Abstract
Objectives. We examined the efficacy of a brief behavioral intervention to promote condom use among female sex workers in Tijuana and Ciudad Juarez, Mexico.
Methods. We randomized 924 female sex workers 18 years or older without known HIV infection living in Tijuana and Ciudad Juarez who had recently had unprotected sex with clients to a 30-minute behavioral intervention or a didactic control condition. At baseline and 6 months, women underwent interviews and testing for HIV, syphilis, gonorrhea, and chlamydia.
Results. We observed a 40% decline in cumulative sexually transmitted illness incidence (P = .049) in the intervention group. Incidence density for the intervention versus control groups was 13.8 versus 24.92 per 100 person-years for sexually transmitted illnesses combined (P = .034) and 0 versus 2.01 per 100 person-years for HIV (P < .001). There were concomitant increases in the number and percentage of protected sex acts and decreases in the number of unprotected sex acts with clients (P < .05).
Conclusions. This brief behavioral intervention shows promise in reducing HIV and sexually transmitted illness risk behaviors among female sex workers and may be transferable to other resource-constrained settings.
Although commercial sex is quasi legal in the zonas rojas (red light districts) of most Mexican cities, HIV prevalence among female sex workers has been low in most parts of the country.1 For example, among female sex workers in Mexico City and at the Mexico–Guatemala border, HIV prevalence was 0.6% in 19972 and in 1998.3 More recently, reports suggest that HIV prevalence is rising among female sex workers on the Mexico–US border. In a sample of 415 female sex workers in Tijuana in 1991, HIV prevalence was 0.5%,4 but in our more recent study of 924 female sex workers in Tijuana (bordering San Diego, CA) and Ciudad Juarez (bordering El Paso, TX), HIV prevalence was 6.0%.5 In the same study, prevalences of gonorrhea, chlamydia, and syphilis titers consistent with active infection were 6.4%, 13.0%, and 14.2%, respectively.
In Mexican culture, sex with female sex workers is tolerated as a display of virility and machismo at all socioeconomic levels.6–8 The sexually permissive environment in Mexican border cities also attracts large numbers of “sex tourists” from the United States and other countries who are attracted by newspaper advertisements and Web sites advertising prostitution.9 More than two thirds of female sex workers in Tijuana and Ciudad Juarez report having US clients, and these women have a higher prevalence of sexually transmitted infections (STIs) and higher levels of transmission risk behaviors, including sex without a condom and injection use of drugs.10
Mexican law requires female sex workers to obtain a permit if they wish to work without prosecution in Tijuana's zona roja, but more than half work without permits. In Ciudad Juarez, permits are not required. In both cities, female sex workers operate out of multiple venues including cantinas, bars, nightclubs, motels, and street corners. Seventy percent of these female sex workers were born in other Mexican states or Central America,11 and most entered prostitution out of economic necessity. Extreme poverty may compromise their ability to turn down offers of higher payments for unprotected sex.12,13
In border regions, where sex work is tolerated and HIV prevalence is rising, there is a clear need to ensure that condoms are consistently used during commercial sex transactions to prevent outbreaks of HIV or STIs in both countries. We developed a brief behavioral intervention to enhance condom use negotiation among Mexican female sex workers.11 Here, we report its first results and examine its efficacy in Tijuana and Ciudad Juarez.
Tijuana, Mexico's largest city on its border with the United States, is home to an estimated 1 410 700 persons. Tijuana and its US neighbor, San Diego, California, form the world's largest binational metropolis. Roughly half of Baja California's population lives in Tijuana, although over half of the city's inhabitants were born outside the state.14 In 1999, the per capita gross regional product in Tijuana was $6800, less than one quarter of that for San Diego.15 The border crossing between Tijuana and San Diego is the busiest in the world. In 2005, border officials tallied over 45 million northbound crossings from Tijuana to San Diego County,16 and 42 000 persons who live in Tijuana head north to San Diego every day.17 Baja California ranks second among Mexico's 32 states in terms of cumulative AIDS incidence.18 Estimates of the number of female sex workers in Tijuana vary, but the most widely cited figure is 9000.15
Ciudad Juarez, the largest city in the state of Chihuahua, has a population of 1 313 338.14 In 2000, 36% of Ciudad Juarez's inhabitants had been born outside Chihuahua. In 1999, the per capita gross regional product of Ciudad Juarez was only 41.2% that of El Paso, Texas ($7074 vs $17 216).15 In 2005, officials recorded 29 million border crossings from Ciudad Juarez to El Paso.16 As in Tijuana, the main employers in Ciudad Juarez are the maquiladoras (assembly plants). Chihuahua ranks 14th in cumulative AIDS incidence among Mexican states.18 It is estimated that there are at least 4000 female sex workers in Ciudad Juarez.19
Our sample consisted of 924 female sex workers we recruited into a behavioral intervention study to increase condom use in Tijuana and Ciudad Juarez between January 2004 and January 2006.11 Eligibility requirements included being aged at least 18 years, giving informed consent, and having traded sex for drugs, money, or other material benefit within the previous 2 months. Participants were also required to have had unprotected vaginal sex with at least 1 client in the previous 2 months. They were excluded if they reported they had previously tested HIV positive. Outreach workers and municipal and community health clinics recruited the participants. Trained female counselors administered baseline and follow-up interviews in Spanish in private clinic rooms or outreach offices. Participants also provided a blood draw and cervical swab and were compensated US $30 at baseline and follow-up visits.
