Midwifery 30 (2014) 512518

Contents lists available at ScienceDirect

Midwifery
journal homepage: www.elsevier.com/midw

Estimating maternal mortality and causes in South Africa:

National and provincial levels
Eric O. Udjo, PhD (Research Director, Demographic Research Division)a,n,
Pinky Lalthapersad-Pillay, PhD (Professor of Economics)b
a
b

Bureau of Market Research, University of South Africa, P.O. Box 392, UNISA 0003, Pretoria, South Africa
Department of Economics, University of South Africa, P.O. Box 392, UNISA 0003, Pretoria, South Africa

art ic l e i nf o

a b s t r a c t

Article history:
Received 16 October 2012
Received in revised form
20 May 2013
Accepted 27 May 2013

Objectives: maternal mortality estimates for South Africa have methodological weaknesses. This study
uses the Growth Balance Method to adjust reported household female deaths and pregnancy-related
deaths and the relational Gompertz model to adjust reported number of live births and estimate
maternal mortality in South Africa at national and provincial level; examines the potential impact of
HIV/AIDS prevalence; and investigates the recorded direct causes of maternal mortality.
Design: data from the 2001 Census, 2007 Community Survey and death registrations were utilised.
Information on household deaths, including pregnancy-related deaths was collected from the aforementioned census and survey.
Setting: enumerated households in the 2001 Census and a nationally representative sample of 250,348
households in the 2007 Community Survey.
Participants: information about members of households who died in the preceding 12 months was
collected, and of these deaths whether there were women aged 1549 who died while pregnant or
within 42 days after childbirth.
Findings: maternal mortality ratio of 764 per 100,000 live births in 2007, ranging from 102 per 100,000 live
births in the Western Cape province to 1639 in the Eastern Cape. Maternal infections and parasitic diseases as
well as other maternal diseases complicating pregnancy, childbirth and the puerperium are the major causes.
The study found a weak correlation between provincial HIVprevalence and maternal mortality ratio.
Conclusion: despite strategies to improve maternal and child health, maternal mortality remains high in South
Africa and it is unlikely that the Millennnium Developmemnt Goal of reducing maternal will be achieved.
& 2013 Elsevier Ltd. All rights reserved.

Keywords:
Pregnancy-related deaths
Maternal mortality ratio
Causes of death
South Africa

Introduction
Maternal mortality is specically an indicator of reproductive
health and socio-economic development in general. The Safe Motherhood Initiative was partly to reduce maternal mortality (Lilijestrand
and Pathmanathan, 2004). Globally maternal deaths decreased by 47%
between 1990 and 2010 (World Health Organization (WHO), 2012).
Shah and Say (2004) have provided the following maternal mortality
ratios (per 100,000 live births): globally in 2005, 400; Sub-Saharan
Africa: 920 in 1990 and 900 in 2005; South East Asia: 450 in 1990 and
300 in 2005; developed regions: 11 in 1990 and 9 in 2005. Despite the
decreasing levels of maternal mortality globally and in Africa in
general, they remain relatively unchanged in Southern African countries (Botswana, Lesotho, Namibia, South Africa and Swaziland). Trend
in maternal mortality ratio (250 in 1990 and 360 in 2005) in Southern
Africa (World Health Organization (WHO), 2012) is in adverse contrast
to international trends. The issue of maternal mortality has been

Corresponding author.
E-mail addresses: udjoe@unisa.ac.za, bororue@yahoo.com.(E.O. Udjo)

0266-6138/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.midw.2013.05.011

thrust again to the fore by its inclusion in the Millennium Development Goals (MDGs) and has provided further impetus to studies on
maternal mortality in recent years (see for example, Human Rights
Watch, 2011; Horton, 2012; Hsu et al., 2012).
Changes in health legislation, health policy and delivery of health
services in post-apartheid South Africa have led to reforms in
reproductive health (National Committee for the Condential
Enquiry into Maternal Deaths, 1998; Cooper et al., 2004). Despite
these reforms, the high rate of maternal mortality is one of the
country's major population concerns (Department of Welfare, 1998).
Rationale for the study
The rationale for this study is as follows:
(1) There is dearth of reliable estimates for monitoring maternal
mortality in South Africa. Although several studies have
provided estimates, the studies have weaknesses. Garenne
et al.'s study (2008) used the general pattern of the UN model
life table system in estimating and assessing the plausibility of

