Summary

In summary, risk ratio and odds ratio are better for meta-analysis than the risk difference. Risk ratios are easier to understand than odds ratios, but require some care in choosing whether to analyse the risk ratio of the event or non-event.

This overall view is summarized in the following table:

Once the analysis is performed the results require careful interpretation. Odds ratios must not be interpreted as if they were risk ratios unless events are very rare. It will be helpful to a reader if any relative effects are re-expressed as absolute effects (RD or NNT) - maybe working out these figures for a range of possible scenarios.

Some review groups will prefer you to use a particular measure to give uniformity across their reviews, and you should check with them or on The Cochrane Library to see if this is the case. If, after reading this module, you don't agree with their policy, you could always challenge them to justify it.

In the next module we will discuss how results from individual trials are combined to give an overall estimate of treatment effect.