Abstract
Many Haitian adolescents are highly vulnerable to HIV infection. Among 3,391 sexually active 13-25-year-olds in our Voluntary Counseling and Testing (VCT) Center in Port-au-Prince from October 2005 to September 2006, we assessed associations between demographic and behavioral factors and HIV status using multivariable logistic regression analyses. We diagnosed HIV infection in 6.3% of 2,533 females and 5.5% of 858 males. Age-specific prevalence was 3.4% for 13-15-year-olds, 4.7% for 16-19, and 6.8% for 20-25 (P=0.02). Poor education, not residing with parents, currently or formerly married, having a child, and being self-referred to VCT services by others were significant predictors of HIV in females. HIV infection was associated with considering oneself at higher risk, though most youth did not recognize this risk. HIV in females was also associated with suspected/confirmed sexually transmitted infection (STI), especially genital ulcers (ORadj=2.28, 95%CI:1.26-4.13), years of sexual activity (Ptrend=0.07), and suspicion that partners had other partners or an STI. Among males, HIV was associated with drug use (though uncommon), as well as sexual debut with a casual/unknown person (ORadj=3.18, 95%CI:1.58-6.42). HIV-infected young people were more likely to be RPR positive and less likely to use condoms. Young Haitians are a key target for HIV prevention and care and avail themselves readily of youth-focused VCT services.
Keywords: HIV, sexual behavior, adolescent, youth, Haiti, counseling, HIV testing
Introduction
Haiti's per capita annual income of US$330 and population of 8.5 million (≈1/3 in greater Port-au-Prince) make it the poorest and one of the most densely populated countries in the Western hemisphere.1 Haiti has the highest number of people living with HIV (≈170 000) in the Caribbean region with adult prevalence rates ranging between 2.2% (population-based survey) and 3.8% (UNAIDS) in 2005; the lower rate is mainly due to the decreasing levels of HIV infection in major cities from 5.5% to 3% during 2000 and 2005, further evidenced by data from urban antenatal clinic attendees whose prevalence has fallen from 5.9% to 3.1% during 1996 and 2004.2,3 True HIV prevalence may be higher as hard-to-reach persons may be at higher risk.4 As is true in Africa but not in the rest of the Americas, heterosexual transmission is responsible for most Haitian HIV/AIDS cases.5,6
Since the early 1980's, the Haitian Study Group on Kaposi's Sarcoma and Opportunistic Infections (GHESKIO) has provided care for HIV/AIDS, other sexually transmitted infections (STI), diarrhea, and tuberculosis through its community and clinic-based services.7 GHESKIO's voluntary counseling and testing (VCT) program began in 1985 and has experienced a steady and dramatic increase in the number of young people seeking services; adolescents ages 13-19 years constituted <1% of VCT clients in 1985, 9% in 1999, and 15% in 2005. About 23,000 young people ages 13-25 years sought VCT at GHESKIO from 2001-2005.7 We explored sociodemographic and sexual behavioral characteristics associated with HIV infection among young people seeking VCT in urban Haiti with the goal of better designing age-appropriate services.
Methods
Study population
GHESKIO serves a densely populated and impoverished area in central Port-au-Prince. The VCT service is integrated into other services (TB, STI management, reproductive health, HIV/AIDS treatment and care) and the clinic is located along a pothole-filled road in a large, gang-infested slum. It serves the area's low-income residents free of charge. We used the World Health Organization's definition of young people (adolescence ages 10-19 years, and youth ages 20-25 years), but we excluded participants ages 10-12 years given their very small numbers in our clinic, as well as excluding youth with very probable perinatally acquired disease (e.g., ages <15 years with relatively advanced illness or with a history of mother's HIV-related illness).8 All VCT attendees aged 13-25 years who attended GHESKIO for VCT from October 2005 to September 2006 were included if they had an HIV test and self-reported sexual intercourse.
Data collection and variables
During an initial VCT visit, we completed a sociodemographic and behavioral questionnaire with clients prior to HIV testing, as per GHESKIO standard procedures. Demographic variables included sex, date of birth, education, occupation, marital or cohabiting status, living with parents, parental relationship, residential area, and number of children. The interviewers were trained in gaining clients' confidence, maintaining an unbiased attitude, and assuring clients' confidentiality. After obtaining informed verbal consents, the information on clients' knowledge, attitudes, behaviors, and clinical symptoms related to HIV were collected using an interviewer-administered structured questionnaire. Questions to test knowledge of HIV covered mode of transmission, prevention methods, sources of knowledge, and willingness to share knowledge with others. We asked sexually experienced females and males about the following HIV-related risk behaviors: age at sexual debut, with whom youth had their sexual initiation, years of sexual activity, experience with rape, genital-anal or oral-genital sex, past and current unprotected sex, age of first and current sexual partners, homosexual experiences, number of sexual partners in the past 6 months, suspicion that one's regular partner has other sex partner(s) or an STI, illicit drug use, condom use, history and/or current symptoms of STI, current pregnancy status, use of family planning methods, and perceived risk of HIV infection. Young people were asked to quantify their self-perceived risk for HIV as none, low, medium, or high. They were also asked to share their reasons for seeking VCT including being referred by a partner or pastor, a partner having become sick or died, marriage, weight change, desire to know HIV status, broken condom, having unprotected sex, a partner with another sex partner, experiencing signs/symptoms of STI or TB and blood loss/transfusion, someone in the family with HIV or having died, and any current clinical signs and symptoms such as fever, diarrhea, genital ulcer, or discharge.
Details of our VCT services are described elsewhere.9 HIV serostatus was tested by ELISA, Murex® HIV-1.2.0 (Abbott Laboratories, Abbott Park, IL, USA) or one of two rapid latex-agglutination tests, Capillus® HIV-1/2, (Trinity Biotech, Wicklow, Ireland) or Determine® HIV-1/2 (Abbott). Confirmation was conducted with a second, separate rapid test, using a different company's product whenever available. Indeterminant ELISA or rapid test results were categorized with the use of a Western blot, Cambridge Biotech HIV-1 WR® (Calypte Biomedical, Rockville, MD, USA).10 Syphilis was detected by rapid plasma reagin (RPR) test using standard techniques.11 Positive RPR was considered an indication for syphilis treatment with penicillin, though confirmatory tests were performed using the fluorescent treponemal antibody-absorption (FTA-ABS) test whenever deemed helpful for clinical management (confirmatory data not reported).