E.O. Udjo, P. Lalthapersad-Pillay / Midwifery 30 (2014) 512518

their estimates of maternal mortality for South Africa. This life

table is inappropriate for South Africa as it disregards the effects of
the HIV/AIDS epidemic (see Udjo (2008)). Furthermore, they
suggest that mortality in the preceding 12 months to the 2001
Census (Statistics South Africa, 2003) was overestimated whereas
the application of the Growth Balance Method to the data shows
that deaths were underreported. Also, they noted that the number
of births they projected backwards appeared to be low and argued
that the number of women who delivered in the preceding
12 months appeared to be high, which compelled them to make
adjustments. The study on which this article is based shows that it
was not about too many women giving birth in the preceding
12 months but rather that some births that occurred in the
reference period were not reported. A follow-up study by
Garenne et al. (2009) has similar limitations. Although they
provided condence intervals for their estimates, the condence
intervals do not resolve the biases in their estimates.
One of the sources utilised in the Hogan et al. (2010) study was
vital registration data. Their estimates for South Africa are biased
as they did not adjust for incomplete registration that is often the
case in vital registration data (see Brass (1971), Hill (1987) and
Udjo (2006)). Secondly, an examination of the UN Population
Division data base used in their estimates indicated that, in the
case of South Africa, the denominator in their estimate for 2008
was the number of live births obtained from the 2007 Community
Survey (Statistics South Africa, 2007b). Besides the fact that the
universe of the estimates (2008 death registration) and reported
live births in the 2007 Community Survey are different, the
number of live births (in the preceding 12 months) was not
adjusted for reference period error.
WHO, UNICEF, UNFPA and the World Bank (2010) estimates used
global adjustment factors for a group of countries (including South
Africa) they considered lacking good vital registration, to adjust for
mis-classication and incomplete registration. Such global adjustments may produce estimates that are either too low or too high
as some countries may lie towards the extreme end of the median
value. Also, the magnitude of underreporting of deaths tends to
get smaller over time due to improvements in the registration
system (Udjo, 2006). In providing additional explanation on the
WHO, UNICEF, UNFPA and the World Bank estimates, Wilmoth
et al. (2012) noted that the evidentiary basis underlying the
assumptions in the estimates is fairly weak. They further noted
that the model underlying the WHO, UNICEF, UNFPA and World
Bank estimates is clearly an enormous simplication of reality.
In view of these limitations, a different approach is needed to
determine maternal mortality levels in South Africa to impart a
better understanding of their magnitudes.
(2) Both national and provincial estimates of maternal mortality are
needed to ensure that suitable interventions are appropriately
targeted. Most studies provide estimates at national level.
(3) There are challenges when using currently available data, such as
changes to provincial boundaries (in 2005, 2008 and 2011) and
the possibility that the high prevalence of HIV/AIDS may mask
other issues contributing to maternal mortality. HIV/AIDS is likely
to be an indirect cause of maternal death rather than a direct
cause. It is important to distinguish between direct and indirect
causes of maternal death.

Study objectives
The objectives of this study therefore are (1) to use the Growth
Balance Method to adjust reported household female deaths and
pregnancy-related deaths, and the relational Gompertz model to
adjust reported number of live births so as to provide estimates of
maternal mortality in South Africa at national and provincial

513

levels; (2) to examine the potential impact of HIV/AIDS prevalence

on maternal mortality at provincial levels in South Africa; and
(3) to examine the recorded direct causes of maternal mortality in
South Africa.

Methods
Data and subjects
Censuses, sample surveys and vital registration are the primary
sources of nationally and provincially representative mortality data.
Censuses and surveys usually do not provide information on perinatal
mortality but they may be obtained from vital registration. Censuses,
sample surveys and vital registration in South Africa have several
weaknesses as is the case in many other countries. Besides coverage
errors in censuses, these sources of data have content errors and, in
the context of maternal mortality, include underreporting of deaths as
well as errors in the number of live births. The District Health
Information System (DHIS) in South Africa collects mortality data
but these are hospital-based and, therefore, cannot be used to generate
nationally and provincially representative mortality estimates. Furthermore, the quality of the DHIS data varies from one province to another.
Thus, estimates of maternal mortality in South Africa rely heavily on
censuses and surveys and to a lesser extent, on vital registration.
This study therefore utilised the 2001 South African Census, 2007
Community Survey and Death registrations for 1997, 2001 and 2007
(Statistics South Africa, 2007a). The 69 questions in the 2001 Census
covered demographic and socio-economic proles of the population
and households. The overall undercount in the 2001 Census was 18%.
The 2007 Community Survey was the largest sample survey ever
conducted by Statistics South Africa. The objectives of the survey were
(a) to provide information at lower geographical levels; (b) to build
human, management and logistical capacities towards the 2011
Census; and (c) to facilitate the linkage between the establishment
of the national address system and database of dwelling units. A twostage stratied cluster sampling method, comprising 947,331 individuals from 250,348 households, was used. Institutions were excluded
from the sampling. There were 88 questions in the 2007 Community
Survey and like the 2001 Census, it covered demographic and socioeconomic proles of the population and households. The overall
response rate in the 2007 Community Survey was 93.9%. (Statistics
South Africa, the government department responsible for ofcial
statistics in South Africa, instituted the 2001 Census and the 2007
Community Survey. Statistics South Africa makes census and survey
data available to the public through various media soon after the
results have been ofcially released).
Regarding mortality, and aside from the orphanhood questions,
the 2001 Census and the 2007 Community Survey included questions
about the number of deaths in the household in the preceding
12 months, sex of the deceased, age of the deceased and whether the
cause of death was natural or unnatural. If the deceased was a female
aged 1250 the question was posed as to whether she was pregnant
at the time of death or whether death occurred within six weeks
after childbirth. The date of the last live birth was also asked of
females aged 1250 years. These questions constituted the basis for
the computation of maternal mortality ratios.
The analysis of the direct causes of maternal deaths was based
on the death records for 1997, 2001 and 2007. Death records
in South Africa are obtained through vital registration. The medical
certication includes the immediate and underlying causes of
death. The certication forms are processed by Statistics South
Africa using ICD-10 coding. An inherent problem in vital registration data in less developed countries hinges on completeness of
registration. The adjustments carried out on the data are described
in another section of this article.

