review saving maternal lives in resource-poor settings
TRANSCRIPT
A
OrMsitRucaDmdth©
K
C
T
0
Health Policy 89 (2009) 131–148
Available online at www.sciencedirect.com
Review
Saving maternal lives in resource-poor settings: Facing reality
Ndola Prata ∗, Amita Sreenivas, Farnaz Vahidnia, Malcolm Potts
Bixby Program in Population, Family Planning, and Maternal Health, School of Public Health,University of California, Berkeley, United States
bstract
bjective: Evaluate safe-motherhood interventions suitable for resource-poor settings that can be implemented with currentesources.
ethods: Literature review to identify interventions that require minimal treatment/infrastructure and are not dependent onkilled providers. Simulations were run to assess the potential number of maternal lives that could be saved through interventionmplementation according to potential program impact. Regional and country level estimates are provided as examples of settingshat would most benefit from proposed interventions.esults: Three interventions were identified: (i) improve access to contraception; (ii) increase efforts to reduce deaths fromnsafe abortion; and (iii) increase access to misoprostol to control postpartum hemorrhage (including for home births). Theombined effect of postpartum hemorrhage and unsafe abortion prevention would result in the greatest gains in maternal deathsverted.iscussion/conclusions: Bold new initiatives are needed to achieve the Millennium Development Goal of reducing maternalortality by three-quarters. Ninety-nine percent of maternal deaths occur in developing countries and the majority of these women
eliver alone, or with a traditional birth attendant. It is time for maternal health program planners to reprioritize interventions inhe face of human and financial resource constraints. The three proposed interventions address the largest part of the maternalealth burden.
2008 Elsevier Ireland Ltd. All rights reserved.
eywords: Maternal mortality; Family planning; Postpartum hemorrhage; Abortion; Interventions; Misoprostol
ontents
1. Maternal mortality worldwide . . . . . . . . . . . . . . . . . . . . . . . . . . .2. Skilled attendants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. Emergency obstetric care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
∗ Corresponding author at: 229 University Hall, 2200 University Avenue,el.: +1 510 643 4284; fax: +1 510 643 8236.
E-mail address: [email protected] (N. Prata).
168-8510/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights resedoi:10.1016/j.healthpol.2008.05.007
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
University of California, Berkeley, CA 94720-7360, United States.
rved.
132 N. Prata et al. / Health Policy 89 (2009) 131–148
5. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1365.1. What will make a difference in resource-poor settings? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1365.2. Family planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1365.3. Safe abortion and post-abortion care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1375.4. Postpartum hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1405.5. Potential maternal deaths averted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
6. Discussion/conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145. . . . . .
1
daMcwsaadparacccpit[
puvtima1Atraad
pta
btiHittsroa
2
awfWaenrshatho
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. Maternal mortality worldwide
Ninety-nine percent of maternal deaths occur ineveloping countries and most of these women deliverlone or without a skilled birth attendant (SBA) [1].ore than 500,000 women die from pregnancy and
hildbirth around the globe annually [2]. For everyoman who dies, an estimated 30 women live to
uffer severe morbidities including infertility, fistula,nd incontinence [1]. In Sub-Saharan Africa (SSA)lone, women are 250 times higher risk of dyinguring pregnancy or childbirth than their counter-arts in industrialized countries [2]. Maternal deathsre highest in the most difficult to access areas (i.e.ural settings)—where women are unable to reachn SBA or appropriate facilities in times of compli-ation during labor or delivery [3]. This burden isompounded by the shear reality that most compli-ations occur at the time of delivery and cannot beredicted beforehand. For example, many pregnanciesdentified as high risk deliver normally and most life-hreatening complications occur in low risk women4].
The main direct causes of maternal death includeostpartum hemorrhage (PPH), sepsis, eclampsia,nsafe abortion, and obstructed labor (Table 1). Usingarious country datasets, Khan et al. (2006) estimatehat hemorrhage is the main cause of maternal mortalityn Asia and Africa—accounting for 30% or more of all
aternal deaths [5]. Deaths due to eclampsia, unsafebortion, and obstructed labor combined account for7.7% and 26.7% of maternal deaths in Africa andsia, respectively. However, these estimates are likely
o be higher as the datasets used do not represent a
andom sample of the population at risk, but simplyvailability of data [5]. Misclassification has also beenn issue in accurately assessing causes of maternaleath [6]. Figures related to abortion, for example, areqiep
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
robably much higher as social stigma and legal restric-ions associated with abortion in many countries meansbortion-related deaths may likely go unreported [7].
To date, the primary strategies to reduce MMRy the international community have been to ensurehat every woman has ready access to an SBA dur-ng delivery and EmOC in case of complications [1].owever, poor settings are often plagued by electric-
ty outages, lack of transportation or viable roadwayso reach higher level facilities, absence of potentialrainees for professional clinical positions, and profes-ional migration. The devastating shortage in humanesources and necessary infrastructure across the devel-ping world makes these interventions impossible tochieve quickly.
. Skilled attendants
Safe-motherhood service packages to reduce MMRre shown to be highly cost-effective, yet a nurse, mid-ife, or doctor for every delivery is still a distant goal
or those regions with the highest MMRs [8]. Theorld Health Organization (WHO) estimates that there
re 57 countries with a critical deficit of health work-rs adding up to 2.4 million doctors, midwives andurses. SSA and South & South East Asia are cur-ently experiencing the greatest deficit [9]. In placesuch as Africa, where global burden of disease is asigh as 24%, only 3% of the world’s available providersre living and working in Africa. This is a sharp con-rast to the Americas (incl. Canada and the US) whichas 10% of the global burden of disease, but has 37%f available health workers [9]. This has dire conse-
uences on maternal health. The lowest MMRs occurn those countries where the highest proportion of deliv-ries are attended by an SBA and where contraceptiverevalence (CPR) is high (Table 2). In developing coun-N. Prata et al. / Health Policy 89 (2009) 131–148 133
Table 1Distribution of causes of maternal deaths by region and available priority interventions
Average percent distribution 5 Available priority intervention(s)
Africa Asia Latin America, the Caribbean
Hemorrhage 33.9 30.8 20.8 AMTSLa
Uterotonic agents: Oxytocin, Ergometrine, MisoprostolBlood and IV fluidsAnti-shock garmentBalloon catheterSurgical procedures
Hypertensive disorders 9.1 9.1 25.7 Blood pressure monitoringAntihypertensivesMagnesium sulfateSupportive air management
Infections/sepsis 9.7 11.6 7.7 STI treatmentClean deliveryAntibioticsVitamin A supplementation
Unsafe abortion 3.9 5.7 12 Family planningSafe abortion carePost-abortion care
Obstructed labor 4.1 9.4 13.4 PartographMisoprostol for inductionInstrumental deliveryCesarean section
Other direct 7.4 2.1 4.9Indirectb 32.0 31.4 15.7
a AMTSL: active management of third stage of labor (oxytocin; cord traction; uterine massage).b Including anemia, HIV/AIDS, and unclassified deaths.
Table 2Maternal mortality ratio (MMR), percent births attended by a skilled professional, percent rural population, and contraceptive prevalence byregion
Region Maternal mortality ratio (a, b)(deaths per 100,000 live births)
% Births with skilledattendants (b)
% rural (b) Contraceptive prevalence(%) (b)
World total 402 62 50 54Developed regions 9 99 25 57Europe 17 99 28 53Africa 824 47 61 21Northern Africa 157 70 48 45Sub-Saharan Africa 905 53 59 21Asia 329 58 59 58Eastern-Asia 53 97 54 81South-Central Asia 581 39 69 42South-Eastern Asia 280 69 55 51Western Asia 127 73 35 29Latin America/Caribbean 132 83 22 63Oceania 427 84 29 57
Source: (a) Hill et al. (2007); (b) State of the World’s Population (2007).
