This is a joint post with Yuna Sakuma
My recent editorial in the Guardian highlighted the importance of better defining and verifying results as part of the next Global Fund funding cycle. In particular, I’m critical of using indicators like “bed nets distributed” to convey anything about the impact of the program on disease.
The Global Fund reports 30 million bed nets distributed in the first half of 2013. However, country-specific data from the Malaria Indicator Surveys (MIS) and Demographic and Health Surveys (DHS) shows that the number of bed nets distributed is not directly related to the share of women or children sleeping under a net, much less the number of malaria cases prevented which is a function of the appropriate use of the net and the cumulative effectiveness of other malaria control measures (see table below).
Let’s look at malaria-endemic Malawi as an example, where almost all households are at risk of the disease and could benefit from using bed nets. The Global Fund and the President’s Malaria Initiative (PMI) report that 11.8 million insecticide-treated bed nets (ITN) and long-lasting insecticide-treated nets (LLIN) have been distributed. This figure suggests that – on average – each household in that country might own 3.1 bed nets each.
Yet according to 2012 MIS results, only 55% of households in Malawi report owning at least one ITN. Only 56% of children under age 5 and 51% of women ages 15–49 slept under an ITN the night before they took the survey. Not only are over 40% of households not in possession of an ITN, those who do own a net are not using them (at least not consistently). The story repeats in many other countries.
There are lots of caveats with this simple comparison: the distribution numbers are cumulative over a period of time not specified by the Global Fund website (for PMI, the time period is 2007–2012); bed nets are also sold in the private sector; bed nets wear out and are replaced; in some countries classified as malaria-endemic, not all households are actually in malaria-endemic areas; and so on.
But this simple comparison does illustrate the discrepancy between the number of products distributed and their actual use and effectiveness. And the bottom line is that it just makes more sense to track what matters for disease impact – ITN use, rather than distribution.
This approach is backed up by analyses like this one: a paper by Gabriel Demombynes and Sofia Karina Trommlerová showing how household ITN ownership is associated with child mortality declines, which notes that “the widespread ownership of insecticide-treated bed nets in areas of Kenya where malaria is rare suggests that better targeting of insecticide-treated bed net provision programs could improve the cost-effectiveness of such programs.”
This is really the point of all our work on value for money: we should measure what matters for impact – like bed nets use, not distribution – and we should assess how we target our interventions so that the money goes as far as it can to improve health.
At a recent CGD event, the Global Fund's Christoph Benn said, “We did get [your] message.… In the future we will go more into outcome indicators and measurement and report that.” I was thrilled to hear that and look forward to seeing the fruits of this progress from the Global Fund and other global health funders in the (near) future.
Note: Countries selected if DHS conducted in 2010 or later; ITNs per household = number of distributed ITNs / (population / average household size); *PMI focus countries
Sources: 1. The Global Fund and PMI websites; 2. World Bank population estimates; 3. MIS / DHS