Hiv Communication between Husbands

In northern Thailand, where substantial male-to-female transmission of HIV has occurred in stable partnerships, the relationships between counseling, communication, and HIVpreventive behaviors in married couples have not been well studied. In a study of HIV incidence among women in northern Thailand, each participant was advised to learn her husband’s HIV-infection status and was asked to bring him for an interview at the final 12-month followup visit. Of the 337 men interviewed, 58% reported having ever had an HIV test. More men reported testing following their wives’ enrollment: 12% in the year prior to enrollment vs 22% during the 1-year study (p < 0.001). In the univariate analysis, men’s HIV testing during the 6 months before being interviewed was associated with communication about HIV testing with their wife and extra marital sex with non-FSW while married. Testing following their wife’s request was the most common reason reported. Agreement between husband’s and wife’s reports was poor for most issues, such as whether HIV-related communication had occurred, but agreement as to whether the husband had ever been tested for HIV was relatively high (kappa = 0.62). However, in the logistic regression analysis, only sex with non-FSW while married remained associated with HIV testing (p = 0.02). The results suggest a relationship between counseling, communication, and husband HIV testing. Better communication by couples may result in more effective use of HIV testing, which is already prevalent in this population, to prevent HIV transmission. 2000). HIV prevalence was highest in northern Thai male conscripts who reported having had first sexual experience with FSWs more than their counterparts from other regions (Kits ir ipornchai et al , 1998). HIV seroprevalence among these young men in Chiang Rai peaked at 17% in 1992 (Kilmarx et al, 2000). As these cohorts of young HIV-positive men in the early 1990s got married, they consequently transmitted HIV to their wives. This was supported by rapid increase of HIV seroprevalence among primigravidas in Chiang Rai, aged < 24 years, which peaked INTRODUCTION Thailand is one of the Asian countries hardest hit by the AIDS epidemic. Chiang Rai, Thailand’s northernmost province, has among the highest rates of heterosexual HIV infection in Asia, largely due to male patronage of female sex workers (FSWs) (Kilmarx et al, SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH 314 Vol 38 No. 2 March 2007 at 11% in 1994 (Bunnell et al, 1999). Several studies have suggested that most HIV-infected women probably acquired HIV infection from their primary partner (Siriwasin et al, 1998; Xu et al, 2000a). HIV intervention for non-commercial partners is necessary, particularly if evidences continue to show that HIV transmission among this group still occur. In response to the rapid spread of the HIV epidemic among commercial sex partners, Thailand launched a 100% Condom Use Program in 1991. This program successfully reduced HIV transmission between FSWs and their clients (Mastro and Limpakarnjanarat, 1995; Nelson et al, 1996; Rojanapithayakorn and Hanenberg, 1996). Outside the commercial sex setting, condoms are infrequently used (Nagachinta et al, 1997). Low condom use with non-commercial partners was reported among young male conscripts (Van Landingham et al, 1993; Kitsiripornchai et al, 1998). In Chiang Rai, only 2% of women reported consistent condom use with steady male partners (Xu et al, 2000). Major barrier to condom use among Thai men was the belief that condoms interfere with sexual pleasure, and condoms are hardly used as a contraceptive method among marital partners in Thailand (Knodel and Pramualratana, 1996). The low condom use among non-commercial partners could have an implication for the HIV epidemic in the future, as this HIV risk behavior may help maintain the epidemic (Kitsiripornchai et al, 1998). The governmental and non-governmental campaigns that coincide with the 100% Condom Use Program often emphasizes non-promiscuity or condom use with FSWs as a mean to protect wives and families (Maticka-Tyndale et al, 1994; Ungphakorn and Sittitrai, 1994; Lyttleton, 1996). It appears that preventing HIV transmission in non-commercial partners has not been as successful as in the commercial sex setting. In addition, there has been relatively little information about risk perception and HIV preventive behaviors among married couples (Nagachinta et al, 1997). The low condom use among noncommercial partners implies that HIV strategies focusing on condom use alone may not be the best HIV prevention option for noncommercial and/ or married partners (Frerichs, 1996). Voluntary testing of HIV among persons with high risk of HIV infection, in combination with condom use in extramarital sex has been advocated as possibly a more practical strategy for HIV prevention among married couples (Knodel and Pramualratana, 1996). With increasing access to anti-retroviral therapy in Thailand, it is also possible that HIV testing may become more appealing than in the past. However, few studies in Thailand to date have explored the prevalence of HIV testing and factors related to it. Studies in other countries have suggested that HIV-related communication; eg, talking about risk of getting HIV from one’s partner, discussing condom use and HIV testing including test results, and asking one’s partner to reduce risk behavior, could decrease risk of transmission between noncommercial heterosexual partners (Moore et al, 1995; van der Straten et al, 1995; Saul et al, 2000). In Thailand, however, cultural norms prevent women to openly