1 h Polic
tw
waidGeswup8rmhng
3
bwemtcfc(diarniia
F(p[iE
thtomnEta
teaTbcalpst
4
iWucnbotitanw
USle
34 N. Prata et al. / Healt
ries more than half of pregnant women deliver at homeithout an SBA [10].Even if efforts to increase the number of SBAs
ere moving ahead quickly, developing countries facenother issue—brain drain, the tendency of trainedndividuals to migrate to urban areas and moreeveloped nations. The International Federation ofynecology and Obstetrics and International Confed-
ration of Midwives recommend that there be onekilled provider for every 5000 individuals. However,ith current brain drain, the average physician to pop-lation ratio is as low as 1 doctor for every 100,000eople in developing countries. In the rural areas, where0% of populations from the developing world live, thisatio is even lower [11]. Thus, although there are safe-otherhood interventions which have been shown to be
ighly cost-effective, the doctors, nurses or midwiveseeded at every delivery to care for them is still a distantoal for those regions with the highest MMRs [8].
. Emergency obstetric care
Barriers to reducing MMR are further compoundedy the reality that even if there were available healthorkers, there is a significant shortage in facilities and
quipment for clinical personnel to provide emergencyaternal care. EmOC refers to a set of interven-
ions prescribed for women who are suffering fromomplications during labor or delivery that require pro-essional assistance. It is divided into two levels ofare: Basic EmOC (BEOC) and Comprehensive EmOCCEOC). BEOC are performed in a health center, butoes not require an operation theatre. Interventionsnclude administration of antibiotics, oxytoxics, andnticonvulsants, manual removal of placenta or otheretained products of pregnancy, and an assisted vagi-al delivery. CEOC requires an operation theatre ands often performed in district hospitals. Interventionsnclude BEOC services as well as caesarean sectionnd safe blood transfusion [1].
According to the WHO, United Nations Children’sund (UNICEF), and United Nations Population FundUNFPA), the minimum number of EmOC facilities per
opulation is 4 BEOC and 1 CEOC per 500,000 people2]. A recent study done by Paxton et al. (2006), exam-ned 39 countries to determine the global availability ofmOC facilities. The study found that CEOC facilitiesCift
y 89 (2009) 131–148
end to be available even in the poorest of countries,owever BEOC are consistently unavailable in coun-ries with high and moderate MMRs. The authors pointut that even though most countries have access to theinimum required CEOC facilities there remains a sig-
ificant fear about the quality of care being provided atmOC facilities, whether care is being provided equi-
ably, and how accessible facilities are geographicallynd financially to healthcare seekers [12].
EmOC requires skilled staff, a reliable transporta-ion system, adequate supply of drugs and functioningquipment (including operation theatres for CEOC),nd are thus difficult to scale up in poor settings [13].he reality is that even if there is the minimum num-er of CEOC facilities per 500,000 individuals, theseenters are rarely fully functioning in the places theyre needed most and are typically only built in urbanocations, making them completely inaccessible to theoorest of the poor living in rural areas. Therefore weuggest it is imperative to explore alternative strategieso reduce MMR.
. Methods
We reviewed literature available on public databasesncluding PubMed, Population Information Online,
eb of Science, and the Scholarly Journal Archive,sing the terms: health worker shortage, developingountries, maternal mortality, causes of mater-al death, traditional birth attendants, misoprostol,irth attendant, postpartum hemorrhage, third stagef labor, developing world, community interven-ions, oxytocics, active management, maternal healthnterventions, abortion, unintended pregnancies, con-raception. We also searched the reference lists from therticles found using these terms. Publication period wasot specified and all studies the search engine returnedere considered.In addition, relevant documents from the WHO,
NICEF, and UNFPA were reviewed. We employedtat Compiler to gather country-specific data from the
ast available Demographic Health Survey (DHS) forach country. The Population Reference Bureau Data
ompiler was also used for further country-specificnformation. Specific MMR information was gatheredrom DHS reports, Hill et al. (2005), and the State ofhe World Population 2007 report.
h Polic
itobohts
sPottRaTcnT
nmii
Fpw[BEoteFbt
TM
R
P
F
S
S(t
N. Prata et al. / Healt
To determine the three suggested maternal healthnterventions on which program planners should focusheir efforts, we first tabulated the distribution of causesf maternal mortality and the interventions that haveeen proven to avert those causes (Table 1). We focusedn those interventions that could be scaled up withoutaving to increase existing health delivery infrastruc-ure and/or human capacity and require a low level ofkilled provider to carry out the intervention.
After deciding on the three interventions thathould be prioritized, we used the WHO Mother Babyackage (MBP) to assess the relative effectivenessf the intervention, and Monte Carlo simulationso assess the potential number of maternal liveshat could be saved after program implementation.egional and country level estimates are calculateds examples of settings that would benefit the most.
he MBP prescribes a set of 18 basic interventionsonsidered essential for decreasing maternal andeonatal mortality in resource poor settings [14].he package describes simple, effective interventions(bdp
able 3onte Carlo simulation assumptions
egion/country Number of maternaldeaths (a, b)
Proportion of attributalcause of maternaldeath (%) (c)
ostpartum hemorrhageSub-Saharan Africa 270,500 25Asia 240,600 25Niger 12,269 25Rwanda 5,199 25Burkina Faso 4,662 25Cambodia 2,022 25
amily planningSub-Saharan Africa 270,500 13Asia 240,600 13Niger 12,269 13Rwanda 5,199 13Burkina Faso 4,662 13Cambodia 2,022 13
afe AbortionSub-Saharan Africa 270,500 13Asia 240,600 13Niger 12,269 13Rwanda 5,199 13Burkina Faso 4,662 13Cambodia 2,022 13
ource: (a) Hill et al. (2005); (b) State of the World’s Population (2007); (ce) 40–80% possible program dissemination for family planning services wo (b) is 19%.
y 89 (2009) 131–148 135
eeded before, during, and after delivery for theother and newborn. It also outlines the maximum
mpact on averting maternal deaths that individualnterventions can have after their implementation.
Using the information provided by the MBP andortney’s study on the potential impact of familylanning services (1987), Monte Carlo simulationsere run for each intervention (10,000 trials for each)
14,15]. Simulations were conducted using Crystalall 7 software, a stochastic modeling supplement forxcel. We performed multivariate sensitivity analysisn the results of each model to assess contributiono variance given our underlying assumptions. Mod-ls were run for SSA, Asia, Niger, Rwanda, Burkinaaso, and Cambodia. The four countries were chosenased on their high MMRs, the proportion of popula-ion living in rural regions (>75%), and their low CPRs
<20%). Assumptions used in the models are the num-er of maternal deaths, proportion of deaths due to airect cause of maternal death, and range of possiblerogram dissemination into the population and poten-Estimated numberof maternal deaths
Programcoverage (%)
Estimated impact ofintervention (%) (c, d)
67,625 20–80 55–8260,150 20–80 55–823,067 20–80 55–821,300 20–80 55–821,166 20–80 55–82
506 20–80 55–82
35,165 20–80 26–5831,278 20–80 26–581,595 20–80 26–58
676 20–80 26–58606 20–80 26–58263 40–80 (e) 26–58
35,165 20–80 75–9531,278 20–80 75–951,595 20–80 75–95
676 20–80 75–95606 20–80 75–95263 20–80 75–95
) Source: WHO Mother Baby Package; (d) Source: Fortney (1987);as used as current contraceptive prevalence in Cambodia according
1 h Polic
tta(
5
5s
aaitaitsspcappib
wuschHphffiter
5
a2pww
F(
36 N. Prata et al. / Healt
ial program impact. These assumptions were then usedo calculate the number of maternal deaths that could beverted with varying degrees of program disseminationTable 3).
. Results
.1. What will make a difference in resource-poorettings?
