The prevention of risk practices and maintenance of safe behaviors continues to be the basic pillar in curbing the spread of HIV. HIV-negative individuals with a stable HIV-infected sexual partner have high risk for acquiring HIV, in some communities equalling or surpassing the risk of men who have sex with men or that of female sex workers.1–3
The time at which HIV testing is performed provides a good opportunity to give information and advice on prevention. Moreover, knowing their HIV status can make people more receptive to messages about prevention and risk reduction.4 Previous studies suggest that HIV testing produces a reduction in risk behavior in individuals with HIV infection and in serodiscordant couples.5 Counseling combined with testing may be more effective when both members of the couple participate6 and when one or both members of the couple are infected with HIV.7 This effect is not so clear in seronegative individuals8 or in those who have been tested several times.9
A number of evaluations has been published on the efficacy of interventions to prevent sexual risk behavior and sexually transmitted diseases (STD) in different populations10–14; however, few studies have evaluated health interventions that focus on reducing risk practices in serodiscordant couples in developed countries,6,15,16 and practically none of them refer to the era of highly effective antiretroviral treatments (HAART).
The objective of this study was to describe changes in sexual risk behavior among HIV serodiscordant steady heterosexual couples participating in an HIV counseling and testing program in Madrid over the period 1989 to 2007.
PATIENTS AND METHODS
Setting and Description of the Program
The study was conducted in a multidisciplinary clinic in Madrid, Spain, which provides HIV counseling and diagnosis as well as diagnosis and treatment of other STDs; the care is free, provided “on demand” and can be received anonymously. Since 1989, there has been a program of counseling and testing focusing on heterosexual couples in which 1 member has been diagnosed with HIV infection (index case). From its start until the end of 2007, 909 couples have attended this program. The program includes comprehensive medical consultation for both members of the couple, with checkups scheduled every 6 months. At each visit, clinical follow-up and testing of the index case (CD4, viral load) is conducted, the HIV test is recommended to the partner, STDs are ruled out in both members, and the woman is offered a gynecological examination. An interview with both members of the couple before the test includes a review of the history of HIV risk behavior and use of condoms in sexual relations. Before and after each test, information is given on the risks associated with different practices and advice on alternatives to avoid or lessen this risk, and free condoms are offered. In response to an increasing number of consultations on reproductive issues, information is provided on options for natural or assisted insemination and their associated risks. When the index case meets criteria for antiretroviral treatment, he/she is referred to a reference hospital. Two practitioners were responsible for this program and performed both the interviews and the counseling. The same 2 people remained in the jobs throughout the study period.
Study Subjects and Variables
The study units were heterosexual couples treated in the program between 1989 and 2007, who had maintained an uninterrupted relationship lasting over 6 months, in which one of the members was diagnosed with HIV infection (index case) and the nonindex partner was HIV negative at the initial test. To reduce the complexity of the risk exposure assessment, this study was limited to couples whose nonindex partner had not had sexual contacts with another partner, had not injected drugs, and had not any other type of exposure involving possible HIV risk. We included all eligible couples who returned for at least 1 follow-up visit in the program and who agreed to participate in the study. Couples where index case had other sexual partners were not excluded. We imposed these strict selection criteria as we aimed to describe the effect of an intervention which focused on the 2 members of the couple. A third person who could be acting as a potential exposure would have introduced additional problems when interpreting the results.
Both in first and successive visits, a practitioner used a structured questionnaire to collect epidemiologic and sexual behavior information, as well as data concerning other parenteral or sexual risk exposures. Individuals were explicitly asked about the frequencies of practices of each type of sexual intercourse (vaginal, anal, and oral) during the previous 6 months. For each type of practice, the number of protected (with condom) and unprotected contacts was estimated, based on a semiquantitative scale by asking about condom use and classified as “never: 0; less than a half of the times: 0.33; more than a half of the times: 0.66; and always: 1.” Subjects were also asked if breaking or slipping of condom had ever occurred during penetration. “Couples without risk practices” were considered to be those who had not had either vaginal or anal coitus or who had always used a condom in such relations.
The clinical, immunologic, and virologic status of the index case was analyzed. The partner was tested every 6 months to determine serum antibodies to HIV by the enzyme-linked immunoadsorbent assay technique, with reactive samples confirmed by Western Blot.
Follow-up was considered to begin on the date of the first negative HIV test within the program, and the endpoints were HIV seroconversion, failure to return for checkups during more than 18 months, and breakup of the couple. For the remaining couples, the follow-up period was calculated up to the last checkup before December 31, 2007 (administrative censoring date).