The baseline interview covered a range of topics, including sexual risk behaviors, working conditions, financial need, victimization and trauma, use of alcohol and illicit drugs, social support, social influence, life experiences, mood, self-esteem, social cognitive factors, sociodemographic characteristics, and physical and psychiatric health variables. It also addressed such behavioral outcomes as frequency of unprotected sexual acts with clients and spouse or steady partner, number of clients, number and type of nonclient sexual partners, self-reported number and type of STIs, and alcohol and drug use. The follow-up survey was nearly identical but referred to behaviors taking place over the previous 6 months.
We randomized the participants to either the intervention (Mujer Segura [Healthy Woman] counseling session) or the didactic control group weekly within each city using a fixed, computer-generated randomization scheme.20
To develop the intervention materials, we drew extensively on our experience from a pilot study conducted in Tijuana21 and from other sexual risk reduction interventions among high-risk populations in the United States.22,23 The content of the Mujer Segura intervention was tailored to the needs, values, beliefs, and behaviors of our target population; specially trained, local health care staff and indigenous promotoras (outreach workers) performed the intervention. The intervention took about 35 minutes. A detailed description of the protocol has been published elsewhere.11 In keeping with the intervention's basis in social cognitive theory,24 we aimed to increase participants’ knowledge, self-efficacy, and outcome expectancies regarding safer sexual behavior.
In addition, we used motivational interviewing techniques (e.g., key questions, reflective listening, summarization, affirmation, and appropriate use of cultural cues) to elicit information on the participants’ current situation and motivations and to increase their motivation to practice safer sex.25 Four main areas were addressed: (1) motivations for practicing safer sex (e.g., to protect one's own health, to avoid STIs, to feel clean) versus those for practicing unsafe sex (e.g., financial gain); (2) barriers to condom use (e.g., threats of physical violence); (3) techniques for negotiating safer sex with clients; and (4) enhancement of social supports. Our pilot findings guided the emphases of this study; for example, female sex workers in the pilot indicated that the most important motivator of behavior change was the desire to protect their health to be able to continue supporting their children.21
As we addressed each area, we used role-playing to help participants improve their self-efficacy in interactions with clients. Then, in an application of the “decisional balance” approach of Miller and Rollnick,25 we asked participants to weigh the advantages and disadvantages of practicing or negotiating the practice of safer sex. Once the participant began to exhibit awareness of the problem and a shift toward positive change, the counselor helped her develop a plan of action that best suited her personal situation; then together they identified barriers to implementation of the plan and ways to overcome them.
Because of the high-risk nature of this population, we chose an active comparison group that provided information critical to HIV and STI prevention. The comparison group was a face-to-face, time-equivalent didactic presentation of prevention materials extracted from the US Centers for Disease Control and Prevention revised guidelines for HIV counseling, testing, and referral26 and from Mexico's National Center for AIDS Studies.27 The focus of the counseling session was on personal risk assessment, cultural identity assessment, and strategies for reducing personal risk. During the personal risk assessment, the counselor helped the participant identify, understand, and acknowledge behaviors and circumstances that put her at risk for contracting HIV and STIs, explored the participant's previous attempts to reduce personal risk, and provided positive reinforcement for steps already taken.
