Morineau, Guy MD, MSc*; Prybylski, Dimitri PhD, MPH†; Song, Ngak MD, MPH*; Natpratan, Chawalit MD, MSc*; Neilsen, Graham MBBS, MM†
CAMBODIA IS AMONG THE FEW COUNTRIES that has successfully turned around its HIV epidemics. Between 1998 and 2003, HIV prevalence among brothel-based female sex workers (FSWs) halved from 42% to 21%. Concomitantly, reported consistent condom use by FSWs with clients dramatically increased from 42% to 96% while the proportion of FSWs seeking treatment at a sexually transmitted diseases (STD) clinic for their last STI episode improved from 36% to 69%.1 Nevertheless, with an estimated HIV prevalence of 1.9% among adults aged 15 to 49 years in 2003, Cambodia still has the highest national HIV prevalence in Asia.2
The simultaneous use of multiple condoms (SUMC), or so-called double bagging, has been described as a common practice among American men who have sex with men.3 Cambodian FSWs first reported using more than 1 condom per intercourse in 1998, but it apparently became common among FSWs in the new millennium.1 In 2003, 71% of FSWs reported using more condoms than the number of episodes of sexual intercourse in their last working day.1 However, behavioral surveillance data did not distinguish whether this use of multiple condoms was sequential or simultaneous.
Although we could not find any published paper demonstrating that SUMC increases the risk of condom breakage, it is commonly believed and widely described on websites providing education to condom users that using more than 1 condom at a time leads to frequent breakage due to friction between the 2 condoms.4–8 A study conducted in neighboring Northern Thailand in 1995 found that 51% of FSWs reported SUMC with clients. In addition, this study found that SUMC use was associated with a reduced frequency of condom breakage.9 Interviews conducted with FSW after completion of the survey suggested that this behavior was initiated by FSW in 75% of the cases.9 Researchers hypothesized that SUMC was a spontaneous community-based coping strategy related to fears about HIV transmission during commercial sex.10
In Cambodia, military personnel serve as an occupational proxy population for clients of FSWs in the national behavioral surveillance system. Indeed, in 2003, 39% of military men reported visiting FSW in the past year,1 whereas a household survey conducted during the same year estimated that 8% of Cambodian men had done so in the past year.11 The vulnerability of Cambodian military personnel has justified the implementation of a national peer-education program among the armed force. Reported consistent condom use with FSWs in the past 3 months derived from behavioral surveillance survey data among military men rose steadily from 53% in 1998 to 89% in 2003.1 After 4 years of ongoing program activities, a survey was conducted in 2005 to evaluate the role of peer education on observed behavior change, the results of which are described elsewhere.12 This survey showed that privates were less exposed to peer education than ranked military, and those trained as peer educators received more peer education than others. In addition, those who had received more peer-education sessions had better knowledge of HIV transmission, were more likely to use condoms consistently with nonmarital partners, and were more likely to have sought their HIV serostatus. A secondary objective of this survey, and the focus of this article, was to estimate the prevalence of SUMC between FSWs and their military clients. As SUMC is expected to change the level of protection in condom use, it is important to assess the existence and levels of such behavior before conducting further research.
Materials and Methods
The study design was a cross-sectional survey among military personnel in 1 out of the 6 Cambodian military regions deliberately selected for its easy access from the capital city. All military men from the region were eligible regardless of their rank or military division. The size of survey population could not be determined because this is a sensitive national defense issue. The sampling design was a 1-stage stratified simple random sampling, with the strata consisting of platoons of approximately 60 men each. All platoons in the region were selected and participants were sampled through simple random sampling using those present in the barracks at the time of the survey as the sampling frame. A fixed number of men was selected from each platoon, regardless of the platoon size. The sampling design was meant to provide a representative sample of military men from the surveyed region and cannot be generalized to Cambodian military. The number of men present at the barracks on the survey day was recorded to calculate the sampling weights. Witnessed verbal informed consent was obtained from participants.
Data were collected using a structured questionnaire that was administered through face-to-face interviews in a private setting. Interviewers were military HIV trainers from other regions who dressed in civilian clothes for the survey and did not acknowledge their profession to participants. Informed consent for voluntary participation in the study was obtained from participants before interviews and all information was collected in an anonymous fashion; no names or personal identifiers were recorded. This study, which was conducted primarily to evaluate the military peer-education program and serve as baseline for evaluation of future interventions gathered information including: demographic and socioeconomic characteristics; the types, intensity, and duration of exposure to HIV prevention interventions; sexual risk behaviors; knowledge about HIV; drinking habits; attitudes regarding alcohol consumption; and attitudes regarding family. In addition, a specific question was included on SUMC (“The last time you used a condom with a FSW, how many condoms on the top of one another did you use?”). Knowledge of HIV transmission was measured as a composite score variable on a scale of 10 using questions that included kissing, sharing a meal, hugging, sharing toilets, unprotected vaginal sex, unprotected anal sex, shaking hands, unsterilized injections, mosquito bites, and mother-to-child contact through delivery and breastfeeding.