A descriptive analysis was made of the characteristics of all the couples that fulfilled the inclusion criteria at the first visit (n = 564). Among those who returned for successive visits and continued to meet the inclusion criteria (n = 399), we calculated the percentage in whom HIV transmission had occurred and the transmission rate, taking as the denominator the number of couple-years of follow-up.
To evaluate the behavior modification achieved by the program, we compared the practices reported in the first visit (preintervention) with those reported in the second visit (postintervention). We also conducted this analysis separately for couples whose index case had been diagnosed of HIV infection more than 6 months before the first visit and for the rest. For the 143 couples who remained in the program for at least 6 visits, we evaluated the change in behavior between the second and sixth visit.
Proportions were compared using the χ2 test and Fisher exact test, and means were compared using Student t test. To compare condom use and sexual practices between first and successive visits, we used McNemar χ2 for proportions for related samples, and Wilcoxon matched pairs test for median of sexual relations and sexual risk practices.
Nonconditional logistic regression techniques were applied to identify the independent effect of each covariate on the presence of risk behaviors after the intervention (second visit). The effect was measured by the odds ratio (OR) with its 95% confidence interval (CI).
Between 1989 and 2007, 564 steady heterosexual serodiscordant couples who fulfilled the inclusion criteria were seen for a first visit. In 82.3% (n = 464) of cases, the nonindex partner was female and the index case was male. The mean age of nonindex partners was 29.4 years (standard deviation [SD]: 7.2) for women and 32.9 years (SD: 7.7) for men, while for index cases it was 32.3 years (SD: 6.7) for men and 30.3 years (SD: 6.7) for women.
About 55.2% (n = 311) of the nonindex partners had secondary or higher education versus 28.2% (n = 159) of the index cases (P <0.0001). Some 27.5% (n = 155) had been in the couple relationship for less than 1 year at the time of study inclusion. In 54.1% (n = 305) of the couples, the index case had been diagnosed with HIV infection after beginning the couple relationship, and 60.3% (n = 340) had been diagnosed for over a year at study entry. At the time of the first visit, 30.3% (n = 171) of the index cases were receiving antiretroviral treatment and, of these, 81.3% (n = 139) were receiving HAART. About the clinical characteristics of the index case, 56.7% had CD4 counts over 350 cells/mm3 (data available in 457 cases) and the median HIV blood viral load was 405 copies/mL, although this information was only available in 212 subjects, beginning in 1997. Some 60.4% (n = 324) were in the asymptomatic stage.
Of the 564 couples, 399 (70.7%) returned for a second visit. Of the remaining 165 couples, 153 did not return and the reason was unknown, and the other 12 had either ended their relationship or had sexual contacts with other partners. Some characteristics of the couples who returned differed from those who did not: they were younger, had higher educational level, had been in the relationship for a shorter time, had used condoms in a greater percentage in their last sexual relation, and reported a higher frequency of sexual relations. In addition, the index case had been diagnosed more recently and was less frequently receiving antiretroviral treatment (Table 1).
Comparison of Preintervention and Postintervention Sexual Behavior
For the 399 couples who returned for a second visit, we compared behavior in the 6 months before counseling and testing with behavior in the 6 months before the second visit. A significant decrease was seen in total number of coital relations that mainly evolved highest sexually active couples; the interquartile range (IQR) of coital relations decreased from 24–84 (median 48) to 24–72 (median 48) (P = 0.001). The number of couples who had not had coital sexual relations with their partner increased from 4 to 7. The median of sexual risk practices in the previous 6 months dropped from 2.6 (IQR: 0–31.7) to 0 (IQR: 0–11.1) (P <0.001) (Table 2).
Vaginal coitus was the sexual practice reported by most couples, and this percentage did not change significantly between the first and second visit. However, vaginal coitus during menstruation declined from 46.4% to 29.1% (P <0.0001), and anal coitus dropped from 21.6% to 12.8% (P <0.0001).
In couples who engaged in vaginal coitus, systematic use of condoms (100% of coitus) increased from 49.4% to 68.9%. An increase in systematic condom use was also observed in couples who practiced vaginal coitus during menstruation (48.1%–72.4%, P <0.0001) and anal coitus (39.5%–72.5%, P <0.0001). The percentage of couples who reported having used a condom in the last sexual relation increased from 78.4% to 87.7% (P <0.001). In the first visit, 46.1% of couples had no sexual risk practices, a percentage which rose to 66.7% at the second visit (P <0.0001). Of the couples who had used a condom, the percentage of those who had had an accident (breaking or slipping of condom) during use decreased significantly after the intervention (Table 2).
