To many people, the major health problem for sub-Saharan Africa is infectious disease. The avoidability of deaths from AIDS, tuberculosis and malaria are a valid justification for the major drives to fundraising to tackle these. But there is another spectre waiting in the wings: non-communicable diseases (NCDs) are now the major cause of death in every continent other than Africa. Even in Africa, absolute death rates from NCD exceed those in industrialised countries1—it is only the fact that infection mortality is still higher in Africa that prevents NCD deaths from heading the list.

The extraction of insulin and its purification in Canada in 1922 was followed, within months, by a transformation in the prospects for a newly diagnosed child with diabetes in North America and Europe. Yet over 80 years later, across much of Africa, the life expectancy for such a child is less than 1 year,2 with the unreliable availability of diagnostic tools, insulin and the healthcare needed for its safe use being largely to blame. The world’s poorest countries, as defined by the World Bank, are mostly in sub-Saharan Africa, where 10% of the world’s population live, surviving on an average of around US$2 per day. When national healthcare priorities, and donor support, are focused on infectious diseases and on the targets set as the Millennium Development Goals, services and drugs for NCDs are sometimes neglected by health ministries. So when insulin is not in stock at the local hospital, patients and their families may have to find the US$10–20 needed to buy a month’s supply of insulin from a private pharmacy, quite aside from other costs, such as syringes, testing strips, transport to health facilities, or consultation fees.

Until recently, much of the evidence needed to document patterns of diabetes care in resource-poor countries …