Two questions, not one.

A tale of two drugs.

Take two drugs. Both have been through a full cost-effectiveness analysis and both come out at an ICER of £20,000 per QALY — comfortably cost-effective, identical value for money. On the "is it worth it?" axis, they are the same drug.

Now look at what they cost the system.

• Drug A treats a rare condition. Each patient's treatment costs £8,000 a year more than what they'd have received otherwise, and about 200 patients are eligible. Annual budget impact: 200 × £8,000 = £1.6 million.
• Drug B treats a common condition. Same £8,000 per patient, same £20,000 ICER — but 200,000 patients are eligible. Annual budget impact: 200,000 × £8,000 = £1.6 billion.

Identical value. A thousand-fold difference in the bill. Drug A is a rounding error; Drug B could consume a meaningful slice of an entire national drug budget. The cost-effectiveness analysis — which said they were the same — is completely silent on this thousand-fold gap. Something else has to speak to it.

The ICER divides; the BIA multiplies.

Why can two identically cost-effective drugs cost so wildly differently? Because the two analyses do opposite arithmetic.

The ICER divides. It's incremental cost per QALY — a cost spread over a unit of health, calculated for a single patient against a single comparator. It's a rate: the price of one unit of value. Dividing by the health gained deliberately washes out the scale — a rate doesn't care whether one patient or a million receive it.

The BIA multiplies. It's the extra cost per patient, times the number of patients, summed into a total cash figure. It's not a rate — it's a sum. And multiplication is exactly where scale re-enters: the population size that the ICER divided away is the very thing the BIA multiplies by.

So a low ICER and a colossal budget impact aren't in tension at all. The drug buys health cheaply per unit — and buys an enormous amount of it. Cheap per unit, vast in total: exactly like an inexpensive item bought by the millions. The rate is excellent; the bill is still staggering.

Four things that differ.

Cost-effectiveness analysis and budget impact analysis aren't two flavours of the same thing — they differ on almost every axis that matters:

• The question. CEA asks "is it good value?" BIA asks "can we afford it?" Value versus affordability.
• The currency. CEA speaks in £ per QALY, weighed against a threshold. BIA speaks in plain pounds — a cash total, no QALYs, no threshold. One is a health-value metric; the other is a financial forecast.
• The scope and horizon. CEA is per patient, over their lifetime, discounted. BIA is the whole eligible population, over a short horizon — typically 1 to 5 years, the span a budget-holder actually plans for — and usually undiscounted, because it's about near-term cash.
• The comparator and perspective. CEA compares against one comparator and reports an increment. BIA compares "the world with this technology" against "the world without it" across the payer's entire current spend on that population — strictly from one payer's budget, often a narrower perspective than the CEA's.

Same technology, two analyses, two genuinely different pictures. A payer needs both before saying yes.

Value versus affordability.

Two independent axes, two independent verdicts. The horizontal axis is value — the ICER, judged against a £30,000 threshold. The vertical axis is affordability — the annual budget impact, judged against an illustrative £20 million ceiling. Move the sliders and watch which of the four quadrants the technology lands in.

(This is not the cost-effectiveness plane from Module 7 — the axes here are the verdicts of two whole analyses, not ΔEffect and ΔCost.)

Keep the ICER fixed at a healthy £20,000 and just drag the population up. The value verdict never changes — it's still cost-effective at every population — yet the technology slides from comfortably affordable into wildly unaffordable. That single movement is the whole lesson: value and affordability are separate axes, and a drug can be excellent on one while catastrophic on the other. The ICER slider can't rescue the budget, and the population slider can't spoil the value.

Now you.

Why cost-effective drugs get rejected.

Here's the consequence that surprises people new to HTA: a drug can be waved through on cost-effectiveness and still be turned down — or admitted only on conditions — because of budget impact. That isn't a contradiction or a mistake. It's the two axes doing their separate jobs.