Questionnaire data were linked to HIV and syphilis test results and medical records in a confidential manner, removing all personal identifiers including names, birthdates, addresses, and other potentially compromising data. The Institutional Review Boards at GHESKIO, Weill Medical College of Cornell University, and Vanderbilt University approved our study.
Statistical analyses
We used Microsoft Access® 2003 for data entry and SAS® version 9.1 for statistical analyses (descriptive, bivariable, and multivariable). Means and standard deviations were calculated for quantitative variables and analyzed using the student's t-test. Proportions for categorical variables were assessed for statistical significance (P value) using the Chi-square test with Yates' correction. Bivariable cross-tabulations measured the association of outcome (HIV infection status) with each independent variable (risk factor). Odds ratios (OR) and 95% confidence intervals (CI) were calculated for each association in females and males separately, while controlling for the effects of other variables in logistic regression analyses. We considered four potential confounding or interacting variables in our multiple regression analyses – age, education, income, and occupation –by assessing co-linearity of these variables. All were ultimately included in our regression analysis models.
Tests for trend (Ptrend) were performed by entering selected categorical variables as continuous parameters in the model. For categorical variables in the field of HIV knowledge and risky sexual behaviors, missing values (10-15% of the total potential data points) were coded as “do not know” or “did not answer.”
Results
Prevalence of HIV infection
In the 13-25 year age group, 4,022 people came for VCT in the recruitment interval and 3,869 (96.2%) accepted VCT and had a test performed. Among those receiving an HIV test, 3,391 (87.6%) self-reported as sexually active and represent our study population. Their mean age was 20.8 years (20.7 for females and 21.1 years for males). Most young people (84.2%) self-referred to the health center and 65.8% of youth came with the specific intent to obtain HIV testing (data not shown). We diagnosed HIV infection in 6.3% of 2,533 females and 5.5% of 858 males. Females accounted for 74.7% of young people tested and 77.2% of all HIV-infected young people (Table 1). Age-specific prevalence was 3.4% for 13-15-year-olds, 4.7% for 16- to 19-year-olds, and 6.8% for 20- to 25-year-olds (P=0.02, Figure), and HIV risk was positively associated with age (Ptrend=0.02, Table 3).
Table 1. Characteristics of sexually active young people seeking HIV Counseling and Testing, Port-au-Prince, Haiti.
| Characteristics | All n=3 391 (%) |
Females n=2 533 (74.7%) |
Males n=858 (25.3%) |
P value† |
|---|---|---|---|---|
| Age at time of interview, y (mean ± SD) | 20.8 ± 2.9 | 20.7 ± 3.0 | 21.1 ± 2.9 | <0.01 |
| Age, years (%) | ||||
| 13-15 | 148 (4.4) | 110(4.3) | 38 (4.4) | 0.003 |
| 16-19 | 986 (29.1) | 776 (30.6) | 210 (24.5) | |
| 20-25 | 2 257 (66.6) | 1 647 (65.0) | 610 (71.1) | |
| Education (%) | ||||
| No education | 210 (6.2) | 183 (7.2) | 29 (3.2) | <0.01 |
| Some or all primary | 1 032 (30.5) | 832 (32.9) | 200 (23.3) | |
| Some or all secondary | 1 906 (56.2) | 1 380 (54.6) | 526 (61.3) | |
| Some or all college/professional | 239 (7.1) | 134 (5.3) | 105 (12.2) | |
| Income in Haitian Gourde per year, 5:1 US$ (%) | ||||
| No | 2 257 (67.2) | 1 698 (67.7) | 559 (65.6) | 0.01 |
| 150-1000 | 693 (20.6) | 528 (21.1) | 165 (19.4) | |
| 1001-5000 | 331 (9.9) | 232 (9.3) | 99 (11.6) | |
| >5000 | 77 (2.3) | 48 (1.9) | 29 (3.4) | |
| Occupation (%) | ||||
| Student | 864 (25.5) | 592 (23.4) | 272 (31.7) | <0.01 |
| Trader/business person | 748 (22.1) | 618 (24.4) | 130 (15.1) | |
| Manual worker § | 544 (16.0) | 286 (11.3) | 258 (30.1) | |
| None | 1 235 (36.4) | 1 037 (40.9) | 198 (23.1) | |
| Living with parents (%) | ||||
| Both parents | 314 (9.3) | 211 (8.3) | 103 (12.0) | <0.01 |
| Father only | 87 (2.6) | 62 (2.4) | 25 (2.9) | |
| Mother only | 556 (16.4) | 390 (15.4) | 166 (19.6) | |
| Relatives/Friends | 1 483 (43.7) | 1 177 (46.5) | 306 (35.7) | |
| Husband or wife/Partner | 951 (28.0) | 693 (27.4) | 258 (30.1) | |
| Quality of relationship between respondent and parents (%) | ||||
| Good | 1 941 (57.2) | 1 473 (58.1) | 468 (54.5) | 0.09 |
| Satisfactory | 335 (9.9) | 253 (1.0) | 82 (9.6) | |
| Did not answer or not applicable | 1 115 (32.9) | 807 (31.9) | 308 (35.9) | |
| Residential area where the respondents live (%) | ||||
| Suburb | 1 493 (44.0) | 1 982 (42.7) | 411 (47.9) | 0.008 |
| Inner city | 1 898 (56.0) | 1 451 (57.3) | 447 (52.1) | |
| Age at onset of sexual activity, y (mean ± SD) | 16.0 ± 2.9 | 16 5 ± 2.5 | 14.6 ± 3.5 | <0.01 |
| Marital or cohabiting status (%) | ||||
| Never married | 2 241 (66.1) | 1 477 (58.3) | 764 (89.0) | 0.008 |
| Married/cohabited/divorced/widowed | 1 150 (33.9) | 1 056 (41.7) | 94 (11.0) | |
| Ever had a child (%) | 1,065 (30.6) | 962 (37.0) | 103 (11.7) | <0.01 |
| Self-referral to VCT for HIV testing (%) | 2 855 (84.2) | 2 131 (84.1) | 724 (84.4) | 0.86 |
| HIV seropositivity (%) | 206 (6.1) | 159 (6.3) | 47 (5.5) | 0.40 |
| History of syphilis seropositivity (%) | 151 (4.5) | 129 (5.1) | 22 (2.6) | 0.002 |
Derived from t-test for continues variables and chi-square test for categorical variables
Skilled or non-skilled worker, housekeeper, odd jobs holder, and commercial female sex worker
Note: Missing values were <1.2% in all variables that excluded from percentage calculation
Figure.
