Feature story

Results of recent population-based surveys

05 July 2007

1. How have results of population-based surveys been used in refining AIDS estimates?

Since 2000 several countries have conducted nationally representative population-based surveys with HIV prevalence measurement. The first ones were conducted in Mali and Zambia in 2001-2002. By mid 2007 a total of 26 countries in sub-Saharan Africa had conducted surveys. In addition the two countries in the Caribbean with the largest HIV burden, Haiti and the Dominican Republic, as well as India and Cambodia in Asia had also conducted population-based surveys.

The information from available surveys has helped to refine UNAIDS published estimates. Where surveys were available prior to the publication of each of the UNAIDS biannual Global Report, the prevalence estimates from the survey have been used to inform the estimates in the UNAIDS Global Reports and consequently the two are very close to each other. As shown in the attached Table, all but one of these surveys (Uganda) has shown a lower prevalence in the survey compared to previously published estimates based on sentinel surveillance data.

The results of a number of surveys however have become available after the 2006 UNAIDS Global Report. Most of these have already been incorporated in the regional estimates published in the 2006 AIDS Epidemic Update, notably Cote d’Ivoire, Cambodia and Haiti. It is noteworthy that the recent survey in the Central African Republic showed a prevalence of 6.2% in 2006 compared to the published estimate of 10.7% for 2005. In other countries in sub-Saharan Africa recent estimates from the population based surveys are close to previously published estimates (Benin, Mali (2006), Niger (2006), Zimbabwe). In Asia, besides India, a population-based survey has also been conducted in Cambodia, suggesting a lower prevalence (0.6% in 2005) than previously estimated (1.6% in 2005).

2. Where can we expect significant differences in the results of population-based surveys and AIDS estimates?

Population-based surveys with HIV prevalence measurement are currently planned or ongoing in several countries in sub-Saharan Africa including the Democratic Republic of Congo, Eritrea, Gabon, Gambia, Liberia and Madagascar. There are no population-based surveys planned in other countries outside of sub-Saharan Africa, as most countries have low prevalence and would not benefit from this type of survey.

3. What is the impact of the differences in the results of population-based surveys and AIDS estimates?

According to the 2006 AIDS Epidemic Update report, the estimated number of people living with HIV in 2006 was 39.5 million people (range 34.1 - 47.1 million). Although several countries report lower prevalence, the total number is still well within the estimated global range for 2006 and remains large.

Cambodia was previously estimated to have 130,000 people living with HIV (range 74,000-210,000). Any changes in Cambodia’s estimates would therefore only have a minor impact on regional and global estimates. In sub-Saharan Africa too, it is expected that the new figures will only have a minor impact on regional prevalence (currently estimated at 5.9% (24.7 million)) and are well within the regional range (5.2% - 6.7% (21.8 - 27.7 million)).


Table of results of recent population-based surveys
More on revised estimates in India
More on Epidemiology
More on Methodology

Related feature stories