How did Nepal Reduce the Maternal Mortality ? A Result from Analysing the Determinants of Maternal Mortality

Nepal reportedly reduced the maternal mortality ratio by 48% within one decade between 1996-2005 and received the Millennium development goal award for this. However, there is debate regarding the accuracy of this fi gure. On the basis of framework of determinants of maternal mortality proposed by McCarthy and Maine in 1992 and successive data from Nepal demographic health survey of 1996, 2001 and 2006, a literature analysis was done to identify the important factors behind this decline. Although facility delivery and skilled birth attendants are acclaimed as best strategy of reducing maternal mortality, a proportionate increase in these factors was not found to account the maternal mortality rate reduction in Nepal. Alternatively, intermediate factors particularly women awareness, family planning and safe abortion might have played a signifi cant role. Hence, Nepal as well as similar other developing countries should pay equal attention to such intermediate factors while concentrating on biomedical care strategy. _______________________________________________________________________________________


BACKGROUND
In September 2010, Nepal received the Millennium Development Goal (MDG) award from the United Nations for its exceptional achievement in reducing maternal mortality. 1Nepal had one of the highest maternal mortality ratio (MMR) in the past decades.World Health Organisation (WHO) and United Nation Development Programme (UNDP) estimated it as 1500 in 1990, 826 in 1995 and 670 in 2005 while National Demographic Health survey (NDHS) calculated it as 515 in 1991, 539 in 1996 and 281 in 2006.Further, a recent maternal mortality and morbidity study estimated the fi gure as 229 in 2010, using a surveillance system where community informants reported births and deaths of women of reproductive age. 2 Since Nepal does not have a functioning complete death registration system of maternal deaths, maternal mortality estimation is diffi cult.NDHS 2006 of Nepal used direct sisterhood method that asks respondents to provide information about their sisters; while WHO in 2005 estimated MMR by means of regression analysis.Estimation of maternal mortality is methodologically varied and may be affected by possible under-reporting or chances of underestimation in poor surveillance system. 3 the other hand, there is debate regarding the extent of decline in the maternal mortality during the civil war period of 1996-2005. 2,4,5During the war period, provision of adequate health services in rural areas was affected due to destruction of health posts and threatening to health workers.This had a negative impact on population health. 6Nepal has still large gender disparity, with women often experiencing sexual violence in their own houses and thus, losing their control over their own reproductive life. 7Further, Nepal's hilly geographical terrain has made it diffi cult to provide health services to rural villages.Since the mortality reduction is an impact indicator of maternal health improvement, the reduction must come through input as well as process indicators i.e the various determinants of maternal mortality.The aim of this paper is to analyse the determinants of maternal mortality in the context of maternal mortality reduction in Nepal so as to explore the important factors behind the signifi cant decline of MMR in Nepal over the past decade.

METHODS
Electronic databases of Medline, PubMed, Science Direct, CINAHAL, and Biomed Central were searched with key words: Nepal, maternal health improvement, maternal mortality, maternal health service, safe motherhood, progress, MDGs.Grey literature was searched with Google search engine to retrieve relevant governmental and non-governmental websites, study reports, documents, and policies.All types of articles including original, review, editorial, commentary, and national newspaper reports were considered.Data was derived mainly from successive Demographic Health Survey of Nepal 1996, 2001 and 2006 for comparison on related determinants of maternal mortality.Framework of determinants of maternal mortality proposed by McCarthy and Maine in 1992 was used. 8planation about determinants of maternal mortality was given based on the framework of McCarthy and Maine and then, analysis of progress on those determinants in Nepal was done based on data (Table 1) followed by a conclusion with plausible factors associated with the decline.

Determinants of maternal mortality
McCarthy and Maine provided a detailed framework of determinants to analyse maternal mortality and morbidity. 8They identifi ed three types of determinants: distal, intermediate, and proximal or immediate outcomes.Distal determinants include women's status in family and community such as education, occupation, income, and social and legal autonomy; family's status in community such as family income, land, education and occupation of family members; and community's status such as community tradition and resources.Intermediate determinants include health and reproductive status of women, and accessibility and utilisation of health services.These two types of determinants operate to produce immediate cause of maternal mortality, which starts from pregnancy and might end in pregnancy related complications such as haemorrhage, infection, pregnancy-induced hypertension, obstructed labour.According to this framework, all strategies to improve maternal health should target either: i) to reduce the likelihood that a woman will become pregnant; or ii) to reduce the likelihood that a woman will experience a serious complication of pregnancy or childbirth; or iii) to improve the outcomes for women with complications.Therefore, it follows that family planning or late marriage and early detection, management and preparedness of birth can contribute to reduction of MMR and improvement of maternal health.
On the other side, most maternal deaths in developing countries were reported during labour, delivery or the fi rst 24 hours of postpartum, with obstetric haemorrhage being the main immediate cause. 10A recent 2009 study of maternal mortality and morbidity in eight districts of Nepal reports that haemorrhage is the leading cause and that percentage contribution of eclampsia, abortion related complication, gastroenteritis and anaemia have more than the similar study in Nepal in 1998. 2 So, if the immediate causes are managed, maternal mortality can be reduced signifi cantly.Accordingly, the universal maternal mortality reduction strategy is mainly focussed on intrapartum and complication related care.This care is best provided at health facility, especially at primary level, whereby women can have access of skilled care easily by skilled birth attendants. 11,12 hile such biomedical strategy is helpful to save the mother's life, the real sustainable success criteria should also include the indicators that show the reduction in the likelihood that a woman will become pregnant or that she will experience a complication.In developing countries like Nepal, early marriage, unmet need of family planning and unsafe abortion is still high. 13,14 his means, we must look further in distal and intermediate determinants as well to make a sustainable achievement rather than narrow focus on immediate outcome curative strategy.

