Integrating Heat-Stable Carbetocin and Tranexamic Acid for Prevention and Management of Postpartum Hemorrhage in Sub-Saharan Africa: A Five-Country Pilot Implementation Study

Background of the Study

Postpartum hemorrhage (PPH), defined as blood loss of 500 ml or more after birth, is the leading direct cause of maternal mortality globally, responsible for about a quarter of all maternal deaths, and disproportionately impacts women in Sub-Saharan Africa (SSA) and South Asia (∼80%).^[1,2] The most important component to prevent PPH is the use of uterotonics like oxytocin and misoprostol during active management of the third stage of labor (AMTSL). Oxytocin has been used for decades for the prevention and treatment of PPH, and it remains the drug of choice for both indications according to the World Health Organization (WHO) recommendations.^[3]

However, substantive progress to reduce maternal mortality is lagging behind meeting the Sustainable Development Goal 3.1 target by 2030. In SSA countries, the burden of PPH is compounded by the use of substandard or falsified medicines. A systematic review assessing studies on the quality of oxytocin in low- and middle-income countries (LMICs) found a high prevalence of substandard oxytocin and misoprostol, and many LMICs face issues with maintaining a consistent cold chain and transport conditions for oxytocin use in maternity care, which impacts its efficacy.^[4]

Thus, in the absence of oxytocin or when its quality cannot be assured, heat-stable carbetocin (HSC) has been recommended for PPH prevention in the 2018 WHO recommendations.^[5] HSC is a long-acting uterotonic that has been shown to be non-inferior to oxytocin and may be more effective for some outcomes in the prevention of PPH without an increase in side effects.^[6,7] HSC is stable for a minimum of three years at 30°C without refrigeration and therefore holds promise for use in countries with cold chain challenges.^[8] Currently, HSC is recommended for PPH prevention only, and to date, there is limited published evidence of HSC use in resource-challenged settings. A recent rapid scoping review on implementing HSC in low-resource settings showed limited evidence of healthcare provider (HCP) acceptability in lower-level maternity care, such as in Basic Emergency Obstetric and Newborn Care (BEmONC) facilities in rural areas as well as the health system’s capacity for an HSC-inclusive PPH intervention package.^[9]

Meanwhile, tranexamic acid (TXA) is an antifibrinolytic agent that has been available since the early 1960s and is indicated in several hemorrhage-related conditions for controlling bleeding.^[10] In 2017, following the publication of the WOMAN trial results, WHO recommended intravenous TXA administration for PPH treatment as an adjunct therapy to oxytocin and other standard treatment options.^[11,12] Subsequently, both HSC and TXA were included in the 2019 WHO Essential Medicines List (EML).^[13]

Despite the 2017 and 2018 WHO recommendations, both drugs are still underused or not used at all for PPH prevention and management in many countries with a high burden. HSC is currently being registered in LMICs with overall limited availability. TXA is available in many countries but is not registered for the PPH treatment indication. HSC and TXA are not yet widely used in many LMICs and there is variability among countries as to whether these drugs are recommended or included in the national EML.

The transition from recommendations to practice is contingent upon multiple steps along the product introduction pathway, including the update of national policies and consistent availability of quality-assured, affordable medicines. Between 2019 and 2022, Concept Foundation implemented a project to accelerate national availability and end-user access to HSC and TXA in 15 SSA countries by supporting Ministries of Health in updating their national guidelines, EMLs, and essential packages of health services to include these medicines, as well as developing clinical protocols on PPH prevention and management with supportive job aids. Dissemination of global recommendations is insufficient on its own to change country practice, hence, implementation research is important in helping to bridge the gap between evidence and practice in real-world settings to support the integration and scale-up of new interventions into national health systems. Efforts to reduce PPH-related morbidity and mortality require these real-world evidence-based recommendations to transition into national adoption of clinical policies and practice.

