Doi:10.1016/s0140-6736(06)69386-0

Meeting MDG-5: an impossible dream?
Reduction of the maternal mortality ratio by three- A crucial development during the past decade is Published Online
quarters by 2015 is the target for one of the eight the growth in the number of programme models September 28, 2006 Millennium Development Goals (MDGs) set by available. In many developing countries, governments, 6736(06)69386-0
189 countries in 2000. That this goal (MDG-5) is the one non-governmental organisations, and international See Comment pages 1129,
towards which the least progress has been made,1 despite agencies are working together to upgrade existing the launch nearly 20 years ago of the Safe Motherhood governmental services. Major donors, such as the Initiative,2 is widely acknowledged. Nonetheless, we UK Department for International Development and believe that substantial progress can be achieved. Indeed, the US Agency for International Development, have a 2003 World Bank report3 on the success of several supported intensive projects in several countries, developing countries in (including China, Sri Lanka, and including Bolivia, the Dominican Republic, Egypt, Malaysia) reducing maternal mortality rates concluded Guatemala, Indonesia, Malawi, and Nepal. The that “maternal mortality can be halved in developing largest coordinated network of projects, which use countries every 7–10 years…regardless of income level and common strategies, indicators, and tools, are those growth rate”. To make real progress by 2015, substantial, that have received fi nancial and technical support fl exible, medium-term funding for fi eld programmes and from Columbia University’s Mailman School of Public related research is needed, with a clear focus on important Health, fi rst under the Prevention of Maternal Mortality programme elements, implemented with commitment programme,14 funded by the Carnegie Corporation, to the crucial goal of strengthening national health and then under the Averting Maternal Death and systems.
Disability programme, funded by the Bill & Melinda Our optimism is based on what has been accomplished Gates Foundation. Between 1999 and 2005, more in the past few decades. The human rights implications than 80 projects were supported by the Averting of maternal deaths are now widely appreciated, forcing Maternal Death and Disability programme in more than attention to be drawn not only to broader social injustice, 50 countries. These projects were in partnership with but also to faltering, sometimes even abusive, health local governments and with the UN Children’s Fund, the systems. A broad consensus exists about appropriate UN Population Fund, CARE, Save the Children, Regional strategies to reduce rates of maternal mortality.4,5 A Prevention of Maternal Mortality Network, Reproductive growing body of data and practical tested tools are Health Response in Confl ict Consortium, and several available to guide programmes.6–10 And, most of all, action human-rights non-governmental organisations. The is fi nally being taken, as replication of successful projects main fi eld projects in 17 countries covered a total gathers momentum across south Asia, sub-Saharan Africa, and parts of Latin America.
After considerable debate, a much more focused approach to the reduction of maternal deaths has taken shape. Diff erent groups focus on diff erent aspects—emergency obstetric care, skilled care by skilled attendants, unmet obstetric need—but all of them have at their core a recognition that without the ability to treat women with obstetric complications, maternal mortality cannot be substantially reduced.11–13 Moreover, consensus is growing that these life-saving services should be integrated into the local health system to deliver continuous care from community education and services (which are especially important for neonatal health care and early intervention with misoprostol for postpartum haemorrhage) to the fi rst referral facility for emergency obstetric care.
www.thelancet.com Vol 368 September 30, 2006
population of nearly 180 million people, and averted an they provide a unique opportunity to tackle fundamental estimated 9500 maternal deaths.
health-systems problems in a focused, measurable, and These successful projects have built on the capacity pragmatic way—with implications well beyond maternal for problem-solving and innovation that is an under- health. Indeed, met need for emergency obstetric care appreciated resource even in countries with the highest off ers a useful tracer for overall strengthening of health mortality rates. One area in which country-led innovation systems.18will be crucial is in addressing the crisis in human 21 years ago in The Lancet, we posed the question, resources. The initiatives in Mozambique, Tanzania, “Where is the M in MCH?”19 Today, we have much of the and Malawi (where non-physicians are trained and answer. What is needed now is the determination, focus, empowered to provide obstetric surgery and other life- and resources to fi nally reduce rates of maternal mortality saving procedures), and in India (where general practice in developing countries. physicians are being trained with the support of the Federation of Obstetrics and Gynaecological Societies of Allan Rosenfi eld, Deborah Maine, *Lynn FreedmanIndia in emergency obstetric care to do caesarean sections Mailman School of Public Health, Columbia University, New York, and to give anaesthesia), are among the examples that NY 10032, USA (AR, LF); and School of Public Health, Boston deserve serious study for eff ectiveness, equity of service [email protected] We declare that we have no confl ict of interest.
