Epidemiology of Maternal Mortality in Malawi - 2 Edition

Chapter 5 The epidemiology of maternal mortality 2
nd
edition 2013

Page 1 of 53
Epidemiology of Maternal Mortality in Malawi 2
nd

edition
Cameron Bowie & Eveline Geubbels
Department of Community Health, College of Medicine, Malawi
Table of Contents
2.1 Measuring ........................................................................................................................ 3
2.1.1 Measuring maternal mortality ................................................................................. 3
2.1.2 Measuring the process of emergency obstetric care ............................................... 5
2.2 Global maternal mortality ................................................................................................ 5
2.2.1 Distribution .............................................................................................................. 5
2.2.2 Historical trends ....................................................................................................... 6
2.3 Maternal mortality in Malawi .......................................................................................... 7
2.3.1 The published studies .............................................................................................. 7
2.3.2 Trends in maternal mortality ................................................................................... 8
3.1 Global causes of maternal mortality ................................................................................ 9
3.2 Causes of maternal mortality in Malawi........................................................................ 10
3.2.1 Haemorrhage ......................................................................................................... 11
3.2.2 Sepsis ..................................................................................................................... 12
3.2.3 Obstructed labour and ruptured uterus (RU) ......................................................... 12
3.2.4 Abortion ................................................................................................................. 12
3.2.5 Hypertensive disorders of pregnancy (HDP) ......................................................... 13
3.2.6 Anaemia ................................................................................................................. 13
3.2.7 HIV / AIDS............................................................................................................ 13
3.3 Global determinants of maternal mortality .................................................................... 14
3.4 Determinants of maternal mortality in Malawi ............................................................. 15
3.4.1 Biological determinants ......................................................................................... 15
3.4.2 Behavioural determinants ...................................................................................... 16
3.4.3 Socio-cultural and socio-economic determinants .................................................. 21
4.1 Mortality & life expectancy ........................................................................................... 30
4.2 Infant and child morbidity and mortality in households affected by the disease .......... 31
4.3 Economic & social ........................................................................................................ 31
4.4 Health services .............................................................................................................. 31
5.1 Biological interventions ................................................................................................. 33
5.1.1 Family planning ..................................................................................................... 33
5.1.2 Iron and/or folate supplementation ........................................................................ 33
5.1.3 Presumptive treatment of malaria .......................................................................... 34
5.1.5 External cephalic version (ECV) for breech presentation at term ......................... 34
5.1.6 Active management in the third stage of labour .................................................... 35
5.1.7 Antibiotic prophylaxis for caesarean section ......................................................... 35
1 Definition ................................................................................................................................ 2
2. Distribution of maternal mortality ...................................................................................... 3
3. Causes and determinants of maternal mortality ................................................................ 9
4. Impact of maternal mortality ............................................................................................. 30
4. Effective interventions ......................................................................................................... 32
nd
edition 2013

Page 2 of 53
5.1.8 Umbilical vein injection for management of retained placenta ............................. 35
5.1.9 Vacuum aspiration versus sharp curettage for incomplete abortion ...................... 36
5.1.10 Choice of anticonvulsant for eclampsia ................................................................. 36
5.1.11 Antibiotics for preterm premature rupture of membranes ..................................... 36
5.2 Behavioural interventions .............................................................................................. 37
5.2.1 IEC for family planning ......................................................................................... 37
5.2.2 IEC to reduce delays in decision making and reaching a health facility ............... 37
5.3 Socio-cultural and socio-economic interventions .......................................................... 37
5.3.1 Improvement of socio-economic status of women ................................................ 37
5.3.2 Provision of safe, legal abortion services [136] .................................................... 38
5.3.3 Improving transport and referral systems .............................................................. 38
5.3.4 Improve availability of maternity waiting homes .................................................. 38
5.3.5 Training of traditional birth attendants .................................................................. 39
5.3.6 Improving access to EMOC facilities .................................................................... 39

1 Definition
In the 10th revision of the International Classification of Diseases (ICD-10), the WHO
defines a maternal death as:
A maternal death is the death of a woman while pregnant or within 42 days of
termination of pregnancy, irrespective of the duration and the site of the
pregnancy, from any cause related to or aggravated by the pregnancy or its
management, but not from accidental or incidental causes.[1]
A late maternal death is the death of a woman from direct or indirect obstetric
causes more than 42 days but less than one year after termination of pregnancy.
A death occurring during pregnancy, childbirth and puerperium is the death of a
woman while pregnant or within 42 days of termination of pregnancy,
irrespective of the cause of death (obstetric and non obstetric).
Maternal deaths are subdivided into two groups:-
1. Direct obstetric deaths: those resulting from obstetric complications of the
pregnant state (pregnancy, labour and puerperium), from interventions, omissions,
incorrect treatment, or from a chain of events resulting from any of the above.
2. Indirect obstetric deaths: those resulting from previous existing disease or
disease that developed during pregnancy and which were not due to direct
obstetric causes, but which were aggravated by physiologic effects of pregnancy.
In practice, it is often impossible to determine the exact cause of death of a pregnant or
recently pregnant woman particularly when deaths occur outside health facilities. For this
reason, WHO and others working in this field often use a broader definition, namely
pregnancy-related death. This dispenses with the need to determine cause of death and
classifies as pregnancy-related all deaths of women of reproductive age in which the
woman was pregnant at the time of death or had recently been so. This is more akin to the
definition of infant death, which is defined solely in terms of the timing of the death. For
all practical purposes, the difference between the two measures is minimal because only a
very small proportion of deaths of pregnant or recently pregnant women are unrelated in
7 References ............................................................................................................................ 41
nd
edition 2013

Page 3 of 53
some way to the pregnancy itself. In other words, the proportion of all deaths among
these women that are incidental is very small in almost all settings.
Pregnancy-related death - the death of a woman while pregnant or within 42 days of
termination of pregnancy, irrespective of the cause of death.
This is the definition used in the Malawi demographic health surveys. It is not the
definition used by the UN in calculating maternal mortality for the Millennium
Development Target 6 of Goal 5 - reduce by three quarters, between 1990 and 2015, the
maternal mortality ratio. The UN tries to estimate the proportion of pregnancy-related
deaths due to AIDS and other causes not due to either direct or indirect obstetric causes
[2]. It is likely that pregnancy is incidental at the time of some deaths due to AIDS. The
UN estimates would be expected to be lower that DHS derived estimates. However, the
sisterhood method of data collection in the DHSs assumes that the family knows if their
sister is pregnant at the time of death. It is likely therefore that pregnancy related deaths
are under-reported. The under-reporting of pregnancy-related deaths may be
counterbalanced by the over-reporting of AIDS deaths unrelated to pregnancy. This
seems to be the case in neighbouring countries with high levels of AIDS and Malawi is
likely to be the same (see Figure 1). An important feature of Figure 1 is the huge
uncertainty interval associated with the MMR estimates.
Figure 1 - UN estimates compared to national estimates using DHS data for Kenya
and Zimbabwe [2]

Kenya Zimbabwe

2. Distribution of maternal mortality
2.1 Measuring
2.1.1 Measuring maternal mortality
Maternal mortality incidence is usually expressed as the number of maternal deaths per
100,000 live births (the maternal mortality ratio), as the number of maternal deaths per
year per 10,000 women of reproductive age (the maternal mortality rate) or as the
lifetime chance of dying from pregnancy (the lifetime risk of maternal death).
nd
edition 2013

Page 4 of 53
The most widely used measure is the maternal mortality ratio, which only depends on the
risk of dying once a woman is pregnant, i.e. the risk of developing a complication
multiplied by the risk of dying from that complication. The maternal mortality rate also
takes into account the annual probability of becoming pregnant for women of
reproductive age. The lifetime risk is the most comprehensive measure as it incorporates
the annual probability to become pregnant, the length of the reproductive period and the
risk of dying a maternal death once pregnant. It is thus a cumulative incidence.
There are five possible sources of information on maternal deaths:
1. Vital registration systems, or death notification systems
2. Hospital-based surveys, including health management information statistics
(HMIS)
3. Population-based surveys, including the sisterhood method
4. Community-based continuous surveillance systems
5. Reproductive Age Mortality Studies (RAMOS)
1. Vital registration systems are seldom available on a wide scale in developing countries.
Even where they exist, they tend to under-report death or provide no information on
cause of death or pregnancy status, which makes it impossible to classify a death as
maternal.
2. Hospital-based surveys or HMIS statistics use data about patients who deliver in a
health facility. For Malawi, this means that 27% of births are missed, because they occur
outside a health facility [2]. In addition, in-hospital deliveries usually concern a
selection of high-risk women or emergency admissions. Together, these lead to a
considerable but unknown bias in the estimate. They are very useful however to
investigate the factors contributing to in-hospital maternal deaths.
3. Population-based surveys are less biased, but because a maternal death is a rare event,
they require very large samples (often > 50,000 births), which makes them very costly
or when a large sample-size is not feasible, they produce imprecise estimates. A way to
overcome this problem is to use the sisterhood-method [3]. Because one respondent
provides information about several other women, the sample size can be reduced to less
than 4,000 households. In the indirect sisterhood method adult female respondents are
asked four simple questions about how many sisters reached adulthood, how many have
died, and whether they were pregnant around the time of death. The overall estimate
relates to a point in time around 10-12 years prior to the survey. The direct sisterhood
method asks more elaborate questions about age at death and time of death and therefore
allows estimating maternal mortality for a narrower time period. The most reliable
estimate is around seven years before the survey. This latter one is used in the Malawi
DHS.
4. Community-based surveillance systems (i.e. longitudinal studies) are also costly, but
have the ability to provide current estimates, and insight into the determinants of maternal
death.
5. RAMOS surveys assess the extent and causes of maternal mortality by identifying and
investigating the causes of death of all women of reproductive age, using a variety of
sources of information on maternal deaths, e.g. civil registers, health facilities,
community leaders, schoolchildren, religious authorities, undertakers, cemetery officials
nd
edition 2013

Page 5 of 53
etc. These are an economical way of measuring maternal death provided it is possible to
trace deaths in women of reproductive age.
In the 2000s the IMMPACT programme (Initiative for Maternal Mortality Programme
Assessment), based at the University of Aberdeen, worked on enhanced methods and
tools for measuring maternal mortality and its underlying processes and on generating
evidence on effective and cost-effective strategies for reducing maternal mortality [4].
IMMPACT collaborated with seven developing countries, of which Malawi was one.
Since then three other initiatives led by UNFPA have been funded to improve the
evidence on maternal mortality [57].
2.1.2 Measuring the process of emergency obstetric care
Because of the difficulties in timely and accurate measurement of maternal mortality, it is
not a suitable indicator to evaluate the effectiveness of obstetric interventions. WHO
suggests the use of the following 6 process indicators [8]:
1. Number of facilities: at least 4 basic emergency obstetric care (EmOC) facilities
i
and
1 comprehensive EmOC facility
ii
per 500,000 population
2. Geographical distribution: proper distribution so that 4 basic EmOC facilities and 1
comprehensive EmOC facility serve a catchment area of 500,000 population
3. Proportion of births in EmOC facilities
iii
: at least 15% of all births in the community
4. Met need for EmOC services: all women with emergency obstetric complications are
treated in an EmOC facility
iv

5. Caesarean Section as percentage of all births: between 5 and 15%
6. Case fatality rate: proportion of women with obstetric complications admitted to a
facility that dies is less than 1%.
Malawi was one of the first countries to evaluate their use in practice, i.e. in the Safe
motherhood Project [911].

Since then three EMOC surveys have been undertaken, the
first in 2003 [12] the second in 2006 [13] with a summary published in 2006 [14] and a
third in 2010 [15]. As part of a five country study three districts in the southern region
were also surveyed in 2010 [16].
2.2 Global maternal mortality
2.2.1 Distribution
In 1990 more than one woman died every minute from complications of pregnancy and
childbirth somewhere in the world, i.e. 546,000 women annually [2]. Global maternal
mortality has declined to 358,000 in 2008. Similarly, the global maternal mortality ratio
(MMR) declined from 397 maternal deaths per 100,000 live births in 1990 to 263 in

i
A facility that can provide treatment with antibiotics, oxytocics and anticonvulsants, manual removal of
placenta, removal of retained products and assisted vaginal delivery. In Malawi, health centers and rural
hospitals should be able to offer these services.
ii
A facility that provides the care of a basic EmOC plus obstetric surgery, anesthesia and blood transfusion.
In Malawi, district, mission and central hospitals should be able to offer these services.
iii
Minimum acceptable level set locally
iv
Expected to be 15% of live births
nd
edition 2013

Page 6 of 53
2008, an average annual decline of 2.3%. Less than one percent of these deaths occur in
developed countries, demonstrating that they could be avoided if resources and services
were available. The lifetime risk for women in Malawi is 1 in 36 and in Africa is 1 in 39,
as compared to 1 in 3,800 in developed countries. This makes maternal mortality one of
the health indicators with the widest disparity between developed and developing
countries [17].
Figure 2 - Map with countries by category according to their maternal mortality
ratio (MMR, death per 100 000 live births), 2010

2.2.2 Historical trends
Developing countries can learn from the historical trends of Europe and America in the 1850-
1950s and the reasons for the huge decline in maternal mortality over this period (Figure 3).
Whilst child mortality fell with improved nutrition and sanitation and a reduction in poverty from
about 1900, maternal mortality did not. It began a rapid decline in 1930s, which was the time
when aseptic techniques were introduced and sulphonamides became available to treat puerperal
sepsis (Figure 4). The previous high mortality was due to unnecessary interference by physicians,
such a forceps for each delivery using non-aseptic techniques. Once midwives trained in aseptic
techniques were delivering mothers in their homes the rates fell dramatically [18].
nd
edition 2013

Page 7 of 53
Figure 3 Trend in maternal mortality England and
Wales, 1880 1980 [18]

Figure 4 - Annual maternal mortality rates
attributable to puerperal fever and to all other causes
(logarithmic scale), in England and Wales, 19201945.

