AIDS and Behavior

, Volume 12, Issue 4, pp 570–577

High Levels of Risk Behavior Among People Living with HIV Initiating and Waiting to Start Antiretroviral Therapy in Cape Town South Africa

Authors

    • Department of International Health and DevelopmentTulane University School of Public Health and Tropical Medicine
  • Catherine Mathews
    • Health Systems Research UnitMedical Research Council
    • School of Public Health and Family MedicineUniversity of Cape Town
  • Mickey Chopra
    • Health Systems Research UnitMedical Research Council
    • School of Public HealthUniversity of the Western Cape
  • Lisanne Brown
    • Department of International Health and DevelopmentTulane University School of Public Health and Tropical Medicine
  • Eva Silvestre
    • Department of International Health and DevelopmentTulane University School of Public Health and Tropical Medicine
  • Vanessa Daries
    • Health Systems Research UnitMedical Research Council
    • School of Public Health and Family MedicineUniversity of Cape Town
  • Carl Kendall
    • Department of International Health and DevelopmentTulane University School of Public Health and Tropical Medicine
    • College of Health and Human ServicesNew Mexico State University
Original Paper

DOI: 10.1007/s10461-007-9279-7

Cite this article as:
Eisele, T.P., Mathews, C., Chopra, M. et al. AIDS Behav (2008) 12: 570. doi:10.1007/s10461-007-9279-7

Abstract

Baseline data were collected in Cape Town during 2006 to study if patients on combination antiretroviral therapy (ART) experience decreased inhibition to avoid risky sexual behavior. A total of 924 HIV-positive individuals were recruited; 520 who initiated ART within 3 months and 404 waiting for ART. Nearly half of men (40.1%) and women (46.3%) reported having unprotected sex their last time. Men and women who did not disclose their HIV status to their partner [Odds ration (OR) = 2.57 (95% CI: 1.22–5.50) and 2.84 (95% CI: 1.84–4.39), respectively], and those with ambivalent perception about the relationship between ART and HIV transmission [OR = 2.08 (95% CI: 1.00–4.30) and 2.39 (95% CI: 1.50–3.84), respectively], were twice as likely to have had unprotected sex their last time. Results suggest an urgent need to strengthen prevention interventions among HIV-positive individuals on and about to start ART in this setting.

Keywords

HIV/AIDSAntiretroviral therapyRisk behaviorSouth AfricaPeople living with HIV/AIDS

Introduction

South Africa has rapidly scaled up the availability of combination antiretroviral therapy (ART) to patients with AIDS to almost 200,000 in 2006 (Nattrass, 2006). Accordingly, there has been a rapid scale-up of free ART provision within the Western Cape Provence, with 16,234 patients on ART by March 2006 within 43 health facilities (Western Cape Department of Health, 2006). Monthly enrollment of new ART patients per month now exceeds 1,000 in the Western Cape Provence, with most new patients women and most attending 16 ART facilities in the Cape Town area.

However, to date there have been limited data on assessing if HIV-positive individuals on ART in urban areas of resource poor settings experience diminished inhibition to avoid risky sexual behavior (Bunnell et al., 2006). Data from developed countries among high risk groups such as men who have sex with men suggest the possibility that they may experience diminished inhibition to avoid risky sexual behavior after initiating ART, although such data are limited for heterosexual individuals living with HIV/AIDS on ART (Centers for Disease Control and Prevention, 1999, 2003; Crepaz, Hart, & Marks, 2004; Dilley, Woods, & McFarland, 1997; Dodds, Nardone, Mercey, & Johnson, 2000; Katz et al., 2002; Stolte & Dukers, 2003; Yamey, 2001). Any increase in unprotected sex among the HIV-positive population following the introduction of ART may have unintentional consequences on the HIV/AIDS epidemic if not appropriately addressed through prevention interventions (Salomon et al., 2005).

Furthermore, there have also been limited data to this point on the determinants of risky sexual behavior among men and women aware of their HIV-positive status in sub-Saharan Africa, especially comparing those on ART with those not yet on treatment. Such studies among HIV-positive individuals in South Africa have shown determinants of unprotected sex to include: shorter duration of illness, denial of HIV status, alcohol use at sex, nondisclosure of HIV status to sex partner, HIV-positive partner, forced sex and having a casual partner (Kiene et al., 2006; Olley, Seedat, Gxamza, Reuter, & Stein 2005; Simbayi et al., 2007a). The characteristics associated with unprotected sex among HIV-positive populations aware of their status outside sub-Saharan Africa include alcohol/drug use during sex, HIV disclosure, having a casual partner, unknown HIV status of partner and depression (Crepaz and Marks, 2003; Eich-Hochli, Niklowitz, Clement, Luthy, & Opravil, 1998; Kalichman, 1999; Kelly et al., 1993; Kennedy et al., 1993; Marks and Crepaz, 2001).