The five countries included in this study were part of the 2019–2022 Concept Foundation project and all have relatively high maternal mortality ratios (per 100,000 live births): Burkina Faso (198), Ethiopia (267), Ghana (263), Sierra Leone (443), and Uganda (284).^[14,15] The countries’ Ministries of Health, in an attempt to implement the 2017 and 2018 WHO PPH recommendations, updated their latest national guidelines to include HSC and TXA. The latest EMLs in Burkina Faso, Ghana, Sierra Leone, and Uganda include these two medicines (Ethiopia’s EML currently includes TXA and will be updated to include HSC). Despite these advances, neither medicine is widely available in any of the countries. To generate evidence on implementing new and lesser-used WHO-recommended medicines in high PPH burden settings, this implementation research assessed the feasibility, acceptability, and appropriate use of HSC for PPH prevention and TXA for PPH treatment in five SSA countries.

Study Setting and Design

We conducted a pre-post study in 19 facilities from five countries (Burkina Faso, Ethiopia, Ghana, Sierra Leone, and Uganda). During the study period (May 1 to December 31, 2022), we collected baseline data retrospectively from patient registers in the labor ward for two months at each facility. Data collected included uterotonics use patterns, patient outcomes, HCP knowledge, attitudes, and practices (KAP) pertaining to PPH prevention and management, and facility characteristics related to the availability of electricity and obstetric medicines stock-out. All identifiable patient data was anonymized. We selected a convenience sample of two BEmONC and two CEmONC facilities from each country, one each from an urban and a rural area. Uganda was the only country with one BEmONC facility due to its large number of deliveries, hence a total of nine BEmONC and ten CEmONC facilities were included in the study. The facilities were selected based on willingness and ability to integrate HSC for PPH prevention and TXA for PPH treatment into routine care. Carbetocin Ferring (100 µg/ml, Ferring Pharmaceuticals) and Tranexamic Acid (100 mg/ml, Medochemie Ltd) were supplied by IDA Foundation, Netherlands, to each of the study countries.

In each study country, HCPs in the selected facilities were trained on the safe and appropriate use of HSC and TXA using a mix of theoretical and practical simulation training. Following the training, both medicines were introduced into each facility, which marked the beginning of data collection for the two-month implementation period (Ghana was the exception, whereby data collection was extended for two weeks due to a slower-than-normal delivery rate). There was regular supportive supervision throughout the implementation phase. All patient data were extracted from data that were routinely collected in clinical practice. The questionnaires did not contain any identifiable patient information.

For the KAP assessment, we purposively recruited a sample of HCPs from each health facility. Maximum variation sampling was used to achieve a stratified sample without random selection and to ensure heterogeneity of research participants. In each of the selected facilities, HCPs were sampled based on their cadres, such as doctor/midwife or nurse/nurse aide.

Statistical Analysis

The clinical data were entered into REDCap software and analyzed using descriptive statistics in the form of mean (± standard deviation) for continuous variables, and proportion (for categorical variables) expressed as a percentage, as described in the literature.^[16] The KAP questionnaire had multiple-choice questions which were also analyzed descriptively. The open-ended questions were analyzed qualitatively through thematic analysis to identify emerging themes from the responses using Dedoose software version 9.0. This method is comprehensively used for analyzing qualitative data.^[17]

Ethical Approval and Informed Consent

Institutional Review Board approval was obtained in all countries before the start of the study. We employed broad participation criteria to be as inclusive as possible of all cadres of HCPs. All potential participants received information about the study in the national language, conforming to ethical requirements for research involving human subjects. Those who consented to participate in the study were requested to sign the informed consent form and it was made clear that they were free to withdraw from the study at any stage without risk of any negative consequences. The clinical data were retrieved from patient records and no contact was made with women delivering during the study period.

Summary of Total Number of Deliveries Captured by Facility Type

During the study period, 25,116 deliveries were recorded across the five countries, out of which 12,153 were captured during the baseline period and 12,963 during the implementation period. There was a total of 2,309 deliveries at BEmONCs and 22,807 deliveries at CEmONCs.

