10.1 Summary
10.1 Summary
10.1 Summary
1 Summary
Control over fertility and access to safe maternity
care are fundamental health and human rights and
are strongly influenced by social determinants. Using
a variety of methods, this chapter examines determinants
of unintended pregnancy and its outcomes and
of maternal risks from childbearing, including access to
care by a skilled birth attendant.
For unintended pregnancy, the analysis in this chapter
was based on a broad review of the literature, supplemented
by commissioned articles produced by experts.
For pregnancy outcome, the analysis focused on determinants
of receiving care from a skilled birth attendant
because the proximate causes of maternal morbidity
and mortality can usually be successfully treated when
women have access to basic health care. This included a
comparison of country-by-country statistics on access
to skilled birth attendance, pregnancy outcome and
various social determinants.
Worldwide, 40% of all pregnancies are unintended.
Comparison of desired family size to actual fertility
demonstrates that, in almost all countries, the burden
of unintended pregnancy disproportionately affects
the poor. Other disadvantaged groups that have higher
rates of unintended pregnancy in many settings include
young people, the uneducated, ethnic minorities and
migrants.
Women with an unintended pregnancy may be faced
with a choice between terminating the pregnancy or an
unwanted birth. Unsafe abortion accounts for 13% of
maternal deaths worldwide, and disadvantaged women
are less likely to have access to safe abortion services
and to proper care to treat complications. Poor women
also suffer disproportionate consequences of unwanted
childbearing, including health and social consequences
for themselves and their children. Vulnerability to unintended
pregnancy is strongly influenced by access to
and use of effective contraception and by exposure to
unwanted sex through child marriage and sexual violence.
These all have strong social determinants.
The proportion of births with skilled attendance and
per capita health expenditure alone account for 90%
of between-country variation in maternal mortality.
At given levels of health expenditure, achieving
equity by income level in coverage with skilled birth
attendance is strongly correlated with high levels of
overall coverage, as are education for women, higher
levels of public (versus private) expenditure on health
and an efficiently performing government. Vulnerability
to maternal mortality and morbidity despite access
to skilled birth attendance depends on the quality of
skilled birth attendant services and the availability of
backup treatment (especially blood transfusion and
caesarean section) for major obstetric complications.
Addressing unintended pregnancy and improving
pregnancy outcome will require interventions specifically
designed to achieve equity in the availability of all
related health services, especially targeting the poor and
disadvantaged for access to contraceptive and skilled
birth attendant services. Such efforts will be most
effective when combined with addressing upstream
determinants, such as improving education for women
and the effective functioning of the health sector and
of government services in general. For future progress,
it will be essential to rigorously measure the impact of
interventions.
10.2 Introduction
Background
The ability of women and couples to control their
fertility and to have basic, safe maternity care is a fundamental
health and human right. This has been endorsed
by the World Health Assembly (1), and the World Health
Organization (WHO) affirms that “sexual and reproductive
health is fundamental to individuals, couples
and families, and the social and economic development
of communities and nations” (2). As stated by the International
Conference on Population and Development
in 1994 (3): “All couples and individuals have the basic
right to decide freely and responsibly the number and
spacing of their children and to have the information,
education and means to do so.”
The broader field of sexual and reproductive health
covers many areas that go beyond pregnancy and its
outcomes to include, for example, human immunodeficiency
virus and other sexually transmitted infections.
These are certainly areas of great importance in which
social determinants have long been recognized to play
a major role, and the entire field is too broad to be covered
in a single chapter of this volume. This chapter
therefore focuses on one aspect of sexual and reproductive
health – the social determinants of unintended
pregnancy and of pregnancy outcome.
Despite significant improvements in the lives of
women (4), high rates of unintended pregnancy continue
to detrimentally impact women’s and children’s
health and restrict opportunities for women (5). Selection
of unintended pregnancy as a focus of this chapter
was based on five main principles:
• Ensuring the ability to choose the number and spacing
of children as a means of achieving health and
development goals has been neglected as part of key
international and national development frameworks
(6–12).
• The burden of unintended pregnancy affects a large
proportion of society. The growing demand for
smaller families, decreasing age at first sex (in some
countries) and increasing age of marriage has meant
that many women spend much of their adult lives
attempting to avoid an unintended pregnancy (13,
14).
