Efficacy of tranexamic acid for the prevention of post-partum haemorrhage among women undergoing caesarean section in Harare, Zimbabwe: A randomized controlled trial

Post-partum haemorrhage (PPH) is a significant cause of maternal mortality and accounts for 25% of all annual global maternal mortality cases [1]. Approximately 99% of all the cases of maternal mortality due to PPH occur in developing countries and 66% of them occur in the Sub-Saharan Africa region [2, 3]. PPH can occur during delivery, either vaginal delivery or caesarean section, and in both cases significant blood loss can happen. Caesarean section has been shown to increase the odds of PPH by at least 7 times compared to vaginal delivery, and it also causes uterine atony a significant risk factor for PPH [4].

To manage PPH, uterotonics such as oxytocin, misoprostol, and ergometrine given after delivery have been proven to be effective for PPH prevention resulting in a 50% reduction in the incidence of PPH [3, 5]. In addition to the commonly used agents for the prevention of PPH, tranexamic acid (TXA) is an antifibrinolytic that has been recently investigated for prophylactic efficacy against PPH [6]. The World Health Organization (WHO) currently recommends the use of TXA for the treatment of PPH where oxytocin and other uterotonics fail to stop bleeding or if it is thought that the bleeding may be partly due to trauma, although WHO cautions that this is a weak recommendation supported by moderate-quality evidence [3]. The WOMAN Trial found that TXA reduces mortality due to bleeding in women with PPH after caesarean and vaginal births with no adverse effects or an increase in the incidence of thromboembolic events [7]. The study showed a significant reduction in bleeding and laparotomy to control bleeding when TXA was administered soon after the onset of PPH [7]. A multicentre randomized placebo-controlled trial, TRAAP (TXA for preventing PPH after vaginal delivery), in France investigated the efficacy of TXA after vaginal delivery. In this study, TXA did not result in a significantly lower incidence of PPH compared to the use of placebo [8]. A clinical trial by Milani et al. also showed a reduction in risk for PPH among those who received TXA before incision, PPH measured as blood loss greater than 500 mL [9]. A systematic review and meta-analysis (SRMA) by Simonazzi et al. of studies involving 2,345 caesarean deliveries showed that TXA administration resulted in significantly lower blood loss and a lower incidence of PPH and severe PPH [10], and similar results in an SRMA were observed by Alam and Choi [11]. Wang et al. conducted a large meta-analysis of 11 randomized control trials where TXA was administered before the start of scheduled caesarean deliveries. They showed that TXA intervention significantly reduced intraoperative blood loss during and after caesarean delivery without an increase in thromboembolic events [12]. Systematic reviews by Novikova et al. and another by Cheema et al. showed that for scheduled elective caesarean deliveries, TXA use was effective in reducing the incidence of PPH and blood transfusion requirements [13, 14]. Low to medium income countries bear the biggest burden of mortality during delivery [2, 3]. In Bangladesh, a clinical trial by Milani et al. showed that TXA administration before surgical procedure reduced blood loss after delivery, measured as less than 500 mL [15]. In Egypt, several studies have been done that showed that TXA reduced blood loss even though the sample size from these studies was small to have enough power to show a difference [16, 17]. Similar results have been observed in clinical trials conducted in Nigeria on the efficacy and safety of TXA to prevent excess blood loss and PPH [18, 19, 20].

PPH is the leading cause of maternal mortality in Zimbabwe contributing to 43% of maternal deaths in 2018 [21]. In the last decade, there has been a 25%–30% increase in caesarean birth deliveries in developing countries, including Zimbabwe [6]. As caesarean section deliveries are a cause of PPH, it becomes imperative that methods to reduce PPH be continuously explored. In 2017, an updated WHO guideline recommended the use of TXA for the treatment of PPH regardless of the cause of bleeding [22]. Other clinical guidelines from competent authorities also recommend the use of TXA administration for prevention of PPH before delivery [6, 23]. A study by Pacheco et al. in a patient population scheduled for caesarean section showed that prophylactic administration of TXA did not significantly reduce the risk of blood transfusion due to excess blood loss compared to those who received the placebo [24]. Unlike other studies that assessed the use of TXA after caesarean section to prevent blood loss, this is one of the few studies that showed that perhaps TXA is not effective if used for prophylaxis.

