DECADES of assumptions about the lethality of malaria have been overturned by a paper from an academic institute in Seattle which says the disease kills twice as many as everybody thought. Even more extraordinary — it would seem that conventional wisdom about the disease has been wrong all this time. It does not just kill babies and children under five — it kills adults too, in nearly as large proportions.

The Institute of Health Metrics and Evaluation has astounded the global health community by claiming it has been fighting malaria apparently with one hand behind its back. The death toll has come down since 2004, thanks to huge efforts to get insecticide-impregnated bednets to households and treat those who are sick with better drugs, but all the while an older age group has been neglected.

“These are certainly results which surprised us when we first did the analysis,” said Steve Lim, one of the authors of the Lancet paper published today. “It is new to what is taught in public health and medical school, which is that when kids are exposed to malaria at a very young age, it conveys immunity.”

Only last year the World Malaria Report gave mortality figures which are half those the institute has found — 655,000 deaths compared to 1.2 million. It is an extraordinary gulf and there will be lots of debate about the statistical methods used by the Seattle team.

But the institute has form. This is part of a five-year project, funded by the Bill and Melinda Gates Foundation, to obtain the best possible data for the toll of death and disease from various causes in the developing world.

Its director Christopher Murray and colleagues have already published on maternal mortality. They found that the half a million deaths on pregnancy and childbirth that had been trotted out for years was wrong — the real figure was nearer a third of a million.

Estimates can go up as well as down. For malaria, they say there has been under-reporting. One of the main reasons is that assumption that older children and adults do not die of it — so deaths may be classified as ‘fever’ when malaria was to blame.

Good national data on mortality in poor countries is extremely difficult to find. Deaths, like births, go unrecorded in most of sub-Saharan Africa. There are no registers. The institute has gone to great lengths to find reliable data as possible that can be relied on from a wide variety of sources.

Mostly these are studies carried out by scientific researchers. But they have incorporated studies that others might not, such as ‘verbal autopsies’, for instance where a relative, usually, tells the researchers about the health of the deceased, their symptoms and the way they died.

It is not a doctor’s diagnosis and there are no blood or tissue tests for confirmation, but some causes of death can be fairly securely identified in this way.

Lim describes what they have been doing as “a really systemic effort to collect all the world’s data and all the studies that measure mortality”. That data all has to be validated, in so far as it can be.

He points out that because they are working on all causes of death, they do not have a vested interest in malaria. And if malaria cases go up, then other causes of death must go down (as they did for women in childbirth) — although they also recognise that the total number of deaths in the world is not known.

Asked why other reports have not come up with similar figures, he said: “It is a little bit difficult for us to know exactly what’s been done by other groups. Their methods are not always particularly described.” — The Guardian, London