2.3 Maternal mortality in Malawi
2.3.1 The published studies
In Malawi, the number of maternal deaths has been estimated using hospital-based
surveys and population-based surveys. The results of these studies, sorted by setting and
year of the study, are presented in Table 1.
In addition, large community-based surveillance systems are currently being
implemented by the Karonga Prevention Study of the London School of Hygiene and
Tropical Medicine in the South of Karonga District and by the Centre for Reproductive
Health of the College of Medicine in the Lungwena area in Mangochi District.
Estimates from hospital-based studies, concerning the period 1977 to 1990 range from 32
to 1027 maternal deaths per 100,000 live births. More recent institutional studies refrain
from offering maternal mortality rates.
Table 1 - Maternal mortality rate, population based and hospital-based studies in
Malawi
Reference Setting Method Year Maternal deaths MMR
Population-based surveys
Chiphangwi,
1992 [19]
Community, Thyolo Indirect sisterhood method 1989, but refers to
approx. 1978
150 409
McDermott, 1996
[20]
Four ANC in Mangochi district Prospective population based
survey among ANC attendees
September 1987
July 1989
15 398
Malawi DHS,
1992 [21]
Random sample within 6 region
/ urban-rural strata
Direct sisterhood method 1992, but refers to
1986-92
71 620
Malawi DHS,
2002 [22]
Random sample, but with over
sampling for 11 districts
1994-2000
344 1120
Malawi DHS
2006 [23]
Random sample, but with over
sampling for 10 districts
1998-2004
240 984
Multiple indicator
cluster survey
(MICS) 2006
2 stage random sample Direct sisterhood method 2006 but refers to
2001-2006
469 861
Beltman 2011
[24]
Random sample Thyolo District,
Southern Region
Direct sisterhood method 2006 84 558 (260-820)
Malawi DHS
2012 [25]
2 stage random sample Direct sisterhood method 2010, but refers to
2004-2010
331 675
nd
edition 2013

Page 8 of 53
2.3.2 Trends in maternal mortality
The MMR estimated from community surveys in the eighties to early nineties varied
from 398 to 620. For the late nineties and early twenties the MDHS estimates of the
MMR are available, augmented by the MICS study and a small sample survey in Thyolo.
A recent analysis of trends of the Millennium Development Goal (MDG) health targets
for the Health Sector Strategic Plan [39] finds a reduction in MMR since 2000, but less
than needed to meet the MDG target [40]. The maternal mortality rate decreased from
984 per 100,000 live births in 2004 to 675 per 100,000 in 2010, with a projected rate of
435 in 2015 against a MDG target of 155 (Figure 5). The 2010 MMR estimate used by
the UN for Malawi is 460 (290-710) per 100,000 live births and a lifetime risk of
maternal death of 1:36 [17].
The maternal mortality ratio found in the DHS2010 makes it more apparent than previous
ratio findings that the 1992 result was aberrant and should be discounted when assessing
the MMR trend. There are two consequences. Firstly, the MDG target which is partly
based on the 1992 figure is probably too low and should have been set higher. Secondly
the trend since 2000 is linear and the projection of 435 maternal deaths per 100,000 live
births is probably realistic. In retrospect, the MDG target was set far too low.
However other reproductive health targets (skilled delivery attendants and contraceptive
prevalence) are also unlikely to be met. The road map has been implemented only
recently and the results may take time to materialise [41].
Community-based surveys and audits
Hofman 2005 [26] TA Nankumba, Mangochi
district
Community based maternal
death reviews
1999-2001 43 Not specified
Hospital-based surveys and audits
Bullough, 1981
[27]
All health facilities, Central
Region
Retrospective hospital survey 1977 118 263
Knowles, 1988
[28]
Ekwendeni Hospital Hospital survey 1976-1985 30 344
WHO, 1985 [29] Six district hospitals,
countrywide
? 1983 34 269
Keller, 1987 [30] Kamuzu Central Hospital,
Lilongwe
Knowles, 1989 Ekwendeni Hospital Hospital survey 1986-1988 1 32
Driessen, 1990
[31]
Two central, 5 district and 5
mission hospitals, country-wide
Kempf, 1990 [32] Mulanje Mission Hospital Hospital survey 1989 21 411
Wiebenga, 1992
[33]
Queen Elizabeth Central
Hospital, Blantyre
Retrospective (1989) /
prospective (1990) hospital
survey
1989-1990 151 529
Sangala, 1992
[34]
Kamuzu Central Hospital,
Lilongwe
Retrospective hospital survey 1990 74 Not specified
Lema, 2005 [35] Queen Elizabeth Central
Hospital, Blantyre
Retrospective hospital survey 1999-2000 204 1027
Ratsma, 2005
[36]
9 hospitals and 5 health centres
in Southern Region
Retrospective survey 2001 312 Not specified
Kongnyuy, 2009
[37]
9 hospitals in 3 districts in
Central Region
Retrospective hospital survey 2007 43 Not specified
Van den Akker
2011 [38]
Thyolo District Hospital,
Southern Region
Retrospective hospital audit Sept 2007 Sept
2009
46 Not specified
nd
edition 2013

Page 9 of 53
Figure 5 - Maternal mortality in Malawi - trends and projection

Summary - incidence
The current best estimate of maternal mortality for 2010 is 460 maternal deaths per
100,000 live births. An estimated total fertility rate of 6.0 means the lifetime risk of
maternal death is 1 in 36. The MDG target will not be met but was probably set
incorrectly high in the first place.
3. Causes and determinants of maternal mortality
3.1 Global causes of maternal mortality
Across the globe the causes of maternal deaths are strikingly similar, although their relative
importance varies between countries (Figure 6). Eighty percent of deaths are due to direct
causes, i.e. obstetric complications, interventions, omissions or incorrect treatment. Out of all
maternal deaths over sixty percent occur post-partum especially in the first week after delivery.
Figure 6 Causes of maternal deaths, globally in 2005 [42]
0
200
400
600
800
1000
1200
1400
1600
1990 1995 2000 2005 2010 2015 2020
M
a
t
e
r
n
a
l

m
o
r
t
a
l
i
t
y

p
e
r

1
0
0
,
0
0
0

l
i
v
e

b
i
r
t
h
s
Year of survey
MDG
target
155
projection
435
SWAp 1
2004 - 2011
nd
edition 2013

Page 10 of 53

3.2 Causes of maternal mortality in Malawi
The same causes of maternal deaths that are found globally are seen in Malawi.
Table 2 presents the results from three hospital-based studies in the late eighties and early
nineties [31,33,34], one hospital based study from 2001 [36], one community-based study
ccovering 1998-2001 [26] and two hospital based studies from the 2000s [38,43]. One
study was not included because information about cause of death was available for only 6
patients [20].
A maternal death is often characterised by a chain of events leading to death. For
example, a woman might undergo a caesarean section for obstructed labour and die of
postoperative sepsis. The cause of death is then the first event, in this case obstructed
labour. In one of the studies [34], this classification method seemed not to have been
practiced throughout. Its results were therefore retabulated from the reported data (Table
2).
Table 2 Causes of maternal mortality in Malawi
Reference Direct deaths (% of all deaths) Indirect deaths (% of all deaths) Unknown
or
fortuitous
(% of all
deaths)
Haemorr-
hage.
Sepsis Ruptured
Uterus &
Obstructed.
Labour.
Abortion Other Anaemia Meningitis. AIDS Other
Wiebenga*, QECH,
1989-1990 [33]
4 % 13 % 7 % 17 % 5 % 5 % 9 % 4 % 24 % 15 %
Driessen, 12
hospitals, 1989 [31]
10 % 12 % 14 % 18 % 6 % 8 % 3 % 2 % 17 % 10 %
Sangala*, KCH,
1990 [34]
14 % 24 % 15 % 18 % 11 % 8 % 3 % - 8 % -
Ratsma*, 18
hospitals, 2001 [36]
11 % 20 % 15 % 6 % 13 % 9 % 7 % 9 % 9 % 3 %
Hofman,
community, 1998-
2001 [26]
30 % 5 % 30 % 14 % 7 % 7 % 5 % - - 2 %
nd
edition 2013

Page 11 of 53
Reference Direct deaths (% of all deaths) Indirect deaths (% of all deaths) Unknown
or
fortuitous
(% of all
deaths)
Haemorr-
hage.
Sepsis Ruptured
Uterus &
Obstructed.
Labour.
Abortion Other Anaemia Meningitis. AIDS Other
Farish*, community,
2003
33% 7% 7% - 7% 13% 7% 7% 13% 7%
Lema, QECH,
1999-2000 [35]
11 % 29 % - 24 % 6 % - - - 20%
Kongnyuy, 9
hospitals 2009 [37]
26 % 16 % 7 % 7 % 9 % 7 % 0 % 16 % 12 % 0 %
Van den Akker,
Thyolo hospital
2011 [38]
22 % 17 % 11 % - 9 % 0 % 26 % 7 % 9%
Vink, Nkhoma
Hospital, 2012 [44]
9 % 2 % 7 % 5 % 17 % 12 % 22 % 3% 21 % 2 %
* Percentages do not add up to 100% due to rounding
The three most important causes of death in the three earlier hospital studies were sepsis,
complications of abortion and obstructed labour, sometimes resulting in ruptured uterus.
In the later hospital study, the relative importance of deaths from abortion appears to
have declined, possibly because of higher uptake of family planning methods and
consequently fewer unwanted pregnancies [25].
In the community based study by Hofman, which tracked both deaths that occurred in the
community and in-hospital, haemorrhage and ruptured uterus were a much more
important cause of death. This is a reflection of the acute nature of these complications.
Often acute care (within 2-12 hours for haemorrhage and within one day for ruptured
uterus) is not available in time and thus a higher proportion of these maternal deaths
occur in the community. The Farish study identified deaths through interviews with
community members only and may have underreported those maternal deaths that
occurred in the hospital, e.g. through RU and obstructed labour. As for the indirect
causes of death, no major changes seemed to have occurred within the last ten years, with
the exception of AIDS that seems to be diagnosed more often. However, the real
contribution of AIDS to the incidence of maternal mortality cannot be estimated from
these studies. Already in 1990 Wiebenga commented that in more than half of all
puerperal sepsis cases HIV was thought to have contributed to death. Also more than
half of meningitis cases and all pulmonary TB and septicaemia cases were thought to be
HIV-associated [33]. Indeed, there is circumstantial evidence from the 1992 and 2000
demographic and health surveys [22,23] that the HIV epidemic has contributed
substantially to the rise in maternal mortality in the 1990s [17]. It has been suggested
that this is not only a causal effect but that the epidemic also leads to decreased quality of
care resulting from crowded health facilities due to HIV-related illnesses and loss of staff,
see HIV/AIDS chapter. Sepsis is becoming the leading cause of death where HIV is
prevalent. The latest estimate for 2010 from the UN for Malawi is that 29% of maternal
deaths are AIDS related indirect maternal deaths [17].
3.2.1 Haemorrhage
Patients with haemorrhage should be treated as emergencies in health facilities. Both
types, ante partum haemorrhage (APH) and postpartum haemorrhage (PPH), require
urgent intervention because the estimated average time to death is short (i.e. 12 hours for
APH and 2 hours for PPH). Therefore, factors that delay timely access to appropriate care
nd
edition 2013

Page 12 of 53
often contribute to death, e.g. limited knowledge of danger signs, delayed decision
making, lack of transport and money for fees and delay in receiving care at the health
facility. The patient needs to be resuscitated and the bleeding stopped, if necessary
operatively. Lack of blood is often an important contributing factor to deaths from
haemorrhage, especially in small or anaemic women, who tolerate blood loss less well.
3.2.1.1 Ante partum haemorrhage (APH)
APH is defined as bleeding from the genital tract after 28 weeks gestation and before the
birth of the baby. Causes include placenta praevia (when the placenta covers all or part of
the cervical opening), abruptio placentae (when the placenta loosens from the uterus wall
before the baby is born) and ruptured uterus. Data about the occurrence of these separate
causes for APH in Malawi were not found.
3.2.1.2 Postpartum Haemorrhage
The most common site of bleeding in PPH is the placental bed, because the uterus does
not contract well. This can happen because of too rapid separation of the placenta (e.g.
by pulling on the umbilical cord), a retained placenta (a placenta or placental part that is
not delivered within two hours after delivery of the infant), high parity, prolonged labour,
twin delivery, polyhydramnios, anaesthesia or even a full bladder. PPH might also
originate from lacerations that developed during childbirth, e.g. vulval or vaginal tears,
from breakdown of the uterine wound after caesarean section or ruptured uterus or from
sloughing of dead tissue following obstructed labour. It is not known for Malawi what is
the contribution of these separate causes for PPH.
3.2.2 Sepsis
Puerperal sepsis can be caused by different micro-organisms, e.g. sexually transmitted
micro-organisms, large bowel bacteria or skin bacteria. Risk factors for sepsis include
poor hygiene during delivery (see paragraph 3.5.2), manipulations high in the birth canal,
premature rupture of membranes, delivery through caesarean section and the presence of
dead tissue in the birth canal after delivery.
3.2.3 Obstructed labour and ruptured uterus (RU)
Obstructed labour may lead to ruptured uterus. Phillips, in an overview of 194 women
treated for a ruptured uterus in Mulanje Mission Hospital between 1974 and 1982, found
a mortality rate of 10.3%; this figure was due largely to ruptured uterus, primigravidity,
and being in shock when admitted [45].
3.2.4 Abortion
The main risks associated with induced or spontaneous incomplete abortion are the same
as those of delivery, i.e. infection and bleeding. However, for induced abortion, lack of
adequate family planning can be seen as a risk factor, because without unwanted
pregnancies, no induced abortions would be carried out.
A study in Kamuzu Central Hospital reported that admissions for abortions constituted
40% of all admissions to the gynaecology ward. Seventy-one percent occurred in first
trimester. Incomplete abortions made up 85% of total and of these women, 38% were
septic on arrival. One third needed blood transfusion. Thirty-three cases (5.5%) were
nd
edition 2013

Page 13 of 53
obvious induced abortions, 30 had sticks introduced in the vagina by traditional healers
and three admitted having taken home medicine. This is most likely an underestimation.
Of the women with induced abortions, 72% were schoolgirls. The case fatality rate
among women with obviously induced abortions was 9% and the overall case fatality rate
was 0.5% [46]. In another study, performed in Queen Elizabeth Central Hospital in
Blantyre, Kamuzu Central Hospital in Lilongwe, Ekwendeni Mission Hospital in Mzuzu
and Mangochi District Hospital, a total of 1325 incomplete abortions were recorded. No
distinction was made between spontaneous and induced abortions. Again, two thirds
occurred in the first trimester, the median age of women was 25.5 years (28% was
younger than 20 years), median parity was 2 and a large majority of women were
married. The facility based abortion case fatality rate was 2%. The most frequent
occurring presenting complications according to service providers were localised
infection, haemorrhage and sepsis [47].
A study in 2009 identified the characteristics of 2028 women who had sought post-
abortion care in health facilities in Malawi over a 30 day period in 2009 [48]. 21% were
adolescents and 29% were aged between 20-24 years. More than half of adolescents and
almost 80% of young adults were married. Less than 5% of adolescents and 22.5% of
young adults reported using contraception when they became pregnant.
3.2.5 Hypertensive disorders of pregnancy (HDP)
HDP include pregnancy-induced hypertension (hypertension that occurs after 20 weeks
gestation in a normally normotensive woman), pre-eclampsia (pregnancy-induced
hypertension with proteinuria and possibly oedema, headache and visual distortions) and
eclampsia (pregnancy-induced hypertension with convulsions) [49]. The risk of HPD is
increased for primiparous women, women who have suffered proteinuric pre-eclampsia
in a previous pregnancy, in women with a positive family history, obesity, multiple
gestation or excessive weight gain in pregnancy. However these factors in combination
cannot adequately predict which individual woman will develop HDP. The wide variety
in case fatality rates between countries (e.g. 7-25% in Africa and 1.4% in Sweden)
implies that differences in care influence outcome. Confidential enquiries suggest that
women who die of HPD have usually received substandard care, including failure to
diagnose the condition until late in pregnancy. Measurement of a rise in arterial blood
pressure is the most sensitive screening test for diagnosing HDP. Because baseline initial
blood pressure measurements are often lacking in Malawi, the combination of proteinuria
and repeated diastolic pressure > 90 mm Hg, or >110 mm Hg at a single reading is
considered to represent mild and severe pre-eclampsia respectively. Women with
convulsions are all treated as having eclampsia, unless another diagnosis is confirmed.
No information was found on the incidence of HDP in Malawi.
3.2.6 Anaemia
See chapter on Anaemia in Pregnancy for further information.
3.2.7 HIV / AIDS
Complications of both early and late pregnancy have been reported to occur more often in
women infected with HIV, e.g. spontaneous abortion, ectopic pregnancy, preterm labour,
preterm rupture of membranes, abruptio placentae and postpartum infectious
nd
edition 2013

