Marseille 2007
Marseille 2007
Abstract book
Go Back

Abstract #470  -  HIV Prevention in Africa: A Meta-Analysis
Session:
  45.2: Prevention in the new millenium (Parallel) on Tuesday @ 16.30-18.30 in PR Chaired by Yolande Obadia, Araceli Rousaud
Authors:
  Presenting Author:   Prof Blair Johnson - University of Connecticut, United States
 
  Additional Authors:  Prof Natalie D. Smoak, Dr Chandra Y. Osborn, Prof Michael P. Carey,  
Aim:
To provide a summary of the efficacy of studies that have evaluated HIV prevention strategies in African countries.
 
Method / Issue:
Relevant studies were located using database search strategies, HIV-related listservs, and conference proceedings. Studies were included if they (1) examined a deliberate HIV-risk-reduction intervention in an African nation; (2) used a controlled trial including a true control group; and (3) measured a sexual risk reduction marker (i.e., condom use, number of sexual occasions or partners, or STD/HIV prevalence) at least once following the intervention. Studies were excluded if they (1) focused on perinatal transmission contexts or behaviors; (2) did not emphasize HIV content; or (3) included a structural or environmental intervention (e.g., a mass media intervention). Retrieved studies were coded along numerous dimensions and their outcomes were converted into standardized effect sizes Analyses followed fixed- and random-effects assumptions in terms of evaluating the mean tendencies for each outcome variable and followed fixed-effects assumptions in terms of testing whether features of the studies could explain variability in the magnitude of effect sizes.
 
Results / Comments:
Application of the selection criteria resulted 23 studies with 31 separate interventions. Mean HIV prevalence was 24.3%. These interventions began with a total of 63,342 participants; the average retention rate was 69%. The majority of the studies (61%) included skills training, but only 16% included participants and their partners. About one-third (36%) of the interventions were bio-behavioral; 32% were social/cultural, and 13% were a mix of biological, social, and cultural. Measures were taken at a mean of 43.9 weeks after the intervention. Effect sizes showed wide variability on nearly every outcome. Generally, studies showed efficacy on the most commonly reported outcome, condom use. Greater condom use occurred to the extent that interventions provided behavioral strategies, provided HIV information, used counseling and testing, provided HIV education in schools, and allied the intervention with a structural change. Sexual-frequency outcomes did not appear to change as readily as did condom use, but these were not as frequently reported.
 
Discussion:
HIV prevention trials conducted in Africa are showing efficacy but of widely varying magnitude. Efficacy is larger on condom use than sexual frequency outcomes such as number of sexual partners. Greater condom use change can occur when interventions provide behavioral strategies, give HIV information, used counseling and testing, provide HIV education in schools, and ally the intervention with co-occurring structural change. Discussion also centers on the issue of quality of methods.
 
Go Back

  Disclaimer   |   T's & C's   |   Copyright Notice    www.AIDSImpact.com www.AIDSImpact.com