Access to an SBA, as achieved in countries suchs Sri Lanka, must not be abandoned in any way,nd improved access to EmOC is imperative where its achievable. However, in the intermediate term forhe 50 million women who do not have SBAs andre beyond the reach of EmOC, only a few prioritynterventions (Table 1) are realistic. After reviewinghe literature, we found that achievable, cost-effective,calable strategies to reduce MMR in resource-poorettings include: decreasing the number of unintendedregnancies by increasing access to contraception;onfronting the public health implications of unsafebortion by increasing access to contraception, com-
rehensive abortion care, safe abortions (SA), andost-abortion care (PAC); as well as controlling PPH byncreasing misoprostol availability (including for homeirths).thn[
ig. 1. Maternal mortality and contraceptive prevalence for countries with aDHS), country specific data for the last five.
y 89 (2009) 131–148
If we examine the direct causes of maternal death,e will find that interventions aimed at preventingnwanted pregnancies, unsafe abortion, and PPH shareomething in common which others do not. Theyan be implemented outside of the hospital, in theands of lower-level providers, and even at home.ypertensive disorders and obstructed labor for exam-le, require highly trained assistance in a facility thatas sufficient equipment, including magnesium sul-ate solutions for eclampsia, and an operating roomor cesarean-sections. Sepsis/infection interventions,ncluding STI treatment and antibiotic administra-ion are also facility-based in today’s global medicalstablishment and are thus difficult to scale up in poor-esource settings.
.2. Family planning
Two hundred and five million pregnancies occurnnually worldwide, 35% of which are unintended and2% of which end in an induced abortion. Most of theseregnancies (182 million) happen in the developingorld. Two-thirds of these pregnancies occur amongomen who are not using any method of contracep-
ion, making FP a significant contributor to maternalealth. Contraceptive use decreases the risk of mater-al death by decreasing the odds of being pregnant15]. Fig. 1, a graph assembled using MMR and CPR
DHS in the last 5 years. Source: Demographic and Health Surveys
h Polic
dc
aoAnontawcideAwbtttta[
itubnFop[
cmtfmibhotat
fnwAcci[icp
5
hwpaoiprslTaii[fdv[
nnScwbiai
N. Prata et al. / Healt
ata from the DHS, demonstrates a strong negativeorrelation.
Today, there are over 100 million married womenlone who have an unmet need for contraception. Mostf these women live in SSA and South & Southeastsia [16]. Among those women who become preg-ant and elect to have an abortion, they are at riskf severe complications, including death if they doot have access to a skilled provider. It is estimatedhat 11% of induced abortions in developing countriesre done in unsafe conditions. This is over 20 millionomen who are at risk of dying every year from a
ompletely preventable condition [16]. Yet, by increas-ng modern contraceptive use, the rate of abortion willecline dramatically. The Center for Population report,ntitled The Role of Family Planning in Preventingbortion, acknowledges that there have been countrieshere use of contraception and abortion rates appear toe rising at the same time. However, authors reiteratehat historical examples across the globe demonstratehat this “phenomenon may occur because the desireo have smaller families may be increasing faster thanhe availability of contraceptive information/servicesnd the consistent use of effective contraceptives”17].
Contraception also has a differentially positivempact on the MMR because contraceptives are some-imes adopted first by the women over age 35 andnder 20. Adolescence, older age, high-parity and shortirth intervals are all known risk factors for mater-al mortality [18]. It is estimated that the impact ofP on reducing MMR among women below the agef 20, above the age of 39, or among those with aarity of more than 5 ranges from a 25% to 58%15].
Provision of family planning (FP) is a solution thatan be achieved quickly and is sustainable. There areany reasons why women may not choose to use con-
raceptives. It is important, therefore, to address anyears or preferences that women may have in choosing aethod to control their own fertility. Common concerns
nclude possible health and side effects, a woman’selief that they are not at risk of getting pregnant, andaving to negotiate contraceptive use with a partner
r her own religious beliefs. However, in communi-ies where several choices are made available, womenre more likely to find a method that is appropriate forhem. Injectable contraceptives, the method of choicetmc
y 89 (2009) 131–148 137
or many women in rural areas, such as Depo, haveow given program planners a way of reaching womenith an unmet need for contraception more quickly.lthough Depo requires an injection, studies are indi-
ating that low level providers, such as paramedics andommunity health workers, are able to properly admin-ster Depo injection to women at the community level19]. Furthermore, studies have shown high acceptabil-ty of injectables as it is both a long-term method andan be done privately without negotiation with one’sartner [20].
.3. Safe abortion and post-abortion care
When speaking of the impact which abortion canave on MMR, it is not abortion which causes aoman to die. It is having one that is unsafe, whichuts her at risk of complications such as hemorrhagend sepsis. According to the WHO, the definitionf a unsafe abortion is “a procedure for terminat-ng an unintended pregnancy carried out either byerson lacking the necessary skills or in an envi-onment that does not conform to minimal medicaltandards, or both.” It is estimated that around 5 mil-ion women are hospitalized every year for PAC [21].his accounts for the largest proportion of hospitaldmissions for gynecological services in the develop-ng world [7]. Among those women who have children,t is estimated that 220,000 children lose their mothers21]. It is believed that worldwide, 67,000 women dierom PAC—the equivalent of 13% of total maternaleaths [16]. It is estimated that the impact of SA ser-ices on reducing maternal mortality ranges 75–95%14].
Unsafe abortion is an important cause of mater-al death. Developed countries, for example, wouldot have achieved very low MMRs without access toA. Furthermore, by preventing unwanted pregnan-ies, unsafe abortion is easily preventable [11]. Evenhere abortion is an illegal practice, several steps cane taken to reduce MMRs attributed to it. The first, is tomprove access to contraception (as explained above)nd PAC (including FP counseling) [22]. The seconds to provide SA to the extent allowed by the law.
There are several options for first trimester abor-ion and PAC. These include both surgical (e.g.
anual vacuum aspiration (MVA) and dilation andurettage) and medical methods (e.g. mifepristone
138 N. Prata et al. / Health Policy 89 (2009) 131–148
Table 4Misoprostol for safe abortion (incl. missed abortion) and post-abortion care
Author Year Country n Study Results [RR (95% CI)] and conclusions
Post-Abortion careCoughlin 2004 UK 131 Effectiveness of 400 �g oral misoprostol
for first trimester incomplete miscarriageSuccessful treatment was achieved in77.7% of womenSingle dose of 400 �g of oralmisoprostoal was an effective treatmentfor women presenting with anincomplete miscarriage
Murchison 2004 USA 41 Study of 800 �g misoprostol intravaginaladministration for uterine evacuation formissed abortion
Overall success rate was 78%.
‘Intravaginal miso is a safe, effectivealternative to surgical currettage for thetreatment of missed abortion
Bluma 2007 – – Review of all English language articlespublished before October of 2007 usingmisprostol to treat incomplete abortionand miscarriage in the first trimester
Sufficient evidence exists to supportmiso as a safe and effective means ofnon-surgical uterine evacuation
Recommendation: 600 �g oral miso totreat incomplete abortion in womenduring the first trimester.
Gemzell-Danielssonb 2007 – – Review of published literature onmisoprostol to treat missed abortion infirst trimester.
800 �g vaginal miso is as effective andsafe as traditional surgical treatment
600 �g sublingual miso can also be usedfor same indication.
Sharma 2007 India 50 Prospective study of 600 �g misoprostolevery 3 h for a maximum of 3 doses forwomen up to 13 weeks gestation formissed abortion.
