Improving Emergency Obstetric Care in a Context of Very High Maternal Mortality: The Nepal Safer Motherhood Project 1997–2004

Abstract

The Nepal Safer Motherhood Project (1997–2004) was one of the first large-scale projects to focus on access to emergency obstetric care, covering 15% of Nepal. Six factors for success in reducing maternal mortality are applied to assess the project. There was an average annual increase of 1.3% per year in met need for emergency obstetric care, reaching 14% in public sector facilities in project districts in 2004. Infrastructure and equipment to achieve comprehensive-level care were improved, but sustained functioning, availability of a skilled doctor, blood and anaesthesia, were greater challenges. In three districts, 70% of emergency procedures were managed by nurses, with additional training. However, major shortages of skilled professionals remain. Enhancement of the weak referral system was beyond the project’s scope. Instead, it worked to increase information in the community about danger signs in pregnancy and delivery and taking prompt action. A key initiative was establishing community emergency funds for obstetric complications. Efforts were also made to develop a positive shift in attitudes towards patient-centred care. Supply-side interventions are insufficient for reducing the high level of maternal deaths. In Nepal, this situation is complicated by social norms that leave women undervalued and disempowered, especially those from lower castes and certain ethnic groups, a pattern reflected in use of maternity services. Programming also needs to address the social environment.

Résumé

Le projet népalais pour une maternité à moindre risque (1997–2004) a été l’un des premiers grands projets à se centrer sur l’accès aux soins obstétricaux d’urgence, couvrant 15% du Népal. Six facteurs de réduction de la mortalité maternelle ont été choisis pour évaluer le projet. La satisfaction des besoins en soins obstétricaux d’urgence a augmenté en moyenne de 1,3% par an, atteignant 14% en 2004 dans les centres publics visés par le projet. L’infrastructure et les équipements se sont améliorés, mais le fonctionnement, la disponibilité d’un médecin qualifié, de sang et d’anesthésie demeuraient problématiques. Dans trois districts, 70% des procédures d’urgence étaient pratiquées par des infirmières ayant suivi une formation complémentaire. Néanmoins, les professionnels qualifiés continuent de manquer cruellement. L’amélioration du système d’aiguillage n’entrait pas dans le champ du projet qui a préféré apprendre à la communauté à reconnaître les signes de danger pendant la grossesse et l’accouchement et à y réagir rapidement. La création de fonds communautaires pour les complications obstétricales constitue une initiative clé. Le projet a également encouragé un changement de comportement en faveur de soins à l’écoute des patientes. Les interventions centrées sur l’offre ne suffisent pas à réduire le niveau élevé de décès maternels. Au Népal, cette situation est compliquée par des normes sociales défavorables aux femmes – particulièrement quant elles appartiennent à des castes inférieures ou à certains groupes ethniques – qui se reflètent dans beaucoup de services de maternité. La programmation doit aussi agir sur l’environnement social.

Resumen

El proyecto de Maternidad más segura (1997–2004), en Nepal, fue uno de los primeros proyectos de gran escala que se centró en el acceso a los cuidados obstétricos de emergencia, y abarcó el 15% de Nepal. A fin de evaluar el proyecto se aplican seis factores para lograr disminuir la tasa de mortalidad materna. En 2004, hubo un aumento anual promedio de un 1.3% para cubrir la necesidad de cuidados obstétricos de emergencia, y se alcanzó el 14% en establecimientos del sector público, en distritos del proyecto. Se mejoraron la infraestructura y el equipo para lograr el nivel de atención integral, pero resultó más difícil lograr funcionamiento continuo, así como disponibilidad de un médico calificado, sangre y anestesia. En tres distritos, el 70% de los procedimientos de emergencia eran manejados por enfermeras, con capacitación adicional. Sin embargo, aún existe una gran escasez de profesionales calificados. Quedó fuera del alcance del proyecto mejorar el sistema de referencia deficiente. En vez, se trabajó para aumentar la información de la comunidad respecto a los signos de alarma del embarazo y el parto, y la toma de medidas con prontitud. Una iniciativa importante fue establecer fondos comunitarios de emergencia para complicaciones obstétricas. Además, se realizaron esfuerzos por fomentar un cambio positivo en actitudes hacia la atención centrada en la paciente. Las intervenciones relacionadas con los suministros son insuficientes para disminuir el alto índice de muertes maternas. En Nepal, esta situación se complica por las normas sociales que dejan a las mujeres subvaloradas y sin poder, especialmente aquéllas de clases más bajas y determinados grupos étnicos, un patrón reflejado en el uso de los servicios de maternidad. Los programas deben tener en cuenta el ambiente social.