• Safe and highly effective means of primary prevention
(contraception) (15) and secondary prevention
(termination of pregnancy) (16, 17) can reduce the
burden of unwanted births.
• While reporting of unintended pregnancy raises
some methodological concerns, ample data are
available for examination (10, 18).
• Assisting women in avoiding unintended pregnancies
improves the health of women, children and
families, and represents a pledge to the right of all
women to control their fertility.
Another focus of this chapter is the risk to women associated
with childbirth and with unsafe abortion. The
morbidity and mortality associated with pregnancy and
childbirth is remarkable among health conditions in
the extent to which it can be minimized by access to
relatively simple care. One of the targets of the Millennium
Development Goals is to provide all women with
access to a skilled birth attendant. This chapter includes
an examination of the social determinants of access to
skilled birth attendance.
Methods
This chapter represents work conducted by two units
of WHO: Reproductive Health and Research and
Making Pregnancy Safer. Instead of trying to cover the
entire broad topic, the approach was for different teams
to choose their own focus and analytical methods. This
chapter attempts to present and synthesize their findings
within the analytical framework of this volume
(see Chapter 1).
For unintended pregnancy, the analysis began with
a broad review of the literature. The search strategy
included studies examining the determinants and
effects of “unintended”, “mistimed” or “unwanted”
pregnancies and births. In addition, nine commissioned
articles were written by identified experts in the field.
The theme and scope of these articles were defined by
an internal working group involving participants from
multiple departments within WHO. Bibliographic
databases, topic-specific journals and Internet searches
were conducted to identify reports and publications
within and outside peer-reviewed journals relevant to
the analysis. Data from the Demographic and Health
Surveys were used to examine gradients of inequity
within countries (19 ).
Previous reviews have noted the methodological difficulties
in measuring unintended pregnancy (18 ).
Authors note the lack of available evidence on pregnancy
intention, particularly in developing countries
(5, 18). Much of the evidence in this review is from
surveys that ask women to retrospectively classify their
pregnancies as “wanted” or “unwanted”. Publications
have described the limitations of this approach (20,
21), including the inherent bias in recall of intention,
underreporting of pregnancies that did not result in a
live birth, the tendency to transform past intention to
match current realities of parenthood and the influence
of culture in classification of pregnancies (18, 22, 23).
In addition to self-reported pregnancy intentions, two
alternative means of measuring unwanted pregnancy
are applied in the research literature: rate of induced
abortion and “excess fertility”. Pregnancies that are
voluntarily terminated are generally considered unintended.
A small number of induced abortions may be
among women whose conception was intentional, but
this number is unlikely to significantly skew observed
disparities in incidence or outcome. Excess fertility,
another measure of unwantedness, is calculated as the
difference between women’s reported ideal family size
and total fertility rate (number of children a woman is
likely to have in her lifetime).
For pregnancy outcome, there were two levels of analysis.
The first (presented mainly as a webannex) covers
a broad range of proximate causes of adverse pregnancy
outcomes. This was based on a consensus process
involving staff of the Making Pregnancy Safer Unit of
WHO and consultants, and includes estimates of prevalence
and risk from the scientific literature. It also
includes an appraisal for each health issue of both the
strength of its association, if any, with social determinants
and of the evidence that the association is causal,
based on generally accepted criteria for causality (24 ).
It was clear from this exercise that relatively few factors
account for most of the variation in women’s chance
of giving birth safely. For this chapter, it was decided
to focus on the second level of analysis: social determinants
of access to skilled birth attendance. The method
for this focus was to conduct an original analysis of
cross-national (or “ecological”) data. The data presented
here are mostly drawn from reports published by
United Nations agencies, the United States Agency for
International Development (USAID) and the World
Bank. The principal data sources are the 2006 Human
Development Report and the 2006 World Health Report,
and where no other reference is given data were taken
from these compilations (25, 26).