514

E.O. Udjo, P. Lalthapersad-Pillay / Midwifery 30 (2014) 512518

Denition of maternal mortality

WHO denes a maternal death as 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 (Graham, 1991: 102). The questions in
the 2001 Census and 2007 Community Survey were phrased as
follows: 2001 Census: Has any member of this household died in the
past 12 months, i.e. between 10 October 2000 and 10 October 2001
If the deceased was a woman under 50 years, did (the person) die
while pregnant or within six weeks after delivery and 2007 Community Survey: Has any member of this household passed away in
the last 12 months between February 2006 and March 2007? The
estimates of maternal mortality in this study were based on the
pregnancy-related deaths computed from the above questions
adjusted for a 42 day puerperium. The magnitude of the direct
causes of maternal mortality advanced by the authors of this article
was based on the ICD-10 coding as provided by Statistics South Africa
on registered deaths.
Estimating maternal mortality
If the data are perfect, unadjusted maternal mortality ratio,
MMR is computed as
MMR

D
k
B

where MMR is the period maternal mortality ratio, D is the

number of maternal deaths in the period, B is the number of live
births in the period whereas k is a constant, usually 100,000. But
because the data are not perfect, adjusted maternal mortality ratio
MMR is estimated as
MMR

D
k
B

where D is the adjusted number of pregnancy-related deaths in

the period, and B is the adjusted number of live births in the
period. Thus, the steps involved in estimating the maternal
mortality ratios in this study are as follows.
Regarding the numerator of the MMR, rstly, the numbers of
reported household female deaths and pregnancy-related deaths
in the reference period were tabulated by ve-year age group
nationally and provincially (based on the 2001 provincial boundaries) from the 2001 Census and 2007 Community Survey data.
Completeness of reporting of household female deaths nationally and provincially was then assessed using the Growth Balance
Method so as to derive an adjustment factor for the number of
pregnancy-related deaths (see Box 1 for details).
Regarding the denominator of the MMR, rstly, the numbers of
women and reported live births in the preceding 12 months were
tabulated by reproductive ve-year age group nationally and provincially from the 2001 Census and 2007 Community Survey data. Next,
the relational Gompertz model was tted to the data to assess the
accuracy and adjust the reported number of live births (see Box 2 for
details). Lastly, MMR was computed based on the adjusted numbers.

Findings
Estimated completeness of reporting of deaths in households using
the Growth Balance Method
Fig. 1 illustrates the result of the application of the Growth Balance
Method to the reports on female deaths at national level in 2001. The
scatter plot shows deviation from a straight line at some ages,
indicating errors in the completeness of reporting of female deaths.

A similar pattern was observed at provincial levels. The estimated

completeness of reporting is summarised in Table 1. The estimates for
the Free State, Northern Cape and the Western Cape should be treated
with scepticism as they imply large adjustments of maternal mortality
ratios upward in the three provinces. The estimates indicate that the
completeness of reporting of female deaths in households was higher
in the 2001 Census compared to the 2007 Community Survey
nationally and provincially except in the Northern Cape and Western
Cape provinces where the reverse was the case. The differences in the
completeness of reporting of female deaths in the 2001 Census and
the 2007 Community Survey were statistically signicant nationally
and provincially (po0.05) except in the Northern Cape and North
West provinces (p40.05).
Assessment and adjustment of reported births in the preceding
12 months
If there were no errors in the reporting of births in the
preceding 12 months, all the points in the scatter plot in Fig. 2
(depicting the national level analysis) should lie in a straight line
but as seen in the graph, some of the points deviate from a straight
line. The points to the left of the tted line indicate omission of
some births in the preceding 12 months whereas the points
corresponding to the oldest reproductive age groups indicate age
errors. A similar pattern was observed at provincial level. The
observed and adjusted total fertility rates summarised in Table 2
indicate that births in the preceding 12 months were underreported during the 2001 Census and 2007 Community Survey
(with the exception of the Eastern Cape in 2007). The magnitude
of the underreporting was greater in the 2001 Census compared
with the 2007 Community Survey. Statistical tests of the
Box 1. Estimating adjusted number of household and
maternal deaths, D using the Growth Balance Method
The Growth Balance Method by Brass (1971) was designed to
estimate the completeness of, and hence adjust for incomplete or overreporting of deaths in households during a
census or survey. This is based on the linear relationship
between deaths and age distributions expressed as
Nx =Nx r Dx =Nx

