Tranexamic Acid for Postpartum Hemorrhage Treatment in Low-Resource Settings: A Rapid Scoping Review

Tranexamic acid (TXA) effectively reduces bleeding in women with postpartum hemorrhage (PPH) in hospital settings. To guide policies and practices, this rapid scoping review undertaken by two reviewers aimed to examine how TXA is utilized in lower-level maternity care settings in low-resource settings. Articles were searched in EMBASE, MEDLINE, Emcare, the Maternity and Infant Care Database, the Joanna Briggs Institute Evidence-Based Practice Database, and the Cochrane Library from January 2011 to September 2021. We included non-randomized and randomized research looking at the feasibility, acceptability, and health system implications in low- and lower-middle-income countries. Relevant information was retrieved using pre-tested forms. Findings were descriptively synthesized. Out of 129 identified citations, 23 records were eligible for inclusion, including 20 TXA effectiveness studies, two economic evaluations, and one mortality modeling. Except for the latter, all the studies were conducted in lower-middle-income countries and most occurred in tertiary referral hospitals. When compared to placebo or other medications, TXA was found effective in both treating and preventing PPH during vaginal and cesarean delivery. If made available in home and clinic settings, it can reduce PPH-related mortality. TXA could be cost-effective when used with non-surgical interventions to treat refractory PPH. Capacity building of service providers appears to need time-intensive training and supportive monitoring. No studies were exploring TXA acceptability from the standpoint of providers, as well as the implications for health governance and information systems. There is a scarcity of information on how to prepare the health system and services to incorporate TXA in lower-level maternity care facilities in low-resource settings. Implementation research is critically needed to assist practitioners and decision-makers in establishing a TXA-inclusive PPH treatment package to reduce PPH-related death and disability.


Introduction
Primary postpartum hemorrhage (PPH) is usually defined as bleeding within the first 24 h equaling 500 mL or more after vaginal birth and 1000 mL or more after cesarean birth [1]. It is the major cause of maternal mortality globally and accounts for around 100,000 deaths each year [2]. Almost all the mortality burden falls on low-income and middle-income countries [3].
The causes of primary PPH include uterine atony, retained placenta, trauma, uterine rupture, clotting disorders, and uterine inversion [4]. Most deaths occur shortly after birth, and, therefore, prompt management of PPH is critical [5]. The World Health Organization (WHO) recommends a multidisciplinary approach with a first line of actions that includes

Materials and Methods
We undertook a rapid scoping review of the literature to summarize recent TXA implementation practices and health-system implications. A rapid review synthesizes knowledge by omitting or reducing elements of the systematic review process in order to collect information in a shorter time [14]. Scoping reviews aim to provide academics, decision-makers, and practitioners with an overview of a topic in order to identify key concepts, evidence types, and knowledge gaps in developing issues [15]. The rapid scoping review was completed in eight weeks by two reviewers. We used the PRISMA statement to report the outcomes (see Appendix A).

Protocol
A protocol was established and can be accessed on the Open Science Framework (https://osf.io/z47ep/).

Literature Search Strategy
We used Ovid to search the following five bibliographic databases: EMBASE, MED-LINE, Emcare, the Maternity and Infant Care Database (MIDIRS), and the Joanna Briggs Institute Evidence-Based Practice (JBI EBP) Database. In addition, the Cochrane Library was searched using tranexamic acid, postpartum hemorrhage, and developing countries as the main search terms. Developing countries is a Medical Subject Heading (MeSH) term in MEDLINE and was used as a proxy for low-resource settings. We also broadened the search to include all countries, which the World Bank classifies as low-income or lowermiddle-income (Appendix A). The search was limited to publications published between 1 January 2011 and 15 September 2021 in English, French, and Spanish. The final MEDLINE search procedure is available in Appendix A. EndNote was used to remove duplicates.

