As a heart surgeon, you regularly perform aortic valve replacement, using a median sternotomy technique. However, you are aware that less-invasive options are available, and you wonder whether these may be as effective and safe as your current practice. After searching the Cochrane Library, you find a systematic review, ‘Limited versus full sternotomy for aortic valve replacement1’. The authors of the review conclude that ‘upper hemi‐sternotomy may have little to no effect on mortality versus full median sternotomy (risk ratio (RR) 0.93, 95% CI 0.45 to 1.94)’. What does this statistic mean in practice?

The risk ratio (also known as RR) illustrates the risk of an event occurring in the intervention group compared with the risk of the same event occurring in the control group. For example, in a hypothetical randomised controlled trial (RCT), 50 out of 100 people who received the experimental drug experienced at least one side effect: the risk (or probability) of the event was 50/100, or 0.5. In the control group, 20 out of 100 people who received a placebo experienced at least one side effect: the risk of the event was 20/100, or 0.2. The risk ratio was therefore 0.5/0.2, or 2.50. This means that the probability of experiencing at least one side effect was more than double in the experimental group compared with the control group.

It is important to remember that the risk ratio is a relative effect measure and does not tell us about the actual numbers of people who may experience an event. In the example of the RCT above, if only five people in the experimental group and two people in the control group experienced an adverse event, the risk ratio would still be 2.50 (0.05/0.02). Understanding this may result in a different clinical decision.

The risk ratio …