Perhaps most encouraging is the fact that many 1 Simwaka BN, Theobald S, Amekudzi YP, Tolhurst R. Meeting Millennium fi eld projects and innovations are being replicated by Development Goals 3 and 5. BMJ 2005; 331: 708–09.
AbouZahr C. Safe Motherhood: a brief history of the global movement 1949– governments and their partner agencies. For example, 2002. Br Med Bull 2003; 67: 15–16.
in Rajasthan, India, the UN Population Fund/Averting 3 Padmanathan I, Liljestrand J, Martins J, et al. Investing in maternal health in Malaysia and Sri Lanka. Washington, DC: The World Bank, 2003.
Maternal Death and Disability project was implemented 4 WHO. The World health report 2005: make every mother and child count. in more than 80 facilities in seven districts with a budget Geneva: World Health Organization, 2005.
5 Wagstaff A, Claeson M. The Millennium Development Goals for health: rising of US$1·4 million. Now the approach has been adopted to the challenges. Washington, DC: The World Bank, 2004.
by the Rajasthan state government for use as a model 6 UN Children’s Fund, WHO, UN Population Fund. Guidelines for monitoring the availability and use of obstetric services. New York: UN Children’s Fund, 1997.
in a project to strengthen state health systems, fi nanced 7 Averting Maternal Death and Disability. Using the UN process indicators of through a World Bank loan of more than $100 million.15 emergency obstetric services: questions and answers. Averting Maternal Death and Disability Workbook. 2003: http://cpmcnet.columbia.edu/dept/ Furthermore, elements of the Rajasthan project are being sph/popfam/amdd/resources.html (accessed Sept 6, 2006).
incorporated into India’s national Reproductive and Child 8 JHPIEGO, Averting Maternal Death and Disability. Emergency obstetric care for doctors and midwives, and anesthesia for emergency obstetric care for Health Program16 (with a budget of billions of dollars). doctors and midwives. 2003: http://www.reproline.jhu.edu/english/2mnh/2mnh.htm (accessed Sept 6, 2006).
Replication and scale-up are taking place in many other 9 WHO, UN Population Fund, UN Children’s Fund, World Bank. Managing countries, sponsored by many other agencies. Even so, complications in pregnancy and childbirth: a guide for midwives and doctors. Geneva: World Health Organization, 2003.
funding for maternal health programmes is still scarce 10 EngenderHealth, Averting Maternal Death and Disability. Quality and inconsistent, and should be rapidly and substantially improvement for emergency obstetric care: leadership manual and toolbook. 2003: http://www.engenderhealth.org/res/off c/mac/emoc/index.html increased if this important human right of women is to be met. The funding should also be fl exible enough to 11 de Bernis L, Sherratt D, AbouZahr C, Van Lerberghe W. Skilled attendants for pregnancy, childbirth and postnatal care. Br Med Bull 2003; 67: 39–57.
encourage local tailoring of programmes, while conserving 12 Paxton A, Maine D, Freedman L, Fry D, Lobis S. The evidence for emergency obstetric care. Int J Gynaecol Obstet 2005; 88: 181–93.
13 Ronsmans C, Brouwere V, Dubourg D, Dieltiens G. Measuring the need for But replication alone is not enough to yield the lifesaving obstetric surgery in developing countries. Br J Obstet Gynaecol 2004;
111: 1027–30.
substantial progress that we believe is possible, even 14 Editorial. Int J Gynaecol Obstet 1997; 59: S3–6.
in countries with high mortality rates.17 Programmes 15 World Bank. Project appraisal document on a proposed credit in the amount of SDR 61·O million (US$89·0 million equivalent) to the to reduce maternal mortality must be joined with Raebublic of India for Rajasthan Health Systems Development Project. bold eff orts to overcome the steep systemic barriers to Feb 9, 2004: http://www-wds.worldbank.org/external/default/WDSContentServer/IW3P/IB/2004/02/24/000160016_20040224122704/ equitable access that have been created during decades of Rendered/PDF/268230IN.pdf (accessed sept 6, 2006).
