How safe is sex with condoms?: an in-depth investigation of the condom use pattern during the last sex act in an urban area of Bangladesh.

The policy of condom intervention is based on achieving ejaculation inside a condom, a "mechanical" goal of sexual interaction, a social act. However, most research on condom use has focused upon a simplistic reliance on survey results of condom use during the last sex act. Interviews with 20 hotel-based, female sex workers and 15 (male) clients were conducted to explore patterns of claimed condom use during the last sex act. The Health Belief Model guided this study and was found deficient in providing an understanding of condom use. The clients' perceptions of dominating sexuality and "the male's right" to enjoy sexual intercourse in commercial settings increased partial condom use. The invisibility of AIDS reduced participants' perceived susceptibility to and severity of suffering from the disease, while using condoms at any time during intercourse was perceived as being beneficial. Condom interventions need to be based on deeper understanding of the complexity of people's lives.

Key Words: condom use, female sex workers, male clients, sexual intercourse, Bangladesh

**********

Condom use during sexual intercourse, whether vaginal or anal, is a prime prevetion tion tool against sexually transmitted infections (STIs). Correct and consistent condom use can also decrease the spread of the human immunodeficiency virus (HIV) (Pinkerton & Abramson, 1997). Many organisms, including HIV, cannot be transmitted through an intact condom worn during sexual intercourse (Fiumara, 1972; Roper, Peterson, & Curran, 1993). Findings from 10 cohort studies conducted in Western countries, which evaluated the efficacy of condom use among heterosexual couples, showed that consistent condom use could protect people against HIV infection (Feldblum, Morrison, Roddy, & Cates, 1995). The most convincing evidence of condom effectiveness comes from studies of HIV-discordant couples, in which one partner is infected with HIV and the other not (Centers for Disease Control [CDC], 1998; Feldblum et al., 1995). Studies of such couples have found significantly lower risks of HIV infections among consistent condom users (Allen et al., 1992; Laurian, Peynet, & Verroust, 1989; Ngugi et al., 1988; Plummer et al., 1991; CDC, 1998). A multi-European country study of 256 HIV-discordant couples followed for an average of 20 months did not find a single new infection occurring among couples using condoms during every sex act (De Vincenzi, 1994). Condoms may help prevent AIDS over the long term not only by blocking transmission of HIV but also protecting against other STIs. For instance, people with genital ulcerative STIs such as chancroid, genital human papillomavirus, herpes simplex, and syphilis are two to seven times more likely to become infected with HIV than people who do not have STIs (Diallo et al., 1992; Laga, Nzila, & Goeman, 1991). If a condom is not worn before penetration, the pre-ejaculatory fluid that contains HIV and sperm may be secreted inside the vagina. This pre-ejaculatory fluid increases the risk of HIV as well as pregnancy (Ilaria et al., 1992; Pudney, Oneta, Mayer, Seage, & Anderson, 1992; Trussell, 1998). In addition, the contact of the infected mucosal surface with an uninfected surface also facilitates transmission of a number of STIs (Sparling, 1990). Therefore, condoms can only be highly effective if they are used correctly and continuously throughout intercourse (De Visser & Smith, 2000).

Baseline, cross-sectional sexual behavioral surveys or post-intervention evaluation surveys are conducted using quantitative instruments. Calculation is made by measuring the percentage of men who reported condom use at the last episode of sex with a sex worker and those who reported sexual intercourse with a sex worker in the last 12 months preceding the survey. Researchers and program managers use this indicator as evidence of risk behaviors and the success or failure of condom promotion interventions as well. The question generally asked is, "Did you or your client use a condom during the last sex act?" When people report condom use, they do not necessarily mean that they used a condom consistently from the beginning of sexual intercourse to the end. This potential lapse in condom use questions the validity of the condom-use measure even when positive answers are provided on the survey questions. A number of studies have explored an alarming phenomenon of the delayed use of condoms during an act of sexual intercourse (Browne & Minichiello, 1994; De Visser & Smith, 2000; Quirk, Rhodes, & Stimson, 1998).

Qualitative studies on the dynamics of condom use among Bangladeshi males are quite limited. However, the few reported studies have primarily involved quantitative measurement and revealed a significantly low level of use among all groups studied (Mitra, Ali, Islam, Cross, & Saha, 1994; National AIDS/STD Program, 2001; Sarkar et al., 1998). These quantitative studies failed to attach meaning to the low rate of condom usage. A person is generally not asked about the pattern of his claimed condom use if he reports condom use in the last week or during the last sex act.

This present study explored the patterns of condom use among a sample of hotel-based, female sex workers (HBFSWs) and their male clients who claimed to have used a condom during the last commercial sex act. This study also addressed the underlying contextual issues and motivations for condom use. The in-depth contextual meaning of condom use tan contribute to designing and re-designing condom promotion interventions for targeted populations, such as sex workers and their clients.

