Keywords
postpartum hemorrhage - tranexamic acid - maternal mortality - sub-Saharan Africa
Postpartum hemorrhage (PPH), the leading cause of maternal mortality, is associated
with up to 35% of maternal deaths worldwide, the vast majority of which occur in low-income
countries, where a significant proportion of births still occur in community settings.[1]
[2]
[3] PPH is caused by three main conditions: uterine atony, lacerations, or retained
products of conception. It also may result from complications of cesarean section,
ruptured uterus, disseminated intravascular coagulopathies, and other pregnancy complications.
One intervention, first studied in the early 1960s, that has recently been investigated
for PPH reduction is tranexamic acid (TXA), a fibrinolysis inhibitor now used routinely
to decrease bleeding during surgical procedures.[4]
[5]
[6]
[7] TXA has proven to be useful for treating a wide range of hemorrhagic conditions.
The drug reduces postoperative blood loss and transfusion requirements in several
types of surgery and dental procedures and appears to reduce rates of mortality and
urgent surgery in patients with upper gastrointestinal hemorrhage. A recent trial
found that the early administration of TXA significantly reduced mortality among trauma
patients.[6]
[7] Since this trial, TXA has been added to the WHO essential medicines list.[8]
[9]
TXA reduces menstrual blood loss and is a possible alternative to surgery in menorrhagia
and is sold in the United States in an oral preparation called Lysteda.[10]
[11] TXA is marketed in many countries, and in some countries, such as Great Britain
and Sweden, because of its low complication rate, it is sold as a nonprescription
medication. TXA has been shown to be stable for at least 12 weeks under a wide variety
of conditions and does not require refrigeration.[12] Following a single oral administration of two 650 mg tablets of Lysteda, the recommended
treatment for heavy menstrual bleeding, the peak plasma concentration occurred at
∼3 hours; however, the time to reduction of bleeding after oral administration is
unknown.
In obstetric trials, TXA has been used both prophylactically during labor to reduce
hemorrhage with cesarean section and as a treatment for PPH, once diagnosed.[5]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20] While these studies have been conducted in a variety of settings, in most studies,
a substantial decrease in blood loss was reported (several hundred mL). Oral administration
methods may be essential for use in community settings in low-income countries, and
for its use as both a prevention and treatment of hemorrhage in these settings.
Taken together, although the numbers are small, these studies provide evidence regarding
the efficacy of TXA to reduce important adverse maternal outcomes including blood
loss, blood transfusion, and the need for reoperation.[4]
[5]
[6]
[7]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20] Given the promising results of TXA and its potential for use in low-resource settings,
we sought to explore the potential value of TXA to reduce maternal mortality in sub-Saharan
Africa through conducting a systematic review of TXA to reduce PPH and a mathematical
model to evaluate its impact in home, clinic, and community settings in sub-Saharan
Africa.
Materials and Methods
The model used for these analyses was the Maternal and Neonatal Directed Assessment
of Technology (MANDATE) model. The model was populated with estimates of maternal
mortality and available interventions derived from many sources (as described in detail
elsewhere).[21] Additionally, a systematic literature review was conducted which included all literature
related to TXA in PubMed published in English since 1980, the Cochrane Library, and
the World Health Organization (WHO) database. Additional databases, including the
Demographic and Household Surveys (DHS) and United Nations (UN), estimated mortality
rates, rates of birth, and utilization of various interventions and mortality rates.
The review was conducted using search terms “tranexamic acid,” “pregnancy,” “postpartum
hemorrhage,” “obstetric hemorrhage,” and “maternal mortality.” Reference papers were
reviewed by at least two investigators for relevance to this analysis. A modified
algorithm was applied to use estimates of efficacy from the highest quality sources
as primary and, when unavailable, with support from other sources as secondary.[22] Finally, the results were validated by several investigators experienced in PPH
in low- and middle-income countries and incorporated into the MANDATE model for this
analysis.