Success rate was 86% (up to 72 h)
Acceptability of the method was 69.7%
Vejborg 2007 Denmark 355 Study of effectiveness of single dosevaginal misoprostol 400 �g to reduce thenumber of surgical interventions in earlypregnancy failure
Single dose misoprostol reduced thenumber of surgical interventions
Success rates of miso treatment variedaccording to ultrasonographic definitionsof pregnancy failure, time of assessmentand criteria for success
Safe abortion in the first trimesterHerabutya 1997 Thailand 84 Compare 200 �g misoprostol vs. placebo
in women with a missed abortion200 �g vaginal misoprostol producedspontaneous expulsion in women with amissed abortion and reduced the need forsurgical treatment83.3% in miso prostol group abortedspontaneously compared to 17.1% in theplacebo group (p < 0.0001)
N. Prata et al. / Health Policy 89 (2009) 131–148 139
Table 4 (Continued )
Author Year Country n Study Results [RR (95% CI)] and conclusions
Moreno-Ruizc 2006 – – Review published reports offirst-trimester medical abortion regimensthat do not include mifepristone with atleast 100 participants and were publishedsince 1990. Author examinesmisoprostol alone.
Alone or in combination withmethotrexate, misoprostol is anefficacious alternative to mifepristone forthe medical termination of pregnancy
Faundesd 2007 – – Review of publications reporting onmisoprostol alone for pregnancytermination within the first 12 weeks ofpregnancy.
Vaginal administration of 800 �grepeated up to three times has an 85 to90% effectiveness in most studies
Authors compare vaginal administrationof 800 �g of misoprostol repeated up tothree times at 6, 12 or 24 h interval to (i)oral administration at the same dose; and(ii) sublingual administration at 3 hintervals.
Oral administration is less effective
Sublingual administration is as effectiveOral and sublingual administrationappear to be better accepted than vaginaladministrationMost studies are limited to first 9 weeksof pregnancy. More research is neededfor >9 weeks
Prasad 2008 India 140 Comparison between saline-soakedmisoprostol (800 �g) administeredvaginally and surgical evacuation inhospital setting.
Single does vaginal misoprostol hascomparable success rate as surgicalmethod for termination.
Side effects were fewer in miso groupand had higher acceptability rate
a Blum et al. examine twenty-six studies: Chug (1999); Chung (1999); Pang (2001); Gronlund (2002); Ngai (2001); Demetroulis (2001); deJonge (1995); Pandian (2001); Agostini (2005); Bagratee (2004); Creinin (1997); Davis (2004); Graziosi (2004); Henshaw (1993); Moodliar(2005); Muffley (2002); Zhang (2005); Ngoc (2005); Blanchard (2004); Weeks (2005); Shwekerela (2007); Dao (2007); Bique (2007); Trinder(2006); Rakotovao (2006).
b Gemzell-Danielsson et al. examine eight additional studies: Kovavisarach (2002); Wood (2002); Heard (2002); Kovavisarach (2005); Ngoc(2004); Tang (2003); Herbutya (1997); Lister (2005).
c Moreno-Ruiz et al. examine twelve additional studies on misoprostol alone: Carbonell (2003); Singh (2003); Zikopoulos (2002); Jain (2002);J bonell
isoprost( ai (2000
atirnssseo[
reimtt
ain (2001); Carbonell (2001); Bugalho (2000); Velazco (2000); Card Faundes et al. examine eight additional studies which analyze m
1999); Bugalho (1996); Blanchard (2005); von Hertzen (2007); Ng
nd misoprostol combination as well as misopros-ol alone). Among options available, women livingn low-resource settings would benefit most fromegiments of misoprostol alone, backed by MVA ifeeded, or MVA, as they usually require fewer sub-idies. Moreover, both technologies can be used touccessfully manage incomplete abortions. They are
afe, effective, cheap, can be performed without gen-ral anesthesia, and usually preempt the need forvernight stay, saving hospital and individual costs23].wstt
(1999); Carbonell (1997); Carbonell (1997); Wiebe (2004).ol alone for first trimester induced abortion: Carbonell (1998); Jain); Salakos (2005); Carbonell (2003).
Table 4 is a list of studies and their results/ecommendations on using misoprostol alone for bothlected and missed abortions as well as PAC. Evidencendicates that misoprostol-alone is a safe and effective
eans to induce abortion or treat abortion complica-ions [24–33]. Furthermore, researchers have foundhat misoprostol is as effective and as acceptable to
omen as MVA, and thus well-suited for low-resourceettings [34–36]. Overall, misoprostol has the potentialo help significantly reduce the number deaths relatedo unsafe abortion.
140 N. Prata et al. / Health Policy 89 (2009) 131–148
Table 5Options to prevent/treat postpartum hemorrhage
Place of delivery Type of assistance Options
Comprehensive obstetric care facility Highly skilled AMTSLUterotonicsOxytocinErgomethrinMisoprosol
Fluid replacementBlood transfusionSurgery
Basic obstetric care, i.e. primary health care facility Skilled AMTSLUterine massagea
Cord tractiona
UterotonicsOxytocin: IVb, Unijectb,c
Ergomethrinb,c
MisoprostolFluid replacementb
Antishock garmentd
Hydrostatic condom balloon cathetera,d
Household level Skilled AMTSLUterine massagea
Cord tractiona
UterotonicsOxytocin: Unijecta,c
MisoprostolTraditional birth attendant HBLSS
Uterine massagea
UterotonicsMisoprostol
None HBLSSUterotonicsMisoprostol
HBLSS: home-based live saving skills.a Requires special skills and training.
5
wpc5adpoU
wiamM
oW
b Requires injection, skills and training.c Requires electricity and special storage.d Under study, requires high resources.
.4. Postpartum hemorrhage
PPH is the leading cause of maternal mortalityorldwide. An estimated 14 million women experienceregnancy-related hemorrhage each year. It is clini-ally defined as blood loss greater than or equal to00 mL. Eighty percent of PPH cases are due to uterinetony, whereby the uterus fails to contract after the baby
elivers. Other causes of PPH include retention of thelacenta or pieces of the placenta, lacerations or tearsf the cervix, vagina, or perineum, and uterine rupture.nfortunately ninety percent of women do not presenttaci
ith risk factors, making PPH prevention extremelymportant especially in low-resource settings, whereccess to care is scarce or non-existent [1,37]. It is esti-ated that the impact of PPH management on reducingMR, ranges from 55% to 82% [14].Options to prevent or treat PPH depend largely
n place of delivery setting and resources (Table 5).e focused on what is feasible, affordable, and sus-
ainable in home deliveries with limited resourcesnd minimally trained delivery attendants. The mostommonly recommended methods of managing PPHnclude active management of the third stage of labor
N.P
rataetal./H
ealthPolicy
89(2009)
131–148141
Table 6Misoprostol and the prevention of postpartum hemorrhage
Author Year Country n Study Results [RR (95% CI)] and conclusions
Bhullar 2004 USA 848 RCT comparing Buccal 200 �g misoprostol tablet or placebo attime of cord clamping
Miso vs. placebo for blood loss >500 mL: RR0.65 (0.33–1.29)
Buccal miso at cord clamping is not more effective than placebo inreducing PPH
Langenbacha 2005 – 30017 Meta-analysis of 22 RCTs of misoprostol vs. placebo ormisoprostol vs. other uteronics
Miso vs. Placebo for blood loss >500 mL and >1000 mL: RR0.85(0.63, 1.14)Miso vs. Oxytocics for blood loss >500 mL: RR1.4 (1.22, 1.56))Miso vs. Oxytocics for blood loss >1000 mL: RR1.25 (0.94, 0.17)Miso is not an inferior drug; Conventional uterotonic drugs shouldnot be used in settings where they are entirely unsuitable
Prata 2006 Egypt 2532 Pre-/Post- study comparing current AMTSL practices withoxytocin vs. with misoprostol 600 �g oral
Miso vs. current practices for blood loss ≥500 mL: RR0.30 (0.16,0.56)Where oxytocin and/or ergometrine are not consistently usedduring AMTSL, misoprostol should be considered for inclusion inAMTSL
Zachariah 2006 India 2023 RCT comparing 2 mg intravenous ergometrine, 10 U IMoxytocin, and 400 �g misoprostol
No significant difference between three groups, p = 0.623 for bloodloss >500 mL, p = 0.146 for blood loss >1000 ml, p = 0.234 for theneed of additional oxytocicsOral misoprostol is as effective as conventional oxytocin agents inpreventing PPH and can be recommended for low-resource settings
Alfirevicb 2007 – 22749 Systematic Review of 7 RCTs of misoprostol 600 �g oral orsublingual vs. injectible uterotonics
Miso vs. Placebo in community settings: RR0.59 (0.41, 0.84); inhospital settings: RR1.34 (1.23, 1.74)Miso vs. Injectible Oxytocin: RR1.34 (1.16, 1.55)Conc: Use misoprostol when conventional injectible uterotonicsare not available
Baskett 2007 Canada 622 Double-blind RCT comparing 400 �g oral misoprostol with 5unites of intravenous oxytocin
No difference between two groups, p = 0.98
The routine use of 400 �g of oral misoprostol was no less effectivethan 5 U of intravenous oxytocin in reducing blood loss afterdelivery for blood loss of >1000 mL
Enakpene 2007 Nigeria 864 RCT comparing oral misoprostol 400 �g with 500 �g IMmethylergometrine.
