Skip navigation

Meta-analysis

 

A systematic review that uses quantitative methods to summarise the results. A downloadable paper is available.

A meta-analysis is where we pool all the information we have from a number of different (but similar) studies. It should not be about adding small piles of rubbish together to make a big pile of rubbish. It is only worth doing when individual trials are themselves of sufficient quality and validity. What meta-analysis does is to give enough size to have the power to see the result clearly, without the noise of the random play of chance.

Any meta-analysis must have enough events to make sense. Combining small, poor, trials, with few events will mislead.