RESEARCH

Estimated global resources needed to attain universal coverage of maternal and newborn health services

Estimation des ressources nécessaires à l’échelle mondiale pour atteindre une couverture universelle par les services de santé maternelle et néonatale

Estimación de los recursos mundiales necesarios para lograr la cobertura universal de servicios de salud de la madre y del recién nacido

Benjamin Johns^{I, 1}; Kristjana Sigurbjörnsdóttir^II; Helga Fogstad^III; Jelka Zupan^IV; Matthews Mathai^IV; Tessa Tan-Torres Edejer^V

^IWorld Health Organization, 9F Bina Mulia – I Bldg, Jl Rasuna Said Kav. 10, Kuningan Jakarta 12950, Indonesia
^IIHIV/AIDS Practice Team, UNDP Regional Centre, Colombo, Sri Lanka
^IIIGlobal Health and AIDS Department, Norwegian Agency for Development Cooperation, Oslo, Norway
^IVDepartment of Making Pregnancy Safer, WHO, Geneva, Switzerland
^VDepartment of Health Systems Financing, WHO, Geneva, Switzerland

ABSTRACT

OBJECTIVE: To estimate the amount of additional resources needed to scale up maternal and newborn health services within the context of the Millennium Development Goals, and to inform countries, donors and multilateral agencies about the resources needed to achieve these goals.
METHODS: A costing model based on WHO’s clinical guidelines was used to estimate the incremental resource needs for maternal and newborn health care in 75 countries. The model estimated the costs for care during pregnancy, childbirth, the neonatal period and the postpartum period, as well as the costs for postpartum family planning and counselling, abortion and post-abortion care; programme-level costs were also estimated. An ingredients-based approach, with financial costs for the years 2006 to 2015 as the output, allowed estimates to be made of country-specific and year-specific populations, unit costs and scale-up rates. Two scenarios using different scale-up rates were used (moderate and rapid).
FINDINGS: The results show that a minimum yearly average increase in resources of US$ 3.9 billion is needed, although annual costs increase over the time period of the model. When more rapid rates of scale-up are assumed, this minimum figure may be as high as US$ 5.6 billion per year. The 10-year estimated incremental costs range from US$ 39.3 billion for a moderate scale-up scenario to US$ 55.7 billion for the rapid scale-up scenario.
CONCLUSION: These projections of future financial costs may be used as a starting point for mobilizing global resources. Countries will have to further refine these estimates, but these figures may serve as goals towards which donors can direct their plans. Further research is needed to measure the costs of health system reforms, such as recruiting, training and retaining a sufficient number of personnel.

RÉSUMÉ

OBJECTIF: Estimer les moyens supplémentaires nécessaires pour la mise à l’échelle des services de santé maternelle et néonatale dans le cadre des objectifs du Millénaire pour le développement et informer les pays, les donateurs et les agences multilatérales des ressources requises pour atteindre ces objectifs.
MÉTHODES: Un modèle d’évaluation des coûts s’appuyant sur les Directives cliniques de l’OMS a permis d’estimer les besoins financiers supplémentaires pour délivrer les soins de santé maternelle et néonatale prévus dans 75 pays. Ce modèle a évalué les coûts des soins dispensés pendant la grossesse, l’accouchement, la période néonatale et la période postpartum, ainsi que ceux des services de planification familiale et de conseil et ceux liés à un éventuel avortement et à la période post-avortement. Les coûts programmatiques ont également été estimés. En appliquant une approche par composants et en considérant comme des extrants les coûts financiers pour les années 2006 à 2015, on a obtenu une estimation des couts unitaires et des coefficients de mise à l’échelle par pays et par année/habitant. Deux scénarios de mise à l’échelle (rythme modéré ou rapide), utilisant des coefficients de mise à l’échelle différents, ont été utilisés.
RÉSULTATS: Les résultats indiquent qu’en moyenne, une augmentation des ressources de US $ 3,9 milliard par an au minimum est nécessaire, sans compter l’augmentation annuelle des coûts sur la période prise en compte par le modèle. Si l’on suppose une mise à l’échelle plus rapide, cette augmentation minimale des ressources peut atteindre US $ 5,6 milliard par an. Sur 10 ans, on estime que les coûts supplémentaires représenteront de US $ 39,3 milliard, pour un scénario à vitesse moyenne, à US $ 55,7 milliard pour un scénario prévoyant une mise à l’échelle rapide.
CONCLUSION: Ces projections de coûts financiers sont utilisables comme point de départ pour mobiliser les ressources au niveau mondial. Les pays auront ensuite à affiner ces estimations, mais elles pourront servir d’objectifs aux donateurs pour la définition des grandes lignes de leurs plans. Des études plus poussées sont nécessaires pour évaluer les coûts des réformes à apporter aux systèmes de santé, telles que : recrutement, formation et recyclage de personnel en effectif suffisant.

