Assessment of Sexual Behaviour of Men in Bangladesh

In Bangladesh, the prevalence of HIV is still low (<1.0%) among the most-at-risk populations. So far, most intervention programmes for HIV prevention have been targeted to these groups. Few data are available on sexual behaviour and sexual networks among the general male population and these are required to understand the potential trends and patterns of HIV transmission. Again in community surveys, answers to sensitive questions on sexual behaviour are often under-reported due to social desirability bias. To address this, a confidential data-collection method is needed for developing countries.

This one-year project (September 2004–August 2005) was designed (1) to understand the sexual behaviour of the general male population (aged 18–49 years) in Bangladesh; (2) to compare the response rates to sensitive questions on sexual behaviour using two different interviewing techniques: (a) asking sensitive questions in a face-to-face interview (FTFI), and (b) administering the sensitive questions by a pre-recorded audio system while collecting the responses using a ballot-box and ballot-slips (slips of paper to provide responses) to maintain confidentiality called the Modified Ballot-Box Method (MBBM); and (3) to measure knowledge, risk-perceptions, substance abuse, and prevalence of symptoms of Sexually Transmitted Infections (STI), and healthcare-seeking behaviour among the general male population in Bangladesh.

A pilot study was conducted in one urban area and in one rural area during December 2004–January 2005 to test how interviews including sensitive questions using a ballot-box with audio system compare with those with the Simple Ballot-Box Method (SBBM) (by asking sensitive questions in a face-to-face interview and collecting the responses with the use of ballot-box and ballot-slips). After piloting the MBBM was found to be effective and was planned for use in the main study. A cross-sectional descriptive survey was conducted among men aged 18-49 years in three purposively selected urban areas (Dhaka metropolitan, Chittagong metropolitan, and Bogra town) and three rural areas (Faridpur, Rajshahi, and Cox’s Bazar districts) of Bangladesh. A 30 cluster sampling method with segmentation at the second stage and systematic selection of subjects in the third stage was used. In each cluster, from the master list of eligible respondents, identified through household listing, 50 were systematically selected and randomly allocated to either of the two interviewing methods (FTFI or MBBM). During February–August 2005, 24 trained interviewers and six supervisors, divided into six teams, were employed for data collection. Of 4497 and 4498 respondents who were approached for interviewing in each FTFI and MBBM, 3623 (80.6%) and 3499 (77.8%) respectively were successfully interviewed.

Overall, 17.5% of the respondents ever had pre- or extra- marital vaginal/anal sex (non-marital sex) in the past year. The corresponding figures were 9.9%, 8.6%, and 2.2% respectively for sex with female sex workers, casual female partners, and males/transgenders. Using the MBBM, 11.7%, 8.8%, and 2.9% ever had vaginal/anal sex in the past year with female sex workers, casual female partners, and males/transgenders respectively. The corresponding figures using the FTFI were 8.1%, 8.5%, and 1.4%. The MBBM elicited higher responses from those who had sex with female sex workers and males/transgenders than the FTFI. The MBBM did not produce substantially different responses for sex with casual female partners. However, for any of the afore-mentioned partners, the response in the MBBM was higher than for the FTFI. There was no statistical difference in response to condom-use and number of non-marital sexual partners by the two interview methods. Overall, condom-use rate during last sex with female sex workers, casual female partners, and males/transgenders was 40.1%, 30.0%, and 8.7% respectively. Among the respondents who had non-marital sex in the past year, overall, 44.4%, 35.9%, and 19.7% had 1, 2-3 and >=4 partners respectively.

Most (92%) respondents had heard of AIDS, 86% knew that some diseases are transmitted through sexual intercourse, and 85% knew that a healthy looking person may be HIV-infected. Thirty-five percent knew that STI symptoms can differ between the sexes but 46% could not mention any STI symptoms and 18% did not know any modes of transmission. Approximately 76% of the respondents perceived that they were at low risk for acquiring HIV. About 8% suffered from STI symptoms during the past year; 59% of them went for treatment, and of them 37% visited qualified physicians and 63% went to informal providers. About one-third had sex at times when they also had STI symptoms, most (85%) without condoms and 51% did not inform partners about the presence of symptoms during sex.

Non-marital sex with high-risk partners is common among the general male population in Bangladesh. Low condom use further augments the risk of HIV transmission in the community. The MBBM is more effective than the FTFI in eliciting higher responses regarding non-marital sexual contact from males, especially sex with female sex workers and males/transgenders. The MBBM may be useful in improving response rates to sensitive questions on sex behaviour among the general male population. Overall knowledge of the general male population on HIV infection was high and the majority perceived themselves at low risk of contacting infection. Although the prevalence of STI symptoms was high in the general male population, care-seeking for STI symptoms was low. To prevent a generalized epidemic, behaviour change communication programming with males should be implemented urgently to reduce rates of sexual partners in increase condom use and seek care for STI symptoms.

A report on this project is available here