Facility Characteristics

Overall, 55.6% of the BEmONCs reported electricity being available sometimes and 44.4% reported electricity being available all the time. Half of the BEmONCs reported refrigeration being sometimes available for oxytocin and the other half reported availability all the time, and 88.9% reported no stock-out of oxytocin over a six-month period.

As for the CEmONC facilities, 15.0% reported electricity being available sometimes and 85.0% reported electricity being available all the time. A quarter of the CEmONCs reported refrigeration being sometimes available for oxytocin and 75.0% reported availability all the time, and 90.0% reported no stock-out of oxytocin over a six-month period.

Obstetric Characteristics

Out of the 25,116 recorded deliveries during the study period, the median age of women who delivered was 27 years. Throughout the study, the percentage of nulliparous deliveries in all study BEmONCs ranged from 14.5% (n = 11) in Ethiopia to 37.4% (n = 92) in Ghana, while nulliparous deliveries at CEmONCs ranged from 28.2% (n = 574) in Ghana to 55.1% (n = 2287) in Uganda. There was 100% vaginal delivery in BEmONCs of all countries at baseline and during implementation. In all CEmONCs, the proportion of cesarean deliveries ranged from 23.3% (n = 293) in Burkina Faso to 37.7% (n = 678) in Ethiopia at baseline, and from 13.6% (n = 237) in Burkina Faso to 43.2% (n = 2104) in Uganda during implementation. During baseline and implementation, uterotonic administration for labor induction/augmentation at BEmONCs was <4% in Ethiopia, Ghana, Sierra Leone, and Uganda. In CEmONCs, uterotonic administration for labor induction/augmentation ranged from 4.7% to 43.6% at baseline and from 2.9% to 27.1% during implementation in Ethiopia, Ghana, Sierra Leone, and Uganda. Burkina Faso results are not included due to unavailable data. A summary of the main demographic characteristics captured is presented in Table 1.

Main Outcome Results

Quantitative Analysis

Appropriate use of HSC

Uterotonic administration for PPH prevention by country, facility type, and study phase

The percentage of women receiving prophylactic uterotonics at baseline was in general very high in all countries and in both facility types (most were more than 90%). Therefore, there was no room for improvement after implementation, except in Uganda, which improved from almost 80% to more than 90%. Unexpectedly, in Burkina Faso, there was a 14% reduction in prophylactic uterotonic administration in CEmONC facilities. During implementation, HSC was the most used uterotonic for PPH prevention in both BEmONCs and CEmONCs. In BEmONCs, HSC was administered to >99% of women in all study countries. In CEmONCs, HSC was administered to >97% of women in Sierra Leone, Burkina Faso, and Ghana and >87% in Ethiopia and Uganda. These findings are presented in Figures 1 and 2. No adverse events were reported from the use of HSC during implementation.

Close

Figure 1*:

Uterotonic administration for PPH prevention by country, facility type, and study phase.

Data are presented as percentage of prophylactic uterotonic administered at BEmONC and CEmONC facilities in each study country at baseline (pre-intervention) and during implementation (post-intervention). At baseline, the uterotonics administered were either oxytocin, misoprostol, or oxytocin in combination with misoprostol. During implementation, the uterotonics administered were either HSC, oxytocin, misoprostol, oxytocin in combination with misoprostol, HSC in combination with oxytocin or misoprostol, or carboprost. B = Baseline and I = Implementation

*HSC was administered in combination with other uterotonics for PPH prevention to 3.1% (55) of women in CEmONCs in Ghana, 0.5% (7) in Burkina Faso, 0.5% (22) in Uganda, and 0.1% (1) in Sierra Leone. Across all BEmONCs, one patient in Ghana received HSC in combination with other uterotonics. In Ghana, 53 cases in CEmONCs and 1 case in BEmONCs were administered HSC in combination with misoprostol for PPH prevention.

Export to PPT

Close

Figure 2:

HSC administration for PPH prevention during study implementation by country and facility type.

Data are presented as percentage of HSC administered at BEmONC and CEmONC facilities in each study country during implementation.

