Uterine balloon tamponade for the treatment of postpartum hemorrhage: a systematic review and meta-analysis

Am J Obstet Gynecol. 2020 Apr;222(4):293.e1-293.e52. doi: 10.1016/j.ajog.2019.11.1287. Epub 2020 Jan 6.


Objective: To assess the efficacy, effectiveness, and safety of uterine balloon tamponade for treating postpartum hemorrhage.

Study design: We searched electronic databases (from their inception to August 2019) and bibliographies. We included randomized controlled trials, nonrandomized studies, and case series that reported on the efficacy, effectiveness, and/or safety of uterine balloon tamponade in women with postpartum hemorrhage. The primary outcome was the success rate of uterine balloon tamponade for treating postpartum hemorrhage (number of uterine balloon tamponade success cases/total number of women treated with uterine balloon tamponade). For meta-analyses, we calculated pooled success rate for all studies, and relative risk with 95% confidence intervals for studies that included a comparative arm.

Results: Ninety-one studies, including 4729 women, met inclusion criteria (6 randomized trials, 1 cluster randomized trial, 15 nonrandomized studies, and 69 case series). The overall pooled uterine balloon tamponade success rate was 85.9% (95% confidence interval, 83.9-87.9%). The highest success rates corresponded to uterine atony (87.1%) and placenta previa (86.8%), and the lowest to placenta accreta spectrum (66.7%) and retained products of conception (76.8%). The uterine balloon tamponade success rate was lower in cesarean deliveries (81.7%) than in vaginal deliveries (87.0%). A meta-analysis of 2 randomized trials that compared uterine balloon tamponade vs no uterine balloon tamponade in postpartum hemorrhage due to uterine atony after vaginal delivery showed no significant differences between the study groups in the risk of surgical interventions or maternal death (relative risk, 0.59; 95% confidence interval, 0.02-16.69). A meta-analysis of 2 nonrandomized before-and-after studies showed that introduction of uterine balloon tamponade in protocols for managing severe postpartum hemorrhage significantly decreased the use of arterial embolization (relative risk, 0.29; 95% confidence interval, 0.14-0.63). A nonrandomized cluster study reported that use of invasive procedures was significantly lower in the perinatal network that routinely used uterine balloon tamponade than that which did not use uterine balloon tamponade (3.0/1000 vs 5.1/1000; P < .01). A cluster randomized trial reported that the frequency of postpartum hemorrhage-related invasive procedures and/or maternal death was significantly higher after uterine balloon tamponade introduction than before uterine balloon tamponade introduction (11.6/10,000 vs 6.7/10,000; P = .04). Overall, the frequency of complications attributed to uterine balloon tamponade use was low (≤6.5%).

Conclusion: Uterine balloon tamponade has a high success rate for treating severe postpartum hemorrhage and appears to be safe. The evidence on uterine balloon tamponade efficacy and effectiveness from randomized and nonrandomized studies is conflicting, with experimental studies suggesting no beneficial effect, in contrast with observational studies. Further research is needed to determine the most effective programmatic and healthcare delivery strategies on uterine balloon tamponade introduction and use.

Keywords: Bakri balloon; cesarean delivery; condom UBT; hysterectomy; maternal mortality; placenta previa; uterine atony; uterine bleeding; uterotonics; vaginal delivery.

Publication types

  • Meta-Analysis
  • Systematic Review

MeSH terms

  • Cesarean Section / statistics & numerical data
  • Female
  • Humans
  • Maternal Mortality
  • Parturition
  • Placenta Accreta / etiology
  • Placenta Previa / etiology
  • Placenta, Retained / etiology
  • Postpartum Hemorrhage / etiology*
  • Postpartum Hemorrhage / therapy*
  • Pregnancy
  • Uterine Artery Embolization / statistics & numerical data
  • Uterine Balloon Tamponade* / adverse effects
  • Uterine Inertia / etiology