HIV prevalence rate by sex and age among sexually active Haitian adolescents and youth seeking voluntary counseling and testing
Table 3. Association between demographic characteristics and HIV infection status among sexually aactive young people at VCT center, Port-Au-Prince, Haiti.
| All (n=3 391) | Females (n=2 533) | Males (n=858) | |
|---|---|---|---|
| OR (95% CI)** | OR (95% CI)* | OR (95% CI)* | |
| Age, years | |||
| 13-15 | 1.00 | 1.00 | 1.00 |
| 16-19 | 1.34 (0.52-3.45) | 1.81 (0.54-6.00) | 0.46 (0.09-2.54) |
| 20-25 | 2.03 (0.80-5.13) | 2.38 (0.73-7.80) | 1.39 (0.29-6.71) |
| Ptrend =0.02 | Ptrend =0.06 | Ptrend =0.09 | |
| Education | |||
| No education | 1.00 | 1.00 | 1.00 |
| Some or all primary | 0.81 (0.47-1.38) | 0.82 (0.46-1.45) | 0.81 (0.16-4.07) |
| Some or all secondary | 0.68 (0.40-1.15) | 0.65 (0.37-1.16) | 0.85 (0.18-3.93) |
| Some or all college/professional | 0.25 (0.09-0.67) | 0.30 (0.10-0.95) | 0.21 (0.03-1.68) |
| Annual Income | |||
| No | 1.00 | 1.00 | 1.00 |
| 150-1000 | 0.69 (0.41-1.18) | 0.68 (0.37-1.25) | 0.65 (0.22-1.91) |
| 1001-5000 | 1.59 (0.97-2.61) | 1.46 (0.82-2.60) | 1.60 (0.57-4.50) |
| >5000 | 1.23 (0.58-3.50) | 1.75 (0.65-4.73) | 0.61 (0.07-5.36) |
| Occupation | |||
| Student | 1.00 | 1.00 | 1.00 |
| Trader/business person | 1.47 (0.81-2.69) | 1.82 (0.89-3.74) | 1.05 (0.32-3.41) |
| Manual worker | 1.44 (0.80-2.60) | 1.73 (0.82-3.64) | 1.13 (0.41-3.10) |
| None | 1.56 (1.00-2.41) | 1.96 (1.14-3.45) | 0.75 (0.30-1.84) |
| Living with parents | |||
| Both parents | 1.00 | 1.00 | 1.00 |
| Father only | 0.30 (0.04-2.18) | 0.53 (0.10-3.75) | - |
| Mother only | 1.03 (0.50-2.12) | 1.25 (0.50-3.11) | 0.69 (0.20-2.34) |
| Relatives/Friends | 1.43 (0.77-2.68) | 1.77 (0.80-3.95) | 0.83 (0.28-2.46) |
| Husband or wife/Partner | 1.93 (1.01-3.69) | 2.30 (1.00-5.28) | 1.41 (0.48-4.10) |
| Quality of relationship between respondent and parents | |||
| Good | 1.00 | 1.00 | 1.00 |
| Satisfactory | 1.15 (0.69-1.91) | 1.29 (0.73-2.26) | 0.80 (0.23-2.78) |
| Did not answer or not applicable | 1.49 (1.08-2.05) | 1.54 (1.06-2.23) | 1.42 (0.74-2.71) |
| Residential area where the respondent live | |||
| Suburb | 1.00 | 1.00 | 1.00 |
| Inner city | 1.13 (0.84-1.51) | 1.03 (0.74-1.43) | 1.49 (0.80-2.78) |
| Marital or cohabiting status | |||
| Never married | 1.00 | 1.00 | 1.00 |
| Married/live with partner | 1.44 (1.03-2.01) | 1.42 (0.99-2.04) | 1.33 (0.55-3.23) |
| Ever had a child (%) | |||
| No | 1.00 | 1.00 | 1.00 |
| Yes | 1.61 (1.17-2.20) | 1.67 (1.18-2.41) | 1.27 (0.56-2.88) |
| Referred to VCT clinic for HIV testing | |||
| by respondent | 1.00 | 1.00 | 1.00 |
| by others | 1.68 (1.17-2.41) | 1.90 (1.28-2.83) | 1.16 (0.46-2.91) |
Adjusted for age (categorical), education, income (categorical) and occupation, except for the same variable
Additionally adjusted for gender
Demographics, medical history, and HIV status
In general, females had lower education and income than males, were less likely to be students, and were less likely to live with their parents (P≤0.01 for all). They were more likely than males to live in suburbs, to be married or cohabiting, and to have children (P<0.01 for all). There were no gender differences in declared quality of relationship with their parents (Table 1).
HIV-infected females were less educated, more likely to be unemployed or out of school, poorer, to be married or cohabited, to have a child, to live apart from or have worse declared relationships with their parents than were HIV-uninfected females (P<0.05 for all, except for education P=0.09; Table 2). Our multivariable regression model for all, predominantly for females indicated that lower prevalence of HIV infection was associated with better education (adjusted odds ratio [ORadj]=0.30, 95% CI 0.10-0.95 for college or professional education), while higher prevalence of HIV was associated with not residing with or having a poor relationship with parents, being or having been married or cohabiting, having a child (ORadj=1.67, 95% CI 1.18-2.41), and being referred to VCT services by others (ORadj=1.90, 95% CI 1.28-2.83). Similar associations were observed among males but may have been due to chance (all P>0.05; Table 3).