We recently completed baseline data collection for a prospective cohort study of HIV-positive individuals in Cape Town to assess if patients on ART experience decreased inhibition to avoid risky sexual behavior. We present baseline results of the prevalence of risky sexual behavior among patients just initiating and waiting to initiate ART. In order to help health officials in the Western Cape Provence design effective secondary prevention interventions among the growing population on ART, sociodemographic and behavioral characteristics associated with unprotected sex were also explored among this population.

Methods

Participants

Baseline measures of risky sexual behavior were ascertained from a census of HIV-positive patients who recently initiated ART and from HIV-positive persons not yet on treatment attending five of the largest ART public health facilities in the Cape Town metropolitan area from August–October 2006. The inclusion criteria for ART participants were having initiated ART within the previous 3 months and being at least 18 years old, while the inclusion criteria for non-ART participants were known HIV-positive status for at least 6 months, being at least 18 years old, and not having initiated ART. The majority of the study participants not yet on ART were making one of several visits to the health facility/clinic in preparation for the initiation of ART at some point in the future.

Measures

Following verbal and written consent, measures of health-related quality of life (Jelsma and Ferguson, 2004; Jelsma, Maclean, Hughes, Tinise, & Darder, 2005), beliefs on the effects of ART on HIV transmission risk, stigma around HIV status, sociodemographic characteristics and self-reported sexual behaviors were obtained using specially designed electronic instruments on hand-held computers that allowed respondents to answer sensitive questions in private using a combination of self-administered and interviewer-assisted techniques. Questionnaires were designed using the Personal Data Collection Toolset (PDACT) which was developed to allow the Questionnaire Development System (QDS™; NOVA Research Company) to operate on Palm™ Pilot PDAs. Prior to data collection, the questionnaire was revised based on results from pretesting among a convenience sample of ART patients in Cape Town.

The primary outcome for measuring risky sexual behavior used in this research is the prevalence of unprotected sex during a respondent’s last sexual encounter, categorized by partner type and whether this episode happened in the past 4 weeks. This outcome was chosen as it was inclusive of all respondents, with the majority (61.7%) of such episodes having occurred in the past 4 weeks.

A wealth index representing socioeconomic status was developed based on a principal components analysis of assets possessed by each respondent (Filmer and Pritchett, 2001). The raw factor scores were then categorized at the median value into the lower and upper socioeconomic categories.

Four Likert scale questions addressing the beliefs about the effect of ART availability on HIV transmission and the need for using a condom to reduce the risk of transmitting HIV were grouped into a composite index (Cronbach coefficient alpha raw score = 0.81). An example of one such question is “Because of the availability of ARVs for HIV, I am less concerned about always using a condom when I have sex”. Response categories ranged from 0 to 4 for strongly disagree to strongly agree, respectively, with unsure scored as 2. The composite index was calculated as the mean of the four Likert scale questions combined, dichotomized into a bivariate variable of not being positive (ambivalent perception) that the availability of ART make safe sex unnecessary versus strongly believe safe sex is necessary even with ART availability.

Drawing from previous research on stigma among people living with HIV/AIDS in Cape Town (Simbayi et al., 2007b), a set of 13 Likert scale questions addressing the perception of stigma and HIV were grouped into a composite index (Cronbach coefficient alpha raw score = 0.83). An example of one such question is “I feel that other people might blame me for being HIV positive”, with response categories ranging from not at all to often. The composite index was calculated as the mean of the 13 Likert scale questions combined, dichotomized into a bivariate variable of perception that HIV causes them stigma versus perception there is limited stigma associated with their HIV status.

Data Analysis

Descriptive statistics were used to summarize sociodemographic and sexual/risk characteristics of respondents, with t-test and Chi-square statistics used to assess bivariate differences between those on and waiting to start ART. Logistic regression was used to model the determinants of unprotected sex at last sex, with Wald statistics and log-likelihood ratios used to identify variable significance and model fit with the probability of committing a type-1 error (alpha) set at 0.10. Standard errors were not adjusted for the effect of ART clinic site after it was found that the interclass correlation coefficient for sites was close to 0 for most outcomes, although ART clinic site is adjusted for in all models.