We restricted this assessment to sub-Saharan Africa. [Table 1] defines the key underlying variables held constant over the analysis. Based on UN
data, there were ∼32,000,000 births annually in sub-Saharan Africa through 2012. An
estimated 50% of all deliveries in sub-Saharan Africa occurred at the home, 35% in
a health clinic, and 15% in a hospital facility.[23] Rates of PPH were estimated to be ∼11% of all pregnancies.[24]
[25]
[26] Of the PPH, ∼83% is estimated to be associated with uterine atony and the remainder
associated with lacerations or retained products of conception. Finally, the model
assumes that most women with PPH (∼90%) will cease bleeding without intervention,
but with further reductions in hemorrhage possible through use of existing interventions
(e.g., uterine massage and oxytocics).[27] The model assumes that those women who do not stop bleeding go on to have severe
hemorrhage which may result in maternal death. The condition map ([Fig. 1]) for prevention and treatment of PPH illustrates the possible interventions in the
clinical scenario of a PPH. The penetration and utilization for each of the standard
interventions described in the condition map and their associated mortality rates
are referenced elsewhere (details published on the web model).[28]
Fig. 1 MANDATE condition map for postpartum hemorrhage.
Table 1
Estimated pregnancy rates and incidence and case fatality rates for postpartum hemorrhage
in sub-Saharan Africa, 2012
|
Total
|
|
Number of pregnancies (N)
|
32,951,656
|
|
Site of prenatal care
|
|
Hospital
|
5%
|
|
Health clinic
|
65%
|
|
Home/none
|
30%
|
|
Location of delivery
|
|
Hospital
|
15%
|
|
Health clinic
|
35%
|
|
Home
|
50%
|
|
Estimated rates of PPH by delivery location
|
|
Hospital
|
8%
|
|
Health clinic
|
10%
|
|
Home
|
2%
|
|
Efficacy of TXA to treat or prevent PPH
|
30%
|
|
Postpartum hemorrhage (incidence)
|
11%
|
|
Percent of PPH by etiology
|
|
Uterine atony
|
83%
|
|
Lacerations
|
7%
|
|
Retained products of conception
|
10%
|
Abbreviations: PPH, postpartum hemorrhage; TXA, tranexamic acid.
Next, we assessed the impact of TXA in sub-Saharan Africa on PPH using an estimated
efficacy of TXA based on the review. First, we examined the estimated number of cases
of PPH and associated maternal mortality if no interventions were available for PPH.
Next, the baseline status of PPH, which does not include TXA in sub-Saharan Africa,
was modeled. Finally, we constructed a model to include TXA to evaluate the potential
impact of TXA. We modeled TXA both as prophylactic and as a treatment for PPH to assess
the range of potential impacts of the intervention based on the review.
Results
Regarding the efficacy of TXA and blood loss during pregnancy, our review found a
total of 52 papers, of which 38 full papers were reviewed and 14 studies and meta-analyses
were used to determine the estimate of the reduction in PPH[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[29]
[30]
[31] ([Fig. 2]). Additional review papers were used to inform potential use and conditions for
which TXA may be effective.[15]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40] In most studies, a significant decrease in blood loss was reported with TXA use.
Additionally, several trials reported reduced PPH associated with TXA, with a 25%
reduction in severe PPH reported in one trial.[5] Furthermore, studies evaluating TXA use with cesarean section suggested about a
one-third reduction in PPH associated with the intervention. A 2010 Cochrane review
concluded that TXA decreased postpartum blood loss after vaginal and cesarean delivery
based on two trials; however, it suggested that further investigation was needed to
confirm these findings.[29] While most PPH is due to uterine atony, the review suggested that TXA may be effective
in reducing hemorrhage from all major causes of PPH, including uterine atony, lacerations,
and retained products. Based on the review, we estimated that TXA used either prophylactically
or as treatment of PPH could be effective in reducing PPH rates by 30% (range, 25–40%).
Fig. 2 Systematic review of tranexamic acid for postpartum hemorrhage.