Table 2. Characteristics of sexually active young people seeking HIV Counseling and Testing, by HIV infection status, Port-Au-Prince, Haiti.
| All (n=3 391) | Females (n=2 533) | Males (n=858) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| HIV- | HIV+ | P value† | HIV- | HIV+ | P value | HIV- | HIV+ | P value | |
| n=3 185 93.9% |
n=206 6.1% |
n=2 374 93.7% |
n=159 6.3% |
n=811 94.5% |
n=47 5.5% |
||||
| Age at time of interview, years (mean ±SD) | 20.8±2.9 | 21.5±2.8 | <0.01 | 20.7±3.0 | 21.3±2.8 | <0.01 | 21.1±2.9 | 22.1±2.8 | 0.02 |
| Age, years (%) | |||||||||
| 13-15 | 4.5 | 2.4 | 0.02 | 4.5 | 1.9 | 0.08 | 4.4 | 4.3 | 0.07 |
| 16-19 | 19.6 | 22.3 | 31.0 | 25.8 | 25.3 | 10.6 | |||
| 20-25 | 65.9 | 75.2 | 64.5 | 72.3 | 70.3 | 85.1 | |||
| Education (%) | |||||||||
| No education | 6.0 | 8.8 | 0.03 | 7.0 | 10.1 | 0.09 | 3.1 | 4.3 | 0.32 |
| Some or all primary | 30.3 | 34.2 | 32.6 | 38.0 | 23.4 | 21.3 | |||
| Some or all secondary | 56.4 | 54.1 | 54.9 | 49.4 | 60.8 | 70.2 | |||
| Some or all college/professional | 7.3 | 2.9 | 5.5 | 2.5 | 12.7 | 4.3 | |||
| Annual Income (%) | |||||||||
| No | 67.6 | 62.1 | <0.01 | 68.0 | 63.5 | 0.03 | 66.1 | 57.5 | 0.08 |
| 150-1000 | 20.8 | 18.0 | 21.3 | 18.3 | 19.5 | 17.0 | |||
| 1001-5000 | 9.4 | 17.0 | 8.9 | 15.1 | 10.9 | 23.4 | |||
| >5000 | 2.2 | 2.9 | 1.8 | 3.1 | 3.5 | 2.1 | |||
| Occupation (%) | |||||||||
| Student | 26.1 | 15.5 | <0.01 | 24.1 | 11.9 | <0.01 | 31.9 | 27.7 | 0.72 |
| Trader/business person | 21.9 | 24.3 | 24.3 | 26.4 | 15.1 | 17.0 | |||
| Manual worker | 16.0 | 17.5 | 11.2 | 12.0 | 29.7 | 36.2 | |||
| None | 36.0 | 42.7 | 40.4 | 49.7 | 23.3 | 19.1 | |||
| Living with parents (%) | |||||||||
| Both parents | 9.5 | 5.8 | <0.01 | 8.6 | 4.4 | 0.01 | 12.1 | 10.6 | 0.10 |
| Father only | 2.7 | 0.5 | 2.6 | 0.6 | 3.1 | 0 | |||
| Mother only | 16.7 | 10.7 | 15.7 | 10.1 | 9.7 | 12.8 | |||
| Relatives/Friends | 43.7 | 44.7 | 46.3 | 49.1 | 36.0 | 29.8 | |||
| Husband or wife/Partner | 27.4 | 38.3 | 26.8 | 35.8 | 29.1 | 46.8 | |||
| Quality of relationship between respondent and parents (%) | |||||||||
| Good | 57.9 | 47.1 | <0.01 | 58.8 | 47.8 | 0.01 | 55.1 | 44.7 | 0.15 |
| Satisfactory | 10.0 | 9.2 | 10.0 | 10.1 | 9.8 | 6.4 | |||
| Did not answer or not applicable | 32.2 | 43.7 | 31.2 | 42.1 | 35.1 | 48.9 | |||
| Residential area where the respondent live (%) | |||||||||
| Suburb of city | 44.3 | 40.3 | 0.26 | 42.8 | 40.9 | 0.63 | 48.5 | 38.3 | 0.17 |
| Inner city | 55.7 | 59.7 | 57.2 | 59.1 | 51.5 | 61.7 | |||
| Marital or cohabiting status (%) | |||||||||
| Never married | 66.9 | 53.4 | <0.01 | 59.2 | 44.6 | <0.01 | 89.4 | 83.0 | 0.17 |
| Married/cohabited/divorced/widowed | 33.1 | 46.6 | 40.8 | 55.4 | 10.6 | 17.0 | |||
| Ever had a child (%) | 29.8 | 45.6 | <0.01 | 35.9 | 53.5 | <0.01 | 11.8 | 19.1 | 0.14 |
| Self-referral to VCT for HIV testing (%) | 84.6 | 78.6 | 0.02 | 84.6 | 76.7 | <0.01 | 84.3 | 85.1 | 0.89 |
| History of syphilis seropositivity (%) | 4.0 | 11.2 | <0.01 | 4.5 | 13.2 | <0.01 | 2.5 | 4.3 | 0.45 |
Derived from t-test for continues variables and chi-square test for categorical variables
HIV-related beliefs and perception and HIV status
Almost all young people (98.9%) had heard about HIV/AIDS (98.8% for females and 99.4% for males). The main sources of HIV knowledge were television and radio, health organizations, and peers (98.3%, 53.2% and 38.1%, respectively). Parents were the least likely sources of HIV/AIDS education (6.4%). Sources of HIV knowledge did not differ among young people by sex or by HIV status, except that uninfected females were more likely to get information from their religious teachers than were the HIV-infected females (14.7% vs. 7.8%, P<0.01). Most young people (85%) were willing to talk about HIV/AIDS with others (e.g., sister or brother, friends, sexual partners and religious teachers). Among females, HIV risk was less frequent among those who discussed HIV openly than those who did not (ORadj=0.63, 95% CI: 0.42-0.94; Table 4). General knowledge of the main modes of HIV transmission was high, and was a bit higher in males than females (90.4% vs. 86.5%, P<0.01). However, young people had poor knowledge of mother-to-child transmission of HIV, including only 51% of female youth. Other lacunae in the HIV knowledge base of the young people were more common among those infected, especially among females. HIV infected females were more likely to believe that HIV can be cured (P<0.01), and less likely to believe that HIV infection can be prevented by having only one sexual partner (P<0.01) than were uninfected females (Table 4). There was a strong association in both sexes between considering oneself at HIV risk and being infected, though majority of uninfected youth considered themselves at no or low risk compared to infected youth. In contrast, HIV-infected youth were more likely to perceive themselves as being at high-risk (P≤0.05 for both sexes). The increased level of perception of personal risk was positively associated with increased prevalence of HIV (ORadj=1.88, 95% CI: 1.52-5.44 for females and ORadj=3.06, 95% CI: 1.12-8.32 for males, Table 4). Higher prevalence of HIV infection was in males who intended to obtain HIV testing (ORadj=5.02, 95% CI 1.19-21.2) or who sought medical consultation (ORadj=10.3, 95% CI 2.04-52.0) than among those who visited the clinic by accompanying others (Table 4).
