Article Text
PDF
Abstract
Objectives The South African government recently launched a national campaign to test 15 million South Africans for HIV by 2011. Little is known about how receipt of HIV testing might influence interpersonal communication. To explore these questions, the authors examined the effects of prior HIV testing on sexual health communication among South Africans.
Methods Adults (N=1284; 98% black, 36% women, mean age 31) residing in a South African township completed street-intercept surveys.
Results Of the 1284 participants, 811 (63%) had been tested for HIV. Among those who had been tested, 77% tested negative, 12% tested positive, and 11% did not know their test result or refused to answer. Compared with those who had not been tested, participants who had been tested for HIV were more likely to communicate with community members about (a) HIV/AIDS, (b) getting tested for HIV, and (c) using condoms. Testing positive for HIV was associated with communication with sexual partners about condom use. Among participants who had been tested for HIV, exploratory analyses revealed that those who had engaged in sexual health communication with community members or sexual partners reported more condom-protected sex than those who had not engaged in sexual health communication.
Conclusions HIV testing is associated with sexual health communication among South African community members and sexual partners. Offering HIV testing to all South Africans may increase communication and lead to reductions in sexual risk.
- HIV testing
- communication
- HIV
- South Africa
- sexual risk
- Africa
- sexual behaviour
Statistics from Altmetric.com
Introduction
South Africa has the highest rate of HIV in the world with an estimated 5.7 million South Africans living with HIV.1 Among South African adults over the age of 15, an estimated 380 000 new infections were diagnosed in 2008. Women are disproportionately affected by HIV, with an estimated 59% of South African women living with HIV.1 2 To reduce the spread of HIV, the WHO and Joint United Nations Programmes on HIV/AIDS (UNAIDS) have joined together to expand access to voluntary HIV counselling and testing (VCT).3
VCT has been shown to be an effective strategy for reducing the spread of HIV by making people aware of their HIV status and promoting safer sex behaviours.4 5 VCT also facilitates earlier access to support services, healthcare and antiretroviral treatment. VCT among South Africans has increased from 47% to 61% between 2006 and 2009.6 Although rates of VCT have increased, many South Africans remain unaware of their HIV status. Barriers to VCT include basic needs such as availability, transportation and childcare, but may also include fears associated with the possibility of discrimination and/or rejection by sexual partners.7–11
In April 2010, the South African government launched a major campaign to increase HIV counselling and testing with an overarching goal of ensuring that every South African is aware of his or her HIV status.12 In support of President Zuma's commitment to nationwide HIV counselling and testing, Michel Sidibé, Executive Director of UNAIDS, suggested that, in addition to increasing awareness of individual HIV status, this campaign will be a catalyst for ‘innumerable conversations…about issues that are sometimes difficult to talk about’ (Sidibe M, p2).13 To date, only a limited number of studies have examined whether HIV testing leads to increased sexual health communications.
Social networks have been shown to have considerable influence on individuals′ perceptions of attitudes, intentions, and behaviours acceptable by a group.14 Most of this research has focused on the negative impact of social networks rather than the positive role of social networks in promoting preventive behaviours, including VCT.15 Communication theory, such as the diffusion of innovations,16 17 suggests that as ‘trend setters’ begin to endorse, model, and discuss HIV prevention, including VCT, within their community, norms shift towards further uptake of the innovation. Recent research across four sites (Tanzania, Zimbabwe, South Africa and Thailand) found that frequent communications regarding HIV-related protection, risk awareness, testing, and assisting people living with HIV/AIDS were associated with HIV testing.18 Moreover, communication about HIV has been shown to increase HIV-related knowledge and reduce HIV/AIDS stigma.19 Thus, a potential benefit of VCT includes increased sexual health communication within a community.
The primary purpose of this study was to examine the effects of HIV testing on sexual health communication among South Africans. We expected that participants who had been tested for HIV would be more likely to communicate with community members about HIV, getting tested, and using condoms and would discuss condom use with sexual partner(s). A secondary purpose was to explore the effects of testing positive for HIV; we expected that those who tested positive would be less likely to engage in sexual health communication because of concerns of stigmatisation associated with a positive test.7 20 21 Examining the potential benefits of VCT on sexual health communications may signify changes regarding sexual health norms among South Africans.
