How We Know What We Know

For Young Learners

What This Is

Every treatment in use today was once an idea somebody had. The process that turns an idea into something a surgeon offers you is called evidence, and it has a shape worth knowing.

The shape matters because the same claim can be true, half true or wrong depending entirely on how it was tested.

Knee surgery has a history of operations that were standard for years and then stopped being standard. In most of those cases nothing new was discovered. Somebody simply ran the test that should have been run first.

The same applies to anything you read about knees online. Learning to ask how do they know that is more useful than memorising any particular answer.

How It Works

Evidence comes in kinds, and they are not equally strong.

Somebody noticed something. One patient, one surgeon, one observation. Useful for generating a question and worthless for answering it.

A series of cases. Twenty patients who had an operation, and how they got on. Better, but with nothing to compare against you cannot tell whether the operation helped or whether they would have improved anyway.

A comparison. Two groups, one treated one way and one another. Now the question is answerable, provided the groups were genuinely similar to begin with.

A randomised trial. The groups are assigned by chance, which is the only reliable way to make them similar in the things nobody thought to measure.

A systematic review. Every trial on a question found, assessed and combined. This sits at the top because it is the only kind that tries to account for the trials that disagreed.

Most knees get better with time. That single fact is why comparison matters so much in this field. Almost any treatment looks effective if you do not compare it with doing nothing.

How to Prepare

Four questions work on almost any claim about health.

Compared with what? If there is no comparison, there is no finding.

How many people? Five is an anecdote. Five thousand is a study.

Who was left out? A trial of fit young men tells you less about a sixty year old woman than it appears to.

Who paid for it, and did they know the answer they wanted? This does not make a study wrong, and it is worth knowing.

Common Mistakes

Treating a single dramatic story as evidence. Our brains are built to find one convincing, and it is not.

Confusing two things happening together with one causing the other. More people drown in months when more ice cream is sold, and ice cream is not the cause.

Assuming newer means better. Many new operations have been tested less than the old ones they replaced.

Trusting a percentage with no denominator. A treatment that halves your risk has halved it from something, and that something is the part worth knowing.

Where to Find Out More

Section 12 of this site is the Research Academy and covers all of this properly.

Cochrane publishes systematic reviews written in plain language summaries alongside the technical versions, and those summaries are readable at school level.

The National Institute for Health and Care Excellence publishes the guidance that turns evidence into practice in England, and its documents show the reasoning.