Socio-cultural environment in Nepal and distal factors
Nepal can be described as a diversifi ed country in terms of geography, ethnicity, languages and culture.Many deeply rooted cultural practices are still prevalent, which might have negative implication for expectant mother and new child.For examples, postpartum women are forbidden from leaving the house or touching anyone for at least 11 days after delivery; and there is superstitious belief that women should not cross a river during pregnancy.Care seeking in pregnancy and childbirth is often perceived shameful. 2More than 80% of land in Nepal is rural area covered with hills and mountains, which are very diffi cult to access and provide challenges for health service provision. 15he diffi cult terrain is particularly challenging while accessing maternity services.Though many rural areas have recently been connected with road network, these roads are often not functional throughout the year and reliable and effi cient vehicles may not be available when required.The transport of women just before and after the delivery to the nearest equipped birthing centre is expensive and troublesome because the pregnant women should be accompanied by her relatives, which increases the logistics in various ways such as lodging if the facility is not near.
Overall, Nepalese society is patriarchal and male dominated, where women have limited autonomy for household decision including their own reproductive health decision and resources handling. 16ome women of lower caste are denied access to skilled birth attendance.Bruson, from her ethnographic study in a semi-urban area of Kathmandu concludes that women still cannot decide and demand the need of obstetric care and men are not involved in the delivery issue until there is urgency. 17There is no signifi cant increase in women's literacy from 1996 to 2006.Similarly, age at marriage and age at fi rst birth of women has also not changed much (Table 1).Early marriage and early motherhood are still quite common among Nepalese women, especially in rural and marginalised areas. 18oor nutrition contributes to poor maternal health with various obstetric complications.Women's and children's nutritional condition have progressed very little over the decade (Table 1).The 2006 Demographic survey reports that still 42% of pregnant women are anaemic in Nepal.Christen et al. reports that low maternal age, malnutrition (thin arm) and illness in early to mid-gestation were associated with risk of death during pregnancy and the fi rst year postpartum in rural Nepal. 19wever, women's situation has also progressed in past few years.Radio programmes on maternal health, maternal health information and mother's education programme has contributed to raise women's status. 20omen's empowerment has increased in Nepal and can be attributed to the adaptation of inclusive gender policy, which allow women in civic participation and in household property.Women having land are signifi cantly more likely to infl uence the household decisions with improved nutrition for their children. 21However, much benefi t may not be guaranteed until such policies are truly translated into action under prevailing social and cultural environments.
Total fertility rate has reduced while contraceptive prevalence rate has increased from 1996 to 2006.One of the plausible explanations is that the spouses of many married couples in Nepal are abroad for years for employment, which might have played role in fertility reduction and thus less pregnancy.However, unmet need for family planning has reduced slightly.Many pregnancies in Nepal are unplanned and unwanted. 22 hospital-based study conducted in 1984-85 at fi ve major hospitals in and around Kathmandu valley reported 1576 cases of abortion related complications. 23ccording to NDHS survey 2006, about one-third of pregnancies is unplanned and thus might seek abortion.Chances of abortion are high when the foetus is female. 24Informal and unmanaged abortion is one of leading causes of maternal mortality. 25Legalisation of abortion in 2002 in Nepal, therefore, has a larger role in reduction of maternal mortality. 26A major decline in maternal mortality in Matlab region of Bangladesh was due to fall in abortion related deaths. 27Similarly, maternal mortality ratio fell by 50% in Romania after legalisation of abortion since 86% of maternal mortality ratio before were caused by complications from unsafe abortion. 28