RESUMEN

OBJETIVO: Estimar la cantidad de recursos adicionales necesarios para expandir los servicios de salud de la madre y del recién nacido en el contexto de los Objetivos de Desarrollo del Milenio, e informar a los países, donantes y organismos multilaterales sobre los recursos necesarios para alcanzar esas metas.
MÉTODOS: Se empleó un modelo de cálculo de costos basado en las directrices clínicas de la OMS a fin de estimar las necesidades progresivas de recursos para asegurar la atención de salud de la madre y del recién nacido en 75 países. El modelo calculó los costos de la atención correspondiente al embarazo, el parto, el periodo neonatal y el periodo posparto, así como los costos de la planificación familiar y el consejo tras el parto, y la atención de aborto y posaborto; también se estimaron los costos de programas. El uso de un método basado en componentes, con costos financieros para los años 2006-2015 como resultado, permitió hacer estimaciones para poblaciones, costos unitarios y ritmos de expansión en países específicos y años específicos. Se usaron dos escenarios con diferentes ritmos de expansión (moderado y rápido).
RESULTADOS: Los resultados muestran que se requiere un incremento medio anual mínimo de los recursos del orden de US$ 3900 millones, aunque los costos anuales aumentan a lo largo del periodo considerado en el modelo. Si se asumen ritmos más rápidos de expansión, esa cifra mínima puede llegar a ser de US$ 5600 millones anuales. Los costos adicionales estimados a 10 años varían entre US$ 39 300 millones para un escenario de expansión moderada y US$ 55 700 millones para el escenario de expansión rápida.
CONCLUSIÓN: Estas proyecciones de los futuros costos financieros pueden servir de punto de partida para movilizar recursos mundiales. Los países tendrán que seguir refinando estas estimaciones, pero las cifras pueden ser utilizadas por los donantes como metas para orientar sus planes. Es necesario realizar nuevas investigaciones para medir los costos de las reformas de los sistemas de salud, como las relacionadas con la contratación, formación y conservación de personal en número suficiente.

Introduction

An estimated 15% of pregnant women in developing countries experience pregnancy-related complications, 7% require care at centres with surgical capacity (referral care) and 2% to 3% require surgical care.¹ Nearly 530 000 women die from pregnancy complications annually.² Furthermore, each year an estimated 4 million babies die within the first 4 weeks of life, around three-quarters during the first week of life.¹ Deaths among neonates account for almost 40% of deaths occurring among children aged < 5 years and for more than half of all deaths among infants.^{3, 4} An additional 3.3 million babies are stillborn, a quarter of them dying during birth.

In developing countries, interventions that are known to be effective in lowering maternal and perinatal mortality and morbidity are not universally provided. These interventions can be delivered by skilled health personnel providing care during pregnancy, childbirth and the postpartum and postnatal periods; health-care personnel may also provide early detection and management of complications that occur during pregnancy, birth and the postpartum period. However, in the countries included in this analysis, only about 43% of births take place in health facilities where there are skilled health personnel.^{5, 6} The skill levels of these personnel vary, and only a fraction of mothers and babies have access to a full range of maternal and neonatal health interventions. Thus, scale up of these interventions must occur on two fronts: the quality and range of health services for mothers and neonates must be improved, and access to skilled birth attendants must be increased.

Scaling-up the coverage of maternal and newborn health care provided by skilled personnel is expected to have a considerable impact on achieving the Millennium Development Goals, specifically goal 5 (which aims to improve the health of mothers), large parts of goal 4 (which focuses on reducing child mortality) and parts of goal 6 (which seeks to combat AIDS, malaria and other diseases). For these goals to be achieved, countries, donors and multilateral agencies must mobilize resources. To determine the resources needed, it is necessary to estimate the cost of scaling-up to universal coverage.