We conducted this clinical trial to assess the efficacy of prophylactic administration of TXA (1 g) at the time of skin incision for a caesarean delivery followed by standard of care prophylactic oxytocin (intravenous 5 IU) administration after delivery, compared to placebo at skin incision followed by standard of care prophylactic oxytocin (intravenous 5 IU) in a double-blind, closed label randomized clinical trial. Our hypothesis was that prophylactic administration of TXA at the time of incision followed by oxytocin after delivery would reduce the risk of PPH compared to oxytocin alone after delivery.

We conducted the efficacy of TXA in preventing PPH after caesarean section trial, a 2-centre, randomized, placebo-controlled, double-blind clinical trial with 2 parallel groups. Women undergoing elective/emergency caesarean section were randomly assigned to either receive TXA (1 g) or placebo before skin incision for caesarean section delivery.

Women were eligible to participate if they were undergoing caesarean section (emergency or elective) at Sally Mugabe Hospital and Parirenyatwa Hospital. These are the 2 largest university teaching hospitals in Zimbabwe located in the capital city of Harare. Inclusion for the study was considered if a woman was at estimated gestational age of 37 weeks or more, had a live intrauterine foetus, was looking to have an elective or emergency caesarean section and could provide informed consent. Women were excluded from the study if they had a history of coagulopathies or conditions predisposing them to thromboembolic phenomena, seizure history, autoimmune disease, placental abruption, placenta praevia, abnormally adherent placentae if identified on prenatal ultrasound, eclampsia or HELLP syndrome, known hypersensitivity to TXA, planned general anaesthesia, caesarean delivery for the second twin or second/third triplet(s) after vaginal birth of the first twin, poor understanding of English/Shona languages, those who have received anticoagulants in the week before delivery or on investigation for Covid-19 and confirmed Covid-19 positive women.

Participants in the intervention arm received intravenous TXA (1 g), administered slowly (over 30–60 s) at the time of skin incision and standard of care 5 IU of oxytocin intravenously administered after delivery. Participants in the control group received intravenous placebo at the time of incision and standard of care 5 IU of oxytocin intravenously at birth of the baby. All other aspects of management of the caesarean section were identical in both arms.

The primary outcome of the study (PPH) was considered to have occurred if the estimated blood loss (EBL) was greater than 1,000 mL. EBL was determined using laboratory haematocrit values, an objective measure to estimate blood loss compared to using laboratory haemoglobin values which tend to be inaccurate [25, 26]. The formula for EBL using haematocrit levels:

The estimated blood volume (mL) was calculated as body weight in kilograms × 85. Preoperative haematocrit was determined 1 day before caesarean or on the day of emergency caesarean delivery. Post-operative haematocrit was measured by blood sampling on the second morning after caesarean section. If haematocrit was not available on that day, then day 3 values were used [8]. Women who received red blood cell transfusion for PPH between delivery and day 2 post-partum were defined to have PPH and met the criteria for the primary outcome. Secondary outcomes were adverse events such as changes in haemodynamic parameters, nausea, vomiting, dizziness, pulmonary embolism, deep vein thrombosis, myocardial infarction, seizure and renal failure requiring dialysis.

We calculated the sample size assuming a proportion of 2.1% PPH in the experimental group and 5.8% in the control group at 95% confidence interval (CI) and 90% power [26]. Using the Fleiss formula gave a minimum sample size of 1,162 total participants with 581 participants per group.

Women eligible for participation and who had provided informed consent were randomly assigned to receive either 1 g TXA or placebo in a 1:1 ratio. The trial statistician produced computer generated random numbers linked to the group allocation. The study nurse enrolled participants into the study and assigned the random sequence numbers to participants.

TXA was kept in the theatre to ensure an uninterrupted supply of the study drug in the custody of research assistants. Intravenous oxytocin for standard of care for participants in both arms was made available every week from the pharmacy. Cold chain storage was maintained up to the administration of the drug to the patient.