Page 14 of 53
complications [50]. Infectious morbidity after caesarean section in HIV positive women
was increased in a Rwandan study, with associated higher maternal mortality [51]. See
chapter on HIV/AIDS chapter for more information on the distribution of HIV in
pregnant women.
Summary - Causes
Distribution of causes of maternal death in Malawi are similar to worldwide
distribution
The majority of maternal deaths have direct causes and occur post-partum
The most common causes are sepsis, haemorrhage and obstructed labor/ruptured
uterus
Hospital-based studies underestimate complications that rapidly lead to death (e.g.
haemorrage)
3.3 Global determinants of maternal mortality
The MMRs presented in paragraph 2.3 express the risk of dying a maternal death once a
woman is pregnant. Her lifetime risk of maternal death will also depend on the number
of times she becomes pregnant.
Basically, three determinants together constitute the risk of maternal mortality:
1. The likelihood of a pregnancy occurring
2. The likelihood of a complication arising
3. The successful management of a complication
Globally, women in Africa have the highest life time risk of maternal death because high
mortality rates are coupled with high fertility.
Maternal death is nearly always multi-causal. Even where only one clinical cause of
death can be identified, often several behavioural or socio-cultural/ economical factors
have contributed to that death.
Ad A. Likelihood of pregnancy occurring
The likelihood of pregnancy depends on the fertility pattern in a country or region. This
pattern is determined by the fertility preferences and the unmet need for family planning.
Both of these depend on womens educational status and rural / urban location.
Obviously the difference between wanted and actual number of children is largest for
those women who have the greatest unmet need for fertility control.
Ad B. Likelihood of a complication arising
Although every pregnant woman runs the risk of developing a complication during
pregnancy, delivery or puerperium, complications are more common among teenagers,
older women, women in their first pregnancy, women in their fourth or higher pregnancy,
women with short birth intervals, short women and women who had a complication
during a prior pregnancy.
Most of these factors are mere proxies for underlying problems that increase the risk of
maternal death.
Whereas these risk factors and proxies are useful in establishing which groups of women
are at higher risk, they fail to correctly predict the risk for individual women. Although
women in the high risk group have relatively more complications, in absolute numbers,
nd
edition 2013

Page 15 of 53
more complications occur in the low-risk group. Also, many women who have one or
more risk factors never develop a complication and conversely, of women who do
develop a complication many had not been identified as being at high risk. Furthermore,
most complications cannot be prevented. This so-called high-risk approach has therefore
been abandoned. Whereas some risk screening is still useful (e.g. asking about
complications in previous pregnancy), presently the general consensus is that every
pregnancy faces risk and thus availability of and access to essential obstetric care is
paramount.
Ad C. Successful management of a complication.
Most obstetric complications cannot be predicted nor prevented, but nearly all can be
successfully treated. Availability, accessibility and quality of care are therefore crucial
determinants for the prevention of maternal mortality. To describe the factors associated
with acquiring timely care, the conceptual framework of the Three Phases of Delay is
useful [52]. These are:
1. Delay in the decision to seek care
2. Delay in the arrival at a health facility
3. Delay in the provision of adequate care
Ad 1. Factors that delay the decision to seek care on the part of the individual, the
family or both include: the centre of decision making (individual/spouse/relative/family);
the status of women; illness characteristics; distance from the health facility; financial
and opportunity costs; previous experience with the health care system; perceived quality
of care.
Ad 2. Delay in reaching the health facility depends on distribution of facilities, travel
time, availability and cost of transportation and condition of roads.

Ad 3. Receiving adequate care at the health facility can be delayed by inadequacy of the
referral system, shortage of supplies, equipment and trained personnel and competence of
the available personnel.
3.4 Determinants of maternal mortality in Malawi
3.4.1 Biological determinants
3.4.1.1 Cephalopelvic disproportion (CPD) and malpresentations
CPD and malpresentations are often the underlying cause of prolonged and obstructed
labour, which may lead to ruptured uterus (see paragraph 3.3.3). For the women who
ruptured an intact uterus in Phillips study [45], factors contributing to the uterus rupture
included cephalopelvic disproportion, malpresentation, prolonged labour, and herbal
oxytocics.
In general, the following patients are at higher risk of CPD: nulliparae who are short,
have a deformity of the spine or leg or who carry a large baby, multiparae with a history
of caesarean section, symphysiotomy, vacuum extraction or forceps delivery, prolonged
or difficult labour, repeated fresh stillbirths or neonatal deaths at term, or a large baby.
WHO recommends to refer short nulliparae, women with a poor obstetric history and
nulliparae younger than 17 years, especially if they are single, to deliver in an appropriate
facility [53,54]. In a study conducted in the Lower Shire Valley, CPD was present in
nd
edition 2013

Page 16 of 53
2.3% of pregnant women [55]. Nulliparous women, babies weighing more than 3400
grams and women shorter than 155 cm were at increased risk for CPD.
3.4.1.2 Nutritional status
Chronic protein energy malnutrition throughout childhood and adolescence leads to
stunting which is associated with impaired pelvic growth and thus contributes to cephalo-
pelvic disproportion (see paragraph 3.4.1). The micronutrient deficiencies that appear to
be most common and most important in pregnancy are iron deficiency and folate
deficiency, which both are an underlying cause of anaemia in pregnancy (see Anaemia in
Pregnancy chapter).
Refer to nutrition and anaemia chapters for a distribution of these macro- and
micronutrient deficiencies in Malawi.
3.4.2 Behavioural determinants
3.4.2.1 Fertility preferences and use of family planning methods
Because every pregnancy carries a risk, a lower fertility would result in less overall risk
of maternal mortality. The ideal family size as reported in the 2010 MDHS was 4 for
women, which is more than one child less than the actual fertility rate of 5.7 children
[25]. The wanted fertility was highest in uneducated women (4.9) and lowest in those
with more than secondary education (2.8). The difference between wanted and actual
number of children was lower in urban (0.7) than in rural areas (1.3). The current total
fertility rate of 5.7 constitutes a decline of 25 percent over the last 23 years (Figure 7).
The decline in fertility was more pronounced in urban (18 percent) than in rural areas (3
percent). The actual number of children in a family correlated well with the familys
fertility preferences [56].
Figure 7 Trend of total fertility rates, Malawi 1977 - 2010

A wide variety of family planning methods exists, which can be grouped into modern
methods and traditional methods. Modern methods include short-term methods (female
y = -0.0603x + 126.97
R = 0.9471
0
1
2
3
4
5
6
7
8
9
1970 1980 1990 2000 2010 2020
T
F
R

Total fertility rate
projection
5.5%
SWAp 1
2004 -
2011
nd
edition 2013

Page 17 of 53
or male condom, oral contraceptives, spermicides, diaphragms), long-term methods
(injectable contraceptives, contraceptive implants, intrauterine devices, Lactational
Amenorrhea Method (LAM)) and permanent methods (male or female sterilisation).
Traditional methods include periodic abstinence, withdrawal and local methods such as
herbs and strings.
Acquiring knowledge about fertility control is an important step toward gaining access to
and then using a suitable contraceptive method in a timely and effective manner.
According to the 2010 MDHS, knowledge of several modern methods of family planning
(i.e. the pill, injectables, condoms and female sterilisation) was high (99.7%) among
married women and men and among women and men who were unmarried but had had
sex before. Other methods like male sterilisation, diaphragm and implants were less well
known. 75% of married women and 55% of married men had ever used a modern method
of family planning, in order of decreasing frequency: injectables, the pill, the male
condom and LAM. The percentage of ever users of contraceptive methods was
marginally higher among unmarried men and women, and the choice of methods was
slightly different, i.e. in order of decreasing frequency: male condom, injectables and the
pill.
There has been a quadrupling of modern contraceptive use over the last decade: from 7%
in the 1992 MDHS and 14% in the 1996 MKAP survey to 26% in the 2000 MDHS and
33% by 2010 in women (Figure 8) [2123,25,56]. This increase was mainly due an
increase in use of injectables. An additional 4% of married women and 1% of unmarried
women used traditional contraceptive methods. The current use of modern contraceptives
was higher in urban (50%) than rural areas (41%). Uptake also increased with higher
education and a higher number of living children. 72% of women want, 46% are getting
(20% for spacing and 26% for limiting) and 26% want but are not getting family
planning.
Figure 8 Trends in use of contraceptives, Malawi, 1992 - 2010

0
10
20
30
40
50
60
70
1990 1995 2000 2005 2010 2015 2020
P
e
r
c
e
n
t

Contraceptive prevalence rate - all forms of contraception of
married women aged 15-49 years of age Malawi
projection
55%
MDG
target
65%
SWAp 1
2004 - 2011
nd
edition 2013

Page 18 of 53
3.4.2.2 Decision making to go to health facility
Delay in decision making to deliver in a health facility was a contributing factor to 77%
of maternal deaths in the community based Nankumba study [26]. Qualitative research
by the SMP suggested that decision making patterns vary between districts in South
Malawi. In many matrilocal societies (like the Yao in Mangochi, the Manganja in
Chikwawa, and the Lomwe and Manganja in Phalombe, Chiradzulu and Zomba) where
the husband goes to live with his wifes family after marriage and thus his relatives often
live far away, the wifes relatives (especially the uncles) are important decision makers.
In the latter three districts, the husband was rated as second most important decision
maker by women, but as most important by men [57]. However, among the Yao in
Mangochi, the husbands relatives also play an important role: they need to witness
whether the labour is prolonged, because this is believed to be proof that the child is not
the husbands. Especially in Mangochi, where girls marry and get pregnant at a very
young age and thus experience prolonged labour more often, this leads to many
complicated deliveries occurring at home [58]. In complicated deliveries, husbands are
the main decision makers in 40% of cases in Blantyre and 68% in Nsanje (with a
predominant patrilocal Sena population) [59]. In normal deliveries however, the woman
herself most frequently decides where to go for delivery. In many societies the husbands
play an important role in providing material and financial support during delivery, and
may therefore in practice have a large influence on delivery site. Apart from these gender
and hierarchical aspects, other factors also influence the decision to go to the health
facility, i.e. distance to health facility (see 3.4.3.4), perception of danger signs (see
3.4.2.6), previous experience (see 3.4.2.4) and perceived quality of care (see 3.4.2.5).
3.4.2.3 Infection prevention practices
With sepsis as a major cause of in-hospital maternal deaths, unhygienic practices are an
important determinant of maternal mortality. A 2000 report by the East, Central and
Southern Africa College of Nursing (ECASON) of seven health facilities and three
schools of nursing in Malawi showed that the only national infection prevention
guidelines that existed were those of the National AIDS Commission, but these were not
known to most healthcare workers. In most hospitals, infection prevention (IP)
committees were lacking, no monitoring systems were in place for nosocomial infections,
knowledge about IP was limited and practices were inadequate. Although some staff had
had training on IP (mostly as part of family planning training), follow up training and
supervision (e.g. by an infection control coordinator or committee) were absent [60].
Another study by the Safe Motherhood Project done in 2000 in the maternity wards of
health facilities in the 12 districts in the Southern Region confirmed these findings [61].
An in-depth study of infection prevention perceptions and practices was done at
Mangochi District Hospital, after the Safe Motherhood Program provided training on IP
to improve staffs knowledge and awareness. It revealed that general knowledge was
good, but knowledge about details and implementation of knowledge was poor. Health
care workers believed that most infections arise outside of the hospital after home
deliveries. Much emphasis was being put on availability of supplies rather than on their
appropriate use. Shortage of staff led to overworked nurses whose most pressing concern
was to attend to all the deliveries occurring at the same time, with no time or attention for
infection prevention. Also nursing duties were delegated to support staff (e.g. changing
nd
edition 2013

Page 19 of 53
intravenous fluid), who had not been trained on infection prevention. Supervision was
weak and hospital management had little or no role in the implementation of infection
control [62].
MOH in cooperation with JHPIEGO started training staff in IP and in the quality
improvement process in 2001 in the four central hospitals and in Chiradzulu District
Hospital, St. Johns hospital and Likuni Hospital, resulting in steady progress in
achievement of performance standards in all hospitals and three sites achieving standards
for external recognition. This program was extended to Chitipa, Rumphi, Nkhata Bay,
Mzimba, Karonga, Salima and Mulanje District Hospitals in 2004 and will be further
expanded to include more hospitals and NGOs in the near future.
In the 2010 EPOC survey IP guidelines were missing in 55% facilities, infection
prevention materials were often unavailable (chlorhexidine 36%, cidex 21%, povidone
iodine 50%, Moyo stand 42%) and only 17% of trained staff knew the next step in tested
IP activity [15].
3.4.2.4 Previous experience with health care system
Women in focus group discussions in Zomba, Phalombe and Chiradzulu Districts who
had previously delivered in a health facility expressed a preference for institutional
delivery for future pregnancies, because of perceived better management of
complications. Women who had only delivered at home did so because they had never
experienced any problems, because they wanted to avoid rude nurses, were afraid for a
caesarean section and because of the unlimited support available at home [57].

Findings
from Thyolo, Mangochi and Chikwawa were similar [58].
3.4.2.5 Perceived quality of care
Like anywhere in the world, the perceived quality of care in delivery services in Malawi
consists of perception about technical quality and the interpersonal communication with
which it is delivered. Clients in Blantyre and Nsanje ranked good reception/staff
behaviour and prompt, appropriate care as the most important factors for quality obstetric
care. Whereas reception was often received as poor, still more than 90% were satisfied
with the care provided. Many women had delivered alone or felt they had been insulted
by staff. Also, the majority of staff themselves admitted they were often rude, reasons for
which were poor working conditions, inadequate staffing levels and long working hours
[63]. In an SMP needs assessment study in Zomba, Phalombe and Chiradzulu districts,
women mentioned health workers unwillingness to assist pregnant women, beating
pregnant women especially when they are in labour, rudeness, performing operations
while drunk, use of abusive language, discrimination of poor women, delays in treating
women and lack of privacy and confidentiality as important barriers. Reception was said
to be better at CHAM health facilities than government facilities and also TBAs were
said to humbly treat and assist pregnant women [57].

CHAM was also perceived to offer
cleaner facilities and always have drugs, but the user fees are a strong deterrent [63].

Similar to Nsanje and Blantyre, respondents did perceive the quality of technical care to
be better at health facilities than at TBAs [57].