180 Equity, social determinants and public health programmes
Compared to what is available in high-income countries,
very few data on pregnancy outcomes and the
factors that affect them are available from low- and
middle-income countries. The quality of data available
is also uncertain and variable, and this should be
considered when evaluating our results. The United
Nations agencies present, as far as possible, data collected
using consistent methods and adjusted for
well-defined sources of error. Many of the data, however,
are collected by national governments, and
methods vary. Even in rich countries, official estimates
of maternal mortality may be inaccurate: in the United
Kingdom, for example, the official rate, estimated from
death certificates, is half the true rate (27 ). In low- and
middle-income countries estimates based on officially
registered deaths systematically and very substantially
underestimate maternal mortality, sometimes only
including deaths that take place in facilities (28 ).
A key variable in this discussion is the percentage of
births attended by a skilled birth attendant. Data collected
by different countries are not based on a single
definition of “skilled birth attendant”, or on any definition
of “attended”. The WHO definition of a skilled
birth attendant is “someone trained to proficiency in
the skills needed to manage normal (uncomplicated)
pregnancies, childbirth and the immediate postnatal
period, and in the identification, management and
referral of complications in women and newborns”
(29 ). However, data in the World Health Report on the
percentage of births attended by skilled birth attendants
in several countries are based on definitions inconsistent
with that of WHO.
No analysis was undertaken of data for countries for
which data on the proportion of births with skilled
attendance were not available. Most other variables
were unavailable for at least some countries. United
Nations reports include data for most variables from
the great majority of countries in sub-Saharan Africa
and from the larger (in population) countries of South
and Central America, Asia and North Africa. Smaller
countries outside Africa are those for which data are
most often unavailable.
Statistical analyses used proprietary statistical software.
In keeping with the limitations of ecological data, the
analysis was exploratory and hypothesis generating.
Spearman rank correlation coefficients are reported
for most bivariate correlations to avoid difficulties
with variables not normally distributed. Multivariate
analysis was used sparingly because of multicollinearity
and other characteristics of the data that violate
model assumptions. Because a large number of comparisons
were made, an arbitrary conservative threshold
of P < 0.005 was used.
10.3 Analysis
Global burden of unintended
pregnancy: context and position
Of all pregnancies worldwide, 40% are unintended.
Approximately 20% of pregnancies worldwide are voluntarily
terminated. In 2003, an estimated 42 million
abortions were induced, 35 million (26 million excluding
China) of which occurred in developing countries
(30 ).
Women who are unable or choose not to terminate
an unwanted pregnancy are faced with an unwanted
birth. Analysis of fertility data from 20 low- and middle-
income countries estimated that on average 22%
of all births were unwanted and that, for most countries,
the proportion of unwanted births has grown
(31 ). In developing countries where data were available,
researchers found that between 14% and 62% of recent
births were reported as unintended (18 ).
Within countries, the burden of unintended and
unwanted pregnancy is not equally distributed. In the
United States of America, for example, rates of unintended
pregnancy are consistently higher for poor
women, ethnic minorities, women aged 18–24 years,
women who have not completed high school and
unmarried women (14, 32–34). The overall rate of unintended
pregnancy in the United States has remained
constant for almost a decade, with almost half (49%)
of all pregnancies reported as unintended (32 ). Among
subpopulations, however, this rate fluctuates. Between
1994 and 2001, the rate of unintended pregnancy
declined among adolescents, college graduates and the
wealthiest women, but increased among poor and less
educated women (32 ). Limited data from other countries
have shown similar patterns of disparities, with
rates of unintended pregnancy markedly higher among
the poor (35 ), migrants (35 ), unmarried (35, 36) and
adolescents (36 ).
Substantially more evidence is available to examine
differences in actual births. Demographic and Health
Survey data substantiate higher levels of excess fertility
among poor women in developing countries. In 41
countries where data were available, poor women from
all countries outside Africa and the majority of African
countries reported higher levels of unintended births
than women from wealthier households (37 ). Figure
10.1 shows women’s ideal family size compared to their
estimated total fertility rate (TFR) by wealth quintile
in selected countries. In the countries shown, there is
substantially less difference in ideal family size between
women from the poorest households and those from
the wealthiest households than there is difference in
the number of children they are likely to bear, given
Consequences of unintended
pregnancy
Unsafe abortion
Women with an unwanted pregnancy are faced with
a difficult decision. Deciding whether to terminate
an unwanted pregnancy or have an unwanted child is
influenced by many factors, including the availability
and accessibility of induced abortion services, the social
acceptability of childbearing and induced abortion, and
support from social structures. Either choice has social,
financial and health consequences that are not equally
experienced among women.