The women’s demographic characteristics and standard risk factors for PPH were compared between the 2 groups. The main analysis of the primary and secondary outcomes was performed using modified intention-to-treat analysis, that is the population of women who were randomized and had a caesarean delivery. Quantitative variables were expressed as means with standard deviations. Chi-square or Fisher’s exact tests of association were used for categorical variables. For variables that showed a significant P value in the chi-square or Fischer’s exact test (P value ≤ 0.05), we added them to the adjusted regression model to control them as potential confounders. We used generalized linear regression model to analyse the efficacy of TXA on PPH, we also carried out a test for effect modification by emergency or elective caesarean section. Statistical analysis was done using Stata version 16 [27].

A total of 1,226 participants were recruited into the clinical trial and randomly allocated to either receive prophylactic TXA (613 participants) or placebo (613 participants) before incision in either emergency or elective caesarean section. Two participants were not included in the analysis from the arm that was given TXA because their gestational age was less than 37 weeks; therefore, total sample size was 1,224 participants (Figure 1).

Most of the participants’ characteristics were independent of random group allocation (age category, gestational age, HIV status, anaemia, placenta previa, abnormal placentation, uterine fibroids, polyhydramnios, history of PPH, BMI, previous caesarean sections) with P values >0.05. Participants with a history of multiple pregnancies (P value = 0.006), comorbidities (P value = 0.04), and a preference of a type of caesarean section (P value < 0.001) showed an association with random group allocation (Table 1).

Haematocrit levels before caesarean section in the 2 groups (36.8 ± 4.2 for TXA and 36.5 ± 4.2 for placebo) were not significantly different (P value = 0.23) (Table 2). After caesarean section procedure, haematocrit levels in the 2 groups were not significantly different (P value = 0.75) (Table 2). Estimated mean blood loss between the 2 groups was not significantly different (P value = 0.88).

Table 3 shows the results of PPH from adjusted generalized linear model. The risk of PPH between the 2 groups was not statistically significant (adjusted relative risk [RR] = 1.07; 95% CI = 0.87–1.31) after controlling potential confounders (comorbidities, multiple pregnancies, type of C-section). We also carried out a test for effect modification by type of caesarean section, emergency or elective. Among participants who had an emergency caesarean section, use of TXA did not significantly reduce the risk for PPH compared to placebo (adjusted RR = 1.06; 95% CI = 0.79–1.42), and participants who had elective caesarean did not have a significant reduction in PPH (adjusted RR = 1.07; 95% CI = 0.80–1.44). Having comorbidities or multiple pregnancies did not confound the risk for PPH in the combined group as well as those who had emergency C-section and those who had elective C-section.

Table 4 shows adverse events that were recorded after TXA or placebo administration. There was no statistically significant risk for nausea (RR = 0.6; 95% CI = 0.25–1.3), vomiting (RR = 1.6; 95% CI = 0.9–1.9) or dizziness (RR = 0.86; 95% CI = 0.4–1.8) occurrence between placebo and TXA groups. Few adverse events were observed in the participants and they were not significantly related to use of either TXA or placebo.

In this clinical trial of assessing the efficacy of prophylactic administration of TXA for the prevention of PPH for women undergoing caesarean section, the risk of PPH in participants who received TXA was not significantly lower compared to participants who received placebo. An analysis of potential effect modification by the type of caesarean section (emergency or elective) showed no significant risk reduction with prophylactic use of TXA before incision. A clinical trial by Pacheco et al. assessing efficacy of TXA to prevent PPH showed that TXA after caesarean section did not have a significant effect [24]. While it is similar in overall interpretation with our study, the major difference is that we administered TXA before procedure while Pacheco et al. administered after delivery and Pacheco assessed maternal death and need for blood transfusion as outcomes. Clinical trials that have been carried out in Africa and other countries in the low to medium income region have shown a reduction in blood loss with the use of TXA. A study by Soliman et al. showed that TXA reduced the risk of blood loss after caesarean section [17]. Like our study they administered TXA before incision, however, they measured blood loss at 500 mL and had a small sample size of 100 participants. Kamel et al. had a smaller sample size than our study, but they also showed that TXA administration before incision resulted in lower blood loss [16]. Several studies carried out in Nigeria showed that prophylactic administration of TXA also reduced blood loss [18, 19, 20]. Unlike our study, these clinical trials from Nigeria had a relatively small sample size and measured blood loss at 500 mL as their primary outcome while we used the standard blood loss greater than 1,000 mL as the definition for PPH.