In an evaluation of delivery preferences
among women attending ANC at Nankumba health centre in Mangochi district, 95% of
women indicated they wanted to deliver in the health centre, mainly because they
expected quick referral to the hospital in case of problems. However, only 29% actually
nd
edition 2013

Page 20 of 53
delivered there, mainly because they realised too late they were in labour (62%) or
because the health centre was too far (27%) [64].
In 2007 staff perceptions of quality of care in Mwanza were still poor [65].
In a JPEIGO study undertaken in 2009 to evaluate a quality improvement programme the
performance scores give an indication of the quality of care both in the intervention and
control facilities studied [66]. Quality of care was generally good in Family Planning:
both study groups complied with at least 80% of the verification criteria for 9 of the 16
standards. The intervention group scored significantly higher than the comparison group
on two standards: establishing a cordial relationship with the client and identifying her
needs (99% and 84% respectively), and identifying the need for protection against
sexually transmitted infections (STIs), including HIV (73% and 26%, p<01). Both study
groups performed less than half of the verification criteria for the four standards related to
inserting implants. The findings also show strong performance in ANC: both study
groups achieved 80% or more of the verification criteria for 11 out of the 18 standards.
Both study groups performed less than half of the verification criteria for requesting
laboratory tests.
In the labour and delivery area, both study groups complied with at least 80% of the
verification criteria for only 7 out of the 16 standards. In both study groups, scores were
nearly perfect (ranging from 96% to 99%) for conducting the obstetric exam and for
performing infection prevention practices during labour according to standards. Both
study groups performed only about one-third of verification criteria for monitoring
postpartum women during the 2 hrs after the birth. Quality of care was not as good for
PNC: both study groups complied with 80% or more of the verification criteria for just 4
of 20 standards. Scores were especially high for assessing the condition of the neonate
(96% and 93%, respectively).
In summary, quality of care is a problem particularly in the labour ward and postnatal
period.
Table 3 Reproductive health services in Malawi: An evaluation of a quality
improvement intervention [66]

3.4.2.6 Perception of danger signs
In several needs assessment reports for SMP, knowledge about danger signs in the
population was low: only 15% recognised bleeding and 0% recognised sepsis / fever as a
danger sign. Community members with knowledge about danger signs were twice as
likely to deliver in a health centre as those with minimal knowledge [67].
nd
edition 2013

Page 21 of 53
In a study in Mchinji in 2003 womens groups were asked to identify the most important
maternal health problems [68]. The results are shown in Table 4.
Table 4 - The ten maternal health problems most commonly identified by women
attending MaiMwana womens groups in Mchinji District across all periods (n=172)

A study looking specially at womens perceptions of pre-term births was undertaken in
Namitambo, Southern Malawi in 2007 as an adjunct to a clinical trial under way in
southern Malawi (including Namitambo), which addresses the effectiveness of antibiotic
prophylaxis to prevent preterm delivery [69]. Similarities and differences exist in
understanding between healthcare providers and the community. The conclusion was
that additional dialogue and action is needed within the health sector and community to
address the problem of preterm births.
3.4.3 Socio-cultural and socio-economic determinants
3.4.3.1 Womens socio-economic status
Poverty level has been shown to have a strong association with maternal mortality. In
Indonesia, the risk of maternal death in the poorest group was 3-4 times that of the richest
group [70]. In contrast, a review of poverty and health in Malawi finds little difference in
maternal mortality rates between the rich and poor (Figure 9) [71].
nd
edition 2013

Page 22 of 53
Figure 9 Maternal mortality by wealth quintile

For this analysis, called the familial method of assessing the effect of wealth on maternal
mortality, an asset score is compiled using data from the DHS survey for each household,
which is ranked and placed in one of five equal groups from a poorest quintile group to a
richest group. The survival history of each sister is recorded in the previous five years
and maternal mortality calculated. The method is called familial because the method
categorises the wealth of the sister and not the women who died. This equality occurs
despite the fact, as shown in the Integrated Household Survey 2004, that poor women
have more children than non-poor women, even when adjusted for education level
(Figure 10) [72].
Figure 10 The association between poverty, education and fertility, HIS 2004,
Malawi

nd
edition 2013

Page 23 of 53
Education level itself is an important determinant of access to family planning (see
paragraph 3.4.3.3) and obstetric care services. In the SMP, higher institutional delivery
rates were observed in districts with higher female literacy rates. Number of years of
education showed a dose response relation with maternity service attendance: women
with primary, secondary or higher education were 4, 5 and 7 times more likely to use
services than women without education [73].
It has been stipulated that a womans social status also influences how much control she
takes over her own fertility. From the MDHS 2000 it appeared that the use of family
planning slightly increased with an increased number of decisions in the household in
which a woman has a final say [22].
3.4.3.2 Urban / rural residence
Residence influences several determinants of maternal mortality, e.g. use of family
planning (see 3.4.2.1, unmet need for family planning (see 3.4.3.3), access to health
facilities (see 3.4.3.4) and access to skilled attendants at birth (see 3.4.3.6).
3.4.3.3 Access to family planning
According to the 2010 MDHS, the unmet need for family planning among married
women was 18%, 10% for child spacing and 8% for limiting child bearing [25]. The
unmet need among unmarried women was much lower, i.e. 3 percent. The unmet need
for family planning was similar for all age groups, except that it was a bit lower for
women approaching the end of their reproductive life. Unmet need was higher in rural
(19%) than urban (16%) settings and for women with lower education and varied
between regions, with northern having the lowest unmet need (17%) and Central the
highest (19%). Of the total demand for family planning, only 66% was met, although this
represents a dramatic increase from the 26% measured in 1992, indicating that the
coverage by family planning services has improved considerably. Demand satisfied was
highest in the richest quintile (73%) and lowest in poorest (60%).

In the abortion study
described in paragraph 3.2.4, most providers thought that offering post-procedure FP
services was important, but comprehensive services did not exist in these hospitals.
Whereas 25-78% of providers said that information about where to get FP in the
community was offered to patients, only two of the 50 patients interviewed received such
information [47].

3.4.3.4 Distance to health facility, availability of transport and transport costs
Distance to a health facility can be an obstacle to reaching the facility, or it can act as a
disincentive to try seeking care [52]. A distance decay curve exists, meaning that the
farther away people live, the less likely they are to use a health facility. However,
utilisation of health services is not a function of distance alone, but also of costs and
quality (see rest of this paragraph).
Patients delay (this may include a delay in the decision to go to the hospital) was the
principal avoidable factor in 19% of the maternal deaths in Driessens study [31], in 15%
of deaths in the Safe Motherhood Study [36]

and in 16% in the Wiebenga study [33].

Although the question was not specifically about problems during delivery, the 2000
MDHS asked women about big obstacles in accessing healthcare for themselves [22].
The time required getting to the health facility, the availability of transport and the cost of
nd
edition 2013

Page 24 of 53
transport were mentioned by 56, 52 (54 in 2010) and 60 percent respectively. A higher
number of children, low education, rural residence, living in the Central region or
Southern region were all associated with reported less access to health facilities due to
time and cost constraints. An SMP needs assessment in the 12 districts in the Southern
region showed that most women lived 2 to 5 km from a health facility, and that walking
was the main form of transport, irrespective of the severity of the obstetric condition. One
third of women took longer than 2 hours to reach a health facility and one tenth took
more than 4 hours [73].
A study in Mzuzu in 2007 assessed the accessibility of birth services for pregnant women
and found that 46% said they could reach a health facility within two hours, 24% would
pay more than $1 for transport, and while 82% intended to give birth in a health facility
only 49% actually did in a previous pregnancy while 32% used a TBA [74].
3.4.3.5 Availability of blood for transfusion
Lack of blood contributes to death from haemorrhage, especially in anaemic women. In
the SMP study in 18 hospitals in the Southern Region, lack of blood for transfusion was
the principal avoidable factor in 18% of maternal deaths [36]. In the Nankumba Safe
Motherhood Project it was a contributing factor in 32% of maternal deaths [26].
In a 2005 study in Thyolo of 133 cases of obstetric haemorrhage 37 had an antepartum
haemorrhage and of these over half received at least one unit of blood. Of the 95 who
had a postpartum haemorrhage 50 received at least one unit of blood. The maximum
number of units given to any one patient was two [75].
In the facility-based maternal death review in three districts in the central region of
Malawi in 2007, of the 43 deaths reviewed lack of blood was a contributing factor in nine
(21% of) patients [43].
A one month study of blood transfusions in Mangochi hospital in 2010 in which 104
units of blood were used 18 were pregnancy related. Blood was continuously in stock
during the study period and available for transfusion within 1 hour of requisition [76].
3.4.3.6 Availability and quality of skilled attendants
A Millennium Development Goal is to have 80% of deliveries assisted by an attendant
with midwifery skills by 2015. Because home delivery by skilled attendants is rare in
Malawi (TBAs are not considered skilled attendants), this percentage roughly
corresponds to the percentage of institutional deliveries. However, due to staff shortages,
deliveries in health centres and hospitals might not always be performed by a person with
midwifery skills, but by e.g. a ward attendant. According to the 2000 MDHS, 55% of
deliveries in the five years preceding the survey were assisted by a skilled birth attendant
[22]. This proportion varied hugely per district, from less than 45% in Karonga and
Kasungu to over 80% in Blantyre. Urban residence was associated with a higher
percentage of pregnancies assisted by a skilled attendant (82% versus 52% in rural areas).
These data are self-reported by interviewed women and might therefore be subject to the
womans interpretation of what constitutes a skilled attendant. The survey also found
higher rates in richer families with little change since the previous DHS in 1992.
In the 2003 HMIS data, which uses reports by hospitals, the estimated percentage of
deliveries assisted by a skilled attendant was 41 %, ranging from a low of 22% in
nd
edition 2013

Page 25 of 53
Chiradzulu to 84% in Rumphi. There may have been considerable underreporting in
some districts. The SMP project reported that the percentage of institutional deliveries in
the Southern Region increased from 34% in 1998 to 57.4% in 2001 [11].
By 2010 the DHS, also reporting deliveries in the preceding 5 years, found a SBA level
of 71% (84% in urban and 69% in rural areas) [25]. There has been a steady rise in the
proportion of skilled attendant deliveries of 20 years (Figure 11), but probably
insufficient to meet the MDG target of 80% by 2015.
Figure 11 Trend in births attended by a skilled attendant, Malawi 1992 - 2010

Staff shortages severely affect the quality of care in Malawi. The vacancy rate for
registered nurses was 47% in 1998 and has likely grown bigger. Ostergaard described
that the two main forms of losses are that midwives die or go abroad. The latter is also a
loss in quality as it is the more experienced midwives who emigrate. Inadequate
remuneration was the main push factor and the midwives who did stay used multiple
coping strategies to supplement their income, e.g. accessing training allowances and
emergency loans, running small businesses, sell food or pharmacy drugs to patients and
doing shifts in the private sector. These often had negative impact on the quality of care
due to the opportunity costs. Poor working conditions, lack of career structure and lack of
job satisfaction were also found to contribute to poor retention. The main pull factors
that attract midwives to stay with the MOH were the retirement package, access to
postgraduate training, flexible leave policy and job security [77]. Further studies on staff
retention have confirmed these findings [7883]. There is some good evidence that the
brain drain is reduced and that the Emergency Human Resources Plan has succeeded in
reducing staff vacancies [84]. The recent EPOC studies have a chapter on staff
availability and their skills [15] (see also section 2.1.2).
The effects of a course on life saving skills by SMP were evaluated by MacLean for SMP
in 2000 [85]. The course resulted in increased knowledge, but understanding of critical
concepts and clinical skills remained substandard. For example, although two thirds of
respondents accurately prioritised the first action in the management of PPH (rub up a
uterine contraction), only 9% correctly prioritised all subsequent actions. MacLean
54.8
55.6
57
62
66
57
75
73
0
10
20
30
40
50
60
70
80
90
100
1990 1995 2000 2005 2010 2015 2020
P
e
r
c
e
n
t
Births attended by skilled attendant
projection
72%
MDG
target
80%
SWAp 1
2004 - 2011
nd
edition 2013

Page 26 of 53
concluded that the SMP course laid a good foundation for improving quality, but that to
be effective, this course should be followed-up by teaching on clinical practice and
analytical thinking that leads into action at the practical level.
3.4.3.7 Adequacy of referral system
The UN indicator target for women with obstetric complications who receive appropriate
treatment is 100%. From the 2003 HMIS data it appeared that only 40% of the 83,000
women who were expected to experience an obstetric complication actually delivered in
an emergency obstetric facility, but again underreporting most likely existed [86]. From
this and the previous paragraph it can be concluded that although the percentage of
women that delivers in an EmOC facility is higher than the UN target of 15%, this does
not meet the need of all of those with complications. Thus the referral system is not
adequate.
3.4.3.8 Availability, accessibility and quality of antenatal care services
According to the 2010 DHS data, antenatal care attendance is nearly universal in Malawi,
with 95% of women attending at least once during pregnancy [25]. The majority attended
2-3 times (49%) or 4 or more times (the recommended frequency, 46.0 %). However,
only 12% came in the first trimester of their pregnancy, as recommended. The median
pregnancy duration at the first visit was 5.6 months. Standard ANC in Malawi consists
of:
IEC for danger signs
blood pressure measurement
urine sampling to detect protein and diabetes
blood sampling for syphilis
injection with tetanus toxoid
provision of iron and antimalarial tablets
Good quality care would mean that all these services are provided to all women. The self-
reported coverage of these services according to women in the MDHS data showed that
approximately 90% of women were protected against tetanus toxoid and given iron, 84%
had their blood pressure measured, approximately 80% of women received IEC, but
blood and urine samples were only taken in 82% and 27% respectively. The HMIS 2002-
2003 annual report showed a different picture, with ANC coverage of 100% [86]. This
figure conceals the variation between districts which ranged from 45% in Chiradzulu to
165% in Mwanza. However, in some cases, not all antenatal cases were captured in the
report and in other cases repeat visits were counted as new visits. In 2009 ANC coverage
was over 90% with an average of three visits per pregnant woman [87].
3.4.3.9 Availability, accessibility and quality of emergency obstetric care services
Whereas at the end of the SMP the availability of basic EmOC and comprehensive
EmOC units in the Southern region in Malawi exceeded the UN target of 4 and 1 facility
respectively per 500,000 population on paper, basic EmOC units with the six signal
functions were rare and their quality was not always up to standard. In the SMP study in
2001, substandard intramural care was responsible for 49% of deaths [36], an increase of
nd
edition 2013

Page 27 of 53
10% from the 32% reported in 1989 in the study by Driessen [31]. Lack of clinical
history taking, diagnostic tests, errors in clinical judgment, deficient nursing care and
lack of blood transfusion were the main problems. The case fatality rate for obstetric
complications was 2.5% in 2003 falling to [86], i.e. higher that the UN indicator of 1%
(see 2.1.2). Interestingly though, it was 5.3% in the Northern and Central regions, but
only 1.4% in the Southern region (where the Safe Motherhood Project took place),
possibly indicating that the in-hospital quality of care has improved. Only 24% of women
came for a postnatal check up.
The 2005 random sample survey of EmOC facilities at 27 hospitals and 94 health centres
by the MOH finds [13]:-
Malawi has almost double the recommended number of Comprehensive EmOC
facilities (1.8 facilities per 500,000 population) and only 2% of the recommended
number of basic EmOC facilities (0.1 facilities per 500,000 population).
Of the 94 health centres assessed 92 did not qualify as basic EmOC facilities as
they were not providing all six basic EmOC signal functions (see Figure 5 below
for the proportion of health centres offering these functions)
The met need for EmOC was 18.5 % which is below the UN recommended level of
100%.
Of expected births in Malawi 2.8% are by caesarean section which is below the
recommended minimum of 5% indicating that many women are not receiving the
care that they need.
Major barriers mentioned include:
o Lack of decision making power of the women with complications,
o Inadequate transport and communication linkages between community and
health facilities, and between health facilities,
o High cost related to service delivery
o Problems related with the service delivery (e.g. staff attitude, inadequate
equipment, drugs and supplies)
o Quality of EMOC services was generally poor as evidenced by the high
maternal mortality within the facilities and high case fatality rate of (3.4%)
which is much higher than the UN recommended level of less than 1%
o Low staffing levels in all facilities
Two further EMOC surveys have been undertaken since 2005, both taking place in about
2010. The 2010 found similar results to the 2005 survey for both hospitals and health
centres. The 2003 and 2009/11 surveys were regional sample surveys with limited
coverage. The inability to undertake Ventouse extraction to assist delivery and removal
of retained placenta are the common reasons why our health centres are not classified as
EMOC facilities (Table 5,
Year Total
population for
survey area
Minimum
recommended
number of CEOC
facilities (1 per
500,000
population)
Minimum
recommended
number of BEOC
facilities (4 per
500,000
population)
Total number of
facilities
available and
surveyed
Number and
proportion of CEOC
facilities providing 9
signal functions
Number and
proportion of BEOC
signal functions
2005

