newborns: no longer going unnoticed 83 Box 5.1 Explaining variations in maternal, neonatal and child mortality: care or context? The debate over the contribution of maternal, various contextual or health systems indicators around 50% of the differences in maternal and newborn and child health programmes to sav-(21-24). There are 67 developing countries for neonatal mortality, 37%of those in postneona- ing lives is not new. Historical analyses have which reliable estimates are available of the mortality, and 50% of those in child mor- often indicated the important role of contex- levels of maternal, neonatal, postn and tality, with human resource density the mai ctors such as a hea nent, child mortality in 2000. For each of these coun- single explanatory factor within the care scon women's empowerment, education and poverty tries a care score can be constructed through This suggests that care, and particularty human in reducing mortality levels. It can be difficult to principal components analysis, reflecting finan- resources, plays a larger role in explaining the disentangle these contextual effects from the cial inputs( total and government expenditure inter-country differences in mortality than di contribution of the care provided through health on health per inhabitant), human resource ferences in context. systems. Poverty, for example, is is often part and dens Bond y(midwives and doctors per head of A significant proportion of the variability in parcel of poorly functioning health systems as population) and responsiveness(determined mortality levels is explained by the interaction ers and children live. The current consensus is the same technique it is also possible to con- sis suggests that where the context is particu that both health systems and the environment struct a context score for each country, using larly challenging even strong health systems care and context-play their part, but that the following indicators: income per inhabitant, can have only a limited effect on mortality the balance may be different for the health of female income, female literacy, sanitation and conversely, where there is an enabling con- mothers from that of their children, maternal access to safe water (25 text for health in terms of education. wealth rtality depending more on health systems Variations in country context scores explain environment and women s empowerment, the efforts and less on contextual factors than child between 10% and 15% of the differences a poor health system could hold back mortal tween countries in maternal, neonatal and ity reduction substantially On the whole, the One way to disentangle the relative contri- postneonatal mortality in a series of multiple analysis confirms the importance of investing ution of care and context to mortality is to regressions. They explain 24% of the differ- in health systems to reduce mortality. relate mortality levels across countries with ences in child mortality. Care scores explain Proportion of inter-country variation in levels of mortality explained by indicators of care and context Unexplained variation variation explained by financial and responsiveness on explained by human resource density Variation explained by interaction en care and cont Variation explained by contextual Maternal Neonatal Postneonatal Child mortality ortal
newborns: no longer going unnoticed 83 various contextual or health systems indicators (21–24). There are 67 developing countries for which reliable estimates are available of the levels of maternal, neonatal, postneonatal and child mortality in 2000. For each of these countries a care score can be constructed through principal components analysis, reflecting financial inputs (total and government expenditure on health per inhabitant), human resource density (midwives and doctors per head of population) and responsiveness (determined through individual satisfaction ratings). Using the same technique it is also possible to construct a context score for each country, using the following indicators: income per inhabitant, female income, female literacy, sanitation and access to safe water (25). Variations in country context scores explain between 10% and 15% of the differences between countries in maternal, neonatal and postneonatal mortality in a series of multiple regressions. They explain 24% of the differences in child mortality. Care scores explain The debate over the contribution of maternal, newborn and child health programmes to saving lives is not new. Historical analyses have often indicated the important role of contextual factors such as a healthy environment, women’s empowerment, education and poverty in reducing mortality levels. It can be difficult to disentangle these contextual effects from the contribution of the care provided through health systems. Poverty, for example, is often part and parcel of poorly functioning health systems as well as being part of the context in which mothers and children live. The current consensus is that both health systems and the environment – care and context – play their part, but that the balance may be different for the health of mothers from that of their children, maternal mortality depending more on health systems’ efforts and less on contextual factors than child mortality. One way to disentangle the relative contribution of care and context to mortality is to relate mortality levels across countries with around 50% of the differences in maternal and neonatal mortality, 37% of those in postneonatal mortality, and 50% of those in child mortality, with human resource density the main single explanatory factor within the care score. This suggests that care, and particularly human resources, plays a larger role in explaining the inter-country differences in mortality than differences in context. A significant proportion of the variability in mortality levels is explained by the interaction between care and context. More detailed analysis suggests that where the context is particularly challenging even strong health systems can have only a limited effect on mortality; conversely, where there is an enabling context for health in terms of education, wealth, environment and women’s empowerment, then a poor health system could hold back mortality reduction substantially. On the whole, the analysis confirms the importance of investing in health systems to reduce mortality. Box 5.1 Explaining variations in maternal, neonatal and child mortality: care or context? Proportion of inter-country variation in levels of mortality explained by indicators of care and context % variance explained Unexplained variation Variation explained by financial inputs and responsiveness Variation explained by human resource density Variation explained by interaction between care and context Variation explained by contextual factors Maternal mortality Child mortality Postneonatal mortality Neonatal mortality 100 90 80 70 60 50 40 30 20 10 0
