M12 · REGULATION & REIMBURSEMENT

Where does a price even come from?

Through the last lesson, the decision gate turned on a single number we kept treating as given: the price. A drug is cost-effective at some price, affordable at some price, negotiated to a price. But we never asked the obvious question — where does that number come from in the first place? Who decides that a new cancer drug costs £15,000 a course rather than £5,000 or £50,000?

The intuitive answer, borrowed from almost every other product you buy, is cost-plus: work out what it costs to make, add a margin, that's the price. And for a toaster, that's roughly right. For a medicine, it's almost irrelevant. The cost of manufacturing one more pill is often trivial — pennies. The real costs are the research and development poured in beforehand (much of it on drugs that failed), and, far more importantly, the value the drug delivers. A medicine that saves a life and one that soothes a mild rash might cost the same pennies to produce, and nobody thinks they should carry the same price.

So a drug's price isn't found by tallying costs — it's constructed, deliberately, according to a chosen logic. And there are two fundamentally different logics for constructing it, answering two completely different questions. This lesson is about both — starting with the one you already, secretly, know how to compute.

Logic 1: price from value.

The first logic is value-based pricing, and its question is: how much is this drug worth to the health system?

The intuition is beautifully simple, and you already have every piece of it. A health system, we've seen, is willing to pay up to a certain amount for a unit of health — the cost-effectiveness threshold, say £30,000 per QALY. Now suppose a new drug delivers half a QALY of extra health per patient over the best current alternative. How much is that worth to the system? Half a QALY, at up to £30,000 each, is worth up to £15,000. So — at least as a ceiling — that's what the system should be willing to pay for the drug: the health it produces, valued at the system's own rate for health.

That's the whole idea of value-based pricing: the price is derived from the value the drug creates, measured in the currency the system already uses to value health. Not cost-plus, not guesswork — a price that falls straight out of the drug's health benefit and the system's willingness to pay for health. And the tool that makes this precise is one you've already met, running in reverse.

Reversing the ICER.

Recall the ICER from Module 7:

ICER = (Cost of new drug − Cost of comparator) / (Effect of new − Effect of comparator) = ΔCost / ΔEffect

All through Module 7, you used it forwards: a price was given, you computed the ICER, and you checked it against the threshold. But look at the equation again — it runs both ways. Instead of asking "given this price, what's the ICER?", ask "what price makes the ICER exactly equal the threshold?" Set ICER = threshold (λ), and solve for the drug's price.

Do it in full first. If λ is the threshold and ΔEffect the QALYs gained:

Price of drug (max) = λ × ΔEffect − (other new costs) + (comparator costs)

In words: the most the system can pay for the drug while staying cost-effective is the value of the health gained (λ × ΔEffect), adjusted for whatever other costs change. If we make the clean simplifying assumption that the other medical costs roughly cancel between the drug and its comparator, it collapses to something you can hold in your head:

Value-based price (max) = threshold × QALYs gained

That's the reversed ICER. £30,000 × 0.5 QALY = £15,000 — the exact price at which this drug's ICER lands right on the threshold. Charge less, it's a bargain; charge more, it fails cost-effectiveness. The price stopped being an input to the assessment and became its output. That single move — treating the threshold and the health gain as knowns and solving for price — is the engine of value-based pricing. (Keep the full formula in mind for real cases; the clean version assumes other costs cancel, which they rarely do exactly — but it captures the logic exactly.)

Compute the value-based price.

You set three things: the system's threshold (what it'll pay per QALY), the health gain (QALYs the drug adds over the comparator), and how the drug changes other medical costs (does it save on hospitalisations, or add to them?). Watch the value-based price — the reversed ICER — fall straight out. Then flip on what reference pricing would say instead, and watch the two logics diverge.

Value-based price (max) = (threshold × QALYs) − other cost change = (£30,000 × 0.5) − £0 = £15,000 per patient

Notice what each logic responds to. The value-based price moves — it climbs with the health gain and the threshold, and shifts with other cost changes, because it's derived from this drug's value to the system. The reference prices sit still — they know nothing about how well this particular drug works; they only know what other drugs cost. That's the whole distinction in one picture: one logic prices the drug by what it's worth, the other by what its neighbours charge. For a highly effective drug, value-based pricing justifies a high price reference pricing can't see; for a marginal me-too, reference pricing props up a price the drug's own value can't earn. Same drug, two logics, two prices.

Logic 2: price from comparison.

The second logic doesn't ask what the drug is worth at all. Reference pricing asks a flatter, more market-like question: what is paid for something comparable? — and sets the price by that external benchmark. It comes in two flavours.

Internal reference pricing looks sideways within the country. Drugs are grouped into therapeutic clusters — medicines treating the same condition in a similar way — and the price is pegged to others in the cluster (often the average, or the cheapest). The logic: if these drugs do broadly the same job, they should cost broadly the same, and a new entrant shouldn't be paid more than the existing options for equivalent effect.