Eligibility Criteria
The first round of screening consisted of a review of the titles and abstracts. The criteria in Table 1 were used to find potential papers and evaluate them for inclusion, guiding the final selection of full-text publications. In addition to the research design, the population, concept, and context (PCC) framework was used to create the eligibility criteria. The JBI recommends the PCC framework for scoping reviews as a better option than the population-intervention-comparator-outcome mnemonic used in systematic reviews [16]. Women who had a vaginal or cesarean birth C (concept) Postpartum hemorrhage; feasibility; acceptability; health system considerations C (context) Low-income countries; lower-middle-income countries

Study Selection and Data Abstraction
For additional screening, we imported non-duplicated papers into JBI SUMARI. We reviewed the full texts of eligible publications after the titles and abstracts were evaluated for eligibility. The data from the studies that were included were retrieved using tables pilot-tested on two random publications. Study characteristics were included in one table (e.g., first author, publication year, location, aim, method, study population, intervention type and outcome measures, and relevant findings). We created another table to summarize concepts related to our study topic, such as country income categorization, levels of healthcare (basic (BEmOC) vs. comprehensive (CEmOC) emergency obstetric care), and desired outcomes (feasibility, acceptability, and effectiveness). Effectiveness was also included because it was a recurrent characteristic among the identified papers and allowed comparison with the other outcomes. Health system considerations were also included, using the WHO framework of the six-health system building blocks, i.e., (i) government and policy alignment, (ii) health staff awareness, motivation, and training, (iii) service delivery, (iv) procurement channels and commodity security, (v) financing, and (vi) health information system. Two reviewers performed separately the article screening, with a third reviewer resolving any differences.

Methodological Quality Appraisal
There was no assessment of the methodological quality of the included papers, as per standards for conducting scoping reviews [17].

Synthesis
Tables on study characteristics and key programmatic aspects are used to present the review findings in a descriptive form.

Quantity of Research Available
We identified 129 citations in the electronic database search, 55 of which were duplicates ( Figure 1). Based on titles and abstracts, 45 were eliminated from the remaining 70 because they were commentaries (2), systematic reviews (4), protocols (14), or unrelated to our topic (25). From the 25 records suitable for full-text screening, two were excluded: one was a duplicate of an already-included study but with a different title, and the other did not occur in an eligible country (Thailand). Eight records turned out to be conference abstracts that were not followed by a full-length paper. Based on the relevance of all eight records and the scoping nature of our review, they were nonetheless included in the final set of 23 references. government and policy alignment, (ii) health staff awareness, motivation, and training, (iii) service delivery, (iv) procurement channels and commodity security, (v) financing, and (vi) health information system. Two reviewers performed separately the article screening, with a third reviewer resolving any differences.

Methodological Quality Appraisal
There was no assessment of the methodological quality of the included papers, as per standards for conducting scoping reviews [17].

Synthesis
Tables on study characteristics and key programmatic aspects are used to present the review findings in a descriptive form.

Quantity of Research Available
We identified 129 citations in the electronic database search, 55 of which were duplicates ( Figure 1). Based on titles and abstracts, 45 were eliminated from the remaining 70 because they were commentaries (2), systematic reviews (4), protocols (14), or unrelated to our topic (25). From the 25 records suitable for full-text screening, two were excluded: one was a duplicate of an already-included study but with a different title, and the other did not occur in an eligible country (Thailand). Eight records turned out to be conference abstracts that were not followed by a full-length paper. Based on the relevance of all eight records and the scoping nature of our review, they were nonetheless included in the final set of 23 references.  Table 2 quantitatively summarizes the publication type, study country by income group, study type, and outcome of interest. Two-thirds of the references were full-text  Table 2 quantitatively summarizes the publication type, study country by income group, study type, and outcome of interest. Two-thirds of the references were full-text articles and one-third conference abstracts. The majority originated from lower-middleincome countries, with India, Iran, and Tunisia among the top three. There were 20 TXA effectiveness studies, two economic evaluations, and one maternal mortality modeling. Acceptability was addressed in one article, operational feasibility in three, and financial feasibility in three.
Overall, the trials tested TXA not only for PPH treatment but also for PPH prevention. While TXA was most of the time compared with routine care and placebo, a few studies used other comparators or combined TXA with other medications, including misoprostol sublingual [20,21], misoprostol per rectum [29], a prostaglandin analog [31,34], and fibrinogen [33].
Among the two economic evaluations, Li examined the cost-effectiveness of early administration of TXA within 3 h after vaginal and cesarean births, using a decision tree and healthcare provider perspectives [38]. Resch developed a cost-effectiveness model to compare three PPH intervention packages: (i) a routine bundle with intravenous fluids, uterotonics, and uterine massage, (ii) a strengthened bundle consisting of the routine bundle plus TXA as well as manual removal of placenta and suturing when indicated, and (iii) and an enhanced bundle reinforcing (ii) with non-surgical interventions for managing refractory PPH [39].
McClure modeled the impact of TXA used for PPH prevention and treatment on PPH-related maternal mortality in sub-Saharan Africa, assuming an overall 30% efficacy of TXA to reduce PPH [40].  Pre-operative use of TXA is associated with reduced blood loss during and after elective c-section. This could benefit anemic women or those who refuse blood transfusion. Mean total blood loss was 241.6 (SE 6.77) mL in the TXA group versus 510 (SE 7.72) mL in the control group. The mean drop in hematocrit and hemoglobin levels were statistically significantly lower in the TXA group than in the control group. There were no statistically or clinically significant differences in other outcomes. The study was not powered to assess the efficacy of TXA in prevention of severe PPH or to assess its safety especially thromboembolic complications.     Prophylactic use of IV TXA decreased blood losses from both placental deliveries to the end of c-section and from end of c-section to 2 h postpartum were significantly lower in the study group (p < 0.05). Total amount of oxytocin required was significantly less in TXA group (p < 0.05) also the number of women requiring other uterotonics (injectable methyl ergometrine, injectable carboprost and misoprostol per rectum) was significantly less in TXA group (p < 0.05). The amount of intra-operative fluid required were significantly less in TXA group (p < 0.005).   Primary outcome: blood loss TXA had comparable effects with prostaglandin analog on reducing PPH in women with uterine atony and in those undergoing C section or vaginal delivery: postoperative bleeding did not significantly differ between the two groups (68.2 ± 6.1 mL and 69.1 ± 175.73 mL, respectively, p = 0.6). Moreover, hemoglobin declines were 1 ± 0.4 g/dL and 1.2 ± 0.5 g/dL in TXA and prostaglandin group respectively, indicating that the difference was not statistically significant (p = 0.7) IM: intramuscular; IU: international unit; IV: intravenous; PPH: postpartum hemorrhage; RCT: randomized controlled trial; SL: sublingual; TXA: tranexamic acid.