Survey respondents received a food incentive package for participating in the study. Field teams were trained on survey methods and sampling for a full week before the beginning of the survey. The study was approved by the Protection of Human Subjects Committee of Family Health International. No adverse events were reported during the conduct of the survey.
Data were double entered and compared using Epi-Data software.13 Analysis was performed using Stata version 8 (Stata Corporation, College Station, TX). Sampling weights were calculated and used to correct for differences in the selection probabilities among the platoons. Among those who had ever had sex with FSWs, we first compared demographics between those who used condom(s) at last paid sex and those who did not. Second, we explored associations between SUMC and behavioral outcomes among those who reported ever having had sex with sex workers. Differences in proportions were tested using the Pearson χ2 test. Comparison of means was ascertained using the Wald unpaired t test. All statistical tests were double-sided and a P value <0.05 was considered statistically significant. Simple logistic regression was used to evaluate risk factors associated with SUMC, and the significance of the resulting odds ratios was assessed using a Wald test. Factors that were significantly associated with SUMC in univariate analysis were fitted in a multiple regression model. Using a Wald test to determine significance at 0.05, a backward stepwise regression approach was used to determine the final model.
A total of 1644 male military personnel were recruited in the study, of whom 2 (0.1%) refused to participate in the survey, and 4 (0.2%) did not provide information on their condom use behaviors and were excluded from analysis. From the 1638 remaining participants, 40% reported never having had sex with FSWs, 55% had used condom(s) at last sex with FSW, and 5% had unprotected sex at their last encounter with an FSW.
The Cambodian military participants were mostly married men (85%) with a mean age of 37 years (95% CI, 36.3–37.7 years) who reported an average of 6.6 years of schooling (95% CI, 5.1–6.1 year). The most common rank of the personnel was petty officer (corporal to sergeant), reported by 47% of respondents. Overall, 23% of military had been trained as peer educator.
Table 1 compares the demographics among personnel who had never had sex with sex workers, personnel who used a condom at last sex with an FSW, and those who were unprotected at last sex with an FSW. Respondents aged 30 to 39 years were significantly more likely to have visited brothels than other age groups combined (47% vs. 37%, P <0.001). However, participants who had unprotected sex with sex workers were significantly older (mean age 39.3 years vs. 36.8 years, P = 0.002) and had their sexual debut in the more distant past than condom users (mean number of years since first sex 18.8 vs. 15.3, P <0.001). No significant difference was found regarding other demographic characteristics. Condom users were more likely than nonusers to have answered correctly to all 10 questions about HIV transmission (83% vs. 71%, P = 0.02).
Among military men who acknowledged ever having had sex with FSWs, 21% reported having used more than 1 condom at a time at their last intercourse with an FSW. Of this subsample, 91% had used 2 condoms and 9% had used 3 condoms.
Univariate analysis of the factors associated with the SUMC among respondents who acknowledged having had sex with FSWs (n = 997) is presented in Table 2, together with the multiple regression analysis model. Compared with single users of condoms, multiple condom users were younger (31% of multiple condom users were aged less than 30 years compared to 15% of single condom users, P <0.001), more likely to have had sex for the first time in the past 6 years, more likely to have visited more than 1 FSW in the past month, and more likely to have ever sought their HIV serostatus. In addition, multiple condom users were less likely to know all correct answers to the 10 questions on HIV transmission, less likely to be willing to talk about family planning with their wives, and less likely to have been trained as a peer educator. Moreover, multiple users of condoms were more likely to be private than those ranking lieutenant or higher and more likely to have fewer than 2 years of schooling compared with those who had more than 6 years of schooling. Finally, multiple users had been exposed to fewer HIV peer-education sessions organized by peers, as they were less likely than others to have ever received at least 6 peer-education sessions.
Multiple Logistic Regression Results
A regression model on the risk of SUMC is presented in Table 2. Variables adjusted for in the model include age, military rank, number of FSWs visited in the past month, ever sought own HIV serostatus, knowledge of HIV transmission, and whether respondent had ever been trained as peer educator. Respondents younger than 30 years were 2.54 times more likely than those aged 40 years and older to simultaneously use more than 1 condom at a time. Respondents who had visited more than 1 sex worker in the past month were 2.44 times more likely to use multiple condoms at a time than those who had visited 1 or no FSW in the past month. Those who had ever sought their HIV serostatus were 1.79 times more likely to use multiple condoms than those never tested for HIV. Those who had been trained as peer educators were 0.50 times less likely to have used multiple condoms at last sex with FSWs than the others.
