When the answer is "we can't afford it."

Affordability isn't a hard threshold.

First, unlearn something. Cost-effectiveness has an explicit threshold — £20,000–£30,000 per QALY, a published line you compare the ICER against. It's tempting to assume budget impact has an equivalent: a fixed "affordability ceiling" above which the answer is automatically no. It usually doesn't.

Affordability is softer, and more political, than that. What makes a budget impact "unaffordable" isn't crossing a magic number — it's the opportunity cost inside the budget: to fund this, what would the payer have to stop funding elsewhere? A £200 million budget impact isn't unaffordable in the abstract; it's unaffordable if the £200 million would have to come out of services the payer is unwilling to cut. This is the Module 7 idea — opportunity cost — made concrete at the level of a real annual budget. So "can we afford it?" is really "is this worth more than whatever we'd displace to pay for it, and can we find the cash flow to do so this year?" That's a judgement, not a threshold — which is exactly why it's negotiable.

The spectrum of responses.

Because affordability is negotiable, the payer's decision is not a switch but a dial. Faced with a valuable but unaffordable technology, a payer can:

• Approve — if, after everything, the budget can absorb it.
• Restrict to a subgroup — fund it only for the patients where value is highest or need is greatest, shrinking both the budget impact and, often, improving the average value.
• Phase it in — approve it but manage the rate of rollout, spreading the cost over more years.
• Negotiate price and access — the big one: change the terms until the numbers fit, through discounts, caps, or conditional deals.
• Reject — the genuine last resort, not the default.

Each of these targets a different problem. Restriction and phasing reshape who and when; negotiation reshapes the price and the exposure. A skilled payer diagnoses which part of the budget impact is the obstacle and reaches for the matching response — which is why treating the decision as a blunt yes/no throws away most of the actual craft of the job.

List price is a fiction; the effective price decides.

Now the single most important thing to understand about how these decisions really work — and the one most often missed by people reading published appraisals.

The list price of a drug — the official, public, headline price — is frequently not the price anyone actually pays. Behind it sits a confidential discount, negotiated between payer and manufacturer, that produces the effective price: the real, lower price the system pays. Decisions are made on the effective price. Yet the ICER and budget impact figures in a public appraisal document are often calculated on the list price, because the effective price is commercially confidential. So the published "£45,000 per QALY, rejected" may conceal a private "£22,000 per QALY, approved" — same drug, different price, opposite verdict.

Why keep the real price secret? Because of international reference pricing: many countries set their own prices by looking at what other countries pay. If a manufacturer openly cut its UK list price, every country referencing the UK would demand the same cut. A confidential discount lets the manufacturer offer one country a low effective price without collapsing its price everywhere else. The consequence for an assessor is profound: a headline ICER or budget impact computed on list price can be almost meaningless, and "what's the effective price?" is often the only question that matters — even when it can't be publicly answered.

Negotiate to yes.

Here's a drug that's good value but unaffordable: at list price its ICER is £24,000 (cost-effective ✓) but its budget impact is £60 million a year against a £20 million ceiling (affordable ✗). You're the payer. Use the three negotiation levers to get both verdicts to green — and notice which lever moves which axis.

There's no single "right" answer — there are several routes to yes: a modest discount plus a cap, or a cap alone, or restriction to the subgroup. Which route a payer takes depends on which problem the drug actually has. This drug's problem was never its value — it was pure scale, and the levers that fix scale (caps, restriction) aren't the same as the levers that fix value (price). Diagnosing the obstacle is the whole game.

Now you.

Matching the lever to the problem.

The hero showed it; here's the principle. Different problems demand different levers, and reaching for the wrong one wastes the negotiation.

• A value problem — the ICER is above threshold — is fixed by a lower unit price (or better evidence of benefit). The drug genuinely costs too much per unit of health, so the per-unit price has to come down.
• A pure affordability problem — the ICER is fine, but the total is too big — often shouldn't be fixed by cutting the unit price at all. The value was never in question; the exposure is. So the right lever is a spend cap or a price–volume agreement that limits the payer's total outlay while leaving the headline unit price intact. Manufacturers often prefer this, because it protects the list price that other countries reference — the payer gets affordability, the manufacturer keeps its international price.
• An uncertainty problem — you're not sure the benefit is real (recall Module 9) — is fixed by an outcomes-based or coverage-with-evidence deal: pay only if the drug works, or fund it while collecting the data that will settle the question.