Export to PPT

Additional uterotonic administration for PPH following prophylaxis

Additional uterotonics are administered for continued or heavy bleeding after prophylactic uterotonics have been given. Overall, additional uterotonic use was low in all study countries. In BEmONCs, additional uterotonics were used in ≤5% of women in Ethiopia, Sierra Leone, and Uganda at baseline. During implementation, the use of additional uterotonics decreased in Uganda. No additional uterotonic use was recorded in Ethiopia and Sierra Leone during implementation, and also at baseline and implementation phases in Ghana. In Burkina Faso, additional uterotonic data was not captured in patient records.

In CEmONCs, additional uterotonics were used in ≤10% of women in Ethiopia, Uganda, Sierra Leone, and Ghana at baseline. There was a general decrease in the use of additional uterotonics in these countries during implementation, except for Ghana which remained the same. Data for additional uterotonic use were not available in Burkina Faso. Oxytocin alone or oxytocin in combination with misoprostol were the most common additional uterotonics used in all countries. HSC was not administered as an additional uterotonic during implementation in all study facilities. The findings are summarized in Figure 3.

Close

Figure 3:

Additional uterotonic administration by country, facility type, and study phase.

Percentage of additional uterotonic administered at the study facilities at baseline (pre-intervention) and implementation (post-intervention). B = Baseline and I = Implementation

Export to PPT

Appropriate use of TXA

TXA administration for PPH treatment by country, facility type, and study phase

At baseline, TXA was only readily accessible in Ghana, Ethiopia, and Uganda. Burkina Faso and Sierra Leone did not have TXA readily accessible.

Once TXA was made available during implementation, use in BEmONCs and CEmONCs without prior availability was at ≤2% in Burkina Faso, and ≤8% in Sierra Leone. In Uganda and Ethiopia CEmONCs, there was minimal to no difference in TXA administration. In Ghana, TXA use decreased in both facility types during implementation. There were no cases of TXA administration in Ethiopia BEmONCs at baseline or during implementation, and only three cases of TXA administration in Uganda BEmONCs during implementation. The results are demonstrated in Table 2. No adverse events were reported from the use of TXA during implementation.

Table 2: TXA administration for PPH treatment and time from delivery to administration by country, facility type, and phase

		Burkina Faso				Ethiopia				Ghana				Sierra Leone				Uganda
		BEmONC		CEmONC		BEmONC		CEmONC		BEmONC		CEmONC		BEmONC		CEmONC		BEmONC		CEmONC
		B	I	B	I	B	I	B	I	B	I	B	I	B	I	B	I	B	I	B	I
TXA administered for PPH treatment Time from delivery to TXA administration for PPH treatment (mins)	n	0	8	0	18	0	0	2	23	10	4	88	22	0	8	0	120	0	3	219	243
	N	252	366	1256	1732	92	76	1802	1863	229	246	2049	1799	130	129	1618	1480	399	388	4302	4867
	%	0.0	2.2	0.0	1.0	0.0	0.0	0.1	1.2	4.4	1.6	4.3	1.2	0.0	6.2	0.0	8.1	0.0	0.8	5.1	5.0
Time from delivery to TXA administration for PPH treatment (mins)	n	0	8	0	18	0	0	2	23	10	4	88	22	0	8	0	109	0	3	39	192
	Median	-	28.0	-	31.5	-	-	121.0	26.0	10.5	20.5	11.0	9.5	-	57.5	-	24.0	-	12.0	55.0	7.0
	IQR	-	30.3	-	24.8	-	-	182.0	29.0	8.0	38.8	10.8	8.8	-	51.8	-	27.5	-	5.0	229.0	19.8

Proportion of TXA administration for PPH treatment. The n (%) represents the number of women who received TXA and the proportion as a percentage, with the N denominator being the total number of deliveries captured within the study period. Time from delivery to TXA administration in minutes. The n represents the number of cases with recorded TXA administration. Data are presented as n (%) or median (IQR). B = Baseline and I = Implementation

Time from delivery to TXA administration for PPH treatment in BEmONC and CEmONC facilities

Once TXA was made available during implementation, the median time from delivery to TXA administration for PPH treatment in Burkina Faso, Sierra Leone, and Uganda BEmONCs and CEmONCs was less than one hour. In Ethiopia and Uganda CEmONCs, there was a clear decrease from baseline to implementation in the time from delivery to TXA administration (and a minimal difference for Ghana CEmONCs). In Ghana BEmONCs, there was a ten-minute increase in the median during implementation. Overall, in the sites with TXA available at baseline, the median time of administration was within 30 minutes during implementation. The results are demonstrated in Table 2.

