THE HIV/AIDS PANDEMIC CONTINUES to grow. UNAIDS estimates that there were 39.4 million persons living with HIV/AIDS as of December 2004 with 4.9 million persons newly infected and 3.1 million deaths during 2004.1 Among countries most affected, the adverse socioeconomic and demographic consequences are considerable.2 As many as 45 million people globally could be newly HIV-infected by 2010 unless prevention efforts are improved.3
The Caribbean is the second most affected region in the world after sub-Saharan Africa with an estimated 440,000 persons living with HIV/AIDS and an average adult HIV prevalence rate of 2.3%.1 AIDS is the leading cause of death among adults 15 to 44 years in the Caribbean and a leading cause of death among children 1 to 4 years of age in countries like Jamaica.4 The adult HIV prevalence rate for adults 15 to 49 years in Jamaica is estimated at 1.5%. HIV transmission in Jamaica is predominantly heterosexual. Homosexual or bisexual transmission is estimated at 8% of reported male AIDS cases but is probably higher because many of the males classified in the unknown transmission category (17% of all reported cases) may be bisexual.5
Jamaica has made significant progress in addressing the HIV/AIDS epidemic.6–8 A comprehensive HIV/sexually transmitted infection (STI) control program has been established since the late 1980s. General awareness of HIV and knowledge on how to prevent it is high among the population. The blood supply is safe and HIV transmission through blood is uncommon. STI services were expanded and reported cases of infectious syphilis have declined from 90 per 100,000 in 1987 to 6.2 per 100,000 in 2001.8 Congenital syphilis cases have declined from 68 in 1994 to 21 in 2001.8 Condom distribution has increased 4-fold from 2.5 million in 1985 to 10.8 million in 1999 and 9.4 million in 2002 with an increasing proportion being sold (from 20–70% in the same period).8,9 Nationally representative KAP surveys in Jamaica show that between 1992 and 2000, approximately 75% of men having sexual intercourse with a nonregular partner used a condom at last sex. Among women having sex with a nonregular partner, condom use at last sex increased from 37% in 1992 to 67% in 2000.10
Despite these achievements, the HIV/AIDS epidemic in Jamaica continues to spread. HIV prevalence rates are high among persons most at risk, thus providing a pool of infection, which continues to feed the spread of HIV/AIDS. For instance, HIV seroprevalence among persons attending public clinics for sexually transmitted diseases (STDs) in 2002 was 5.8%; among commercial sex workers, 10% in Kingston and 20% in Montego Bay; among prisoners, 12%; and among homo-/bisexual males, 25% in one survey.8 It is therefore not surprising that the risk behaviors associated with AIDS in Jamaica include multiple sex partners, having a STI other than HIV, commercial sex, and use of crack cocaine.5,8
A variety of approaches have been used to promote safe sexual behavior among Jamaicans. These include mass media, targeted interventions, peer educators, community drama, condom campaigns, social marketing, interactive activities, school-based programs, and counseling. Although there is clear evidence of some behavior change such as increased condom use and a decline in the proportion of men reporting more than one sexual partner in the past year, this is not sufficient to control the spread of HIV. As many as 24% of men and 35% of women who report having sexual intercourse with a nonregular partner did not use a condom at last sex. In fact, behavior change among Jamaican adults appears to have reached a plateau and new methods are needed to achieve further progress.
It is in this context that a relatively new approach to targeting cost-effective HIV/AIDS prevention interventions known as the PLACE method11 was applied in St. James, the parish with the highest cumulative AIDS case rate (806.5 per 100,000 population as of December 2004).5 The PLACE (Priority Locations for AIDS Control Efforts) method is a new monitoring tool to identify areas likely to have a higher incidence of HIV infection and specific sites where AIDS prevention programs should be focused. Site-based indicators of sexual activity and AIDS prevention programs are provided by the method to monitor whether interventions are reaching key target groups.
The PLACE method was developed at the University of North Carolina and pilot-tested in 1999 in Cape Town in collaboration with the University of Cape Town. USAID has supported development of the method through MEASURE Evaluation Project.