Several procedures are currently available to assist
women with safe termination of pregnancy. Expansion
of safe induced abortion services into remote and rural
areas is possible largely as a result of advances in medical
technologies, which have reduced cost and simplified
procedures (16, 17, 42, 43). Complication rates for these
procedures are extremely low, with almost all abortionattributable
morbidity and mortality resulting from
untrained providers, use of harmful procedures or failure
to use appropriate infection prevention procedures
(17, 44). “Unsafe abortion” is defined as a procedure for
terminating pregnancy carried out by attendants without
appropriate skills, or in an environment that does
not meet minimum standards for the procedure, or
both (17 ). Unsafe abortion is a major cause of maternal
mortality, accounting for an estimated 13% of maternal
deaths worldwide (16 ). The highest estimated rate of
unsafe abortion is in Latin America and the Caribbean,
where there are 33 unsafe abortions per 100 live births,
followed by Africa (17 per 100 live births) and Asia (13
per 100 live births) (30 ).
In 2005, an estimated 5 million women were hospitalized
for treatment of complications from unsafe abortion
(45 ). Rates of unsafe abortion are highest among young
women (46–48), with almost 60% of unsafe abortions in
Africa occurring among women under age 25 (46 ). A
number of studies have documented higher complication
rates and mortality resulting from unsafe abortion
among women of low socioeconomic status (49–52).
Factors that contribute to observed differentials in
abortion complications include the health status of
women (53 ), longer delay in seeking induced abortion
(53, 54), use of less skilled providers (43, 53), use of more
dangerous methods (43, 53) and longer delay in seeking
care for complications (43 ). Figure 10.2 illustrates the
differences in care-seeking behaviour among women
of varying socioeconomic status (40, 55). Women from
more affluent households are more likely to obtain an
induced abortion from a physician or nurse, while poor
women living in rural areas are more likely to use a traditional
practitioner or self-induce an abortion.
Women report that socioeconomic concerns are a primary
consideration in deciding whether to seek an
induced abortion (54, 56). Poorer women and adolescents
are less likely to have the financial resources (54 );
less likely to have the knowledge of when, where and
from whom to seek an induced abortion (54 ); or lack
the social support to secure safe abortion services (57 ).
The principal social determinant of recourse to unsafe
abortion is real or perceived legal restriction on safe
abortion (58 ). Developing countries are much more
likely to restrict access to legal abortion than developed
countries, and the restrictions disproportionately
affect poorer women (58 ). While abortion is allowed to
preserve women’s physical or mental health in 86% of
developed countries, only 55% of developing countries
allow this. Many conditions that make pregnancy dangerous,
however, such as valvular heart disease, are more
common in developing countries and more common
among poorer women within those countries, and
women in those countries and poorer women within
them are less likely to have access to effective treatment.
Affected women are then forced to make an invidious
choice between a high-risk pregnancy and an unsafe
abortion.
10.4 Discussion
Social determinants play a key role in both unintended
pregnancy and pregnancy outcome. Women from disadvantaged
social circumstances are more likely to
experience an unintended pregnancy than women
with greater financial and social resources. When
faced with an unwanted pregnancy, women with less
means are also more likely to face more severe consequences
from an unsafe abortion or an unwanted birth
than more advantaged women (100). These disparities
in unintended pregnancy and its consequences are the
result of social, political and economic systems that do
not provide access to correct knowledge of sexual and
reproductive health and to necessary services.
Unintended pregnancy and pregnancy outcome are
affected by social determinants that operate at all five
levels of the analytical framework used in this volume
(see Chapter 1). At the level of socioeconomic context
and position, women living in poorer countries
and poorer women within countries clearly do worse
on all counts. They have less access to modern contraception,
more unintended pregnancies, less access to
pregnancy care and worse pregnancy outcomes. Other
aspects of context and position are also crucial. These
include broad gender issues, especially the importance
of education for girls.