Key Words:

• maternal mortality
• emergency obstetric care
• human resources
• health financing
• provider–patient relations
• Nepal

The Nepal National Safe Motherhood Programme was initiated in 1997 with the goal of reducing the maternal mortality ratio, at that time among the highest in South Asia at 539 per 100,000 live births (Nepal Family Health Survey 1996). It was designated a national priority programme, receiving financial and technical support through a wide range of donor projects. The Nepal Safer Motherhood Project (NSMP), which operated from 1997 to 2004, was a collaborative intervention between the Nepal Ministry of Health and Population and the UK Department for International Development (DFID), managed by Options Consultancy Services. It was one of the first large-scale projects to focus on access to quality emergency obstetric care.

High maternal mortality and morbidity in Nepal are associated with both socio-economic and health system issues. The low status of women, their lack of voice in reproductive health matters and adherence to cultural practices and taboos around pregnancy and childbirth continue to have significant repercussions. Education and literacy levels, though now rising, are generally low, especially among women, making it difficult to change established practices and improve understanding of safe motherhood issues.

Nepal’s challenging terrain and poor communication network mean that travel to a referral centre is often expensive and difficult. During the last ten years, the armed conflict has further exacerbated the situation. A project study in 2003 found that the armed conflict had affected women’s access to emergency obstetric care through raised financial and travel barriers.¹

Efforts to improve maternal health need to be informed by critical accounts of the complexities of programme implementation. Case studies of safe motherhood programmes from a range of developing countries with successful maternal mortality reduction are instructive. Koblinsky and Campbell,² for example, identified six factors for success in reducing maternal mortality ratios in a review covering Bolivia, China, Egypt, Honduras, Indonesia, Jamaica and Zimbabwe:

•	high availability of birthing facilities,
•	increased availability of a skilled birth attendant located near the home,
•	formalised referral links between facilities, beginning with providers at the community level,
•	free or reduced costs for services and transport to services,
•	public accountability for providers’ performance, and
•	strong government policy guidance.

We use Koblinsky and Campbell’s framework to analyse the Project’s contribution to safe motherhood in Nepal. In Phase 1, the Project focused mainly on improving midwifery and emergency obstetric services in selected health facilities in three districts (Baglung, Kailali and Surkhet). Two main components were developed: i) management of service provision for women of reproductive age, including improvements to the physical infrastructure of hospitals, equipment and supplies, and training of personnel; and ii) increasing access to midwifery and obstetric services by improving the social context to enable women to utilise services. Following a mid-term review in 2000, Phase 2 extended the Project to six more districts (Parbat, Myagdi, Rupendehi, Nawalparasi, Jumla and Dailekh), covering in total approximately 15% of the population. Project districts were located in the poorer regions of the mid-and far west regions of the country.

This paper draws upon extensive project documentation: seven internal evaluations and activity-to-output reviews, ten external evaluations and studies of key project components^3–12 as well as research studies commissioned during the Project,^1,13 other published papers,^14–17 and a project evaluation synthesis report.¹⁸

Findings

Availability of birthing facilities

Birth at home is the prevalent form of delivery in Nepal. In 2001, 88.9% of deliveries occurred at home with little change over the previous decade¹⁹ and the proportion was still high at 81.0% in 2006.²⁰ Pregnant women conceive of hospital care as emergency care, and home births tend to be preferred because of the flexibility of payment, convenience and the comfort of a familiar birth attendant.⁷

The central aim of the Nepal Safer Motherhood Project was to bring about an increase in utilisation of quality basic and comprehensive essential obstetric care to avert deaths from complications. Progress during the project showed a slow but steady advance in meeting the need for emergency obstetric care, as measured by the UN process indicators. Met need for emergency obstetric care was <5% in the Phase 1 districts in 1997.²¹ The average annual increase in met need has been 1.3% per year over the intervention period, bringing it to the 2004 level of 14% in public sector facilities in project-supported districts,²² In a further four districts supported by UNICEF, met need increased from 1.9% to 16.9% between 2000 to 2004.²³ Pooling district data masks large inter-district variation in annual met need increases, which ranged from -0.1% to 3.3% in NSMP districts and from 1.1–5.9% in UNICEF project districts in the same period.²²

The total population of the nine intervention districts was about three million, requiring, according to UN guidelines, an estimated 26 basic and six comprehensive essential obstetric care sites. Five hospitals were successfully upgraded to be comprehensive facilities and seven to basic level and the capacity of participating health facilities to respond to and resolve obstetrical complications opportunely increased appreciably. Project interventions improved the skills of different types of health care providers in the facilities,¹¹ strengthened infection control,¹⁶ blood banks¹⁷ and procurement systems, with resultant benefits not just for maternity care but wider acute health care provision. When the Project identified a lack of anaesthetic services, it addressed this by training nurses and auxiliary nurse-midwives as anaesthetic assistants. An evaluation demonstrated their competence, especially in providing spinal anaesthesia for caesarean section.⁸

