Reviewing Quantitative Studies: Meta-Analysis and Other Approaches

Many questions are answered using quantitative data. The main options are to analyse data using a narrative approach of tables and narrative explanation, or the formal pooling of data using meta-analysis. In this chapter, we outline a four-step approach to understanding quantitative data: what is the point estimate, how much variability or uncertainty is there about this, what is the clinical significance, and what is the statistical significance? When undertaking a meta-analysis, there are key decisions such as the outcome measure and meta-analysis model to be used. Key results include the point estimate with its confidence interval, prediction interval, and measures of heterogeneity. When the key assumption of studies being independent is violated, there are different approaches, namely, multilevel and multivariate analyses.