Trends in national and subnational wealth related inequalities in use of maternal health care services in Nepal: an analysis using demographic and health surveys (2001–2016)

Maternal health affects the lives of many women and children globally every year. Global maternal mortality remains unacceptably high as a consequence of pregnancy and childbirth-related problems [1]. Two-thirds of the countries had a maternal mortality ratio (MMR) of 420 per 100,000 live births or greater as of 2015 [1, 2]. The overwhelming 80-fold difference in the estimated lifetime risk of maternal mortality in low-income countries, as compared to high-income countries, points to the diligence of insightful inequality that must be addressed [3]. Inequalities in access to care particularly for vulnerable populations; poor quality of available care; grave deficiencies in health system infrastructure and workforce; and the impact of economic, political, socio-demographic and environmental factors all contribute significantly to the risk of poor maternal health outcomes and hinder progress toward reduction in mortality and morbidity [4‐9]. In this situation, attaining the sustainable development goal (SDG) will not be possible without reducing the burden of the deprived in all population subgroups.

Nepal has made remarkable progress with respect to improving the situation of maternal health in the last two decades [10, 11]. Percentage of pregnant women with at least four antenatal care (ANC) visits has increased from 14% in 2001 to 69% in 2016. Similarly, more women (58%) were assisted during delivery by a skilled provider in 2016 compared to the situation in 2001 (11%) [12]. Though the government is committed to promoting equity in distribution and utilization of health services through its health policies and sector strategies, the studies have found the unfair distribution of benefits from public investment on health across different population sub-groups [11, 13]. Inequitable access to and utilization of maternal health services due to financial, socio-cultural, and geographical barriers, among others are key challenges [14].

Maternal health has been one of the high priority programs of the Government of Nepal (GoN). The Government aims to reduce the maternal mortality ratio to 70 per 100,000 live births by 2030 as part of the SDGs target from 239 in 2016 [15]. To improve the situation of maternal health, GoN has implemented various demand and supply-side interventions. Regarding demand-side intervention, the maternity incentive scheme was initiated in 2005 that made a provision to pay a fixed amount to mother/family to help with the cost of transportation. In 2009 the benefit under the scheme was broadened and the user fee associated with all types of delivery care was removed (Aama program) [13]. Further, in 2012 incentives for completing four ANC with the recommended schedule followed by institutional delivery (introduced in 2009 as separate demand-side financing) was added to the Aama program [16, 17]. Regarding supply-side interventions, extensive expansion of service delivery sites was made through the establishment of health posts and sub-health posts throughout the country following the endorsement of National Health Policy 1991. The Safe Motherhood Policy and Plan of Action (1994–1997); National Safe Motherhood Policy 1998; National Safe Motherhood and Newborn Health Long Term Plan (2006–2017); and National Policy on Skilled Birth Attendants (2006) introduced key policy interventions regarding the availability of maternal health care in rural and remote areas. The examples of such interventions are: establishment of birthing centers (BCs) in health posts, establishment of emergency obstetric centers (EOC) in primary health care centers and district hospitals, availability of skilled birth attendant training to nursing staffs and doctors of BCs and EOCs in training sites, and the strengthening of the referral services.

The National Health Policy 2019 envisions equitable access to healthcare with respect to all the population groups and emphasizes on providing a high priority to the poor and vulnerable [18]. Similarly, Nepal Health Sector Strategy (NHSS) (2015–2020) has recognized inequity as one of the key challenges of the health sector in Nepal [14], though the government has implemented special incentives such as free health care programme and safe delivery incentive scheme with an objective to reduce inequity in healthcare utilization. Further, equity and access are identified as one of the four strategic pillars by NHSS to move towards universal health coverage, other three being: quality, health sector reform and multi-sectoral collaboration. Nepal Health Sector Plan [19, 20] has planned to strengthen the health system so that the poor and vulnerable communities have the priority for access. National Health Policy 1991 established primary health care in Nepal (sub-health post in every ward) and provided special emphasis to the population of rural areas [21]. SDGs in Nepal, structured on three dimensions: economic, social and environment, are set with a vision of transforming Nepal into a prosperous nation with equity and social justice [15]. With all these efforts and many others, the equity gap in accessing and using maternal health in the general population has decreased over the period of 2001 to 2016 as presented by different papers [11, 13]. In this paper we have analyzed whether the equity gap in utilization of key maternal health services has decreased across seven provinces over this period as the geography of Nepal has already been identified as one of the predictors of availability, access and utilization of maternal health services [13, 14]. The constitution of Nepal promulgated in 2015 has institutionalized Nepal as a Federal Democratic Republic country with seven provinces. Evidence on the status of equity gap and progress over time related to maternal health indicators across the provinces (subnational) will help policymakers and planners at the federal and provincial level equitably allocate scarce resources. Therefore, it is high time we measure whether the distribution of health indicators at the sub-national level is equitable and provide evidence to pursue the agenda of Universal Health Coverage (UHC) at different level.

