SGLT2 Inhibitors in UK Primary Care 2026: What NICE NG28 Means in Practice

The February 2026 NICE NG28 update moved SGLT2 inhibitors to first-line therapy for most adults with type 2 diabetes. We covered the headline shift in our NG28 update summary. This post is the operational companion piece — the agents, the eligibility, the cautions, and the workflow at the GP-appointment level.

It is written for UK GPs, GP registrars, GP pharmacists, advanced clinical practitioners, and trainees in primary care and general medicine. The NICE guideline and the BNF entry for each licensed agent remain the authoritative references. This post summarises and links; it does not substitute for clinical judgement.

The licensed agents in the UK

Four SGLT2 inhibitors are licensed in the UK for type 2 diabetes: dapagliflozin, empagliflozin, canagliflozin and ertugliflozin. Each has its own licensing nuances and renal cut-offs, but for primary-care prescribing the relevant distinction is local-formulary status rather than head-to-head agent choice.

In most UK integrated care boards and health boards, dapagliflozin and empagliflozin are the preferred agents. Both have the strongest cardiovascular and renal outcome trials behind them and the widest range of licensed indications (type 2 diabetes, heart failure with reduced or preserved ejection fraction, and chronic kidney disease). For most patients in most settings, agent choice is straightforward: prescribe the local-formulary preferred SGLT2 inhibitor unless there is a clinical reason to do otherwise.

Renal function matters at initiation. Each agent has a minimum eGFR below which initiation is not licensed, and a lower threshold below which continuation is not recommended. The BNF and SmPC are the authoritative sources for these thresholds and they update intermittently; checking before initiation, particularly for patients with stage 3 or 4 CKD, is essential.

Who is now eligible for first-line SGLT2 therapy

Under the 2026 NG28 framework, almost every adult newly diagnosed with type 2 diabetes is eligible. The pathway is no longer "metformin first, escalate later." It is "SGLT2 first-line, with metformin co-initiated or used as the alternative where appropriate."

The eligibility framework prioritises four overlapping groups for early or simultaneous-with-diagnosis SGLT2 initiation.

Established cardiovascular disease. Type 2 diabetes plus prior MI, stroke, angina, peripheral arterial disease, or revascularisation. The trial evidence is strongest here, and these patients gain the most.

Heart failure (any ejection fraction). Empagliflozin and dapagliflozin have established licensed indications in heart failure independent of diabetes. For a patient with both, SGLT2 therapy is core treatment for the heart failure as well as the diabetes.

Chronic kidney disease. Patients with type 2 diabetes and CKD with eGFR within the licensed range benefit from SGLT2 inhibition at the level of progression, not only glycaemia. The 2026 NG28 update treats this benefit as a primary indication, not a secondary consideration.

High pre-treatment HbA1c. Patients with HbA1c well above target at diagnosis benefit from earlier intensification. Combination initiation with metformin at diagnosis is now explicitly supported in this group.

For patients who do not fall into one of these groups — younger adults, low cardiovascular risk, modestly elevated HbA1c — SGLT2 monotherapy at diagnosis remains the new default. Metformin can be added if glycaemic targets are not met, or co-initiated if patient preference and clinical judgement support it.

When metformin still leads

The 2026 NG28 update repositions metformin; it does not retire it. Metformin remains the first-line agent in several circumstances.

SGLT2 contraindication. Active DKA, type 1 diabetes wrongly classified, severe foot disease (particularly with prior amputation), or recurrent severe genitourinary infection. Pregnancy and breastfeeding remain metformin-led where pharmacological therapy is indicated.

SGLT2 intolerance. Recurrent thrush, recurrent UTI, intractable volume depletion, or symptomatic euglycaemic ketosis on previous exposure.

Patient preference. After an informed conversation about cardiovascular and renal benefit, some patients will prefer metformin. The 2026 NG28 frames patient preference as a legitimate first-line determinant, not an obstacle to be overcome.

Frailty and limited life expectancy. For older adults with significant frailty, polypharmacy, or limited life expectancy, the calculus shifts. Long-term cardiorenal benefit is less relevant; tolerability and simplicity matter more. Metformin (or no pharmacological therapy at all) is often the right answer.

The three cautions that matter most in practice

Euglycaemic diabetic ketoacidosis is rare but serious. SGLT2 inhibitors lower blood glucose by increasing urinary glucose excretion; patients can be in DKA at glucose levels that look reassuring. The risk rises with intercurrent illness, fasting, low-carbohydrate diet, alcohol excess, and perioperative periods. The sick-day rule is to pause the SGLT2 inhibitor during acute illness until eating and drinking are restored. Patients should know this at initiation.

Genitourinary infections. Genital thrush is common, particularly in the first three months and particularly in women. Uncomplicated UTI is more common but rarely a reason to discontinue therapy. Personal hygiene advice at initiation is worthwhile; pre-emptive antifungal treatment is not routinely indicated. Recurrent or severe GU infection is a reason to consider discontinuation.

Foot care and amputation risk. The early canagliflozin trial signal around lower-limb amputation prompted ongoing vigilance across the class. NG28 2026 codifies a documented foot examination at SGLT2 initiation and continued foot vigilance throughout therapy. For patients with prior ulcer, neuropathy, or amputation, the risk-benefit balance shifts and deserves a deliberate conversation.

The initiation workflow in primary care

A practical sequence for a new type 2 diabetes diagnosis under NG28 2026:

1. Confirm the diagnosis and exclude type 1 diabetes (autoantibodies if clinically indicated; consider GAD-65 and C-peptide where there is doubt).
2. Baseline assessment: HbA1c, eGFR, urine ACR, lipid profile, BP, weight, foot examination, structured education referral.
3. Explain the prescribing choice. For most patients, the conversation is "the first-line drug for diabetes has changed, and this one has heart and kidney benefits as well as lowering your blood sugar."
4. Initiate the formulary-preferred SGLT2 inhibitor. Co-initiate metformin if HbA1c is well above target or cardiovascular comorbidity supports it.
5. Document the sick-day rules conversation. Some practices issue a written summary card; this is good practice.
6. Schedule a follow-up at six to eight weeks: HbA1c trajectory, tolerability, GU symptoms, foot status.
7. Review and titrate as usual.

For existing patients on metformin monotherapy, the workflow is different. Targeted register review — patients with cardiovascular comorbidity, CKD, or rising HbA1c — is the highest-yield use of clinical time. Not every patient on stable metformin monotherapy needs their regimen changed because the guideline has moved. Many of them, however, will benefit from SGLT2 addition once the conversation is had.

A note on what this post is — and is not

This is a practical summary of the 2026 NICE NG28 update aimed at UK primary-care clinicians. It is not a substitute for the NICE guideline, the BNF entry for each licensed agent, or local prescribing policy. Clinical decisions remain the responsibility of the prescribing clinician. Where doubt exists, the NICE visual summary at nice.org.uk/guidance/ng28 and the BNF are the references to consult.

How MCB fits in

We publish a weekly summary of the most important papers and guideline updates across 31 UK medical specialties. Type 2 diabetes is one of them; primary care more broadly is another. The 2026 NG28 update is the kind of practice-changing news that lands in a busy clinical week and gets missed by the clinicians it most affects. That is the problem we are built to fix.

Subscribe to The Monday Clinical Brief — £20 a year, no advertising, no sponsored content. One email every Monday morning with the practice-changing material from your specialty, summarised and linked.