Postpartum hemorrhage drills or simulations and adverse outcomes: a systematic review and Bayesian meta-analysis
Hector Mendez-Figueroa1, Cynthia S Bell2, Steve M Wagner3
1Department of Obstetrics, Gynecology and Reproductive Sciences, McGovern Medical School at The University of Texas Health Science Center at Houston (UTHealth), Houston, TX, USA.
Objective:
To compare the rates of adverse outcomes with postpartum hemorrhage (PPH) before and after implementation of drills or simulation exercises.
Study Eligibility Criteria:
We included all English studies that reported on rates of PPH and associated complications during the pre- and post-implementation of interventional exercises.
Study Apprasial And Synthesis Methods:
Two investigators independently reviewed the abstracts, and full articles for eligibility of all studies. Inconsistencies related to study evaluation or data extraction were resolved by a third author. The co-primary outcomes were the rate of PPH and of any transfusion; the secondary outcomes included admission to the intensive care unit (ICU), transfusion ≥ 4 units of packed red blood cells, hysterectomy, or maternal death. Study effects were combined by Bayesian meta-analysis and reported as risk ratios (RR) and 95% credible intervals (Cr).
Results:
We reviewed 142 full length articles. Of these, 18 publications, with 355,060 deliveries-150,562 (42%) deliveries during the pre-intervention and 204,498 (57.6%) deliveries in the post-interventional period-were included in the meta-analysis. Using the Newcastle-Ottawa Scale, only three studies were considered good quality, and none of them were done in the US. The rate of PPH prior to intervention was 5.06% and 5.46% afterwards (RR 1.09, 95% CI 0.87-1.36; probability of reduction in the diagnosis being 21%). The likelihood of transfusion decreased from 1.68% in the pre-intervention to 1.27% in the post-intervention period (RR 0.80, 95% Cr 0.57-1.09). The overall probability of reduction in transfusion was 93%, albeit it varied among studies done in non-US countries (96%) versus in the US (23%). Transfusion of 4 units or more of blood occurred in 0.44% of deliveries before intervention and 0.37% afterwards (RR of 0.85, 95% CI 0.50-1.52), with the overall probability of reduction being 72% (76% probability of reduction in studies from non-US countries and 49% reduction with reports from the US). Surgical interventions to manage PPH, which was not reported in any US studies, occurred in 0.14% before intervention and 0.28% afterwards (RR 1.29; 95% CI 0.56-3.06; probability of reduction 27%). Admission to the ICU occurred in 0.10% before intervention and 0.08% subsequently (RR 0.92, 95% CI 0.58-1.43), with the overall probability of reduction being 65% (81% in studies from non-US countries and 27% from the study done in the US). Maternal death occurred in 0.17% in the pre-intervention period and 0.09% during the post-intervention (RR 0.62, 95% CI 0.33-1.05; probability of reduction 93% in studies from non-US countries and 82% in one study from the US).
Conclusions:
Interventions to reduce the sequelae of PPH are associated with decrease in adverse outcomes. The conclusion, however, ought not to be accepted reflexively for the US population. All of the studies on the topic done in the US are of poor quality and the associated probability of reduction in sequelae are consistently lower than those done in other countries.
Synopsis:
Since the putative benefits of PPH drills or simulation exercises are based on poor quality pre- and post-intervention trials, policies recommending them ought to be revisited.
