One jab a week: what the NICE once-weekly insulin decision means

Insulin has been keeping people alive for a little over a century, and for almost all of that century it has meant frequent injections. In August 2026, NICE recommended insulin efsitora alfa — a basal insulin from Eli Lilly, injected once a week rather than once a day — for use in type 2 diabetes on the NHS. How quickly it reaches clinics is a separate matter. As of late August 2026 the final guidance, local formularies and supply were each still running on their own clock.

The decision itself is a gift to any medicine applicant, because it walks through the whole machinery of how the NHS decides what to pay for. One body concluded the drug works and is safe to sell. A separate body then asked a harder question: is it worth the money, set against what the NHS already does? Understand why those are two questions answered by two institutions, and you can handle almost any interview question about NHS spending, rationing or new treatments.

Applicants routinely blur two institutions that do very different jobs. The Medicines and Healthcare products Regulatory Agency (MHRA) is the licensing gate: it examines trial evidence and manufacturing standards and decides whether a medicine is safe, effective and of acceptable quality. An MHRA licence means a drug may be sold in the UK. It says nothing about whether the NHS will buy it.

That second question belongs to NICE, the National Institute for Health and Care Excellence. Its usual route for a single new medicine is a technology appraisal: a structured comparison against current NHS care that weighs clinical effectiveness and cost effectiveness together. The route matters, because a positive technology appraisal carries legal force in a way NICE’s ordinary clinical guidelines do not. Once final appraisal guidance is published, the NHS in England is normally required to fund the treatment for eligible patients within three months. That funding mandate is why a NICE decision makes the news while an MHRA licence rarely does. If the division of labour is new to you, our guide to how the NHS works sets out the full map of who does what.

One wrinkle worth knowing: NICE technology appraisals formally cover England. Wales usually adopts them, Northern Ireland generally follows, and Scotland runs its own assessment through the Scottish Medicines Consortium, so "the NHS has approved a drug" is shorthand for a more devolved reality. NICE itself was created in 1999, then called the National Institute for Clinical Excellence, precisely because access to new treatments varied with postcode. Whether it fixed that entirely is a fair debate; that it made the variation visible and forced decisions through a published method is not.

How do you compare a diabetes drug with a cancer drug with a hip replacement? NICE’s answer is the quality-adjusted life year. One QALY is one year of life in full health; time lived in poorer health counts as a fraction of a year. Divide a treatment’s extra cost by the extra QALYs it delivers and you get a single number, cost per QALY, that lets utterly different treatments compete for the same fixed budget on common terms.

The commonly cited benchmark is roughly £20,000 to £30,000 per QALY. It is a convention rather than a statute. NICE has flexibility at the margins, and its current methods apply a severity modifier that allows a higher cost per QALY for the most severe conditions. As a rule of thumb, though, a treatment costing far above that range per QALY gained will struggle, however impressive its science.

The number that matters most never appears in a headline: the health forgone elsewhere. The NHS budget is fixed in any given year, so recommending one treatment quietly moves resources away from others — health economists call this opportunity cost. Framed that way, NICE is not the body that denies patients treatment. It is the body that makes sure the trade-off, which would happen anyway, happens in daylight and by a method anyone can read and challenge.

Why might a weekly insulin justify a premium over daily versions that already work? Three arguments, and only the first is obvious.

The first is adherence. Basal insulin only works if it is actually injected, and doses get missed in ordinary life in a way they rarely do inside a tidy trial protocol. Fifty-two scheduled moments a year are easier to keep than 365. The second is workforce. Some people with diabetes, often older or living with frailty or dementia, cannot self-inject, so community nurses visit to do it for them. Turn a daily visit into a weekly one and you release nursing time across a stretched service, which is the kind of pressure our guide to NHS pressures and the workforce examines.

The third argument resists a spreadsheet altogether. Fewer injections mean less daily reminder of disease, and for some patients a regimen they can finally live with. All three feed the economic model, which is why "is it worth the money?" is a subtler question than the list price suggests.

Now the other side, argued properly. Type 2 diabetes is one of the most common long-term conditions in the country, so even a modest per-patient premium multiplies into serious money, and total budget impact matters alongside cost per QALY. Prices are usually softened by confidential commercial discounts, which means the public rarely knows what the NHS truly pays; that sits awkwardly with an otherwise open method. Long-term outcome data for a newly licensed weekly insulin is inevitably thinner than for daily insulins used across decades. And convenience only becomes saving if the released nurse hours are genuinely redeployed rather than absorbed into the next crisis.

The strongest interview answers hold both halves at once. This is a promising decision, and it is also a live test of whether modelled savings turn into real ones.

This story rarely arrives as "tell me about efsitora alfa". It arrives in general forms:

• "Should the NHS pay for expensive new drugs?"
• "What is NICE and why does it matter?"
• "Is it ever acceptable to refuse a patient a treatment that works?"
• Follow-ups on rationing: "Who should make these decisions — doctors, politicians or the public?" and "What would you say to the patient who is refused?"

The depth expected is mechanism, not memory. A panel wants four things: the two-gate distinction between MHRA licensing and NICE appraisal; a one-sentence definition of a QALY; the commonly cited £20,000 to £30,000 range, offered as a convention rather than a law; and one concrete example — which is where a once-weekly insulin earns its place. Practise compressing all four into ninety seconds, because the follow-up will pull on whichever thread the interviewer finds interesting. When you are ready to rehearse under realistic pressure, our interview preparation pages are built for exactly that.

Hot topics decay; mechanisms compound. The efsitora headline will eventually be displaced by the next recommendation, but the machinery — licence, appraisal, mandate, rollout — will still be deciding what the NHS buys when you are the one prescribing. Work through the rest of the system on our interview reading path, then move from reading to speaking: the candidates who score on this topic are the ones who have said their answer out loud before the day it counts.