HEALTH BELIEF MODEL

The Health Belief Model (HBM) (Janz & Becker, 1984; Rosenstock, Strecher, & Becker, 1988) provides useful elements for examining the determinants of sexual decision-making, especially behavioral aspects of condom use. The HBM was developed in the early 1950s by a group of social psychologists in the United States Public Health Service in order to understand the failure of people to engage in preventive activities (Rosenstock, 1974). The HBM has been used to study patients' responses to symptoms of illness (Kirscht, 1974; Kirscht & Joseph, 1989) as well as patients' compliance with medical recommendations (Becker, 1974). The HBM is the most widely researched and accepted theory exploring why people do and do not practice health, illness, and sick-role behaviors (Sarafino, 1990). According to Rosenstock (1974), the core concept of HBM is based on: (1) the individual's psychological "readiness to take action" related to any preventive health behaviors, which is determined by both the person's perceived "susceptibility" to a particular condition and the "severity" of the consequences of contracting that condition; and (2) the individual's evaiuation of the recommended health behavior in terms of estimating the potential "benefits" weighed against psychological or other "barriers" or "costs" in order to reduce perceived susceptibility and severity (i.e., a cost-benefit analysis of an action). An external or internal stimulus may act as a "cue to action" to trigger the appropriate behavior. In practicing a preventive behavior, perceived susceptibility to a disease and perceived benefits of and barriers to adopting a preventive behavior play crucial roles more so than the perceived severity of that disease. However, in studies of sick-rule, perceived severity of a disease significantly contributes to illness and compliance behaviors (Rosenstock, Strecher, & Becker, 1988). The elements of the HBM can thus be applied to understanding of safer-sex behaviors (condom use) before or during an illness episode to prevent new or further infections of STIs/HIV.

METHOD

The present qualitative study explored the exact specifics of condom use at the last commercial sex act. Twenty hotel-based, female sex workers (HBFSWs) located in a port city of Bangladesh, who claimed to have used condoms in the last sex act, were interviewed by snowball sampling. A local, non-government organization (NGO), the Health and Education for Less-privileged People (HELP), had created satisfactory rapport with HBFSWs. This close relationship made it possible to conduct interviews with the HBFSWs. Two HELP staff members were interviewed and acted as key-informants to understand their condom intervention strategies. We approached clients through our personal networks and were able to interview 15 clients who claimed to have used a condom during the last commercial sex act with any type of female sex workers (FSWs) working either in hotels, boarding houses, on streets, or at residences.

We prepared two separate semi-structured, open-ended interview guidelines for interviewing the HBFSWs and their clients. We gathered information about the meaning and pattern of condom use, exact timing of putting on a condom and taking it off, the exact time span when the condom was worn, and the underlying emic explanations of each situation. The open-ended questions in the flexible interview guideline encouraged participants to raise issues about condom use not initially included in the interview. Subsequently, these "raised issues" were incorporated in the guideline for later interviews. Some HBFSWs allowed audio recording of the conversations, while others were hesitant. We did not record the conversations of this latter group. For this group, we depended on hand-written notes. None of the clients allowed the use of a tape recorder. Therefore, we took hand-written notes, using them as the basis for writing full reports at the earliest possible time following the interview.

Instead of written consents, a verbal affirmation was considered acceptable to begin sensitive interviews. A verbal testimony was tape-recorded every time at the beginning of the in-depth interview. Each participant was informed of the study's objectives, use of tape recorder, and the intimate nature of interview questions. They were told about their rights and role in the research before beginning the interviews. Participants were informed that they could stop the interview at any time. In addition, the participants were told not to answer any question they perceived impolite or insensitive. The participants chose the location of the interviews. Initially anxious, the clients took a longer-than-expected time to develop rapport with the interviewer. One male and one female interviewer were trained for conducting interviews with clients and FSWs, respectively. We assumed that FSWs would be more comfortable interacting with female interviewers. We maintained strict confidentially for all collected information, the locations of the interview venues, and the participants' identity.

Tape-recorded interviews were transcribed and later destroyed. Ongoing data analysis was manually performed in the framework of line-by-line content, contextual, and thematic analysis. We provided the interpretive contextual meaning of data in the context of "thick" descriptions (Ezzy, 2002; Geertz, 1973; Miles & Huberman, 1994) in order to identify diverse patterns of condom use. Findings were interpreted within the framework of the Health Belief Model as well as focusing on gender and masculinity.

In this study, correct condom use refers to using a condom properly as conventionally suggested including the careful opening of the package, unrolling the condom on to an erect penis, and wearing the condom throughout the entire act of intercourse. The consistent condom use generally refers to the use of a condom before penetration and until the end of a single episode of intercourse. Therefore, in the broadest sense, correct condom use includes consistent condom use in the sense of its complete and full time use during sexual intercourse. In this article, consistent condom use refers to the correct use of a condom from erection before penetration to the end of ejaculation.

RESULTS

Both clients and hotel-based, female sex workers reported using condoms during their last sex act. We explored the exact timing of condom use, the underlying reasons and meanings of condom-use behaviors and perceived barriers to and experiences of condom use. Three specific patterns of condom use are identified and discussed below.