This section begins with a descriptive analysis of the variables used in the analysis. Table 1 provides the national and provincial level statistics for the three indicators. Figures 2, 3 and 4 and Table 2 provides detailed findings on concentration curves and indexes disaggregated by the provinces. Finally, Table 3 ends with an explanation of temporal changes in concentration index at the national and provincial level.

The main aim of this study was to find the trends and inequalities at national and subnationall level with respect to the use of maternal health services in Nepal using the nationally representative cross-sectional study (Demographic and Health Survey) over the period of 2001 to 2016. In general, the study found that the status of inequality in utilization of maternal health services in Nepal depicted by three indicators (at least four ANC visits, Institutional delivery and CS delivery) has improved remarkably over the survey period. However, the progress is not proportional across seven provinces. Similar findings were obtained in other studies based on NDHS data [11, 13].

The increase in the uptake of maternal health services during this period, such as institutional delivery, has been reported to be associated with the mix of supply and demand side investments made by the GoN [17]. Strengthening of the healthcare delivery system (supply side) specifically since the late 1990s (Safe motherhood long term plan) has made maternal health more accessible to the poor and women from hard to reach geography [36]. The example includes the construction and expansion of birthing centers (BCs) in most disadvantaged areas, strengthening of BEOC/CEOC, production of auxiliary nurse midwives (ANMs) and training of skilled birth attendant (SBA to nursing staffs and advanced SBA to medical officers), training of nurse-midwives as anesthetic assistants and strengthening of blood banking [17, 36, 37]. Similarly, Demand Side Financing (DSF) policy in maternal health, introduced first in 2005 and revised in subsequent years, is also reported to be associated with the increase in access and utilization of maternal health care in Nepal [17]. Studies conducted in other low and middle-income countries (India, Bangladesh) to understand the effect of demand side financing on maternal health outcomes has shown that cash incentives can contribute in increasing the utilization of ANC and institutional delivery services [38, 39]. The targeted voucher programs implemented in Cambodia and Pakistan showed improved access to institutional deliveries by poor women and thus reducing inequalities in utilization [40, 41]. An assessment conducted by the Ministry of Health and Population has shown that demand side financing in Nepal was associated with an increase in uptake of maternal health services such as institutional delivery [17, 42]. Similarly, another analysis has shown that the demand side financing in maternal health had a major effect on utilization of at least four ANC and institutional delivery [16].

With respect to the use of at least four ANC visits, we observed that the concentration curves are gradually moving towards the line of equality with each subsequent survey period (2001, 2006, 2001, 2016). The shift towards the line of equality is more prominent between the survey period of 2011 and 2016. The movement towards equality is progressive and statistically significant. The combined effect of both supply and demand side interventions in maternal health can be attributed to such a change. Talking about the Provinces, somehow consistent improvement in the status of inequality with respect to the use of at least four ANC services was observed in Province one, Bagmati Province, Lumbini Province and Sudhurpaschim Province. Bagmati Province has witnessed a huge and significant decrease in inequality, particularly during the survey period 2001–2006 and 2011–2016. Similarly, Province one, Lumbini and Sudurpaschim have witnessed a statistically significant decrease in concentration index between 2006 and 2011. Particularly, statistically significant and progressive change is observed at both national and provincial level during the survey period of 2011–2016, however, Karnali province and Province two are comparatively lagging behind. The studies have also shown that the benefits of DSF policies are skewed towards areas that are comparatively better off in terms of wealth and geography [17], such as Province one, Bagmati Province and Lumbini Province that are more accessible and wealthy. In contrast, the women from mountain areas (Karnali Province) have found to be comparatively least benefited from the DSF interventions [43]. So the finding in our study can be linked with earlier studies that indicated limitation of DSF to sufficiently address geographical remoteness and poor transport links [41, 44‐54].

With respect to the institutional delivery, the greatest decrease in inequality was observed between 2011 and 2016 at the national and subnational level, except for Province two and Karnali. It has been shown that due to lack of various supply side factors such as birthing center services, trained health workers and lifesaving surgical provision in hard-to-reach geography (high hill and mountain area of Nepal) the progress towards the utilization of maternal health services is relatively poorer compared to more accessible areas [17]. It is also argued that since the health facilities in these Provinces have to be accessed only with hours (in some cases days) of walking on foot due to lack of transport facilities, the opportunity cost associated with this plays a crucial part in reducing the demand [55]. Though Province two is plain, this province is one of the poorest with respect to the overall socio-economic status [56, 57]. So, available evidence supports our finding that the progress in addressing the level of inequality is sluggish in Karnali Province and Province two. Bagmati Province, Lumbini Province and Sudurpaschim Province have made comparatively better progress towards reducing inequality, which is as expected because these are comparatively better off regions in terms of socio-economic development and geographical positioning [56, 57]. There was no significant improvement in the level of inequality between 2001 and 2006, even at the national level. Possible reasons could be that the DSF in maternal health was started in 2005 and rapid maternal health service expansion through BCs/EOCs accompanied with SBA training was started after 2006. Similarly, Safe Motherhood and Neonatal Health Long Term Plan (2006–2017) initiated after 2006 and associate interventions at community level that focused on birth preparedness and institutional delivery could also explain this finding. However, some improvement was observed during 2006–2011 at national and provincial level (Province one only).