Blood transfusion and other PPH treatments received

In CEmONCs, 58 women in Burkina Faso (4.8%) received blood transfusions at baseline and 75 (4.3%) during the implementation period (the reduced percentage during implementation is due to the 40% increase in the number of delivery records captured). Seventeen women (1.0%) were transfused at baseline and 12 (0.6%) during implementation in Ethiopia; 24 (1.5%) were transfused at baseline and 23 (1.5%) during implementation in Sierra Leone; and 76 (1.8%) were transfused at baseline and 39 (0.8%) during implementation in Uganda. Ghana results are not included due to unavailable data. There were no blood transfusions recorded in BEmONCs of all study countries at baseline or during implementation.

In CEmONCs, the intrauterine tamponade was used only at baseline and in three countries—two patients in Burkina Faso, three in Ethiopia, and three in Uganda. A non-pneumatic anti-shock garment (NASG) was used in one patient at baseline in Uganda and one patient during the implementation phase in Ethiopia. No intrauterine tamponade or NASG was used in any BEmONC facility during the study period.

Safe use of HSC and TXA

Monitoring of heat-stable carbetocin and tranexamic acid ampoules

Facilities monitored ampoule counts for potential misuse of HSC and TXA and recorded the number of ampoules that were supplied to them, the number administered to patients during implementation, and the number remaining at the end of study implementation. They also recorded the number of ampoules of HSC and TXA that were broken or wasted, missing, or used for reasons other than PPH prevention or treatment.

For HSC, no countries reported use for reasons other than PPH prevention. For TXA, Ethiopia, Sierra Leone, and Uganda reported the use of the medication for reasons other than PPH treatment—nine patients in Ethiopia (0.7%) received TXA empirically to prevent PPH and eight patients in Sierra Leone (2.0%)received TXA empirically to prevent PPH (five), or for gynecologic surgery (two), and for a patient undergoing surgery for a motor vehicle accident (one). Uganda reported three cases of TXA administered to prevent PPH in high-risk patients.

Healthcare provider KAP assessment

Healthcare provider methods of measuring blood loss

We asked a sample of 140 HCPs to report on how they measure blood loss, and the cumulative result from the five countries showed that 75.0% of HCPs visually estimate blood loss, which remained unchanged post-implementation. The results are presented in Table 3.

Table 3: Healthcare providers’ self-reported methods of measuring blood loss

Methods of measuring blood loss	Burkina Faso		Ethiopia		Ghana		Sierra Leone		Uganda		Baseline Total	Implementation Total	Total
	n		n		n		n		n		n (%)	n (%)	n (%)
	B	I	B	I	B	I	B	I	B	I			B + I
I don’t measure blood loss	0	0	0	0	1	0	0	0	0	0	1 (0.7%)	0	1(0.4%)
I visually estimate blood loss	20	19	23	21	35	35	17	20	10	9	105 (75.0%)	104 (74.3%)	209 (74.6%)
I measure the volume of blood collected in drapes in milliliters	2	4	1	0	0	2	1	0	1	0	5 (3.6%)	6 (4.3%)	11 (3.9%)
I measure by weighing sponges/packs soaked with blood	1	1	3	5	1	0	3	2	1	2	9 (6.5%)	10 (7.1%)	19 (6.8%)
I measure the volume of blood in drapes + weighing sponges/packs soaked with blood	1	1	3	2	2	2	2	1	2	3	10 (7.1%)	9 (6.4%)	19 (6.8%)
I measure by counting the number of sponges/packs used	1	2	2	3	2	2	3	2	1	2	10 (7.1%)	11 (7.9%)	21 (7.5%)
I visually estimate + count the number of sponges/packs used	0	0	0	0	0	0	0	0	0	0	0	0	0
I do something else	0	0	0	0	0	0	0	0	0	0	0	0	0
TOTAL	27	27	32	31	41	41	26	25	15	16	140	140	280

Answers from self-reported methods of measuring blood loss by a sample of 140 healthcare providers collected during baseline and implementation. Data are presented as n (%).