The results of this analysis demonstrate that SUMC is a common practice among Cambodian military men from region-2 who visit FSWs.
Compared with users of a single condom at last sex with an FSW, multiple condom users were younger, were more likely to be in the lowest socioeconomic strata (as measured by military rank), had initiated their sexuality more recently, and had visited more FSWs in the past month. In Thailand, SUMC was estimated to be 50.1% based on analyzed sex acts and was reported as a spontaneous strategy to reduce incidence of condom breakage.10 However, the Thai study was based on information collected from FSW exclusively and did not explore the characteristics of clients practicing SUMC. In Cambodia, condom failure among sex workers is frequent, as 27% of FSWs’ clients reported having experienced condom breakage/leakage.14 Other studies have shown that a small proportion of condom users are more prone to recurrent condom breakages.14,15 Men experiencing condom breakage in the literature present some similarities with the multiple users of condoms reported in this study, including young age, low educational status, low socioeconomic status, short time since first sexual experience, and having multiple partners.16–23 However, our study could not show any relationship between SUMC and condom breakage because information on past experience of condom breakage was not collected. Additional data are needed to determine whether SUMC is related to previous experience of condom failures in Cambodia.
Peer educators were more educated, of higher rank, and older age than other military because of selection criteria within the military hierarchy.12 Nevertheless, multiple condom users remained less likely to have been trained as peer educators after adjusting, for age, education and military rank. Although the peer educators’ training curriculum did not address SUMC, peer educators may follow the norms they were taught when participating in condom demonstrations using a single condom.
Uptake of VCT services is usually associated with high levels of HIV transmission knowledge and educational status.24–26 Similar patterns were observed among Cambodian military personnel; overall, those who answered 9 out of 10 questions correctly on HIV transmission were significantly more likely to have ever sought an HIV test than those with lower scores (22% vs. 11%, P = 0.02).12 However, compared with single condom users, those reporting SUMC were more likely to have sought their HIV serostatus despite having a lower knowledge of HIV transmission. The uptake of VCT services in developing countries where these are not widely available is mostly motivated by having engaged in high-risk behavior or appearance of HIV-related symptoms.27,28 These findings suggest that simultaneous users of multiple condoms felt vulnerable to HIV despite a limited knowledge of HIV transmission.
The study did not assess whether FSWs or military clients proposed SMUC before intercourse. SMUC is an economic burden to clients because they usually purchase the condoms at the brothel. If the FSW proposed SMUC, it is hypothesized that only clients with specific personal characteristics would agree to have sex in this manner. In our study, if simultaneous users of multiple condoms indeed fear HIV infection, they may be doing so because they more frequently report engaging in sex with FSWs than single condom users. As their natural coping strategy is not geared toward avoiding high-risk multiple partnerships, they have instead adapted a measure that has never been promoted by any HIV prevention program, i.e., SUMC. The study from Thailand found a lower incidence of condom breakage among simultaneous multiple condoms users than among single condom users.9 The researchers believed they had captured the result of a spontaneous “natural experiment” by which FSWs and their clients had reduced the incidence of condom breakages through the SUMC.10 Such a community response to condom promotion messages and to fear of HIV could have been imported from Thailand by migrant workers or could have been reinvented by Cambodians. In any case, the widespread diffusion of this prevention message appeared to have happened informally through word of mouth. Regardless of whether FSWs or military clients typically initiate SUMC, the study results indicate that individuals with low education attainment are capable of empirically discovering a protective mechanism to help cope with the HIV epidemic.
This study has several limitations. First, SUMC may be overestimated, as participants might have been reporting sequential use of condoms instead of SUMC, although the wording of the question clearly specified that condoms were used 1 on top of the other. Second, this study was restricted to an occupational group from a Cambodian region, and the results are probably not generalizable to all clients of Cambodian FSWs.
Although a single observational study9 found that SUMC was associated with decreased risk for condom breakage, it is commonly accepted in the scientific community that friction between 2 condoms used on the top of one another would lead to increased condom failure including breakage. Guidelines available on the Internet recommend avoiding such condom use behavior because it is believed to decrease the protective effectiveness of condoms.5,29–31 Further studies are urgently needed to help understand the issues around multiple condom use and to advise condom users and planners of condom promotion programs accordingly.