Patient Population
Most of the clinical trials enrolled women with a singleton pregnancy who delivered vaginally or by cesarean section. Their sample sizes were relatively small to moderate (n = 40 to 300) [29,32] except in the studies by Abdel-Aleem (n = 740) [18] and Joudeh (n > 12,000) [36].
The cost-effectiveness study by Resch included 1 million women [39]. Li did not provide such information [38]. McClure's mathematical model used baseline birth rates and mortality estimates drawing from on a review of existing PPH interventions in sub-Saharan Africa [40].

Context
Most clinical and non-clinical studies took place in tertiary care and often university hospitals in the country capital or a major urban setting. Exceptions include studies by Joudeh (district hospitals) [36], McClure (homes and clinics in addition to hospitals) [41], and Resch (homes, community health centers, primary health clinics, and district hospitals) [39].
Out of the 23 studies, only McClure's included low-income countries (its mortality modeling focused on all sub-Saharan Africa-without specific mention of countries, which presumably encompassed low-income and middle-income countries) [41]. The other studies focused on lower-middle-income countries (see Table 1).

Concepts
As quantitatively described in Table 2 and further expanded in Table 4, the concept most frequently studied was the effectiveness of TXA not only to treat PPH (9 references) but as well prevent it (16 references). McClure's mathematical model also examined TXA effectiveness but in decreasing PPH-related maternal mortality. Financial feasibility [20,38,39] and operational feasibility in terms of commodity security [20], health staff capacitation [36], and service delivery [39] were addressed by a handful of studies. The acceptability of TXA by women was only examined by Diop [21]. None of the studies looked at the perspectives of providers in terms of acceptability, nor did they examine the health system components related to governance and health information systems.

Effectiveness PPH Treatment
As a medication to treat PPH, TXA was found effective for both vaginal and cesarean births in the studies by Ajroudi [32], Briki (in combination with fibrinogen) [33], and Dimassi (in combination with sulprostone) [34]. Diop found that the addition of sublingual TXA to misoprostol resulted in similar clinical and acceptability outcomes in vaginal births compared to treatment with misoprostol alone [21]. In a large study involving 22 district hospitals in Bihar, India, Joudeh assessed the effectiveness of a comprehensive emergency obstetric and neonatal readiness program, which included TXA in its PPH management package [36]. The study results showed an insufficient level of PPH diagnosis but a significantly positive trend in the number of patients diagnosed with PPH who received TXA (from 6% to 14%).

