Men and women in developing countries who are offered personalized voluntary HIV counseling and testing are likely to avail themselves of those services, and they subsequently reduce their practice of risky behaviors more than their peers who receive HIV education in a group setting, according to findings from a multicenter study.1 The proportion of participants who had unprotected intercourse with someone other than a primary partner dropped by about 35-39% among men and women receiving individualized services, but by only 17% among those receiving group education; patterns of behavior change differed by participants' HIV infection status. A related analysis demonstrates that voluntary counseling and testing is highly cost-effective, particularly when offered to couples and to HIV-infected people.2 Together, these studies bolster the case for personalized HIV prevention services in developing countries, a strategy that has been opposed because of questions about cost, logistics, and whether people are motivated to learn their HIV status and to modify their behavior.
Effectiveness
The effectiveness study was conducted at clinics in Kenya, Tanzania and Trinidad in 1995-1998, and was open to men and women who were at least 18 years old and were not known to be infected with HIV. (The prevalence of infection is estimated at 8-13% in Kenya, 10-12% in Tanzania and 1-2% in Trinidad.) Participants could enroll individually or with a partner, although members of enrolled couples were interviewed individually. In all, 3,120 individuals and 586 couples enrolled. On average, participants were in their late 20s or (in the case of men enrolled as part of a couple) early 30s, and the majority had had at least a primary education. Participants who enrolled as individuals were predominantly unmarried, whereas about two-thirds of couples were married.
Upon enrollment, participants completed a baseline survey that asked about their sexual behavior in the previous two months. Their last five partners were all categorized by type: For men and women enrolled individually, partners were classified as primary (spouses, boyfriends and girlfriends) or nonprimary. For participants enrolled as part of a couple, the partner who was also in the study was distinguished from all other partners.
Participants were randomly assigned to receive either personalized counseling and testing or a group education intervention. (Half of participants were assigned to each intervention, and the two groups had similar background profiles.) Personalized counseling included risk assessment and the development of a risk reduction plan, tailored to participants' needs. Those receiving this intervention were offered HIV testing and, if they agreed to be tested, were asked to return for the results two weeks later. By contrast, men and women in the group education intervention watched a 15-minute video and participated in a discussion about HIV transmission and condom use. All participants were given 25 condoms and a brochure illustrating how to use them, and were invited to return for more condoms at any time.
At an initial follow-up visit (which occurred an average of 7.3 months after randomization), participants were interviewed about their sexual behavior in the previous two months and were tested (and, if necessary, treated) for sexually transmitted diseases (STDs). All participants who had had an HIV test were offered a retest, and those in the group intervention were offered personalized counseling and testing. At a second follow-up visit (occurring, on average, 13.9 months after randomization), participants were reinterviewed and were offered an HIV retest.
Virtually all participants who received personalized services agreed to take an HIV test, and most (roughly 60-90% of both men and women) returned to get their results. Furthermore, most of those who made a follow-up visit chose to be retested at that time. Similarly, the vast majority of participants in the group intervention opted for HIV testing when they made their first follow-up visit, and most of these (about 70-90%) returned for the results. In both groups, the rate of testing dropped at the second follow-up.
In the baseline interview, 31% of men who enrolled as individuals in the counseling and testing intervention reported that they had recently had unprotected sex with a nonprimary partner; by the first follow-up visit, the proportion had fallen to 20% (a 35% decline, which was statistically significant). Among men in the group intervention, by contrast, the proportion reporting this behavior did not change. Women in both interventions reported a significant decrease in the likelihood of having unprotected sex with a nonprimary partner, but the change was larger among those who had received personalized services--a 39% drop, from 23% to 14%--than among those in the group intervention--a 17% reduction, from 24% to 20%. For both men and women, further reductions occurred between the first and second follow-up visits.
Among participants enrolled as part of a couple, rates of unprotected intercourse with a partner other than the one in the study did not change between the baseline and follow-up visits for either intervention group. However, unprotected sex with the partner in the study declined significantly over time; again, participants in the counseling and testing intervention reported greater declines (24% for men and 21% for women) than did those who re- ceived group education (15% and 18%, respectively).
Data on STDs acquired between baseline and the first follow-up visit were available for 89% of participants enrolled as individuals and for 85% of those enrolled as members of couples. By applying logistic regression techniques to these data, the investigators calculated that individuals who had had unprotected intercourse with a nonprimary partner were twice as likely to have contracted an STD as were those who had not engaged in this behavior (odds ratio, 2.2). In addition, women who had had unprotected sex with a primary partner had an elevated STD risk (1.9). For couples, unprotected sex with a partner not in the study tripled the risk of STD infection (3.0).
A final set of calculations revealed that for men who received personalized services, the proportions reporting unprotected intercourse with primary and nonprimary partners at the first follow-up visit were lower among those who had tested positive for HIV when they entered the study than among those who had not been infected. For women, however, an HIV diagnosis was associated with a reduction in intercourse with primary partners only. Couples in which one or both members had tested positive for HIV at baseline were significantly less likely to have engaged in unprotected intercourse than were couples in which neither member had been infected. HIV-infected individuals and participants who had unprotected sex with a nonprimary partner had elevated odds of STD infection at the first follow-up (odds ratios, 3.2 and 2.7, respectively).
Cost-Effectiveness
Using results of the effectiveness study and cost information supplied by the clinics, the researchers assessed the cost-effectiveness of providing voluntary HIV counseling and testing in Kenya and Tanzania. (Trinidad was excluded from this analysis because it differs greatly from the African countries in terms of HIV prevalence and economic factors.) They estimated that for every 10,000 men and women obtaining personalized services, 1,104 infections would be prevented in Kenya and 895 would be averted in Tanzania during the following year. Costs per infection averted would total US$249 and $346, respectively; the investigators point out that the costs of treating an infected individual, "not to mention the noneconomic costs in suffering and social impact on families and communities," would likely be substantially higher.
The researchers also estimated the costs of counseling and testing per disability- adjusted (i.e., fully productive, healthy) life-year saved. These costs would average $12.77 in Kenya and $17.78 in Tanzania (totals that compare favorably with those of other public health interventions); they would be lowest for men and women who enrolled in the intervention with a partner and for HIV-infected persons. Under alternative assumptions about HIV prevalence and program costs and effectiveness, the total estimated cost per healthy life-year saved ranged from $5.16 to $27.36 in Kenya and from $6.58 to $45.03 in Tanzania. According to the investigators, $50 is a recommended threshold for this measure in developing countries; therefore, under even the costliest scenario in Tanzania, the intervention is cost-effective.
Conclusion
Although voluntary HIV counseling and testing has not been widely accepted in developing regions, the researchers urge that it become "part of a standard package of prevention strategies." Echoing this recommendation, the author of a commentary on these studies writes: "The challenge is no longer the need to show the efficacy of [counseling and testing] but to make it accessible to those who desperately need it and to expand it and render it more acceptable, innocuous, and less expensive."3--D. Hollander
References
1. The Voluntary HIV-1 Counseling and Testing Efficacy Study Group, Efficacy of HIV-1 counselling and testing in individuals and couples in Kenya, Tanzania, and Trinidad: a randomised trial, Lancet, 2000, 356(9224): 103-112.
2. Sweat M et al., Cost-effectiveness of voluntary HIV-1 counselling and testing in reducing sexual transmission of HIV-1 in Kenya and Tanzania, Lancet, 2000, 356(9224):113-121.
3. Van de Perre P, HIV voluntary counselling and testing in community health services, commentary, Lancet, 2000, 356(9224):86-87.