University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

The impact of HIV treatment on risk behaviour in developing countries: A systematic review
Global Health Sciences Literature Digest
Published July 13, 2007
Journal Article

Kennedy C, O'Reilly K, Medley A, Sweat M. The impact of HIV treatment on risk behaviour in developing countries: A systematic review. AIDS Care 2007 Jul;19(6):707-20.


To examine the strength of evidence for the impact of medical treatment for HIV-infected individuals on behavioral outcomes in developing countries.


Studies were included if they met the following criteria: 1. data were presented from a developing country, defined by combining the World Bank categories of low-income, lower-middle income, or upper-middle income economies; 2. participants received a diagnosis of HIV or were presumed to be HIV-infected based on clinical signs and symptoms; 3. clinical treatment was provided, including treatment for opportunistic infections and antiretroviral therapy; 4. an evaluation design was employed that compared post-intervention outcomes using either a pre/post- or multi-arm study design (including post-only exposure analysis); 5. behavioral, psychological, social, care, or biological outcome(s) related to HIV prevention were presented; and 6. the article appeared in a peer-reviewed journal from January 1990 through January 2006. No language restrictions were used. When an article in a language other than English was found, it was translated into English, and then coded.

Search Strategy

The authors searched electronic databases including the U.S. National Library of Medicine's (NLM) Gateway system, PsycINFO, Sociological Abstracts, EMBASE, and the Cumulative Index to Nursing & Allied Health Literature (CINAHL) for articles published from January 1990 through January 2006. To identify articles not obtained from electronic database searching, study staff handsearched the table of contents of the following journals: AIDS, AIDS and Behavior, AIDS Care, and AIDS Education and Prevention. Finally, the authors examined the reference lists of articles selected to further identify potential articles for inclusion. This process was iterated until no new articles were found.


Patients in developing countries who had received a diagnosis of HIV or were presumed to be HIV-infected based on clinical signs and symptoms, and were receiving clinical treatment, including treatment for opportunistic infections and antiretroviral therapy.


Clinical HIV treatment, including treatment for opportunistic infections and antiretroviral therapy.

Outcome Measures

Sexual abstinence, frequency of sexual intercourse, condom use, type of partner, longitudinal risk and HIV transmission, and sexually transmitted disease (STD) treatment.


Of the 166 identified articles, 104 were excluded for not meeting inclusion criteria and 59 were used for background material. The following three remaining studies met the inclusion criteria: Bateganya 2005 (Uganda);(1) Bunnell 2006 (Uganda);(2) and Moatti 2003 (Côte d'Ivoire).(3) Bateganya and Moatti were cross-sectional studies comparing patients who had received ART to those who had not, and Bunnell was a before/after study following a cohort of patients initiating ART and examining behavioral outcomes six months later. The overall rigor of these studies was low.

Sexual abstinence: In all three studies, approximately half of the participants reported practicing sexual abstinence. In the two cross-sectional studies, these high rates of sexual abstinence were reported for all HIV-infected patients in the previous six months, regardless of ART status. In the before/after study, the percentage of ART patients who were sexually abstinent in the previous three months did not change from baseline to follow-up.

Frequency of sexual intercourse: Frequency of sexual intercourse was not different between ART and non-ART patients in the one cross-sectional study that measured it. It also did not change from baseline to follow-up in the before/after study. Condom use: Condom use was significantly higher among ART patients compared to non-ART patients in both cross-sectional studies. Moatti found that condom use at last sexual intercourse, regardless of partner type, was 80.2% for ART patients vs. 58.8% for non-ART patients (p<0.001). Bateganya found that condom use at last sexual intercourse with a spouse was 71% for ART patients vs. 47% for non-ART patients (95%CI: 1.7-4.6). In the before/after study, condom use at last sexual intercourse increased significantly from baseline to follow-up among participants with HIV-negative or unknown partners (59% to 82%, 95% CI: 1.7-5.8) and with HIV-infected partners (58% to 74%, 95% CI: 1.4-3.7).

Type of partner: Moatti found that ART patients were more likely to report that the last sexual encounter was with a main partner compared to non-ART patients (95.6% vs. 86.8%, p=0.02); however, Bateganya reported no difference in the unadjusted percent of ART and non-ART patients reporting multiple sexual partners in the last six months (34% vs 35%, UOR 0.96).