11,937,934 24 96 186 42 100% 2 1.4%
nd
edition 2013

Page 28 of 53
2010

13,077,160 26 105 131 42 100% 10 11%
2009/11

1,972,563 4 16 39 8 100% 2 6%
Figure 12 and Figure 13).
Table 5 - Minimum recommended number of health facilities expected to provide
Emergency Obstetric Care (EmOC), number of health facilities available and
number providing required signal functions for Basic and Comprehensive EOC by
country [12,13,15,16].
Year Total
population for
survey area
Minimum
recommended
number of CEOC
facilities (1 per
500,000
population)
Minimum
recommended
number of BEOC
facilities (4 per
500,000
population)
Total number of
facilities
available and
surveyed
Number and
proportion of CEOC
signal functions
Number and
proportion of BEOC
signal functions
2005

11,937,934 24 96 186 42 100% 2 1.4%
2010

13,077,160 26 105 131 42 100% 10 11%
2009/11

1,972,563 4 16 39 8 100% 2 6%
Figure 12 Availability of individual signal functions at CEmOC (hospital) level, 4
EMOC surveys, Malawi, 2003-2011 [12,13,15,16]

0
10
20
30
40
50
60
70
80
90
100
P
e
r
c
e
n
t

individual signal functions
2003
2005
2010
2009/11
nd
edition 2013

Page 29 of 53
Figure 13 - Availability of individual signal functions at BEOC (health centre) level,
4 surveys, Malawi, 2003-2011 [12,13,15,16]

0
10
20
30
40
50
60
70
80
90
100
P
e
r
c
e
n
t

individual signal functions
2003
2005
2010
2009/11
nd
edition 2013

Page 30 of 53
Summary - Determinants
Decreased number of pregnancies will decrease number of maternal deaths
o Met need for family planning increased to 47% in 2010 but unmet demand stands
at 26%.
o Ideal family size (4) is still more than one child less than actual family size (5.7);
unmet need for family planning is highest among rural and less educated women.
Complications cannot be predicted, but screening for some conditions is possible
o Access to antenatal care is good, but quality of care substandard, especially
screening for diabetes and syphilis is often not performed
Patient delay is the primary avoidable factor in 15-20% of deaths.
o Time and costs constraints to get to health facility are obstacles for 2/3 of women.
o Even where women take decision about place of delivery, in practice they often
depend on husband providing material and financial support.
o Women think quality of technical care in health facilities is better than at TBAs,
but staff attitude and service are bad.
o Knowledge of danger signs increases likelihood of institutional delivery, but is
low.
o Districts with higher female literacy levels have higher institutional delivery rates.
o Three-quarters of all women but only some of those with complications deliver in
health facility.
Once at the health facility, timely care of good quality determines outcome.
o Availability of CemOC facilities is sufficient, but quality is substandard.
Availability of BEmOC facilities is grossly insufficient.
o Staff shortage is still high but slowly improving. Main push factor is low salary;
main pull factors are retirement package, postgraduate training, and flexible leave
policy and job security.
o Infection prevention practice was inadequate before quality improvement program
started.
o Life skills theory was improved by training, but clinical skills need improvement.
o Lack of blood was an important contributing factor to 18-32% of deaths but may
be improving.
o Case fatality rate in complicated institutional deliveries in the Northern and
Central regions was 5.3% and in the Southern region (where SMP was carried
out) 1.4%.

4. Impact of maternal mortality
4.1 Mortality & life expectancy
The mortality rate from maternal causes rose from 1.4 in the DHS 1992 survey to 2.4 in
2000 to 2.0 in 2004 to 1.3 in /1000 women between 15-49 years in the DHS 2010 survey
[25].

Given that the rise in non-maternal mortality is mainly AIDS driven, one would
nd
edition 2013

Page 31 of 53
expect a disproportionate rise in non-maternal mortality and thus a decrease of the
proportion of female deaths that is maternity-related. However, the proportion of all
maternity-related female deaths has remained constant at 20-21% between the late 1980s
and the late 1990s. This means that measured maternal mortality has risen at roughly the
same pace as non-maternal mortality. Several explanations for this phenomenon exist:
underestimation of the maternal death component of all female deaths in the 1992
MDHS
overestimation of the maternal death component in the 2000 MDHS, e.g. over-
reporting of an AIDS-related death as a maternal death because of stigma
a real rise in HIV-related indirect obstetric deaths
underreporting of AIDS deaths in the 2000 MDHS
a real rise in maternal deaths due to deterioration of health services associated with
the AIDS epidemic.
Because the AIDS epidemic affected urban populations earlier and more severely then
rural populations, one would expect a higher increase in maternal mortality in urban
populations if AIDS was to have an impact on maternal mortality. Bicego and colleagues
showed that this was indeed the case for Malawi [88].
4.2 Infant and child morbidity and mortality in households affected by the
disease
A maternal death has a detrimental effect on the wellbeing of the family, but especially
on young infants. McDermott described that among infants of mothers who died a
maternal death in Mangochi, the mortality was 3.7 times higher than among infants
whose mother survived. Only 31% of children born to deceased mothers survived
through infancy [20].
4.3 Economic & social
Women are the backbone of the agricultural part of Malawis economy. Their average
work day lasts 15 hours as compared to 6 hours for men. They carry out 70% of all the
farm work in the small-holder agricultural sub-sector [89]. From a societal perspective,
women are the main carers, not just for their own family, but also within the community.
Over one quarter of all households is female-headed [25]. It can therefore be expected
that a woman dying a maternal death has a large impact on the economic and social status
of her family, but no data have been found.
4.4 Health services
Assuming that all women who deliver in hospital are admitted, deliveries constitute just
under half (420,000) of 923,000 total admissions to hospitals and health centres in
2010/11 (HMIS). Any change to the rate of institutional deliveries or the structures and
procedures of maternity services thus impacts significantly on the health system.
Summary - Impact
1.3 out of 1000 women between 15 and 49 die a maternal death
Maternal mortality has risen at the same rate as (mainly AIDS driven) non-maternal
mortality in the early 2000s but it is difficult to say how much is due to AIDS and
nd
edition 2013

Page 32 of 53
4. Effective interventions
THIS SECTION HAS NOT BEEN FULLY UPDATED. THE REFERENCES OF
RECENT COCHRANE REVIEWS ARE INCLUDED HERE AND SHOULD BE
CHECKED BEFORE ACCEPTING THE CONCLUSIONS OF THE REVIEW [90119]
Interventions should focus on one of the three main determinants mentioned under
paragraph 3.2:
1. the likelihood of a pregnancy occurring
2. the likelihood of a complication arising
3. the successful management of a complication
Ad A. This concerns family planning methods because they reduce overall fertility and
increase child spacing.
Ad B. These concern 1) family planning methods because they reduce the risk of
unwanted pregnancy and therefore induced abortion and 2) interventions to increase
womens overall well-being and health through prevention of and screening for existing
problems that contribute to poor maternal health, including anaemia and deficiencies of
essential nutrients.
Ad C. Even when the interventions mentioned under B are adequately implemented,
many women will experience pregnancy complications that could not have been
predicted or prevented.
Therefore, important safe motherhood interventions focus on:
educating women and their families about the fact that all women face the risk of
pregnancy complications, and the actions they should take if and when there is a
problem;
providing adequate care as close as possible to where women live, including skilled
attendance at delivery, prompt recognition and referral of complications, and
adequate treatment for women with complications;
ensuring functioning systems of communication and transport that link community-
based health workers, health centres, and hospitals to ensure that women receive
needed care quickly.
Below, all interventions are described that have proven efficacy in preventing either
maternal deaths or the complications that often precede maternal deaths. The highest
quality evidence comes from (systematic reviews of) randomised trials such as done by
the Cochrane Collaboration and included in the WHO Reproductive Health Library
(RHL) [120]. Because nearly no behavioural or socio-cultural/ socio-economic
interventions have been subject of such thorough studies, we included evidence from
observational studies as well. WHO has published a global review of the key interventions
related to reproductive, maternal, newborn and child health [121]. A one page summary is
provided in the appendix.
how much due to obstetrics
Mortality in infants of deceased mothers is 3.7 times the mortality in infants of living
mothers
Deliveries form half of all hospital admissions
nd
edition 2013

Page 33 of 53
There are two WHO publications which summarise and offer guidance about maternal
health:-
Packages of Interventions for Family Planning, Safe Abortion care, Maternal,
Newborn and Child Health, 2010 [122]
Managing complications in pregnancy and childbirth: a guide for midwives and
doctors, reprint 2007 [54]
They provide authoritative advice and some good references.
5.1 Biological interventions
5.1.1 Family planning
Aspects of two family planning methods are included in the RHL, i.e. emergency
contraception (EC) and the surgical approach to tubal ligation.
5.1.1.1 Emergency contraception
Emergency contraception is the use of a drug or device as an emergency measure to
prevent pregnancy after unprotected intercourse. In many developing countries the lack
of access to emergency contraception may subject women to unsafe abortions. Currently,
there are several different interventions available for emergency contraception, such as
estrogen+progestogen, progestogen alone (levonorgestrel), mifepristone, danazol etc. A
recent Cochrane review concluded that levonorgestrel and mifepristone offered the
highest efficacy with an acceptable side-effect profile[104]. In Malawi, a nationwide EC
program is being implemented through JHPIEGO. First choice for Malawi is the Postinor
2 regimen (levonorgestrel) which requires the woman to take one pill within 72 hours of
unprotected intercourse and another one 12 hours later. EC services are now available in
11 health facilities in the Northern region, 13 in the Central region and 23 in the Southern
region. The development of IEC messages and materials is in progress.
5.1.1.2 Surgical approach to tubal ligation
Worldwide, the most commonly used method of fertility regulation is tubal sterilisation.
In developing countries, where the resources are limited for the purchase and
maintenance of the more sophisticated laparoscopic equipment, minilaparotomy is a more
common approach than laparoscopy. A Cochrane review showed that major morbidity
seems to be a rare outcome for both laparoscopy and minilaparotomy. Personal
preference of the woman and/or of the surgeon and of course practical considerations
(cost, maintenance and sterilisation of the instruments) can guide the choice of technique
[110]. In Malawi, minilaparotomy is the method of choice. Tubal ligation at Caesarean
Section is also common.
5.1.2 Iron and/or folate supplementation
According to three Cochrane reviews, supplementation of iron or folate or both during
pregnancy prevents low haemoglobin in late pregnancy or at delivery [123,124].

However, no conclusions could be drawn in terms of any effects, beneficial or harmful,
on clinical outcomes for mother and baby. The reviewer concluded that routine iron and
folate supplementation could be warranted in populations where deficiency for these
micronutrients is common (as is the case in Malawi: refer to Nutrition and Anaemia in
nd
edition 2013

Page 34 of 53
pregnancy chapter). The reduction in the proportion of women with low haemoglobin
levels at term should in theory reduce the need for blood transfusion, which is relevant in
countries with high HIV prevalence like Malawi. The reviewer further commented that
data from countries where anaemia is a serious clinical problem and iron and folate
deficiency is common are scarce, and called for trials in these populations that also are
large enough to establish the effects of iron and folate supplementation on clinical
outcomes in mother and child.
In Malawi iron, but not folate supplementation is part of routine antenatal care. However
data from the 2000 DHS show that only 70% of pregnant women received iron tablets.
This figure ranged from 63% in Machinga to 86% in Karonga. Among women going for
antenatal care, older mothers, mothers with low education and multiparae were less likely
to have received iron tablets.

The 2002-2003 HMIS data suggest 84% of pregnant
women received iron tablets.
5.1.3 Presumptive treatment of malaria
The use of antimalarial drugs for the prevention of malaria during pregnancy reduces
severe antenatal anaemia in the mother and is associated with higher birth weight of the
baby and probably fewer perinatal deaths. This effect appears to be limited to women of
low parity [97]. The policy in Malawi is to give SP twice during pregnancy. According to
the MDHS 2000 72% of women received antimalarials and according to the 2002-2003
HMIS data 90%. Refer to the anaemia chapter for more details.
5.1.4 Calcium supplementation during pregnancy for preventing hypertensive disorders
and related problems.
Calcium supplementation was shown to be beneficial to women at high risk of gestational
hypertension and in communities with low dietary calcium intake, in a Cochrane review
including eleven clinical trials [125]. In women at high risk (i.e. teenagers, women with
previous pre-eclampsia, women with increased sensitivity to angiotensin II, and women
with pre-existing hypertension) the reduction in risk of hypertension with or without
proteinuria was 53%, in women with low baseline dietary calcium intake it was 62%.
The risk of pre-eclampsia was reduced with 78% for high-risk women and with 71% for
those with low baseline dietary calcium intake. No side effects were noticed in any of the
trials. Data were inadequate to establish the effect of calcium supplementation on
maternal deaths. Optimum dosage also requires further investigation. Baseline calcium
intake in Malawi is unknown but is thought to be reasonable in view of the quantity of
fish consumed. No programmes have been identified in Malawi that offer calcium
supplementation.
5.1.5 External cephalic version (ECV) for breech presentation at term
The six studies included in the Cochrane review on external cephalic version at term
showed that this intervention reduced the risk of a non-cephalic birth by 58% and the risk
of caesarean section by 48% [101]. The direct effect on maternal mortality was not
assessed. The studies were too small to assess any risks associated with the procedure.
Another Cochrane review showed that giving tocolytic drugs before ECV (to relax the
uterus) results in an increased number of successful external cephalic versions [126].
Attempts have been made to introduce ECV in Malawi, but when initial results were
negative this approach was abandoned.
nd
edition 2013

Page 35 of 53
5.1.6 Active management in the third stage of labour
Expectant management of the third stage of labour involves allowing the placenta to
deliver spontaneously or aiding by gravity or nipple stimulation. Active management
involves administration of a prophylactic oxytocic before delivery of the placenta, and
usually early cord clamping and cutting, and controlled traction of the umbilical cord. A
Cochrane review concluded that active management is superior to expectant management
in terms of blood loss, post partum haemorrhage and other serious complications of the
third stage of labour, but was associated with increased risk of nausea and vomiting and
hypertension, where ergometrine was used [127]. On average, maternal blood loss was
79 millilitres less, third stage of labour lasted 9.8 minutes less and the risk of PPH was
reduced by 62%. Maternal death was not included in the outcomes studied. Active
management of third stage of labour requires economic and human resources for drugs,
needles and syringes, training of health personnel and availability of refrigerators. Active
management of third stage of labour is in the midwifery curriculum and is considered
standard practice, but data about its use in practice were not found
5.1.7 Antibiotic prophylaxis for caesarean section
Caesarean delivery is the most important risk factor for postpartum maternal infection.
Prophylactic antibiotics were shown to reduce the risk of endometritis by two thirds to
three quarters in a systematic review of 66 trials [93]. They also reduced the risk of
wound infections by 60% and of other serious infectious morbidity by 56%. Another
review, which analysed the best antibiotic regimen, concluded that ampicillin and first
generation cephalosporins are as good as a more broad-spectrum agent. Multi-dose
regimens showed no additional benefit. There appeared to be no difference in efficacy
based on whether the antibiotic was administered systemically or by a lavage route.
There was insufficient evidence regarding the optimal timing of administration [128].
The current obstetric protocols in Malawi are to give chloramphenicol 1g IV, or
ampicillin 2g IV or cefotaxime 1g IV, but no data exist on how well these protocols are
followed in practice.
5.1.8 Umbilical vein injection for management of retained placenta
If a retained placenta is left untreated, there is a high risk of maternal death. However,
manual removal of the placenta is an invasive procedure with its own serious
complications of haemorrhage, infection or genital tract trauma. Any management simple
and safe enough to be performed at the place of delivery, which reduces the need for
manual removal of placenta, could be of major benefit to women world-wide. In a
Cochrane review of 12 trials, umbilical vein injection of saline solution plus oxytocin
compared with expectant management reduced manual removal by 14%, although this
was not statistically significant. Saline solution with oxytocin compared with saline
solution alone showed a 21% statistically significant reduction [116].

