Anti-vascular endothelial growth factor for diabetic macular edema: a Bayesian network analysis

Aims Comparison of the efficacy of six Anti- vascular endothelial growth drugs in the treatment of diabetic macular edema Methods This network meta-analysis has been registered on the PROSPERO platform (CRD42022295684).A comprehensive search of eight databases without language restrictions.PubMed, EMBASE, Web of Science, Cochrane Library, CBM, CNKI, VIP, and Wanfang were used to search for randomized controlled trials on anti- vascular endothelial growth factor of diabetic macular edema, no language restrictions and publication time restrictions. According to the inclusion and exclusion criteria, literatures were screened, data were extracted and literature quality was evaluated, and the mean changes in best-corrected visual acuity (BCVA) and central macular thickness (CMT) were obtained. Using the Gemtc 1.0-1 package in R 4.1.2 to call the JAGS and the Markov Chain-Monte Card (MC-MC) method for Bayesian network meta-analysis. Risk of bias was assessed using RevMan 5.3. Stata 14.0 draws funnel plots to assess publication bias. RESULTS Sixteen studies including 3651 eyes were included, all with treatment follow-up over 6 months. The overall heterogeneity in the network analysis was low(I ² =0), and there was no inconsistency. For the efficacy of reducing CMT, ranking according to the cumulative probability: Faricimab (0.9) > Brolucizumab (0.87) > Aflibercept (0.58) > Conbercept (0.37) > Ranibizumab (0.29)>Bevacizumab (0), but there was no statistically significant difference between Conbercept and Faclibercept, Aflibercept, and Buloxizumab. For the efficacy of improving BCVA, ranking according to the cumulative probability: Conbercept (0.87) > Faricimab (0.79) > Aflibercept (0.61) > Brolucizumab (0.51) > Ranibizumab (0.2)>Bevacizumab (0.02), but there was no statistically significant difference between Conbercept and Faclibercept, Aflibercept, and Brolucizumab.