where N(x) is the number of persons at exact age x, N(x+) is the

total number of persons above age x, D(x+) is the total number
of deaths occurring to persons aged x and over and r is the
growth rate. Since there would be some error patterns in the
completeness of death distribution, the equation may be
rewritten as
Nx =Nx r kDx =Nx

where r (the intercept of the straight line tted to the data) is an

estimate of the growth rate, k (the slope of the tted line) is a
coefcient that is used to adjust for the completeness of death
reporting. The completeness of death reporting c, is estimated as
c 1=k

c is for all causes of deaths and was used to adjust the reported
number of pregnancy-related deaths. The procedure involved the
following: (a) N(x)/N(x+) and D(x+)/N(x+) values were computed from the tabulated age distribution of the population of
females and reported female deaths, (b) the computed N(x)/N(x
+) values were plotted against the computed D(x+)/N(x+)
values, (c) a straight line was tted to the best points
(i.e. ignoring outliers) of the plotted values; (d) r and k were
determined from the tted line using the least squares method;
(e) the reciprocal of k was then used to adjust the reported
number of pregnancy-related deaths.

E.O. Udjo, P. Lalthapersad-Pillay / Midwifery 30 (2014) 512518

0.12

Box 2. Estimating adjusted number of live births, B using the

relational Gomperzt model

F x F e

e abY sx 

where F(x) is the cumulated age-specic fertility rate up to age

x, and F is the total fertility rate. Ys(x) is dened as ln[ln Fs
(x)/F] where Fs(x) is a standard cumulative fertility rate up to
age x (Brass, 1981). The and parameters measure the location
and spread of the fertility distribution (Brass, 1981). In the usual
application, F is unknown but can be separated from the
estimation of and with 1 using a linear transformation (Zaba,
1981) from the following equation:
zx ex a 0:48b12 bgx

0.1

N(x)/N(x+)

The number of live births to women aged 1549 from the

information on the reported date the last live birth child was
born, may be under- or overreported (reference period error:
see Brass (1971)). The relational Gompertz model was
designed to detect and adjust for such errors. The model is
expressed as

515

0.08
0.06
0.04
0.02
0
0.00

0.02

0.04

2001
Census

2007
Community
Survey

where f(x) is the decumulated model age-specic fertility rate

for women aged x and F. F(x) is the model cumulative fertility
up to age x. The adjusted number of live births for a specic
period was estimated as
B

49

w x f x

x 15

where B is the adjusted number of live births and w(x) is the

number of women aged x in a ve-year reproductive age group.
The procedure involved in adjusting the reported number of
live births therefore was as follows: (a) ln[ln F(x)/F(x+5)]
values were computed from the distribution of reported
number of live births in the last 12 months by age of women
to obtain observed z values; (b) the observed z values were
plotted against standard z values; (c) a straight line was tted to
the points using the group average method; (d) and were
computed algebraically from the tted line and applied to the
standard series of z's to obtain a single estimate of F (by
averaging the series of F's that were of consistent level);
(e) model age-specic fertility rates were computed using the
estimated , and and applied to the standard series of z's, and
(f) the model age-specic fertility rates were applied to the
distribution of the number of women to obtain adjusted
number of live births.

differences indicate signicantly lower levels of reported fertility

than estimated at national and provincial levels (p o0.05) except
in the Eastern Cape in 2007, where the reported level of fertility
was signicantly higher (p o0.05) than the estimated level.
Estimated maternal mortality ratios
Table 3 shows the estimated maternal mortality ratios after
adjustments for errors and suggest that maternal mortality ratio
increased from about 473 per 100,000 live births in 2001 to about
764 per 100,000 live births in 2007. This indicates an annual
increase of about 11% between 2001 and 2007. At provincial level,
the results indicate that maternal mortality increased between

South Africa
Eastern Cape
Free State
Gauteng
KwaZulu-Natal
Limpopo
Mpumalanga
Northern Cape
North West
Western Cape

74.5
75.0
60.2
72.4
73.2
75.3
74.7
59.5
70.9
61.6

0.12

Statistical test of
difference in
completeness
of reporting between
2001 and 2007: z value
16.9n
18.4n
5.3n
3.71n
17.4n
9.6n
23.2n
1.4
1.5
29.2n

73.1
71.0
58.1
71.7
70.0
72.8
67.5
60.1
70.4
69.6

n
Statistically signicant (p o 0.05): standard normal table value at 95% level of
condence 1.96.

6
5
4
3

z(x)-(ex)

0.10

Table 1
Estimated percentage completeness of reporting of female deaths in household by
province, 2001 and 2007.
Source: Computed from 2001 Census data.

where F(x) is the cumulated ASFR up to age x, e(x) and g(x) are
tabulated standard values; 0.48 is a constant. Model ASFRs were
estimated as
f x F F x

0.08

Fig. 1. Plot of partial birth rates, N(x)/N(x+), against partial death rates, D(x+)/N(x+),
2001: Females. .
Source: Computed from 2001 Census data.

where
zx lnlnF x =F x 5

0.06

D(x+)/N(x+)

2
1
0
-4

-2

-1

-2

g(x)

-3
2001 Census

2007 CS

2007 CS Fitted

Fig. 2. Fitting the relational Gompertz model to current fertility, 2001 and 2007:
National .
Source: Computed from 2001 census and 2007 Community Survey (2007 CS).