Methods
Participants and procedures
Participants were 1284 residents (823 men, 460 women) of Gugelethu township, a primarily Xhosa-speaking African community just outside Cape Town, South Africa. Field workers used street-intercept methods to sample from six, non-contiguous, neighbourhoods within the township. Two neighbourhoods were surveyed from 3 June to 29 June 2008, two were surveyed from 2 November to 7 December 2008, and two were surveyed from 9 May to 28 June 2009.i Field workers consisted of 12 ethnically matched residents of the communities who spoke both Xhosa and English. Surveys were administered in the participants' preferred language. Field workers approached pedestrians or shebeen patrons and requested their participation in a survey, described as ‘research regarding HIV/AIDS…that may benefit your community’. If individuals were interested, they provided verbal consent and then completed a nine-page anonymous survey in exchange for a small non-monetary gift (a keychain). Surveys were typically self-administered; however, if a prospective participant did not possess adequate literacy skills, the surveys were administered by the field worker as an interview. All procedures were approved by the institutional review boards of the Human Sciences Research Council, University of Connecticut, and Syracuse University.
Measures
Surveys assessed (a) demographic information (eg, age, gender, ethnicity), (b) HIV testing history, (c) sexual health communication and norms, and (d) additional measures (eg, alcohol use) as part of a larger study.
HIV testing history
History of HIV testing was assessed using two items: (a) have you ever been tested for HIV (yes, no), and (b) if you have been tested, what were the results of your most recent test (positive, negative, don't know, refuse to answer). Responses to these two questions were used to categorise people into groups based on whether they had ever been tested for HIV (1=yes, 0=no) and HIV serostatus (1=positive, 0=negative).
Sexual health communication
Communication regarding HIV testing and sexual risk behaviours was assessed using several items. Participants were asked about the last time they talked with community members about (a) HIV/AIDS, (b) getting tested for HIV, and (c) using a condom. Response options were never, yes but not in the past 30 days, and in the past 30 days. To assess communication within the community, we recoded these responses to reflect any communication in the past or present (1) or no communication (0). Communication with sexual partner(s) was assessed by asking participants ‘in the past 30 days, how many times have you talked with a sex partner about using condoms?’ (total number of events).
Communication norms
Norms regarding communication about HIV were assessed using a single item: ‘People in our community can talk openly about HIV/AIDS’. Response options ranged from strongly disagree (1) to strongly agree (4).
Data management and analysis
Summary statistics (means and standard deviations, frequencies) were used to describe sample characteristics and communication for the overall sample. Differences between individuals who had been previously tested or not tested for HIV were examined using a series of univariate logistic regression analyses. Demographic variables that differed significantly between those who had or had not been tested for HIV were included in subsequent analyses. To test our predictions, we used logistic regression analyses to examine the association between prior HIV test and sexual health communication with community members and sexual partners. Specifically, the probability of communication regarding HIV/AIDS, getting tested for HIV, and advising condom use with community members (yes, no) or condom use communication with sexual partner (yes, no) was predicted from prior HIV test (yes, no). Four separate models were used to examine the effects of prior HIV test on each type of communication (HIV/AIDS, getting tested for HIV, and advising condom use with community members; condom communication with sexual partners). Continuous variables were examined for skewness. The variable ‘number of times participant talked with sexual partner about using condoms’ was skewed and clustered at zero, with the variance surrounding the means indicating a high degree of overdispersion; therefore, negative binomial regression analysis was used to model this variable. In order to control the overall type I error rate at the 0.05 level, an adjustment of the p values was performed because of the multiple comparisons (eight). Following procedures presented by Holm,22 we ranked and sequentially adjusted the p values for the number of statistical tests performed. All data analyses were conducted using Stata V.11.0.
Results
Characteristics of the sample
We restricted our analyses to the 1284 (out of 1294) participants who responded to the question regarding HIV testing history. Table 1 reports the overall characteristics of the sample and by HIV testing behaviour. Most (98%) of the participants were black Africans with a mean age of 31 (SD=8.45; IQR=25–36). Some (20%) participants reported being married and had a median of one child (IQR=one to two). Over half were employed (54%), 43% had completed the equivalent of high school, and 12% had attended at least some college.
Characteristics of the sample overall and by prior HIV testing*
Overall, 30% of participants reported having two or more sexual partners in the past 30 days (table 1). Most participants reported having sex with a partner of the opposite sex (82%), 3% (25 men, 12 women) reported sex with a same-sex partner, 2% (19 men, four women) reported having sex with both men and women; 12% (151/1284) of the sample did not have a sexual partner in the past 30 days (data not tabled). Condoms were used for 49% of vaginal or anal sex events. Nearly two-thirds of participants (38%) reported being diagnosed with an STI at least once in their lifetime.