At the level of differential exposure, poor and disadvantaged
women are more likely to be exposed to
unwanted sex, including through sexual violence and
child marriage. At the level of differential vulnerability,
they are at higher risk of unintended pregnancy
because they are less likely to have the necessary knowledge,
access and skills to use contraception when they
do not wish to become pregnant. Even when they seek
such services, the poor, the young and the disadvantaged
often receive inferior care.
Poor women are especially vulnerable because they are
less likely to deliver under the care of a skilled birth
attendant, sometimes resulting in rates of maternal
morbidity and mortality orders of magnitude higher
than for richer women. Even when they do have skilled
birth attendance, they may still suffer from differential
outcomes of care because not all “skilled” birth attendants
have the same level of skill or the same access to
hospital back-up when complications arise. For women
who choose not to keep an unwanted pregnancy, the
lack of access to safe abortion services can also increase
risk by orders of magnitude. Further, certain groups of
women are more likely than others to receive differential
treatment, including being subject to provider
biases and value judgements not necessarily in line with
official policy.
Differential consequences add to the burden of the
poor and disadvantaged. While an unintended pregnancy
can be a life-changing event for any woman,
poor women have fewer resources with which to cope
with resulting health, social and economic strains. This
can quickly turn to tragedy if an abortion or childbirth
results in serious maternal morbidity or mortality or if
another unintended child means not enough food for
that child or its siblings.
10.5 Interventions
This section describes structural interventions to
improve the accessibility, availability and acceptability
of services at the micro and macro levels (101). Beyond
service provision, avoiding unintended pregnancy
involves complex behaviours that require consistent
contraceptive use over an extended period of time.
Macro-level approaches
Within the health sector, programmes can shift human
and financial resources to reach underserved populations,
increasing overall availability of services. Policies
can improve the accessibility and acceptability of services
by protecting reproductive rights and expanding
knowledge of sexual and reproductive health. Also,
communities can reduce gender inequity by ensuring
equal access to educational and financial opportunities
for women.
Redistribution of health sector resources
One of the most ambitious attempts at extending
coverage to underserved populations involves the redistribution
of health system resources to the periphery.
The Matlab experiment in Bangladesh is perhaps the
most widely known example of this approach. Beginning
in the 1970s, the government, with support from
international donors, sustained nearly 20 000 female
community health workers whose jobs involved visiting
households, meeting with residents, caring for
the health needs of mothers and children and offering
contraceptives (injectable, oral and barrier methods)
(102). Doorstep services were supported by clinic-based
professionals who offered permanent contraceptive
methods along with basic primary health care services.
Evaluations of the programme have shown improvements
in maternal mortality, contraceptive use and
child survival indicators (51, 102, 103). Although the
programme has not been directly linked to equitable
availability of family planning services, nationally
representative surveys show little variation in contraceptive
use among socioeconomic groups (85 ).
10.7 Conclusion
Adequate funding of services that increase the safety
of pregnancy and delivery is essential, and the level of
funding is certainly a useful indicator. When assessing
whether funding for maternal health services is adequate,
the focus should be on public health systems.
Given that many low- and middle-income countries
find it difficult to spend enough on their public health
systems to ensure a high level of access to skilled birth
attendance, the efficiency of service provision is also
an important issue. Adequate and consistent funding is
also essential to assure equitable access to contraception.
This includes not only providing facilities and
health personnel but also programmes that reach out
to poor and disadvantaged communities. And no family
planning programme can be successful without an
uninterrupted flow of basic contraceptive commodities.
Reproductive health services provided to women by
the health sector are often not equitably distributed and
are determined by social factors. In theory, it should be
within the power of the health care system to substantially
reduce disparities in pregnancy outcome. But in
practice, an inadequate or inequitable health care system
may only serve to widen these disparities.
Almost everyone would agree that healthy mothers and
families should be a high priority for any society. The
means to greatly reduce unintended pregnancy and
morbidity and mortality associated with pregnancy
are well within our knowledge and not overly expensive.
Because the burden falls so disproportionately on
the poor and disadvantaged, it is impossible to make
significant strides in improving overall rates without
concentrating on reaching poorer women. This means
that a broader social perspective will be essential to
achieve the results we all desire.