The counselor also helped the participant to set small, achievable risk-reduction goals and offered concrete suggestions. Basic educational information (e.g., HIV and STI transmission modes) was provided if the woman's level of knowledge was low; however, the primary focus was on the discussion of transmission risk associated with specific behaviors or activities that were relevant to the participant's personal risk. Lower risk alternatives (e.g., oral sex) were promoted. Motivational techniques, role-playing, and discussion of specific techniques for negotiating safer sex were avoided.
Specimen testing was conducted at either the San Diego County Health Department (for Tijuana STI samples and all HIV confirmatory tests) or the El Paso County Health Department (for Ciudad Juarez STI samples). We administered the Determine Rapid HIV Antibody (Abbott Pharmaceuticals, Boston, MA) test to determine the presence of HIV antibodies. We then tested all reactive samples using HIV-1 antibody by enzyme immunoassay (EIA) and Western blot. For syphilis serology, we used the rapid plasma reagin test (Macro-Vue, Becton Dickenson, Cockeysville, MD).
We subjected all rapid plasma reagin–positive samples to confirmatory testing using the Treponema pallidum hemagglutinin assay (Fujirebio, Wilmington, DE). Neisseria gonorrhea and Chlamydia trachomatis were detected from vaginal swabs that trained nurses collected using the Aptima Combo 2 collection device (Genprobe, San Diego, CA), which allows a direct target-amplified nucleic acid probe test. We provided participants with their HIV and STI test results and referred those testing positive to municipal health clinics for free medical care. Study protocols were approved by the institutional review boards in the United States and Mexico.
Behavioral outcomes focused on the following behaviors with clients: frequency of unprotected sex, protected sex ratio (number of protected sex acts to total number of sex acts), number and type of STIs, and incidence of HIV and STIs. Female sex workers also reported the number of times they engaged in vaginal, oral, and anal sex without a condom.
We first examined group differences in demographic and baseline characteristics of our sample using the t test and χ2 analysis to compare groups on linear and dichotomous variables, respectively. To examine change in sexual activity outcomes from pre- to postintervention, for each outcome, we conducted a repeated measures analysis of covariance with intervention condition as our primary independent variable. To control for preintervention sexual activity, baseline values of our outcome variable were entered as a covariate. Intervention site (Tijuana vs Ciudad Juarez) was also entered as a covariate.
We calculated the Cohen d effect size as an indication of the magnitude of effect, whereby positive values indicated superiority of the intervention.28 The Cohen d effect size represents the magnitude of differences between 2 groups in standard deviation units. An effect size (standard deviation difference) of 0.2 is traditionally considered small, 0.5 is medium, and 0.8 is large.11 Because both groups were active interventions, d values of 0.2 or greater were interpreted as clinically meaningful, as reported in the literature.29,30
Our second set of analyses examined the efficacy of the intervention for reducing incident HIV and other STIs. To take the most conservative approach, the χ2 test compared only those participants who had negative test results for the respective STI at the baseline assessment. As an indication of effect size, we calculated the number needed to treat for each STI outcome. Number needed to treat represents the number of participants who would need to be treated before 1 fewer person contracted the STI than would be the case had all participants been in the control group.31
Finally, we compared groups in terms of cumulative incidence and incidence density. In both cases, the numerator was the number of women who acquired the STI in question during follow-up. For cumulative incidence, the denominator was the total number of at-risk women, whereas for incidence density, the denominator was the total number of person-years for at-risk women. For each HIV and STI outcome, Poisson regression was used to determine if group differences were statistically significant.
A total of 924 eligible female sex workers were enrolled (474 in Tijuana and 450 in Ciudad Juarez), of whom 55 (6%) tested HIV positive and 869 (94%) tested HIV negative. The average age was 33.5 years (interquartile range: 26.5–40.5), a minority spoke English (19.3%), and the average duration of formal education was low (mean: 6 years; interquartile range: 4–8). Most women were married or living in common-law relationships, and 93.7% had children (Table 1). Participants in the intervention group were significantly more likely to have syphilis titers consistent with active infection and reported fewer children than did those in the control group but did not differ with respect to other characteristics (all other P values > .05).