Table 4. Prevalence and association between HIV-related beliefs and risk perception and HIV status among sexually active yyoung people at VCT center, Port-Au-Prince, Haiti.
| Females (n=2 533) | Males (n=858) | |||||
|---|---|---|---|---|---|---|
| HIV- | HIV+ | OR (95% CI)* | HIV- | HIV+ | OR (95% CI)* | |
| n=2 374 (%) |
n=159 (%) |
n=811 (%) |
n=47 (%) |
|||
| Willingness to share HIV information with someone | ||||||
| No | 14.5 | 22.0 | 1.00 | 10.0 | 12.8 | 1.00 |
| Yes | 85.5 | 78.0 | 0.63 (0.42-0.94) | 90.0 | 87.2 | 0.75 (0.30-1.89) |
| P † =0.01 | P =0.54 | |||||
| Believe that HIV can be acquired by mosquito bite | ||||||
| No | 57.8 | 56.0 | 1.00 | 69.1 | 78.7 | 1.00 |
| Yes | 28.3 | 30.8 | 1.03 (0.71-1.48) | 21.8 | 12.8 | 0.46 (0.19-1.13) |
| Did not know | 13.9 | 13.2 | 0.96 (0.58-1.59) | 9.1 | 8.5 | 0.78 (0.26-2.29) |
| P =0.78 | P =0.31 | |||||
| Believe that HIV can be infected by supernatural events | ||||||
| No | 70.2 | 61.6 | 1.00 | 76.7 | 78.7 | 1.00 |
| Yes | 16.5 | 22.0 | 1.41 (0.94-2.13) | 12.3 | 12.8 | 0.94 (0.38-2.33) |
| Did not know | 13.3 | 16.4 | 1.46 (0.93-2.31) | 11.0 | 8.5 | 0.75 (0.25-2.34) |
| P =0.07 | P=0.87 | |||||
| Believe that HIV can be prevented by having only one sexual partner | ||||||
| No | 0.6 | 2.5 | 4.75 (1.49-15.2) | 1.5 | 0.0 | - |
| Yes | 71.2 | 66.0 | 1.00 | 74.6 | 78.7 | 1.00 |
| Did not know | 28.2 | 31.5 | 1.17 (0.81-1.69) | 23.9 | 21.3 | 0.92 (0.43-1.97) |
| P =0.008 | P =0.63 | |||||
| Believe that HIV infection can be cured | ||||||
| No | 94.5 | 88.7 | 1.00 | 93.9 | 95.7 | |
| Yes | 0.8 | 6.3 | 8.48 (3.77-19.1) | 1.5 | 2.1 | |
| Did not know | 4.7 | 5.0 | 1.36 (0.63-2.91) | 4.6 | 2.2 | 0.46 (0.20-3.52) |
| P <0.01 | P =0.69 | |||||
| Perceived risk level for HIV infection | ||||||
| Not at all | 15.5 | 9.4 | 1.00 | 17.1 | 12.8 | 1.00 |
| Low | 38.6 | 27.0 | 1.27 (0.69-2.34) | 42.9 | 31.9 | 1.03 (0.39-2.76) |
| Moderate | 33.5 | 42.1 | 2.07 (1.16-3.71) | 23.6 | 23.4 | 1.37 (0.49-3.86) |
| High | 12.9 | 21.4 | 1.88 (1.52-5.44) | 16.4 | 31.9 | 3.06 (1.12-8.32) |
| P<0.01 | P =0.05 | |||||
| Reason for visiting VCT | ||||||
| To accompany others | 26.3 | 25.2 | 1.00 | 19.4 | 4.3 | 1.00 |
| To know my HIV infection status | 63.2 | 58.5 | 0.93 (0.63-1.38) | 74.2 | 80.8 | 5.02 (1.19-21.2) |
| To get medical consultation | 10.5 | 16.3 | 1.48 (0.87-2.50) | 6.4 | 14.9 | 10.3 (2.04-52.0) |
| P =0.07 | P =0.006 | |||||
Derived from chi-square test
Adjusted for age (categorical), education, income (categorical) and occupation
Note: Missing values excluded from percentage calculations
HIV-related risk behaviors and HIV status
Several risk behaviors and history and/or symptoms of STIs were associated with HIV seropositivity (Table 5). Median age of sexual debut was 16 years for girls and 15 for boys. Among girls, the prevalence of sexual debut at age of ≤16 years was high (≈ 42%) and did not differ by HIV status; sexual debut was not associated with a risk of HIV infection. Unexpectedly, and in contrast to young women, more of the HIV-uninfected young men were sexually active age of ≤16 years compared to HIV-infected males (70.9% vs. 51.1%, P <0.01), such that sexual debut after the age of 16 years was positively associated with HIV infection (ORadj=2.55, 95% CI 1.32-4.92).
Table 5.
