Module 1 · Efficacy vs Effectiveness

A regulator just said yes. Are you done?

You've already seen that "approved" never means "funded." Time to find out why — and the answer turns out to be about evidence, not just money.

A new drug clears the regulator. The verdict is clean: it works, and it's safe enough to sell. The file lands on your desk for a funding decision.

Two gates

Every funded treatment passes through two separate gates — not one.

Different gatekeepers. Different questions. Clearing the first tells you nothing about the second — and the evidence that opens Gate 1 was built for Gate 1's question, not yours.

The regulator's question

Stand at the first gate. What does a regulator actually want to know?

Two things: Can this drug work at all? and Is it safe enough to allow? That's it. Not "is it worth the money," not "is it better than what we already use" — just can it work, and is it acceptably safe.

To answer that cleanly, regulators rely on a tightly controlled trial: carefully chosen patients, ideal conditions, close monitoring. The whole design exists to give a clear yes/no on one narrow question — does this drug, under good conditions, do something real?

That's a perfectly good question. It's just not the one you're being asked to decide.

The HTA question

Now stand at the second gate — your gate. Your question is bigger, and messier:

Will this deliver enough real health, in our actual patients, beyond the care they already get, to be worth what it displaces?

Notice three things smuggled into that sentence that the regulator's trial never had to address: our patients (not a hand-picked sample), real health delivered (not health under ideal conditions), and beyond current care (not "beyond nothing").

So before you trust the trial's impressive result, you have to ask a harder question: how much of it survives contact with reality?

Stress-testing the trial

Meet Drug Z. In its trial, patients on Drug Z did far better than those given a dummy pill — a big, clear benefit. The regulator was satisfied: it works, it's safe, it's approved.

Now it's your decision. Below are four ways the real world differs from that trial. For each one, predict what happens to Drug Z's benefit when it meets reality.

Same three options every time. Trust your instinct — you'll be right more than you expect.

Efficacy vs effectiveness

Two words for two questions.

Efficacy — does it work under ideal, controlled conditions? Hand-picked patients, perfect adherence, expert centres, often measured against a dummy pill. This is what the regulator's trial is built to show.

Effectiveness — does it work in routine, real-world practice? Ordinary patients, imperfect adherence, ordinary clinics, measured against the care people already get. This is what HTA needs to know.

The distance between them has a name: the efficacy–effectiveness gap. As you just saw, it almost always runs one way — effectiveness is usually lower than efficacy, because every real-world condition chips away at the ideal-world result.

A drug can have excellent efficacy and disappointing effectiveness. Same drug. Different question.

The sharpest gap: the comparator

One of those four rounds wasn't like the others. Three of them — patients, adherence, setting — quietly shrink a benefit. The comparator round can erase it.

Beating a dummy pill tells you the drug does something. It says nothing about whether it does more than the treatment your patients already receive. A drug can be genuinely efficacious and still add almost nothing over current care — and "almost nothing extra" is rarely worth funding, however clean the trial.

So HTA fixes on a question the regulator can wave through: compared to what?

In the next lesson you'll meet the tool that forces that question into the open for every study you read — PICO. It turns a vague "does it work?" into a question precise enough to actually answer.

Why two gates?

So why have two gates at all? Isn't HTA just second-guessing the regulator?

No. They're not asking the same question twice — they're asking two different questions, and neither answers the other.

HTA isn't redoing the safety review or distrusting the science. It's answering the question the trial was never designed to ask. "Approved" was always the answer to a different question.

In practice the line is blurrier than two tidy gates — sometimes even efficacy is uncertain at approval, and regulators and HTA bodies increasingly talk to each other. We'll get into that machinery later. The two-question logic underneath it doesn't change.

Two gates, two questions

• A treatment passes two separate gates: the regulator (may it be sold?) and HTA (should we pay?).
• The regulator's trial shows efficacy — does it work under ideal conditions?
• HTA needs effectiveness — does it work in real practice, against the care patients already get?
• The efficacy–effectiveness gap runs downhill: real patients, real adherence, real comparators and real clinics all chip away at the trial's result.
• The sharpest gap is the comparator: better than nothing is not better than what we do now.

"The regulator approved it" tells you the drug can work. It doesn't tell you it will work — for your patients, in the real world, better than what they already get. Closing that gap is HTA's whole job.

Next, the instrument that pins it all down — turning "does it work?" into a question sharp enough to answer: PICO.