2001 and 2007 in all provinces except in the Western Cape. In

2001, the province with the highest level of maternal mortality
was the Free State but in 2007, it had shifted to the Eastern Cape.
In both years, the Western Cape had the lowest maternal mortality
ratio and was the only province that experienced a decline
between 2001 and 2007.
HIV/AIDS and maternal mortality in South Africa
If HIV/AIDS was responsible for the increasing level of maternal
mortality in South Africa, then the provincial distribution of HIV
prevalence should be similar to the provincial distribution of
maternal mortality ratio and result in a high correlation (r0.5)

516

E.O. Udjo, P. Lalthapersad-Pillay / Midwifery 30 (2014) 512518

Table 2
Reported and estimated total fertility rate by province, 2001 census and 2007 Community Survey.
Source: Computed from 2001 census and 2007 community survey.
2001 Census
Reported
South Africa
Eastern Cape
Free State
Gauteng
KwaZulu-Natal
Limpopo
Mpumalanga
Northern Cape
North West
Western Cape
n

2.2
2.2
2.1
1.9
2.2
2.6
2.4
2.3
2.4
2.1

Estimated

2007 Community Survey

Statistical test of difference: z value

n

715.4
369.6n
272.8n
272.4n
653.6n
450.6n
252.7n
69.1n
245.5n
180.4n

3.0
3.3
3.3
2.6
3.8
4.0
3.4
2.8
3.4
2.7

Reported

Estimated

Statistical test of difference: z value

2.5
2.9
2.5
2.1
2.5
2.8
2.6
2.6
2.8
2.2

2.7
2.8
2.6
2.5
3.0
3.1
2.9
2.8
3.2
2.4

34.8
3.2
2.50
28.3
37.1
15.9
12.1
5.3
15.5
10.5

Statistically signicant (p o 0.05): standard normal table value at 95% level of condence 1.96.

Table 3
Provincial levels of maternal mortality ratios per 100,000 live births, 2001 and
2007 (2001 Provincial boundaries).
Source: Computed from 2001 census and 2007 Community Survey.

Maternal deaths per 100,000

live births

South Africa
Eastern Cape
Free State
Gauteng
KwaZulu-Natal
Limpopo
Mpumalanga
Northern Cape
North West
Western Cape

2001 Census

2007 Community Survey

463
539
619
372
579
223
468
396
545
179

764
1639
1080
484
969
587
477
610
698
102

y = 1.7626x + 301.3
R = 0.1128

1400
1200
1000
800

that in 1997, it accounted for about 16% of the total immediate

causes of maternal death whereas in 2007 it accounted for about
8% (Table 4). On the other hand, other maternal conditions
account for over 50% of the total immediate causes of maternal
death and the trend is upward. Of these other maternal conditions, maternal infections and parasitic diseases as well as other
maternal diseases complicating pregnancy, childbirth and the
puerperium are the major contributors. In 2007, these groups of
conditions accounted for about 62% of the maternal deaths
(Table 4).
As seen in Table 5, the magnitude of the direct causes of death
exhibits differences at provincial level. In 2007, the relative
contribution of maternal haemorrhage, maternal sepsis and abortion to maternal deaths was highest in the Northern Cape and
lowest in the Western Cape (Table 5); the latter also exhibited the
lowest maternal mortality ratio in 2007. However, the relative
contribution of other maternal conditions was highest in the
Western Cape compared with other provinces. Due to small
numbers, further disaggregation of the other maternal conditions
could not be carried out at provincial level.

600
400
200

Discussion

0
0

50

100

150

200

250

300

HIV prevalence per thousand population aged 15-49

Fig. 3. Relationship of maternal mortality ratio (2007) and HIV prevalence (2008)
in South Africa's provinces*. Source of HIV prevalence: Shisana et al. (2008).
*
It would have been more appropriate to use HIV prevalence for women aged 1549
in the general population instead of women and men combined in the general
population, but the information is not available at provincial level from the
published gures. HIV prevalence among females aged 1549 in the general
population was estimated as 20.2% and for males of the same age as 16.2 in the
general population (Shisana et al, 2005).

and high R2 (0.5) between HIV prevalence and maternal mortality

ratio in a scatter plot of the two variables by province. The scatter
plot of the 2008 HIV prevalence and 2007 maternal mortality
ratio, however, indicated a weak correlation (r 0.34). The R2
value of 0.1128 suggests that only about 11% of the differences in
maternal mortality ratio in South Africa's provinces in 2007 were
explained by differences in HIV prevalence in the provinces (see
Fig. 3).
Causes of maternal deaths
Of the direct causes, hypertensive disorder is the highest
contributor to maternal deaths in South Africa, though its magnitude as a direct cause appears to be declining. The data suggest