In 330 of the 399 couples (82.7%), the index case was male. The changes in behavior between the first and second visit were very similar when couples were analyzed separately by sex of the index case; however, some changes did not reach statistical significance when the index case was female because of the smaller number of such cases. Among the 138 couples whose first visit took place in the period 1998 to 2007, behavior modification observed at the second visit was similar to what has been described, except that the proportion of those who reported anal coitus did not decrease (22.5%–24.6%, P = 0.629), albeit systematic condom use during this practice was more frequent (64.5% and 76.5%, respectively).
Couples whose index case had been diagnosed with HIV infection more than 6 months before the first visit showed a lower proportion of sexual risk behaviors at the baseline than couples with index case recently diagnosed (45.1% vs. 71.6%; P <0.001). Both groups reduced the frequency of risk practices in the second visit; however, this reduction was more pronounced in the last, and hence the differences among the 2 groups in the percentage of couples with risk practices disappeared (35.5% vs. 29.1%, respectively; P = 0.202) (Table 2).
Comparison of Behavior Between the Second and Sixth Visit
By the end of 2007, 143 couples had a total of at least 6 visits to the program, which represents 462.9 couple-years of follow-up. The changes in behavior between the first and second visit were similar to those already described. These changes were maintained between the second and sixth visit, although hardly any additional reduction in risk behavior was achieved. The only exceptions were that the median of sexual relations in the 6 months before the visit continued to decline in successive visits, from 54.9 to 46.0 (P = 0.001), and the percentage of couples who had vaginal relations during menstruation dropped from 28.6% to 17.5% (P = 0.011). At the second visit, 69.9% of the couples reported no sexual risk behavior compared to 65.7% at the sixth visit (P = 0.471) (Table 3).
Factors Predictive Of Risk Behavior After the Intervention
We analyzed those variables associated with risky sexual practices at second visit. Among the couples who reported sexual risk behavior before the intervention, 56% ceased these practices after intervention, in comparison, among couples who reported no sexual risk behavior before the intervention, 21% reported risk practices after the intervention (McNemar test for repeated observations, P <0.001). Consistent with the associations found in the bivariate analysis, the multivariate analysis showed that sexual risk practices after the intervention were more frequent in those couples who also had them before the intervention (OR: 2.8, 95% CI: 1.7–4.4), in those in which the index case was over age 35 (OR: 2.0, 95% CI: 1.2–3.3) and in women who were pregnant at the time of the second visit (OR: 3.1, 95% CI: 1.6–6.3). In the bivariate analysis, HAART in the index case was also associated with continued risk behavior, but this association disappeared in the multivariate analysis (Table 4).
Transmission of HIV Infection
Of the 399 couples in the cohort, follow-up time ranged between 0.6 and 18.1 year (mean: 3.2 years; median: 2.1 year). During the 1279 couple-years of follow-up, HIV seroconversion occurred in 5 nonindex partners (1.3%). In 4 of 330 couples (1.2%), transmission was from the male index case to the woman, and in 1 of 69 couples (1.4%), it was from woman to man. The HIV seroconversion rate was 3.9 per 1000 couple-years of follow-up (95% CI: 1.4 – 9.7) (Table 5). None of the seroconversions took place in couples in whom the index case was on HAART. Also, none of the women who became pregnant during the study follow-up seroconverted to HIV.