Prevalence and association between risky behaviors and HIV status among sexually active young people at VCT center, Port-au-Prince, Haiti
| Females (n=2 533) | Males (n=858) | |||||
|---|---|---|---|---|---|---|
| HIV- | HIV+ | OR (95% CI)* | HIV- | HIV+ | OR (95% CI)* | |
| n=2 374 (%) |
n=159 (%) |
n=811 (%) |
n=47 (%) |
|||
| Age at onset of sexual activity, y (median) | ||||||
| ≤16 | 41.9 | 43.2 | 1.00 | 70.9 | 51.1 | 1.00 |
| >16 | 58.1 | 56.8 | 0.90 (0.60-1.35) | 29.1 | 48.9 | 2.55 (1.32-4.92) |
| P† =0.78 | P <0.01 | |||||
| Years of sexual activity | ||||||
| ≤2 | 32.5 | 22.6 | 1.00 | 18.5 | 21.3 | 1.00 |
| ≤5 | 16.1 | 35.9 | 1.25 (0.79-1.98) | 26.3 | 27.7 | 0.80 (0.33-1.91) |
| >5 | 32.4 | 41.5 | 1.52 (0.92-2.52) | 55.2 | 51.0 | 0.54 (0.24-1.22) |
| P <0.01 | Ptrend =0.08 | P <0.01 | P <0.07 | |||
| Age of first sexual partner, y | ||||||
| ≤15 | 9.2 | 6.4 | 1.00 | 56.0 | 43.5 | 1.00 |
| 16-20 | 46.8 | 54.1 | 1.40 (0.74-2.65) | 37.9 | 47.8 | 1.66 (0.88-3.14) |
| >20 | 44.0 | 39.5 | 1.01 (0.52-1.94) | 6.1 | 8.7 | 1.80 (0.57-5.64) |
| P =0.16 | Ptrend =0.32 | P =0.24 | Ptrend =0.17 | |||
| Sexual debut with: | ||||||
| Friends and close relatives | 7.7 | 10.1 | 1.00 | 48.5 | 23.4 | 1.00 |
| Casual or unknown person | 92.3 | 89.9 | 0.74 (0.43-1.28) | 51.5 | 76.6 | 3.18 (1.58-6.42) |
| P =0.28 | P =0.008 | |||||
| Number of sex partners in past 6 months | ||||||
| 0-1 | 76.5 | 74.2 | 1.00 | 49.7 | 53.2 | 1.00 |
| 2 | 6.2 | 8.8 | 1.42 (0.77-2.61) | 26.4 | 21.3 | 0.74 (0.34-1.59) |
| >2 | 17.3 | 17.0 | 1.14 (0.73-1.79) | 23.9 | 25.5 | 0.97 (0.47-2.01) |
| P =0.44 | Ptrend =0.61 | P =0.74 | Ptrend =0.93 | |||
| Current partner has other partner | ||||||
| No | 31.1 | 20.1 | 1.00 | 25.7 | 21.3 | 1.00 |
| Yes | 15.4 | 20.2 | 1.92 (1.15-3.20) | 9.7 | 10.6 | 1.36 (0.44-4.18) |
| Did not know | 53.5 | 59.7 | 1.80 (1.18-2.72) | 64.6 | 69.1 | 1.23 (0.58-2.61) |
| P =0.01 | P =0.79 | |||||
| Engaging in casual sex | ||||||
| No | 77.3 | 73.6 | 1.00 | 50.5 | 51.1 | 1.00 |
| Yes | 3.8 | 6.3 | 1.64 (0.80-3.35) | 20.1 | 17.0 | 0.92 (0.40-2.13) |
| Did not answer | 18.9 | 20.1 | 1.31 (0.86-1.99) | 29.4 | 31.9 | 1.19 (0.60-2.35) |
| P =0.27 | P =0.85 | |||||
| Experience with genital-anal sex | ||||||
| No | 80.3 | 78.6 | 1.00 | 68.3 | 63.8 | 1.00 |
| Yes | 1.2 | 1.9 | 1.48 (0.44-4.96) | 1.5 | 4.3 | 3.16 (0.64-15.7) |
| Did not answer | 18.5 | 19.5 | 1.28 (0.84-1.95) | 30.2 | 31.9 | 1.21 (0.63-2.32) |
| P =0.72 | P =0.30 | |||||
| Condom use | ||||||
| Never | 64.8 | 66.1 | 2.05 (0.94-4.45) | 52.8 | 46.8 | 1.66 (0.64-4.31) |
| Sometimes | 26.2 | 28.9 | 1.64 (0.77-3.46) | 30.5 | 40.4 | 1.12 (0.44-2.88) |
| Regular | 9.0 | 5.0 | 1.00 | 16.7 | 12.8 | 1.00 |
| P =0.21 | P =0.34 | |||||
| Ever used family planning methods | ||||||
| No | 84.1 | 83.6 | 1.00 | 76.8 | 66.0 | 1.00 |
| Yes | 15.9 | 16.4 | 0.99 (0.63-1.55) | 23.2 | 34.0 | 1.79 (0.93-3.45) |
| P =0.89 | P =0.09 | |||||
| Ever used illegal drugs | ||||||
| No | 99.7 | 98.7 | 1.00 | 97.4 | 89.4 | 1.00 |
| Yes | 0.3 | 1.3 | 3.91 (0.79-19.4) | 2.6 | 10.6 | 5.21 (1.76-15.4) |
| P =0.07 | P =0.002 | |||||
| Ever exchanged sex for money or gifts | ||||||
| No | 91.8 | 92.4 | 1.00 | 91.0 | 93.6 | 1.00 |
| Yes | 4.0 | 4.5 | 0.86 (0.36-2.04) | 3.2 | 4.3 | 1.44 (0.31-6.67) |
| Did not answer | 4.2 | 3.1 | 1.00 (0.39-2.58) | 5.8 | 2.1 | 0.40 (0.10-2.99) |
| P =0.80 | P =0.54 | |||||
| Current sexual partner has history of STIs | ||||||
| No | 55.6 | 49.1 | 1.00 | 37.4 | 36.2 | 1.00 |
| Yes | 3.8 | 5.7 | 1.82 (0.88-3.78) | 6.0 | 2.1 | 0.61 (0.27-1.41) |
| Did not know | 40.5 | 45.3 | 1.48 (1.05-2.08) | 56.6 | 61.7 | 1.17 (0.26-5.27) |
| P =0.19 | P =0.50 | |||||
| Self-reported genital ulcer at VCT | ||||||
| No | 95.7 | 91.1 | 1.00 | 92.5 | 93.6 | 1.00 |
| Yes | 4.3 | 8.9 | 2.28 (1.26-4.13) | 7.5 | 6.4 | 0.97 (0.29-3.27) |
| P =0.007 | P =0.54 | |||||
| History of latent syphilis by serology | ||||||
| Negative | 95.4 | 86.8 | 1.00 | 97.5 | 95.7 | 1.00 |
| Positive | 4.6 | 13.2 | 2.77 (1.66-4.63) | 2.5 | 4.3 | 1.65 (0.35-7.65) |
| P <0.01 | P =0.45 | |||||
Derived from chi-square test
Adjusted for age (categorical), education, income (categorical) and occupation
Note: Missing values excluded from percentage calculations
Factors prominent in female risk included longer years of sexual activity (Ptrend=0.07), suspicion that male partners had other partners or had had an STI, and evidence of an STI (ORadj=2.28, 95% CI 1.26-4.13 for genital ulcer). Among males, drug use (though rare) and sexual debut with a casual or unknown person (people other than friends, relatives or neighbors) (ORadj=3.18, 95% CI 1.58-6.42) were risk factors for HIV infection (Table 5). In this study, 1.4 % (n=13) of young men reported having had sex with other men, and 15.5% (n=142) of young men reported having had sex with CSWs (data not shown). Young women were twice as likely as men to have syphilis (5.2% vs. 2.5%, P =0.001). HIV-infected young people were more likely to have a positive RPR test than were the uninfected (13.2% vs. 4.6%), with this association statistically significant in females (ORadj=2.77, 95% CI 1.66-4.63, Table 5).