Maternal mortality is a public health issue and one of the

Millennium Development Goals. Estimates of maternal mortality
for South Africa have been bandied around, both in national and
international publications. There is a need to critically examine the
estimates to attest to the veracity of the problem in South Africa.
This study's examination of the methodologies of the estimates
indicates that there are a number of issues in the numerator and
denominator of the estimates that were inadequately addressed,
which may have resulted in biases in the estimates. This study
attempted to improve on the estimates. Its national estimate of
maternal mortality ratio of 463 per 100,000 live births in 2001 is
lower than the estimate by Garenne et al. (2008) (542 per 100,000
live births) for the same period. However, its estimate for 2001 is
higher than that of Hogan et al. (2010) (237 per 100,000 live
births) for the period 19802008. Regarding 2007, this study's
national estimate for the period (764 per 100,000 live births) is
higher than that of Garenne et al. (2009) (700 per 100,000 live
births) for the period but lower than the WHO, UNICEF, UNFPA and
the World Bank (2010) estimate (425 per 100,000 live births) for
the same period. South Africa's Health Data Advisory and Coordination Committee (National Department of Health, 2012)
recommended to the National Department of Health that, because
pregnancy-related mortality reported by households in censuses is
twice as high as that from vital registration, cause of death data

E.O. Udjo, P. Lalthapersad-Pillay / Midwifery 30 (2014) 512518

517

Table 4
Contribution of direct immediate causes and other maternal conditions to maternal death 19972007: South Africa.
Source: Computed from registered deaths 1997, 2001, 2007 based on ICD-10 codes.
Per cent contribution to maternal deaths
1997

2001

2007

Direct causes of death : South Africa

Maternal haemorrage
Maternal sepsis
Hypertensive disorder
Obstructed labour
Abortion
Other maternal conditions

10.4
8.0
15.8
0.2
9.5
55.9

10.0
11.5
13.4
0.4
10.5
53.2

6.5
5.9
8.3
0.05
5.0
74.3

Other maternal conditions:

Disorders related to pregnancy
Maternal care related to fetus, amniotic cavity and childbirth problem
Complications of labour and childbirth
Complications related to puerperium
Other obstetric conditions

3.1
0.7
4.6
8.0
42.1

3.0
0.6
4.1
6.2
40.4

5.0
0.4
2.7
3.5
63.4

Other obstetric conditions:

Obstetric death of unspecied cause
Death from obstetric cause 442 days and o 1 year after childbirth
Maternal infections and parasitic diseases complicating pregnancy, childbirth and the puerperium
Other obstetric conditions

0.3
0.2
4.9
36.6

0.8
0.2
7.1
32.3

1.0
0.1
18.0
44.3

HIV/AIDS is an indirect cause of maternal death and so is not included in the table.

Table 5
Per cent contribution of direct immediate causes and other maternal conditions to maternal death by Province, 2007.
Source: Computed from registered deaths, 2007 based on ICD-10 codes.
Direct causes of death

EC

FS

GT

KZN

LP

MP

NC

NW

WC

Maternal haemorrage
Maternal sepsis
Hypertensive disorder
Obstructed labour
Abortion
Other maternal conditions

9.0
9.0
7.2
0.0
4.5
70.1

5.6
2.5
8.8
0.0
3.8
79.4

6.9
4.5
7.9
0.0
4.1
76.6

4.7
6.6
9.6
0.2
5.9
73.0

9.0
6.4
7.1
0.0
5.8
71.8

5.9
10.8
8.1
0.0
5.4
69.9

10.8
5.4
10.8
0.0
10.8
62.2

6.3
4.9
7.0
0.0
3.5
78.3

3.6
0.0
5.5
0.0
1.8
89.1

EC Eastern Cape; FS Free State; GT Gauteng; KZN KwaZulu-Natal; LP Limpopo; MP Mpumalanga; NCNorthern Cape; NW North West; WC Western Cape.

from vital registration with adjustments be used to monitor

maternal mortality. Accordingly, the Committee adopted a maternal mortality estimate of 310 per 100,000 live births for South
Africa for 2010. The scientic logic of this recommendation is
bafing, given that a number of studies (Groenewald and
Bradshaw, 2005; Groenewald et al., 2005; McKerrow and
Mulaudzi, 2010; Birnbaum et al., 2011) have expressed concern
about the accuracy of cause of death information from vital
registration in South Africa. One study in particular, noted that
in terms of monitoring the health status of the nation and
understanding the burden of disease, the extent of ill-dened
causes together with 9.0% of deaths being due to garbage codes,
highlights the urgent need to improve the quality of cause of death
certication (Bradshaw et al., 2010: 27).
This study also focused on the direct causes of maternal
mortality as well as provincial differentials and came to the same
conclusion as other studies that maternal mortality increased
during the period 20012007 in South Africa and may have begun
in the late 1990s. The 2011 South African Census data became
available as this study was nearing completion. Using the methods
described above, the study estimated maternal mortality ratio to
be 692 per 100,000 live births in 2011 from the data, which
suggests that maternal mortality declined during the period 2007
2011. It has been argued that the increase in maternal mortality
levels in the 1990s2007 in South Africa is due to HIV/AIDS
(Garenne et al., 2009). The study on which this article is based