PATTERN I:

STARTED INTERCOURSE WITHOUT CONDOM, BUT PUT ONE ON BEFORE EJACULATION

Some HBFSWs and clients reported that sex began without a condom but that condoms were put on before ejaculation. One sex worker stated, "It is far better to use a condom whatever stage of intercourse than complete non-use, which is risky." Another sex worker noted," ... since semen was not ejaculated inside the vagina, I believe this [put the condom on just before ejaculation] can also protect."

Here are some of the comments sex workers made when asked about the possibility of an infected sex worker infecting a client:

   Our clients do not bother to protect us by using condoms. Why should
   we care about the safety of our clients?

   As long as semen is not ejaculated inside, I am safe. I never
   consider whether I can infect my client, especially, since I believe
   I have no disease.

   I do not have any idea how HIV can be transmitted from my body to my
   client. If I have a disease, I think it will be transmitted through
   sexual intercourse, kissing, and close body contact. Therefore, a
   condom alone cannot protect completely.

A sex worker perceived that she would be sale as long as semen was not discharged inside her vagina. The perceived benefit of using a condom at least before ejaculation was given comparatively higher importance than the perceived costs of convincing a client to use a condom consistently. These statements suggest that some sex workers perceived themselves free from STIs and AIDS. They, therefore, believed they could not infect clients. Others obviously were not concerned with their clients' safety because they believed that clients would never think about their safety. Furthermore, some of the sex workers perceived themselves powerless in terms of expressing their rights and choices to their clients. Sex workers sell their bodies to clients, who by virtue of the financial transaction gain the right over the sex worker's body. As such, the foundation of the relationship is clearly based on the power differentials in the context of a male-dominated gender perspective. Therefore, the mutual respect and sense of complementary and responsible sexual relations were absent and probably will be difficult to achieve especially through promoting condom-only messages.

On the other hand, clients framed their statements differently concerning condom use.

   I know I have to use a condom, and I do. No one ever told me exactly
   when and how to wear a condom. Is there a correct time? I use one
   during the middle of the sex act, so that I get much pleasure and can
   also ensure protection.

   I do not like condoms. However, I have no choice but to use a condom
   to protect myself. I have my own style. I like to enjoy sex without a
   condom for a few minutes and then, finally, I wear it for protection
   before I ejaculate. I do not think an infected sex worker can infect
   me if I have a few minutes of sex without a condom. I do not
   ejaculate inside the vagina. There is therefore no chemical
   interaction of mal [semen] with vaginal fluid and no risk.

   Sex workers insist that I use a condom. So I trick the sex workers. I
   begin sex without a condom. Then I put on the condom for a few
   minutes [and] secretly taking it off after a few minutes. Sometimes,
   I ejaculate before putting on the condom. I prefer to ejaculate
   inside the vagina, especially if the sex worker is not so adamant
   about condom use.

   I always enjoy sex for a few minutes without a condom. When I find
   myself ejaculating, I pull out and put on a condom. This break
   decreases my sexual excitement and postpones my ejaculation. Then I
   am able to sustain the sex act. Condom use can increase my
   performance and pleasure, and I feel like a really powerful man. To
   tell you the truth, I do not bother that my semen may infect sex
   workers or that I can be infected from them. Condoms help to prolong
   sex, and that's why I use them. That's all.

These statements convey diverse and deeper contextual meanings. These include (1) little knowledge and gaps in the understanding of the disease transmission process; (2) individuals' perception of pleasure; (3) sense of responsibility to engage in safer sex; (4) symbolic meaning of birjoban purus or "real man" within the context of masculine sexual performance; (5) the personal techniques of making sex enjoyable and prolonged; (6) sex workers' request for condom use in the context of increased awareness for safer sex from NGO interventions; and (7) clients' various strategies to avoid condom use by giving a false sense of security to sex workers. Perceived benefits of using a condom only part-time are considered greater than the perceived drawbacks in reducing pleasure in full-time condom use. However, this calculation is not a straightforward one. The cost-benefit analysis actually takes place in the different layers of the complex framework of personal perception of pleasure, sense of masculine sexuality, and responsibilities to perform safer sex with a sex worker. All these contextual issues are equally crucial to anticipate any final action. Using a condom to increase the duration of sexual intercourse is considered a "real man's sex act" and can contribute to the development of an innovative message for condom promotion.

PATTERN II:

STARTED INTERCOURSE WITH A CONDOM, BUT TOOK IT OFF AFTER FEW MINUTES OR JUST IMMEDIATELY BEFORE EJACULATION

Some sex workers and clients reported that the last intercourse was initiated with a "condom on the penis," but after a few minutes or just before ejaculation, condoms were taken off. Again, here are some sex workers' comments with regard to this situation.

   Within a minute of beginning sexual intercourse, the client's lingo
   [penis] became norom [flaccid], and he became angry with me. He took
   out his lingo, took off the condom, and started sexual intercourse
   again until he ejaculated. What can I say then? He paid money, so he
   deserves to ejaculate.