This paper presents the methods used to construct a model and the results of two scenarios used in the model to estimate the financial costs of expanding the coverage of skilled maternal and newborn health care at facilities. The two scenarios reflect uncertainty around the speed and scope of future scale-up rates. Both scenarios assume that resources supporting current maternal and newborn health-care programmes will remain available; thus, they include only costs additional to current resource use.

The cost projections are based on country-specific and year-specific estimates for 75 low- and middle-income countries. The costing includes clinical interventions critical to improving health outcomes during pregnancy for mothers and neonates, and it is not limited to interventions that save lives. It includes the provision of information, advice and counselling to help women and their families with home care, birth planning and emergency preparedness as well as community support for women with special needs. It assumes that existing primary- and secondary-level health-care services deliver maternal and newborn health services. The cost estimates include the investments needed to strengthen the health system infrastructure and human resources as well as the costs of the management and support needed to ensure that high-quality care is provided, that there is a demand for these health services and that they are accessible.

Methods

To estimate financial needs until 2015 of expanding the coverage of maternal and newborn care provided by skilled personnel, we selected countries for the analysis; derived country-specific estimates and projections of pregnancies, births and abortions; selected the activities to be scaled up; estimated scale-up scenarios; and calculated country-specific unit costs (for additional information, see http://www.who.int/ whr/2005/td_two_en.pdf).

The cost of tetanus toxoid immunization was estimated using a costing model developed by WHO’s Department of Immunization, Vaccines and Biologicals. The costs of intermittent preventive therapy for malaria and the distribution of insecticide-treated bednets were estimated using the model developed by Roll Back Malaria and WHO. The methods and data sources for these models are similar to those of the maternal and newborn health model that we developed. Prevention of mother-to-child transmission of HIV was not included in this model; it was included in a parallel costing exercise.⁷

Countries

Countries were selected in parallel with a costing exercise for child health interventions⁷ and included those ranked highest in terms of both rates and gross numbers of maternal deaths, neonatal deaths and children’s deaths. Additional countries were added to ensure that all WHO regions were represented; this resulted in a list of 75 countries (this list is available on request from the corresponding author). Together, these countries account for more than 75% of the world’s population and almost 86% of births (more than an estimated 122 million births in 2006, rising to more than 137 million in 2015). These countries also represent approximately 97% of maternal deaths, 93% of fetal and neonatal deaths, and 94% of deaths occurring among children aged < 5 years.

Maternal and newborn health services

Clinical interventions to be included were selected based on proven effectiveness in improving the health outcomes of mothers and neonates, and whether they could feasibly be implemented in resource-poor settings. For example, "screening all pregnant women for blood group isoimmunization" and "postpartum administration of anti-D immunoglobulin" are included only for countries classified as having strong health systems using the classification described below. Interventions that require advanced technology, such as assisted ventilation for neonates, and services that require tertiary-level health care are not included. The interventions, and their required technologies, were obtained from WHO’s clinical guidelines.^8–10

Table 1 (available at http://www.who.int/bulletin) presents the interventions used in the model as well as the necessary inputs and the number of times the intervention is used during pregnancy, birth and the postnatal period.

In this model, health facilities provide all services. The primary level of the health-care system administers preventive and diagnostic assessment and treatment from the start of pregnancy until the end of the postpartum period (6 weeks) for the mother and the neonatal period (first 28 days of life) for the infant. This includes four antenatal visits, care during childbirth and the immediate postpartum period for mother and newborn, and one postpartum visit. Pre-referral care is also included when needed. In areas where certain diseases, such as malaria, are endemic, preventive care is included. In this model, the secondary level of health care (to which patients are referred from the primary level) includes the provision of emergency care. Referral includes services for managing complications in a hospital with a laboratory, surgical facilities and blood transfusion facilities. It also includes the cost of transport and a maternal stay in hospital in order to breastfeed a preterm infant.

In addition to an early postpartum visit, the model also includes a late postpartum visit for counselling on family planning. These services include contraception; the usage rates and types of contraception are based on current patterns in each country. The treatment of post-abortion complications is included for all countries. Abortion care is included in countries where it is currently legal.

In addition to implementing the clinical interventions described above, countries’ abilities at the programme levels (national, provincial and district administration levels, directly above patient care) may need upgrading to support scale-up. These costs include providing training for additional health-care providers, such as midwives and doctors, to ensure coverage of services. The areas of activity included in programme-level support are shown in Table 2.