indicates a weak correlation between provincial HIV prevalence

and maternal mortality ratio. Moszynski (2011) attributes South
Africa's rising maternal mortality to health system failures.
Direct causes of maternal deaths are haemorrhage, sepsis,
hypertensive disorder, obstructed labour and complications of
poorly performed abortions (Winikoff et al., 1991). Analysis of
the death records from vital registration (in the study under
discussion) suggest that over 50% of maternal deaths in South
Africa are currently due to other maternal conditions. These
comprise mainly maternal infections and parasitic diseases as well
as other maternal diseases complicating pregnancy, childbirth and
the puerperium. This requires further investigation as this study
does not proffer a strict conclusive answer as to the high level of
maternal mortality due to other maternal conditions: did some of
these women give birth in hospital or at home? This cannot be
investigated from the vital registration data. Also, the data investigating possible confounding factors are limited. It is suggested
that, whereas interventions should continue to focus on reducing
maternal mortality due to hypertensive disorder, maternal haemorrhage, maternal sepsis and obstructed labour, the thrust of
intervention should focus on reducing maternal infections and
parasitic diseases as well as other maternal diseases that complicate pregnancy, childbirth and the puerperium.
This study's provincial level estimates suggest that the Eastern
Cape currently has the highest maternal mortality ratio. Eight of
the nine provinces showed a rising trend in maternal mortality

518

E.O. Udjo, P. Lalthapersad-Pillay / Midwifery 30 (2014) 512518

ratio between 2001 and 2007. The Western Cape had the lowest
maternal mortality ratio and was the only province that experienced a decline between 2001 and 2007. In all provinces, other
maternal conditions are currently the largest contributors to
maternal deaths. Although further disaggregation of these conditions could not be carried out at provincial level due to small
numbers, the same intervention measures should be applied at
provincial level.
The main limitation of this study is that, in adjusting for
underreporting of maternal deaths, it was assumed that underreporting/overreporting of deaths in households was constant for
all causes of death. As there is no methodology comparable to the
Growth Balance Method for estimating the completeness of death
reporting by specic causes, this assumption provided perhaps a
minimum adjustment factor for the reported pregnancy-related
deaths. Other possible limitations have to do with stillbirths and
multiple births. Regarding stillbirths, although the 2001 Census
and 2007 Community Survey fertility questions related only to live
births, some women may have reported a still birth as a live birth.
But some women may also have reported a live birth as a stillbirth
if the child died a short moment after birth. The resulting biases
would be in the opposite direction though they may not necessarily cancel out each other. Regarding multiple births, births in the
preceding 12 months were derived from date of birth of last child
so multiple births would be present in the data (though likely to
be relatively small). Multiple births were not given a specic code
in the data. Theoretically, stillbirths and multiple births would
have a potential impact on the denominator of the estimate.
However, this study argues that any possible bias arising from
this is most likely to be negligible because of how the denominator
was derived: stillbirths and multiple births in the data are unlikely
to produce signicant shifts in the and parameters of the
Gompertz model and hence unlikely to produce signicant impact
on the model age-specic fertility rates derived from the and
parameters. Despite these limitations, the results in this study
suggest that, despite strategies by the National Department of
Health to improve maternal and child health, maternal mortality
remains a public health challenge in South Africa. Whilst the
accuracy of maternal mortality estimates for South Africa might be
debatable, one irrefutable conclusion from different studies is that
maternal mortality increased between 2001 and 2007. Although
the 2011 South African Census data indicate that maternal mortality declined between 2007 and 2011, the level remains high and
South Africa is unlikely to achieve the MDG of reducing maternal
mortality by three quarters by 2015. This should be a focus for the
relevant government departments.
Conict of interest statement
We hereby afrm that there is no conict of interest that could
inappropriately inuence this study. We did not receive any
nancial support from any individual or organisation for this
study.
Acknowledgement
We wish to thank Statistics South Africa for providing access to
its data. The views expressed in this paper are, however, those of
the authors.
References
Birnbaum, J.K., Murray, C.J., Lozano, R., 2011. Exposing misclassied HIV/AIDS
deaths in South Africa. Bulletin of the World Health Organization 89, 278285.