However, implementation of the basic essential obstetric care satellites and comprehensive essential obstetric care hospitals had limitations. The referral hierarchy was sometimes bypassed by household members, who felt unable to predict the level of care needed once a complication arose and preferred to go directly to the highest level of facility. However, to increase the proportion of comprehensive to basic level facilities would have been expensive, with both the capital costs (NR 7.81 million (US$104,133^*) versus NR 6.02 million (US$80,267)) and the recurrent costs of a comprehensive care facility considerably higher.⁶ Improving the infrastructure and equipment of facilities to achieve comprehensive level care proved to be relatively easy. Ensuring sustained functioning with a doctor with adequate obstetric skills, together with availability of blood and anaesthesia, was a far greater challenge. Even during its lifetime, the Project could not secure these essential elements for all the facilities upgraded, although it made considerable advances.

The experience showed that efforts to increase availability of health facilities to provide skilled birthing care face complex challenges – on the supply side because they require attempts to make good a weak health system, on the demand side because of underlying limitations on care-seeking behaviour in a highly unequal society.

Availability of a skilled birth attendant near the home

The basic essential obstetric care sites that were successfully developed are a good example of how nurses, with additional training, can manage many non-surgical obstetric emergencies. Monitoring data from the three Phase 1 districts showed that 70% of 1,030 emergency procedures were managed by nurses. Retaining doctors is difficult at primary care level, and enhancing nursing skills has been an effective way of increasing provision of skilled birthing care closer to where women live. However, the national shortage of skilled professionals is a major constraint. The 2001 Demographic and Health Survey (DHS) found that only 3.1% of deliveries of the approximately 900,000 births per annum were attended by an auxiliary nurse-midwife or nurse.¹⁹ This had increased to 8.3% in the 2006 DHS.²⁰ In 2004, there were just under 800 certificate nurses working in the public sector (56% of the national total) and 1,254 auxiliary nurse-midwives (78% of all those trained), who receive 18 months’ training.²⁴ As there is a concentration of staff with higher skills in urban centres, vacant rural posts and absenteeism are common.¹¹ Widespread dual practice in public and private sectors also reduces the general availability of public sector care.²⁵

With too few nurses and auxiliary nurse-midwives, posts for these cadres are restricted to the higher level facilities. The challenge of how then to expand skilled attendance quickly, fuelled by the pressure to achieve the MDG goal of 60% skilled attendance by 2015, is one that Nepal shares with many other low-resource countries. During the life of the Nepal Safer Motherhood Project there was considerable debate in Nepal about upgrading and certifying of skills of intermediate staff as an interim strategy.²⁶

In Nepal the nearest point of access to health care for women in rural areas are the 3,132 sub-health posts, staffed by mother and child health workers. This cadre was once seen as Nepal’s response to increasing the availability of skilled attendants. However, recent government policy (2006) recognises only doctors, nurses or auxiliary nurse-midwives who have received specific training in internationally-defined skills as skilled birth attendants.²⁷ Eligible mother and child health workers (those who meet the educational requirements) are being re-trained as auxiliary nurse-midwives. Internationally, discussion continues on how to scale up availability of professional care rapidly,²⁸ and Nepal may still need to consider different options if it is to meet its target.

In Nepal, caste and ethnicity are central in determining access and use. One study found 76% of the 51 mother and child health workers interviewed were from high caste groups (Brahmin/Chhetri), which is likely to have limited uptake of their services by lower caste and Janajati groups.¹⁰ Thus, whilst the skill level of the attendant is undeniably important, equal focus needs to be given to all the other factors that determine provider effectiveness, ranging from selection and employment policies to social and cultural issues, and the power structures of society that pervade the health system, communities and families, and constrain women’s empowerment.

Formalised referral links between facilities

Koblinsky and Campbell’s review² suggests that substantial reduction in maternal mortality is possible with a range of different combinations of home and institutional birth patterns, and of different categories of health workers – if functioning referral links to essential obstetric care are a central feature of the maternity system.