16 Government of India Ministry of Family Health and Welfare. The national harmful economic policies and political neglect. Because rural health mission framework for implementation, 2005–2012. eff ective maternal mortality programmes must include New Delhi: Government of India, 2005. http://www.mohfw.nic.in/NRHM%20-%20Framework%20for%20Implementation%20- facility-based services to treat obstetric complications, %20%203.3.06.pdf (accessed Sept 6, 2006).
www.thelancet.com Vol 368 September 30, 2006
UN Millenium Project. Who’s got the power? Transforming health systems for 18 Olsen OE, Ndeki S, Norheim OF. Complicated deliveries, critical care and women and children: fi nal report of the UN Millennium Project Task Force on quality in emergency obstetric care in Northern Tanzania. Int J Gynaecol Obstet Child Health and Maternal Health. New York: UN Development Programme, 2004; 87: 98-108.
2005: http://www.unmillenniumproject.org/reports/tf_health.htm (accessed 19 Rosenfi eld A, Maine D. Maternal mortality: a neglected tragedy: where is the M in MCH? Lancet 1985; 326: 83–85.
Japan: are statins still good for everybody?
Use of statins has become almost mandatory in a range pravastatin in women. The hazard ratio for coronary See Articles page 1155
of disorders, from coronary heart disease to is chaemic heart disease (0·71) did not reach statistical signifi cance.
cerebrovascular disease, and several other high-risk The male minority was responsible for the fi nal positive
conditions, such as diabetes and hyper
outcome. Some other fi ndings are worthy of attention. Although the guidelines for risk assessment have been There seemed to be no eff ect of the presence or absence somewhat softened—eg, the coronary heart disease risk of diabetes, being overweight, or smoking. Results were equivalence of diabetes4—the tendency to widen statin better in individuals older than 60 years and with an indications still remains.
LDL cholesterol concentration of more than 4·01 mmol/L.
Despite the widened indications, some questions The positive outcome of MEGA emphasises the about primary risk prevention remain open. Women importance of population versus individual approaches, are poorly represented in most studies and generally and of social choices for cardiovascular prevention. It achieve a lower risk-reduction than men, frequently not is a general rule to support the validity of the “lower reaching statistical signifi cance.5 Another example is the better” approach by cholesterol lowering in individuals at low risk—eg, Japanese, in whom the need graphic form, as best exemplifi ed in fi gure 4 of the for cholesterol-lowering drugs might possibly seem article by Opie and colleagues.9 MEGA underlines the overstated.
importance of calculating an absolute, rather than a A formal answer from Japan comes from the MEGA relative, risk-reduction after treatment. A 30% relative trial in today’s Lancet.6 Haruo Nakamura and colleagues risk-reduction might hide some important requirements: did a randomised open-label study, investigating such reduction for a population with a 1% global risk is pravastatin (10–20 mg per day) added to a low- actually only 0·3%, whereas if the disease occurs in all lipid diet in primary prevention in individuals with of the population a 30% risk-reduction is, well, 30%. In moderate hypercholesterolaemia (5·69–6·98 mmol/L). lipid-lowering trials, absolute risk varies widely, from very Of about 8000 participants, 70% were women. The high—eg, exceeding 22% in the 5 years of the 4S trial10 (with study provided overall positive results, reaching the an absolute risk-reduction of 7·7%)—to studies showing a gold standard of about 30% reduction in the diet plus far lower risk. These fi ndings allow the calculation of the pravastatin group versus the diet only group, with a number needed-to-treat as 1/absolute risk reduction—coronary heart disease endpoint. This risk reduction ie, 1/0·077=13 for the 4S trial. Low-risk trials, such as occurred despite modest total and LDL cholesterol AFCAPS,7 had an absolute baseline risk of only 5·5%, lowering by attributable reductions of 9·4% and 14·8% a risk reduction of 2%, and a consequent number needed-to-treat of 50. MEGA shows an absolute Most noteworthy from the results was the low number risk reduction of less than 1% and number needed- of coronary heart disease events—101 in the diet group to-treat of 119. In studies with a low absolute risk versus 66 in the drug-treated group. Such fi ndings would reduction, a paradoxical increase in non-cardiovascular be surprising in about 8000 patients with moderate deaths might occur, which happened in AFCAPS with cholesterol elevations followed up for over 5 years in the lovastatin and in the Helsinki and FIELD studies with USA or Europe,7,8 particularly since 40% of the Japanese fi brates,7 but not in MEGA. patients had hypertension and over 20% of patients were The reduction of events with lipid-lowering should thus not only be correlated to changes in total or MEGA does not yet provide the fi nal answer for LDL cholesterol, but also to the absolute risk in the placebo women, as it did not show a reduction of events after group.8 Risk-reduction is highest in trials with a high www.thelancet.com Vol 368 September 30, 2006

Source: http://www.old.givewell.net/files/Cause1-2/+UNICEF/Lancet%20MDG5%20Rosenfield%20Dream.pdf