Miso vs. Ergometrine for blood loss >500 mL: RR0.14(p-value < 0.0001)Oral miso was more effective in reducing blood loss during thirdstage of labor than ergometrine
a Study examines fifteen studies: Amant (1999); Bamigboye (1998); Bamigboye (1998); Benchimol (2001); Bugalho (2001); Caliskan (2003); Caliskan (2002); Cook (1999);El-Refaey (2000); Gerstenfeld (2001); Gulmezoglu (2001); Hofmeyr (2001); Hofmeyr (1998); Karkanis (2002); Kundodyiwa (2001); Lam (2004); Ng (2001); Oboro (2003); Ray(2001); Surbeck (1999); Vilmala (2004); Walley (2000).
b Study examines three additional studies than Langanbach: Derman (2006); Hoj (2005); Walraven (2005).
1 h Policy 89 (2009) 131–148
(pcdotrtFob
AaaAtiaaaheai
eboalgsiottaattwm[cfro
Table 7Estimated potential number of maternal deaths averted by interven-tion according to region and country
Region Mean Range [min–max]
Postpartum hemorrhageSub-Saharan Africa 23,225 7443–44427Asia 20,630 7179–39441Niger 1,052 372–2056Rwanda 448 160–829Burkina Faso 400 142–769Cambodia 174 61–342
Safe abortionSub-Saharan Africa 14,903 5663–28395Asia 13,286 4891–24495Niger 677 254–1213Rwanda 287 107–535Burkina Faso 259 91–466Cambodia 111 41–208
Family planningSub-Saharan Africa 7,408 2213–16411Asia 6,547 1947–14199Niger 336 97–746Rwanda 142 42–307Burkina Faso 128 37–280Cambodia 63 27–126
Postpartum hemorrhage, family planning, and safe abortionSub-Saharan Africa 45,577 23401–75424Asia 40,456 18598–68415Niger 2,063 958–3247Rwanda 876 429–1380Burkina Faso 784 396–1239Cambodia 386 265–582
Postpartum hemorrhage and family planningSub-Saharan Africa 30,485 12691–56362Asia 27,158 11543–48331Niger 1,388 592–2474Rwanda 587 223–1023
mt
5
m
42 N. Prata et al. / Healt
AMTSL), which involves the administration of a pro-hylactic uterotonic, delivery of the placenta withontrolled cord traction, and uterine massage afterelivery of the placenta. AMTSL reduces the lengthf the third stage of labor and boosts uterine contrac-ions, decreasing the risk of PPH [38]. However, itequires a skilled attendant, a viable supply of utero-onics, and items needed for injection administration.urthermore, once PPH is established, the treatmentf PPH – through administration of a uterotonic –ecomes extremely important [38].
Oxytocin is the uterotonic drug of choice forMTSL, and should be made available wheneverskilled attendant is present and appropriate stor-
ge conditions exist [39]. Ergometrine, also used forMTSL, is less stable in field conditions than oxy-
ocin, and has increased risk of serious side effectsn hypertensive women [40]. Both drugs need to bedministered parentally. Therefore, without a skilledttendant present to make the injection, ergometrinend oxytocin cannot be chosen for PPH management inome births in developing countries where most deliv-ries occur. Alternative methods like nipple stimulationnd oral ergometrine have shown no advantage overnjectable uterotonics [41,42].
Misoprostol has long been recognized as a safe andfficacious drug for controlling PPH. It can be usedy itself or as a part of AMTSL, in place of oxytocinr ergometrine whenever an SBA is not available andppropriate storage conditions do not exist [43]. Itsow cost, ease of administration (rectal, oral, sublin-ual) and stability in tropical climates make it an idealolution for home births [44]. Table 6 reviews stud-es examining the use of misoprostol in the preventionf PPH in community and hospital settings. Despitehe relative superiority of conventional injectable utero-onics, to date nobody has rejected misoprostol as anppropriate drug where injectable uterotonics are notvailable or cannot be administered [43,45–52]. Fur-hermore, systematic reviews of randomized controlledrials found decreased need for additional uterotonicshen comparing misoprostol to placebo, and recom-end misoprostol to be used in developing countries
45,51,52]. Results from a recent study, for example,
omparing misoprostol with a placebo in home birthsound that misoprostol was associated with a 50%eduction in PPH in a resource-poor setting [53]. Onef the most recent review studies even concludes thatsibo
Burkina Faso 529 223–929Cambodia 259 157–414
isoprostol should not be branded an inferior drug forhe prevention of PPH [51].
.5. Potential maternal deaths averted
According to our simulations, we found that imple-enting FP, SA and PAC, as well as PPH management
ervices in low-resource settings will have a significantmpact on reducing MMR. Table 7 shows the num-er of maternal lives that can be saved by providingur three suggested priority interventions to women in
N. Prata et al. / Health Policy 89 (2009) 131–148 143
l Mater
SRdcSFie
waoilomtM
oeMp
gaw
cogiCtpwr
iIpPi
Fig. 2. Sensitivity analysis: Sub-Saharan Africa Mode
SA, Asia, and four low-resource countries—Niger,wanda, Burkina Faso, and Cambodia. The table alsoepicts the number of lives that could be saved aombination of the interventions were implemented.pecifically, we examined the result of combining (i)P and PPH management services; and (ii) all three
nterventions. We refer to program dissemination tostimate the potential coverage of the interventions.
Implementing PPH programs, including thosehich will reach women who are delivering at home
nd/or without an SBA, will avert the greatest numberf maternal deaths. In SSA, where maternal mortalitys the highest in the world, regional gains in maternalives saved will be the greatest. At the country level,ur examples demonstrate that Niger will benefit theost in terms of potential maternal death averted due
o PPH. This is in comparison to countries with lowerMR, such as Cambodia, where gains will be smaller.The intervention impact and potential for coverage
f SA services has the possibility of averting the great-st number of maternal deaths in countries with highMR, such as Niger. Preventing unsafe abortions and
roviding PAC services will, however, save an even
2ma4
Fig. 3. Sensitivity analysis: Sub-Saharan Africa Model. Mat
nal Deaths Averted due to Family Planning Services.
reater number of maternal lives. Similarly discussedbove, countries with lower MMR, such as Cambodia,ill benefit but gains will not be as large.The majority of gains from family planning program
overage will be in countries with the lowest CPRs. Inur regional examples, SSA demonstrates the largestain in potential maternal lives saved. This is the samen our country-level examples, where Nigeria, with aPR of 4% and 17% in rural and urban areas, respec-
ively, has the potential to benefit the most from familylanning interventions. Cambodia, on the other hand,ith a CPR of 19% and 21% in rural and urban areas
espectively, will again see smaller gains.Combining interventions is important in maximiz-
ng the number of maternal lives that can be averted.n places where safe abortion services are not legallyossible, governments should focus their resources onPH, FP, and PAC. In SSA and Asia, implement-
ng both interventions would save around 30,485 and
7,158 maternal lives, respectively. However, imple-enting all three suggested services (i.e. including safebortion and PAC) would prevent about 45,577 and0,456 maternal deaths in SSA and Asia, respectively.
ernal Deaths Averted due to Safe Abortion Services.