[Table 2] summarizes the specific numbers for each scenario starting from the number of cases
of PPH, the PPH cases averted with TXA, the number of cases of severe PPH, the number
of cases of severe PPH averted with TXA, the number of deaths from PPH, and the number
of deaths averted with TXA. The cases and deaths averted compare the baseline scenario
with various TXA treatments.
Table 2
Summary of scenarios evaluating impact of tranexamic acid (TXA) on postpartum hemorrhage
(PPH) in sub-Saharan Africa, 2012
|
N
|
Scenario[a]
|
Prevalence of PPH
|
PPH cases prevented by TXA
|
Prevalence of severe PPH
|
Severe PPH cases prevented by TXA
|
Deaths from PPH
|
PPH deaths prevented by TXA versus baseline
|
|
1
|
No interventions for PPH
|
3,700,703
|
NA
|
370,070
|
NA
|
95,108
|
NA
|
|
2
|
Baseline prevalence with current interventions to reduce mortality from PPH
|
3,075,759
|
NA
|
292,887
|
NA
|
71,797
|
NA
|
|
3
|
TXA treatment (in addition to current treatments) in hospitals at 99% penetration and utilization
|
3,075,759
|
NA
|
292,887
|
NA
|
70,403
|
1,394
|
|
4
|
TXA treatment (in addition to current treatment) in hospitals, clinics, and home at 99% penetration and utilization
|
3,075,759
|
NA
|
292,887
|
NA
|
61,840
|
9,957
|
|
5
|
TXA treatment in home, clinic, and hospital at 65, 85, and 95% utilization and penetration at home, clinic, and hospital
|
3,075,759
|
NA
|
292,887
|
NA
|
65,264
|
6,533
|
|
6
|
TXA prophylaxis of PPH (in addition to current treatments) in hospitals, clinics, and at home at 99% penetration and utilization
|
2,515,849
|
559,910
|
238,732
|
54,154
|
58,346
|
13,452
|
|
7
|
TXA prophylaxis and PPH treatment (in addition to current treatments) in hospitals, clinics, and at home at high rates of penetration and utilization
|
2,515,849
|
559,910
|
238,732
|
54,154
|
50,210
|
21,587
|
|
8
|
TXA prophylaxis and PPH treatment (with current treatment) in hospitals, clinics, and at home at more plausible rates of penetration and utilization
|
2,787,238
|
288,721
|
265,192
|
27,695
|
59,159
|
12,638
|
a For TXA, high penetration and utilization are defined as 99% for each, while more
plausible rates are defined as 65% in the home, 85% in health clinic, and 95% in hospitals
for penetration and utilization.
The first scenario summarizes the estimated mortality if there were no interventions
used to prevent, diagnose, or treat PPH. In this scenario, we estimate that there
would be ∼95,000 maternal deaths in sub-Saharan Africa if no interventions were available.
The second scenario provides the estimated mortality and cases of severe PPH in sub-Saharan
Africa in 2012, given the currently available interventions, such as uterine massage,
oxytocics, transfusions, and surgical interventions. We estimate that in 2012 there
were ∼293,000 cases of severe PPH and 72,000 deaths from severe PPH.
The third scenario adds TXA at high (99%) penetration and utilization to treat severe
PPH only for women in hospitals, in addition to currently available interventions.
In this scenario, the number of severe PPH cases did not change but the proportion
of women who ultimately died from PPH was reduced. We estimate 1,394 maternal lives
would be saved by this treatment. The fourth scenario evaluates TXA given as a treatment
for severe PPH at home, clinic, and hospitals at high (99%) penetration and utilization.
Again, the number of severe PPH cases has not changed, but the number of women who
ultimately die is estimated to be reduced by 9,957 from the current care model because
∼85% of the deliveries in sub-Saharan Africa occur at home or community health clinics.
In the fifth scenario, we modeled TXA use at home, clinic, and hospitals, but at more
plausible rates of penetration and utilization, and the number of lives saved decreased
nearly by half, compared with the previous scenario.
