Analysing data and undertaking meta-analyses
Deeks JJ., Higgins JPT., Altman DG.
© 2019 The Cochrane Collaboration. This chapter describes the principles and methods used to carry out a meta-analysis for a comparison of two interventions for the main types of data encountered. A very common and simple version of the meta-analysis procedure is commonly referred to as the inverse-variance method. This approach is implemented in its most basic form in RevMan, and is used behind the scenes in many meta-analyses of both dichotomous and continuous data. Results may be expressed as count data when each participant may experience an event, and may experience it more than once. Count data may be analysed using methods for dichotomous data if the counts are dichotomized for each individual, continuous data and time-to-event data, as well as being analysed as rate data. Prediction intervals from random-effects meta-analyses are a useful device for presenting the extent of between-study variation. Sensitivity analyses should be used to examine whether overall findings are robust to potentially influential decisions.