Major enhancement of the very weak formal sector referral system was considered beyond the scope of a district-based project like the Nepal Safer Motherhood Project. Instead the project set out to increase information in the community about danger signs in pregnancy and delivery and the need for prompt action at household and community level. Ability to act quickly would be facilitated by emergency fund and transport schemes. Whilst there were some local successes,⁵ this approach could not fully compensate for the weakness of the public referral system, nor did it take sufficient account of the extensive use of the formal and informal private sector. A study in a later stage of the project⁷ revealed the frequency with which women visited private drug retailers (21% of the comprehensive care users sampled) and private clinics (24%) before reaching the comprehensive care facility where they finally delivered. Nepal has now embarked on a Health Sector Programme which should be able to tackle the challenge of improving the public sector referral system. Project evidence suggests that the importance of formal and informal private referral networks needs to be taken into account and be included in overall plans for strengthening referral systems.

Free or reduced costs for services and transport

During the Project period, user fees in public hospitals were equivalent to almost 100% of the cost of providing emergency obstetric care because of low health budget allocations. There is also strong pressure on hospitals and staff to supplement their incomes; unethical commercial practices between doctors and pharmacies exacerbate the high cost of drugs and supplies to families. A Project study of the costs of maternity care found that for many households, delivery in a facility was unaffordable. Transport in Project and non-Project districts alike, on average, represented almost 60% of the total cost associated with a normal delivery in hospital. This is unsurprising as the average time spent reaching a facility varied from a mean of 2.8 hours in terai (plains) districts to 8.3 hours in mountain districts.⁷ The average total cost of a normal delivery was NR 693 (US$9) at home and NR 5,457 (US$73) in hospital, and a caesarean section cost nearly double a normal delivery in hospital. Facility delivery cost approximately three months of a poor household’s income for a normal birth and six months for caesarean section. The study also found that hospital exemption policies were ineffective; the poorest women paid the same as the least poor. Families were willing to pay for emergency care, but they were unable to do so without borrowing money or selling assets.

One of the project’s key financing initiatives was to support and encourage communities to establish emergency funds from which families could borrow in the event of an obstetric complication. Many of these community-based funds were an extension of existing endeavours such as forest-user groups, saving and credit groups and mother’s groups. An evaluation indicated that communities valued these funds and that they increased confidence in being able to cope with emergencies.⁵⁷ However, the constituency for whom loan funds provide a viable way of bearing costs can be limited.²⁹ The loans were found to be least accessible to the poorest community members, who were the least creditworthy. Thus, while the funds may have facilitated emergency access for some, they did not reduce existing social inequities.

Debriefing on needs assessment in Surkhet district. Nepal, 1997

This experience suggested a national initiative was required to prevent cost from deterring care-seeking among the poor. In Nepal, although central policies dictated allocations for the indigent, there was no central reimbursement of such funding. Implementation is a challenge even within somewhat stronger health systems. In Indonesia, women who could qualify for social safety nets did not always access facilities when complications occurred, and in Bolivia poor people remained outside the health system even with free care provision.² As Parkhurst et al³⁰ point out in their recent four-country comparison, access barriers work in dynamic and mutually reinforcing ways. How formal and informal costs can be effectively and fairly reduced, and how this can then be translated into improved utilisation, needs far more attention in the future.

Subsequent to the Project, the Nepal Government instituted a policy to reduce financial barriers to institutional delivery through cash payments to subsidise transport costs and free institutional delivery in the poorest districts. Such reforms will need to be carefully targeted so that they protect the poor without extinguishing the willingness of the non-poor to pay for services or encouraging elite-capture. What is clear is that increased subsidies for the poor demands increased financing from central Government, as this cannot feasibly be supported by local hospital initiatives. However, the Health Budget is itself heavily dependent on external funds, so whilst the Ministry of Health is now strongly supportive of subsidies for essential obstetric care, resourcing for the long run is likely to pose a problem. The importance of ensuring adequate cash flow for the cost-sharing policy has been emphasised.³¹

Public accountability

In a health system with low public accountability, the Project focused on improving the interaction between staff and patients to increase the value accorded to patient satisfaction. The two main tools introduced were an “appreciative enquiry” approach to change management (known locally as “Foundations for Change”) and use of Quality of Care Guidelines focusing on both technical and patient-centred perspectives. At the same time, an adapted set of UN process indicators increased the capacity of facilities to monitor utilisation of services (a measure of demand for services) and at the community level a participatory ethnographic monitoring tool was used to record changes in community perceptions of the quality of services.³²

Recent discussions in the international literature have highlighted the importance of understanding what motivates or de-motivates health workers and managers,³³³⁴ and the difficulties of determining the best balance between inputs into infrastructural vs. human resources development. Project experience was that both are important and demonstrated the value and acceptability of using appreciative enquiry and team-building approaches within a traditionally hierarchical health care bureaucracy. Whilst this fell short of creating a system of public accountability, the positive shift in attitudes towards patient-centred care set the foundation for systems change. The Quality of Care¹¹ evaluation found that more attention was needed on making users aware of their health care rights, another vital area for future intervention towards better accountability.

