BACKGROUND:Postpartum haemorrhage is a leading cause of maternal mortality, particularly in low-income and middle-income countries. Several pharmacological agents, such as oxytocin, ergot alkaloids, prostaglandins, and tranexamic acid, have been used prophylactically to prevent postpartum haemorrhage. However, the optimal prophylactic regimen and the comparative efficacy of these agents and their combinations have not been fully elucidated for individuals undergoing caesarean delivery. We aimed to conduct a network meta-analysis to assess different agents for postpartum haemorrhage prophylaxis in caesarean deliveries.
METHODS:In this systematic review and meta-analysis, we conducted a Bayesian network meta-analysis of randomised controlled trials (RCTs) evaluating the relative effectiveness of different prophylactic agents and their combinations for postpartum haemorrhage in caesarean deliveries. We searched MEDLINE, the Cochrane Central Register of Controlled Trials, Embase, and Web of Science from database inception to Nov 7, 2023, for RCTs that enrolled adult pregnant women (ie, older than 18 years) undergoing a caesarean delivery; compared prophylactic strategies (monotherapy or combination drug therapy) with placebo or another active prophylactic regimen; administered prophylactic strategies of any parenteral dosage or regimen systemically before surgical incision or immediately after birth for preventing postpartum haemorrhage; and reported our prespecified endpoints of interest. Quasi-randomised trials, trials evaluating prophylactic strategies exclusively comparing different dosages, routes, or regimens of the same prophylactic agent, trials that included vaginal delivery, single-arm studies, conference abstracts, studies not published in English, and studies with overlapping populations were excluded. Ten authors reviewed study reports and supplementary materials and extracted the data. Two authors performed these tasks independently for each study. For data reported in graphical format, extraction was performed with graph digitising web software. The primary outcome was postpartum haemorrhage (ie, blood loss of ≥1000 mL following caesarean delivery). We fitted a Bayesian random-effects network meta-analysis model to compare multiple regimens simultaneously, with results presented as risk ratios (RRs) and their respective 95% credible intervals (CrIs). Only strategies reported by two or more studies were included in the network. If a prophylactic strategy was reported by only one study, it was included if at least 1000 patients were allocated in each study group. We also synthesised head-to-head RCTs separately to assess differences between regimens with league tables. To assess the hierarchy of treatments based on efficacy, we estimated surface under the cumulative ranking curve (SUCRA) probabilities. This review is registered at PROSPERO, CRD42023488236.
FINDINGS:The search strategy yielded 3339 studies. After removing duplicates, 2241 studies remained, of which a total of 2022 were excluded on the basis of title or abstract screening. After full-text review, 167 RCTs (with 44 817 patients) evaluating monotherapy with or various combinations of oxytocin, carbetocin, carboprost, ergot alkaloids, misoprostol, and tranexamic acid were included in the final analysis. Across all 167 studies, 12 868 patients received oxytocin monotherapy, 5849 patients received tranexamic acid monotherapy, 2964 patients received carbetocin monotherapy, 1773 patients received misoprostol monotherapy, and 100 patients received carboprost monotherapy. The most common combination therapy was tranexamic acid plus oxytocin (n=5331) followed by misoprostol plus oxytocin (n=2983). Oxytocin plus tranexamic acid (RR 0·44 [95% CrI 0·33-0·58]) and carbetocin (0·54 [0·37-0·74]) were the only interventions that were more effective than oxytocin alone in reducing postpartum haemorrhage. Oxytocin plus tranexamic acid ranked as the most effective intervention for postpartum haemorrhage prophylaxis with a SUCRA probability value of 0·85. Most prophylactic combinations reduced intraoperative blood transfusions and the need for additional uterotonics. Two maternal deaths were reported among 29 412 patients. No significant heterogeneity was detected for postpartum haemorrhage (I2=6%), blood transfusion (I2=0%), and additional uterotonics (I2=7%).
INTERPRETATION:Carbetocin alone and oxytocin plus tranexamic acid were superior to oxytocin monotherapy for preventing postpartum haemorrhage in caesarean deliveries. Oxytocin plus tranexamic acid ranked as the most effective intervention for postpartum haemorrhage prevention. These results are crucial in highlighting the comparative efficacy and hierarchy of prophylactic agents for postpartum haemorrhage prevention, especially given the widespread availability and low cost associated with oxytocin and tranexamic acid.