Provided that
health workers are trained in the technique it would be feasible to implement this
intervention. Umbilical vein injection for retained placenta is currently not standard
practice in Malawi.
nd
edition 2013

Page 36 of 53
5.1.9 Vacuum aspiration versus sharp curettage for incomplete abortion
Surgical evacuation of the uterus for incomplete abortion usually involves vacuum
aspiration or sharp curettage. Vacuum aspiration utilises a vacuum source and can be
performed on an outpatient basis with local anaesthesia or analgesics. It can be used
without electricity using a hand-held vacuum syringe (Manual Vacuum Aspiration,
MVA). In sharp metal curettage (also called dilatation and curettage, D&C) a metal
curette is used to evacuate the contents of the uterus. It is often performed in an operating
room under general anaesthesia. In two trials included in a Cochrane review, vacuum
aspiration was associated with on average 17 millilitres less blood loss, 26% less pain and
a 1.2 minutes shorter procedure [129]. The main advantage of MVA however is that
provided health care workers are adequately trained, it can be performed at health centre
level, and thus increases the access of women to quality post abortion services. In
Malawi, the RHU and JHPIEGO instigated a comprehensive post abortion care (PAC)
programme, that not only includes emergency treatment of incomplete abortion and its
complications (through MVA if in first trimester and through sharp curettage if in second
or third trimester), but also provides family planning counselling and service, counselling
for emotional and reproductive health concerns, medical conditions, social issues and
legal and sexual rights issues. Comprehensive PAC services are now available in eight
sites in the Northern region and 13 each in the Central and Southern regions. These are
planned to be expanded with 14 services.
5.1.10 Choice of anticonvulsant for eclampsia
A number of different anticonvulsants are used to control eclamptic fits and prevent
future seizures. In three Cochrane Reviews, Duley et al reviewed the evidence on
treatment with magnesium sulphate, diazepam (valium), phenytoin and lytic cocktail (a
mixture of chlorpromazine, promethazine and pethidine) [91,106,107].

Magnesium
sulphate was superior to the three other drugs in preventing recurrent convulsions (by
56%, 69% and 91% respectively) and maternal mortality (by 41%, 50% and 75%
respectively), be it that the latter was not statistically significant. In Malawi magnesium
sulphate is the drug of choice, but diazepam is still used at health centre level to control
fits prior to referral.
5.1.11 Antibiotics for preterm premature rupture of membranes
Preterm prelabour rupture of membranes not only results in neonatal morbidity and
mortality, but also increases maternal infectious morbidity. A Cochrane review concluded
that the use of antibiotics in this situation statistically significantly reduced the risk of
chorioamnionitis by 43% [94]. Because of an increased risk of necrotising enterocolitis
associated with beta lactams (co-amoxiclav), macrolides (erythromycin) are
recommended as the drug of choice. The standard protocol in Malawi is to give
erythromycin 250 mg by mouth 3 x per day for 7 days plus metronidazole 400 mg by
mouth 3x per day for 7 days, but again information about how well these protocols are
followed is lacking.
nd
edition 2013

Page 37 of 53
5.2 Behavioural interventions
5.2.1 IEC for family planning
The provision of education on contraceptive use to postpartum mothers has come to be
considered a standard component of postnatal care in many countries, but data from the
SMP show that only one third of women in Blantyre and Nsanje districts received
information on family planning at discharge. A Cochrane review showed that such
education may be effective in increasing the short-term use of contraception [130]. There
are only limited data examining a more important longer-term effect on the prevention of
unplanned pregnancies. Research on the effectiveness of various aspects of family
planning programs in developing countries is needed, examining the content, timing,
range and organisation of postpartum education on contraceptive use.
5.2.2 IEC to reduce delays in decision making and reaching a health facility
Only very limited evidence exists from developing countries about the efficacy of IEC
for reducing maternal mortality (or any other health problem for that matter). SMP based
their IEC strategy on a review of 81 articles [131] which concluded that:
1. The IEC strategy should focus on the topics danger signs in pregnancy and the post
partum period, use of health services especially with obstetric complications, immediate
referral to health facilities and antenatal care
2. Pregnant women, opinion leaders, health professionals and policymakers should all be
targeted
3. Radio would be the most cost effective communication channel to reach the public, to
be reinforced with messages through mobile video units, pictographs at health centres
and interactive health talks.
Three years into the SMP, after the IEC campaign had been implemented, over 80% of
women and around 60% of men in Blantyre and Nsanje districts reported to have
received information about danger signs, mostly at health facilities or through the radio.
However, when asked, only half of men and women named bleeding as a danger sign and
other complications were reported even less frequently [132].
5.3 Socio-cultural and socio-economic interventions
5.3.1 Improvement of socio-economic status of women
As expected, no trials or systematic reviews have been done on this subject. However,
history teaches us some lessons. Improving socio-economic status in itself does not lead
to a reduction in maternal mortality, as evidence from the early 20
th
century in
industrialised countries has shown. Until the 1930s, maternal mortality ratios in the USA
and UK were as high as they are now in Africa and they had not declined with increased
education levels or with economic development. They only dropped when obstetric care
improved, especially when antibiotics and blood transfusion became widely available
[133]. In Holland and Scandinavia MMRs dropped even earlier due to the establishment
of a professional midwifery cadre that was taught and supervised to implement strict
asepsis. This is further illustrated by the example of a religious sect in the USA called the
Faith Assembly of God, whose members are well fed, well educated and relatively well
nd
edition 2013

Page 38 of 53
off, but who refuse modern medical services. Their MMR stood at 872 per 100.000 live
births in 1983, more than 100 times as high as the average in the USA [134]. In Malawi,
improved educational status of women is associated with better access to obstetric care
and, if the quality of obstetric care would be sufficient, improving womens educational
status might reduce maternal deaths by ensuring that they get care in time.
The traditional cultural beliefs in Malawi vary from place to place and from tribe to tribe.
Many relate to childbirth. An excellent description is found in the report on Cultural
Practices of Malawi [135].
5.3.2 Provision of safe, legal abortion services [136]
Millennia of experience worldwide have shown that despite religious, cultural or
government rules and regulations prohibiting induced abortions, women have always
sought ways to terminate their unwanted pregnancies. Many risk their lives doing so out
of powerful reasons such as concern for their own health, for the wellbeing of the
children they already have or their ability to continue schooling or work. Whereas access
to modern contraceptives does reduce the demand for induced abortion, it does not
eliminate it. Even if all women who do not want to become pregnant were to use a
contraceptive, there would still be need for abortion services, because no contraceptive is
100% effective. The efficacy of abortion services has not been formally evaluated, but
again, historical evidence suggests that providing safe, legal abortion services reduces
maternal mortality, both by reducing the number of illegal abortions and by improving
the quality of legal abortion services. E.g. in the USA, the number of deaths from illegal
and legal abortions two years after introduction of legal abortion services dropped by
84% and 75% respectively [134] and in Romania [137]. In Malawi, abortion is only
permitted on medical grounds, not on social grounds or on demand.
5.3.3 Improving transport and referral systems
In an SMP study, provision of bicycle ambulances or establishment of community
transport schemes did not reduce transport time from home to the health centre. Most
women still preferred to walk, because the cultural belief that publicising the onset of
labour summons evil spirits deterred many pregnant women from using the transport
schemes [138].
SMP also evaluated the effect of an upgraded radio communications system on the
referral of obstetric emergencies in a before-after study. The percentage of women who
were transported from health centre to hospital in an ambulance rose from 54% to 82%
and the median transport time decreased from 3 to 2 hours. Although this is an
improvement, this is still too long, especially for someone with PPH [139].
Motorbike ambulances have been evaluated in Mangochi [140]. It was concluded that in
resource-poor countries motorcycle ambulances at rural health centres are a useful means
of referral for emergency obstetric care and a relatively cheap option for the health sector.
5.3.4 Improve availability of maternity waiting homes
The purpose of maternity waiting homes is to provide a setting where high-risk women
can be accommodated during the final weeks of their pregnancy near a hospital with
essential obstetric facilities. Some have expanded their purpose to include improved
maternal and neonatal health. In these homes additional emphasis is put on education and
nd
edition 2013

Page 39 of 53
counselling regarding pregnancy, delivery and care of the newborn infant and family.
Whereas common sense and anecdotal evidence suggests these are effective in reducing
maternal deaths, a formal evaluation has not been found. For Ekwendeni hospital in
Malawi, Knowles reported that the maternal death rate was reduced to zero, partly
because women have been persuaded to use an antenatal shelter situated about 50 metres
from the delivery ward [28]. SMP reported that maternity waiting huts were available in
4 of 9 CHAM hospitals and none of the government hospitals in the Southern region. The
huts were acceptable for women, but an evaluation showed that staff attitude, supervision
by midwives and provision of cooking and washing facilities needed to be improved if
they were to be fully utilised [141].
5.3.5 Training of traditional birth attendants
Evidence from Malawi suggests that TBAs are well used but not effective in saving lives
[142]. Training of TBAs is intuitively appealing because many women deliver at home,
TBAs are already in the rural areas where women have least access to health facilities,
TBAs are acceptable to the women, they are reimbursed by the women and their families
and training them is relatively cheap [143]. However, up till now, no demonstrable effect
of training TBAs on maternal mortality exists. Two often proposed roles of TBAs in the
reduction of maternal mortality are to educate women about nutrition and hygiene and to
screen them and refer high risk women for medical attention. However, these
interventions have inherent problems that are not solved by moving the process further
into the community. The most valuable training of TBAs would be in the field of
recognising complications and referring the women to EmOC facilities and to train them
and supply them with means to treat some of the complications. The Malawi government
has decided not to train any new TBAs until a revised policy has been agreed.
5.3.6 Improving access to EMOC facilities
The success of the health service over the last two decades in improving the lives of
Malawians has been impressive. Despite limited finances the effectiveness of the health
sector has been profound [144]. The detailed review of the sector in anticipation of the
new Health Sector Strategic Programme (HSSP) revealed that reproductive health gains
had not been as forthcoming as gains in other parts of the health service [39]. MDG5 is
one of the few goals which will not be reaching in 2015, despite the Road Map for
Reproductive Health [41]. This is one of the reasons for the drive to increase the number
and quality of health centres. The plan includes the following assessment of
accessibility:-
2.7.10 Universal access
The MoH is committed to ensuring that services in the EHP are available with universal
coverage for all Malawians. The signing of Service Level Agreements (SLAs) with CHAM
facilities for the delivery of Maternal and Neonatal Health (MNH) services is one way of
ensuring that the services are accessed by everyone regardless of their socio-economic
status. Evidence shows that the removal of user fees in CHAM facilities has resulted in an
increase in the number of patients seeking care in these facilities. Universal coverage
also includes geographical coverage. An analysis of the proportion of Malawis
population living within an 8km radius of a health facility (Annex 3) shows that there are
certain districts that are better served than others.
nd
edition 2013

Page 40 of 53
In some rural places, the health infrastructure is absent or dysfunctional. In others, the
challenge is to provide health support to widely dispersed populations. In high density
urban areas, health services can be physically within reach of the poor and other
vulnerable populations, but provided by unregulated private providers who do not deliver
EHP services.
The plan to improve access is impressive:-
Page 103 - access is planned to be improved from 65% in the first year to
100% in the fifth year with the building of 76 new health centres and the
upgrading of 130 dispensaries to health centres particularly with the intention
of improving the coverage of basic emergency obstetric (BEMOC) services.
The reproductive health services will be transformed if this increase in infrastructure is
mirrored by an improved supply of drugs and supplies (as also planned in the HSSP) and
by staffing as vacancies are filled by well trained midwives (also planned).
Summary - Interventions
Effective interventions currently implemented in Malawi
o Emergency contraception with an adequate regimen (Postinor 2) is currently
available in some centers in the country.
o Minilaparotomy used for tubal sterilisation is an adequate and safe method.
o Active management of third stage of labour reduces risk of PPH with 62%. This
method is included in the midwifery curriculum, but use in practice is uncommon.
o Prophylactic antibiotics for caesarian section reduce risk of endometritis, wound
infection and other serious infectious morbidity by approximately 60%. Malawis
current antibiotic regimen is adequate, but use in practice is unknown.
o Manual Vacuum Aspiration has less secondary effects than dilatation and
curettage and has the advantage that it can be performed at health centre level.
Comprehensive post abortion care programmes (also including counselling and
family planning) were started in 34 centres and coverage is expanding.
o Antibiotic prophylaxis reduces the risk of maternal infectious morbidity in
preterm rupture of membranes by 43%. Malawis recommended antibiotic
regimen is adequate, but again information about use in practice is lacking.
Effective interventions that are implemented but can be improved
Magnesium sulphate is the drug of choice to control eclamptic fits; but diazepam
is still being used at health centre level to control fits before referral.
o BEmOC facilities exist but cannot perform all functions and save lives
o Iron deficiency is common in Malawi and iron supplementation clearly prevents
low haemoglobin at delivery. Not all women at ANC receive iron tablets yet.
o In low parity women, intermittent presumptive malaria therapy reduces severe
antenatal anaemia in the mother and is associated with higher birth weight of the
baby and probably fewer perinatal deaths. Not all women receive antimalarials
yet.
o Improving educational status improves womens access to obstetric care 4-7
times.
o Upgrading radio communication systems for ambulance use increases the
nd
edition 2013

Page 41 of 53
proportion of women with obstetric complications transferred to hospital and
reduces transport time.
o Maternal waiting huts improve access to timely obstetric care.
Effective interventions that are not implemented
o Calcium supplements reduce risk for pre-eclampsia for high risk women or
women in areas with low calcium intake. Calcium intake in Malawi is unknown.
o External Cephalic Version by well trained staff halves the risk of caesarian
section. Method was abandoned in Malawi, after initial negative results.
o Umbilical vein injection with saline + oxytocin reduces the need for manual
removal of placenta by 20%.
o Safe effective abortion services reduce maternal deaths from abortion by 84%.
7 References
1. World Health Organization (1993) International statistical classification of diseases and
related health problems. Vol 2, Instruction manual. World Health Organization. p.
2. Wilmoth JR, Mizoguchi N, Oestergaard MZ, Say L, Mathers CD, et al. (2012) A New
Method for Deriving Global Estimates of Maternal Mortality. Statistics, Politics, and Policy
3.
Available:http://www.who.int/entity/reproductivehealth/publications/monitoring/newMetho
d_deriving_globalEstimates.pdf. Accessed 4 January 2013.
3. World Health Organization. Division of Reproductive Health, UNICEF (1997) The
sisterhood method for estimating maternal mortality: guidance notes for potential users.
Geneva: World Health Organization. 22 p.
4. Immpact (Initiative for Maternal Mortality Programme Assessment) | Home (n.d.).
Available:http://www.immpact-international.org/. Accessed 3 January 2013.
5. Women Deliver: Safe Motherhood, Gender Equality, Girls Education, Poverty Reduction,
Combating HIV/AIDS, Maternal and Newborn Health (n.d.).
Available:http://www.womendeliver.org/. Accessed 4 January 2013.
6. UNFPA - Measuring Progress (n.d.).
Available:http://www.unfpa.org/public/cache/offonce/home/mothers/pid/4389;jsessionid=D
AF694FD726E72E54B0E1C801498E448.jahia01. Accessed 4 January 2013.
7. Technical Resources | www.amddprogram.org (n.d.).
Available:http://www.amddprogram.org/d/content/technical-resources. Accessed 7 January
2013.
8. World Health Organization, United Nations Population Fund, UNICEF, Mailman School of
Public Health. Averting Maternal Death and Disability (2009) Monitoring emergency
obstetric care: a handbook. Geneva: World Health Organization. 152 p.
9. Goodburn EA, Hussein J, Lema V, Damisoni H, Graham W (2001) Monitoring obstetric
services: putting the UN guidelines into practice in Malawi. I: developing the system. Int J
Gynaecol Obstet 74: 105117; discussion 118.
nd
edition 2013