Our results suggest that participation in a program of counseling and testing that specifically targets both members of HIV serodiscordant couples may have resulted in a considerable proportion of couples ceasing sexual risk practices and a decline by half in the overall frequency of unprotected coital relations. Couples in which the index case had been diagnosed with HIV infection before joining the programme presented a lower frequency of sexual risk behavior at baseline suggesting that the HIV diagnosis itself may have already modified their practices. For all couples, however, joining the programme was associated with a reduction of sexual risk behavior, more pronounced in the ones with a recent HIV diagnosis. All these changes were evident after the first visit and were maintained over the various years of follow-up. Together with other factors, it probably helped to keep the HIV transmission rate very low. As in other studies conducted exclusively with seropositive people, counseling and testing programs, using existing clinic staff and implemented during clinical routine, can be effective in reducing unsafe sex.17–19
Our findings have to be interpreted in the context of the limitations inherent to this type of design and to the lack of a control group, as we are evaluating behaviors before and after an intervention. Our study included only couples that elected to join the intervention; however, we do not consider probable that couples with ongoing risk or partners who acquired HIV infection may have selectively dropped out since this clinic is one of more accessible HIV testing sites in Madrid and this cohort was one of the groups of patient with lower HIV seroincidence. The most important bias to be taken into account is that of the social desirability derived from the fact that the counselor was the same person collecting information about behavioral change. Although this information bias may have overestimated the beneficial effect of the intervention, we consider it unlikely to account for most of it, given that the programme involved the same 2 dedicated physicians in a confidential and counseling setting over a long period of time, and involved interviewing the 2 members of the couple. As well as the programme intervention, environmental messages regarding HIV transmission are likely to have had an impact on the observed changes, as well as a decreased sexual activity in the couples attributable to the passage of time. Finally, although all these factors may be overestimating the effect of the intervention, the beneficial effects detected occurred rapidly, between the first and second visit, and not between the second and successive visits, suggesting that they can largely be attributed to the intervention, HIV diagnosis, and the first counseling session. Other studies have also observed that most behavioral change occurs between intake and the first follow-up visit.6,16,20
Despite the ease of access to the program, the couples who participated may not be representative of all HIV serodiscordant couples. Couples who participated in the program probably had a previous awareness of and positive motivation about prevention and are, therefore, a self-selected group. Besides, selection criteria excluded nonindex partners with any risk exposures other than sexual relationship with the index case. We may have excluded those with a higher risk of HIV infection but in order to describe the results of the programme, aimed at the 2 members of the couple, a third person would also introduce additional misclassification of exposure when interpreting the results. All the couples received a first session of counseling and testing, but 27.7% of them did not return for later visits, therefore they could not be included in the evaluation of the intervention. Nevertheless, the level of compliance with the program can be considered good, given that participation was voluntary and the absence of reinforcement mechanisms, which could also introduce biases in the evaluation. Moreover, some of the losses to follow-up could be for justifiable reasons for leaving the study, such as termination of the couple’s relationship or HIV transmission. In any case, in the couples who did not return, the index case was older, had been diagnosed longer with HIV, and was more frequently in antiretroviral treatment, indicating that they were being followed in another health center. Of more concern is the fact that the couples who did not return had a higher frequency of some sexual risk practices. We did not consider the effect of the intervention on the reduction of risk in oral sex practices, since this practice has been associated with a considerably lower risk of HIV transmission than vaginal or anal intercourses.21
The reduced risk in the couples was due in part to decreased frequency of relations, to their avoidance of some practices associated with higher risk, such as sex during menstruation and anal sex, and to increased use of condoms. Various studies have shown that HIV information and counseling for serodiscordant couples produces a substantial increase in condom use. In 2 African studies, in Zambia and Congo, condom use increased from 3% to 80% and from 5% to 71%, respectively, after provision of health information and counseling.6,22 In our case, the couples had a relatively high educational level and nearly half of them were already systematically using condoms at recruitment, thus the improvement achieved by the intervention was more modest.
In a similar study carried out by Padian et al. in the United States,23 about 97% of the change in behavior was seen between the first and second visit; however, consistent condom use during follow-up was more variable. In our study, the behavioral changes achieved were maintained for at least 6 visits. This would indicate that there is no relaxation in prevention measures over time, which is probably reinforced by the prevention message given at each visit.