Our results differ from what has been observed in other clinical trials assessing the efficacy of TXA for PPH after caesarean or vaginal delivery. Gohel et al. showed that TXA significantly reduced the amount of blood loss after caesarean section compared to placebo; however, their study used blood loss estimation that was prone to human error [28]. A meta-analysis of clinical trials by Wang et al. also showed that TXA was effective in the prevention and treatment of PPH, a finding different from our study [12]. However, a review of evidence by Sentilhes et al. had cast doubt on the effectiveness of the drug since most of the studies were underpowered from small sample sizes or used estimation methods that were prone to error [1]. A large international clinical trial (WOMAN Trial Collaborators) also showed that early treatment with TXA reduces death and adverse outcomes in women with confirmed PPH after vaginal or caesarean section [7]. The WOMAN Trial was a landmark trial as it also confirmed the safety of TXA as shown in our study; however, they used TXA as treatment for PPH and not a prophylactic approach. An SRMA by Cheema et al. showed that there was a potential for TXA to prevent PPH in high-risk populations if given before incision [14]. Simonazzi and Novikova also carried out SRMA that showed reduced blood loss after TXA, though there were differences in studies compared; with some administering TXA before incision and some after delivery [10, 13]. While this analysis came from studies in high-risk women, the authors inferred that stronger evidence still needed to be collected to show real efficacy of TXA for PPH. Similar to other clinical trials, we observed a good safety profile of TXA. No serious adverse events or life-threatening events occurred during and after administration of TXA. The larger multicentre international clinical trial on efficacy of TXA (WOMAN Trial) showed that TXA was safe to use as a fibrinolytic agent for prevention and treatment of blood loss and PPH after caesarean delivery [7].

Our study considered participants who were not at high risk for PPH, and this may have affected the overall efficacy of TXA compared to the placebo.

While other studies have shown the efficacy of preoperative administration of TXA, our study did not show a significant relation and thus may not be generalizable at this point in clinical practice. However, we also observed that TXA remains safe for use in PPH among pregnant women.

TXA did not significantly reduce the risk for PPH in women undergoing caesarean section when administered before procedure but has a good safety profile. A larger study may still be able to show that TXA is effective as a prophylactic agent administered before incision.

Data are available upon request from the University of Zimbabwe College of Health Sciences in line with their IT and Research Policies. Data can be requested through Dr Chipo Gwanzura ([email protected]).

No funding was provided for the preparation and submission of this manuscript. The study was funded by the Fogarty International Center of the National Institute of Health award D43TW009343 (U.S. NIH Grant/Contract).

The authors declare that they have no competing interests.

CG conceptualized the study. CG, MGM, PW, DW and ZMC implemented the study. VC, SG, CG and MGM carried out statistical data analysis. CG, SG and MGM wrote the first version of the manuscript. PW, DW and ZMC provided methodological guidance and all authors contributed to all final manuscript preparation. All authors read and approved the final manuscript.

Participants in the study provided written informed consent for publication of the results. Deidentified patient information was used for statistical analysis.

Participants were provided informed consent prior to enrolment into the main study, and all study procedures were done according to the ethical principles of human participants involved in medical research of the Declaration of Helsinki. Only those participants who understood and signed the written informed consent were then enrolled for study procedures. Approvals for the research were provided by the Medicines Control Authority of Zimbabwe (reference CT160/2017), the Medicines Research Council of Zimbabwe (reference MRCZ/A/2691) and the Harare Central Hospital Ethics Committee (reference HCHEC191020/49).

Full protocol can be assessed at clinicaltrials.gov.

The clinical trial is registered in the U.S. National Library of Medicine clinicaltrials.gov registration number NCT04733157 in February 2021.

How to cite this article: Gwanzura C, Madziyire MG, Chikwasha V, Gavi S, Wright P, Walker D, Chirenje ZM. Efficacy of tranexamic acid for the prevention of post-partum haemorrhage among women undergoing caesarean section in Harare, Zimbabwe: A randomized controlled trial. Adv Glob Health. 2024;3(1). https://doi.org/10.1525/agh.2024.2307229

Editor-in-Chief: Craig R. Cohen, University of California San Francisco, CA, USA

Senior Editor: Andres G. Lescano, Cayetano University, Peru

Section: Improving Health and Well-Being