144 N. Prata et al. / Health Policy 89 (2009) 131–148
Fig. 4. Sensitivity analysis: Sub-Saharan Africa Model. Maternal Deaths Averted due to Postpartum Hemorrhage Management.
F aths Av
Atmw
itl
ddtt
FH
ig. 5. Sensitivity analysis: Sub-Saharan Africa Model. Maternal De
s with the implementation of individual interventions,hose regions with higher MMRs will experience a
uch higher reduction in maternal mortality than thoseith lower MMRs.
Sensitivity analyses of models which combined thenterventions demonstrated that PPH management ishe largest contributor to the variance in maternalives saved. In all settings, the variance was mostly
ig
v
ig. 6. Sensitivity analysis: Sub-Saharan Africa Model. Maternal Deathsemorrhage Management Services.
erted due to Postpartum Hemorrhage and Family Planning Services.
ue to changes in the PPH management programissemination. When combining FP and PPH interven-ions, program coverage of PPH services contributedo 61–79% of the variance. When combining all three
nterventions, FP, SA/PAC, and PPH services, PPH pro-ram coverage contributed to 47–59% of the variance.Results from the sensitivity analysis for the indi-idual interventions suggested demonstrate that for
Averted due to Safe Abortion, Family Planning and Postpartum
h Polic
etcuFowvF9ev
6
Nmea(paso2arpewtts
rt1oipi
awdA
fiphSih
aiua(tcwptlatawlwg
incmf(itocsc
ncStm
N. Prata et al. / Healt
ach program, the major contributor to variance washe extent of program coverage. Figs. 2–6 show theontribution to variance in deaths averted for each sim-lation run for SSA. In all models which tested onlyP services, program coverage contributed to 69–71%f the variance (with the exception of Cambodia,here program coverage contributed to 44% and inter-ention impact contributed to 49% of the variance).or SA/PAC models, program coverage contributed to1–93% of the variance. For PPH management mod-ls, program coverage contributed to 87–89% of theariance.
. Discussion/conclusion
Since the 1987 Safe Motherhood Conference inairobi little progress has been made in reducingaternal mortality ratio (MMR). The primary strat-
gy to date has been to promote delivery with a skilledttendant and the provision of emergency obstetric careEmOC) [1]. However, progress is largely stalled and inarts of Africa the ratio of women to trained profession-ls is falling [54]. It is imperative to explore alternativetrategies if the Millennium Development Goal (MDG)f reducing MMR by three-quarters between 1990 and015 is to be met. This paper examines the evidence onvailable interventions aimed at decreasing MMR, andeassesses what is feasible with existing resources inoor settings. We identify interventions with the great-st potential to reduce MMR on a large scale. Facedith human and financial resource constraints, it is
ime for maternal health program planners to selecthose interventions that are both quickly achievable andustainable.
By 2015, there will be 350 million more women ofeproductive age in the developing world than there areoday. If we extrapolate the limited progress made since990 (the baseline for calculating the MDG), then notnly will this MDG go unfulfilled but it could be thatn absolute numbers, more women will have died fromregnancy, abortion, and child birth than in any similarnterval in human history.
From available priority interventions, only a few
re realistic for the 50 million women who deliverithout a skilled birth attendant and the millions whoeliver at home, away from emergency obstetric care.s the developing world faces significant human andqaps
y 89 (2009) 131–148 145
nancial resource shortages, program planners andolicymakers must reassess and reprioritize maternalealth interventions to include access to contraception,A and PAC as well as PPH prevention. These key
nterventions will address a large part of the maternalealth burden.
This paper has some limitations which need to becknowledged. As is the case for any modeling stud-es, our results were dependent on the assumptions wesed. Ours were based mostly on the MBP spreadsheets well as Hill et al.’s estimate for maternal mortality2007). We assumed intervention effectiveness to havehe same range throughout each region. However, suc-essful implementation of the proposed interventionsill depend on the cultural and religious beliefs of eachopulation. Furthermore, the efficacy of the interven-ions assumes the same level of standard and certainevel of quality to produce the results we report. Inddition, individual countries represented here fit a par-icular model of having high MMR, low CPR, as wells low percentage of SBA during delivery. Countriesith similar levels in MMR, CPR, and SBA are more
ikely to show similar gains in comparison to countriesith better levels which are more likely to show smallerains in maternal deaths averted.
Family planning can potentially make a significantmpact on MMR for many reasons including the highumbers of unintended pregnancy and unmet need forontraception, the low cost of modern contraceptiveethods, and the lack in need for highly skilled pro-
essionals to explain and administer most forms of FPincluding condoms, oral contraceptive pills, and Deponjection). For poor communities that cannot affordhe full cost of modern contraceptives, some degreef subsidy is needed. Fortunately, social marketing andommunity-based distribution of contraceptives at sub-idy prices can be used in resource-poor settings to meetommunity needs [55].
Providing access to SA and PAC will also sig-ificantly reduce MMRs experienced by developingountries. Not all countries will be able to provideA services. In such situations, SA should be allowed
o the extent of the law. PAC, however, should beade available regardless, due to the severe conse-
uences that are likely to follow should the incompletebortion go untreated. The misoprostol regiment is sim-le and straight forward and studies have successfullyhown that PAC can even be delegated to community-
1 h Polic
ltpr
tA(octantwTtt
tsteetdrPfbtl[appi[aph
dgsapt
cpotcita
tdettbaaapoppo[
aatw
R
46 N. Prata et al. / Healt
evel providers. The potential for lower level providerso help women with elected abortion and PAC isromising given the current human resources crisis inesource-poor settings.
Misoprostol is an important method for PPH preven-ion among women who live in low-resource settings.t present, assistance by a traditional birth attendant
TBA), a family member, or delivering alone are thenly options for millions of pregnant women espe-ially in Asia and Africa [56]. Misoprostol changeshis equation. Currently there is some disillusionmentbout training TBAs, but the fault in prior efforts mayot have been so much that TBAs cannot be trained, buthat without adequate referral systems current trainingould not significantly reduce MMR. Even illiterateBAs are able to incorporate their knowledge into prac-
ice and provide basic maternal care and they have donehat in relation to misoprostol use [57–61].
There is evidence for the safe use of misoprostol athe community level. Walraven et al. (2005) demon-trated in a randomized, controlled trial in the Gambia,hat misoprostol in the hands of TBAs was safe andffective at home births. Their study, which empow-red TBAs to administer 600 mcg of misoprostol orallyo women after delivery for the prevention of PPH,emonstrated lower blood loss among those womeneceiving misoprostol in comparison to placebo [60].rata et al. (2005) also showed that TBAs can success-ully diagnose PPH using a local garment to measurelood loss after delivery 58 [59] and administer rec-ally 1000 mcg misoprostol when 500 mL of bloodoss was reached (2 kanga) for the treatment of PPH61]. Furthermore, studies done in Indonesia, Nepal,nd Afghanistan showed that self-administration of arophylactic dose of misoprostol, distributed duringregnancy by trained community healthcare workerss a viable option that can produce promising results62]. Thus, in order to significantly reduce MMR, atpolicy level it will be essential that community levelroviders and women themselves, where appropriate,ave access to misoprostol.