TABLE 1 Study Participant Demographics and Baseline Characteristics, by Intervention Condition: Mujer Segura Intervention Study, Tijuana and Ciudad Juarez, Mexico, January 2004–January 2006
| Variable | Intervention (n = 409) | Control (n = 460) | P |
| Age, y, mean (SD) | 33.6 (9.5) | 33.4 (8.9) | .820 |
| Education ( > 9 y), no. (%) | 43 (10.6) | 41 (9.0) | .435 |
| Total children, mean (SD) | 2.8 (1.8) | 3.1 (1.9) | .022 |
| Has child(ren), no. (%) | 376 (92.4) | 435 (95.0) | .115 |
| Years worked in sex trade, mean (SD) | 6.7 (7.1) | 5.9 (6.6) | .094 |
| Years lived in study location, mean (SD) | 15.8 (13.0) | 16.8 (13.2) | .268 |
| Ever injected drugs, no. (%) | 66 (16.1) | 88 (19.1) | .249 |
| Hours worked per week, mean (SD) | 42.2 (23.8) | 40.8 (23.0) | .373 |
| Outcomes | |||
| Had vaginal sex past month, mean (SD) | 57.0 (59.8) | 54.8 (55.2) | .587 |
| Had anal sex past month, % (SD) | 80 (19.8) | 94 (20.5) | .791 |
| Had anal sex past month,a mean (SD) | 8.4 (13.3) | 9.6 (16.5) | .603 |
| Had vaginal and anal sex past month, mean (SD) | 58.6 (60.7) | 56.8 (56.6) | .651 |
| STIs, no. (%) | |||
| HIV positive | 30 (6.8) | 25 (5.2) | .281 |
| Lifetime syphilis | 132 (30.2) | 117 (24.4) | .047 |
| Syphilis titers ≥ 1:8 | 60 (13.8) | 34 (7.1) | .001 |
| Gonorrhea | 25 (7.0) | 26 (5.9) | .501 |
| Chlamydia | 48 (13.5) | 56 (12.6) | .714 |
| Any STIb | 131 (29.8) | 118 (24.3) | .059 |
Note. STI = sexually transmitted infection.
aMean values are for participants who reported having anal sex.
bIncludes women with HIV, syphilis titers at or above 1:8, chlamydia, or gonorrhea.
Because the primary outcome was to reduce HIV incidence, the 55 HIV-positive female sex workers were excluded from subsequent analysis. Of the remaining 869 women, 709 (81.6%) were assessed at the 6-month postintervention assessment (intervention: 341; didactic control: 368; see the figure available as a supplement to the online version of this article at http://www.ajph.org). Participants who did not complete the postintervention assessment were significantly younger (31.0 vs 34.0 years; P < .001), had worked fewer years in the sex trade (4.2 vs 6.8 years; P < .001), and had lived fewer years in their respective study location (13.2 vs 16.9 years; P = .001) but did not differ in terms of any other demographic or sexual risk behaviors (all other P > .05).
Individuals in the intervention condition reported significant improvements in their risk behaviors (Table 2). Compared with those in the didactic control condition (N = 337), participants in the intervention condition (N = 307) demonstrated significant increases in total number of protected sex acts (P < .05; Cohen d = .16), significant decreases in the total number of unprotected sex acts (P < .05; Cohen d = .17), and significant increases in the percentage of sex acts during which a condom was used (P < .05; Cohen d = .25). Repeating these analyses with the 55 HIV-infected women included in the sample led to no appreciable differences.
TABLE 2 Pre- and Postintervention Sexual Activity Outcomes, by Intervention Condition: Mujer Segura Intervention Study, Tijuana and Ciudad Juarez, Mexico, January 2004–January 2006
| Intervention (n = 307)a | Control (n = 337)a | |||||
| Preintervention | Postinterventionb | Preintervention | Postinterventionb | Fc | P | |
| Total protected sex, mean (SD) | 39.1 (47.9) | 51.6 (49.1) | 38.7 (49.4) | 43.9 (49.1) | 3.98 | .047 |
| Total unprotected sex, mean (SD) | 23.8 (32.0) | 8.2 (28.2) | 20.9 (29.5) | 12.9 (28.2) | 4.41 | .036 |
| Condom use, % (SD) | 56.3 (32.1) | 83.7 (32.9) | 58.0 (30.5) | 75.5 (32.9) | 9.78 | .002 |
aTotal numbers reflect cases with both baseline and follow-up (6-month) data; 101 participants in the intervention and 122 participants in the control group were missing follow-up sexual risk data. Also, 1 participant in the intervention group and 1 participant in the control group were missing baseline sexual risk data.
bPostintervention values represent estimated marginal means after adjusting for baseline values and study site.
cF test values are for treatment multiplied by interaction effects after we controlled for baseline values and intervention site.