Read that middle case again, because it's the counter-intuitive one: when a drug is cost-effective but unaffordable, the solution is frequently not "make it better value." It's "cap the bill." The instinct to always push the unit price down misses that a good-value drug can be made affordable without touching its value at all.

What these deals cost.

These arrangements — confidential discounts, caps, price–volume and outcomes-based deals — are collectively managed entry (or risk-sharing) agreements, and they're powerful. They're also not free, and a clear-eyed assessor weighs their costs:

• They distort the information environment. Confidential effective prices mean the published ICER and budget impact — the numbers everyone analyses — may not reflect reality. Transparency, and cross-country learning, suffer.
• They're administratively heavy. Outcomes-based deals require measuring, verifying, and enforcing whether the drug "worked" for each patient — data infrastructure many systems lack, and disputes when the answer is ambiguous.
• They can paper over a bad decision. A deal that makes the numbers fit isn't the same as a technology that's worth it; "we negotiated it down to affordable" can quietly fund something whose underlying value or evidence was never sound.

The mechanics of these agreements — the legal and contractual forms, how they're implemented in real reimbursement systems — are a Module 12 topic. What matters here is the logic: they exist because affordability is negotiable, they solve real problems of scale and uncertainty, and they carry a real price in transparency and complexity. A payer's "yes" is very often one of these deals wearing the mask of a simple approval.

What negotiated solution fits best?

Why this matters for HTA

The payer's decision is where all the analysis of the preceding modules turns into an actual outcome for patients — and reading it well means seeing past the binary verdict.

• Never read a rejection as final, or an approval as unconditional. Most real decisions are conditional — a subgroup, a cap, a confidential discount. "Rejected at list price" and "approved with a deal" are often the same drug. The interesting question is always on what terms, not just yes or no.
• Ask what price the numbers were computed on. A public ICER or budget impact built on list price can be a fiction; the effective price is what the decision actually turned on. When you can't see the effective price — and often you can't — treat the headline figures with according caution.
• Diagnose which axis fails before judging the fix. A value problem and an affordability problem look similar from a distance and need opposite remedies — a lower unit price versus a cap on total spend. Matching the lever to the problem is the core skill; a good-value drug made affordable by a spend cap is a well-handled decision, not a fudge.

The last question in health economics is rarely "yes or no." It's "at what price, for whom, how fast, and on what terms?" A payer's real power isn't the veto — it's the ability to reshape a technology's price and exposure until an unaffordable "no" becomes a workable "yes."

Affordability and the payer's decision, in one breath.

• Affordability isn't a fixed threshold like the WTP line — it's the opportunity cost inside the budget: what you'd have to stop funding to pay for this. That makes it a judgement, and therefore negotiable.
• The payer's decision is a spectrum, not a yes/no: approve, restrict to a subgroup, phase in, negotiate price and access, or reject. Each targets a different obstacle.
• List price is often a fiction; decisions are made on the confidential effective price. Public ICERs and budget impacts computed on list price can mislead — and effective prices stay secret because of international reference pricing.
• Match the lever to the problem: a value problem needs a lower unit price; a pure affordability problem needs a spend cap or price–volume deal (protecting the unit price); an uncertainty problem needs an outcomes-based deal. A price discount is the one lever that moves both value and affordability at once.

"We can't afford it" is where the analysis stops and the negotiation starts. The whole apparatus of value and budget impact exists to set the terms of that negotiation — not to end it with a single word.

That closes Module 10 — and with it, the core analytical machinery of HTA: value, models, uncertainty, and affordability. But every number we've fed that machine came from somewhere, and mostly from trials — selected populations, protocol conditions, short follow-up. What happens when we want evidence from the messy real world instead: registries, claims data, routine practice? That's real-world evidence, and it opens Module 11.