Overall, the combined effect of saving mothers fromying from PPH and UA would result in the greatestains in the decline in MMR. However, of the three
uggested interventions, FP and confronting unsafebortion are probably the most contentious. Yet, exam-les of Bangladesh, Tunisia and Iran are a testamenthat policy makers and the public health communityy 89 (2009) 131–148
an together find a prudent way to confront theseroblems [63–65]. The evidence suggests a varietyf service delivery modes and financing mechanismshat combine public and private partnerships to suc-essfully implement the proposed interventions. Thesenclude social marketing; social franchising; accredita-ion; contracting; vouchers systems; and output-basedssistance.
Immediate action is needed. The three interven-ions most likely to significantly reduce MMR allepend more on the attitudes and policies set by gov-rnments, international agencies, and donors than onhe willingness of those living in resource-poor areaso help themselves. Ministries of Health must stille committed to upgrading district health facilities,nd employ every effort to increase access to EmOCs part of basic medical care. However, they mustlso recognize that until there are enough higher levelroviders in rural areas, it is crucial that they focusn training non-physicians and medical technicians torovide EmOC, including caesarian sections. Accom-lishments in Mozambique, Malawi, and what wasnce Zaire are evidence that this is safe and efficacious66–69].
In order to achieve these practical goals, women’sccess to reproductive health services must be vieweds a universal and unequivocal human right. Withouthis fundamental recognition, few genuine advances inomen’s health will ever be accomplished.
eferences
[1] United Nations Population Fund. Maternal mortality update2002, a focus on emergency obstetric care. New York. 2003.
[2] World health Organization, Department of Reproductive Healthand Research. Maternal mortality in 2000: estimates developedby WHO, UNICEF and UNFPA. Geneva. 2004.
[3] Fortney JA, Susanti I, Gadalla S, Saleh S, Feldblum PJ, PottsM. Maternal mortality in Indonesia and Egypt. InternationalJournal of Gynecology & Obstetrics 1988;26(1):21–32.
[4] Greenwood AM, Greenwood BM, Bradley AK, Williams K,Shenton FC, Tulloch S, et al. A prospective survey of the out-come of pregnancy in a rural area of the Gambia. Bulletin ofthe World Health Organization 1987;65(5):635–43.
[5] Khan KS, Wojdyla D, Say L, Gulmezoglu AM, Van Look PF.
WHO analysis of causes of maternal death: a systematic review.The Lancet 2006;367(9516):1066–74.[6] Hill K, AbouZahr C, Wardlaw T. Estimates of maternal mor-tality for 1995. Bulletin of the World Health Organization2001;79(3):182–98.
h Polic
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[Internet]. Available at http://www.maqweb.org/techbriefs/
N. Prata et al. / Healt
[7] Singh S. Hospital admissions resulting from unsafe abor-tion: estimates from 13 developing countries. Lancet 2006;368(9550):1887–92.
[8] Adam T, Lim SS, Mehta S, Bhutta ZA, Fogstad H, Mathai M,et al. Cost effectiveness analysis of strategies for maternal andneonatal health in developing countries. British Medical Journal2005;331(7525):1107.
[9] World Health Organization. The world health report 2006:working together for health 2006. Geneva.
10] World health Organization, Department of Reproductive Healthand Research. Reproductive health strategy to acceleratetowards the attainment of international development goals andtargets. Geneva. 2004.
11] Rosenfield A, Schwartz K. Improving the health of women indeveloping countries: the time is now. Journal of Midwifery andWomen’s Health 2005;50(4):272–4.
12] Paxton A, Bailey P, Lobis S, Fry D. Global patterns in avail-ability in emergency obstetric care. International Journal ofGynecology & Obstetrics 2003;93:300–7.
13] Pearson L, Shoo R. Availability and use of emergency obstet-ric services: Kenya, Rwanda, Southern Sudan, and Uganda.International Journal of Gynecology & Obstetrics 2005;88(2):208–15.
14] World Health Organization. Mother-baby package: implement-ing safe motherhood in countries; practical guide. Geneva.1994.
15] Fortney JA. The importance of family planning inreducing maternal mortality. Studies in Family Planning1987;18:109–14.
16] Guttmacher Institute. Facts on induced abortion worldwide;worldwide incidence and trends. http://www.guttmacher.org/pubs/fb IAW.html. 2007.
17] US Agency for International Development; Center for Popu-lation, Health, and Nutrition. The role of family planning inpreventing abortion. Washington, DC. 1996.
18] Marston C, Cleland J. The effects of contraception on obstet-ric outcomes. Geneva: Department of Reproductive Health andResearch, World Health Organization; 2004.
19] Stanback J, Mbonye A, Bekiita M. Contraceptive injectionsby community health workers in Uganda: a nonrandomizedcommunity trial. Bulletin of the World Health Organization2007;85:768–73.
20] Lande R, Richey C. Expanding services for injectables. Popu-lation Reports. Baltimore. 2006;K(6):1–23.
21] Guttmacher Institute. Abortion: worldwide levels and trends.[Presentation]. October 2007.
22] Diggory P, Peel J, Potts M. Preliminary assessment ofthe 1967 Abortion Act in practice. Lancet 1970;1(7641):287–91.
23] Blumenthal PD, Remsburg RE. A time and cost analysis ofthe management of incomplete abortion with manual vacuumaspiration. International Journal of Gynecology & Obstetrics1994;45(3):261–7.
24] Faundes A, Fiala C, Tang OS, Velasco A. Misoprostol forthe termination of pregnancy up to 12 completed weeks ofpregnancy. International Journal of Gynecology & Obstetrics2007;99(Suppl 2 (December)):S172–7.
[
y 89 (2009) 131–148 147
25] Murchison A, Duff P. Misoprostol for uterine evacuation inpatients with early pregnancy failures. American Journal ofObstetrics and Gynecology 2004;190(5 (May)):1445–6.
26] Coughlin LB, Roberst D, Haddad NG, Long A. Medicalmanagement of first trimester incomplete miscarriage usingmisoprostol. Journal of Obstetrics and Gynaecology 2004;24(1(January)):67–8.
27] Prasad S, Kumar A, Divya A. Early termination of pregnancyby single -dose 800mug misoprostol compared with surgicalevacuation. Fertility and Sterility 2008. Feb 21 [Epub ahead ofprint].
28] Gemzell-Danielsson K, Ho PC, Gomez Ponce de Leon R,Weeks A, Winikoff B. Misoprostol to treat missed abortionin the first trimester. International Journal of Gynecology &Obstetrics 2007;99(Suppl 2 (December)):S183–5 [Epub 2007October 24].
29] Blum J, Winikoff B, Gemzell-Danielsson K, Ho PC, SchiavonR, Weeks A. Treatment of incomplete abortion and miscar-riage with misoprostol. International Journal of Gynecology &Obstetrics 2007;99(Suppl 2 (December)):S186–9 [Epub 2007October 24].
30] Sharma D, Singhal SR, Rani XX. Sublingual misoprostolin management ofmissed abortion in India. Tropical Doctor2007;37(1 (January)):39–40.
31] Herabutya Y, O-Prasertsawat P. Misoprostol in the manage-ment of missed abortion. International Journal of Gynecology& Obstetrics 1997;56(3 (March)):263–6.
32] Moreno-Ruiz NL, Borgatta L, Yanow S, Kapp N, Wiebe ER,Winikoff B. Alternatives to mifepristone for early medicalabortion. International Journal of Gynecology & Obstetrics2007;96:212–8.
33] Vejborg TS, Nilas L, Rorbye C. Medical managementof first trimester miscarriage according to ultrasonographicfindings. Acta Obstetrica et Gynecologia Scandinavica2007;86(5):604–9.
