OTC) medication for pain management, yet has an increased risk of upper gastrointestinal bleed (UGIB) of 20 cases per 1000 person-years, which can lead to substantial increased risk of mortality.

4.2 Coalition safety model and internal risk assessments

Coalition members already implement comprehensive risk assessments and controls designed to support safe prescribing decisions and ongoing patient monitoring.

These cover:

• GLP-1 prescribing for weight loss, including detailed mapping of potential harms (hypoglycaemia, pancreatitis, gallbladder disease, eating disorders, aspiration risk, pregnancy, renal/hepatic impairment, older age, drug interactions, loss of muscle mass), and extensive controls (questionnaire-based inclusion/exclusion, treatment plans, side-effect management pathways, BMI thresholds, age restrictions, escalation routes, and 6-monthly audits).

• Access to Summary Care Records (SCR) and alternative verification pathways, recognising both the limitations of current SCR access for independent digital providers and the risk mitigation offered by robust questionnaires, ID verification, document uploads, audit processes, and clear escalation criteria.

• Scenarios where patients withhold consent to share information with their GP, including legal and ethical analysis, consent processes, repeated non-coercive prompts, and detailed clinical safeguards.

These internal assessments are used to support prescriber judgement, not replace it. Moreover, they demonstrate that, with appropriate controls in place, residual risk ratings are in the low–moderate range (e.g. overall GLP-1 prescribing risk rated 4 on a 5x5 matrix; risk from patients withholding GP-sharing consent reduced to 2).

CoRDH’s position is that such structured, auditable risk management should be recognised as a core indicator of safe practice. However, GPhC inspection practice often appears to place greater weight on a narrow set of process expectations (e.g. GP verification, SCR access) that are not mandated in law or guidance and which may in fact conflict with patient rights.

5. The modern patient profile and regulatory implications

5.1 Patient behaviour and access to care

Today’s weight-management patients are:

• Digitally literate and often prefer online access for sensitive issues such as obesity.

• Used to self-managing data, including uploading photos, home measurements and documents.

• Frequently under-served in traditional GP settings due to time pressure, perceived stigma or lack of specialist support.

If regulated digital providers are over-restricted, patients are more likely to:

• Seek medicines from unregulated suppliers, including social-media-driven grey markets and counterfeit sources (Healthcare Improvement Scotland).

• Avoid seeking help altogether, perpetuating obesity-related morbidity.

Regulation which assumes that the “safe default” is always face-to-face GP-mediated care is therefore out of step with both policy and lived reality. It also risks increasing, rather than reducing, patient harm.

Instead, proportionate standards should:

• Recognise multiple safe routes to care.

• Focus on outcomes and risk controls, not on mandating a particular analogue pathway.

5.2 Autonomy, confidentiality and GP involvement

Patients have a well-established legal right to:

• Choose where they receive care and medicines.

• Control the sharing of their personal health information, including with their GP, provided they are competent and adequately informed.

Our coalition model is:

• Default: strongly encourage and facilitate sharing of treatment information with the patient’s GP.

• Respect: where a competent adult, fully informed of the risks, declines, their decision is respected, and the associated risk is actively managed through enhanced clinical safeguards rather than overriding their autonomy.

Any regulatory expectation that effectively forces GP notification as a pre-condition of care risks undermining these rights and may unintentionally drive patients away from regulated providers and towards unregulated supply.

5.3 Precedent in other regulated healthcare settings

There is established precedent across the healthcare system for managing patient safety, confidentiality, and continuity of care without mandatory GP notification, particularly in areas where stigma, sensitivity or access barriers are recognised.

Sexual health services are a clear example. Patients can access care, including provision of high risk prescription-only medicines (such as anti-retrovirals)with significant interaction and safety considerations, without any requirement to inform their GP or to consent to information sharing. Instead, risk is managed through:

• Specialist clinical teams
• Targeted questioning for triage on relevant medical history
• Clinical assessments and exclusion criteria
• Comprehensive patient education
• Safety-netting and referral pathways

This approach is long established and applied across NHS and independent providers whilst being considered compatible with high standards of patient safety. Digital weight management services operate in a comparable context, where stigma and access barriers are well documented. In each case, patient safety is protected through proportionate clinical controls and governance, rather than compulsory GP involvement.

