ABSTRACT
A graphical method that is used to evaluate heterogeneity of effect among the different primary studies included in a meta-analysis, and which may be employed with binary outcomes (L’ABBE, DETSKY & O’ROURKE, 1987). For each individual study, the risk or odds in the exposed or treatment group is plotted on the y-axis against the risk or odds in the control group on the x-axis, as shown in Figure L.1 (BAX et al., 2009). The area of each plotted circle is proportional to the weight of the corresponding study. A L’Abbe plot often shows the regression line of treatment vs. control results (with greater between-study heterogeneity corresponding to greater scatter around this line), in addition to a central diagonal line that represents absence of treatment or exposure effect (i.e. no difference in risk of outcome between exposed and control groups); both are shown in Figure L.1. See also forest plot, Galbraith plot, index of heterogeneity, meta-regression.