Deriving country-specific estimates and projections

Country-specific mid-range (or medium variant) projections from the United Nations Development Programme were used to calculate total population (interpolated to yearly estimates using MORTPACK, the United Nations demographic software package), crude birth rate¹¹ and maternal mortality.² The number of pregnancies not resulting in birth was taken from Demographic and Health Surveys (DHS) for the latest year available until 2004;⁶ regional averages from the Global Burden of Disease study were used for countries for which data were lacking.¹² The number of pregnancies was estimated using the projected number of live births for each country adjusted for maternal mortality and pregnancies not resulting in birth.

Data on the incidence of induced abortion and abortion care services were incomplete or unavailable for all countries. Thus, in countries where abortion is legal, the number of abortions was derived from the regional estimates from the Global Burden of Disease study of the ratio of abortions to live births; in countries where abortion is not legal, regional estimates of the ratio of unsafe abortions to live births was used.¹² We assumed that 3% of legal abortions had complications requiring medical care;¹³ we assumed that 33–50% of unsafe abortions had complications requiring medical treatment and that 23–35% of these required hospitalization, depending on the Global Burden of Disease region.¹⁴

Estimating country-specific scale-up

Using estimates of current coverage and international goals, two scenarios were established to reflect movement towards providing universal access to maternal and newborn health care. Using experiences from developing countries that have 95% coverage, WHO experts’ opinions were used to draw a moderate scale-up scenario: this estimated that coverage across countries will average 73% in 2015. An alternative scenario, the rapid scale-up scenario, was developed using the assumption that all countries would achieve 95% coverage by 2015.

The first step in developing the scale-up scenarios was to establish a health-system constraint index to reflect the strength of the countries’ health systems in relation to service requirements for maternal and newborn health care. The indicator "percentage of births attended by skilled health personnel in a health facility" was determined to be the most suitable baseline indicator for the ability to scale up, and countries were divided into four categories based on this indicator. (These categories were denoted HSC-1 to HSC-4, with HSC-1 having the greatest constraints on scaling-up.) Using separate growth rates for the four health-system constraint categories, scale up patterns and targets for 2015 were determined. The scenarios are not meant to be prescriptive; they are for estimation purposes only.

Calculating country-specific costs for the activities

The population in need of a specified intervention was multiplied by the unit price of that activity. The median price reported by the organization Management Sciences for Health was used for medicines,¹⁵ and publicly available medical supply catalogues were used for the prices of supplies and equipment. The prices for these goods were adjusted to account for insurance during shipment and freight. The prices for visits to health centres and hospital outpatient departments, inpatient stays,^16,17 salaries, per diems, media costs and construction costs^16,18 were derived from WHO’s Choosing Interventions that are Cost Effective (WHO-CHOICE) studies. The prices for laboratory procedures, referrals to hospital and medical procedures were modelled using country-specific prices and ingredients lists for the disposable inputs necessary to complete procedures.^16,18–20 These unit prices included compensation for health workers at current pay rates. All unit prices are presented with a 3% annual inflation rate. The second column in Table 1 (available at http://www.who.int/bulletin) and the third column in Table 2 list the type of price category used in each activity.

Attempts were made to derive country-specific rates for diseases and conditions. However, data are scant, so default assumptions were based on regional rates from the Global Burden of Disease study and WHO experts’ experiences and opinions. The final column in Table 1 (available at http://www.who.int/bulletin) lists the number of times each intervention is carried out for an individual woman going through pregnancy, childbirth and the postpartum period.

In addition to current rates, the interaction of scaling-up maternal and newborn health-care activities on the incidence of other conditions is considered (Table 3).

The best proxy for the coverage of skilled care during birth and the postpartum, newborn and postnatal periods was the percentage of births delivered at a facility (this is to reflect the need for improvements in quality; for additional information, see http://www.who.int/ whr/2005/td_two_en.pdf). Current coverage for tetanus vaccination was derived from databases at WHO’s Department of Immunization, Vaccine and Biologicals, while data from DHS surveys or regional estimates are used for bednets and intermittent preventive therapy. Information on access to abortion in countries where it is legal was derived from country-specific data and country experts; information on access to hospital care after unsafe abortion was based on experts’ opinions.