The PLACE protocol includes 5 steps:
- To identify high transmission areas in a given geographic area, e.g., parish;
- To identify sites in high transmission areas where people meet new sexual partners;
- To visit, map, and characterize sites in each area;
- To describe the characteristics of people socializing at sites; and
- To use findings to inform interventions.11
Ten communities in the parish of St. James were selected as assessment areas for the study based on the socioeconomic characteristics of the communities and the judgment of health staff that they were likely to be high transmission areas for HIV. The proposed assessment areas were categorized according to their level of urbanization to ensure inclusion of a variety of communities (urban planned, unplanned and mixed, and rural). The following communities were selected: Central Montego Bay, Flankers/Whitehouse, Farm Heights/Rose Heights, Canterbury, Sign, Salt Spring, Norwood, Granville/Pitfour, John's Hall, and Maroon Town.
The study included 3 phases. In phase 1, trained interviewers visited these 10 communities and approached adults for a brief interview aiming to reach a broad cross-section of persons in each community. Persons were asked, “Where do people from St. James meet new sexual partners?” All the sites named as well as the age, sex, and occupation of the community informant were noted. A list of all the reported sites was prepared.
In phase 2, the interviewers visited all the reported sites in the parish. A knowledgeable person at each site (such as a manager or employee) was identified and interviewed using a structured questionnaire. Information was obtained on the characteristics of the site and the customers who visited it. Two persons used Global Positioning System (GPS) units to measure the latitude and longitude of each site for mapping and inclusion in a computerized geographic information system.
In phase 3, 2 samples of sites were selected: first, a random sample representative of all sites and the second, a purposeful sample (Table 1). The criteria for inclusion in the latter “special” sample were sites reported by 14 or more community informants as being important sites where persons go to meet new sex partners (e.g., nightclubs). Three sites were included in both samples. Interviewers visited all the sites selected in the 2 samples and interviewed a cross-section of persons socializing at the site using a structured questionnaire seeking information about their sexual behavior. Interviews were completed in 15 days spread over 3 weeks in June 2003.
The scheduling of individual interviews was determined based on the sites reported busiest time obtained during site visits in phase 2. Interviewers worked in pairs and were assigned to sites when the maximum numbers of respondents were reported to be present. We initially expected to vary the number of interviews per site based on information obtained during phase 2 on the number of persons expected at a site at peak times. Unfortunately, site representatives had difficulty providing estimates of the number of persons at the site during peak attendance. Consequently, interviewers were given a quota of 24 interviews to be conducted at each site regardless of the number of persons at the site. They were instructed to count and record the actual number of men and women socializing at the site at the beginning of the interview period at the midpoint and at the end. At some sites, including a small cook shop, 24 respondents were unavailable. In such cases, interviewers waited for several hours to try to obtain the quota. If the gender quota was unattainable, interviews were conducted with the people present.
Interviewers were trained to select respondents randomly by visually drawing an “X” across the site and by interviewing people located along the “X.” One interviewer would follow one line from corner to corner, whereas the other followed the other line. Respondents were interviewed face to face in a private setting to achieve privacy. If a respondent was socializing with other persons, interviewers were trained to ask the respondent to step outside of the group to conduct the interview. Verbal consent was obtained by reading a standardized paragraph at the beginning of each interview affirming confidentiality and that their responses were both voluntary and anonymous. Respondents were told the nature of the questions and the purpose of the study before asking their consent to participate. Fieldwork was actively monitored by supervisors while in progress and daily debriefings of interviewers were held by the coordinator to promote quality control.
The list of sites obtained in phase 1 was compiled manually. Data from phases 2 and 3 were double-entered. In the analysis, interview data from persons socializing at sites is weighted based on the median number of persons observed at the site during the study period. Persons with a low probability of selection are given a higher weight than persons at sites where the probability of selection was high. The determination of whether differences in the proportion at “random” versus “special” sites was made using the Rao-Scott χ2 statistic estimated by proc surveyfreq in SAS controlling for the site of interview and using the weighted data.
The PLACE study in St. James is a joint initiative between Ministry of Health Jamaica, Carolina Population Centre at the University of North Carolina (UNC) at Chapel Hill, and the Western Regional Health Authority (WRHA). Ethical approval was obtained from the Institutional Review Boards of the Ministry of Health, Jamaica, and the School of Public Health at the University of North Carolina. Fieldwork was conducted between April and June 2003 in St. James. Funding for the assessment was provided by USAID through the MEASURE Evaluation Project. A report of the study was prepared.12
Phase 1: Identification of Sites
Twelve interviewers worked 4 days to interview 560 community informants in the 10 communities identified as likely HIV high-transmission areas in St. James. Community informants identified 421 unique sites and events where they believed people from St. James went to meet new sexual partners. One fifth of these sites were located in Central Montego Bay. A wide cross-section of community informants participated with ages ranging from 18 to 84 years. The majority of respondents were male (56%). The refusal rate was 17% (114 persons).