84 The World Health Report 2005 The reversal of progress in neonatal health in sub-Saharan Africa is both concern ing and unusual. Historically, declines in child mortality have often reversed when the social context deteriorated. Within Europe, these reversals mostly affected older children, while remaining modest for neonatal mortality (13 ) The causes of the poor progress in reducing both neonatal and later childhood deaths in sub-Saharan Africa are likely to be many and complex Economic decline and conflict are likely to have played significant roles through their disruptive effect on access to health ser (14-16). The impact of the HIV/AIDS epidemic on mortality is less well established for newborns than for the post-neonatal period, but infants born to HIv-positive mothers are more likely to be stillborn or premature; they are also likely to have low APGAR scores' and very low birth weights(17, 18) Reductions in child mortality in many countries are at least partly driven by socio economic development: improvements in women's education and literacy, household income, environmental conditions(safe water supply, sanitation and housing), along with improvements in health services and child nutrition (19, 20 ). While neonatal mor tality is affected by these factors, they may have a greater impact in the post-neonatal and early childhood periods than for newborns(see Box 5. 1 ). Historical data further support this hypothesis. There is little evidence that the often dramatic reductions in infant and child mortality in Europe during the first few decades of the 20th century i The APGAR test evaluates a newborn' s physical condition Each year more than 4 million babies die within 28 days of coming into the world and nearly 3.3 million babies are stillborn
84 The World Health Report 2005 The reversal of progress in neonatal health in sub-Saharan Africa is both concerning and unusual. Historically, declines in child mortality have often reversed when the social context deteriorated. Within Europe, these reversals mostly affected older children, while remaining modest for neonatal mortality (13). The causes of the poor progress in reducing both neonatal and later childhood deaths in sub-Saharan Africa are likely to be many and complex. Economic decline and conflict are likely to have played significant roles through their disruptive effect on access to health services (14–16). The impact of the HIV/AIDS epidemic on mortality is less well established for newborns than for the post-neonatal period, but infants born to HIV-positive mothers are more likely to be stillborn or premature; they are also likely to have low APGAR scores1 and very low birth weights (17, 18). Reductions in child mortality in many countries are at least partly driven by socioeconomic development: improvements in women’s education and literacy, household income, environmental conditions (safe water supply, sanitation and housing), along with improvements in health services and child nutrition (19, 20). While neonatal mortality is affected by these factors, they may have a greater impact in the post-neonatal and early childhood periods than for newborns (see Box 5.1). Historical data further support this hypothesis. There is little evidence that the often dramatic reductions in infant and child mortality in Europe during the first few decades of the 20th century Each year more than 4 million babies die within 28 days of coming into the world, and nearly 3.3 million babies are stillborn. N. Behring-Chisholm/WHO 1 The APGAR test evaluates a newborn’s physical condition
newborns: no longer going unnoticed 85 were fuelled by improvements in health care provision, and most studies argue that they resulted from a number of factors including rising standards of living and nutri tion, reduced fertility, safer water, better sanitation, and improved housing(26, 27) During this time, progress in reducing neonatal mortality was limited and was confined to the late neonatal period. Progress did not accelerate until around the time of the Second World War (28), which coincided with greater provision and use of maternal health care, improved quality of professional midwifery and obstetric services, and access to antibiotics. This suggests that, while some limited progress can be made in the late neonatal period as a result of general improvements in standards of living, progress will not accelerate and spread to the early neonatal period until appropriate maternal and neonatal health care is available and widely used IO LONGER FALLING BETWEEN THE CRACKS It is often argued that a radical reduction of the number of newborn deaths is possible nly where very high expenditure on health allows for large investments in sophist- cated technology. But in actual fact, nurses and doctors can easily acquire the neces ary skills without needing to become specialists. Countries such as Colombia and Sri Lanka, with fewer than 15 neonatal deaths per 1000 live births, have demonstrated that expensive technology is not a prerequisite for success. So have Nicaragua and Viet Nam, which lowered their neonatal mortality rates to 17 and 15 per 1000 births respectively, while their spending on health in the 1990s was only US$ 45 and USS 20 per capita, respectively. In northern European countries, well-coordinated antenatal intrapartum and postnatal care for mothers and newborns coincided with reduced rates of mortality before the introduction of neonatal intensive care in the early 1980s (8). Intensive care facilities, specialists and expensive equipment are useful to re duce neonatal mortality even further only after very low levels have already been achieved. Rather than deploying high-tech instrumentation, the challenge is to find a Box 5.2 Sex selection The low value given to women and girls in some juvenile(0-6 years) sex ratio declined from daughters-in-law and mothers, sex determina for boy children. Over the decades, this has with some of the steepest declines occurring gain control over at least one aspect of their translated into many practices that heavily dis- among the better educated and in economi- lives. criminate against girls, such as neglect in feed- cally better-off districts that also have greater This is a conundrum which cannot be ing, education and health care. The practice of access to commercial health services National resolved by focusing only on medical technol- female infanticide has also been documented records on sex ratio at birth in China and South ogy. The most severely affected countries such in some places Korea have shown similar rapid changes that as China, India and South Korea have all banned apidly