PPH Prevention
For use in preventing PPH, TXA was reported to be effective for both vaginal and cesarean births. As an example, in their study with a large sample size of women planned for elective cesarean birth, Abdel-Aleem found that the administration of TXA 1 g intravenously 10 min before the procedure resulted in significantly reduced blood loss (241 mL vs. 510 mL in the control group) [18]. There were, however, no differences in other outcomes of interest, such as additional uterotonic use, additional surgical interventions, admission to the intensive care unit, or hospitalization length. The study was not pow-ered to assess TXA efficacy in preventing severe PPH (≥1000 mL) or safety, especially thromboembolic complications.  Compared to sublingual misoprostol, TXA was shown by Bose to prevent more postcesarean bleeding in women without high risk factors for PPH, but not in women with such risk factors [20]. Tabatabaie found that TXA-oxytocin and misoprostol-oxytocin combinations were both effective in preventing post-cesarean blood loss with misoprostoloxytocin being more effective [29]. Finally, Zargar showed that TXA had comparable effects as a prostaglandin analog (Hemebate) in preventing post-cesarean bleeding [31].

Maternal Mortality
McClure's model in sub-Saharan countries estimated the following proportions of births occurring in different locations: 15% in hospitals, 35% in health clinics, and 50% at home [41]. The study showed that if TXA is available for PPH prophylaxis and treatment only in hospital settings, less than 2% of PPH-related mortality would be reduced. In contrast, if it is availed in home and health clinic settings, where the majority of births occur, a reduction in PPH mortality of nearly 30% (almost 22,000 deaths per year) would be possible. (Note that in the WOMAN trial and subsequent systematic review, Shakur concluded that intravenous TXA given immediately after bleeding onset reduced primary PPH mortality by 19% [8,9]).

Feasibility
Li showed that the use of TXA for PPH treatment could yield an average gain in quality-adjusted life-years (QALYs) of 0.18 at an additional cost of $37.12 per patient in Nigeria and 0.08 at an additional cost of $6.55 per patient in Pakistan [38]. The best-case results for incremental cost-effectiveness ratios (ICERs) were $208 per QALY in Nigeria and $83 per QALY in Pakistan, considered highly cost-effective in both countries.
In Kerala, India, Bose found TXA to be more effective than misoprostol for PPH prevention in cesarean births [20]. They reported the unit cost for each medication: INR (Indian Rupee) 52 for misoprostol-made by Cipla and INR 57 for TXA-made by Ozone. However, they did not perform a comparative economic analysis. The authors still thought that misoprostol might be operationally more feasible in resource-restricted settings in India as it does not require refrigeration, in addition to being freely available from government supplies.
In Uttar Pradesh, India, Resch showed that the use of TXA in all PPH cases would be the most cost-effective if it is integrated into a PPH care package enhanced by non-surgical interventions [39]. These interventions would help manage refractory PPH and include uterine balloon tamponade, aortic compression, and non-pneumatic anti-shock garment. The components of such an enhanced bundle would be implemented in adequation with the health facility level. Taken together, the components of this bundle could generate greater health impact and cost-savings thanks to a larger reduction in the number of required surgeries. In Uttar Pradesh, this would yield an estimated 98% reduction of PPH-related mortality (from 10.7 to 0.3 per 100,000 deliveries), averting 450 deaths yearly.
In the hospital settings in Bihar, India, the increased number of PPH patients who received TXA in the comprehensive emergency obstetric and neonatal readiness program would not have been possible without substantial investment in staff capacitation, as described by Joudeh [36]. It was the result of a high-intensity training and coaching program involving mentors (physicians and nurses), who offered training and supportive supervision to clinicians five days weekly over six consecutive months.

Acceptability
Diop found that the adjunct use of oral TXA with misoprostol to treat PPH had similar acceptability and clinical outcomes when compared to treatment with misoprostol alone [21]. The acceptability reported by women referred to the side effects they experienced, such as shivering, fever, nausea, vomiting, diarrhea, and fainting in decreasing frequency.

