After months of angry disputes between different global health agencies about how best to reduce deaths from malaria in Africa, the disputing parties Unicef, the World Health Organization, and Médecins Sans Frontières came together in New York last week to present a united face and declare that artemisinin based combination therapy (ACT) must urgently be made available across Africa.

In a joint consensus statement, the three agencies together with Columbia University's school of public health, which hosted the meeting acknowledged that "expanding access to ACT is increasingly a matter of life and death for people at risk of malaria," and they committed themselves to "discontinuing support for the use of ineffective medicines."

Dr Jean-Marie Kindermans of Médecins Sans Frontières said that his organisation's call for Unicef and the Roll Back Malaria partnership to implement ACT "has not been smooth sailing.

"He therefore welcomed the statement but cautioned that the rhetoric must be turned into action. Professor Nick White, professor of tropical medicine at Mahidol University, Thailand, said that in Africa "we are failing to roll back malaria" but pointed to Vietnam as an example of how malaria can be controlled by using effective malaria medicines.

Half the population of Vietnam is at risk of malaria. The Vietnamese government, facing a rise in malaria deaths owing to resistance to conventional drugs, introduced ACT in 1991.

There was a very dramatic and rapid fall in malaria deaths, said Dr Le Dinh Cong, former director of Vietnam's national malaria control programme.

Compared with 1991, he said, mortality from malaria has fallen by 97.3% and morbidity by 77%. ACT costs at least 10 times more than the conventional, though largely ineffective, drugs that are used in most of Africa.

Malaria disproportionately affects the poor, who spend a third of their disposable income on buying conventional malaria drugs. ACT will therefore be unaffordable to those who need it. There was unanimous agreement at the meeting that donors must cover the costs.

"Donors are in the vanguard of those who've been reticent to move things forward," said Professor White. Three of those donors the World Bank, the US Agency for International Development, and the UK Department for International Development were at the meeting. All promised their increased support for rolling out ACT in Africa.

But both the bank and the UK Department for International Development could not give any figures on how much they are currently spending on ACT, and the US Agency for International Development has a policy of not purchasing any malaria drugs.

"Donors have never taken malaria seriously," said Professor Jeffrey Sachs, director of Columbia University's Earth Institute. "One to three million [malaria] deaths a year is mass neglect."

Professor Sachs called on the donors to increase massively their spending on malaria control, from the current $100m a year, to $3bn to fund an "appropriate, integrated continent-wide effort."

At the Abuja summit in malaria in 2000, the World Bank announced that it would spend $500m on malaria control, and yet "not much has happened" since this announcement, said Professor Sachs.

"These announcements come and go. This one went." What the bank must do, he said, is to put its money into a separate specific account for malaria control, analogous to its multicountry AIDS programme for Africa. He also said that donor agencies must distribute effective malaria drugs and insecticide-treated bed nets free of charge.

He demanded that the UK Department for International Development and the US Agency for International Development abandon their current policy of asking patients to pay for bed nets. "We're dealing with one of the most dangerous conditions in world history which afflicts the very poorest people, the world's most desperate people, people who don't have enough to pay to eat, let alone buy nets or malaria drugs". 

Dr Dennis Caroll, senior health adviser for the US Agency for International Development, told the BMJ: "I endorse [Sachs'] call for a massive increase in resources," but said there are "differences in opinion" in how best to use the money.

Many speakers at the meeting pointed out that children with malaria are often treated at home, not in medical facilities. So any strategy to get ACT to sick children must include a community based approach, not just an approach that targets medical facilities.

Dr Ronald Waldman of Columbia University's school of public health said that "communities should not be used as dumping grounds for ineffective drugs."

If medical facilities are using ACT, it would be wrong, he said, for ineffective drugs like chloroquine and sulfadoxine pyrimethamine to still be available at the community level. Those at the meeting generally agreed that scaling up of ACT should be accompanied by distribution of rapid diagnostic tests for malaria, which involve dipstick testing of blood samples for the malaria parasite.

In many places in Africa, malaria drugs are given to anyone with a fever. But with ACT being so costly, a strategy of giving it only to those with a positive test should help keep costs to the minimum.

One of the biggest obstacles to rolling out ACT is that a huge increase in the manufacture of these drugs will be needed.

The current production capacity is 30 million annual treatments, but WHO estimates that by 2005, between 131 million and 219 million treatments will be needed worldwide.

Only one fixed dose combination is currently available artemether with lumefantrine, which Novartis is supplying to WHO at cost price though other combinations are available as individual generic drugs that must be coadministered.