HIV testing history
Of the 1284 participants, 811 (63%) reported having been tested for HIV. Among those who were tested, 623 (77%) were HIV negative, 97 (12%) were HIV positive, 52 (6%) refused to answer, and 39 (5%) did not know their test results. Comparisons between participants who had or had not been tested for HIV indicated that individuals who had been tested for HIV tended to be women (40% vs 29%), with at least a high school education (59% vs 46%), and were currently employed (57% vs 48%). Participants who had been tested for HIV reported fewer sexual partners (27% vs 35%), more condom-protected vaginal or anal sex events (mean=51% vs 45%), and were less likely to report being diagnosed with an STI (19% vs 49%) compared with those who had not been tested. No other differences in demographic or risk variables between those who had or had not been tested for HIV were found (table 1).
Sexual health communication
Participants tended to agree that people in the community can talk openly about HIV/AIDS (mean=3.05, SD=0.99). Many participants communicated with community members about HIV/AIDS (61%), getting tested for HIV (64%), and condom use (75%). Participants reported discussing condom use with their partners on average six times (SD=14.58) in the past month. Bivariate analyses indicated that, compared with those who had not been tested, participants who had been tested for HIV were more likely to communicate with community members about HIV/AIDS (66% vs 53%), getting tested (71% vs 51%), and condom use (80% vs 67%). Prior HIV testing was associated with condom use discussions with sexual partner(s)—that is, participants who had been tested for HIV reported discussing condoms with their sexual partner(s) an average of seven times (versus five times for participants who had not been tested) in the past month (table 1).
Communication by prior HIV testing status
Logistic regression analyses were used to predict the probability of sexual health communication (yes, no) from prior HIV testing (yes, no) controlling for demographic variables that distinguished the testing groups (gender, education and employment) and communication norms (table 2). Prior HIV testing was significantly associated with communication among community members about HIV/AIDS (adjusted OR=1.57, 95% CI=1.23 to 2.00), talking with others about getting tested for HIV (adjusted OR=2.24, 95% CI=1.75 to 2.86), and advising someone to use a condom (adjusted OR=1.86, 95% CI=1.42 to 2.44). As shown in figure 1, participants who had been previously tested for HIV were more likely to discuss HIV/AIDS, getting tested for HIV, and condom use with community members compared with those who had never been tested for HIV.
Regression analyses examining the effects of HIV testing or HIV serostatus on sexual risk reduction communication with community members and sexual partner(s)*
Proportion of participants communicating with community members regarding HIV/AIDS, getting tested for HIV, and using condoms as a function of HIV testing history.
To examine communication with sexual partners, we used a negative binomial regression model to predict the probability of communicating with partners about condom use based on prior HIV testing (yes, no) and controlling for gender, education, employment and communication norms (table 2). Prior HIV testing was associated with partner's communication about condom use (incident RR=1.40, 95% CI=1.05 to 1.86), but this did not reach significance after adjustment of the p value for type I error.
Communication by HIV serostatus
Regression analyses were used to examine sexual risk reduction communication among those who had tested positive or negative for HIV (n=720). As shown in table 2, logistic regression analyses evaluated the association between sexual health communication among community members and HIV serostatus after gender, education, employment and communication norms had been controlled for. Testing positive for HIV was not associated with communication about HIV/AIDS (adjusted OR=1.91, 95% CI=1.15 to 3.19) and talking with others about getting tested for HIV (adjusted OR=1.73, 95% CI=1.01 to 2.95) after adjustment of the p value for type I error. HIV-positive participants were more likely to advise community members to use condoms compared with those who were HIV-negative (87% vs 80%; adjusted OR=1.80, 95% CI=0.96 to 3.81); these results were not statistically significant. Negative binomial regression analyses showed that participants testing positive for HIV were more likely to discuss condom use with their sexual partner (incident RR=2.12, 95% CI=1.42 to 3.18). As shown in figure 2, participants diagnosed with HIV discussed using condoms with a sexual partner 11 times (SD=22.29) compared with five times among those who had tested negative for HIV (SD=10.43).
Mean number of discussions about condom use with sexual partners by HIV serostatus.