The quantities needed for the programme section of the costs were derived from interviews with experts at WHO. If an activity was judged to cover a wider area than maternal and newborn health care, only a percentage of the costs were allocated to maternal and newborn health care. Additionally, to control for varying needs, we adjusted the level of investments and activities based on the health system constraint index. The last column in Table 2 shows the general assumptions used.

Findings

The estimated costs for 2006–2015 are summarized in Table 4. Over the 10-year period, the estimated cost ranges from US$ 39.3 billion for the moderate scale-up scenario to US$ 55.7 billion for the rapid scale-up scenario. The average cost per capita per year is US$ 0.73 for the moderate scale-up scenario and US$ 1.03 for rapid scale-up, with an average additional cost of US$ 1.17–1.82 per capita in 2015 when full coverage is achieved. The WHO region requiring the most additional resources is South-East Asia, reflecting the large population in need of services there; the region with the highest cost per capita is the Eastern Mediterranean, reflecting relatively high labour costs. The countries with the lowest health system constraint ranking, mostly in Africa, are projected to need US$ 0.35–0.38 of additional resources per inhabitant in 2006; with increased coverage, this amount is estimated to rise to US$ 2.17–4.19 per inhabitant in 2015. Countries in category HSC-1 are projected to represent 11% of the total population for the year 2015, but they represent 18–21% of costs. Countries in the HSC-2 category represent 14% of the population and 17–21% of costs; countries in HSC-3 have 9% of the population and 9–10% of costs; and countries in HSC-4 account for 65% of the population but only 49–56% of costs.

A comparison with the current general government health expenditure per country reveals that the investments required are equivalent to a weighted average increase in this expenditure of 3% in 2006, rising to 20–34% by 2015 (see WHO’s data at http://www.who.int/nha).

Fig. 1 shows the breakdown of total costs during the 10 years for both scale-up scenarios. Overall, the primary-care level comprises 29% of costs for moderate scale-up and 30% for rapid scale-up, with referral care accounting for 44% of costs in the moderate scale-up scenario and 47% in the rapid scale-up scenario. The remaining 27% in the moderate scale-up scenario and 23% in the rapid scale-up cover programme development and investments in health infrastructure. The costs of direct care – which include essential medicines, procedures (including some staff time) and commodities – account for 38% of total costs for moderate scale-up and 41% for rapid scale-up; 35% of the costs of moderate scale-up and 36% of rapid scale-up consist of non-traded goods such as salaries and utilities. The yearly health infrastructure costs double between 2006 and 2015 in the moderate scale-up scenario but the share of the total drops from 30% to 7%.

Fig. 2 (available at http://www.who.int/bulletin) shows the breakdown of the 10-year costs by scenario and service type. Care during childbirth is the largest cost category. Antenatal care and programme costs each represent about a quarter of total costs; the other broad service types represent 5% or less.

Discussion

These results are intended to be a rough estimate to aid countries, donors and multilateral agencies in determining the amount of resources needed to scale up maternal and newborn health services and to start taking steps to increase resource availability. Although the estimates are built country by country, they are intended to provide a global price tag. The results show that the financial effort needed for such a scale-up in 75 countries represents a modest investment of around US$ 0.73–1.03 per person per year. Countries such as Malaysia, Sri Lanka and Thailand have demonstrated that reducing maternal and neonatal mortality is achievable at this cost, given appropriate management, budget priorities and training.¹

Within the results presented here, some of the costs may be counted in other estimates of need for scaling-up priority health interventions. For example, our modelling found that in 2015, 2–3% of costs may be attributable to the treatment and prevention of malaria and 1.5–2.5% to the provision of vaccines. (The methods used to cost these two interventions are slightly different from those used for the core maternal and newborn health model; vaccine estimates, for example, include the cost of campaigns.)

For some programme costs, only a percentage of the total need was allocated to maternal and newborn health services; this was based on the assumption that the scale-up of these services would be possible only in the context of an overall strengthening of the health-care system. However, most of the costs are specific to pregnancy and childbirth, so resources specifically allocated for to maternal and newborn health are needed. It will take considerable effort for countries with weak health-care systems to upgrade their capacities to reach the desired coverage levels, and fuller assessments of necessary resources are needed.

Some costs not normally borne by health systems, such as the costs of transportation needed to access care, are not included in this analysis. Certain services included in the WHO recommended list of maternal and newborn health interventions are estimated under different disease categories and are not included in this exercise. The most significant example is HIV testing and preventing mother-to-child transmission of HIV. Additionally, the model does not incorporate the impact of developments that do not fall directly under the purview of maternal and newborn health services – such as improvements in the education of girls and women – on the incidence of diseases or conditions.

