AEGiS-15IAC: Incidence of HIV infection among women in Abidjan, Cote d'Ivoire: Estimation from serial antenatal seroprevalence data.

15th International AIDS Conference

Bangkok, Thailand - July 11-16, 2004

Incidence of HIV infection among women in Abidjan, Cote d'Ivoire: Estimation from serial antenatal seroprevalence data.

Int Conf AIDS 2004 Jul 11-16; 15:(abstract no. ThOrC1422)

Sakarovitch C, Msellati P, Leroy V, Becquet L, Atta H, Viho I, Ouassa T, Welffens-Ekra C, Dabis F, Alioum A
Unite EMI INSERM 03 38, Unite INSERM 593, Universite Victor Segalen, Bordeaux, France

BACKGROUND: Incidence of HIV infection is a key indicator for the planning and evaluation of national AIDS control programs, but direct measures are difficult to obtain in Africa. By contrast, seroprevalence data on specific sentinel groups such as pregnant women are widely available. We estimate incidence of HIV infection among women aged under 40, in Abidjan, Cote d'Ivoire, from serial seroprevalence data obtained from pregnant women.

METHODS: From 1995 to 2002, HIV counseling and testing was systematically proposed to pregnant women aged over 18 years attending antenatal clinics where programs of prevention of mother-to-child transmission of HIV were implemented. Incidence was estimated using a maximum likelihood approach described by Ades and Medley (Stat in Med, 1994), adapted to pregnant women in Abidjan. This method takes into account both age and time trends of incidence, and the differential inclusion of infected and uninfected women in the surveys. We modelised the differential inclusion rate as a function of time of the survey, age at inclusion, at infection, and time since infection.

RESULTS: Overall, more than 36000 women were tested during the eight years period. Incidence rate was found to depend strongly on age and calendar time. women aged 17 to 21 year old had the highest incidence rate yielding 2.5%/year on average. The peak of the epidemic was observed between 1989 and 1992 with incidence rate about 3%/year. It then decreased until 1998, and is now stable around 0.5 to 1.5%/year.

CONCLUSIONS: While the peak of HIV transmission rate seems to have occurred 10 years ago, the epidemic is still active. These results have to be considered with caution as estimates are based on differential inclusion of infected and uninfected women modelised from available information in the literature. However, the method is particularly relevant and informative in Africa and could be helpful in monitoring the HIV epidemic.

Keywords: AEGIS, Incidence, HIV Infections, HIV Seroprevalence, Seroepidemiologic Studies, Acquired Immunodeficiency Syndrome, HIV Seropositivity, Cote d'Ivoire, Research Design, Africa, Humans, Female, Child, Pregnancy, epidemiology

040711
ThOrC1422