Exploratory analyses
Exploratory analyses examined whether participants who communicated with community members or sexual partners about sexual risk reduction engaged in more protected sex. After gender, education, employment, communication norms and HIV testing history had been controlled for, analyses show that participants who communicated with community members about HIV/AIDS (β=0.12, p=0.001), getting tested for HIV (β=0.11, p=0.001), and advising condom use (β=0.20, p=0.001) reported more condom-protected events than those who did not communicate with community members. Participants who had discussed condom use with sexual partners were more likely to use condoms than those who had not after demographics, norms and HIV testing history had been controlled for (β=0.20, p=0.001).
Discussion
In the country with the highest per capita rate of HIV in the world, we examined the association between VCT and sexual health communication. Two primary findings emerged. First, history of HIV testing was significantly associated with sexual health communication among community members. Second, testing positive for HIV was also associated with sexual health communication with sexual partner(s). These two findings suggest that VCT provides a secondary benefit, namely, a mechanism to increase interpersonal communication about HIV, getting tested, and promoting protective behaviours.
The findings corroborate prior research showing that HIV testing is associated with increased sexual health communication among community members.18 The cross-sectional nature of these studies prevents strong inferences regarding the directionality of the association between HIV testing and communication. It is possible that increased interpersonal communication prompts individuals to seek HIV testing. Nonetheless, consistent with diffusion of innovation theory,16 we suggest that the ‘trend setters’ are those who seek to improve their sexual health, get tested for HIV, and subsequently discuss sexual health with community and sexual partner(s). Prior research also demonstrates that people who communicate about behaviour change to others are more likely to adopt risk reduction behaviours themselves,23 consistent with cognitive dissonance theory.24 Future research might examine HIV testing behaviour longitudinally to determine the sequence of testing (including the context and motivations associated with obtaining an HIV test) and sexual health communications.
We extended prior research18 19 by demonstrating a consistent effect of testing across specific communication topics (ie, HIV, getting tested, and condom use) among community members. We also examined the association between communication and HIV test result (ie, comparing individuals who tested positive or negative). These findings have implications in the development of sexual risk reduction interventions. Interventionists seeking to improve interpersonal sexual health communication should consider including VCT as part of their sexual risk reduction intervention. Interventionists should also consider providing interpersonal skills training to facilitate communication with community members and sexual partner(s).
Several factors are important in interpreting these findings. First, data were gathered from self-reports, which are imperfect indicators of behaviour. To minimise socially desirable responding,25 field staff emphasised that all data were anonymous. Second, the respondents sampled may not be representative of the larger community; however, we used street-intercept surveys, a methodology that provides access to the widest range of community members and minimises the self-selection bias often seen with clinical or other recruitment strategies.26 Third, a number of HIV-related studies have been conducted in Gugulethu township; thus, our sample may have been exposed to previous public health campaigns that may have affected their serological testing and communication behaviours. Fourth, our data are cross-sectional and do not support causal inferences. Longitudinal data can provide strong evidence of directionality of influences between HIV testing and communication.
Conclusions
One of the South African government's new strategic goals is to reduce vulnerability to HIV by ‘establishing a national culture in which all people in South Africa regularly seek voluntary testing and counselling for HIV’ (South African Government, p57).27 Our study suggests that uptake of VCT could have an added benefit of increasing sexual health communication among South Africans. Despite the continued rise in South Africa's HIV epidemic, the South African government's decision to provide HIV counselling and testing to 15 million individuals by 2011 may prove to be a turning point in the fight against HIV by influencing communication norms associated with HIV in South Africa.
Key messages
HIV testing is associated with sexual health communication among South Africans.
Testing positive for HIV is associated with sexual health communication with both community members and sexual partners.
Offering HIV testing to all South Africans may increase sexual health communication and lead to reductions in sexual risk behaviours and subsequent infections.
Acknowledgments
We thank Regina Mlobeli for her contributions to this project.
References
Footnotes
Funding This project was supported by the National Institute on Alcohol Abuse and Alcoholism (R01-AA017399 to SCK).
Competing interests None.
Patient consent Obtained.
Ethics approval This study was conducted with the approval of the institutional review boards of the Human Sciences Research Council, University of Connecticut, and Syracuse University.
Provenance and peer review Not commissioned; externally peer reviewed.
↵i Data collection for this study occurred before President's Zuma's launch of the national HIV counselling and testing programme in South Africa.