Reliable statistics on HIV/AIDS have been hard to come by in Liberia, slowly recovering from years of conflict, but two recently released surveys indicate a much lower HIV/AIDS prevalence rate than was previously thought.

"This is a reliable survey, involving 7,000 households across the country, where individual blood samples were collected and tested for HIV, and this exercise is the first in 25 years," said Edward Liberty, head of the Liberia Institute of Statistics and Geo-Information Services (LISGIS). Respondents sampled were in the age range of 15 to 49.

The first post-war demographic and health survey, conducted by the LISGIS from December 2006 to July 2007 in nine of the country's 15 counties, indicated an HIV prevalence rate of 1.5 percent of the country's 3.2 million people.

The civil war made most rural areas inaccessible, but when it ended in 2003, UNAIDS put HIV infection at 5.9 percent, while a 2002 Liberian government study estimated 8.2 percent.

According to the new report, "Of the 1.5 percent national prevalence rate, there is high percentage among females, which stands at 1.8 percent, while males accounts for 1.2 percent. There is a higher prevalence rate in urban areas, especially Monrovia, [the capital] than the rural zones."

The results of the first antenatal survey in 15 years, carried out between July and October in 2006 to determine the HIV infection rate among pregnant women attending clinics, have also been released, revealing a 5.7 percent HIV infection rate. Five heavily forested counties along the eastern coast and Cote D'Ivoire border registered the highest prevalence rate at 9 percent.

Countries worldwide use antenatal surveys as a statistical basis for measuring the prevalence of HIV/AIDS, and are now beginning to combine this with more representative data gathered in demographic surveys, because the age groups of the women, and the fact they were having unprotected sex, meant the numbers tended to overestimate HIV infections in the general population.

UNAIDS has warned that the high numbers of people who refused to be tested in household surveys, or were absent from home, could lead to underestimations of HIV prevalence. Sister Barbara Brilliant, a Roman Catholic nun who runs the Church's HIV/AIDS voluntary counselling and testing centre in Monrovia, agreed.

"A less than two percent national prevalence rate seems low, but this figure is from those who willingly accepted to have their blood sample collected. However, had it been series of tests of patients admitted at all health posts and hospitals, I am sure the figure would have been higher than the one LISGIS released."

Walter Gwenigale, Liberia's health minister, is cautious about attributing the lower figures to a decline in HIV infections. "The LISGIS survey was based on household sampling covering a broader-based population than the sentinel survey. However, the fact that the sentinel one differs with the LISGIS survey, which points to a lower rate, does not suggest that the level of HIV prevalence has decreased or dropped."