Module 1 · PICO

"Does Drug X work?" — a question with no answer

Last lesson ended on a sharp little question: compared to what? It turns out that's only the first of four things you have to pin down before "does it work?" means anything at all.

Here's a drug — call it Drug X, a new pill for type 2 diabetes. Its maker says, plainly: it works.

'Does Drug X work?' — is that a question you can actually answer as it stands?

'Does it work?' has no answer until you say: work for whom, instead of what, measured how. Pin those down and the question becomes answerable. Leave them vague and any answer is just a guess wearing a number.

One question, four slots

Imagine two honest analysts, same drug, same data. One concludes Drug X is a clear success. The other concludes we can't tell. Neither lied. They simply answered different questions — because "does it work?" was never one question. It was a blank with four slots, and they filled the slots differently.

HTA's tool for filling those slots — the same one used across evidence-based medicine — is an acronym you'll see on every submission you ever read:

PICO — Population, Intervention, Comparator, Outcome.

It looks like a form. It behaves like a verdict. Let's take the slots one at a time.

The four slots

Before we fill them, here's the whole shape — four questions hiding inside "does it work?":

Three of these hide a choice, and two of them — C and O — are where HTA decisions are quietly won and lost. Watch for them.

You'll also meet variants — PICOS adds study design, PICOT adds a timeframe — but these four slots are the heart of it.

P — Population

Work for whom?

A drug doesn't work "in general." It works, or doesn't, in particular people. And the people a drug was tested on are rarely the exact people who'll receive it — or the exact people the funding decision is about.

Get the population wrong and everything downstream answers the wrong question: who actually benefits, who gets displaced to pay for it, and whether the trial's result even transfers to your patients at all.

Drug X is licensed for 'type 2 diabetes.' Who should the funding question be framed around?

I — Intervention

The what.

This slot looks trivial — it's just the new drug, surely? But "Drug X" still hides choices that change the whole question: at what dose? at which point in the pathway? on its own, or added on top of current treatment?

That last one matters most, because it silently sets up the next slot. "Drug X instead of current care" and "Drug X added on top of current care" are different interventions — and they demand different comparators.

How should Drug X's intervention be specified?

C — Comparator

Better than what?

This is the slot you were promised last lesson — and the one most often quietly rigged. Your patients are never getting nothing. So the only comparison that answers the decision is against the care they'd otherwise receive. Beat that, and you've shown something the system cares about. Beat anything else, and you've shown something irrelevant.

What should Drug X be measured against?

O — Outcome

Better how?

A drug can move many things. The question is whether it moves the thing that matters — to the patient, and to the decision. Trials often lead with whatever is quickest and easiest to measure. That's not always the same as what people actually care about.

Drug X's trial can report several things. Which should the funding decision rest on?

The payer's PICO

Put the four slots together and watch "does it work?" turn into something you can actually answer:

In patients whose type 2 diabetes is uncontrolled on standard care (P), does adding Drug X (I), compared with continuing standard care alone (C), reduce heart attacks, kidney failure and deaths, and improve quality of life (O)?

One sentence. But notice what changed: this question has an answer. "Does Drug X work?" never did. That's the entire job of a PICO — to convert a slogan into a question evidence can speak to.

The flip

Now the part that shows why PICO is power, not paperwork.

Here's the PICO Drug X's maker would love to be judged on:

In fit trial patients (P), does Drug X (I) versus a placebo (C) lower blood sugar (O)?

On that question, Drug X is a dramatic success.

Keep the same drug and the same data — but swap in the payer's comparator (standard care) and outcome (real patient events). What happens to 'dramatic success'?

Same drug. Same trial. Opposite picture — because the PICO, not the drug, decided what 'working' meant. Against standard care instead of placebo, the extra benefit may be small or untested. On hard outcomes instead of blood sugar, it may be unproven. The triumph becomes a shrug, and nothing about the molecule changed.

PICO isn't the question you ask before the analysis. It's where the analysis is quietly already decided.

Whoever sets the PICO, sets the answer

This is why PICO is fought over, not filled in. A manufacturer will, quite rationally, propose the PICO that shows their drug at its best: the flattering population, the weak comparator, the convenient outcome. None of it is fraud — it's framing.

So the HTA body's very first job — before a single result is weighed — is to insist on the right PICO: the real population, the comparator patients actually receive, the outcomes that matter. Win that, and half the assessment is already settled. (You'll do exactly this unpicking yourself in M14, tearing apart a manufacturer's submission.)

One last twist: "standard care" isn't the same in every country, so a single drug can need many PICOs at once. Europe's joint assessments wrestle with precisely this — more when we reach the regulatory landscape in M12.

PICO, in one breath

A sharp question is already half the answer. PICO is how HTA makes the question sharp.

You can now frame the question. The rest of this course is about answering it — and that begins with learning to read the evidence itself. That's where M2 picks up.