Paying Once or Paying Forever: The NHS's CRISPR Budget Conundrum

Ask a haematologist in 2026 what stands between a patient with severe sickle-cell disease and a CRISPR-based cure, and the answer is unlikely to be biology. The science works. The obstacle is the invoice.

A cure that arrives as a single bill

In 2023 the UK medicines regulator became the first in the world to authorise a CRISPR therapy, Casgevy, for sickle-cell disease and transfusion-dependent beta-thalassaemia. The treatment edits a patient's own blood stem cells outside the body, raising fetal haemoglobin and, in many trial participants, ending crises and transfusions. It is a genuine one-off intervention. It is also, by any normal NHS standard, extraordinarily expensive.

Reported list prices for comparable gene therapies run into seven figures per patient. The exact UK price is commercially confidential, but the order of magnitude is not: a course of treatment can cost more than a decade of standard care for the same patient. That creates an accounting problem that no amount of scientific elegance can dissolve.

The NICE arithmetic of time

The National Institute for Health and Care Excellence (NICE) decides whether a medicine represents value for money for the NHS in England. Its standard framework compares the extra cost of a treatment with the extra health it delivers, measured in quality-adjusted life-years (QALYs). A treatment that costs more than about £20,000 to £30,000 per QALY is usually rejected, unless the severity of the condition or other factors justify a higher threshold.

Gene-editing cures struggle in this framework for a simple reason: the benefits arrive late. NICE's reference case discounts future health gains at 3.5% a year. A benefit that accrues in 20 years is worth roughly half its nominal value in the model. A therapy that prevents strokes, organ damage and hundreds of hospital admissions over three decades may therefore look less cost-effective than annual transfusions and pain management, even if the total cash spent over a lifetime is lower.

This is not a flaw in NICE's method so much as a feature of it. Discounting reflects society's preference for health now over health later. But it sits awkwardly with a technology whose entire selling point is to trade a large cost today for a stream of savings tomorrow.

Who holds the budget?

The NHS does not budget in decades. It budgets in financial years. A hospital trust that administers a one-off gene therapy must account for the cost in the year it is given. The savings — fewer emergency admissions, less chronic transfusion, fewer days lost from work or school — may accrue to different budgets, different organisations, or different years. The trust that pays is not always the body that saves.

For rare diseases, specialised commissioning adds another layer. NHS England commissions highly specialised services centrally, but the budget is still finite and still scrutinised annually. Sickle-cell disease affects around 15,000 people in the UK, with the highest prevalence in London, the West Midlands and the North West. Beta-thalassaemia is rarer. Even a modestly sized eligible population, multiplied by a seven-figure price, produces a budget impact that dwarfs many existing specialised programmes.

Risk-sharing and the durability question

NHS England has experience with one-off therapies. It has used outcomes-based agreements and annuity-style payments for other gene therapies, including treatments for spinal muscular atrophy and metachromatic leukodystrophy. The logic is straightforward: spread the cost over several years, and link payment to observable outcomes. If the treatment stops working, the NHS pays less.

For CRISPR therapies, the relevant outcome might be transfusion independence, freedom from vaso-occlusive crises, or a sustained rise in fetal haemoglobin at 12 or 24 months. But durability data are still maturing. Trial follow-up for the first patients extends only so far. An outcomes-based deal can share the risk of early failure; it cannot eliminate the uncertainty of a 30-year benefit stream that no one has yet observed.

What value for money means for patients

For patients, the debate is not abstract. NICE approval is necessary but not sufficient. Specialised commissioning must agree funding, and clinicians must select patients who are fit enough for myeloablative conditioning — the chemotherapy that makes space in the bone marrow for edited cells. That conditioning carries risks, including infertility and organ toxicity, and supportive care is intensive.

There is also an equity dimension. Sickle-cell disease disproportionately affects people of Black African and Caribbean heritage, communities that have long reported delays and neglect in NHS pain management. A cure that is approved but not funded, or funded only in a handful of specialist centres, would risk reproducing those inequalities in a new form.

Can the accounting catch up?

Several fixes are possible. Annuitised payment models can smooth the upfront spike. Risk-sharing agreements can tie payment to durability. Pooled budgets across NHS and social care could capture savings that currently fall outside the health ledger. NICE's severity modifier and highly specialised technologies route already offer some flexibility for very rare, severe conditions. The UK's voluntary medicines pricing scheme could, in principle, accommodate novel payment structures.

None of these solves the underlying tension. A one-off cure is a long-term investment, and the NHS is a short-term accountant. The question is not whether CRISPR works. It is who pays, when, and against which budget. As more one-off therapies arrive — for haemophilia, inherited blindness and metabolic disease — the answer will determine how quickly British patients see the benefits of the gene-editing revolution.

For now, the NHS's CRISPR bill is less a single invoice than a test of how the health service values time. The science has moved faster than the spreadsheet. The spreadsheet will have to catch up.