6. CoRDH’s proposed safety standards for distance prescribing of GLP-1 medicines

We propose that GPhC explicitly recognise the following as reasonable and sufficient safesafety standards for distance-selling pharmacies providing GLP-1s, provided they are properly implemented, documented and audited. These standards are intended to support safe clinical decision-making and should be assessed on the basis of patient safety outcomes rather than mandating a single mode of consultation.

6.1 Patient identification, BMI and anthropometric verification

1. Robust identity verification

   • Approved photo identification (such as a passport, driving licence); and
   • Face-match or equivalent digital Know Your Customer Check (e.g. Persona, Onfido etc.) processes.

2. Independent BMI verification
   GPhC should accept the following as equivalent to in-person measurement, used singly or in combination, following risk assessment by the prescriber:

   • Live full-length photos captured at the point of consultation - this means the photos must be taken in the moment and cannot be old photos uploaded from a photo library. This reduces the risk of edited or AI-generated photos as well as providing contemporaneous evidence of body habitus.
   • Photos should be assessed alongside patient declared height and weight to detect material inconsistencies.
   • Video consultations should be used as part of a risk-based escalation process where the prescriber determines that additional assurance is required. It should not be mandated for all patients, as this may be inappropriate, inaccessible, or disproportionately intrusive, particularly for those with mobility impairments, who cannot safely undertake live measurements, or for whom weight-related stigma makes such interactions distressing.

3. Dose and eligibility controls

   • Treatment cessation when BMI falls below a safety threshold (e.g. 19 kg/m²), with auto-generated clinical review and discontinuation communications.
   • Dose-increase restrictions in lower BMI ranges (e.g. 19–22 kg/m²) with manual prescriber override only where clearly justified and documented.
   • Patient-informed choice regarding dose, on the basis of current tolerance and side effects, and weight loss progress.
   • Additional evidence of prior GLP-1 exposure when patients request doses above the initiation dose.

6.2 Comorbidities and medicines: independent verification

We propose a risk-stratified approach:

1. Baseline for all patients:

   • Structured digital questionnaire aligned to evidence-based inclusion/ exclusion criteria and NICE / SmPC guidance, including specific prompts on:

     • Type 1 diabetes and current antidiabetic medication.

     • History of pancreatitis, gallbladder disease, renal and hepatic impairment.

     • Eating disorders and significant mental health history.

     • Medications with a narrow therapeutic index.

2. Enhanced verification where risk is higher, for example, complex polypharmacy or high-risk co-morbidities.
   Acceptable independent verification should include:

   • Photos of prescribed medicines, including dispensing label.
   • Uploads of NHS documents (repeat prescription list, clinic letters, discharge summaries).
   • Screenshots or PDFs from the NHS App showing active problems or repeat medications.
   • Where technically and contractually feasible, SCR or shared care record access. This is recognised as highly desirable but not yet universally available to independent digital providers, and therefore not a reasonable universal pre-condition at this time.

We also note that it is not standard practice in either NHS or private healthcare to require providers to obtain a patient’s full GP past medical history as a prerequisite of consultation or prescribing. Nor is it standard to verify the absence of all potential contraindicating conditions via GP records. To require this would impose a higher evidentiary threshold on digital providers than exists anywhere else in regulated care, without clear evidence of additional patient safety.

