Meta-regression analysis
The meta-regression analysis was done to investigate whether the variables of geographical location, assessment scales and evaluation time of traumatic birth/PTSD, the type of participants, study design, year of publication, and the number of events (birth trauma and PTSD) had impacted on the effect size (Table 6). A significantly different incidence of postpartum PTSD after birth trauma was identified in Asia (P =0.03, 95% CI: 1.03 ~1.60) compared to Europe, North America, and Oceania. Meta-regression was also conducted to check the difference in rates in the samples sizes of birth trauma and PTSD. And the sample sizes of birth trauma (P =0.02) and PTSD (P =0.02) were potential variables for effect size, while no significant association was found in study design, the year of publication, and the type of participants of studies. Furthermore, scales and time of assessment for birth trauma and postpartum PTSD had no statistical significance in the effect size, indicating that assessment scales and evaluation time were not the sources of heterogeneity.