One-third of the couples engaged in risk behavior after the intervention. This behavior was more frequent among couples who reported risk behavior before the intervention. Older age of the index case was associated with continued risk behavior, which should be kept in mind in prevention activities. The association between risk behavior and pregnancy may indicate that part of the postintervention risk behavior was because the couples were trying to have children. Thus, prevention in serodiscordant couples must respond to their desire for procreation, while minimizing their risks.24,25
The seroconversion rates during follow-up in these couples were low in comparison to those found in other studies of serodiscordant heterosexual couples.26–33 These results are partly attributable to the reduced risk behavior achieved by the intervention. Added to this may have been the influence of various types of prevention and information campaigns that have been carried out in Spain.34 Finally, advances in antiretroviral treatment have managed to considerably reduce the infectivity of the index case, especially since the introduction of HAART.20,35–37
The preventive intervention described in this study seem to have considerably reduced risk behavior but would have not eliminated it entirely. Important changes in risk practices were found to have already occurred after the first visit, therefore special attention should be paid to the couple’s first visit. It is also important that programs for these couples not only offer HIV testing to the seronegative member and follow-up of infection in the seropositive one, but that they also respond to the couples’ concerns about safe sex practices, risk reduction, reproductive counseling, gynecological care, STD screening, access to treatments, and psychological support. The program must create an environment of trust so that couples can resolve all their questions, be aware of the level of risk they are assuming, and know how to change their sexual life to minimize their risk.7
In conclusion, and taking into account the limitations previously discussed, we have shown a significant reduction in risk sexual behaviors in the context of a program of comprehensive care for HIV serodiscordant couples over a 22 years period aimed at HIV serodiscordant heterosexual couples. The programme appears to have been effective in contributing to steadily reducing sexual risk behavior and probably also the risk of HIV transmission, but has not managed to totally eliminate this risk. Diagnosis of HIV-infection of index case and counseling to both members of the serodiscordant couples appear to provide complementary reductions in sexual risk behaviors. The possibilities for improvement detected include earlier participation of couples in such programs, preferably as soon as the relation is started; improving adherence to the program and prevention; reinforcement of preventive messages in order to achieve a greater proportion of couples who cease risk behavior; and investigation of reproductive options, which minimize the risk of transmission.
1. Centro Nacional de Epidemiología. Estudio anónimo no relacionado de la seroprevalencia de VIH en pacientes de consultas de enfermedades de transmisión sexual. 1998–2002. Available at: http://www.isciii.es/htdocs/pdf/anr.pdf.2007
4. Marks G, Crepaz N, Senterfitt JW, et al. Meta-analysis of high-risk sexual behavior in persons aware and unaware they are infected with HIV in the United States: Implications for HIV prevention programs. J Acquir Immun Defic Syndr 2005; 39:446–453.
5. Weinhardt LS, Carey MP, Johnson BT, et al. Effects of HIV counseling and testing on sexual risk behavior: A meta-analytic review of published research: 1985–1997. Am J Public Health 1999; 89:1397–1405.
6. Allen S, Meinzen-Derr J, Kautzman M, et al. Sexual behavior of HIV discordant couples after HIV counseling and testing. AIDS 2003; 17:733–740.
7. Efficacy of voluntary HIV-1 counselling and testing in individuals and couples in Kenya, Tanzania, and Trinidad: A randomised trial. The Voluntary HIV-1 Counseling and Testing Efficacy Study Group. Lancet 2000; 356:103–112.
8. Wolitski RJ, MacGowan RJ, Higgins DL, et al. The effects of HIV counseling and testing on risk-related practices and help-seeking behavior. AIDS Educ Prev 1997; 9:52–67.
9. Matovu JK, Gray RH, Kiwanuka N, et al. Repeat voluntary HIV counseling and testing (VCT), sexual risk behavior and HIV incidence in Rakai, Uganda. AIDS Behav 2007; 11:71–78.
10. Crepaz N, Lyles CM, Wolitski RJ, et al. Do prevention interventions reduce HIV risk behaviours among people living with HIV? A meta-analytic review of controlled trials. AIDS 2006; 20:143–157.
11. Elwy AR, Hart GJ, Hawkes S, et al. Effectiveness of interventions to prevent sexually transmitted infections and human immunodeficiency virus in heterosexual men: A systematic review. Arch Intern Med 2002; 162:1818–1830.
12. Stephenson JM, Imrie J, Sutton SR. Rigorous trials of sexual behaviour interventions in STD/HIV prevention: What can we learn from them? AIDS 2000; 14(suppl 3):S115–S124.
13. Herbst JH, Sherba RT, Crepaz N, et al. A meta-analytic review of HIV behavioral interventions for reducing sexual risk behavior of men who have sex with men. J Acquir Immun Defic Syndr 2005; 39:228–241.
14. Bentley ME, Spratt K, Shepherd ME, et al. HIV testing and counseling among men attending sexually transmitted disease clinics in Pune, India: Changes in condom use and sexual behavior over time. AIDS 1998; 12:1869–1877.
15. Allen S, Tice J, Van de Perre P, et al. Effect of serotesting with counselling on condom use and seroconversion among HIV discordant couples in Africa. BMJ 1992; 304:1605–1609.
16. Padian NS, O’Brien TR, Chang Y, et al. Prevention of heterosexual transmission of human immunodeficiency virus through couple counseling. J Acquir Immun Defic Syndr 1993; 6:1043–1048.