Condoms were poorly utilized (Table 5), with 65% of females and 52% of males reporting never having used a condom. An increase in prevalence of HIV infection was observed with “never use” of condom among females (ORadj=2.05, 95% CI 0.94-4.45) and males (ORadj=1.66, 95% CI 0.64-4.31). The association was significant in the model also adjusted for believing the male partner had other partners (data not shown). Similarly, use of family planning methods was very low among both females and males, regardless of their HIV status; borderline significant increase in HIV prevalence was observed among males who did not use family planning methods (ORadj =1.79, 95% CI 0.93-3.45, Table 5). None of the following known HIV risk behaviors was associated with HIV infection in our study: past history of STI, oral-genital sex, male-to-male sex, exchanging sex for money or goods, engaging in casual sex, age of primary sex partner, having ever shared syringe/needles, current pregnancy, number of sexual partners in past six months, and non-use of family planning methods (Table 5).
Discussion
Our study demonstrates the acute need for VCT services in Haiti that focus on adolescents and young adults. Demand is high, risk behaviors common, and HIV prevalence is higher in these tested youth (6.1%) than for the Haitian population as a whole (3.8%).12 Girls and young women are particularly vulnerable. High HIV prevalence is likely to reflect higher urban rates, particularly among young people seeking VCT.13 Despite familiarity with HIV/AIDS, condom use rates are very low.
The sociodemographic factors that played a major role in HIV infection risk among Haitian female adolescents and youth14,15 are similar to those seen in other Caribbean countries, e.g., widowed/divorced/separated marital status; low educational level; occupations such as merchant, housekeeper, unemployed; having any income; poor relationships with their parents; low economic status; and having a child at a young age.16,17 Parental engagement, late sexual debut and school attendance have been reported to play a role in reducing sexual risk in young females.18,19 Poor knowledge of HIV, having an unfaithful sexual partner, perceived risk of HIV, RPR-positive status indicating current or past syphilis, having unprotected sex as evidenced by STI and current STI symptoms were risk factors for HIV infection among Haitian young people, as has been reported in other settings.20,6
There were gaps in young people's understanding about HIV transmission. The sociodemographic risk factors of poverty, poor family support structures, low educational levels, and poor acceptance of family planning all suggest the benefits of building self-esteem and sexual negotiating skills for young women.21,5
Later sexual debut was not protective from HIV in women or men. In fact, we found a paradoxical effect of later sexual debut as a risk factor for HIV infection in males. Other familiar HIV risk factors were confirmed in young Haitian men: use of illicit drugs, having a high self-perceived risk of HIV, sexual initiation with casual or unknown people, and a history of practicing genital-anal sex.22,17 Social influences on HIV risk may be lower among boys in whom emerging sexual drive and sexually aggressive cultural norms shared with other young men are more dominant behavioral influences 23,24. Other studies have also found stronger risk-related predictors among women than men in disadvantaged communities.25,19
Basic knowledge observed among Haitian young people prior to their seeking VCT was not consistent across subtopics, as with ignorance of mother-to-child transmission among about half the youth. Similar to other studies,26,27 most young people (87.5%) knew about the main methods of HIV/AIDS prevention, including sexual abstinence, condom use, maintaining a monogamous relationship, and avoiding the use of unsafe injecting materials. Being better educated or more knowledgeable about HIV did not correlate with lower HIV rates among young men, though young women were less likely to be infected if they were more knowledgeable. Other studies also suggest that the level of HIV awareness has little effect on risk behaviors.28,29 Also, high awareness of the routes of transmission co-exists with conspicuous misconceptions about HIV causation, especially among females. Similar misconceptions related to HIV transmission and prevention have been reported in Haiti28 and other diverse communities.30
Poverty for young Haitians (≈75% live at or below poverty levels) is likely to foster risk behaviors among young people who may be unemployed, may exchange sex for support, or may lack self-esteem or a sense of future possibilities.20,31 Many adolescents in Haiti, as with other Caribbean countries, have grown up in female-headed, low-income households or in homes with domestic family problems;25 parents may not educate their children about reproductive and sexual health 32,6 and may struggle to offer supervision and education to children.
Condoms were not used often (≈26% of females and 31% of males reported some use), demonstrating the gap from knowledge to practice.22 We also observed a low prevalence of use of family planning (≈24% for both sexes). Birth control methods were not associated with HIV.22 Other Caribbean studies have also found a low rate of condom use among young women despite their relatively high levels of STI/HIV awareness.5,33 A school-based survey in Haiti found that only 18% of sexually active adolescents reported having always or sometimes used condoms.15 Adolescent and young adult men often report at least 2 sexual partners in the past 6 months, suggesting high sexual mixing rates.34 However, our study did not document that having multiple sexual partners increased HIV risk. We found a strong correlation between having a child and having HIV, independent of older age. Both low socioeconomic status and young age are more likely to put young women at higher risk for having unprotected sex, which results in pregnancy or childbearing as well as STI and HIV. Among Haitian couples, there are power disparities. The lack of economic resources available to Haitian women causes them to remain dependent upon men for support. This pattern begins in adolescents as Haitian girls and young women typically leave sexual decision making (as with condom use) to the male partner.5,22
A female predominance was seen in our study, though this gender differential is even more dramatic in southern Africa.35 Previous studies report that Caribbean young women (15-24 years) are more likely to be HIV infected than men.20,5 As seen by others, monogamous young women are often infected through their partners' risky behaviors.36 As a result of their socioeconomic and cultural dependence on men, women often have partners who are older and engage in sexual encounters involving gifts, money, and/or illegal drug use.37 Our VCT screening questionnaire was not detailed enough to investigate these issues thoroughly. Our results are compatible with previous findings that sex with commercial sex workers is common among young men (≈40%) who make most of the sexual decisions.14,34,36 Men have been culturally encouraged to have an early sexual debut, to have multiple sexual partners, and there are peer pressures to use illicit drugs.38 Poverty is thought to be one lid on drug use; persons simply have no money to buy them. High prevalence of STI, especially syphilis, and its relation to HIV infection is well documented in previous reports from Haiti.20,19
As noted elsewhere, self-perceived susceptibility for HIV was a predictor for HIV seropositivity among young Haitians.6 Women in Lusaka, Zambia could not self-assess their risks for HIV as well as Haitian female youth could.30 We speculate that Haitian women better perceive their vulnerability, with a higher awareness of their partners' risky behaviors, e.g., multiple sexual partners, and unprotected sexual activity or non-use of condoms.