   My client was not at fault. He began with a condom, but could not
   continue and had no choice but to complete sex without a condom.
   Never using a condom before, the client felt bad. Most of the clients
   simply will not try a condom. They just refuse to use a condom. My
   last client at least tried. He failed, but I did not force
   him.

   I saw him wearing a condom before he penetrated me, but I did not
   realize when he took it off. He did not tell me; rather he claimed
   that he used a condom. But at the end, I felt his semen in my vagina.
   I checked the condom, and it was empty. I realized he had cheated me,
   but what can I do? Everything was over ... then I thought, well, I
   have been cheated many times in my life. A condom-cheat is nothing
   big or any new incident.

   I know men's semen contains disease, but what can I do if they behave
   like that? Not only me, none of us [HBFSWs] can say anything against
   our clients. If we say anything bad to out clients, and if clients
   complain to the hotel authority, we will no longer be allowed to work
   here. Sex workers are not united. I believe if we were united, then
   we could force our clients to use condoms. I request you
   [interviewer] to talk to the hotel authority, so that they can force
   the clients to use condoms before they enter our rooms. Then we also
   will be powerful enough to say "no" to clients who refuse to use
   condoms.

These statements contain several diverse meanings and important implications for interventions by NGOs. Sex workers' perceptions of their low and subordinate status discourage them from protesting against any of their clients' "cheating behavior." Sex workers perceived themselves as "slaves" to the male authority of the sex trade. Their low self-esteem forces them to be silent and not speak out against their clients' unsafe behaviors. A sex worker's priority is to protect her source of income, not to protect her body from infections. Therefore, despite knowing the dangers of unprotected sex [e.g., ejaculation inside the vagina], the perceived benefits of protection against disease was not given a higher value than the perceived danger of losing one's source of income. A significant message for those working with sex workers can be found in the last statement (above), which demonstrates the importance of sex workers' unity by forming associations and involving the hotel management in HIV intervention.

When clients were asked why they took off condoms during intercourse, their statements reflect other realities.

   I know that I have to use condoms to prevent HIV. However, after I
   began intercourse wearing a condom, I immediately lost pleasure, felt
   very bad, and my erect penis became soft. Then I had no choice but to
   throw it off and to complete sexual intercourse, since I already had
   paid money.

   I begin intercourse with a condom. I continue intercourse with the
   condom on. Just before ejaculating, I love to discharge my semen
   inside her vagina. By wearing the condom I was safe. I know
   ejaculation inside a vagina is safe for me, but not good for her.
   However, that is not my concern. They are already diseased.
   Ejaculation inside the vagina is real sex for a birjoban purus [a
   "real man"]. Visiting a sex worker is a risky behavior, but I am
   unconcerned about risk. I have accepted many risks already in my
   life.

   I have been enjoying sex with sex workers for the last five years.
   Only during the past year, I occasionally used condoms. However, I
   have never suffered from any sexual diseases or AIDS. Therefore, I
   really do not understand what is the role of a condom in protecting
   [from a] sexual disease?

   I had sex with a condom and took off the condom just before
   ejaculation. I do not have HIV or any other sexual diseases. I also
   never saw any AIDS patient in my life. Therefore, I think if I
   ejaculate inside her vagina, she will actually have no problem.

These clients' statements illustrate many significant issues. For example, clients' first-hand experience of reduced pleasure and having a flaccid penis after putting on a condom was perceived as a significant barrier to further use. On the other hand, if discharging semen inside the vagina is symbolized as "real sex" or a "real man's" sex, then the perceived benefits of consistent condom use will never outweigh the perceived costs of ejaculation inside a condom. This is especially true when clients' perceived susceptibility and severity to STIs/AIDS is considered low in the absence of earlier illness experience, current disease-free status, or invisibility of AIDS patients to validate the risk involved in unprotected sex. Moreover, the perceived responsibility of protecting a sex worker has been compromised by clients' lack of concern for sex workers' safety in the framework of male-dominated masculine sexuality and the perceived right of purchasing the sex worker's body in the sex trade dynamics.

PATTERN III:

STARTED INTERCOURSE WITH CONDOM AND CONTINUEr) UNTIL THE END (CONSISTENT CONDOM USE)

A few sex workers and clients claimed a condom was used and worn consistently throughout the last sex act. However, it is noteworthy that although these clients wore a condom during the last sex act, they practiced a non-consistent pattern of condom use at other times. Sex workers, on the other hand, had similar experiences where clients did not use a condom consistently. However, it is crucial to know the underlying reasons for consistent condom use. One sex worker stated:

   I love my clients. I try to enjoy sex with my clients. I do many
   things before penetration. I show that I care for them. Then I
   honestly and gently request them to use condoms before they penetrate
   me. I assure them that I will give them pleasurable sex if they use
   condoms. In this way, I try to convince my clients to use condoms. In
   many occasions, clients use condoms nicely.