Bradshaw, D., Pillay-Van Wyk, V., Laubscher, R., 2010. Cause of Death Statistics for
South Africa: Challenges and Possibilities for Improvement. Burden of Disease
Research Unit, Medical Research Council, Cape Town.
Brass, W., 1981. The use of the Gompertz relational model to estimate fertility.
Paper presented at the International Union for the Scientic Study of Population Conference, Manila, 3, 45-361.
Brass, W., 1971. Methods for Estimating Fertility and Mortality from Limited and
Defective Data. The University of North Carolina, Chapel Hill.
Cooper, D., Morroni, C., Orner, P., et al., 2004. Ten years of democracy in South
Africa: documenting transformation in reproductive health policy and status.
Reproductive Health Matter 12, 7085.
Department of Welfare, 1998. Population Policy for South Africa. Ministry for
Welfare and Population Development, Pretoria.
Garenne, M., McCaa, R., Nacro, K., 2008. Maternal mortality in South Africa in 2001:
from demographic census to epidemiological investigation. Population Health
Metrics 6, 113.
Garenne, M., McCaa, R., Nacro, K., 2009. Maternal mortality in South Africa: an
update from the 2007 Community Survey. Paper prepared for the 26th IUSSP
International Conference, Marrakech, 27 September2 October.
Graham, W.J., 1991. Maternal mortality: levels, trends and data deciencies. In:
Feachem, R.G., Jamieson, D.T. (Eds.), Disease and Mortality in Sub-Saharan
Africa. Oxford University Press, London, pp. 425445.
Groenewald, P., Bradshaw, D., 2005. Ofcial death statistics under-represent AIDS as
a cause in South Africa. AIDS Analysis Africa, AprilMay, pp. 24.
Groenewald, P., Nannan, N., Bourne, D., Laubscher, R., Bradshaw, D., 2005.
Identifying deaths from AIDS in South Africa. AIDS 19, 193201.
Hill, K., 1987. Estimating census and death registration completeness. Asian and
Pacic Population Forum 1, 813.
Hogan, M.C., Foreman, K.J., Naghavi, M., et al., 2010. Maternal mortality for 181
countries, 19802008: a systematic analysis of progress towards Millennium
Development Goal 5. The Lancet 375, 16091623.
Horton, R, 2012. Women's and children's health: no time for complacency. The
Lancet 380, 11231125.
Hsu, J., Pitt, C., Greco, G., Berman, P., Mills, A., 2012. Count down to 2015: changes in
ofcial development assistance to maternal, newborn, and child health in
200910, and assessment of progress since 2003. The Lancet 380, 11571168.
Human Rights Watch, 2011. Stop making excuses: accountability for maternal
health care in South Africa. http://www.hrw.org/norde/100757.
Lilijestrand, J., Pathmanathan, I., 2004. Reducing maternal mortality: can we derive
policy guidance from developing country experiences? Critical elements in
reducing maternal mortality. Journal of Public Health Policy 25, 299314.
McKerrow, N., Mulaudzi, M., 2010. Child mortality in South Africa: using existing
data. South African Health Review, (Part 5), 5972.
Moszynski, P., 2011. South Africa's rising maternal mortality is due to health system
failures. British Medical Journal 343, d5089.
National Committee for the Condential Enquiry into Maternal Deaths, 1998.
Interim Report on the Condential Enquiry into Maternal Deaths in South
Africa. Pretoria.
National Department of Health, 2012. Health Data Advisory and Co-Ordination
Committee Report. Pretoria.
Shah, H.I., Say, L., 2004. Maternal mortality and maternity care from 1990 to 2005:
uneven but important gains. Reproductive Health Matters 15, 1727.
Shisana, O., Rehle, T., Simbayi, L., et al., 2005. South African National HIV Prevalence,
HIV Incidence, Behaviour and Communication Survey, 2005. HSRC Press, Cape
Town.
Shisana, O., Rehle, T., Simbayi, L., et al., 2008. South African National HIV Prevalence,
HIV Incidence, Behaviour and Communication Survey, 2008: A Turning Tide
Among Teenagers. HSRC Press, Cape Town.
Statistics South Africa, 2003. 2001 Census: Census in Brief. Pretoria.
Statistics South Africa, 2007a. Causes of Death in South Africa 19972007. Pretoria.
Statistics South Africa, 2007b. Community Survey 2007: Methodology, Processes
and Highlights of Key Results. Pretoria.
Udjo, E.O., 2006. Estimation of mortality from vital registrations in South Africa.
Current HIV Research 4, 469474.
Udjo, E.O., 2008. A re-look at recent statistics on mortality in the context of HIV/
AIDS with particular reference to South Africa. Current HIV Research 6,
143151.
World Health Organization, 2012. Maternal mortality. http://www.who.int/gho/
maternal_health/mortality/maternal/en/.
World Health Organization (WHO), UNICEF, UNFPA and the World Bank, 2010.
Trends in Maternal Mortality: 1990 to 2010. World Health Organization,
Geneva.
Wilmoth, J.R., Mizoguchi, N., Oestergaard, M.Z., et al., 2012. A new method for
deriving global estimates of maternal mortality. Statistics, Politics and Policy 3,
138.
Winikoff, B., Carignan, C., Bernardik, E., Semeraro, P. 1991. Medical Services to Save
Mothers' Lives: Feasible Approaches to Reducing Maternal Mortality. Population Council, New York: Programs Division Working Paper 4, March.
Zaba, B., 1981. Use of the Relational Gompertz Model in Analysing Fertility Data
Collected in Retrospective Surveys. Centre for Population Studies Working
Paper 81-2, London School of Hygiene and Tropical Medicine.