Overall analysis
On pairwise comparison, the combined treatment of pessary and vaginal progesterone did not reduce the risk of sPTB<34 weeks when compared to no intervention (OR 0.68; 95%CI 0.16-2.91). Also, pessary alone (OR 0.78; 95%CI 0.49-1.25), vaginal progesterone alone (OR 0.79; CI95% 0.45-1.41) and injectable 17-OH progesterone alone (OR 0.85; CI95% 0.26-2.75) did not showed reduction in PTB. None of the pairwise comparisons reached statistical significance. (Table 2 & Figure 1) Compared to the control group, pessary + vaginal progesterone ranked first according to the SUCRA values, with pessary, vaginal progesterone and 17-OH progesterone ranked second, third, and fourth respectively (Table S3).
For overall PTB<34 weeks, the combined treatment pessary + vaginal progesterone did not show a statistically significant reduction in PTB (OR 0.56; 95%CI 0.17-1.82) compared to the control group. Pessary alone (OR 0.76; 95%CI 0.55-1.06) and vaginal progesterone alone (OR 0.88; CI95% 0.63-1.22) also did not show reduction in PTB. Injectable 17-OH progesterone showed a non-statistically significant increase on risk of PTB<34 (OR 1.42; CI95% 0.92-2.20). Again, none of the pairwise comparisons did reach statistical significance. (Table 2 & Figure 1) Compared to the control group, the rank according to the SUCRA values was pessary + vaginal progesterone, pessary, vaginal progesterone and 17-OH progesterone (Table S3).
Table 2. Comparative treatment effect of intervention to prevent preterm birth in twin gestations compared to control group for spontaneous PTB<34 weeks and overall PTB<34 weeks
Figure 1 - Treatment effect of intervention compared to control to prevent preterm birth in twin gestations compared to control group for spontaneous PTB<34 weeks and overall PTB<34 weeks
Figure 2 - Evidence network diagram of network meta-analysis comparisons for spontaneous PTB<34 weeks and overall PTB<34. The width of each edge is proportional to the number of randomized controlled trials comparing each pair of treatments, and the size of each treatment node is proportional to the number of randomized participants
For overall PTB<34 weeks, the p value for overall network consistency test was 0.489, indicating overall satisfactory consistency. When performing node splitting test, the p values for each comparison ranged between 0.489 and 0.998, suggesting minimal local inconsistency. The network consistency test was not possible for sPTB<34 weeks because there was no closed loop in the network (Figure 2).