This statement carries a crucial message for other sex workers and peer educators. Convincing a client to use a condom is a matter of art and tact. Forcing clients to wear a condom on many occasions may not work if the penis really becomes flaccid or if perceptions of masculine sexuality are threatened. Here are some clients' statements that follow up on this theme.

   When I use condoms, I use them full-time because I know I may be
   infected if I have sex without a condom. On the other hand, if I
   ejaculate inside her vagina, she may be infected if I have a disease.
   Therefore, I do not believe [in] part-time condom use.

   I never have sex without a condom, especially since I learned about
   HIV and condom use. My friends told me that pre-ejaculatory fluid
   contains HIV germs, which may infect my sex partners. On the other
   hand, I may also be infected from a sex worker if she is infected. In
   fact, to tell you frankly, I am afraid about my own safety.
   Therefore, I never introduce my penis without a condom inside a sex
   worker's commercially used vagina. I actually do not consider her
   safety.

   If I begin sex without a condom, then it becomes very difficult to
   use a condom in the middle of the sex act. Therefore, if I use a
   condom, I use it from the beginning and, if possible, I continue till
   the end.

   A condom can make me stronger and perform longer. I can prove my
   sexual potency. I noticed that my penis without a condom becomes very
   sensitive when I penetrate. I come quickly. It is better for men to
   use condoms because condoms can reduce sensitivity and pleasure and
   can prolong sexual intercourse in return.

These statements reflect some crucial issues that sex educators should consider. For instance, knowledge of HIV transmission and ways of prevention should be comprehensively delivered to both clients and sex workers. It has to be remembered that partial and incomplete messages sometime make the situation confusing. Peereducation strategies tan work well with clients as well. Overall, the clients in this study were more concerned with their own safety and less about sex workers' safety. More important, the capacity of a condom to prolong the timing of sexual intercourse and offer the opportunity to perform "real man's sex" should be considered in a condom promotion message. Condom use that emphasizes increased masculine sexuality tan positively motivate men to use condoms.

CONDOM PROMOTION MESSAGE REQUIRES COMPLETENESS AND INNOVATION

We worked with a local NGO that has been recently promoting safer sex (condom use) among hotel-based, female sex workers who received the message that AIDS could be prevented if they could convince their clients to use condoms. We found that the danger of men's pre-ejaculatory fluid and the exact mechanism of STIs/HIV transmission from one person to another were not discussed clearly nor emphasized. Further, issues of sexuality and sexual behaviors, gender and masculinity dimensions in sexual behaviors, and social and cultural context of risk and vulnerability were never discussed with sex workers. The core issue that the sex workers got from intervention messages was that if semen could be contained in a condom, they would be disease-free. When asked, the sex workers identified condoms as a primary protection from HIV/AIDS. However, they did not clearly understand how STIs or HIV could be transmitted from one person to another. And few sex workers were curious. One sex worker raised the issue in the following way:

   We heard that men's semen contains the germ of HIV, but we [women] do
   not have semen like men, so how can HIV be transmitted from our body
   to the clients' body? I believe clients actually carry HIV and infect
   us. We are blamed since we are kharap maye [bad women] in our
   society.

It's noteworthy that despite exposure to consciousness-raising intervention programs, sex workers neither received complete information nor were able to completely internalize the messages they heard. Well-intentioned intervention programs have to be designed by considering sex workers' belief systems, their attitudes toward clients, and the overall safety issues in the sex trade.

Although there were no direct educational programs for clients, we noted clients' information about STIs/HIV transmission was better than the sex workers' knowledge we studied. Developing separate materials to educate clients and sex workers could be beneficial in increasing both groups' understanding. Appropriate service facilities need to be available and accessible for clients.

Health promotion messages are based on the assumption that the targeted populations need biomedical knowledge of disease transmission for its prevention. Therefore, awareness-building messages are delivered in order to make the targeted population knowledgeable. An NGO staff member stated:

   We have developed awareness-building messages in brief and
   understandable language. If we deliver complex biomedical
   information, then sex workers will not understand many issues and may
   not respond to out messages. Therefore, we basically deliver
   knowledge about condoms and the danger of semen if it is discharged
   into the vagina. We actually encourage them to talk to their clients
   about condoms and request or even insist clients use condoms. Our NGO
   is a new one and lacking funds; however, we are trying out best to
   help. Now, at least, they know about condoms, and I believe many are
   probably using condoms as well.

Intervention messages should contain simple and brief information and avoid complexity. However, it is necessary to differentiate between simple versus complex messages and brief versus adequate information. For example, in order to make an awareness-building message simple and brief, the potential danger of pre-ejaculatory fluid in transmitting STIs/HIV or pregnancy was not properly addressed. Thus, the perceived simple and brief form of message did not contain crucial information, which not only made the message incomplete but also created a crucial gap. An effective Behavior Change Communication (BCC) material has to consider these issues.