34] Bique C, Usta M, Deobra B, Chong E, Westheimer E, WinikoffB. Comparison of misoprostol and manual vacuum aspira-tion for the treatment of incomplete abortion. InternationalJournal of Gynecology & Obstetrics 2007;98(3 (September)):222–6.
35] Dao B, Blum J, Thieba B, Raghavan S, Ouedraego M,Lankoande J, et al. Is misporsotol a safe, effective, and accept-able alternative to manual vacuum aspiration for postabortioncare? Results from a randomised trial in Burkina Faso,West Africa. British Journal of Obstetrics and Gynaecology2007;114:1368–75.
36] Weeks A, Alia G, Blum J, Winikoff B, Ekwaru P, DurocherJ, et al. A randomized trial of misprostol compared with man-ual vacuum aspiration for incomplete abortion. Obstetrics andGynecology 2005;106(3):540–7.
37] United States Agency for International Development. Prevent-ing Postpartum Hemorrhage. Global Health Technical Briefs.
tb48posthem.shtml.38] Prendiville WJ, Elbourne D, McDonald S. Active versus expec-
tant management in the third stage of labour. Cochrane Databaseof Systematic Review 2000;3. CD000007.
1 h Polic
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
48 N. Prata et al. / Healt
39] World health Organization. Stability of Injectable Oxytocicsin Tropical Climates: Results of Field Surveys and SimulationStudies on Ergometrine, Methylergometrine and Oxytocin. In:WHO (Ed.), Geneva, 1993.
40] McDonald S, Prendiville WJ, Elbourne D. Prophylactic syn-tometrine versus oxytocin for delivery of the placenta. CochraneDatabase of Systematic Review 2000;2. CD000201.
41] De Groot AN, van Roosmalen J, van Dongen PW, Borm GF. Aplacebo-controlled trial of oral ergometrine to reduce postpar-tum hemorrhage. Acta Obstetrica et Gynecologia Scandinavica1996;75(5):464–8.
42] Bullough CH, Msuku RS, Karonde L. Early suckling and post-partum haemorrhage: controlled trial in deliveries by traditionalbirth attendants. Lancet 1989;2(8662):522–5.
43] Prata N, Hamza S, Gypson R, Nada K, Vahidnia F, Potts M.Misoprostol and active management of the third stage of labor.
44] Bradley SE, Prata N, Young-Lin N, Bishai DM. Cost-effectiveness of misporstol to control postpartum hemorrhagein low-resource settings. International Journal of Gyne-cology & Obstetrics 2007;97(1 (April)):52–6 [Epub 2007February 20].
45] Alfirevic Z, Blum J, Walraven G, Weeks A, Winikoff B.Prevention of postpartum hemorrhage with misprostol. Inter-national Journal of Gynecology & Obstetrics 2007;99(Suppl 2(December)):S1198–201 [Epub 2007 October 24].
46] Chandhiok N, Dhillon BS, Datey S, Mathur A, Saxena NC.Oral misoprostol for prevention of postpartum hemorrhage byparamedical workers in India. International Journal of Gyne-cology & Obstetrics 2006;92(2 (February)):170–5 [Epub 2005December 20].
47] Enakpene CA, Morhason-Bello IO, Enakpene EO, ArowojoluAO, Omigbodun AO. Oral misoprostol for the prevention ofprimary post-partum hemorrhage during third stage of labor.Journal of Obstetrics and Gynaecology Research 2007;33(6(December)):810–7.
48] Bhullar A, Carlan SJ, Hamm J, Lamberty N, White L, RichichiK. Buccal misoprostol to decrease blood loss after vagi-nal delivery: a randomized trial. Obstetrics and Gynaecology2004;104(6 (December)):1282–8.
49] Baskett TF, Persad VL, Clough HJ, Young DC. Misoprostolversus oxytocin for the reduction of postpartum blood loss.International Journal of Gynecology & Obstetrics 2007;97(1(April)) [Epub 2007 February 26].
50] Zachariah ES, Naidu M, Seshadri L. Oral misoprostol inthe third stage of labor. International Journal of Gyne-cology & Obstetrics 2006;92(1 (January)) [Epub 2005November 4].
51] Langenbach C. Misoprostol in preventing postpartum hemor-rhage: a meta-analysis. International Journal of Gynecology &Obstetrics 2006;92(1):10–8.
52] Villar J, Gulmezoglu AM, Hofmeyr GJ, Forna F. System-
atic review of randomized controlled trials of misoprostol toprevent postpartum hemorrhage. Obstetrics and Gynaecology2002;100(6):1301–12.53] Derman RJ, Kodkany BS, Goudar SS, Geller SE, Naik VA,Bellad M, et al. Oral misoprostol in preventing postpartum
[
y 89 (2009) 131–148
haemorrhage in resource-poor communities: a randomised con-trolled trial. Lancet 2006;368(9543):1248–53.
54] AbouZahr C, Wardlaw T. Maternal mortality at the end of adecade: signs of progress? Bulletin of the World Health Orga-nization 2001;79(6):561–8.
55] Prata N, Vahidnia F, Potts M, Dries-Daffner I. Revisitingcommunity-based distribution programs: are they still needed?Contraception 2005;72(6):402–7.
56] WHO/UNFPA/UNICEF/World Bank. Reduction of maternalmortality. A joint WHO/UNFPA/UNICEF/World Bank state-ment. 1999.
57] Bij de Vaate A, Coleman R, Manneh H, Walraven G.Knowledge, Attitudes and practices of trained traditionalbirth attendants in the Gambia in the prevention, recogni-tion and management of postpartum haemorrhage. Midwifery2002;18(1):3–11.
58] Prata N, Mbaruku G, Campbell M. Using the kanga to measurepostpartum blood loss. International Journal of Gynecology &Obstetrics 2005;89(1):49–50.
59] Walraven G, Weeks A. The role of (traditional) birth attendantswith midwifery skills in the reduction of maternal mortality.Tropical Medicine & International Health 1999;4(8):527–9.
60] Walraven G, Blum J, Dampha Y, Sowe M, Morison L, WinikoffB, et al. Misoprostol in the management of the third stageof labour in the home delivery setting in rural Gambia: arandomised controlled trial. British Journal of Obstetrics andGynaecology 2005;112(9):1277–83.
61] Prata N, Mbaruku G, Campbell M, Potts M, Vahidnia F.Controlling postpartum hemorrhage after home births inTanzania. International Journal of Gynecology & Obstetrics2005;90(1):51–5.
62] A textbook of postpartum hemorrhage; a comprehensive guideto evaluation, management and surgical intervention. Lan-cashire: Sapiens publishing; 2006. p. 158–9.
63] Hoodfar H, Assadpour S. The politics of population policyin the Islamic Republic of Iran. Studies in Family Planning2000;31(1):19–34.
64] United Nations Population Division. Abortion Policies: AGlobal Review: United Nations, 2002:48–9.
65] Obermeyer CM. Reproductive choice in Islam: gender andstate in Iran and Tunisia. Studies in Family Planning1994;25(1):41–51.
66] White SM, Thorpe RG, Maine D. Emergency obstetricsurgery performed by nurses in Zaire. Lancet 1987;2(8559(September)):612–3.
67] Pereira C, Bugalho A, Bergstrom S, Vaz F, Cotiro M. A Compar-ative Study of caesarean deliveries by assistant medical officersand obstetricians in mozambique. British Journal of Obstetricsand Gynaecology 1996;103(6 (June)):508–12.
68] Duale S. Delegation of responsibility in maternity care inKarawa rural health zone. International Journal of Gynecology
& Obstetrics 1992;38(Suppl (June)):S33–5.69] Fenton PM, Whitty CJ, Reynolds F. Caesarean section inMalawi: prospective study of early maternal and perinatalmortality. British Medical Journal 2003;327(7415 (Septem-ber)):587.