Olivia Roberts: “Lies, damn lies, and statistics”

The recently released World Health Organisation statistics on progress to meeting the health-related Millennium Development Goals reveal little progress in some areas. But closer analysis reveals some huge achievements and underlines the need for greater evaluation of what works and what doesn’t, rather than cutting development aid.

On 21 May 2009, the World Health Assembly received the mid-way report card on health-
related Millennium Development Goals, the first progress report on the MDGs since their
inception in 2000. It concluded that results were “mixed”, but for some goals the results are starker: maternal mortality has barely changed since 1990, two-thirds of 9.7 million people in developing countries who need treatment for HIV/AIDS did not receive it, pneumonia and diarrhoea continue to kill 3.8 million children aged under five each year, although both conditions are preventable and treatable, and in about 30 developing countries show that availability of selected medicines at health facilities was only 35% in the public sector.

So is the WHO damned by its own data? Yes and no.

That the end-point for the MDGS is six years away and seemingly little progress has been made is a cause for concern and, on the face of it, seems to resonate strongly with Dambisa Moyo’s damning account of development assistance in “Dead Aid”. The Zambian economist’s solution, as proposed in Dead Aid, is cold turkey. Her mantra that aid is something for nothing and has created reliance resonates strongly in these times of missed targets and competing pressures for financial resources. So is the answer?

Saving lives is an expensive business and quick solutions seem appealing when faced with 9 years, billions of dollars and seemingly little progress. But as well as being expensive, saving lives is complex, and so is interpreting the data. Releasing this data, the first time in 9 years, is a risky move for the WHO and badly timed considering the rave reviews that Moyo’s theories have received. But the data is both fascinating as well as valuable.

In the first instance, the baseline data dates from 1990, providing a longitudinal perspective that mirrors the social and political upheavals that the world has seen during this period. Mortality of children under 5 years old in Zimbabwe (MDG 4) increased from 95 per 1000 live births in 1990 to 122 per 1000 in 200 to 90 per 1000 in 2007 – demonstrating how, in the words of the Red Queen in Alice’s Adventures in Wonderland, it takes all the running you can do to keep in the same place.

Unpicking other indicators tell a different story. While a child’s chances of living beyond 5 years old has barely changed in some countries, how long the child live beyond this has seen some significant increases. Life expectancy in Bangladesh has seen 10 year increase, from 54 years to 1990 to 64 in 2007 and a staggering 37 years in 1990 to 51 in 2007 in Niger.

The data also provides a succinct snapshot of global injustice – living to 60 years old in Botswana is less likely today than in 1990 whereas a 60 year old celebrating their birthday in Japan is only three-quarters of their way to life expectancy of 83 years. The data also provides examples of almost 20 years of failure with the data from Afghanistan is particular being consistently bleak: unchanged life expectancy – 41 for men, 42 for women – and the mortality of children under 5 virtually unchanged at 257 per 1000 live births in 2007 compared with 260 per 1000 live births in 1990.

So why shouldn’t we take the medicine expounded in Dead Aid and cut off the flow of aid when it seems to have such little impact? For starters, how do you know which aid isn’t working? Changes in child mortality may be attributable to the distribution of bed nets, with estimates for Niger, Gambia and Ethiopia have 59.6%, 49% and 42% of children under 5 sleeping under insecticide treated bednets. The changes, however, could also be due to increased immunisation coverage, use of oral rehydration therapies during episodes of diarrhoea, improved water and sanitation.

It is nothing new to say that the numbers are complicated and contradictory and for some countries paint an unremittingly bleak picture. But to say that they are reason to scrap development in its entirety are, to paraphrase Kevin Watkins, like banning fire engines because they are often found near burning buildings. What must be done is to use the data to look in more detail at those countries where programmes are not working and figure out why. The WHO should, in the years to 2015, dedicate itself to combing through this data and identify the lessons for better planning and evaluation, not less aid. It is also too simplistic to focus on just eight goals and is a recipe for disappointment, which is something the WHO realised and subsequently developed sub categories of “targets” and “indicators” in 2007. The WHO should seriously consider the addition of more indicators – knowing more might make for uncomfortable reading but is vital for improved efficacy and sustainable progress.

Olivia Roberts is senior research officer in the International Department, British Medical Association London WC1H 9JP