The reason is the shape of a real budget. A payer's drug budget is fixed and annual — a pot of money for this year. Cost-effectiveness speaks in lifetime value: the QALYs a patient will accrue over decades, discounted back. But a budget-holder can't pay this year's bills with a patient's future lifetime value. They need the cash, now, and if a cost-effective drug demands £1.6 billion of it in year one, "it's great value" doesn't make the money appear. Something else in the budget would have to give — the opportunity cost from Module 7, made brutally concrete.

This is precisely why a cost-effective but budget-busting drug so often enters through a side door rather than a flat yes: the payer and manufacturer negotiate a managed entry or risk-sharing arrangement — a confidential discount, a cap on total spend, a price-volume deal — that shrinks the budget impact without touching the headline ICER. The value was never the problem; the affordability was. (Those agreements are a Module 12 topic — for now, just note why they exist: to bend the BIA when the CEA already passed.)

The rarer mirror image.

The famous case is "cost-effective but unaffordable." The mirror image is rarer but just as instructive, and it proves the two axes are truly independent — in both directions.

Picture a cheap drug for a tiny population: it costs very little extra per patient, and only a few hundred people are eligible, so its budget impact is trivially affordable — the payer would barely notice it. But suppose its ICER lands just above the threshold — its health gain, per pound, is marginally poor value. Here the affordability verdict is a comfortable yes and the value verdict a narrow no. A payer could reasonably fund it anyway — the sum is so small the value question hardly bites — but strict cost-effectiveness would say reject.

The point isn't which way any particular case tips. It's that the two verdicts can disagree in either direction: cheap-but-poor-value just as much as good-value-but-unaffordable. Value and affordability are genuinely orthogonal — knowing one tells you nothing about the other. Which is exactly why a serious appraisal computes both, and never lets one stand in for the other.

What's the most likely reason?

Why this matters for HTA

Budget impact is where a technically favourable appraisal meets financial reality, and reading the two analyses as separate is one of the most important habits an assessor can build.

• Never let a good ICER answer the affordability question. A favourable cost-effectiveness result tells you the technology is good value — and nothing about whether the budget can take it. The moment you see a strong ICER attached to a large population, the next question is automatic: what's the budget impact? The two are computed separately for a reason.
• Watch the population size — it's where affordability is won or lost. Because BIA multiplies by eligible patients, the population estimate does more to move the budget impact than almost anything else. A drug's affordability lives or dies on how many people qualify — which is exactly why that number is so often contested (the subject of the next lesson).
• Read a high budget impact as a signal, not a verdict. A large BIA doesn't mean "bad drug" or "poor value" — it means "big cash demand." The right response is usually a conversation about phasing, caps, or discounts, not rejection. Diagnosing which axis a technology fails on tells you which lever can fix it.

Cost-effectiveness asks whether a pound spent buys enough health. Budget impact asks whether the pounds exist to spend. A technology has to be true on both — and the whole discipline of the payer's "yes" lives in the gap between value and affordability.

Budget impact: a different question, in one breath.

• HTA answers two independent questions: cost-effectiveness asks "is it good value?" (the ICER, per patient, against a threshold); budget impact asks "can we afford it?" (the total cash cost to the payer). A technology can pass one and fail the other.
• The ICER divides (cost per QALY — a per-unit rate that washes out scale); the BIA multiplies (cost per patient × eligible patients — a total that puts scale back). So a low ICER and a huge budget impact happily coexist.
• They differ on nearly everything: currency (£/QALY vs plain £), scope (per-patient lifetime vs whole-population 1–5 years), comparator and perspective (one comparator, discounted vs current spend, one payer's budget).
• A cost-effective drug can be rejected or restricted on affordability — which is why managed-entry and risk-sharing deals exist: to bend the budget impact when the cost-effectiveness case is already won.

Good value and affordable are not the same sentence. The ICER can tell you a drug is worth buying; only the budget impact can tell you whether you can pay for it.

We've established that budget impact hinges on one number above all — how many patients will actually receive the technology, and how fast. That number is anything but obvious: eligible isn't the same as treated, and uptake builds over years. Pinning it down — population, uptake, and the time horizon — is the machinery of the next lesson.