17. Fisher JD, Fisher WA, Cornman DH, et al. Clinician-delivered intervention during routine clinical care reduces unprotected sexual behavior among HIV-infected patients. J Acquir Immun Defic Syndr 2006; 41:44–52.
18. Gardner LI, Marks G, O’Daniels CM, et al. Implementation and evaluation of a clinic-based behavioral intervention: Positive steps for patients with HIV. AIDS Patient Care STDS 2008; 22:627–635.
19. Richardson JL, Milam J, McCutchan A, et al. Effect of brief safer-sex counseling by medical providers to HIV-1 seropositive patients: A multi-clinic assessment. AIDS 2004; 18:1179–1186.
20. Bunnell R, Ekwaru JP, Solberg P, et al. Changes in sexual behavior and risk of HIV transmission after antiretroviral therapy and prevention interventions in rural Uganda. AIDS 2006; 20:85–92.
21. del Romero J, Marincovich B, Castilla J, et al. Evaluating the risk of HIV transmission through unprotected orogenital sex. AIDS 2002; 16:1296–1297.
22. Kamenga M, Ryder RW, Jingu M, et al. Evidence of marked sexual behavior change associated with low HIV-1 seroconversion in 149 married couples with discordant HIV-1 serostatus: Experience at an HIV counselling center in Zaire. AIDS 1991; 5:61–67.
23. Padian NS, Shiboski SC, Glass SO, et al. Heterosexual transmission of human immunodeficiency virus (HIV) in northern California: Results from a ten-year study. Am J Epidemiol 1997; 146:350–357.
24. Barreiro P, Duerr A, Beckerman K, et al. Reproductive options for HIV-serodiscordant couples. AIDS Rev 2006; 8:158–170.
25. Labarga P, Martinez E, Soriano V, et al. Reproductive advice in HIV-discordant couples. Med Clin (Barc) 2007; 129:140–148.
26. European Study Group on heterosexual transmission of HIV. Comparison of female to male and male to female transmission of HIV in 563 stable couples. BMJ 1992; 304:809–813.
27. Carpenter LM, Kamali A, Ruberantwari A, et al. Rates of HIV-1 transmission within marriage in rural Uganda in relation to the HIV sero-status of the partners. AIDS 1999; 13:1083–1089.
28. Downs AM, de Vincenzi I, for European Study Group in Heterosexual Transmission of HIV. Probability of heterosexual transmission of HIV: Relationship to the number of unprotected sexual contacts. J Acquir Immun Defic Syndr Hum Retrovirol 1996; 11:388–395.
29. Fideli US, Allen SA, Musonda R, et al. Virologic and immunologic determinants of heterosexual transmission of human immunodeficiency virus type 1 in Africa. AIDS Res Hum Retroviruses 2001; 17:901–910.
30. Gray RH, Wawer MJ, Brookmeyer R, et al. Probability of HIV-1 transmission per coital act in monogamous, heterosexual, HIV-1-discordant couples in Rakai, Uganda. Lancet 2001; 357:1149–1153.
31. Melo MG, Santos BR, de Cassia LR, et al. Sexual transmission of HIV-1 among serodiscordant couples in Porto Alegre, Southern Brazil. Sex Transm Dis 2008.
32. Nicolosi A, Correa Leite ML, Musicco M, et al; Italian Study Group on HIV heterosexual transmission. The efficiency of male-to-female and female-to-male sexual transmission of the human immunodeficiency virus: A study of 730 stable couples. Epidemiology 1994; 5:570–575.
33. Quinn TC, Wawer MJ, Sewankambo N, et al; Rakai Project Study Group. Viral load and heterosexual transmission of human immunodeficiency virus type 1. N Engl J Med 2000; 342:921–929.
35. Castilla J, del Romero J, Hernando V, et al. Effectiveness of highly active antiretroviral therapy in reducing heterosexual transmission of HIV. J Acquir Immun Defic Syndr 2005; 40:96–101.
36. Cohen MS, Gay C, Kashuba AD, et al. Narrative review: antiretroviral therapy to prevent the sexual transmission of HIV-1. Ann Intern Med 2007; 146:591–601.
37. Kayitenkore K, Bekan B, Rufagari J, et al. The impact of ART on HIV transmission among HIV serodiscordant couples. Paper presented at: XIV International AIDS Conference; 2007; Toronto.