Absolute Risk Reduction and NNT

Relative measures like the risk ratio and odds ratio can make small effects look huge. Absolute measures keep you grounded in what actually happens to patients. The absolute risk reduction (ARR), also called the risk difference, is simply the control risk minus the treated risk, the real percentage-point drop in events. If a drug cuts risk from 10% to 6%, the ARR is 4 percentage points, or 0.04. From the ARR comes the number needed to treat (NNT), equal to 1 divided by the ARR. It answers a wonderfully concrete clinical question: how many patients must I treat to prevent one event? An ARR of 0.04 gives an NNT of 1 / 0.04 = 25, meaning you treat 25 patients to prevent one event. A small NNT signals a powerful, efficient treatment; a large NNT means many people are treated for each one who benefits. The mirror image for harms is the number needed to harm (NNH), computed the same way from the absolute risk increase. Why do these matter so much? Because the same relative effect can have wildly different absolute value depending on the baseline risk. A 50% relative risk reduction produces an ARR of 10 points (NNT 10) when baseline risk is 20%, but only 0.5 points (NNT 200) when baseline risk is 1%. The relative effect is identical; the value to a patient is not. This is exactly why guidelines and shared decision-making lean on the ARR and NNT rather than the relative risk alone. Two practical habits: always present relative and absolute measures together, and quote the NNT over a stated time horizon, for example 'NNT 25 over 5 years', because preventing one event over five years is very different from preventing one over five weeks. Reporting the NNT with a confidence interval is good practice too.