M7 · COST-EFFECTIVENESS
Every comparison is two numbers.
Strip an HTA decision down to its core and only two things matter about a new technology: how much more health does it buy, and how much more does it cost?
Health is measured in QALYs. Cost is measured in pounds. Everything else in a dossier — the trials, the models, the survival curves — is machinery for pinning down just those two numbers.
We always measure both as a difference from what you're already doing — the comparator. So we talk about the incremental effect (ΔEffect) and the incremental cost (ΔCost).
The comparator sits at zero.
The technology you're already using — standard care — is your reference point. Compared with itself it has no incremental cost and no incremental effect, so it sits at the origin.
Every new technology is then placed by how far it moves you from there: right if it adds health, left if it takes health away; up if it adds cost, down if it saves money.
Good news points down and right.
Put ΔEffect on the horizontal axis and ΔCost on the vertical one and you get the cost-effectiveness plane — the single map the rest of this module is drawn on.
Right means more health. Down means less money. So the friendliest direction is down-and-to-the-right: more health for less money. The plane's whole job is to tell you, at a glance, which kind of decision you're facing — before you've read a single p-value.
Four quadrants, two kinds of decision.
Drag the point. It's a new technology compared with standard care. Watch which quadrant it lands in — and read what each one means.
ΔEffect: +0.30 QALYs · ΔCost: +£6,000
Trade-off (north-east) — more effective but more costly.
Verdict: Needs a rule — that's the ICER, next.
Notice something: two corners give you an answer immediately. The other two just say you need a rule. Hold onto that — it's the shape of the whole module.
The free decisions: dominance.
Two of the four quadrants settle themselves.
Down-and-right — more effective and cheaper — a technology there is dominant. You adopt it. There's nothing to trade off: it wins on both axes, so no threshold, no value judgement, no debate.
Up-and-left — less effective and more costly — a technology there is dominated. You reject it, for the mirror-image reason: it loses on both axes.
Dominance is the strongest thing you can say on the plane, and — this is the point — it needs no notion of what a QALY is worth.
One warning: dominant has a precise meaning here — better on BOTH dimensions. A technology that is merely cheaper, or merely more effective, is NOT dominant. When a dossier reaches for the word, check that both arms actually earn it.
The decisions that need a rule.
That leaves the two diagonal quadrants — and this is where almost all the real work of HTA happens.
North-east (more effective, more costly) is where most new drugs land. You're buying extra health, but you're paying for it. Is it worth the price? The plane can't say. You need an exchange rate between money and health.
South-west (cheaper, less effective) is the mirror image: you'd save money, but you'd give up some health. Worth it? Same missing rule. This is the quadrant of disinvestment and de-adoption — often the most politically charged decisions of all.
Supplying that missing rule is the entire job of the next three lessons: the ICER, the threshold, and net benefit.
Complete the reading.
A redesigned heart-failure care pathway is compared with usual care. Across a patient's lifetime it delivers 0.15 additional QALYs and saves £1,200 per patient. Complete the reading:
This option iseffective andcostly, so the plane says:.
Where does it sit, and what does that tell you?
A generic substitution would cut drug costs by £900 per patient, but comes with slightly worse outcomes: −0.02 QALYs per patient. Where does it sit, and what does that tell you?
Why this matters for HTA
A manufacturer's dossier lands on your desk with one word doing a lot of heavy lifting: cost-effective. Before you engage with any of their modelling, you do one thing — you locate their technology on the plane.
- Most new drugs live in the north-east quadrant — more effective, more expensive — so almost every appraisal is really a question about that one quadrant. The plane tells you that before you read a single number.
- Dominance claims are the easiest to check and the easiest to overstate. A dossier claiming "dominant" is claiming cheaper AND better — that earns scrutiny of both arms, not just the headline.
- The quadrant a technology sits in decides which tool you reach for next — threshold, ICER, net benefit. Reading the plane first stops you applying the wrong one.
Before you ask whether a technology is worth it, ask where it sits. Half the time, the plane has already answered.
The cost-effectiveness plane, in one breath.
- Every comparison is two numbers: how much more health (ΔEffect, in QALYs) and how much more money (ΔCost, in £).
- The comparator sits at the origin; the new technology lands in one of four quadrants.
- Down-right = dominant (cheaper and better) → adopt. Up-left = dominated (costlier and worse) → reject. No threshold required.
- North-east (better, costlier) and south-west (cheaper, worse) are genuine trade-offs — they need a rule.
The plane doesn't tell you what a QALY is worth. It tells you when you need to ask.
That rule — the exchange rate between money and health — is what the rest of Module 7 builds. Next: the ICER, the single number that tries to price the north-east quadrant.