Hypothesized determinants of unprotected sex at last sex were selected based on previous research among similar populations within this setting (Kiene et al., 2006; Olley et al., 2005; Simbayi et al., 2007a), and include age, socioeconomic status (wealth index), disclosure of HIV status to sex partner, beliefs about ART and transmission risk (index), alcohol consumption, perceptions of stigma around HIV, HIV status of sex partner and desire to get pregnant (among women). The primary outcome of unprotected sex at last sexual encounter was modeled separately for all partner types and among partners defined as casual or of unknown status. The length of time since their last sexual encounter, dichotomized as less than versus more than 4 weeks ago, was included in all models to control for recent exposure to HIV prevention messages for those initiating and about to start ART. As it is hypothesized the determinants of unprotected sex would likely differ in effect between men and women, models were stratified a priori by sex.

Results

Overall, 520 HIV patients initiating ART (165 men and 355 women) and 404 HIV patients not yet on ART (150 men and 254 women) were enrolled and completed the questionnaire. While men and women on ART were slightly older than their counterparts not yet on ART, they were similar with respect to race, education and employment (Table 1). Overall, both men and women on ART were better off in terms of socioeconomic status than those not yet on ART (men, χ2 = 12.63, P < 0.01; women, χ2 = 8.22, P < 0.05). Fewer women on ART reported being married or living with a partner compared to their counterparts not-yet on ART (χ2 = 6.51, P < 0.05).
Table 1

Demographic characteristics of respondents

Demographic characteristics

Initiated ARTc ≤3 months

Not yet eligible for ARTc

Total

Test statistic

n

Point estimate (SE)d

N

Point estimate (SE)d

n

Point estimate (SE)d

 

Men

165

 

150

 

315

  

  Mean agea

 

37.8 (0.64)

 

35.8 (0.68)

 

36.9 (0.47)

2.39*

  Percent blackb

 

90.2 (2.34)

 

93.7 (2.04)

 

91.8 (1.57)

0.89

  Mean years of educationa

 

7.5 (0.28)

 

8.4 (0.31)

 

7.9 (0.21)

1.92

  Percent employed at least part timeb

 

41.9 (3.91)

 

41.3 (4.13)

 

41.6 (2.84)

0.67

  Socioeconomic statusb

    Poorest

 

17.8 (3.00)

 

33.3 (3.90)

 

25.2 (2.47)

 

    Middle

 

36.2 (3.78)

 

36.7 (3.99)

 

36.5 (2.74)

 

    Least poor

 

46.0 (3.92)

 

29.9 (3.79)

 

38.4 (2.77)

12.63**

  Currently married/living with partnerb

 

42.9 (3.89)

 

42.0 (4.14)

 

42.4 (2.83)

0.09

Women

355

 

254

 

609

  

  Mean agea

 

32.3 (0.44)

 

31.0 (.042)

 

31.8 (0.31)

2.04*

  Percent blackb

 

94.6 (1.20)

 

93.7 (1.53)

 

94.2 (0.95)

0.24

  Mean years of educationa

 

9.2 (0.17)

 

9.4 (0.18)

 

9.3 (0.12)

0.82

  Percent employed at least part timeb

 

21.6 (2.20)

 

26.5 (2.78)

 

23.6 (1.73)

1.17

  Socioeconomic statusb

    Poorest

 

26.6 (2.35)

 

36.6 (3.03)

 

30.8 (1.87)

 

    Middle

 

31.1 (2.46)

 

30.3 (2.89)

 

30.8 (1.87)

 

    Least poor

 

42.3 (2.63)

 

33.1 (2.96)

 

38.5 (1.97)

8.22*

  Currently married/living with partnerb

 

23.8 (2.27)

 

33.2 (2.97)

 

27.7 (1.82)

6.51*

ART and non-ART respondents significantly different: * P < 0.05; ** P < 0.01

at-Test statistic

bChi-square test statistic

cCombination antiretroviral therapy

dStandard error

Overall, ART and non-ART respondents, within sexes, were similar with respect to their reported sexual behaviors (Table 2). While nearly a third of men (32.7%) and half of women (49.3%) reported having no desire to have sex, well over half the men (59.7%) and women (62.3%) reported at least one sexual partner in the past 4 weeks (Table 2). Fewer men (53.5%) and women (57.7%) on ART reported a sexual partner within the past 4 weeks compared to men (66.7%) and women (70.0%) not yet on ART (men: χ2 = 5.35, P < 0.05; women: χ2 = 9.51, P < 0.01). The vast majority of men and women (89.1%) reported having sex within the past 12 months.
Table 2

