The MamaMiso study of self-administered misoprostol to prevent bleeding after childbirth in Uganda
- September 14th, 2015
- Postpartum Hemorrhage
- Weeks, A.D., Ditai, J., Ononge, S., Faragher, B., Frye, L.J., Durocher, J., Mirembe, F.M., Byamugisha, J., Winikoff, B., Alfirevic, Z.
BMC Pregnancy Childbirth; 14 September 2015; 14;15:219; doi:10.1186/s12884-015-0650-9
Background: 600 mcg of oral misoprostol reduces the incidence of postpartum haemorrhage (PPH), but in previous research this medication has been administered by health workers. It is unclear whether it is also safe and effective when self-administered by women.
Methods: This placebo-controlled, double-blind randomised trial enrolled consenting women of at least 34 weeks gestation, recruited over a 2-month period in Mbale District, Eastern Uganda. Participants had their haemoglobin measured antenatally and were given either 600mcg misoprostol or placebo to take home and use immediately after birth in the event of delivery at home. The primary clinical outcome was the incidence of fall in haemoglobin of over 20 % in home births followed-up within 5 days.
Results: 748 women were randomised to either misoprostol (374) or placebo (374). Of those enrolled, 57 % delivered at a health facility and 43 % delivered at home. 82 % of all medicine packs were retrieved at postnatal follow-up and 97 % of women delivering at home reported self-administration of the medicine. Two women in the misoprostol group took the study medication antenatally without adverse effects. There was no significant difference between the study groups in the drop of maternal haemoglobin by >20 % (misoprostol 9.4 % vs placebo 7.5 %, risk ratio 1.11, 95 % confidence interval 0.717 to 1.719). There was significantly more fever and shivering in the misoprostol group, but women found the medication highly acceptable.
Conclusions:This study has shown that antenatally distributed, self-administered misoprostol can be appropriately taken by study participants. The rarity of the primary outcome means that a very large sample size would be required to demonstrate clinical effectiveness.