The UK Faculty of Public Health has recently taken ownership of the Health Knowledge resource. This new, advert-free website is still under development and there may be some issues accessing content. Additionally, the content has not been audited or verified by the Faculty of Public Health as part of an ongoing quality assurance process and as such certain material included maybe out of date. If you have any concerns regarding content you should seek to independently verify this.

Interpreting meta-analyses

The outcomes considered in meta-analyses are, of course, similar to those reported in randomised controlled trials with the exception of standardised mean difference (SMD), which is sometimes used in meta-analysis where the outcomes are measured in different ways in the included trials. The statistical methods are also similar.

In a meta-analysis, trial results are pooled as a weighted average, with the weighting being inversely proportional to the variance (the standard error squared) of the trial result. This weighting is chosen because the variance is an indicator of the amount of information contributed by the trial. Larger variances correspond to less information and so the inverse of the variance is chosen as the weight, giving more influence to the result of the trials with smallest standard errors (which will usually also have the largest sample sizes). The variances themselves are also pooled and the square-root taken to estimate the standard error of the pooled estimate, allowing p-values and confidence intervals to be calculated as usual.

The results of a meta-analysis are usually presented on a "blobbogram" (or forest plot) and must be interpreted carefully, taking account of any heterogeneity in the results of the individual trials. These two topics are discussed in more detail in the rest of this section.