Increasing scale-up targets from an average of 73% coverage to 95% coverage (a 27% increase) results in around a 42% increase in total costs. There is considerable uncertainty about these results, including the uncertainty inherent in demographic projections and the impact of scaled-up family planning programmes on reducing numbers of pregnancies. It was necessary to use assumption-based estimates owing to variability in treatment protocols and the current quality of services, unreliable data on service utilization and a lack of accurate measures of the incidence and prevalence of maternal and newborn health conditions. This level of uncertainty precluded a formal analysis of uncertainty. Additionally, establishing goals is an uncertain process not only in terms of identifying an achievable target, but also in terms of the pace of scale-up. Thus, the results of this analysis should be regarded as projections of the order of magnitude, and they can be validated only by costing exercises conducted in specific individual countries.

Antenatal care and childbirth care constitute the majority of costs; in these categories, the cost of human resources represents about half of the total costs. The cost of human resources was derived using current levels of remuneration; in many countries these are likely to be insufficient to recruit, retain and deploy enough health workers. The Eastern Mediterranean Region, for example, has relatively high labour costs. This is reflected in the outcomes, despite the common perception that countries in this region have adequate human resources when compared with other regions. Meanwhile, the African Region needs relatively high levels of investment in health infrastructure; this, together with low levels of current coverage, helps account for that region’s higher costs.

These results represent only the share of labour attributable to providing maternal and newborn health services; in many countries any attempt to reform human resources may involve reforming the entire health labour sector or the entire civil service. The entire cost of reforms targeting human resources is not captured here, but the success of reforms in this area will predict the success of scaling-up maternal and newborn care in many countries. Thus we suggest that our projections are a low-end estimate. If the costs of recruiting and retaining staff are, on average, double what is currently paid for wages, then we calculate that costs will increase by roughly a third. However, not all countries face a severe crisis in human resources, and the salary increases needed are uncertain.

Finally, attainment of the Millennium Development Goals demands a reliable and comparable method for estimating resource needs. More robust country-specific information and costing of scaling-up these health services is needed to better inform national decisions about the resources needed. Achieving the Millennium Development Goals may also require reform of health-care systems to allow greater access for the poor; for example, health financing systems may need to be reformed in order to allow the poor to access care. Because the costs of these reforms are unknown, they are uncounted in this analysis. Such information should be systematically collected in conjunction with efforts to scale-up maternal and newborn health services.

Acknowledgements

The authors would like to acknowledge Luc de Bernis, Rita Kabra, Della Sherratt, Annie Portela, Juliana Yartey, Ornella Lincetto, Zoë Matthews and other members of the Making Pregnancy Safer team at WHO for their guidance and technical input in relation to the maternal and newborn health interventions. Special thanks go to Dr Monir Islam, director of the Making Pregnancy Safer Department. Jane Cottingham, Elisabeth Aahman and Iqbal Shah provided technical expertise and data for this analysis; the Child and Adolescent Health team jointly worked in allocating shared costs between the two programmes, notably, Karin Stenberg and Robert Scherpbier. Data, methods and calculations from Lara Wolfson and the Roll Back Malaria costing team were used in the calculations of tetanus vaccinations and the costs of intermittent preventive care and insecticide-treated bednets. The authors acknowledge guidance from David Evans, Taghreed Adam and other members of WHO’s Health System Financing team. Finally, the background for this study was motivated by the World health report 2005: making every mother and child count, and by Wim Van Lerberghe, its editor-in-chief.

Competing interests: None declared.

References

1. The world health report 2005: make every mother and child count. Geneva: WHO; 2005.        

2. Maternal mortality in 2000: estimates developed by WHO, UNICEF and UNFPA. Geneva: WHO; 2004.        

3. Perinatal mortality: a listing of available information. Geneva: WHO; 1996 (WHO/FRH/MSM/96.7).        

4. State of the world’s newborns: a report from saving newborn lives. Washington: Save the Children Fund; 2004. pp. 1-28.        

5. Global monitoring and evaluation. Geneva: WHO; 2004.        

6. Demographic and Health Surveys. STATcompiler, 2004. Available from http://www.measuredhs.com/ statcompiler/start.cfm?action=new_table&userid=155356&usertabid=170332&CFID=139536&CFTOKEN =83650943        

7. Stenberg K, Johns B, Scherpbier R, Tan-Torres Edejer T. A financial road map to scaling up essential child health interventions in 75 countries. Bull World Health Organ 85;4:305-314.        