Discussion
Our rapid scoping review comprised 20 TXA effectiveness studies, 2 economic evaluations, and 1 mortality modeling. All the studies were undertaken in lower-middle-income countries except for the mortality study, which included unspecified low-income countries in sub-Saharan Africa. Most of the studies occurred in tertiary CEmOC referral settings. Overall, TXA was found to be effective in both treating and preventing PPH in vaginal and cesarean birth when compared to placebo or other medications. It has the potential to decrease PPH-related deaths if made available in home and clinic settings. It is potentially cost-effective, notably when integrated into a PPH intervention package enhanced with non-surgical interventions to manage refractory PPH. Capacitating service providers to increase the use of TXA to treat PPH appears to require time-intensive training and supportive supervision but would be necessary for sustainable integration into health service delivery. One study found the acceptability of TXA side effects comparable to that of misoprostol. There were no studies examining TXA acceptability from providers' perspectives and its implications for health governance and health information systems.
The outcomes of the identified PPH treatment studies align with those of the WOMAN trial, which also included low-income countries [8]. It is well established that countries belonging to the low-income group and those beset by fragility and humanitarian disasters carry the largest burden of global maternal fatalities [2]. McClure showed a potential maternal mortality decrease in rolling out TXA in sub-Saharan Africa, especially when it is made available for deliveries in lower levels of facility-based and home-based maternity care [41]. The model may overestimate the reduction in PPH mortality, as it is based on 50% of births occurring at home in sub-Saharan Africa. This figure contrasts with the 31% estimated in the same region using data from 2000 to 2019 [42]. Disaggregated data from the latter study indicate that the countries with the highest proportion of home births in sub-Saharan Africa were Chad (78%), Ethiopia (73%), Niger (70%), Madagascar (64%), Nigeria (59%), and Angola (53%) (there was no data for South Sudan, but another study done at a county level found that 72% of births occurred at home [43]). In countries with such a high proportion of home births, the availability and safe administration of TXA may play a critical role in reducing a third or more of PPH mortality according to McClure's model. However, based on current clinical recommendations, this would require the presence of staff competent in injecting TXA. Therefore, oral TXA could be a game-changer if found safe and effective in treating PPH. Notwithstanding, our review identified limited information on the safety, feasibility, acceptability, and effectiveness of TXA in these lower and often resource-challenged levels of care. Piloting and evaluating the implementation of TXA by trained health care professionals in such contexts would be needed. However, this should be done as part of an overall health system strengthening strategy focused on curbing maternal death and disability. As shown by a secondary analysis of the WOMAN trial, which examined the deaths of 483 women following PPH, other clinical and contextual factors should be tackled, including maternal anemia, delayed referral to higher levels of care, unavailability of blood transfusion, and poor facility infrastructure [44].
With regard to maternal anemia, it is prevalent in low-resource settings and has been long established to affect negatively women experiencing postpartum blood loss, even in a small amount [45]. Maternal anemia is associated with increased risk of PPH, low birthweight, small-for-gestational age babies, and perinatal death [46]. Therefore, preventing PPH is critical for women living in these settings and has been prioritized in numerous studies, including those identified in our scoping review. Owing to the mixed quality of the extant literature, WHO is not recommending TXA for PPH prevention and is awaiting the results the WOMAN-2 trial [47]. The WOMAN-2 trial is a large randomized double-blind controlled trial taking place in Pakistan and Zambia.
In terms of health system considerations useful for TXA planning and implementation in low-resource settings, the scoping review may offer some insights to decision-makers and practitioners on cost-effectiveness, health staff training and support, and integration of TXA into PPH care packages. It would be helpful to document how TXA can be effectively integrated into (i) national technical guidance and health policies, (ii) clinical records and the health information system, and (iii) lower-level maternity care settings with limited infrastructure and staff capacity.
Rapid scoping reviews, by their very nature, have flaws since they focus over a shorter period on finding knowledge gaps and repercussions for decision-making, as well as guiding future research. [17]. Therefore, biases may have been introduced in literature searches, article retrieval, and assessment. We may have missed key papers by restricting our search to English, French, and Spanish articles and speeding the data extraction procedure. We did not assess the methodological quality of the studies and cannot comment on their scientific robustness. The lists of references in the publications may not have been fully scanned due to time constraints, and we did not contact authors for further information.

Conclusions
TXA could be a cost-effective addition to the PPH treatment toolkit in vaginal and cesarean births in hospital settings of lower-middle-income countries. Information on how to prepare the health system and health services to integrate TXA in lower-level maternity care facilities in low-resource settings is scarce. Further implementation research is needed to assist decision-makers and practitioners in developing a TXA-inclusive PPH treatment package to limit bleeding-related deaths and disabilities among women giving birth in resource-challenged settings.