In the next series of scenarios, we evaluated the potential use of TXA as a prophylaxis
against PPH. In the sixth scenario, TXA was available for prophylaxis at high (99%)
penetration and utilization at the home, hospital, and clinic. The number of cases
of severe PPH is estimated to be reduced by 54,154 to 238,732 and maternal deaths
are reduced to 58,346. We estimate that 13,452 maternal lives will be saved.
In the seventh scenario, TXA is used prophylactically as in the sixth scenario, but
in addition, it was also used as a treatment at high (99%) rates for those women who
went on to develop PPH. The maternal mortality was further decreased to an estimated
50,210 deaths and 21,587 maternal lives were saved.
Finally, in the last scenario, the penetration and utilization for TXA prophylaxis
was reduced to more plausible levels. The number of lives saved from TXA was reduced
by nearly half from the most optimistic scenario.
Comment
One of the challenges of reducing maternal mortality associated with PPH in sub-Saharan
Africa has been the low rate of hospital births and the lack of effective interventions
to reduce PPH in community settings.[3]
[41] We sought to explore the potential impact of an intervention that has effectively
reduced bleeding in surgical settings and may ultimately be used in health clinics
or in home deliveries. We found that with near-perfect availability and use of TXA
for both prophylaxis and treatment, one could potentially reduce maternal mortality
attributable to PPH by more than 30%.
Reliable data to estimate mortality and the impact of interventions on mortality remain
limited, especially from low-resource geographic areas in sub-Saharan Africa. Furthermore,
while we used an estimated efficacy for TXA to reduce PPH (30%) which was reasonable
based on available studies, the research into the efficacy of TXA for prevention of
PPH and especially maternal mortality is limited. However, we used a relatively conservative
estimate and ran several different scenarios to find a range of impact. Our results
suggest that there is important potential for maternal mortality reduction for an
intervention that could be given in home settings, especially if it were effective
for different causes of PPH. In preliminary research, TXA appears to have several
qualities that would make it ideal for low-resource settings—including its low cost,
safety profile, and its ability to be given orally.[9]
[10] Given that PPH remains a leading cause of mortality and the interventions available
to date have not substantially reduced PPH, further evaluation of, and potential investment
in, interventions such as TXA appears to be warranted. Regarding safety, there is
a potential for an increased risk of thrombosis since TXA is a fibrinolysis inhibitor.
However, data from both obstetrical and surgical studies have not shown this to be
an issue. There does appear to be a mild increase in short-lived gastrointestinal
symptoms associated with TXA use.
This study has several strengths and some limitations. First, we are aware that data
for some of our variables are limited and estimates are used. However, we have done
extensive literature reviews of published papers and many other sources. We are also
aware that maternal hemorrhage occurs in concert with other morbidities, and even
if the hemorrhage is prevented or treated, death may occur due to other etiologies.
The model accounts for maternal deaths due to other etiologies, and if there is no
death from maternal hemorrhage, it recognizes that mortality from other etiologies
can still occur. That the results are based on a computer model is both a strength
and limitation. Using the model allows us to deal with hypothetical questions and
to present results from several potential “what if?” scenarios. Since these scenarios
are hypothetical, whether in actual practice they can be achieved is unknown. Nevertheless,
the computer model presents an approximation of the deaths that may be prevented using
TXA and helps to determine whether carrying out additional studies on this intervention
is likely to be worthwhile. The model also allows an approximation of the potential
lives saved if TXA is used only for treatment of PPH compared with potential lives
saved if used prophylactically. It also points out that if TXA use is restricted to
hospitals alone, few lives will be saved. Expanding TXA treatment and especially prophylaxis
to clinic and home deliveries will potentially save many more lives. Therefore, randomized
trials of TXA given both as a prophylactic and as treatment to reduce PPH and maternal
mortality in different settings would be important.
Clinical Perspective
-
Community-based interventions are needed to prevent postpartum hemorrhage, which remains
a major cause of maternal mortality in developing countries.
-
A systematic review was conducted and mathematical model developed to evaluate the
potential impact of tranexamic acid, a drug to reduce bleeding that is potentially
suitable for community settings.
-
Tranexamic acid potentially will have large impact to prevent postpartum hemorrhage
in developing countries.