With respect to CS delivery, the concentration index is observed to be decreased over the period of 2001–2016 at national level, however, the changes are not statistically significant. Similarly, the curves are shifting towards the line of equality over the period between 2001 and 2016, a greater shift observed while moving from 2011 to 2016. Similar study using DHS data found an increase in concentration index over this period of survey years [11]. A study conducted in 2016 in western Nepal shows that odds of going through CS delivery was four times more in women from urban areas compared to that of rural areas [58]. Similarly, the likelihood of using CS delivery was found 10 times more in women from the richest quintile compared to poorest [11]. A study from a tertiary hospital in Kathmandu reports a dramatic rise in CS rates, from around 21% in 2004 to 39% in 2014 [59].

Disproportionate in status of utilization of maternal health services and unequal progress towards equality could be linked to various factors. One such factor, as pointed out in other papers as well, is that the Government’s blanket approach in delivering maternal health services across the country [60], with greater emphasis on achieving national targets and less attention to monitor subnational targets and equality [26]. An example is the universal nature of demand side financing in maternal health in Nepal that aims to cater the maternal health care needs of all without specific focus on the women that have barriers due to geographic and socio-economic factors, such as women form Province two and Karnali Province. This also shows that the investment of government in maternal health including the DSF initiatives of GoN could not address the geographical, financial and social barriers present differently across seven provinces [38]. Second is linked with the difference in level of availability and accessibility of health facilities in different provinces (supply side investment). A national level survey on availability and readiness of health services shows that the percentage of facilities providing normal vaginal delivery is lowest in Province two, whereas one of the lowest percentages of health professionals are trained for normal delivery in Karnali Province. Province two has the lowest percentage of health professionals trained for ANC [61]. So the gaps in the situation of supply side investment in different provinces witnessed in terms of availability of services, essential medicines and readiness (in terms of skill, equipment, medicine) may explain the inequality in service utilization [61]. Another study using the nationally representative health facility survey showed that the mean general health service readiness score was lowest in Province two [62]. Third is the underlying socio-economic and geographical position of the provinces. The Human Development Index in 2011 was lowest of Karnali Province (0.475), followed by Suderpaschim Province (0.478) and Province two (0.485), and the lowest annual change was observed in Province two since 1996 [56]. Fourth is lack of special focus of government towards provinces that are geographically challenged and socio-economically lagging behind (already discussed above). So, these evidence points out towards the need of special strategies from the government side tailored across the provinces.

The study has a few limitations that readers should be aware of while interpreting findings. First, this study describes the maternal health services inequalities at national and provincial level and does not take any predictors of inequalities in maternal health services into account. However, the study discussed the observed changes in inequalities in light of government policies during that period. Second, the analysis covers multiple rounds of cross-sectional surveys to analyze the trend in inequality. However, in the survey round of 2016, the strata were different from that of previous surveys. For the comparison purpose, we did not take into account the strata in survey design analysis. This generally affects the standard error of the revised estimates but the approach ensured comparability of methods employed across the survey rounds. Third, the frequency distribution of some of the indicators across provinces is very low in the year of 2001 and 2006. This might be the reason for statistically non-significant results despite a noticeable shift in concentration curve. These limitations, however, do not dilute the relevance of the findings for the policy makers. Fourth, the frequency distribution of CS delivery is very low when stratified across the provinces. Therefore, study reported the concentration curves at national level only. Fifth, the frequency distribution of indicators are very low for Karnali Province in 2001. We therefore could not produce a meaningful concentration curve for Karnali Province for 2001. Despite these limitations, the study reports observed trends in inequalities at national and subnational level, which can be useful for policy makers at both levels. The study has observed a narrowing of inequality in use of CS services. However, the complete analysis also requires the systematic analysis of the fact that CS are effective in saving maternal and infant lives only when they are required for medically indicated reasons. Such analysis requires the Robson classification at admission [63] for the pregnancy cases which is not available in the current dataset. However, inequality in utilization of CS services itself reflects useful information to the policy makers for taking pro-poor actions to bridge the gap.

The study has found a remarkable improvement in the distribution of maternal health indicators across wealth over four consecutive DHS surveys in Nepal. The inequality in utilization of maternal health services among different wealth quintiles is observed to be decreasing with every subsequent survey at national level. However, if the statistics are disintegrated at the subnational level, the progress towards decreasing inequality in utilization of maternal health services among different wealth quintiles is not the same across different provinces; the progress is also not uniform across different survey years. Provinces which are comparatively disadvantaged in terms of underlying factors such as geography and socio-economic status (Province two, Karnali Province) have made minimal to zero progress towards reducing inequality in comparison to other provinces, particularly Province one, Lumbini Province and Bagmati Province. Among the maternal health indicators studied here, CS delivery seems to have made less improvement compared to ANC and institutional delivery. Additional research may be needed to explore the reasons and their implication. Special investment to address barriers to access and utilization in provinces that are lacking to make progress in reducing inequality is urgent. Further studies are needed to understand the strategies required to address the gaps in these provinces and bring about fair improvement.