discuss matters related to sexuality. HIV risk behaviors such as husband’s extramarital sex were never directly and openly discussed among husbands and wives, as women need to maintain the norms of discretion and silence (Ford and Kittisuksathit, 1994; Maticka-Tyndale et al, 1994; Havanon, 1996). There is some evidence that discussion of sexual matters among men and women who are acquainted is possible with appropriate, cultural sensitive intervention (Maticka-Tyndale et al, 1994; Cash et al, 1997). In this study, we determined whether women would be able to talk to their husbands about HIV risk reduction, following our counseling on preventive behaviors. We also determined whether HIV COMMUNICATION BETWEEN HUSBANDS AND WIVES Vol 38 No. 2 March 2007 315 communication has effect on HIV preventive behaviors with a focus on male HIV testing. We interviewed both husbands and wives regarding HIV communication and HIV preventive behaviors and assessed agreement within couples. We also explored the factors related to HIV testing in men. We hypothesize that couples who communicated about HIV infection would be more likely to take HIV preventive action such as HIV testing of the male partner. Information from this study can help shed light into areas of HIV intervention that are appropriate for non-commercial partners in Thailand. MATERIALS AND METHODS This cross-sect ional study was a substudy from a prospective cohort of HIV incidence among 779 HIV-negative women conducted in the north of Thailand in 1998. Women aged 16-45 years were recruited from public family planning clinics and a public hospital postpartum ward by our study nurse. Women were offered enrollment if they met the following eligibility criteria: age 16-45 years, spoke and understood spoken Thai well, and planned to live in Chiang Rai for the following 12 months. HIV-positive women were not followed. Factors related to enrollment were described in Xu et al (2000). In the prospective study, women came for follow-up visits at 6 and 12 months, during which time they received HIV counseling and testing. Pre-test and post-test HIV counseling was provided by trained study nurse-counselors at enrollment and at each follow-up visit. Counseling sessions lasted for about 20 to 45 minutes. Counsel ing messages were tai lored to each woman’s own HIV risk profile based on her risk behavior and her partner’s HIV-infection status and risk behavior. All women were encouraged to talk to her husband about possible risk of getting HIV from him, discuss condom use, and learn her husband’s HIV infection status if not yet known. Condom use was recommended if the husband HIV status was positive. Condom use and partner testing were recommended if the husband’s recent HIV status was unknown. The nurse demonstrated condom use and condoms were given out to women. Reimbursement for husband’s HIV testing was offered. Results of the prospective cohort showed that HIV prevalence of women at baseline was at 3.1%. Most women had a low personal risk profile and HIV-positive women were likely infected by their husbands. HIV seroconversion rate was low; only 1 woman seroconverted during follow-up. Detail results of the prospective study have been reported elsewhere (Xu et al, 2000, 2002). For our substudy, each woman was asked to bring her husband to the 12-month interview. Of the 704 women who came for the 12-month follow-up visit and who were living with their husband at the time, 337 women (48%) brought their husband (Fig 1). Husbands and wives were interviewed separately. Using a structured questionnaire, trained female research nurses fluent in northern Thai language conducted face-to-face interviews. Questions covered sociodemographic characteristics, recent communication about HIV, sexual and HIV preventive behaviors, and perception of HIV risk and HIV prevention responsibility. Interviews lasted about 45 minutes, and questions were similar for both partners. After the face-to-face interview was completed, the study nurse left the respondents alone in the interview room to complete a self-administered questionnaire concerning issues related to extramarital sex. Men and women were reimbursed 200 Thai baht each (~US$ 5.2) for their time and travel. All participants provided written informed consent. The study protocol was approved by the Ethical Review Committee of the Thai Ministry of Public Health and by an Institutional Review Board of the US Centers for Disease Control and Prevention (CDC). SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH 316 Vol 38 No. 2 March 2007 Data analysis Data were double-entered and validated using EpiInfo, Version 6.04 (CDC, Atlanta, GA). Data analysis were performed in SAS, Version 6 (SAS Institute Inc, Cary, NC). We matched the responses from the men with those from their wives and used the unweighted kappa statistic to assess the rate of agreement. Kappa measures the excess of agreement between responses over the level of agreement that would have been obtained by chance alone. The kappa coefficient will equal 1 when agreement is perfect; whereas 0 means that agreement would be expected by chance. We used conventional criteria to describe the level of agreement when interpreting kappas: <0.40 (poor), 0.40-0.59 (fair), 0.60-0.80 (good), and >0.80 (excellent) (Cohen, 1960; Cicchetti and Feinstein, 1990; Feinstein and Cicchetti, 1990). We examined factors associating with HIV testing in men 6 months prior to the interview using chi-squared tests. Factors that were associated univariately at p<0.1 were then included in a stepwise multiple logistic regression model to assess the factors adjusted for one another.