Research
Research explained
The word "study" sounds authoritative, but it covers work of very different weight: from observations of a few patients to trials with thousands of participants and reviews that combine dozens of such trials. This section explains how to make sense of them without specialist training and why some conclusions are more reliable than others.
An honest study accepts in advance that the method may not work. That is exactly why its conclusion deserves more trust than the most glowing testimonial.
Levels of evidence in plain words
| Type of data | What it can show |
|---|---|
| Systematic review | The overall picture from many trials on one question |
| Randomised controlled trial | Whether there is a benefit compared with placebo or usual care |
| Observational study | A link between a factor and an outcome, but not always its cause |
| Case report | An interesting observation worth testing |
| Opinion and personal experience | A reason to ask a question, but not the answer to it |
How to read a report about a study
Identify the question
What exactly was tested: whether the method helps, whether it is safe, for whom and for which condition.
Find the comparison group
If there was no comparison, the result only describes what happened to the participants, not why.
Look at the participants
What ages and conditions the participants had, and whether they are like you.
Tell a key outcome from a surrogate
A change in a lab value does not yet mean that people lived longer or felt better.
Check for reviews
Look for what systematic reviews say on the question, for example in the Cochrane Library.
Why it is easy to be misled without a control group
- The placebo effect: expecting a benefit improves how people feel by itself
- Natural course: many illnesses clear up or settle down on their own
- Regression to the mean: people often see a doctor when symptoms peak, and a decline usually follows
- Simultaneous changes: the person may have changed their routine or diet, or received other care
Related sections
- Glossary
Terms from evidence-based medicine and medical history in plain words: from placebo to systematic review.
- Methods
Evidence-based medicine, historical doctrines and alternative practices: what is known about them and how they are assessed.
- Prevention
How to spot an ad for a "miracle cure" and avoid losing time, money and health: a step-by-step check of the promises.
Sources
- Cochrane Library
- NHMRC — Statement on Homeopathy (2015)
- WHO — information for patients
This page reflects the state of knowledge on the date it was prepared. Scientific assessments and official guidance are revised as new research appears, so make decisions about your health together with your doctor.