DISCUSSION

This study explored the nature and dynamics of condom use. Our purpose was to select only those participants who used a condom in the last sex act. Therefore, we used a snowball sampling among a closed network of clients and sex workers (hotelbased, female sex workers). The known bias of snowball sampling is the strength of this project. In the context of partnership with the intervention NGO, we were able to select sex workers and to create a study based on mutual trust and respect between researchers and participants. We conducted an ongoing analysis, which began during data collection. This enabled us to modify the focused questions based on field experiences. Mutual trust and respect between researchers and participants were absolutely necessary to gain understanding of their thoughts and behaviors concerning condom use. Avoiding written consent and depending on verbal affirmation enhanced this trust.

The Health Belief Model (HBM) is a model of decision-making that attempts to predict whether people will accept medical intervention and treatment or follow preventive behaviors such as condom use during sexual intercourse. A judgment about disease risks, benefits, and barriers of condom use is a social feature that reflects values, perceptions, and conflicts at the societal level. The HBM is mainly concerned with people's cognition and decision-making processes. The model is based on the assumption that people have sufficient knowledge to make decisions according to their perceptions. Although this study does not allow for any conclusive comment on the HBM, we note that clients' condom-using behavior requires complex contextual analyses regarding the perceived benefits and barriers to safer sex. Condom-using behaviors are not constructed within a social and environmental vacuum. Rather the cost-benefit analysis of condom use is socially constructed in the framework of masculine sexuality, gender power relations, and socioeconomic realities in peoples' lives. The main weaknesses of the HBM are its lack of recognition of sociocultural, economic, and environmental factors, including gender and masculinity. These factors have crucial influences on an individual's decisions to undertake any recommended actions (Hulton & Falkingham 1996; Pleck, Sonenstein & Ku, 1993a & b; Wilton, 1997). The realities of peoples' everyday lives are too complex to be captured by the Health Belief Model. Concepts of gender, masculinity, and social environmental contexts need to be understood if culturally appropriate and effective means of safer sexual practices are to be achieved.

During program evaluations, if a sex worker names a condom as the means of prevention or displays that she is carrying a condom, many researchers will [mis]interpret either act as an indication of a successful intervention program. Program managers also become complacent in light of the "success" of intervention. The recent experience of condom promotion in a brothel in Bangladesh suggests that the increasing trend of sex workers' requesting clients to use condoms or even having a condom in their possession during the survey has no correlation to the proportion of actual condom use during each act of sexual intercourse (MAP, 2001). The present study revealed that many outreach workers and peer educators did not have sufficient training to deliver complete safer-sex messages. Condom promotion interventions are formulated without due attention to detailed information about men-to-women and women-to-men sexual transmission of STIs/HIV. The art of developing simple but complete messages is the core of BCC approach. In many cases, BCC materials are based on simple biomedical perspectives ignoring the detailed context of risk-taking behavior. People's knowledge acquired through an awareness-building program may contain gaps that result in potential threats, making the intervention unsuccessful. Studies of these kinds of programs often conclude, perhaps incorrectly, that knowledge alone does not ensure behavior change.

The current condom promotion programs are narrowly focused under the banner of safer-sex campaigns. The way program managers are introducing condoms inherently suffers from limitations since the approach fails to address the complex dynamics of gendered sexual behaviors, the impact of masculinity on men's sexual risk-taking, and the social dimensions of condom use. The "condom solution" is a simple biomedical response to a complex social behavior, which isolates people from their sociocultural contexts and considers men are solely responsible for non-use of condoms. On the other hand, women, as sex workers, are targeted as being accountable to ensure that men "behave" properly by wearing condoms. This simple approach denies the complexity and importance of male/female relationship in sexual intercourse. It makes women responsible for ensuring men's protective behaviors by further accentuating the power relation as an essential human condition. It also ignores sex workers' real life situations and diverse vulnerabilities. The strategy assumes that individuals listen, learn, and perform according to whatever information is delivered to them. This focus at the individual level usually neglects the fact that men are guided by collective sociocultural norms (Connell, 1987, 1990) and gender relations of their social environments (Connell, 2002) where the risks are embedded. Thus, expected behavioral changes become difficult, due to the decontextualization of condom use from both men's and women's real life situations.

