Meta-analyses increase power

Meta-analyses not only provide researchers with a single pooled treatment estimate, but they also increase the probability of finding a significant treatment estimate. But why is this important?

Let’s imagine that we want to know whether treatment A is more beneficial than treatment B. Five small sized studies have been identified which have attempted to answer this question. The majority of the studies found no statistical difference in the effectiveness of the two treatments. Why might this be? It could be related to the number of individuals the researchers included in their studies, known as the “sample size”.

This is due to the concept of the “power of a study”. The power of a study is the probability that it will find a statistically significant result. A statistically significant result is indicated by a p value being less than 0.05. Smaller sample sized studies can be insufficiently powered and are subsequently unable to find results which are statistically significant.

To be able to conclude that treatment A is beneficial as compared to treatment B, we need to have sufficient people and hence power in our analysis.

By combining the results from the smaller studies using a meta-analysis we can increase the overall power of the analysis. Therefore, the results from a meta-analysis will usually have more power than the results from the individual studies.