Page 42 of 53
10. Hussein J, Goodburn EA, Damisoni H, Lema V, Graham W (2001) Monitoring obstetric
services: putting the UN Guidelines into practice in Malawi: 3 years on. Int J Gynaecol
Obstet 75: 6373; discussion 74.
11. Ratsma YE (2003) Using process indicators to measure progress. Blantyre, Malawi: Malawi
Safe Motherhood Project. p.
12. Ministry of Health, WHO (2003) Malawi Obstetric Quality of Care Assessment - 2003.
Lilongwe, Malawi. p.
13. Ministry of Health (2005) Assessment of Emergency Obstetric Care Services in Malawi -
2005. Lilongwe, Malawi. p.
14. Kongnyuy EJ, Hofman J, Mlava G, Mhango C, Van den Broek N (2009) Availability,
utilisation and quality of basic and comprehensive emergency obstetric care services in
Malawi. Matern Child Health J 13: 687694. doi:10.1007/s10995-008-0380-y.
15. Ministry of Health, Malawi, UNFPA, UNICEF, WHO, AMDD (2010) Malawi 2010
EmONC Needs Assessment Final Report. Lilongwe, Malawi. p.
16. Ameh C, Msuya S, Hofman J, Raven J, Mathai M, et al. (2012) Status of Emergency
Obstetric Care in Six Developing Countries Five Years before the MDG Targets for
Maternal and Newborn Health. PLoS ONE 7: e49938. doi:10.1371/journal.pone.0049938.
17. World Health Organization, World Bank, UNICEF, United Nations Population Fund (2012)
Trends in maternal mortality: 1990 to 2010: WHO, UNICEF, UNFPA and The World Bank
estimates. Geneva: World Health Organization. 59 p.
18. Loudon I (2000) Maternal mortality in the past and its relevance to developing countries
today. Am J Clin Nutr 72: 241s246s.
19. Chiphangwi JD, Zamaere TP, Graham WJ, Duncan B, Kenyon T, et al. (1992) Maternal
mortality in the Thyolo District of southern Malawi. East Afr Med J 69: 675679.
20. McDermott JM, Slutsker L, Steketee RW, Wirima JJ, Breman JG, et al. (1996) Prospective
assessment of mortality among a cohort of pregnant women in rural Malawi. Am J Trop
Med Hyg 55: 6670.
21. National Statistics Office (NSO) [Malawi] MNS, OCR Macro (1994) Malawi Demographic
and Health Survey, 1992. Calverton, Maryland: NSO and Macro. p.
22. National Statistics Office (NSO) [Malawi] MNS, OCR Macro (2001) Malawi Demographic
and Health Survey, 2000. Calverton, Maryland: NSO and Macro. p.
23. National Statistics Office (NSO) [Malawi], OCR Macro (2005) Malai Demographic and
Health Survey 2004. Calverton, Maryland: NSO and Macro. p.
24. Beltman JJ, Van den Akker T, Lam E, Moens M, Kazima J, et al. (2011) Repetition of a
sisterhood survey at district level in Malawi: the challenge to achieve MDG 5. BMJ Open 1:
e000080. doi:10.1136/bmjopen-2011-000080.
nd
edition 2013

Page 43 of 53
25. National Statistics Office (NSO) [Malawi], OCR Macro (2011) Malai Demographic and
Health Survey 2010. Calverton, Maryland: NSO and Macro. p.
26. Hofman JJ, Sibande NM (2005) Review of community based maternal deaths in Mangochi
district. Malawi Medical Journal 17: 8184.
27. Bullough CH (1981) Analysis of maternal deaths in the Central Region of Malawi. East Afr
Med J 58: 2536.
28. Knowles JK (1988) A shelter that saves mothers lives. World Health Forum 9: 387388.
29. Government of Malawi, WHO (1985) Joint programme review MCH, EPI, PHC. Geneva:
World Health Organisation. p.
30. Keller ME (1987) Maternal mortality at Kamuzu Central Hospital, 1985. Medical Quarterly
(J Med Assoc Malawi) 4: 1316.
31. Driessen F (1990) Driessen F. Maternal deaths in 12 Malawian hospital in 1989.
32. Kempf M (1990) Mulanje Mission Hospital Annual Report 1989. Mulanje, Malawi:
Mulanje Mission Hospital. p.
33. Wiebenga JE (1992) Maternal mortality at Queen Elizabeth Central Hospital, 1989 to 1990.
Malawi Med J 8: 1923.
34. Sangala V (1992) Maternal Deaths in 1990 at Kamuzu Central Hospital. Malawi Medical
Journal 8: 2428.
35. Lema VM, Changole J, Kanyighe C, Malunga EV (2005) Maternal mortality at the Queen
Elizabeth Central Teaching Hospital, Blantyre, Malawi. East Afr Med J 82: 39.
36. Ratsma EY, Lungu K, Hoffman JJ (2005) Why more mothers die: confidential enquiries
into institutional maternal deaths in the Southern region of Malawi, 2001. Malawi Medical
Journal 17: 7580.
37. Kongnyuy EJ, Mlava G, Van den Broek N (2009) Facility-based maternal death review in
three districts in the central region of Malawi: an analysis of causes and characteristics of
maternal deaths. Womens Health Issues 19: 1420. doi:10.1016/j.whi.2008.09.008.
38. Van den Akker T, Van Rhenen J, Mwagomba B, Lommerse K, Vinkhumbo S, et al. (2011)
Reduction of severe acute maternal morbidity and maternal mortality in Thyolo District,
Malawi: the impact of obstetric audit. PLoS ONE 6: e20776.
doi:10.1371/journal.pone.0020776.
39. Ministry of Health (2012) Malawi Health Sector Strategic Plan 2011 - 2016. Lilongwe,
Malawi: Ministry of Health. p.
Available:http://www.medcol.mw/commhealth/publications/3%20Malawi%20HSSP%20Fi
nal%20Document%20%283%29.pdf. Accessed 29 September 2012.
40. Bowie C (2012) A selection of national research undertaken by the College of Medicine.
Blantyre, Malawi: College of Medicine, University of Malawi. p.
nd
edition 2013

Page 44 of 53
Available:http://www.medcol.mw/commhealth/publications/national%20research/national_
research.htm.
41. Ministry of Health (2005) Malawi National Road Map for Accelerating the Reduction of
Maternal and Newborn Mortality. Lilongwe, Malawi. p. Available:http://www.wcf-uk.
org/attachments/article/251/Malawi_Road_Map.pdf. Accessed 6 January 2013.
42. World Health Organization (2005) The World Health Report 2005 make every mother and
child count. Geneva: World Health Organization. p.
Available:http://search.ebscohost.com/login.aspx?direct=true&scope=site&db=nlebk&db=n
labk&AN=134646. Accessed 6 January 2013.
43. Kongnyuy EJ, Mlava G, Van den Broek N (2009) Facility-based maternal death review in
three districts in the central region of Malawi: an analysis of causes and characteristics of
maternal deaths. Womens Health Issues 19: 1420. doi:10.1016/j.whi.2008.09.008.
44. Vink NM, De Jonge HCC, Ter Haar R, Chizimba EM, Stekelenburg J (2013) Maternal
death reviews at a rural hospital in Malawi. Int J Gynaecol Obstet 120: 7477.
doi:10.1016/j.ijgo.2012.07.028.
45. Phillips JA (1990) Ruptured uterus in Mulanje CCAP Hospital, 1974-1982. Trop Doct 20:
175176.
46. Gardeil F, Sangala V (1991) Retrospective study of 600 consecutive admissions with
abortions at Kamuzu Central Hospital. Malawi Med J 7: 113114.
47. Kinoti SN, Gaffikin L, Benson J, Nicholson LA (1996) Monograph on complications of
unsafe abortion in Africa. Commonwealth Regional Health Community Secretariat for East,
Central and Southern Africa. p. Available:http://pdf.usaid.gov/pdf_docs/PNACB975.pdf.
Accessed 3 January 2013.
48. Levandowski BA, Pearson E, Lunguzi J, Katengeza HR (2012) Reproductive health
characteristics of young Malawian women seeking post-abortion care. Afr J Reprod Health
16: 253261.
49. Carroli G, Rooney C, Villar J (2001) How effective is antenatal care in preventing maternal
mortality and serious morbidity? An overview of the evidence. Paediatr Perinat Epidemiol
15 Suppl 1: 142.
50. World Health Organization. Dept. of Reproductive Health and Research, UNAIDS (1999)
HIV in pregnancy: a review. Geneva: World Health Organization. 66 p.
51. Bulterys M, Chao A, Dushimimana A, Saah A (1996) Fatal complications after Cesarian
section in HIV-infected women. AIDS 10: 923924.
52. Thaddeus S, Maine D (1994) Too far to walk: maternal mortality in context. Social science
& medicine 38: 10911110.
53. WHO | Pregnancy, childbirth, postpartum and newborn care (n.d.).
Available:http://www.who.int/maternal_child_adolescent/documents/924159084x/en/index.
html. Accessed 6 January 2013.
nd
edition 2013

Page 45 of 53
54. WHO | WHO guidelines on maternal and reproductive health (n.d.). WHO.
Available:http://www.who.int/publications/guidelines/reproductive_health/en/index.html.
Accessed 28 December 2012.
55. Brabin L, Verhoeff F, Brabin BJ (2002) Maternal height, birthweight and cephalo pelvic
disproportion in urban Nigeria and rural Malawi. Acta Obstet Gynecol Scand 81: 502507.
56. Office NS, Malawi, Surveys D and H, Development I for R, International M (1997) Malawi
Knowledge, Attitudes, and Practices in Health Survey, 1996. Calverton, Maryland: NSO
and Macro. p.
57. Safe Motherhood Project. Qualitative Needs Assessment Report. Zomba, Phalombe and
Chiradzulu (1999). Blantyre, Malawi: Safe Motherhood Project. p.
58. Matinga P, Safe Motherhood Project (2000) Qualitative Needs Assessment (3). Chikwawa,
Mangochi and Thyolo Districts. Safe Motherhood Project. p.
59. Safe Motherhood Project (1998) Safe Motherhood Project. Participatory Needs Assessment
Results. Nsanje District. Blantyre, Malawi: Safe Motherhood Project. p.
60. Kawonga MC (2001) Study results for infection prevention and control practices in health
care institutions in Malawi. National Maternal Mortality Conference, Mangochi.
61. Bokosi M, Ashwood-Smith H, Makawa J (2003) Infection prevention in Southern Malawi:
an effective intervention to lower maternity mortality. Blantyre, Malawi: Malawi Safe
Motherhood Project. Research Abstracts. p.
62. Ogbonnaya Okey P (2002) Situation analysis of hospital infection prevention practices at
Mangochi District Hospital, Malawi [dissertation]. Liverpool School of Tropical Medicine.
p.
63. Safe Motherhood Project (1998) Safe Motherhood Project. Needs Assessment Results.
Blantyre District. Blantyre, Malawi: Safe Motherhood Project. p.
64. Lule GS, Tugumisirize J, Ndekha M (2000) Quality of care and its effects on utilisation of
maternity services at health centre level. East Afr Med J 77: 250255.
65. Chodzaza E, Bultemeier K (2011) Service providers perception of the quality of
emergency obsteric care provided and factors indentified which affect the provision of
quality care. Malawi Medical Journal 22.
Available:http://www.ajol.info/index.php/mmj/article/view/63946. Accessed 3 January
2013.
66. Rawlins BJ, Kim Y-M, Rozario AM, Bazant E, Rashidi T, et al. (2011) Reproductive health
services in Malawi: An evaluation of a quality improvement intervention. Midwifery.
doi:10.1016/j.midw.2011.10.005.
67. Ashwood-Smith H, Simpson H, Ssembatya-Lule G, Matinga P. (2003) Clinical awareness
of obstetric complications among proiders, users and non-users of services. Blantyre,
Malawi: Malawi Safe Motherhood Project. Research Abstracts. p.
nd
edition 2013

Page 46 of 53
68. Rosato M, Mwansambo C, Lewycka S, Kazembe P, Phiri T, et al. (2010) MaiMwana
womens groups: a community mobilisation intervention to improve mother and child
health and reduce mortality in rural Malawi. Malawi Med J 22: 112119.
69. Tolhurst R, Theobald S, Kayira E, Ntonya C, Kafulafula G, et al. (2008) I dont want all
my babies to go to the grave: perceptions of preterm birth in Southern Malawi. Midwifery
24: 8398. doi:10.1016/j.midw.2006.09.003.
70. Graham WJ, Fitzmaurice AE, Bell JS, Cairns JA (2004) The familial technique for linking
maternal death with poverty. Lancet 363: 2327. doi:10.1016/S0140-6736(03)15165-3.
71. Bowie C (2006) Poverty and Health in Malawi. Malawi Medical Journal 18: 512.
72. Machinjili C, Malawi. National Statistical Office (2005) Integrated household survey, 2004-
2005. Zomba, Malawi: National Statistical Office. p.
73. Ashwood-Smith H, Bokosi M, Matinga P, Ngwale M. (2003) Barriers to utilisation of
maternity services in Southern Malawi. Blantyre, Malawi: Malawi Safe Motherhood
Project. Research Abstracts. Blantyre, Project Management Unit, Jan 2003. p.
74. Chen SC-C, Wang J-D, Yu JK-L, Chiang T-Y, Chan C-C, et al. (2011) Applying the global
positioning system and google earth to evaluate the accessibility of birth services for
pregnant women in northern Malawi. J Midwifery Womens Health 56: 6874.
doi:10.1111/j.1542-2011.2010.00005.x.
75. Beltman J, VAN DEN Akker T, VAN Lonkhuijzen L, Schmidt A, Chidakwani R, et al.
(2011) Beyond maternal mortality: obstetric hemorrhage in a Malawian district. Acta Obstet
Gynecol Scand 90: 14231427. doi:10.1111/j.1600-0412.2011.01219.x.
76. Bugge HF, Karlsen NCT, Oydna E, Rake MM, Wexels N, et al. (2013) A study of blood
transfusion services at a district hospital in Malawi. Vox Sang 104: 3745.
doi:10.1111/j.1423-0410.2012.01628.x.
77. Ostergaard LS (2003) A study identifying factors affecting retention of midwives in Malawi
[dissertation]. Liverpool, UK: Liverpool School of Tropical Medicine. p.
78. Bradley S, McAuliffe E (2009) Mid-level providers in emergency obstetric and newborn
health care: factors affecting their performance and retention within the Malawian health
system. Hum Resour Health 7: 14. doi:10.1186/1478-4491-7-14.
79. Maluwa V (2008) Moral Distress In Nursing Practice In Selected Health Institutions In
Lilongwe District, Malawi Veronica Maluwa. Blantyre, Malawi: College of
Medicine, University of Malawi. p.
80. Manafa O, McAuliffe E, Maseko F, Bowie C, Maclachlan M, et al. (2009) Retention of
health workers in Malawi: perspectives of health workers and district management. Hum
Resour Health 7: 65. doi:10.1186/1478-4491-7-65.
81. McAuliffe E, Bowie C, Manafa O, Maseko F, Maclachlan M, et al. (2009) Measuring and
managing the work environment of the mid-level provider - the neglected human resource.
Hum Resour Health 7: 13. doi:10.1186/1478-4491-7-13.
nd
edition 2013