In Cambodia, SUMC is common practice between FSWs and their military clients. The simultaneous users of multiple condoms present similar demographic characteristics to men who often experience condom breakages. Additional data are needed to confirm if adoption of this strategy is related to previous experience of condom failures in Cambodia. As it is unknown if SUMC increases or decreases condom effectiveness, more research is needed to investigate the effect of SUMC on condom failure rates.
1. Heng S, Morineau G, Gorbach P. Cambodia 2003 Behavioral Surveillance Survey. Phnom Penh: NCHADS/FHI/USAID, 2005.
2. National Center for HIV/AIDS Dermatology and STDs. Report on HIV Sentinel Surveillance in Cambodia 2003. Phnom Penh, Cambodia: NCHADS, CDC-GAP, FHI, 2006.
3. Wolitski RJ, Halkitis PN, Parsons JT, et al. Awareness and use of untested barrier methods by HIV-seropositive gay and bisexual men. AIDS Educ Prev 2001; 13:291–301.
9. Rugpao S, Beyrer C, Tovanabutra S, et al. Multiple condom use and decreased condom breakage and slippage in Thailand. J Acquir Immune Defic Syndr Hum Retrovirol 1997; 14:169–173.
10. Rugpao S, Tovanabutra S, Beyrer C, et al. Multiple condom use in commercial sex in Lamphun Province, Thailand: A community-generated STD/HIV prevention strategy. Sex Transm Dis 1997; 24:546–549.
11. Population Service International/Cambodia. Knowledge, Attitudes and Behaviors Regarding HIV/AIDS in Cambodia: Results of a National Survey. Phnom Penh, Cambodia: PSI, 2004.
12. Thou T, Tan S, Morineau G, et al. Evaluation of HIV/AIDS Peer Education Prevention Program Among the Military in Military Region 2. Phnom Penh, Cambodia: Family Health International/Cambodia Ministry of National Defense, 2005.
13. Lauritsen JM, ed. EpiData Data Entry, Data Management and Basic Statistical Analysis System. Odense, Denmark: EpiData Association, 2000–2006.
14. Hor LB, Detels R, Heng S, et al. The role of sex worker clients in transmission of HIV in Cambodia. Int J STD AIDS 2005; 16:170–174.
15. Sparrow MJ. Condom failures in women presenting for abortion. NZ Med J 1999; 112:319–321.
16. Trussell J, Warner DL, Hatcher RA. Condom slippage and breakage rates. Fam Plann Perspect 1992; 24:20–23.
17. Macaluso M, Kelaghan J, Artz L, et al. Mechanical failure of the latex condom in a cohort of women at high STD risk. Sex Transm Dis 1999; 26:450–458.
18. Grady WR, Tanfer K. Condom breakage and slippage among men in the United States. Fam Plann Perspect 1994; 26:107–112.
19. Messiah A, Dart T, Spencer BE, et al. Condom breakage and slippage during heterosexual intercourse: A French national survey. French National Survey on Sexual Behavior Group (ACSF). Am J Public Health 1997; 87:421–424.
20. Civic D, Scholes D, Ichikawa L, et al. Ineffective use of condoms among young women in managed care. AIDS Care 2002; 14:779–788.
21. Mekonen E, Mekonen A. Breakage and slippage of condoms among users in north Gondar Province, Ethiopia. East Afr Med J 1999; 76:481–483.
22. Spruyt A, Steiner MJ, Joanis C, et al. Identifying condom users at risk for breakage and slippage: Findings from three international sites. Am J Public Health 1998; 88:239–244.
23. Valappil T, Kelaghan J, Macaluso M, et al. Female condom and male condom failure among women at high risk of sexually transmitted diseases. Sex Transm Dis 2005; 32:35–43.
24. Adewole DA, Lawoyin TO. Characteristics of volunteers and non-volunteers for voluntary counseling and HIV testing among unmarried male undergraduates. Afr J Med Med Sci 2004; 33:165–170.
25. Iliyasu Z, Kabir M, Galadanci HS, et al. Awareness and attitude of antenatal clients towards HIV voluntary counselling and testing in Aminu Kano Teaching Hospital, Kano, Nigeria. Niger J Med 2005; 14:27–32.
26. Alwano-Edyegu MG, Marum E, Wheeler M. Knowledge is Power: Voluntary HIV Counseling and Testing in Uganda. Geneva, Switzerland: UNAIDS, 1999.
27. Solomon S, Cecelia AJ, James R, et al. Why are people getting tested? Self-reported reasons for seeking voluntary counseling and testing at a clinic in Chennai, India. AIDS Behav 2006; 22:22.
28. Nuwaha F, Kabatesi D, Muganwa M, et al. Factors influencing acceptability of voluntary counselling and testing for HIV in Bushenyi district of Uganda. East Afr Med J 2002; 79:626–632.