Healthcare provider hesitations with using HSC and/or TXA

We asked the same sample of HCPs in each country a yes/no question on whether they had any hesitation with using HSC and/or TXA. Overall, HCPs had no hesitation about using HSC or TXA post-implementation in all countries, except Burkina Faso where there is a slight increase in the proportion of self-reported hesitation with using TXA post-implementation. The results are shown in Table 4. The reasons for self-reported hesitations at baseline and during implementation are expanded upon in the thematic analysis.

Table 4: Healthcare providers’ self-reported hesitations with using heat-stable carbetocin and tranexamic acid

	Burkina Fason (%)		Ethiopian (%)		Ghanan (%)		Sierra Leonen (%)		Ugandan (%)		Baseline Total n (%)	Implementation Total n (%)
Heat-stable carbetocin	B	I	B	I	B	I	B	I	B	I
Hesitations using heat-stable carbetocin	5 (16.7%)	0	17 (60.7%)	0	0	0	10 (38.5%)	0	8 (50.0%)	0	40 (28.6%)	0
No hesitations using heat-stable carbetocin	25 (83.3%)	30 (100%)	11 (39.3%)	28 (100%)	40 (100%)	40 (100%)	16 (61.5%)	26 (100%)	8 (50.0%)	16 (100%)	100 (71.4%)	140 (100%)
TOTAL	30	30	28	28	40	40	26	26	16	16	140	140
Tranexamic acid
Hesitations using tranexamic acid	4 (13.3%)	5 (16.7%)	10 (35.7%)	1 (3.6%)	0	0	8 (30.8%)	0	1 (6.3%)	0	23 (16.4%)	6 (4.3%)
No hesitations using tranexamic acid	26 (86.6%)	25 (83.3%)	18 (64.3%)	27 (96.4%)	40 (100%)	40 (100%)	18 (69.2%)	26 (100%)	15 (93.7%)	16 (100%)	117 (83.6%)	134 (95.7%)
TOTAL	30	30	28	28	40	40	26	26	16	16	140	140

Add Yes/No answers from self-reported hesitations with using heat-stable carbetocin for PPH prevention and tranexamic acid for PPH treatment by a sample of 140 healthcare providers collected during baseline and implementation. Data are presented as n (%).

Qualitative analysis

Using an inductive thematic analysis approach, three distinct but related themes were identified from responses by a sample of 140 HCPs (the findings are captured in Table 5).

Strengths and Limitations of the Study

We assessed the feasibility of concurrently integrating two new and lesser-used PPH medicines into clinical care in primary and tertiary facilities of five SSA countries. We were able to identify similarities and differences between countries and facility types, which we believe will be helpful in health systems’ decision-making. We demonstrated that despite data recording limitations, HSC and TXA were safely and appropriately implemented in all sites.

Labor induction data for Burkina Faso was unavailable for CEmONCs due to poor record capture (as is the case for unavailable blood transfusion data in Ghana). Overall, poor recording of obstetric conditions and interventions in all study countries is a major limitation of our study. We only collected the data recorded in patient records and birth registers and did not triangulate or confirm use by asking the HCPs. The duration of data capture was short, which is a limitation, but the sample size is large enough to provide useful information on safe and appropriate use.

Key Messages

• HSC and TXA can be safely and appropriately integrated into immediate postpartum care in cold-chain-challenged settings

• Refresher training that encompasses a holistic approach to PPH care and regular supportive supervision is needed to ensure continued safe and appropriate use of HSC and TXA

• Health information systems need to be improved for accurate record capture of all events during childbirth