Phase 2: Characteristics of Sites
In phase 2, 15 interviewers visited 361 sites to interview someone knowledgeable about the site (Table 2). Two hundred eighty-four interviews were completed over 12 days. Only 3% of interviews were not completed as a result of refusals. Twenty sites could not be located, 32 sites were duplicate sites, and 14 were closed. A wide variety of sites were verified including: bar/nightclub 17%, street 14%, market/mall/other commercial 11%, restaurant 10%, church 9%, sports venue/event 8%, go-go club/massage parlor 5%, hotel/hostel/guest house 5%, beach/park 4%, taxi stand/bus park 4%, and other 13%. The majority of sites (82%) had been in operation for more than 2 years and only 4% had opened less than 1 year before the interview. More than 90% of the sites were located within 2 blocks of a busy road or taxi route and approximately half were within 2 blocks of a bus stop or trucking route. Few were located near a port.
Almost 40% of sites reported Friday and Saturday evenings (6–10 pm) as a busy time with 17% reporting only Friday and Saturday late nights (10 pm–6 am) as busy. Twenty percent of sites were busy during the afternoon hours from noon to 6 pm daily.
Site representatives at one third of the sites reported that most of the people who visit their site are residents of the community where it is located and nearly as many reported that most men and women come to the site at least once per week. Primary school students and youth below the age of 18 can be found at the majority of sites. At approximately half of the sites, people visit from all over St. James from the neighboring parishes of Hanover, Westmoreland, and Trelawny as well as from foreign countries. At only 34% of sites was mixing limited to only those living within the parish.
Sixty-two percent of the knowledgeable people interviewed confirmed the community informant reports that people meet new sex partners onsite. At 22% of sites, someone facilitates partners meeting and at 16% of sites, persons have sex on site. Female sex workers solicit customers at 11% of sites and female staff meet new sexual partners at 10% of sites. Other activities common at sites were drinking alcohol (69%), music (59%), smoking marijuana (41%), playing games/dominoes (41%), TV (28%), dancing (17%), and gambling (17%). Go-go dancing occurred at 6% of sites and sex videos at 5%. At approximately 20% of sites, some patrons appear to use crack cocaine and men buy or sell sex, whereas at 10% of sites, some persons appear to be injecting drug users and women sell or buy sex.
Only one fourth of site representatives reported that condoms were available onsite at the time of the interview and almost half said they were never available in the past year. However, 50% of respondents were willing to sell condoms or allow their distribution onsite. Almost one third of sites had ever had any AIDS prevention activities onsite. Educational talks, peer health education, and condom promotion were the most frequently mentioned interventions. Approximately 80% of site representatives were willing to have an HIV/AIDS prevention program onsite.
Phase 3: Interviews With Persons Socializing at Sites
In the random sample of sites, 282 men and 200 women were interviewed at 23 sites. At the sites named by 14 or more community informants, 320 men and 265 women were interviewed at 26 sites. The refusal rates were similar for men and for women at approximately 13%. The characteristics of people interviewed are reported in Table 2. We first report on characteristics of persons socializing at randomly selected sites.
The mean age of men and women at the randomly selected sites was approximately 32 for men and 29 for women. Approximately half of site patrons reported living within the 10 areas included in the study as high transmission areas and over half had lived in the same area at least 10 years. More women than men had lived in the area for 1 year or less. More women than men were currently students. The majority of respondents were employed either full- or part-time, although unemployment was higher among women. Only a few respondents were visitors from other countries.
Sites were visited frequently by a core group of people. More than 50% of respondents visited sites at least once per week with one fourth of men visiting daily. Approximately half visited the site for the first time more than 5 years ago. More women than men reported being a first-time visitor.
The most common reason respondents reported for visiting the site was to socialize with friends; a substantial proportion reported coming to the sites to drink alcohol.
Although few reported visiting the site specifically to meet a sexual partner, most respondents believe that people meet new sexual partners at the site. More men than women (18% vs. 4%) report having themselves met a new partner at the site. Of the men who had ever met a new partner on site, 81% reported using a condom with the most recent new partner from the site.