Page 47 of 53
82. McAuliffe E, Manafa O, Maseko F, Bowie C, White E (2009) Understanding job
satisfaction amongst mid-level cadres in Malawi: the contribution of organisational justice.
Reprod Health Matters 17: 8090. doi:10.1016/S0968-8080(09)33443-6.
83. Mtengezo J (2007) Job satisfaction: factors influencing job satisfaction of nurses at
Machinga District Hospital. College of Medicine, Blantyre, Malawi. p.
Available:http://www.medcol.mw/commhealth/mph/dissertations/mph_dissertations.htm.
84. Ministry of Health, Health SWAp Donor Group, GTZ (2007) Human Resources - Capacity
Development within the Health Sector - Needs Assessment Study. Lilongwe, Malawi. p.
Available:http://www.medcol.mw/commhealth/publications/national%20research/Malawi_
HealthSWAp_HR-CD-NeedsAssessment_GTZ_FinalReport_v09.pdf. Accessed 15
September 2009.
85. Maclean G (2000) Evaluation of life saving skills course for the Malawi Safe Motherhood
Project. JSI, 2000. Blantyre, Malawi: Safe Motherhood Project. p.
86. Ministry of Health and Population, Planning Department (2003) Health Management
Information Unit. Health Management Information Bulletin. Annual Report, July 2002
June 2003. Lilongwe, December 2003. Lilongwe, Malawi: Health Management Information
Unit, MOH. p.
87. Ministry of Health and Population, Planning Department (2010) Health Management
Information Unit. Health Management Information Bulletin. Annual Report, July 2008
June 2009. Lilongwe, Malawi: Health Management Information Unit, MOH. p.
88. Bicego G, Boerma JT, Ronsmans C (2002) The effect of AIDS on maternal mortality in
Malawi and Zimbabwe. AIDS 16: 10781081.
89. Government of Malawi, United Nations (1993) Situation analysis of poverty in Malawi.
Lilongwe, Malawi: Government of Malawi. p.
90. Begley CM, Gyte GML, Devane D, McGuire W, Weeks A (2011) Active versus expectant
management for women in the third stage of labour. Cochrane Database Syst Rev:
CD007412. doi:10.1002/14651858.CD007412.pub3.
91. Duley L, Matar HE, Almerie MQ, Hall DR (2010) Alternative magnesium sulphate
regimens for women with pre-eclampsia and eclampsia. Cochrane Database Syst Rev:
CD007388. doi:10.1002/14651858.CD007388.pub2.
92. Dowswell T, Carroli G, Duley L, Gates S, Glmezoglu AM, et al. (2010) Alternative versus
standard packages of antenatal care for low-risk pregnancy. In: The Cochrane
Collaboration, Duley L, editors. Cochrane Database of Systematic Reviews. Chichester,
UK: John Wiley & Sons, Ltd. Available:http://hinari-
gw.who.int/whalecomonlinelibrary.wiley.com/whalecom0/doi/10.1002/14651858.CD00093
4.pub2/abstract;jsessionid=7D5602BAB2EA16C63A4AC56A636521B1.d01t04. Accessed
28 December 2012.
93. Smaill FM, Gyte GM (2010) Antibiotic prophylaxis versus no prophylaxis for preventing
infection after cesarean section. Cochrane Database Syst Rev: CD007482.
doi:10.1002/14651858.CD007482.pub2.
nd
edition 2013

Page 48 of 53
94. Kenyon S, Boulvain M, Neilson JP (2010) Antibiotics for preterm rupture of membranes.
Cochrane Database Syst Rev: CD001058. doi:10.1002/14651858.CD001058.pub2.
95. Buppasiri P, Lumbiganon P, Thinkhamrop J, Ngamjarus C, Laopaiboon M (2011) Calcium
supplementation (other than for preventing or treating hypertension) for improving
pregnancy and infant outcomes. Cochrane Database Syst Rev: CD007079.
doi:10.1002/14651858.CD007079.pub2.
96. Pea-Rosas JP, De-Regil LM, Dowswell T, Viteri FE (2012) Daily oral iron
supplementation during pregnancy. Cochrane Database Syst Rev 12: CD004736.
doi:10.1002/14651858.CD004736.pub4.
97. Garner P, Glmezoglu AM (2006) Drugs for preventing malaria in pregnant women.
98. Garner P, Glmezoglu AM (2006) Drugs for preventing malaria in pregnant women. In:
The Cochrane Collaboration, Garner P, editors. Cochrane Database of Systematic Reviews.
Chichester, UK: John Wiley & Sons, Ltd. Available:http://hinari-
gw.who.int/whalecomonlinelibrary.wiley.com/whalecom0/doi/10.1002/14651858.CD00016
9.pub2/abstract. Accessed 2 January 2013.
99. Haider BA, Humayun Q, Bhutta ZA (2009) Effect of administration of antihelminthics for
soil transmitted helminths during pregnancy. Cochrane Database Syst Rev: CD005547.
doi:10.1002/14651858.CD005547.pub2.
100. De-Regil LM, Fernndez-Gaxiola AC, Dowswell T, Pea-Rosas JP (2010) Effects and
safety of periconceptional folate supplementation for preventing birth defects. Cochrane
Database Syst Rev: CD007950. doi:10.1002/14651858.CD007950.pub2.
101. Hofmeyr GJ, Kulier R (2012) External cephalic version for breech presentation at term.
Cochrane Database Syst Rev 10: CD000083. doi:10.1002/14651858.CD000083.pub2.
102. Pea-Rosas JP, De-Regil LM, Dowswell T, Viteri FE (2012) Intermittent oral iron
supplementation during pregnancy. Cochrane Database Syst Rev 7: CD009997.
doi:10.1002/14651858.CD009997.
103. Mathanga DP, Uthman OA, Chinkhumba J (2011) Intermittent preventive treatment
regimens for malaria in HIV-positive pregnant women. Cochrane Database Syst Rev:
CD006689. doi:10.1002/14651858.CD006689.pub2.
104. Cheng L, Che Y, Glmezoglu AM (2012) Interventions for emergency contraception.
Cochrane Database Syst Rev 8: CD001324. doi:10.1002/14651858.CD001324.pub4.
105. Cluver C, Hofmeyr GJ, Gyte GM, Sinclair M (2012) Interventions for helping to turn term
breech babies to head first presentation when using external cephalic version. Cochrane
Database Syst Rev 1: CD000184. doi:10.1002/14651858.CD000184.pub3.
106. Duley L, Glmezoglu AM, Henderson-Smart DJ, Chou D (2010) Magnesium sulphate and
other anticonvulsants for women with pre-eclampsia. Cochrane Database Syst Rev:
CD000025. doi:10.1002/14651858.CD000025.pub2.
nd
edition 2013

Page 49 of 53
107. Duley L, Henderson-Smart DJ, Walker GJ, Chou D (2010) Magnesium sulphate versus
diazepam for eclampsia. Cochrane Database Syst Rev: CD000127.
doi:10.1002/14651858.CD000127.pub2.
108. Kulier R, Kapp N, Glmezoglu AM, Hofmeyr GJ, Cheng L, et al. (2011) Medical methods
for first trimester abortion. Cochrane Database Syst Rev: CD002855.
doi:10.1002/14651858.CD002855.pub4.
109. Siegfried N, Irlam JH, Visser ME, Rollins NN (2012) Micronutrient supplementation in
pregnant women with HIV infection. Cochrane Database Syst Rev 3: CD009755.
doi:10.1002/14651858.CD009755.
110. Kulier R, Boulvain M, Walker D, Candolle G, Campana A (2004) Minilaparotomy and
endoscopic techniques for tubal sterilisation. Cochrane Database Syst Rev: CD001328.
doi:10.1002/14651858.CD001328.pub2.
111. Haider BA, Bhutta ZA (2012) Multiple-micronutrient supplementation for women during
pregnancy. Cochrane Database Syst Rev 11: CD004905.
doi:10.1002/14651858.CD004905.pub3.
112. Kramer MS, Kakuma R (2012) Optimal duration of exclusive breastfeeding. Cochrane
Database Syst Rev 8: CD003517. doi:10.1002/14651858.CD003517.pub2.
113. Brown HC, Paranjothy S, Dowswell T, Thomas J (2008) Package of care for active
management in labour for reducing caesarean section rates in low-risk women. Cochrane
114. Tunalp O, Glmezoglu AM, Souza JP (2010) Surgical procedures for evacuating
incomplete miscarriage. Cochrane Database Syst Rev: CD001993.
doi:10.1002/14651858.CD001993.pub2.
115. Abdel-Aleem H, Abdel-Aleem MA, Shaaban OM (2011) Tocolysis for management of
retained placenta. Cochrane Database Syst Rev: CD007708.
doi:10.1002/14651858.CD007708.pub2.
116. Nardin JM, Weeks A, Carroli G (2011) Umbilical vein injection for management of
retained placenta. Cochrane Database Syst Rev: CD001337.
doi:10.1002/14651858.CD001337.pub2.
117. Oliveira-Menegozzo JM, Bergamaschi DP, Middleton P, East CE (2010) Vitamin A
supplementation for postpartum women. Cochrane Database Syst Rev: CD005944.
doi:10.1002/14651858.CD005944.pub2.
118. De-Regil LM, Palacios C, Ansary A, Kulier R, Pea-Rosas JP (2012) Vitamin D
supplementation for women during pregnancy. Cochrane Database Syst Rev 2: CD008873.
doi:10.1002/14651858.CD008873.pub2.
119. Mori R, Ota E, Middleton P, Tobe-Gai R, Mahomed K, et al. (2012) Zinc supplementation
for improving pregnancy and infant outcome. Cochrane Database Syst Rev 7: CD000230.
doi:10.1002/14651858.CD000230.pub4.
nd
edition 2013

Page 50 of 53
120. WHO | Reproductive Health Library (n.d.). Available:http://apps.who.int/rhl/en/. Accessed
8 January 2013.
121. WHO | Essential Interventions, Commodities and Guidelines for Reproductive, Maternal,
Newborn and Child Health (n.d.). WHO.
Available:http://www.who.int/pmnch/topics/part_publications/201112_essential_interventio
ns/en/index.html. Accessed 8 January 2013.
122. WHO | Packages of interventions (n.d.). WHO.
Available:http://www.who.int/reproductivehealth/publications/maternal_perinatal_health/fc
h_10_06/en/index.html. Accessed 6 January 2013.
123. Mahomed K (2006) WITHDRAWN: Iron supplementation in pregnancy. Cochrane
124. Mahomed K (2006) WITHDRAWN: Iron and folate supplementation in pregnancy.
125. Hofmeyr GJ, Lawrie TA, Atallah AN, Duley L (2010) Calcium supplementation during
pregnancy for preventing hypertensive disorders and related problems. Cochrane Database
Syst Rev: CD001059. doi:10.1002/14651858.CD001059.pub3.
126. Hofmeyr GJ, Kulier R (2012) Cephalic version by postural management for breech
presentation. Cochrane Database Syst Rev 10: CD000051.
doi:10.1002/14651858.CD000051.pub2.
127. Prendiville WJ, Elbourne D, McDonald S (2000) Active versus expectant management in
the third stage of labour. Cochrane Database Syst Rev: CD000007.
doi:10.1002/14651858.CD000007.
128. Hopkins L, Smaill FM (2012) WITHDRAWN: Antibiotic prophylaxis regimens and drugs
for cesarean section. Cochrane Database Syst Rev 1: CD001136.
doi:10.1002/14651858.CD001136.pub2.
129. Forna F, Glmezoglu AM (2001) Surgical procedures to evacuate incomplete abortion.
Cochrane Database Syst Rev: CD001993. doi:10.1002/14651858.CD001993.
130. Lopez LM, Hiller JE, Grimes DA, Chen M (2012) Education for contraceptive use by
women after childbirth. Cochrane Database Syst Rev 8: CD001863.
doi:10.1002/14651858.CD001863.pub3.
131. Coombes Y (2001) Information, education, communication and behaviour change
Evidence to support a safe motherhood campaign in Malawi. Blantyre, Malawi: Malawi
Safe Motherhood Project. p.
132. Ngwale M and Safe Motherhood Project (2001) Repeat participatory needs assessment
survey. Blantyre, Malawi: Safe Motherhood Project. p.
133. Maine D, Rosenfield A (1999) The Safe Motherhood Initiative: why has it stalled? Am J
Public Health 89: 480482.
nd
edition 2013

Page 51 of 53
134. McCarthy J, Maine D (1992) A framework for analyzing the determinants of maternal
mortality. Studies in family planning 23: 2333.
135. Malawi Human rights Commission (2006) Cultural practices and their impact on the
enjoyment of human rights, particularly the rights of women and children in Malawi.
Blantyre, Malawi: Malawi Human Rights Commission. p.
136. Grimes DA, Benson J, Singh S, Romero M, Ganatra B, et al. (2006) Unsafe abortion: the
preventable pandemic. Europe 500: 3.
137. Stephenson P, Wagner M, Badea M, Serbanescu F (1992) Commentary: the public health
consequences of restricted induced abortion--lessons from Romania. Am J Public Health
82: 13281331.
138. Lunga K, Kamphose V HJ, Ashwood-Smith H (2001) Are bicycles ambulances and
community transport plans effective? Malawi Medical Journal 13: 1618.
139. Lunga K, Ratsma YE (2007) Does the upgrading of the radio communications network in
health facilities reduce the delay in the referral of obstetric emergencies in Southern
Malawi. Malawi Medical Journal 19: 18.
140. Hofman JJ, Dzimadzi C, Lungu K, Ratsma EY, Hussein J (2008) Motorcycle ambulances
for referral of obstetric emergencies in rural Malawi: Do they reduce delay and what do
they cost? International Journal of Gynecology & Obstetrics 102: 191197.
141. Ashwood-Smith H, Bokosi M, Ratsma E (2003) Maternity waiting huts in Southern
Malawi. Lilongwe, Malawi: National Health Research Dissemination Conference,
Lilongwe. p.
142. Bisika T (2008) The effectiveness of the TBA programme in reducing maternal mortality
and morbidity in Malawi. East Afr J Public Health 5: 103110.
143. Chen SC-C, Wang J-D, Ward AL, Chan C-C, Chen P-C, et al. (2011) The effectiveness of
continuing training for traditional birth attendants on their reproductive health-care
knowledge and performance. Midwifery 27: 648653. doi:10.1016/j.midw.2009.12.008.
144. Bowie C, Mwase T (2011) Assessing the use of an essential health package in a sector wide
approach in Malawi. Health Res Policy Syst 9: 4. doi:10.1186/1478-4505-9-4.

Chapter 5 The epidemiology of maternal mortality v2
Appendix 1 One page summary of WHOs essential interventions
Page 52 of 53

Chapter 5 The epidemiology of maternal mortality v2
Appendix 1 One page summary of WHOs essential interventions
Page 53 of 53

Epidemiology of Maternal Mortality in Malawi - 2 Edition

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Epidemiology of Maternal Mortality in Malawi - 2 Edition

Încărcat de

Drepturi de autor:

Formate disponibile

Chapter 5 The epidemiology of maternal mortality 2

S-ar putea să vă placă și