6.3 Management of GLP-1-specific risks

GPhC should regard the following digital controls as adequate and proportionate:

1. Exclusion criteria and contraindications

   • Automatic exclusion at questionnaire stage for:

     • Severe liver or kidney impairment
     • Severe heart failure
     • History of Pancreatitis
     • Multiple Endocrine Neoplasia Type 2
     • Cancer that requires monitoring or treatment
     • Type 1 diabetes
     • Diabetic retinopathy
     • Personal or family history of thyroid cancer
     • History of an eating disorder (e.g., anorexia, bulimia,)
     • History of gallstones if they still have their gallbladder
     • Severe digestive conditions such as Ulcerative Colitis, Crohn’s disease or gastroparesis

2. Comprehensive treatment plans and patient education

   • Condition and medicine-specific treatment plans covering:

     • Recognition of serious side effects (pancreatitis, gallbladder symptoms, severe GI issues, mental health deterioration).
     • Contraception and pregnancy actions (including stopping treatment and informing GP).
     • How to administer the medication.
     • Information regarding off-label prescribing and what this means for the patient, in line with GMC guidance.
     • In line MHRA safety alert, instructions to patients to withhold weekly injections 7 days pre-anaesthesia and daily preparations 24 hours before, and to inform surgical teams of GLP-1 use.

3. Side-effect management and escalation

   • Access to clinicians via various communication channels such as live chat, telephone hotline and email with defined SLAs. This allows the provision of prompt side effect management advice, including dietary recommendations.
   • Prescribing pathways for symptomatic relief where appropriate, and protocols for emergency signposting.
   • Procedure for the MHRA Yellow Card Reporting by clinicians, and encouragement of patients to complete these online.

4. Audits and continuous improvement

   • Six-monthly audits of:

     • Adherence to prescribing criteria.
     • Adverse event patterns and escalation outcomes.
     • Weight-loss outcomes and engagement with behavioural support.
     • GP-sharing consent rates.

6.4 Information sharing with GPs

We propose the following minimum standard, which supports patient safety, respects patient rights and manages clinical risk in line with established healthcare practice:

1. Ask every patient to consent to sharing treatment details with their GP, at multiple touchpoints (onboarding, consultation, treatment plan), with clear explanation of:

   • The benefits of sharing, including continuity of care and reduced medication interaction risk., emergency safety.
   • The risks of not sharing, including lack of continuity of care and reduced visibility for other healthcare providers

2. Make sharing straightforward and proportionate

   • A simple digital consent process capturing GP practice details.
   • Automated generation and dispatch of letters to GPs or downloadable letters for patients who prefer to share themselves.

3. Respect informed refusal and manage associated risk

   • Where consent is withheld, document:

     • That the patient has been informed of risks.
     • Their reasons (where offered).

   • Treat this as a managed clinical risk as per the GPhC guidance, not an automatic barrier to treatment.

This aligns with established practice in other regulated areas of care where confidentiality and stigma are recognised and where patient safety is maintained without mandatory GP notification. We ask GPhC to confirm that such a model meets standards, and that inspectors should not expect universal GP notification as a condition of safe practice.

5.5 Governance, competency and transparency

CoRDH members commit to:

• Formal clinical governance frameworks that prioritise patient safety, including risk registers, standard operating procedures (SOPs), incident reporting and learning systems.

• Defined competency standards and training for prescribers in obesity medicine and GLP-1 pharmacology.

• Transparent communication of risks, benefits, and service limitations to patients and, where appropriate, other healthcare providers.

We ask GPhC to focus inspections on whether such governance is functioning effectively in practice, rather than on the application of highly prescriptive or variable process requirements that may differ between inspectors.

7. Misalignment with current GPhC approach

CoRDH recognises the GPhC’s commitment to patient safety and the effective regulation of distance-selling pharmacies. However, evidence from inspections suggests that the application of current guidance has, in some cases, led to unintended and inconsistent expectations.

We are concerned, for example, that current inspection practice:

• Over-emphasises analogue proxies (e.g. in-person BMI measurement, GP pre-verification, universal SCR access) and under-values robust digital verification and risk-management processes.

• Results in variation between inspections and between providers, creating uncertainty,undermining trust, and reducing predictability for regulated services.

• Creates expectations that are difficult to reconcile with wider national policy that promotes digital-first, patient-centred care and the use of regulated digital providers to reduce reliance on unsafe informal markets (Healthcare Improvement Scotland+4NHS England+4NHS England Digital+4).