A significant group difference was observed for overall STIs (P = .049; Table 3); however, significant results were not observed for individual STIs. For any STI including active syphilis, the number needed to treat was 20, indicating that 20 women would need to receive the intervention before 1 reduction in STI would be observed.
TABLE 3 Cumulative Incidence of Sexually Transmitted Infections (STIs) at 6-Month Follow-up in Intervention and Control Conditions: Mujer Segura Intervention Study, Tijuana and Ciudad Juarez, Mexico, January 2004–January 2006
| STI | Intervention (n = 341), No. (%) | Didactic Control (n = 368), No. (%) | P | RR (95% CI) | No. Needed to Treat |
| Incident HIV infection | 0 (0.0) | 4 (1.1) | .125 | … | … |
| Lifetime syphilis | 10 (4.2) | 14 (4.9) | .691 | 0.85 (0.39, 1.88) | 143 |
| Syphilis titers ≥ 1:8 | 6 (2.1) | 14 (4.2) | .137 | 0.50 (0.19, 1.28) | 48 |
| Gonorrhea | 9 (3.4) | 14 (4.3) | .602 | 0.80 (0.35, 1.83) | 111 |
| Chlamydia | 13 (5.3) | 17 (5.6) | .883 | 0.95 (0.47, 1.92) | 333 |
| Any STIa | 20 (7.7) | 38 (12.8) | .049 | 0.60 (0.36, 1.00) | 20 |
Note. RR = relative risk; CI = confidence interval. Ellipses indicate that values could not be computed because none of the participants contracted HIV. Degree of freedom for all analyses was 1. Cumulative incidence was defined as the total number of women with incident infections occurring during follow-up divided by the total number of women at risk.
aIncludes women with HIV, syphilis titers at or above 1:8, chlamydia, or gonorrhea.
In terms of cumulative incidence (Table 3), we similarly did not observe significant differences for HIV or specific STIs, likely because of low power; however, a 40% decrease in overall STI incidence was observed in the intervention group relative to the didactic control (i.e., 20% vs 38%; relative risk = .60; 95% confidence interval = 0.36, 1.00). In terms of incidence density (Table 4), HIV incidence was 2.01 per 100 person-years in the didactic control group compared with 0 in the intervention group (P = .001), and overall STI incidence was also significantly lower in the intervention group (13.80 vs 24.92 per 100 person-years; P = .034).
TABLE 4 Incidence Density for Sexually Transmitted Infections (STIs) Outcomes Among Female Sex Workers, by Intervention Condition: Mujer Segura Intervention Study, Tijuana and Ciudad Juarez, Mexico, January 2004–January 2006
| STI | Incidence Density per 100 Person-Years (95% CI) | P |
| Incident HIV infection | <.001 | |
| Intervention | 0.00 (0.00, 0.00) | |
| Control | 2.01 (0.04, 3.97) | |
| Syphilis titers ≥ 1:8 | .114 | |
| Intervention | 3.62 (0.72, 6.51) | |
| Control | 7.75 (3.69, 11.80) | |
| Gonorrhea | .584 | |
| Intervention | 6.33 (2.19, 10.47) | |
| Control | 8.00 (3.81, 12.19) | |
| Chlamydia | .878 | |
| Intervention | 9.92 (4.53, 15.31) | |
| Control | 10.50 (5.51, 15.50) | |
| All STIsa | .034 | |
| Intervention | 13.80 (7.75, 19.85) | |
| Control | 24.92 (17.00, 32.85) |
Note. CI = confidence interval. Incidence density is defined as the total number of at-risk women with incident infections during follow-up divided by the number of person-years at risk.
aIncludes women with HIV, syphilis titers at or above 1:8, chlamydia, or gonorrhea.
In this study, we found that a brief behavioral intervention integrating motivational interviewing and principles of behavior change significantly reduced HIV and overall STI incidence and unprotected sex with male clients among high-risk female sex workers in 2 Mexico–US border cities. Overall, cumulative STI incidence decreased 40% in the intervention group relative to the didactic control. In particular, there was not a single incident HIV infection in the intervention group, whereas HIV incidence was 2.01 per 100 person-years in the control group. It is noteworthy that our intervention demonstrated significant effects, because our didactic control group was itself an educational intervention that was adapted from binational counseling guidelines.