External reference pricing — also called international reference pricing (IRP) — looks across borders. The country defines a basket of reference countries and pegs the drug's price to what those countries pay: the average, the lowest, or the median of the basket. The logic: why should we pay more than comparable nations for the identical product?

Both are administrative-market logics, not value logics. They don't measure QALYs or reverse any ICER — they anchor to a comparison. That makes them simple, transparent, and politically defensible ("we pay what others pay"). It also makes them blind to the one thing value-based pricing is built to see: how much health this specific drug actually delivers.

Now you.

Which pricing logic is each one?

1. Set the price so the ICER exactly meets the £30,000/QALY threshold.

2. Price it at the average of what four comparable drugs in the same class already cost here.

3. Price it at the average of its list price across a basket of other countries.

4. Add a margin to what it costs to manufacture each dose.

5. Work backwards from the QALYs gained and the payer's willingness to pay.

6. Cap it at the cheapest equivalent already reimbursed in the same therapeutic cluster.

When each logic breaks.

Neither logic is "the right one" — each is powerful where it fits and fails where it doesn't, and knowing the failure modes is what separates using them from being fooled by them.

Value-based pricing breaks when value can't be cleanly measured. It needs a credible QALY gain and a threshold. But if the evidence is too weak to quantify the benefit (Module 3's uncertainty, Module 11's confounding), the "value" is a guess dressed as arithmetic. And it strains at the extremes: for a life-saving therapy where patients (and society) resist putting any finite value on survival, the threshold logic feels obscene; for an ultra-rare disease, the numbers get vast. Value-based pricing is only as solid as the value estimate underneath it — and that estimate is often the softest part.

Reference pricing breaks when there's nothing good to reference. A genuinely first-in-class, breakthrough drug has no therapeutic cluster and no comparators — internal reference has nothing to anchor to. It can penalise innovation: a far better drug gets pegged to the price of the mediocre ones it replaces, because they're in the same "cluster." It can prop up expensive me-toos by averaging them into a comfortable class price. And IRP has a subtler flaw: it imports other countries' decisions wholesale, letting a nation's price be set by choices made elsewhere, for other health systems, with other priorities — outsourcing your own valuation. Reference pricing is only as sensible as the comparison it rests on.

So the two logics fail in opposite conditions: value-based struggles when value is unclear, reference struggles when comparison is unavailable or misleading. Which is exactly why real systems use them together — and why the numbers reference pricing rests on turn out to be, quite literally, fictional.

The list-price fiction.

Here's the twist that makes reference pricing stranger than it first looks, and it comes straight from Module 10: the price everyone can see is usually not the price anyone actually pays.

Recall list price versus effective price. The list price is the official, published number. The effective price is what the payer really pays after a confidential discount negotiated behind closed doors — and that discount can be large. Payers and manufacturers both often prefer it hidden: the manufacturer keeps a high headline price, the payer gets a real bargain, and neither has to advertise the true figure.

Now watch what that does to international reference pricing. IRP pegs a country's price to other countries' prices — but the only prices it can see are the list prices, because the real, effective prices are secret. So a whole system of international reference is anchored to numbers that are, in practice, fictions — prices that no one actually pays. This isn't a small technicality; it drives real behaviour. Manufacturers sequence their launches carefully (enter high-price countries first, so the references cascade favourably), fiercely resist cutting any list price (because one low list price contaminates the reference basket everywhere), and the industry ends up managing a set of public phantom prices alongside the private real ones.

This is also how the two logics coexist in practice. Value-based pricing tends to set the ceiling — the most a system should rationally pay for the health delivered. Reference pricing and confidential negotiation then work underneath that ceiling, pulling the effective price down toward what comparators cost and what other countries (appear to) pay. The list price stays up as a reference-basket token; the effective price is where the real deal lives. Price, in the end, is neither purely value nor purely comparison — it's value-based logic setting the sky, reference logic and secret negotiation deciding where under it the drug actually lands.

What's the price, and what does it mean?

A drug delivers 0.8 QALYs over its comparator. The health system's threshold is £30,000 per QALY. Assuming other costs roughly cancel, a manufacturer wants the highest price at which the drug is still cost-effective. What is it — and what does that number represent?

Why this matters for HTA

Pricing is where HTA stops being an assessment and becomes a lever — the one number in the whole system that's genuinely negotiable, and understanding how it's built is core to the practice:

A drug's price isn't discovered by counting what it cost to make — it's constructed, either from the value it creates or from what its neighbours charge. Learn to see which logic built a given price, and a number that looked like a fact turns back into what it always was: a choice.

Pricing, in one breath.

Two questions build every drug price: "what is this worth to us?" and "what do others charge for something like it?" They rarely give the same answer — and the gap between them, half-hidden behind confidential discounts, is where the real price is quietly decided.

We've now priced a drug assuming we know its value. But the recurring theme of this course is that we often don't — the effect is uncertain, the long-term benefit unproven, the value estimate soft. What does a payer do when a drug might be worth a high price but might not, and getting it wrong is expensive either way? The answer is a cleverer kind of deal — one that ties payment to what actually happens. That's the next lesson: risk-sharing and managed-entry agreements.