declining fertility and the trend are unlikely to be sustainable in the long term. prenatal sex determination through the use of limit families to one or two children has The demographic impact of these adverse sex ultrasound or pre-conception techniques; other increased the desire of couples to have a boy. ratios is beginning to be felt in the form of a measures taken include registration and regu The emergence and increased availability of dearth of young women in some communities, lation of genetic laboratories and ultrasound ultrasound equipment, which can detect the thereby making women in general more vulner- machines and self-regulation by the medical sex of a fetus early in pregnancy, has opened able to violence, including sexual coercion and profession. Such policies have so far been p the opportunity for the commercial use of sale of brides. rgely ineffective because demand continues to medical technology to pre-select and terminate Many women s rights organizations and oth- be high. Various nongovemmental organizations pregnancies of female fetuses, thus reinforcing ers, in India and elsewhere, have seen prenatal and civil society organizations are currently the devaluation of girls and women. x selection as another form of discrimination involved in large-scale awareness and sensi Over the last decade, the ratio of girls to against women, and have been active in moves tization campaigns and in organizing a broader yys in the 0-6 year age group has become to have such selection banned. On the other social debate on the devaluation of females and creasingly skewed in a number of countries. hand, in societies where giving birth to sons the consequences of sex preference. For instance, India' s census revealed that the defines women' s status and rights as wives
newborns: no longer going unnoticed 85 were fuelled by improvements in health care provision, and most studies argue that they resulted from a number of factors including rising standards of living and nutrition, reduced fertility, safer water, better sanitation, and improved housing (26, 27). During this time, progress in reducing neonatal mortality was limited and was confined to the late neonatal period. Progress did not accelerate until around the time of the Second World War (28), which coincided with greater provision and use of maternal health care, improved quality of professional midwifery and obstetric services, and access to antibiotics. This suggests that, while some limited progress can be made in the late neonatal period as a result of general improvements in standards of living, progress will not accelerate and spread to the early neonatal period until appropriate maternal and neonatal health care is available and widely used. NO LONGER FALLING BETWEEN THE CRACKS It is often argued that a radical reduction of the number of newborn deaths is possible only where very high expenditure on health allows for large investments in sophisticated technology. But in actual fact, nurses and doctors can easily acquire the necessary skills without needing to become specialists. Countries such as Colombia and Sri Lanka, with fewer than 15 neonatal deaths per 1000 live births, have demonstrated that expensive technology is not a prerequisite for success. So have Nicaragua and Viet Nam, which lowered their neonatal mortality rates to 17 and 15 per 1000 births, respectively, while their spending on health in the 1990s was only US$ 45 and US$ 20 per capita, respectively. In northern European countries, well-coordinated antenatal, intrapartum and postnatal care for mothers and newborns coincided with reduced rates of mortality before the introduction of neonatal intensive care in the early 1980s (8). Intensive care facilities, specialists and expensive equipment are useful to reduce neonatal mortality even further only after very low levels have already been achieved. Rather than deploying high-tech instrumentation, the challenge is to find a juvenile (0–6 years) sex ratio declined from 945 girls per 1000 boys in 1991 to 927 in 2001, with some of the steepest declines occurring among the better educated and in economically better-off districts that also have greater access to commercial health services. National records on sex ratio at birth in China and South Korea have shown similar rapid changes that are unlikely to be sustainable in the long term. The demographic impact of these adverse sex ratios is beginning to be felt in the form of a dearth of young women in some communities, thereby making women in general more vulnerable to violence, including sexual coercion and sale of brides. Many women’s rights organizations and others, in India and elsewhere, have seen prenatal sex selection as another form of discrimination against women, and have been active in moves to have such selection banned. On the other hand, in societies where giving birth to sons defines women’s status and rights as wives, The low value given to women and girls in some countries is reflected in a marked preference for boy children. Over the decades, this has translated into many practices that heavily discriminate against girls, such as neglect in feeding, education and health care. The practice of female infanticide has also been documented in some places. Rapidly declining fertility and the trend to limit families to one or two children has increased the desire of couples to have a boy. The emergence and increased availability of ultrasound equipment, which can detect the sex of a fetus early in pregnancy, has opened up the opportunity for the commercial use of medical technology to pre-select and terminate pregnancies of female fetuses, thus reinforcing the devaluation of girls and women. Over the last decade, the ratio of girls to boys in the 0–6 year age group has become increasingly skewed in a number of countries. For instance, India’s census revealed that the daughters-in-law and mothers, sex determination and sex selective abortion allow women to gain control over at least one aspect of their lives. This is a conundrum which cannot be resolved by focusing only on medical technology. The most severely affected countries such as China, India and South Korea have all banned prenatal sex determination through the use of ultrasound or pre-conception techniques; other measures taken include registration and regulation of genetic laboratories and ultrasound machines and self-regulation by the medical profession. Such policies have so far been largely ineffective because demand continues to be high. Various nongovernmental organizations and civil society organizations are currently involved in large-scale awareness and sensitization campaigns and in organizing a broader social debate on the devaluation of females and the consequences of sex preference. Box 5.2 Sex selection