Denison JA, O'Reilly KR, Schmid GP, Kennedy CE, Sweat MD. HIV voluntary counseling and testing and behavioral risk reduction in developing countries: a meta-analysis, 1990-2005. AIDS Behav 2008;12:363-73.
To perform a meta-analysis assessing the effectiveness of HIV voluntary counseling and testing (VCT) in reducing HIV risk behaviors among people in developing countries
Trained staff searched the following computer databases: National Library of Medicine's Gateway database, Psych INFO, Sociological Abstracts, the Cumulative Index to Nursing and Allied Health Literature, and EMBASE, as well as the following HIV/AIDS-related journals: AIDS Care, AIDS, AIDS & Behavior, AIDS Education and Prevention, and the Journal of AIDS. Additionally, the reference sections of previous review papers and meta-analyses were searched. Studies meeting the following criteria were included: 1. conducted in a developing country or emerging economy (World Bank 2005); 2. evaluated VCT interventions that adhered to the US Centers for Disease Control and Prevention (CDC)/Joint United Nations Programme on HIVAIDS (UNAIDS) guidelines;(1,2) 3. used assessments pre- and post-VCT or in people receiving or not receiving VCT; 4. measured HIV-related outcomes, such as knowledge and HIV-risk behavior; and 5. were published between 1990 and 2005. Identified studies were independently reviewed for eligibility by two senior staff members, and differences of opinion between the reviewers were resolved through discussion.
Five hundred thirty-eight potentially relevant studies were identified, of which 124 were retrieved for more detailed evaluation. Eighteen studies were eligible for systematic review, but only seven met all criteria and were included in the meta-analysis: two prospective cohort studies, one randomized controlled trial (RCT), and four cross-sectional studies. The seven studies were implemented between 1998 and 2002. Studies were evaluated on an 8-point scale of rigor. The rigor scores ranged from 1 to 6. Only the RCT received a score as high as 6 of the possible 8 points.
The seven studies included were conducted in Africa, Asia, and the Caribbean; specifically, Kenya, Rwanda, Tanzania, Thailand, Trinidad & Tobago, and Uganda
Four studies were conducted among people seeking testing at VCT centers or who were offered a community-based VCT approach. The other three were among women alone or women and their partners attending antenatal, pediatric, obstetric, or family planning clinics. Only one study randomly assigned participants to the intervention. Only one of the studies used a probability-based sampling method, and the remaining six employed non-random selection of participants. The number of participants in the included studies ranged from 120 to 6,088.
VCT consists of learning one's HIV status coupled with pre- and post-test counseling. This definition corresponds to international standards initially developed by the US CDC and UNAIDS. These guidelines recommend that during pre-test counseling, trained counselors and those being tested discuss the testing process, assess the client's risk behavior, discuss coping strategies related to the test results, review prevention options, and reaffirm the decision to test for HIV. During post-test counseling, clients not only receive their HIV test results but discuss risk reduction strategies and disclosure of test results; if appropriate, they are provided with referrals to HIV care and support. The follow-up period ranged from 2 weeks to 1 year. Two studies had follow-up rates of 80% or greater.
Three studies reported on the number of sex partners, while all seven studies presented data on unprotected sex. The definitions of unprotected sex and condom use varied across studies. Effect size estimates were standardized to an odds ratio metric for the meta-analysis. For number of sex partners, all studies presented results as a dichotomous outcome (proportion reporting two or more partners); for unprotected sex, either the proportion that did not use condoms or the proportion that had unprotected sex was reported.
Pooling data from the seven studies that presented data on unprotected sex (n=12,348), the authors found that VCT recipients were significantly less likely to engage in unprotected sex after than before receiving VCT, or compared to participants who had not received VCT (odds ratio [OR] 1.69; 95 confidence interval [CI], 1.25-2.31; P<0.01). Although one study found a statistically significant relation between VCT and a decreased number of partners, pooling data from three studies (n=8,803) showed that VCT had no significant effect on the number of sex partners (OR 1.22; 95% CI, 0.89-1.67; P>0.05).The statistically significant Q-statistic for heterogeneity showed that the findings across studies for both primary outcomes were not consistent, and that factors other than sampling error may explain the variability between the results of the studies.
The authors concluded that although these findings provide only moderate evidence to support VCT as an effective prevention strategy, they do not negate the need to expand access to HIV VCT services.
Using the QUOROM checklist for assessing the quality of meta-analyses, this study was of adequate quality. Although the authors assessed the quality of the included studies and performed tests for heterogeneity, the components of the interventions, the target populations and the outcome measures varied widely between the studies. As the average rigor score of the included studies was 3 of a possible 8 points and as none of the included studies assessed the long-term effects of VCT, the conclusions of this meta-analysis should be interpreted with caution. The use of a dichotomous measure of number of sex partners may have been insensitive. All outcome measures were self-reported and were subject to possible reporting bias. Additionally, as all included studied focused on adults, primarily in urban settings, generalizing the results of this meta-analysis to other populations should be avoided.
This study was the first meta-analysis to focus solely on VCT efficacy data from developing countries. A review of VCT studies conducted between 1990 and 1996 showed that the effect of VCT on risk behavior varied based on the target population and the studies' design, with the strongest evidence of behavior change observed among discordant couples.(3) Additionally, a meta-analysis conducted in 1999 examining the efficacy of VCT supported the intervention as an effective behavior change strategy for persons infected with HIV.(4) The lack of a biological outcome measure and the reliance on self-reported behavior are additional weaknesses. Unfortunately, the majority of studies supporting VCT as an effective strategy for behavior change have been conducted in developed countries in North America and Western Europe.
The expansion of VCT as a prevention strategy should be accompanied by rigorous evaluation to maximize the prevention benefits of learning one's HIV status. Although the authors of this study concluded that the modest evidence of effectiveness should not negate the need to expand VCT services, its potential should be weighed against other interventions in allocating prevention funding.
- CDC Revised guidelines fro HIV counseling, testing and referral and revised recommendations for HIV screening for pregnant women. MMWR 2001;50(RR19):1-58.
- UNAIDS. Voluntary counseling and testing (VCT). UNAIDS Best Practices Collection. Technical Update. 2000;1-12.
- Wolitski RJ, MacGowan RJ, Higgins DL, Jorgensen CM. The effects of HIV counseling and testing on risk-related practices and help-seeking behavior. AIDS Edu Prev 1997; 9(3 Suppl):52-67.
- Weinhardt LS, Carey MP, Johnson BT, Bickman NL. Effects of HIV counseling and testing on sexual risk behavior: a meta-analytic review of published research, 1985-1997. Am J Pub Health 1999; 89(9):1397-1405.