Fig. 4From: Pattern-mixture model in network meta-analysis of binary missing outcome data: one-stage or two-stage approach?A panel of scatterplots on the within-trial standard error of log OR for ‘new versus old intervention’ (axis y) against the within-trial log OR for that comparison (axis x) for each simulation scenario. The colour key indicates the magnitude of covariance between the within-trial standard error of log OR and within-trial log OR for that comparison. MOD, missing outcome data; OR, odds ratioBack to article page