CoRDH does not believe this reflects the intent of the updated guidance, but is a consequence of how that guidance is interpreted and applied in practice. (Pharmacy Regulation Authority+2gppharmacy.club+2).

Clarifying acceptable approaches, supported by clear examples of compliant digital models, would support inspectors and providers alike. Moreover it would improve consistency and better align regulatory practice with patient safety objectives and the larger national goals for a healthcare system fit for the future.

The proposals set out at the start of this document are intended to provide that clarity, by setting out a proportionate, evidence-based framework for safe digital prescribing that can be applied consistently in practice.

8. Next steps and data to be added

CoRDH members will now:

• Aggregate real-world safety data (e.g. adverse event rates, hospital admissions, pregnancy exposures, mental-health outcomes) from coalition services.

• Compile process adherence and outcome metrics (e.g. percentage of patients with GP-sharing consent, audit findings, proportion achieving ≥10% weight loss at 6 months).

• Develop case studies demonstrating how the proposed framework has prevented harm or managed serious risk.

These datasets will be appended to this document to provide further reassurance that digital GLP-1 prescribing, when governed as described, offers a high-safety, high-value route for patients to access clinically appropriate care.

9. Appendices

Appendix A: Real world safety and outcomes of GLP-1 prescribing in a regulated digital service

A.1 Study population and exposure

We are able to evidence one of the large-scale, peer-reviewed real-world evidence research on outcomes and safety from Voy, a national digital weight-management service (DWLS) prescribing GLP-1 RAs. In a cohort of almost 107,000 adults treated with tirzeptatide and semaglutide, patients engaged in the digital Voy platform (medications with coaching access; app usage and educational content; and weight tracking) achieved at month 12: 21.5% versus 17.0% weight loss for those in the service that used primarily medication alone. Both weight loss outcomes are either in-line with or superior to clinical trials outcomes at the 12-month follow-up period, which may lead to beneficial clinical and psychosocial outcomes.

A.2 Safety outcomes

The study across 290,050 patient-months of exposure showed equitable and favourable safety profiles:

• Overall clinical incident rate was 1.57 events per 1,000 patient-months.
• With no excess risk for those that utilised the full digital service provided by Voy (incidence-rate ratio 0.83, 95% CI 0.60–1.15) compared to users not engaged fully in the digital platform.
• Most adverse events were no harm or minor harm (88%).
• With a low prescribing error rate of 1.71 per 1000 patients (95% CI, 1.47–1.97) compared to other digital health providers.

The real word picture is consistent with major clinical trials. SELECT, STEP and SURMOUNT show that GLP-1 RAs deliver large, sustained weight loss and reduce cardiovascular risk, with adverse effects dominated by dose-dependent gastrointestinal (GI) symptoms. Serious events are rare relative to the burden of untreated obesity.

A.3 Effectiveness outcomes

When benchmarked against widely accepted long-term medications prescribed over-the-counter or online, the Voy DWLS safety signal sits well within, and often below routine regulatory tolerability thresholds:

• Statins: widely used for primary and secondary prevention, but associated with muscle, liver and renal adverse events and small diabetes risk, however safety events for muscle related events are 190 cases per 100,000 patient years; these are accepted in view of cardiovascular benefit.
• Combined hormonal contraceptives: often prescribed online or in an emergency has a risk as high as 12 per 10 000 women per year of use for developing venous thromboembolism (VTE), yet remain standard of care because benefits outweigh this manageable risk.
• Anti-hypertensives: such as Ramipril can lead to acute kidney injuries (AKI) which can have serious sequelae, with AKI event rates up to 2.54 per 1000 person-years (95% CI, 2.34–2.76), yet is prescribed due to the blood pressure control efficacy and impact on preventing further co-morbidity development.
• NSAIDs: Ibuprofen is one of the commonest consumed over the counter (OTC) medication for pain management, yet has an increased risk of upper gastrointestinal bleed (UGIB) of 20 cases per 1000 person-years, which can lead to substantial increased risk of mortality.