NNT stands for “number needed to treat.” It tells you how many people need to take a treatment before one person benefits from it. An NNT of 5, for example, means that for every 5 people treated, 1 will be helped who wouldn’t have improved without the treatment. The lower the number, the more effective the treatment.
How NNT Works
NNT answers a deceptively simple question: if I take this medication, what are my actual chances of benefiting? Imagine a study where 20 out of 100 people in the untreated group have a heart attack, but only 10 out of 100 people in the treated group do. The treatment prevented 10 extra heart attacks per 100 people. That 10% difference is called the absolute risk reduction. NNT is the inverse of that number: 1 divided by 0.10 equals 10. So 10 people would need to take the drug for 1 person to avoid a heart attack.
This matters because many treatments sound impressive when described in percentages but look different through the lens of NNT. A drug that “cuts your risk in half” sounds powerful, but if your risk was only 2% to begin with, the drug only drops it to 1%. That’s an NNT of 100, meaning 99 out of 100 people who take it won’t benefit from it at all.
What Counts as a “Good” NNT
An NNT of 1 would be a perfect treatment: every single person who takes it benefits. That’s rare in medicine. In practice, NNTs vary enormously depending on the condition and what counts as a “benefit.” A few real-world examples help put the numbers in perspective.
For ear infections in children, antibiotics have an NNT of about 7 to 20 for reducing pain after the first 24 hours. That means you’d treat somewhere between 7 and 20 kids with antibiotics before one child experiences less pain than they would have without treatment. The NNT for preventing a ruptured eardrum is 33, meaning most children prescribed antibiotics for ear infections would have been fine without them.
For statins used in people at low risk of heart disease, the numbers are even larger. The NNT for preventing a nonfatal heart attack over the study period was 217, and for preventing a nonfatal stroke it was 313. These are not small benefits for the individuals who are helped, but they do mean that the vast majority of low-risk people taking statins daily won’t personally avoid a cardiac event because of the medication. In higher-risk populations, the same drug will have a much lower (better) NNT because more events are being prevented.
Why the Same Drug Can Have Different NNTs
One of the most important things to understand about NNT is that it changes depending on who’s being treated. A blood pressure medication tested in people with severe hypertension will show a lower NNT than the same medication tested in people with mildly elevated blood pressure. The drug works the same way in both groups, but there are simply more bad outcomes to prevent when the starting risk is higher.
Time frame matters just as much. A cholesterol-lowering drug studied over 1 year will have a higher NNT than the same drug studied over 5 years, because more heart attacks and strokes accumulate over longer periods. An NNT from a 1-year statin trial can’t be directly compared to one from a 5-year trial. Whenever you see an NNT, it should come with the specific treatment, dose, outcome being measured, and the time period of the study. Without those details, the number is almost meaningless.
NNT vs. NNH: Balancing Benefit and Harm
Every treatment has side effects, and those are measured with a parallel statistic called NNH, or “number needed to harm.” NNH tells you how many people need to take a treatment before one person experiences a significant side effect. The calculation works the same way as NNT, just applied to adverse events instead of benefits.
Comparing the two numbers gives you a practical sense of whether a treatment is worth it. If a pain medication has an NNT of 3 (one in three people gets meaningful relief) and an NNH of 50 (one in fifty gets a serious side effect), the benefit clearly outweighs the risk. But if the NNT is 50 and the NNH is 10, you’re far more likely to be harmed than helped. A low NNT-to-NNH ratio signals a favorable tradeoff; a high ratio signals caution.
How NNT Helps You Make Decisions
NNT is useful because it cuts through the way treatment benefits are often presented. Drug advertisements and even doctors tend to describe benefits in relative terms: “reduces your risk by 50%.” That framing can make a tiny absolute benefit sound dramatic. NNT forces the conversation back to absolute terms by asking the blunt question: how many people like me need to take this for one of us to actually benefit?
If your doctor recommends a medication and you want to understand what you’re signing up for, asking about the NNT is a reasonable place to start. A treatment with an NNT of 5 is very different from one with an NNT of 200, even if both are described as “effective.” Knowing where a particular treatment falls on that spectrum, along with the NNH for side effects, gives you a clearer picture than percentages alone.
NNT values are always rounded up to the nearest whole number, because you can’t treat a fraction of a person. And they always describe averages across a population. There’s no way to know in advance whether you’ll be the 1 person who benefits or one of the others who doesn’t. But having the number gives you a realistic expectation rather than an inflated one.