Effect of external cephalic version in a resource-limited setting on the Thailand-Myanmar border: a retrospective cohort with propensity score analysis

External cephalic version (ECV) is recommended to reduce the risk of breech presentation at birth. This study analysed the effect of external cephalic version (ECV) or not, on breech presentation at birth in a resource-limited setting. Women with ultrasound confirmed breech presentation from 28 weeks gestation at antenatal clinics of the Shoklo Malaria Research Unit (SMRU) and a known pregnancy outcome, from 2008 to 2018 along the Thailand-Myanmar border were included. Propensity score analysis using inverse probability weighting compared breech at birth between women who had ECV offered or not. Pregnancy outcomes were compared between ECV successful and unsuccessful versions. Adverse perinatal outcomes included cord prolapse, fetal distress, Apgar < 7 at 5 min, stillbirth and early neonatal death. Among 504 women with breech presentation between 35–37 weeks, 330 were offered ECV and 174 were not. Breech at birth in women offered ECV was 50.9% (168/330) compared to 47.7% (83/174) in women with no ECV attempt. In other words, the rate of spontaneous version from breech to cephalic in women with no ECV attempt was 52.3% (91/174). Propensity score analysis indicated no association between ECV being offered or not and breech presentation at birth (adjusted Odds Ratio 1.23, 95% confidence interval 0.82–1.83). Caesarean section for breech 22.4% (74/330) vs 20.1% (35/174), p = 0.540; and adverse perinatal outcomes, 5.2% (17/330) vs 7.5% (13/174), p = 0.295, were similar whether ECV was offered or not. Among all women offered ECV (n = 537, range 32–40 weeks), breech at birth (4.4% (12/273) vs 90.9% (240/264), p < 0.001; caesarean section for breech 1.1% (3/273) vs 40.5% (107/264), p < 0.001; and adverse perinatal outcomes 3.7% (10/273) vs 9.1% (24/264), p = 0.010, were significantly lower in the successful vs unsuccessful groups. ECV was safely offered in a resource-limited setting. Comparison of ECV offered or not at 35 to 37 weeks suggested no benefit or harm with respect to presentation at birth in contrast to comparison of ECV success or unsuccessful at 32 to 40 weeks. Improving the ECV success rate of health care practitioners, or task shifting, could positively contribute to optimising the potential benefits of ECV.