NICE recommended 66 new treatments last year and turned down seven. But the number worth pausing on is the third one: 25.
The National Institute for Health and Care Excellence published its annual report and accounts for 2025 to 2026 on 7 July, covering the year to 31 March. It logged 98 new medicine evaluations. Of those, 66 ended in a recommendation, seven in a rejection and 25 were terminated outright because the evidence either wasn’t submitted or didn’t meet specifications.
That last figure reframes the headline statistic NICE quotes elsewhere in the same document. The institute says it recommends 91% of the medicines it evaluates, roughly 70 a year. Run the arithmetic, and you can see where that comes from: 66 out of the 73 appraisals that actually reached a verdict is 90.4%. Measured against all 98 evaluations started, the figure is 67%.
Neither number is wrong, and NICE isn’t concealing anything; both sit on the same page of the report. But the gap between them is the interesting part, and it mostly reflects decisions made on the industry side of the table rather than the committee’s. A quarter of the evaluations NICE opened never got far enough for anyone to rule on value.
The achievement NICE leads with is speed, and it’s a real one. Guidance production is now 30% faster for medicines and 9% faster for healthtech than in April 2024. The supporting metrics back it up: every piece of final guidance was published within 240 working days of the invitation to participate, against a 60% target and 44% the year before. Median time from marketing authorisation to a NICE recommendation fell to 232 days, comfortably inside the 332-day target. The institute reckons it’s on course to make England the third fastest country in Europe for medicines access by 2030.
The mean, though, tells a less flattering story: 367 days against a 335-day target, and worse than last year’s 335. Averages get dragged by outliers, and NICE’s divergent cases still run to 397 days.
Two structural changes matter more to industry than any of this. In December 2025, the government raised NICE’s cost-effectiveness thresholds; committees have been applying the new £25,000-to-£35,000 standard range since April 2026. NICE’s own analysis suggests that it buys an extra three to five medicines or indications a year. Alongside it is a new EQ-5D-5L value set, replacing a dataset that dated to the 1990s and offered three response options instead of five, a change NICE concedes may itself shift cost-effectiveness results.
The second is the MHRA-NICE aligned pathway, live since 1 April 2026 and announced at NICE’s Manchester conference in March. For the first time, licensing and value decisions can be published in parallel rather than in sequence, which NICE says pulls access forward by three to six months. A revamped Integrated Scientific Advice service one entry point, one meeting, one report, one payment sits alongside it.
Then there’s the remit expansion. Under the government’s 10 Year Health Plan, published in July 2025, NICE’s mandate now stretches to high-impact healthtech. The National Healthtech Access Programme launched in February 2026 with a funding mandate attached, meaning approved technologies get commissioning money the way medicines do. First two topics: capsule sponge tests for Barrett’s oesophagus and oesophageal cancer, and AI tools for spotting prostate and breast cancer.
The healthtech scorecard is where the report gets uncomfortable. None of the evaluations moved from prioritisation board decision to the start of guidance development inside 66 working days, 0% against a 40% target. Only 18% went from starting to finishing guidance within nine months, against 35%. Referral to board decision hit 30% against 50%. Healthtech confidentiality breaches came in at seven, one over tolerance. NICE has taken on a bigger job in a domain where its own timeliness targets are, so far, mostly red.
Elsewhere, the year reads well. NICE published 178 pieces of new and updated guidance, recommended ruxolitinib cream as the first licensed vitiligo treatment on the NHS in England, restructured its sepsis guidance, and conditionally backed six AI tools for detecting bowel cancer during colonoscopy. Financially, it landed a £0.32m surplus on £64.5m of DHSC funding plus £28.1m of other income, against £62.4m net expenditure.
It’s also Jonathan Benger’s first report as chief executive, having taken over from Sam Roberts in December 2025 after three years as NICE’s chief medical officer.