Research guide5 min read
How to read risk numbers
A phrase like “cuts the risk in half” can mean a large or a very small difference, depending on the starting risk. Knowing the difference between absolute and relative risk is one of the most useful skills for understanding health news.
NIH example: if a disease affects 2 in every 100 middle-aged men during their lifetime, and a drug reduces this risk to 1 in 100, the relative reduction is 50% because 1 is half of 2, but the absolute reduction is one person in 100 (1%); that is, 100 people would need to be treated with this drug to prevent one additional case.
The way the same number is phrased changes our impression: the NIH notes that the statements "more than 20% of Americans will eventually die of cancer", "about 80% will not die of it" and "more than 1 in 5" express the same information but evoke different feelings. A "statistically significant" difference may also be too small to matter in practice to a patient, which is why statistical significance is distinguished from clinical importance.
The core idea
Absolute risk is the probability that something will happen, such as a health problem that may appear within a certain period of time.
Relative risk compares the absolute risk of one group with the absolute risk of another group.
A worked example
The example: a disease affecting 2 in 100, and a drug that reduces it to 1 in 100. The relative reduction is 50%, the absolute reduction is 1 in 100 (1%), and 100 people would need to be treated to prevent one additional case.
Looking at relative risk alone may give you the impression that the drug is very effective; this is why it is advisable to focus on absolute risk.
Wording changes the impression: "more than 20% will die", "about 80% will not die" and "more than 1 in 5" are one piece of information in three forms, and pictures can help, such as drawing 5 human figures with one in a different colour.
Numbers about broad groups may not apply to you: if you are not similar to their members, or the group is very broad, your risk may differ.
Questions to ask about any number
What is the risk before treatment or exposure, and what is the risk after? And does the number refer to a whole lifetime or a shorter period?
Is the difference large enough to be meaningful to you? A small difference may matter to scientists but not change much in your health or quality of life.
Is the result only "statistically significant"? Statistical significance means the difference between the groups is unlikely to be due to chance, whereas clinical importance concerns the size of the effect; a difference may be statistically significant and very small at the same time.
How many participants were there? Small samples are more prone to results produced by chance, and large samples may give more reliable results and represent people better.
When talking to your doctor about a test or treatment — such as decisions about routine screening — ask: what is my personal risk? What are the possible benefits and side effects? And what is the chance that it will happen to me?
Common pitfalls
Being dazzled by a large percentage without knowing the original risk: "half the risk" may mean going from 2 to 1 in 100.
Confusing "statistically significant" with "important for my health".
Limits of this page
The numbers in the example are illustrative as given by the NIH, and do not relate to a particular drug or disease.
This page does not calculate your personal risk; estimating individual risk needs a validated clinical model and discussion with the doctor.
Common questions
If a news item says a habit "doubles the risk", should I worry?
It depends on the original risk: doubling a very small risk still leaves a small risk. Look for the absolute risk before and after the doubling, and over what period of time, and ask your doctor about your personal risk.1
Do graphics help in understanding risk?
Yes, the NIH notes that pictures may help, such as showing 5 human figures with one in a different colour to express "1 in 5".1
Questions for your doctor
- What is my personal risk, and over how many years?
- What is the expected absolute benefit of this treatment or test, and what are its possible side effects?
- Do the results of this study apply to people in my situation?
References
- 1NIH News in Health. Understanding Health Risks. newsinhealth.nih.gov/2016/10/understanding-health-risksHealth agencies & guidelines · Accessed 2026-10-08
- 2NCCIH (NIH). Know the Science: Checklist for Understanding Health News Stories. www.nccih.nih.gov/health/know-science/facts-health-news-stories/checklist-for-understanding-health-news-storiesHealth agencies & guidelines · Accessed 2026-10-08
- 3NCCIH (NIH). Know the Science: How To Make Sense of a Scientific Journal Article — Results. www.nccih.nih.gov/health/know-science/how-to-make-sense-of-a-scientific-journal-article/resultsHealth agencies & guidelines · Accessed 2026-10-08
- 4NCCIH (NIH). Know the Science: How To Make Sense of a Scientific Journal Article — Size of the Study. www.nccih.nih.gov/health/know-science/how-to-make-sense-of-a-scientific-journal-article/methods/size-of-the-studyHealth agencies & guidelines · Accessed 2026-10-08
Review status: Edited content · Last updated:
Change log
- — Page created.
Educational content only — no diagnosis, and no substitute for a clinician.