Foreword: Clustering of mortality at INDEPTH member Health and Demographic Surveillance Systems.

Under the section Scientific Productivity, the 2010 report of the External Evaluation of INDEPTH (1) recommended that ‘INDEPTH must campaign and work for fuller exploitation of sites’ data, but should not impose specific research topics on sites. As a supplementary strategy to the comparative studies pursued by Working Groups, it might consider commissioning a series of “Illustrative Analyses”. This would entail an analyst working with one site's data, producing a research paper on a strategically important topic that demonstrates to other sites what they can potentially achieve with such data. Backup documentation of how to produce the results would be helpful’.

The above recommendation endorses what INDEPTH started in 2007. With funding from the Hewlett Foundation, I taught a small group of statisticians from Dodowa, Kintampo and Navrongo HDSS centres in Ghana the statistical methods in our paper, Clustering of Childhood Mortality in Rural Burkina Faso (2) based on data from Nouna HDSS in Burkina Faso. After two workshops which I facilitated in Kintampo and in Dodowa, my colleagues were able to work on centre-specific papers. One of these papers on Dodowa has been published in the journal Tropical Medicine and International Health (3). The other two on Kintampo (4) and Navrongo (5) appear in this INDEPTH mortality clustering supplement. It is indeed gratifying!

I then encouraged Elizabeth Awini, a young statistician at Dodowa HDSS, to lead an INDEPTH group on clustering of mortality and pass on the statistical methods to a wider group of INDEPTH statisticians and data analysts. Through core support from Sida/GLOBFORSK, the Hewlett Foundation, the Gates Foundation, the Rockefeller Foundation and the Wellcome Trust, INDEPTH supported a Clustering of Mortality Analysis Workshop in Accra.

This supplement is an outcome of the above efforts and is an example of how illustrative analyses are an effective strategy for increasing INDEPTH's scientific productivity and for strengthening the analytical capacities of young scientists at INDEPTH member centres.

The scientific content of this supplement highlights what is seldom visible to people working outside our HDSS centres (mostly in the rural areas), often termed as small areas. It is that most of the time, in-country variations are much greater than between-country variations. INDEPTH research reveals such inequalities because it is conducted at household/individual levels – obtaining micro-level data – and also highlights the fact that national statistics are seldom ‘national’ since they are often based on sample and/or infrequent national census data. Longitudinal HDSS is the only approach that enables extremely localised differentials to be seen; this has huge public health implications.

I am grateful to the guest editor of this supplement, Professor Heiko Becher, a former colleague at the University of Heidelberg Medical School in Germany and co-author of the paper (2), who we used for the illustrative analysis.

I hope that such success will continue to convince our funders of the invaluable impact made by their core support to INDEPTH.

Congratulations to the authors from the INDEPTH family.

Osman Sankoh

Executive Director, INDEPTH Network