Globally, more than 34 million people are infected with HIV; in sub-Saharan Africa alone, 3 million new infections occur annually.
In an attempt to stop the spread of HIV, governments in the region are considering providing antiretroviral drugs to people who do not have the virus but are at risk for becoming infected, researchers said.
Such drugs are known as pre-exposure prophylaxis, or PrEP.
Although the conventional strategy - attempting to distribute the drugs to people in every city and village - might seem logical and equitable, researchers at University of California, Los Angeles have devised a plan they said would be much more effective in reducing HIV transmission.
The strategy, developed using a complex mathematical model, focuses on targeting "hot zones".
In South Africa, where 17 per cent of the population is infected with HIV, the model predicted that targeting hot zones would prevent 40 per cent more HIV infections than using the conventional strategy - and would therefore be 40 percent more cost-effective.
To develop the strategy, UCLA researchers designed a computer model that calculated and mapped the incidence of HIV in South Africa and identified hot zones.
The model featured three important components: the geographic dispersion of the population, the geographic variation in the severity of the HIV epidemic and the geographic variation in the level of risk behaviour.
The model revealed that two of South Africa's nine provinces are hot zones.
The researchers then used the model to predict where, and how many, new HIV infections would occur based on using either the conventional strategy or a strategy targeting hot zones for distributing the drugs.
"Our results are quite striking. Both strategies would provide PrEP to the same number of people, but using the hot zones plan would prevent 40 per cent more HIV infections than using the conventional plan," said Sally Blower, the paper's senior author.
"The methods we developed can be used to find hot zones in any other sub-Saharan countries that have geographic variation in the severity of their HIV epidemic, such as Lesotho, Botswana, Nigeria and Uganda," said David Gerberry, the study's first author and a former UCLA postdoctoral fellow who now is an assistant professor in mathematics at Xavier University.
"Once the hot zones have been found in these countries, our spatial optimisation algorithm can be used to identify the geographic targeting strategy that would be the most cost-effective," he said.
The study appears in the journal Nature Communications.