The reported rates of recent new, concurrent, and commercial sexual partnerships were higher among men compared with women at these sites. One fourth of men reported a new partner in the past 4 weeks and almost 10% reporting exchanging money for sex. Almost one fifth of men and one third of women reported having no sexual partners in the last 4 weeks. The majority of men and women reporting any new sexual partner in the last 4 weeks also had a regular partner.
Over half of the men and over one third of the women reported having a new sexual partner in the past 12 months with 20% of men having 4 or more. The majority of respondents reported meeting their last new sexual partner at a public place or event. Among respondents who had at least one new sex partner in the last year, 7% of men and 3% of women reported never using a condom.
Men were more likely to have younger partners and women to have older partners. More than one fourth of men interviewed reported their youngest partner to be at least 10 years younger and, similarly, one fourth of women reported their oldest partner to be more than 10 years older. However, it was not uncommon for partnerships among older women and younger men to occur.
Younger men were more likely to report concurrent partners and having a new sexual partner in the last 4 weeks. They were also more likely to have had a partner less than 18 years of age. Over 10% of men reported having a partner in the last 12 months who resided outside Jamaica.
Among people who had a new sex partner in the last year, most men (78%) reported using a condom the last time they had sex with the most recent of these partners. Approximately half of respondents reported using a condom at last sex with a regular partner in the last year. However, few reported using a condom obtained at the site and few were able to show the interviewer a condom when requested. Only 12% of men and 3% of women displayed a condom to the interviewer on request. Youth 18 to 24 years of age of both sexes reported higher rates of condom use than older persons. Almost 20% of women in the random sites had never used a condom; however, most of these women did not have a new partner in the last 12 months.
Sexually Transmitted Infections
Fewer than 10% of men and women reported a symptom of a STI in the past 12 months, although almost one fourth reported having ever had a symptom. Most people sought treatment with public or private doctors.
Men were more likely to have smoked marijuana in the past week than women and users were more likely to report having a new sexual partner in the past 4 weeks than persons who did not smoke. Five persons reported crack cocaine use in the past 4 weeks.
Exposure to AIDS Prevention Programs
Most persons reported exposure to some form of AIDS prevention program in the last 3 months. Radio and TV programs and posters were the forms most frequently mentioned. Almost one third of respondents in the random sample reported being exposed to AIDS prevention interventions at the site of the interview. This was usually hearing a radio program onsite.
Over one third of respondents reported ever having an HIV test. Approximately half of these persons were tested more than 1 year ago.
Differences Among Persons Interviewed at the Special Sample of Sites
Overall, the sociodemographic characteristics of persons socializing at the sites named more frequently as places where people meet new sexual partners were very similar to the characteristics of those interviewed at the random sample of sites. The age distribution, educational attainment, length of time in current residence, student and employment status, and site visitation patterns did not differ depending on whether the site was one of the sites that was most frequently reported or not. There were a few specific differences, however, in the sexual behavior of persons at the 2 different types of sites. Both men and women were more likely to report coming to the more frequently reported sites to socialize. Women at these sites were also more likely to report coming to the site to meet a new sexual partner at the site, to have met a new sexual partner at the site previously, to have used a condom with the partner met at the site, and to have had a new partner in the past 4 weeks. Few women reported commercial sex, but more women at the special sites reported commercial sex that at the random sites.
This study, using the PLACE methodology, does not have a population-based sample and therefore cannot be generalized to the population as a whole. However, the sample from the random sites is likely to reflect the more sexually active population in St. James and therefore is of great relevance to HIV control.
Specific measures were taken to ensure quality control, including close supervision of field work, careful training and daily debriefing of interviewers, review of all questionnaires by supervisors, and only using interviewers who performed adequately in the field. Nevertheless, there were a number of problems that need to be noted. The method of identifying individuals for interview in phase 3 by following the imaginary lines of an “X” across the site was not feasible in some cases, e.g., a taxi stand. Many respondents had difficulty understanding questions regarding the number of regular partners and the total number of sex partners in the past 12 months. Interviewers had to probe respondents to clarify conflicting answers to these questions. (For example, some respondents reported that the number of new sex partners, total number of sex partners, and number of regular partners in the past year were all the same.) Respondents who work at the site of the interview were not clearly identified, making interpretation of some results difficult. The wording of these questions needs to be improved in future PLACE studies in Jamaica, and these indicators for St. James should be interpreted with caution.
