Absolute versus relative risk: the difference that changes everything

Key points#

• Relative risk is the proportional change in an outcome; absolute risk is the real change in your chances.
• A big relative reduction can be a small absolute one if the starting risk was low, which is how the same result can sound dramatic or trivial.
• Number needed to treat translates benefit into concrete terms: how many people must be treated for one to benefit.
• To judge a claim, find the actual rates in each group, not just the percentage reduction.

If you learn only one idea about reading health statistics, make it this one. The difference between absolute and relative risk explains why a single study can produce both a thrilling headline and a shrug from a careful clinician. Once you can see the difference, a great deal of health news becomes easier to interpret, and harder to be misled by.

Two ways to describe the same result#

Imagine a treatment that lowers the risk of some event. There are two honest ways to describe what it does, and they can feel very different.

Relative risk describes the proportional change. If a treatment takes risk from 2 in 100 down to 1 in 100, that is a 50 percent relative reduction. Half the risk. It sounds powerful, and it is the figure most likely to appear in a headline.

Absolute risk describes the actual change in your chances. In that same example, the risk went from 2 percent to 1 percent, an absolute reduction of one percentage point. One fewer person out of a hundred.

Both statements are true. They describe the identical result. But "cuts risk by 50 percent" and "reduces risk by one percentage point" land in completely different places, and only one of them tells you what to expect for yourself.

Why this matters so much#

The catch is that relative figures are detached from your starting point. A 50 percent relative reduction is impressive when the starting risk is high and almost meaningless when it is tiny. Halving a 40 percent risk saves twenty people in a hundred. Halving a 0.2 percent risk saves one person in a thousand. The relative number is the same; the real-world impact is worlds apart.

This is not usually about anyone trying to deceive. Relative figures are legitimate and often reported in good faith. But because they sound larger, they dominate headlines and marketing, and they can make a modest benefit feel like a breakthrough. The fix is simple awareness: whenever you see a percentage reduction, ask the next question.

The question to always ask#

When a claim says a treatment "reduces risk by" some percentage, ask: reduced from what, to what? Find the actual rates. How many people out of a hundred, or a thousand, had the outcome with the treatment, and how many without it? Those two numbers give you the absolute picture, and the absolute picture is the one that tells you whether a benefit is large or small for someone like you.

A related and very useful figure is the number needed to treat, which turns benefit into something concrete: how many people have to take a treatment for one of them to benefit. A number needed to treat of 10 means a sizable benefit; a number of 500 means a much smaller one. It is essentially absolute risk made tangible, and it is one of the most honest ways to express what a treatment does.

A worked example to keep#

Hold onto this example, because it generalizes. Suppose a study reports that a medicine "reduces the risk of a condition by a quarter." If the condition affected 4 in 100 people, a quarter reduction takes it to 3 in 100, an absolute gain of one point, with a number needed to treat of 100. If the same condition affected 40 in 100, the same quarter reduction takes it to 30 in 100, ten points of absolute benefit, with a number needed to treat of 10. Same relative claim, tenfold difference in what it means. The relative number told you almost nothing on its own.

Reading risk numbers from here on#

You do not need statistics training to use this. When you read that something cuts risk by some percentage, treat that as the beginning of the question, not the answer. Look for the absolute numbers, or ask your clinician to translate the claim into them, and ask how many people would need the treatment for one to benefit. This single habit will protect you from a surprising amount of overstatement, and it will help you make decisions based on what a treatment is actually likely to do for you, rather than on how good it sounds.