Reported sexual characteristics and HIV disclosure

Sexual behaviors

Initiated ARTa ≤3 months

Not yet eligible for ARTa

Total

Chi-square test statistic

n

% (SE)b

n

% (SE)b

n

%(SE)b

 

Men

  No desire to have sex

160

35.0 (3.78)

143

30.1 (3.85)

303

32.7 (2.70)

0.83

  At least one sexual partner in last 4 weeks

157

53.5 (4.00)

141

66.7 (4.00)

298

59.7 (2.85)

5.35*

  No condom at last sex with most recent partner

161

39.8 (3.87)

143

40.6 (4.12)

304

40.1 (2.82)

0.02

    Steady partner

116

32.8 (4.38)

102

35.3 (4.76)

218

33.9 (3.20)

0.20

    Casual partner/unsure of partner type

43

58.1 (7.61)

37

54.1 (8.31)

80

56.3 (5.58)

0.24

  Disclosed HIV status to sex partner

163

70.5 (3.58)

147

80.3 (3.29)

310

75.2 (2.46)

3.63

  No condom last sex, among partners with unknown or negative HIV status

103

39.8 (4.85)

95

44.2 (5.12)

198

41.9 (5.12)

0.39

Women

  No desire to have sex

345

51.3 (2.68)

252

46.4 (3.15)

601

49.3 (2.04)

1.38

  At least one sexual partner in last 4 weeks

352

57.7 (2.64)

250

70.0 (2.90)

602

62.3 (1.97)

9.51**

  No condom at last sex with most recent partner

348

46.0 (2.68)

252

46.8 (3.15)

600

46.3 (2.04)

0.04

    Steady partner

270

43.7 (3.02)

200

46.5 (3.54)

470

45.0 (2.30)

0.36

    Casual partner/unsure of partner type

70

50.0 (6.02)

49

46.9 (7.20)

119

48.7 (4.60)

0.11

  Disclosed HIV status to sex partner

354

65.5 (2.53)

254

69.3 (2.90)

608

67.1 (1.91)

0.93

  No condom last sex, among partners with unknown or negative HIV status

267

47.2 (3.06)

190

50.5 (3.64)

457

48.6 (2.34)

0.49

ART and non-ART respondents significantly different: * P < 0.05; ** P < 0.01

aCombination antiretroviral therapy

bStandard error

Regardless of ART status, four in ten men (mean of 40.1%) reported having unprotected sex the last time they had sex (Table 2). Of these last encounters with a casual partner or where the man was unsure of the partner’s status [27% (80/298)], over half were unprotected (56.3%). This was true for those already on (58.1%) and those waiting to start ART (54.1%). Of those men who did not use a condom during this last sexual encounter, the primary reason was they did not feel it was necessary (50.4%), followed by partner refusal (24.4%). For both men already on and those waiting to start ART, half did not know the HIV status of their last sexual partner (49.7 and 48.9%, respectively). Nearly a third (29.5%) of men on ART and one in five (19.7%) men not on ART had not disclosed their HIV status to their sex partners. Sixteen percent of men reported their partner was currently trying to get pregnant.

Similar to men, nearly half (46.3%) of women also reported unprotected sex at their last sexual encounter (Table 2). Of these last encounters that were with a casual partner or where the woman was unsure of the partner status [20% (119/600)], just under half were unprotected (48.7%), regardless of ART status. Of those women who did not use a condom during this last sexual encounter, the primary reason was partner refusal (48.5%), followed by reporting they did not feel it was necessary (31.5%). The majority of women in both groups did not know the HIV status of their last sexual partner (66.5 and 59.7%, respectively). Over a third of women, irrespective of ART status, had not disclosed their HIV status to their sex partners. Fifteen percent of women reported they are currently trying to get pregnant.

Most ART and non-ART respondents were aware that it is still possible for them to transmit HIV when a condom is not used during sex (91.4 and 83.8%, respectively), that they remain a risk of infecting others (90.6 and 85.1%, respectively) and that they are concerned about always using a condom (95.1 and 84.3%, respectively).

Alcohol use in the previous 4 weeks was lower amongst men (16.0%) and women (10.2%) on ART, than those who were not yet on ART (33.6% men and 18.6% women). While non-response was high (12–25%), reported illegal drug use in the past 4 weeks was low among all groups.