8. Pregnancy, childbirth, postpartum and newborn care: a guide for essential practice. Geneva: WHO; 2003. Available from http://www.who.int/ reproductive-health/publications/pcpnc/index.html        

9. Managing complications in pregnancy and childbirth: a guide for midwives and doctors. Geneva: WHO; 2000. Available from http://www.who.int/ reproductive-health/impac/index.html        

10. Managing newborn problems: a guide for doctors, nurses, and midwives. Geneva: WHO; 2003. Available from http://www.who.int/ reproductive-health/publications/mnp/index.html        

11. United Nations Population Division. World population prospects: the 2002 revision, 2004. Available from http://www.un.org/ esa/population/publications/wpp2002/wpp2002annextables.PDF        

12. Murray CJL, Lopez AD. The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020. Cambridge: Harvard University Press; 1996.         

13. Thonneau P, Fougeyrollas B, Ducot B, Boubilley D, Dif J, Lalande M et al. Complications of abortion performed under local anesthesia. Eur J Obstet Gynecol Reprod Biol 1998;81:59-63. doi: 10.1016/S0301-2115(98)00147-X        

14. Unsafe abortion: global and regional estimates of incidence of mortality due to unsafe abortion with a listing of available country data, 3rd edition. Geneva: WHO; 2004. Available from http://www.who.int/reproductive-health/publications/MSM_97_16/ MSM_97_16_table_of_contents_en.html        

15. Management Sciences for Health. International drug price indicator guide, 2004. Available from http://erc.msh.org/dmpguide/        

16. Mulligan J, Fox-Rushby JA, Adam T, Johns B, Mills A. Unit costs of health care inputs in low and middle income regions. Washington, DC: World Bank; 2005. Disease Control Priorities Project Working Paper No. 9. Available from http://www.dcp2.org/file/24/wp9.pdf        

17. Adam T, Evans DB, Murray CJL. Econometric estimation of country-specific hospital costs. Cost Eff Resour Alloc [online journal] 2003;1:3. Available from: http://www.resource-allocation.com/content/1/1/3) doi: 10.1186/1478-7547-1-3        

18. Johns B, Adam T, Evans DB. Enhancing the comparability of costing methods: cross-country variability in the prices of non-traded inputs to health programmes. Cost Eff Resour Alloc [online journal] 2006;4:8. Available from: http://www.resource-allocation.com/content/4/1/8 doi: 10.1186/1478-7547-4-8        

19. Essential technology package software. Tygerberg: WHO Collaborating Centre for Essential Technologies in Health; 2004.        

20. UNFPA, World Health Organization. Essential drugs and other commodities for reproductive health services. Geneva: UNDP, WHO; 2003.        

21. Antenatal syphilis screening in the UK: a systematic review and national options appraisal with recommendations. London: Public Health Laboratory Service Syphilis Working Group; 1998.        

22. Maral I, Cirak M, Aksakal FN, Baykan Z, Kayikcioglu F, Bumin MA. Tetanus immunization in pregnant women: serum levels of antitetanus antibodies at time of delivery. Eur J Epidemiol 2001;17:661-5. doi: 10.1023/A:1015507402480        

23. Kenyon S, Boulvain M, Neilson J. Antibiotics for preterm rupture of membranes. Cochrane Database of Systematic Reviews. 2003;2:CD001058.         

24. McDonald H, Brocklehurst P, Parsons J. Antibiotics for treating bacterial vaginosis in pregnancy. Cochrane Database of Systematic Reviews 2005;1:CD000262. doi: 10.1002/14651858.CD000262.pub2        

25. Prendiville WJ, Elbourne D, McDonald S. Active versus expectant management in the third stage of labour. Cochrane Database of Systematic Reviews 2000; 3:CD000007.        

26. Hofmeyr GK, Kulier R. Interventions to help external cephalic version for breech presentation at term. Cochrane Database of Systematic Reviews 2004:1:CD000184.        

(Submitted: 28 March 2006 – Final revised version received: 5 October 2006 – Accepted: 11 October 2006)

1 Correspondence to Benjamin Johns (e-mail: johnsb@who.or.id).