Condom promotion campaigns cannot be delivered in a social vacuum by ignoring issues of sexuality, masculinity, and gender constructions. The intervention should not ignore or overlook the interconnectedness of human behaviors with overall environmental contexts and personal expectations. This may be the reason policy planners and program managers, despite their well-intentioned efforts in HIV interventions during the last two decades, have not been able to achieve desired outcomes in terms of encouraging men to use condoms. Billions of dollars have been spent to "condomize" men, with questionable success. Most of the programs have adopted indirect approaches and targeted women or female sex workers in the hope that they could encourage their clients to use condoms. Males control decisions about condom use, particularly in the commercial sex setting of Bangladesh. The meaning of condom using behavior is socioculturally constructed, reaching far beyond one's cognitive domain of perceived cost-benefit analysis. Condom use is a social behavior and probably one of the most ambiguous behaviors, since it takes place between at least two persons with an unequal distribution of power including physical, mental, social, economic, gender relations, and acquired knowledge. Therefore, knowledge of condoms and their consistent use is not a straightforward calculation. Asking a person whether he used a condom during his last sex act is not a simple question. Based on out research findings, we suggest that claimed condom use during the last sexual act might have potential hidden threats that impose critical challenges for condom promotion programs. The limitation of traditional surveys can be diminished if questions are designed on the basis of in-depth findings of condom-using complexities. The meanings of condom use held by men and women are no less important than are simple measurements of condom usage. The conventional framework of condom promotion strategy, which targets individuals by providing simple awareness-building messages, and implementing social marketing of condoms through peer encouragement to female sex workers needs to be reevaluated. Innovative strategies have to move beyond individual focus toward structural and social dimensions to deal effectively with the cultural meanings attached to masculine sexuality and gender dimensions in condom use.

REFERENCES

Allen, S., Tice, J., Van De Perre, P., Serufilira, A., Hudes, E., Nsengumuremyi, F., et al. (1992). Effect of serotesting with counselling on condom use and seroconversion among HIV discordant couples in Africa. British Medical Journal, 304, 1605-1609.

Becker, M.H. (1974). The Health Belief Model and sick role behavior. Health Education Monogram, 2, 409.

Browne, J., & Minichiello, V. (1994). The condom: Why more people don't put it on. Sociology of Health and Illness, 16, 229-251.

Centers for Disease Control and Prevention (CDC). (1998). Facts about condoms and their use in preventing HIV infections and other STDs. Atlanta.

Connell, R.W. (1987). Gender and power. Stanford: Stanford University Press.

Connell, R.W. (1990). A whole new world: Remaking masculinity in the context of the environment movement. Gender & Society, 4(4), 452-78.

Connell, R.W. (2002). Gender. Cambridge, UK: Polity Press.

De Vincenzi, I. (1994). A longitudinal study of Human Immunodeficiency Virus transmission by heterosexual partners. New England Journal of Medicine, 331 (6), 341-346.

De Visser R.O., & Smith, A.M.A. (2000). When always isn't enough: Implications of the late application of condoms for the validity and reliability of self-reported condom use. AIDS CARE, 12(2), 221-224.

Diallo, M.O., Ackah, A.N., Lafontaine, M.F., Doorly, R., Roux, R., Kanga, J.M., et al. (1992). HIV-1 and HIV-2 infections in men attending sexually transmitted disease clinics in Abidjan, Cote d'Ivoire. AIDS, 6(6), 581-585.

Ezzy, D. (2002). Qualitative analysis: Practices and innovation. NSW, Australia: Allen & Unwin.

Feldblum, P.J., Morrison, C.S., Roddy, R.E., & Cates, W. (1995). The effectiveness of barrier methods of contraception in preventing the spread of HIV. AIDS, 9(Suppl. A), S83-S93.

Fiumara, N.J. (1972). Ineffectiveness of condoms in preventing venereal disease. Medical Aspects of Human Sexuality, 6, 146.

Geertz, C. (1973). The interpretation of cultures. Basic Book: New York.

Hulton, L., & Falkingham, J. (1996). Male contraceptive knowledge and practice: What do we know? Reproductive Health Matters, 7, 90-99.

Ilaria, G., Jacops, J., Polsky, B., Koli, B., Baron, P., MacLow, C., et al. (1992). Detection of HIV-1 DNA sequences in pre-ejaculatory fluid. Lancet, 340(8833), 1469.

Janz, N.K., & Becker, M.H. (1984). The Health Belief Model: A decade later. Health Education Quarterly, 11, 1-47.

Kirscht, J.P. (1974). The Health Belief Model and illness behavior. Health Education Monogram, 2,387-408.

Kirscht, J.P., & Joseph, J.G. (1989). The Health Belief Model: Some implication for behavior change, with reference to homosexual males. In V.M. Mays, G.W. Albee, & S.F. Schneider (Eds.), Primary prevention of AIDS: Psychological approaches (pp. 111-127). London: Sage.

Laga, M., Nzila, N., & Goeman, J. (1991). The interrelationship of sexually transmitted diseases and HIV infection: Implications for the control of both epidemics in Africa. AIDS, 5(Suppl. 1), S55-S63.

Laurian, Y., Peynet, J., & Verroust, F. (1989). HIV infection in sexual partners of HIV-seropositive patients with hemophilia. New England Journal of Medicine, 320(3), 183.

MAP Report. (2001). The status and trends of HIV/AIDS/STI epidemics in Asia and the Pacific: Provisional report. Melbourne, Australia. Monitoring the AIDS Pandemic (MAP) Network Secretariat: U.S. Census Bureau, Washington, D.C.