Risk factors associated with unprotected sex at last sex, among steady and casual partners, were similar for both male and female respondents, after controlling for partner type, time since last sexual encounter, ART clinic site and sociodemographic characteristics (Table 3). Both men and women who had not disclosed their HIV status to their sex partner were over twice as likely to have had unprotected sex the last time they had sex. Men and women not positive that the availability of ART makes condom use unnecessary were also twice as likely to have had unprotected sex at their last sexual encounter as compared to those who strongly believe safe sex is necessary even with ART availability. Women who were trying to get pregnant were more likely to have had unprotected sex. Women who consumed alcohol, more than a few sips, within the past month were also significantly more likely to have had unprotected sex during their last encounter.
Table 3

Risk factors of unprotected sex at last sexual intercourse

Determinants

Adjusted odds ratio

95% Confidence interval

Adjusted odds ratio

95% Confidence interval

Men (n = 278)

Women (n = 558)

No disclosure of HIV status to sex partner

2.57**

(1.22–5.50)

2.84***

(1.84–4.39)

Ambivalent perception that availability of ART make safe sex unnecessary vs. strongly feel safe sex is necessary even with availability of ART (reference): index score for four questions

2.08**

(1.00–4.30)

2.39***

(1.50–3.84)

More than primary education (primary/no education reference)

0.67

(0.34–1.32)

0.69

(0.35–1.11)

Upper socioeconomic status (low reference)

1.22

(0.60–2.46)

1.20

(0.80–1.81)

Desire to get pregnant

1.03

(0.40–2.61)

1.76*

(0.96–3.26)

Age in years

    18–24

1.00

1.00

    15–34

1.18

(0.13–10.42)

0.91

(0.53–1.56)

    ≥35

1.37

(0.16–12.02)

1.13

(0.62–2.07)

Alcohol consumption in past month, more than a few sips

0.99

(0.46–2.12)

1.81**

(1.01–3.24)

Casual/unknown partner type (steady partner reference)

1.47

(0.72–3.04)

0.76

(0.47–1.23)

ART in last 3 months (waiting for ART reference)

0.61

(0.31–1.18)

0.92

(0.62–1.38)

Perceive HIV as stigmatizing: index score for 13 questions

1.05

(0.56–2.00)

0.97

(0.65–1.46)

Model controls for time since last sexual encounter (less than vs. more than 1 month ago) and ART facility

Significant at * P < 0.10; ** P < 0.05; *** P < 0.01

Among men whose last sex was with a casual partner, defined as either casual or a partner of unknown status, those who reported the availability of ART makes safe sex unnecessary or who reported being unclear about this were five times more likely to have had unprotected sex, as compared to those who strongly believe safe sex is necessary even with ART availability, after controlling for time since last sexual encounter, ART clinic site and sociodemographic characteristics (Table 4). Among women whose last sex was with a casual partner, those who had not disclosed their HIV status to their sex partner were more than twice as likely to have had unprotected sex, after controlling for the same potential confounders.
Table 4

Risk factors of unprotected sex at last sexual intercourse, among casual or unknown partner type

Determinants

Adjusted odds ratio

95% Confidence interval

Adjusted odds ratio

95% Confidence interval

Men (n = 74)

Women (n = 116)

No disclosure of HIV status to sex partner

2.07

(0.60–7.15)

2.64*

(1.09–6.46)

Ambivalent perception that availability of ART make safe sex unnecessary vs. strongly feel safe sex is necessary even with availability of ART (reference): index score for four questions

5.53*

(1.29–23.71)

1.21

(0.43–3.37)

More than primary education (primary/no education reference)

0.49

(0.12–2.01)

0.36

(0.10–1.28)

Upper socioeconomic status (low reference)

2.61

(0.52–13.04)

2.53*

(1.00–6.37)

Desire to get pregnant

0.69

(0.07–6.52)

2.03

(0.29–14.31)

Age in years

    18–24 (no men in this category)

1.00

    15–34

1.00

1.52

(0.46–5.06)

    ≥35

0.90

(0.26–3.05)

1.03

(0.30–3.56)

Alcohol consumption in past month, more than a few sips

0.36

(0.08–1.67)

0.52

(0.12–2.24)

ART in last 3 months (waiting for ART reference)

0.83

(0.20–3.48)

1.05

(0.44–2.51)

Perceive HIV as stigmatizing: index score for 13 questions

0.91

(0.27–3.02)

0.72

(0.29–1.79)

Model controls for time since last sexual encounter (less than vs. more than 1 month ago) and ART facility

Significant at * P < 0.05

Discussion

Patients on ART in Cape Town are a select population; they report little alcohol or illicit drug use, they have disclosed their HIV status, and they have received at least some counseling concerning the need for safe sex. Yet, even in this population, these results suggest that a high proportion of HIV-positive men and women aware of their HIV status in Cape Town are engaging in unprotected sex. This is true for patients who recently started ART within the past 3 months as well as those not yet eligible for ART.