The continuing challenge of how best to empower pregnant women and their families within health systems demands greater attention. For example, whilst utilisation of community emergency funds was relatively low,⁵ self-management of funds by communities was considered to be one of the more important project achievements by communities and project stakeholders.³⁵ This finding resonates with the results of a cluster randomised, controlled trial in Makwampur which indicated that improved birth outcomes could be achieved by providing support to problem identification by local women’s groups, alongside improvements at referral facilities.³⁶

Internationally, a rights-based approach to safe motherhood has gained greater attention.³⁷³⁸ In Nepal, the need for this is starkly apparent from both internal monitoring and external data,¹⁰ which showed, in the three districts studied, that the majority of essential obstetric care users were from higher caste communities, inverse to their proportion in the community. This suggests that despite efforts, the Project was unable to counter the strong social norms which give higher caste groups preferred access to society’s resources.³⁵

Whilst the Project demonstrated considerable success in changing staff mindset at the implementation level, for example enabling the shift from a prevailing negative, fatalistic attitude to a belief in the possibility of positive change, this urgently needs to be echoed in changed attitudes at the policy level. The Nepal Human Development Report 2004 emphasises that “Changing the long-standing institutional culture that governs Nepal’s decision-making processes simply will not take place without radical changes in the mindsets of those who work within them”.³⁹

Strong government policy guidance

Analysts of the dynamics of policy agenda-setting suggest that it is the characteristics of particular issues, combining with the characteristics of political institutions that can lead to the opening – or closing – of policy windows of opportunity.⁴⁰⁴¹ However, the processes by which strong national policies for safe motherhood are influenced and developed are only recently beginning to receive attention.⁴² What the Project’s own experience seems to confirm is the value of conscious policy entrepreneurship and of building close relationships between project managers and policy-making structures from the outset. The Project assisted in the development of government policy by being able to offer both local knowledge about implementation and international lessons for key individuals prepared to take the political agenda forward. Project research on user costs and field knowledge of the inadequate working of the exemption systems, for example, helped to catalyse government discussion of subsidising service delivery for all, announced in 2004. Project experience of contracting NGOs for training and service provision to the public sector also helped develop local policy thinking about working collaboratively with non-state health care providers. During the policy process that led to legalisation of abortion, the Project was able to contribute a key review of current knowledge on abortion policy in other countries, and again to encourage government to work with the private sector in this area.

The Project was thus able to play a constructive role influencing policy development –agreed between the Nepal Ministry of Health and DFID as one of the official Project “outputs”. However, taking policy forward lies with the government and the implementation of Comprehensive Abortion Care post-legalisation is a good example of the strong drive and initiative within the government’s own ranks.⁴³ Current reform of the health sector needs to proactively support ways to gather knowledge and experience from within the system to shape future policy.

Conclusion

Koblinsky and Campbell’s review² was based on case studies in settings with maternal mortality ratios of less than 200 per 100,000 live births. They recognised that where this level of maternal mortality has been reached there is probably momentum already in place towards further reduction. Countries such as Nepal are at a very different starting point. The initiation of sub-sector projects such as the Nepal Safer Motherhood Project required confronting the constraints of an extremely weak health system coupled with widespread poverty and social inequity. That experience raises questions about some of the assumptions commonly underlying the design of Safe Motherhood programmes, in particular, narrowly focused implementation using clinical models derived from the UN process indicators.

What does seem clear from the Project experience is that a supply-side analysis of intervention is by itself insufficient for informing effective maternal mortality reduction in contexts with very high maternal mortality. In Nepal, this situation is further complicated by social norms and structures that leave women undervalued and disempowered, and which marginalise people from lower castes and certain ethnic groups, a pattern reflected in the use of maternity services. Koblinsky and colleagues, in their recent paper on maternal health in poor countries,⁴⁴ stress the importance of co-strategies such as poverty reduction and women’s empowerment for improving maternal health.

The next period of DFID support to the Safe Motherhood Programme in Nepal has taken equity and access issues as a starting point for planning. The challenge will be to ensure that, in a form of programme support in which donor funds are pooled with government finances for maternal health, these sorts of additional “nurturing” activities directed at the social environment are seen as essential and integral to long-term programmatic success of the safe motherhood programme.

Acknowledgements

The NSMP was a collaborative intervention between the Nepal Ministry of Health and Population and the British Department for International Development, managed by Options Consultancy Services. The views expressed are those of the authors.

Notes

* US dollars are at 2004 exchange rates throughout the paper.