Figure 1.
PRISMA diagram of the screening and selection process.
Figure 2.
Study network.
Table 1.
Summary of the proportion of patients gaining ≥10 letters (BCVA) from baseline to month 12 by study and treatment group.
Table 2.
Pair-wise odds ratios (95% credible intervals) from the random treatment effects model with baseline BVCA included as a covariate.
Table 3.
Probability that each treatment is the most efficacious treatment in the network.
Table 4.
Model selection: comparison of the total residual deviance and DIC for Bayesian network meta-analyses with fixed or random treatment effects and baseline BCVA and/or CRT as covariates.