IMCI: what can we learn from an innovation that didn't reach the poor?



Davidson R Gwatkin

6483 Wishbone Terrace, Cabin John, Maryland 20818, USA (email: dgwatkin@comcast.net)



In this month's Bulletin, Cesar Victora et al.1 continue their ongoing assessment of the Integrated Management of Childhood Illness (IMCI) strategy, by looking at how well it reached poor areas within three countries. Their findings are sobering: overall, the strategy seemed to be implemented least energetically in the areas where it was most needed.

This illustrates one of the many cruel ironies of efforts to help the poor: the tendency of service programmes to be much weaker in deprived areas than elsewhere. As a result, the promoters of new initiatives tend to adopt the IMCI strategy of beginning in easier, better-off districts, to achieve the early successes needed to establish credibility. The intent is to expand into more difficult areas as soon as possible. Whether such expansion ever takes place is usually unreported; but the well-documented difficulties encountered by IMCI are consistent with anecdotal evidence about other programmes, and they will sound familiar to anyone who has tried to introduce new approaches.

This suggests that the issue is not unique to IMCI. If it applies to public health interventions in general, what are the implications for the design of initiatives to reach disadvantaged groups? Three stand out.

First, the IMCI experience illustrates the distinction that needs to be made between developing interventions that address the needs of the poor, and reaching the poor with those interventions. The relevance of IMCI interventions is beyond doubt: study after study has demonstrated a much higher prevalence of childhood illness among the poor.2 Yet no matter how relevant, an intervention cannot help the poor unless it gets to them. When the evidence presented here is added to that assembled by Victora et al. in their earlier IMCI work,3 it becomes clear that the IMCI approach has reached very few of the world's poor during the first ten years or so of its existence. A programme addressing a problem less important for the poor that reached them effectively would have produced more benefit.

Second, the striking contrast between the spread of IMCI and of oral rehydration, the earlier focus of WHO's child health efforts, supports the arguments of those preferring such "vertical" initiatives over more "horizontal" efforts to strengthen health systems. Oral rehydration was the epitome of a vertical approach, dealing with one particular intervention against one specific health problem. Whatever one might think of such a vertical approach in principle, oral rehydration’s rapid and widespread acceptance cannot be denied. For instance, in the early 1990s, approximately ten years after oral rehydration was introduced, it was being used to treat over half of childhood diarrhoea cases among the poorest 20% of the population in the nine countries with available data.4 Did the focus on oral rehydration delivery inhibit the longer-term development of health systems? Perhaps. Can one draw firm general conclusions from a comparison of only these two experiences? Certainly not. Yet in at least this one instance a vertical programme clearly performed much better — in terms of acceptance and likely health improvement — than a more horizontal one.

Third, adding a distributional element to the assessment of programme effectiveness increases the challenges that health planners face. With regard to health systems, it means worrying not only about the performance of programmes in districts like those featured in the IMCI early implementation phase, but also about ensuring that the poorest, most difficult areas are equally well served. These are the areas where health personnel are most reluctant to serve, where transport is most difficult to arrange, where basic financial infrastructure is lacking. It is not clear that traditional approaches to health systems development can overcome these difficulties. Striking out in different directions may be necessary. Giving highest priority in human resource planning to the development of lower- and middle-level cadres more likely to work in remote areas, for instance; or emphasizing contracts with nongovernmental and other service providers, seems to have worked well in several difficult settings.5 While the long-term value of such approaches remains to be determined, in the absence of innovative thinking, health systems will almost certainly continue to overlook the poor.


1. Victora CG, Huicho L, Amaral J, Armstrong-Schellenberg J, Manzi F, Mason E, et al. Are health interventions implemented where they are most needed? District uptake of the Integrated Management of Childhood Illness strategy in Brazil, Peru and the United Republic of Tanzania. Bull World Health Organ 2006;84:792-801.

2. UNICEF. End decade assessment: Multiple Indicator Cluster Survey. http://www.childinfo.org/MICS2/MICSDataSet.htm

3. Bryce J, Victora CG, Habicht JP, Vaughan JP, Black RE. The Multi-Country Evaluation of the International Management of Childhood Illness Strategy: Lessons for the Evaluation of Public Health Interventions. Am J Pub Health 2004;94:406-15.

4. World Bank. Health, Nutrition and Population. www.worldbank.org/hnp/povertyandhealth/countrydata

5. Loevinsoh B, Harding A. Buying Results? Contracting for Health Service Delivery in Developing Countries. Lancet 2005;366:676-81.

World Health Organization Genebra - Genebra - Switzerland
E-mail: bulletin@who.int