Lessons for mother and child health research, policy and action in the 21st century

We now know more about the relationship between poverty and health. In particular, we know about the relationship between income inequality and health. This is important for all countries because it illustrates that relative, as well as absolute, poverty matters for health. We know that good income, education, physical environment and social relations (social capital see Harpham et al.2) are associated with good health. But how often do we see arguments for investments in these sectors linked to health? The last century saw failure in inter-sectoral action for health (see for example Harpham et al.3). While we need to avoid the rhetoric that accompanied those earlier attempts, we must capitalize upon the wider resourced which are available for improving health. Let us move out of our sectoral boxes and train younger colleagues to be better at inter-disciplinary work. We also need international champions to highlight the role of poverty in health. In the 1970s and 1980s United Nations Children's Fund (UNICEF) was successful in the promotion of MCH. In the last ten years it seems to have lost its voice and vision. Similarly, the heady days of World Health Organization (WHO) leading the global move towards primary health care are long since gone. We now have an agency which focuses on single topic issues and apIntroduction


Introduction
International health witnessed many changes in the late 20 th century.In terms of health status there was a growth of chronic diseases but a worsening of some infectious diseases like HIV and TB.Within health systems there was experimentation with decentralization and sector wide allocation programmes (SWAPs).The research world called for more evidence based policy while governments debated the percentage of GNP spent on health and argued about how best to allocate expenditure within the health sector itself.What can we learn from these dynamic times in terms of mother and child health (MCH)?Below I present a summarized, partial viewpoint based on my own research and experience of working in international public health for the last twenty years.

Understanding the relationship between research and policy
Great strides have been made in enabling researchers to advocate evidence-based policy.Of particular use are the meta-analyses, previously limited to Europe and North America, which are now being undertaken on health problems which predominate in the South.See for example, a summary of the work on malaria by Omari and Garner. 1 However, many researchers remain naïve about the relationship between research and policy.They feel advocacy is beyond their remit and expect policy makers to be as convinced by the evidence as themselves.Although more health researchers are now studying the policy process there is still little understanding about the pears to be more vertical, medicalized and not engaged in the politics of health.Yet, at the same time, agencies which address poverty, e.g. the World Bank, seem unable to successfully link health into poverty reduction strategies.

The inability to scale up pilot projects
We have seen a move from projects (one off, small scale, often cost-intensive initiatives) to programmes (an array of activities involving day-to-day implementation of regular activities, including management and administration).This was partly a result of the failure to replicate, or scale up, pilot projects.However, we still see a plethora of projects which are not costed and which will never be replicated or extended, due to resource constraints.Basic costings are needed for all projects and a realistic assessment of the feasibility of replication should be made before embarking upon any project.

The need for proper evaluations
Evaluation has become a science of it own. 4And yet MCH professionals are largely unaware of it.As a result, projects and programmes often fail to have specific objectives with measurable outcomes and fail to collect baseline data (against which, change over time can be evaluated).Evaluation needs to be de-mystified, properly planned, resourced, and implemented with rigour.Vast resources have been dedicated to health information systems and management information systems but community-based information is still poor and we are mainly dependent upon the relatively infrequent Demographic Health Surveys and similar exercises when it comes to systematic population based health data.Censuses in many countries remain unreliable and irregular.Lack of such routine data adds to the difficulties of evaluations.

Conclusions
There is inevitably a tension between specialized, focused, narrow approaches to MCH and broader, contextual, integrated efforts.The 1980s were characterized by the latter, the 1990s by the former.We need to take the best from both -a typical post-modern mixture.The challenge is to identify past successes and failures in the relevant cultural and political settings.'One size' will not fit all.However, there are still sufficient similarities in health problems and health systems to allow some common learning.Evaluation needs to inform policy and action, and researchers can make a valuable input to the design and implementation of evaluations.