Two previous studies in South Africa have shown the prevalence of unprotected sex among people living with HIV/AIDS to be as high as 30–50% (Kiene et al., 2006; Olley et al., 2005). Results presented here add to these previous results with more than six times the sample size and resultant statistical power, the inclusion of both ART and non-ART patients and data collection taking place at five of the largest ART facilities in the study area which improves external validity.

The primary reason among men for not using a condom at last sex, cited by just over half, was that they did not feel it was necessary, while the primary reason cited for not using a condom among women, cited by just under half, was partner refusal. These data reinforce the importance of underlying gender inequity within this setting where men are typically in control of negotiating condom use during sex and that when they deem it unnecessary, women typically comply (Preston-Whyte, 1999). Secondary prevention interventions for men within this setting should therefore focus on increasing the awareness and motivation around the need to use a condom during sex, regardless of the availability of ART. In addition to increasing the awareness and motivation to use condoms, these data suggest a clear need for improving the general sexual climate around condom use, as well as the specific skills among women to negotiate the use of a condom during sex with their male partners who may not consider them necessary.

As has been observed previously in a similar population in Cape Town (Simbayi et al., 2007a), these results suggest that not disclosing one’s HIV status to their sexual partner put men and women at a twofold increase of having unprotected sex, after controlling for time since last sexual encounter, partner type, HIV status of partner and sociodemographic characteristics. While the requirements to receive ART in the Western Cape Province include disclosure of one’s HIV status, disclosure to one’s sexual partner is not specifically required. Although disclosure is related to a number of factors, including stigma, an important factor involves a person’s capacity to openly discuss sensitive issues involving sex and relationships, especially among non-steady partners, as well as one’s perception of their own risk (Sullivan, 2005). Nondisclosure as a risk factor for unprotected sex also held true among women whose last sex was with a casual partner. However, while in the hypothesized direction (adjusted odds ratio = 2.07; 95% CI 0.60–7.15), lack of disclosure was not significantly associated with unprotected sex among men whose last sex was with a casual partner. This result is not entirely surprising considering the low statistical power due to a sample size of only 74 men whose last sex was with a casual partner.

Men and women, regardless of their ART status, who were not certain that the availability of ART does not eliminate the risk of HIV transmission were twice as likely to have had unprotected sex their last time, as compared to those men and women who strongly believed this to be true. These data suggest prevention interventions among HIV-positive men and women should focus on increasing the awareness that ART does not eliminate the risk of transmitting HIV and thus condoms should consistently be used to reduce such risk.

These findings should be interpreted within the limitations of this study. First, although it is common practice, reporting on condom use at last sexual intercourse may not always accurately characterize an individual’s overall condom use behavior and HIV transmission risk (Schroder, Carry, & Vanable, 2003; Sheeran and Abraham, 1994). Second, ART facilities were chosen based on the logistics of obtaining the desired sample size and not based on a probability sample of ART facilities in the Cape Town metropolitan area. Nevertheless, the five sites included here covered more than half the ART patients presently on treatment in Cape Town. Third, the study population was limited to those individuals who recently started ART or who were about to start ART, thus they were homogeneous with respect to having low levels of active alcohol/substance abuse and untreated depression, disclosure of HIV status, acceptance of their HIV positive status and insight into the consequences of HIV infection and the role of ART.

These results show great need for an effective risk prevention behavioral intervention for HIV-positive individuals in clinical care for HIV/AIDS in Cape Town. These interventions should focus on building awareness and motivation for practicing safe sex even with the availability of ART. At a minimum this will involve encouraging disclosing one’s HIV status to all sexual partners, and supporting gender appropriate behavioral interventions that recognize the need to address traditional sex roles and gender power differentials.

Acknowledgments

We sincerely thank the men and women in Cape Town who agreed to participate in the study. This research was conducted with financial support from the Centers for Disease Control and Prevention, Atlanta, GA, USA which played no role in the design, analysis or write-up of the study.

Copyright information

© Springer Science+Business Media, LLC 2007