Longitudinal risk and HIV transmission: Bunnell found that six months after initiating ART, the number of unprotected sex acts with a partner of known negative or unknown HIV status declined by 70%. Men experienced a 75% reduction (5.4 acts vs. 1.3 acts, p=0.02) and women experienced a 58% reduction (3.5 acts vs. 1.5 acts, p=0.03). Over 85% of unprotected sex acts with a negative or unknown partner occurred within married couples. The estimated risk of HIV transmission to partners of negative or unknown status reduced from 45.7 per 1000 person-years at baseline to 0.9 per 1000 person-years at follow-up, representing a 98% decrease.

STI treatment: One potentially contradictory finding was reported. After controlling for possible confounders, Bateganya found that ART patients were more than twice as likely as non-ART patients to report STI treatment in the past six months (AOR, 2.6; 95% CI, 1.77-3.82). This finding was inconsistent with all other measures of risk behavior in this study, which found that ART patients were less likely to engage in risky behavior.


The authors state that the available evidence indicates a significant reduction in risk behavior associated with ART in developing countries. However, there are few existing studies and the rigor of these studies is weak. More studies are needed to build an evidence base on which to make programmatic and policy decisions.

Quality Rating

Although the overall methodological rigor of the three included studies was very low, the quality of this systematic review is very high, meeting 13 of the 16 criteria on the QUOROM Statement checklist.

In Context

This paper is the first systematic review to examine the impact of medical treatment on HIV risk behavior in developing countries. Its findings are consistent with a 2004 meta-analysis of the relationship between ART and sexual risk behavior in developed countries, which showed that HIV-infected patients receiving ART did not exhibit increased sexual risk behavior.(4) Despite considerable research on this topic in developed countries, and despite calls for this type of research in developing countries,(5) the authors were able to identify only three studies that met the broad inclusion criteria for this review. The lack of studies is a key finding of this review, as there is currently great speculation but limited data about potential increases in risk behaviors after initiation of ART (called treatment optimism or behavioral disinhibition). The dearth of articles found in this review may reflect the relatively recent expansion of ART in developing countries, and more studies may be forthcoming as time allows for documentation of behavioral change.

Programmatic Implications

The conclusion of this review-that ART is not associated with increases in sexual risk behavior among HIV-infected individuals in developing-country settings-must be interpreted with caution. It is based on the results of three studies with poor methodological rigor. The discrepant outcome of increased self-reported STI treatment among ART patients also deserves further investigation to see if it reflects an actual increase in infections, better diagnosis and treatment, or a socially desirable response bias. Additional studies are needed to better assess the relationship between HIV treatment and risky sexual behavior. Controlled trials in this area are not possible, given the unethical nature of withholding ART; however, rigorous studies can still be performed and should address other types of HIV treatment in addition to ART (i.e. opportunistic infection and STI treatment), people of negative and unknown sero-status, and diverse clinical and cultural settings.


  1. Bateganya M, Colfax G, Shafer LA, Kityo C, Mugyenyi P, Serwadda D, et al. Antiretroviral therapy and sexual behavior: a comparative study between antiretroviral- naive and -experienced patients at an urban HIV/AIDS care and research center in Kampala, Uganda. AIDS Patient Care STDS 2005 Nov;19(11):760-8.
  2. Bunnell R, Ekwaru JP, Solberg P, Wamai N, Bikaako-Kajura W, Were W, et al. Changes in sexual behavior and risk of HIV transmission after antiretroviral therapy and prevention interventions in rural Uganda. AIDS 2006 Jan 2;20(1):85-92.
  3. Moatti JP, Prudhomme J, Traore DC, Juillet-Amari A, Akribi HA, Msellati P, et al. Access to antiretroviral treatment and sexual behaviours of HIV-infected patients aware of their serostatus in Cote d'Ivoire. AIDS 2003 Jul;17 Suppl 3:S69-77.
  4. Crepaz N, Hart TA, Marks G. Highly active antiretroviral therapy and sexual risk behavior: a meta-analytic review. JAMA 2004 Jul 14;292(2):224-36.
  5. Valdiserri RO. Mapping the roots of HIV/AIDS complacency: implications for program and policy development. AIDS Educ Prev 2004 Oct;16(5):426-39. (No abstract available.)