Safe motherhood, service provision and intermediate factors
Following Nairobi Conference of Safe Motherhood in 1987, Nepal formulated the national safe motherhood plan of action and safe motherhood programme was launched by family health division of ministry of health. 2 It now has a national policy for skilled birth attendants (2006), national safe abortion policy, and safe motherhood and newborn health long term plan (2006-2017).The safe motherhood programme has received substantial international support and initiated a variety of promotional activities. 29Improvement in maternal health has been linked to efforts of safe motherhood activities. 30rstly, there is increased availability and accessibility of maternal health services through an extensive network of district health system: district hospital, primary health care centre, health post and sub-health post, with female community health volunteers at the grass root level.Overall, there is substantial growth in health facilities reaching the peripheral area. 31The number of primary health care centres that provides birthing facilities increased more than double from 2004 to 2007.Similarly, number of sites of Basic emergency obstetric care and Comprehensive emergency obstetric care has increased. 32condly, more human resources for skilled birth attendants have been produced in recent years.Female community health volunteers, who link community with health system by providing advice on family planning, maternal and child health, have been extensively deployed.Maternal and child health worker provides antenatal and postnatal care at sub health post, and they are trained for skilled birth attendant after a short training. 33Auxiliary nurse midwife provide antenatal and post natal care at health posts, some of which have birthing centres.Medical offi cer or health assistant or auxiliary nurse midwife provide antenatal, delivery and postnatal care in primary health care centres, some of which provide Basic emergency obstetric care.However, in the mountainous and hilly regions, availability of human resource especially staff nurse and doctors in health centre is still poor and women do not get the care even after visiting the centres. 34irdly, free delivery care and safe maternity delivery incentive has increased fi nancial accessibility of women to health service.Safe maternity delivery scheme was introduced in 2005 and provides cash incentive to women who deliver at public and few private facilities.
The incentive was mainly targeted to compensate the transport expense and alter women and family's behaviour, to motivate them to come to health facility.However, there are challenges in effi cient management of funds, fair distribution and its suffi ciency. 35Besides, just providing a small fund of money is not long-term service oriented solution.

Non-Utilisation and immediate consequences
On the other hand, despite availability and free delivery care, tendency of home delivery continues and institutional delivery has not increased much.For example, only 18% of delivery occurred in health care facility in 2006 whereas the target by 2015 is 60%.The rise in postnatal care is extremely low when compared with antenatal care.There might be either no perceived need, or the facility is far away and care is costly, or peripheral facilities may be perceived of lower quality and not consulted.Since antenatal visits have risen sharply, women just do not go health facilities for delivery because of perceived no benefi t by having delivered in accessible lower level facility.Bypassing of peripheral facilities and lack of effi cient referral system in health care system of Nepal is quite common. 36saggregated data by caste, ethnicity, economic status and place of residence has wide disparity in terms of utilisation of maternity services. 37Utilisation of skilled care varies from a fi gure of 33% in central hill to a low of 8% in western mountain area.Women in the poorest quintile had approximately 12 times less access to skilled care compared to their richest quintile. 38Therefore, even when national data may indicate achievement of MDG, the disaggregate data will be far behind the target.It is mainly due to unequal development among the regions, inequity in resource provisions among the different regions and pocket areas of poverty and marginalised population.
It is well documented that the most maternal deaths occur due to complications around delivery and that institutional help is indispensable to avoid most of the maternal deaths.Many countries, especially the developed ones, reduced maternal mortality signifi cantly by means of providing midwifery care.Sweden developed an expanded network of midwifery service that reached even to household and reduced its maternal mortality signifi cantly early in 19 th century. 39MMR declined sharply in England, the Netherlands and other European countries only after availability of antibiotics for infection, blood transfusions for haemorrhage, and improved surgical techniques for caesarean section. 12owever, Qiu et al. argued that maternal mortality reduction in Zhejiang Province of China corresponded with increase in economic status and improvement in social factors like maternal safety. 40ere are other strategies that have been shown effective in MMR reduction.Sri Lanka, in particular, established a system of health facilities with midwifery care and family planning services around the country. 12MR declined substantially in a district of Bangladesh as a result of community-based family planning and maternity care. 41In Ethiopia, Community based obstetric care that included informed community participation with special attention to prenatal screening through an outreach programme with experienced nurse-midwives, a functionning referral system, and a maternity waiting home for women reduced the number of women going to a tertiary care hospital as well as the maternal mortality. 42It seems that the improvement comes when midwifery service is coupled with family planning and safe abortion, and that such service be available at community or even household level.It is paradox to note that without much increase in health facility utilisation, Nepal has reportedly reduced the maternal mortality substantially.This clearly indicates some factors out of the direct health service related factors might have played the signifi cant role.For example, as shown earlier, legalisation of abortion, increased women autonomy, and family planning methods improved in Nepal during past decade and this, in turn, might have contributed positively to improved maternal health.Decline in fertility and high risk pregnancies has been found strongly associated with maternal mortality reduction in Nepal.

43 CONCLUSION
According to data from successive NDHS of 1996, 2001 and 2006, Nepal has reduced maternal mortality rate signifi cantly but utilisation of facility delivery services has not increased proportionately.The maternal health improvement is a result of various factors.Especially, there is good indication that the maternal health improvement has resulted more from improvement in intermediate or distal factors rather than improvement in immediate casual factors.With continuation of emphasis in intermediate factors such as age of marriage and family planning, Nepal should focus on strategies on increasing utilisation of quality skilled care and referral chain from household to tertiary facility to continue the decreasing trend, and further reduction of maternal mortality to achieve the MDG.