Miles, M., & Huberman, A. (1994). Qualitative data analysis: A sourcebook of new methods. (2nd ed.). Thousand Oaks, CA: Sage

Mitra, S.N., Ali, N., Islam, S., Cross, A., & Saha, T. (1994). Bangladesh demographic and health survey, 1993-1994. Calverton, MD: National Institute of Population Research and Training, Mitra and Associates, and Macro International, Inc.

National AIDS/STD Program. (2001). HIV in Bangladesh: Where is it going? Directorate General of Health Services, MOHFW: Government of the People's Republic of Bangladesh.

Ngugi, E.N., Plummer, F.A., Wimonsen, J.N., Cameron, D.W., Bosire, M., Waiyaki, et al. (1988). Prevention of transmission of Human Immunodeficiency Virus in Africa: Effectiveness of condom promotion and health education among prostitutes. Lancet, 2(8616), 887-890.

Pleck, J.H., Sonenstein, F.L., & Ku, L.C. (1993a). Changes in adolescent males' use of and attitudes toward condoms, 1988-1991. Family Planning Perspectives, 25(3), 106-110,117

Pleck, J.H., Sonenstein, F.L., & Ku, L.C. (1993b). Masculinity ideology: Its impact on adolescent males' heterosexual relationships. Journal of Social Issues, 49(3), 11-29.

Plummer, F.A., Simonsen, J.N., Cameron, D.W., Ndinya-Achola, J.O., Kreiss, J.K., Gakinya, M.N., et al. (1991). Cofactors in male-female sexual transmission of Human Immunodeficiency Virus Type 1. Journal of Infectious Diseases, 162(2), 233-239.

Pinkerton, S.D., & Abramson, P.R. (1997). Effectiveness of condoms in preventing HIV transmission. Social Science and Medicine, 44(9), 1303-1312.

Pudeny, J., Oneta, M., Mayer, K., Seage, G. & Anderson, D. (1992). Pre-ejaculatory fluid as potential vector for sexual transmission of HIV-1. Lancet, 340, 1470

Quirk, A., Rhodes, T., & Stimson, V. (1998). "Unsafe protected sex": Qualitative insights on measures of sexual risk. AIDS Care, 10, 105-114. Roper, W., Peterson, H., & Curran, J. (1993). Commentary: Condoms and HIV/STD prevention-clarifying the message. American Journal of Public Health, 83, 501-503.

Rosenstock, I. (1974). Historical origins of the Health Belief Model. Health Education Monographs, 2, 238-335.

Rosenstock, I.M., Strecher, V.J., & Becker, M.H. (1988). Social learning theory and the Health Belief Model. Health Education Quarterly, 15, 175-183.

Sarafino, E.P. (1990). Health psychology: Biopsychosocial interactions. New York: Wiley.

Sarkar, S., Islam, N., Durandin, F., Siddiqui, N., Panda, S., Jana, S., et al. (1998). Low HIV and high STD among commercial sex workers in a brothel in Bangladesh: Scope for prevention of larger epidemic. International Journal of STD & AIDS 1998, 9(1), 45-47.

Sparling, P. (1990). Biology of Neisseria gonorrhea. In K. Holmes, P.A. Mardh, P. Sparling, & P. Wiesner (Eds.), Sexually transmitted disease (2nd edition; pp.131-147). New York: McGraw Hill.

Trussell, J. (1998). Contraceptive efficacy. In R.A. Hatcher, J. Trussell, F. Stewart, W. Cats Jr., G.K. Stewart, F. Guest, et al. (Eds.), Contraceptive technology (17th edition, pp. 779-845). New York: Ardent Media.

Wilton, T. (1997). Engendering AIDS: Deconstructing sex, text and epidemic, London: Sage.

This study was conducted at ICDDR, B: Center for Health and Population Research with support form SDC. ICDDR, B acknowledges the support of SDC to the Center's research efforts. We extend our thanks to the Social and Behavioral Sciences Unit (SBSU) of Public Health Sciences Division for its supports to the project. We tender out thanks to the Health and Education for the Less-privileged People, (HELP) Chittagong for their support during the fieldwork. The authors also wish to thank the anonymous reviewers for their helpful comments. Further, the interviewers and all other project staff are to be thanked for their hard work. Finally, we convey our deepest gratitude and respect to the interviewed hotel-based, female sex workers and clients for their cordial participation and for their willingness to share intimate issues of their lives on which this article is based.

Correspondence concerning this article should be addressed to Sharful Islam Khan, School of International, Cultural and Community Studies, Edith Cowan University, Mount Lawley, Perth, Western Australia. Electronic mail: [e-mail].

SHARFUL ISLAM KHAN

Edith Cowan University

Western Australia and

ICDDR,B: Center for Health

and Population Research

Bangladesh

ABBAS BHUIYA

ICDDR,B: Center for Health

and Population Research

Bangladesh

M.A. KAMRUL HASAN

Health and Education

for the Less-Privileged People

Bangladesh

NANCY HUDSON-RODD

Edith Cowan University

Western Australia

SHERRY SAGGERS

Edith Cowan University

Western Australia