Persons were generally willing to be interviewed and the refusal rate of 13% in phase 3 is acceptable. However, interviewers felt that the refusal rate would have been lower if a token (besides the pamphlet used) was available to give respondents at the completion of the interview. Overall, the data appear to be valid, although caution should always be used in interpreting self-reported data on sensitive matters such as sexual practice and condom use. Persons younger than 18 years were not interviewed, so the behavior of this important age group visiting sites was not assessed.
This study illustrates the great diversity of sites at which people seek to meet new sexual partners. This was a significant new insight for health staff. Virtually any gathering or location was identified, including malls, street corners, taxi stands, and churches. Persons of all classes and occupations frequented these sites. Such activity was by no means limited to nights and many of the sites included teenagers. Also of interest was the core of regular customers at many sites: 50% to 80% visited at least once weekly.
There was a significant gap between the perception of sexual activity by respondents at sites and the actual sexual activity reported. This is reflected in the difference between the approximately 80% who perceive that people meet new sex partners at the site and those who report actually doing so. There was a core of customers (20–30% of all customers at any site) who appear to be more sexually active. These are the ones most in need of targeting for prevention intervention. Nevertheless, 39% of women and 52% of men have at least one new partner in the past 12 months and are at risk for HIV infection. Most of these persons also have a regular sexual partner. Commercial sex and transactional sex are definite features at many sites. Alcohol and marijuana use are common at sites and could contribute to risk-taking behavior by disinhibiting persons using these drugs. This suggests that the National HIV/STI Control Program and the National Council for Drug Abuse need to coordinate closely.
The special sample appears to be at higher risk for HIV than the random sample, although the difference is smaller than we expected. Dancers at go-go clubs and sex workers soliciting at street sites were more likely to be included in the special sample and therefore account for some of these differences. Almost all women reporting more than 2 new partners in the last 4 weeks were socializing at go-go clubs or streets known for sex. Based on these data, AIDS prevention messages and outreach efforts cannot be focused solely on those sites most frequently reported by informants as places where people meet new sexual partners. At every site, there were persons with recent new and multiple sexual partnerships.
Overall, men appear to be at higher risk for HIV than women except for a small group of women who have many partners and who accept money for sex. This is consistent with national AIDS data showing an adult male to female ratio of 1.6:1.
Young men 18 to 24 years were more likely to have multiple partners than older men, but they were also more likely to report using condoms.
The proportion of people reporting condom use was relatively high and consistent with the 2000 National Knowledge, Attitudes and Practices Survey in which 76% of men and 67% of women reported using a condom at last sex with a nonregular partner and 52% of men and 38% of women reported doing so at last sex with a regular partner.10 Overreporting of condom use in surveys resulting from self-presentation bias is well known,13 and the pressure to give the “correct” answer may be high in St. James given the intense condom promotion and AIDS education campaigns underway. The proportion of respondents able to display a condom at the time of interview was low.
The majority of people visiting sites had been exposed to some form of AIDS prevention in the past 3 months, mostly through the media. This suggests high visibility of prevention messages. More women than men reported being counseled by a health worker regarding HIV/AIDS. This probably reflects men accessing health care less frequently, although health workers may also address issues of sexuality more often with female patients.
The finding that more than one third of all respondents ever had an HIV test is encouraging. However, more than half of these persons had not had a test in the past year and therefore do not know their current HIV status. Respondents were not asked whether they had received the result of their HIV test.
There is a significant gap between opportunities onsite for prevention and actual prevention activities. According to site representatives, 81% of sites were willing to have prevention programs and 41% of sites were willing to sell condoms. This is a challenge that health staff need to address.
This survey has been extremely useful in providing new insights into sexual networking in St. James as well as establishing a baseline for sexual behavior so that prevention interventions can be properly monitored and evaluated. Many new sites have been identified for targeted intervention. The PLACE method has opened up many opportunities for HIV prevention activities that were not being addressed. If these opportunities are properly pursued, it is likely that significant progress will be made in controlling the HIV/AIDS epidemic in St. James.
These findings are relevant for other countries where HIV prevention would benefit from a broader focus than a narrow focus on particular high-risk populations such as sex workers, men who have sex with men, and injecting drug users. A broader focus is particularly relevant in countries where a large proportion of the new infections are outside these risk groups and where targeted programs would experience issues of stigma, where commercial sex has a relatively small role in the larger cultural context of multiple partnerships, and where core group membership is difficult to define. The findings are also relevant for countries interested in using site- or venue-based strategies for HIV prevention efforts.