A more passive control condition would likely have generated greater effect sizes. Although the effect sizes we observed tended to be relatively modest for individual STIs and behavioral outcomes, they are within the same range as that reported in meta-analyses of behavioral HIV interventions.32,33 In populations in which there is considerable sexual mixing and prevalence of ulcerative STIs is high (because of high incidence or prolonged duration of infection), even interventions with modest effect sizes may have an important impact on HIV and STI incidence at the community level.34 Our findings, therefore, have important implications for Mexico and other resource-constrained settings.
Our study adds to the growing body of literature on interventions to reduce HIV and STIs and associated high-risk behaviors among female sex workers. In regions in which STIs are prevalent, presumptive STI treatment among female sex workers has been evaluated as an HIV prevention measure with mixed, and often short-term, results.35,36 The “100% condom campaign”—which requires brothels to enforce strict regulations for consistent condom use or risk losing their license—has been successful in Thailand37 and the Dominican Republic.38 However, marginalized Thai female sex workers who started sex work later and did not work in brothels benefited less.38
A multilevel intervention to reduce HIV and STI incidence among female sex workers showed efficacy in Calcutta38 but has proven difficult to replicate elsewhere.38 These experiences suggest that cultural, social, and contextual conditions that vary between and even within countries must be taken into account in the development of interventions for female sex workers. An advantage to our approach is that it can be tailored not only to the cultural context but also to each woman's personal situation because the woman actively identifies barriers to safer sex in her own life and potential solutions as part of her goal setting with the counselor.11,21
Previous work supports the notion that behavioral interventions can reduce high-risk behavior and STIs among ethnic minorities. For example, in a study by Shain et al.,39 Mexican American and African American women randomized to three 4-hour intervention sessions experienced significantly reduced STI incidence and risk behaviors. DiClemente et al.40 reported that African American adolescent girls who received four 4-hour group interventions reported consistent use of condoms and had reduced STI incidence. In a meta-analysis, Crepaz et al.32 concluded that successful interventions are informed by ethnographic studies, are theory based, focus on skill building, and are tailored to the appropriate cultural context. In their meta-analysis, Johnson et al.33 showed that motivational and skill-based interventions were especially efficacious. Our study extends previous findings by demonstrating that a brief single-session intervention incorporating these elements is efficacious in reducing HIV, STIs, and high-risk behavior among Mexican female sex workers.
In the interpretation of our findings, a number of limitations should be borne in mind. Follow-up was relatively short (i.e., 6 months), and effects may wane over time. However, we recently reported that a similar brief intervention based on the same theoretical principles led to sustained improvements among men who have sex with men22 and HIV-negative heterosexual men and women23 over periods of 12 and 18 months, respectively. Although our study incorporated randomization and achieved balance across most baseline characteristics, participants randomly assigned to the intervention condition had a higher prevalence of syphilis titers consistent with active infection.
In analyzing these data, we took the most conservative approach by considering women at risk of incident syphilis only if they were negative for syphilis at baseline. Although our study recruited female sex workers from 2 Mexico–US border cities, our findings may not be generalizable outside these cities. Because STI incidence remained unacceptably high even in the intervention group, additional efforts are needed, such as interventions aimed at clients and nonpaying partners (e.g., spouses or boyfriends).
Although there are a number of candidate HIV vaccines and vaginal microbicides under development,11 neither is likely to be available in the near future, and it would be unrealistic to anticipate 100% efficacy in either case. The urgent need continues for effective, culturally appropriate interventions that can be used as stand-alone programs or as adjuncts to existing approaches. Our brief intervention, which counselors with little or no formal education can be trained to deliver in a variety of settings, is an inexpensive approach to HIV and STI risk reduction that may be transferable to other resource-constrained settings.
Acknowledgments
Funding for this study was provided by the National Institute of Mental Health (grant R01 MH065849).
The authors gratefully acknowledge the study staff and participants, Brian Kelly for editing assistance, and the following organizations for their cooperation: the Municipal and State Health Departments of Tijuana, Baja California and Ciudad Juarez, Chihuahua; Patronato Pro-COMUSIDA; Salud y Desarollo Comunitario de Ciudad Juarez A. C. (SADEC) and Federación Méxicana de Asociaciones Privadas (FEMAP); and the Universidad Autónoma de Baja California (UABC) and Universidad Autónoma de Ciudad Juárez (UACJ). In addition, we would like to thank the County Health Departments of San Diego, California, and El Paso, Texas, for their assistance with sexually transmitted infection and HIV testing.
Human Participant Protection
The protocol for the research study on which this article was based was reviewed and approved by the institutional review board of the Human Research Protections Program at the University of California, San Diego (project no. 051182).