Our VCT clinic-based study had several strengths. The results of the study may be generalizable to the urban and semi-urban areas because the samples could represent the general population of Haitian youth, many of whom were illiterate (≈40%) with about 50% of households living in extreme poverty, and many not enrolled in school and unemployed (47.4%).39 The adolescent/young adult population and the large sample size permitted a more comprehensive risk assessment than published heretofore among Haitian youth. Nearly all persons in VCT were willing to answer the survey questions (data not shown). The environment was ideal for ethical research since VCT is done within an Integrated Primary Health Care service.
The limitations of our cross-sectional study include the temporal ambiguity inherent in assessing causal or predictive inferences from these associations. We may have encountered social response bias in some subjects since highly sensitive sexual behavioral information relied on self-report. Since the questionnaire was administrated in standardized fashion by trained interviewers, we sought to minimize information bias, but surely could not eliminate it. Whether self-report measures of risk behavior are indeed reliable and valid is debatable and varies by circumstances.27,18 We also were limited by our comparatively smaller number of male participants. We recognize that subjects may not represent the general population since they may have come for HIV testing for a variety of reasons.
Risky sexual behaviours among adolescents and youth in Haiti must be reduced. Combined approaches are essential for success.40,41,42 Education should emphasize the need to use condoms in all sexual encounters outside the primary relationship, but must be accompanied by female skill building for more successful sexual negotiation and risk reduction (e.g., delayed sexual debut, condom use). Changing cultural norms to reduce partner numbers and concurrency is challenging but essential; the adolescent age group is an important target pool. Adolescent-friendly services are very well-received in Port-au-Prince and have increased VCT uptake and adherence to needed care. Since the completion of this study, GHESKIO has created a youth-dedicated VCT clinic to better serve the burgeoning needs of this age group. Nationwide expansion of such services can help control HIV in Haiti by bringing infected persons to care and ART, reducing their risk to others and improving their own health status.
Acknowledgments
The authors thank the participants and research staff of the GHESKIO for their contribution to the study. We also thank Abdis Marcelin and Dale Plummer Jr. for their assistance with the data base, and Julie Lankford and Meredith Bortz for their editorial assistance.
Sources of support: This work was supported in part by a supplement from the President's Emergency Plan for AIDS Relief (PEPFAR) to National Institutes of Health grant U01 AI069923 for the Caribbean, Central America, and South America network (CCASAnet) of the International Epidemiology Databases to Evaluate AIDS (IeDEA) program. The work was also supported in part by received support from the Vanderbilt-Meharry Center for AIDS Research (NIH grant P30 AI054999). UNICEF supports the adolescent HIV clinic at GHESKIO. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or UNICEF.
Footnotes
Presented as Abstract 606 at the 15th Conference on Retroviruses and Opportunistic Infections, February 4–6, 2008, Boston, MA, USA.
Conflicts of interest: None declared.
References
- 1.UNDP Human Development Report. 2008 http://hdr.undp.org/en/statistics/
- 2.Gaillard EM, Boulos LM, ndre Cayemittes MP, et al. Understanding the reasons for decline of HIV prevalence in Haiti. Sex Transm Infect. 2006;82 1:i14–i20. doi: 10.1136/sti.2005.018051. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.UNAIDS. Report on the global AIDS epidemic. 2008. [Google Scholar]
- 4.Calleja JM, Walker N, Cuchi P, et al. Status of the HIV/AIDS epidemic and methods to monitor it in the Latin America and Caribbean region. AIDS. 2002;16 3:S3–12. doi: 10.1097/00002030-200212003-00002. [DOI] [PubMed] [Google Scholar]
- 5.Kershaw TS, Small M, Joseph G, et al. The influence of power on HIV risk among pregnant women in rural Haiti. AIDS Behav. 2006;10:309–318. doi: 10.1007/s10461-006-9072-z. [DOI] [PubMed] [Google Scholar]
- 6.Lerand SJ, Ireland M, Blum RW. Individual and environmental impacts on sexual health of Caribbean youth. ScientificWorldJournal. 2006;6:707–717. doi: 10.1100/tsw.2006.150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Peck R, Fitzgerald DW, Liautaud B, et al. The feasibility, demand, and effect of integrating primary care services with HIV voluntary counseling and testing: evaluation of a 15-year experience in Haiti, 1985-2000. J Acquir Immune Defic Syndr. 2003;33:470–475. doi: 10.1097/00126334-200308010-00007. [DOI] [PubMed] [Google Scholar]
- 8.World Health Organization. Report of a Study Group on Young People and Health for All by the Year 2000. Technical Report Series, No 731. Geneva: World Health Organization; 2003. Young people's health – a challenge for society. 1986. [PubMed] [Google Scholar]
- 9.Fitzgerald DW, Maxi A, Marcelin A, et al. Notification of positive HIV test results in Haiti: can we better intervene at this critical crossroads in the life of HIV-infected patients in a resource-poor country? AIDS Patient Care STDS. 2004;18:658–664. doi: 10.1089/apc.2004.18.658. [DOI] [PubMed] [Google Scholar]
- 10.Severe P, Leger P, Charles M, et al. Antiretroviral therapy in a thousand patients with AIDS in Haiti. N Engl J Med. 2005;353:2325–2334. doi: 10.1056/NEJMoa051908. [DOI] [PubMed] [Google Scholar]
- 11.Schackman BR, Neukermans CP, Fontain SN, et al. Cost-effectiveness of rapid syphilis screening in prenatal HIV testing programs in Haiti. PLoS Med. 2007;4:e183. doi: 10.1371/journal.pmed.0040183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.The Global HIV/AIDS pandemic, MMWR Morb Mortal Wkly Rep. Vol. 55. 2006. pp. 841–844. [PubMed] [Google Scholar]
- 13.Bradley H, Bedada A, Tsui A, et al. HIV and family planning service integration and voluntary HIV counselling and testing client composition in Ethiopia. AIDS Care. 2008;20:61–71. doi: 10.1080/09540120701449112. [DOI] [PubMed] [Google Scholar]
- 14.Bernier M, Ascensio P. Street children and AIDS in Haiti. Sante. 1995;5:125–130. [PubMed] [Google Scholar]
- 15.Holschneider SO, Alexander CS. Social and psychological influences on HIV preventive behaviors of youth in Haiti. J Adolesc Health. 2003;33:31–40. doi: 10.1016/s1054-139x(02)00418-4. [DOI] [PubMed] [Google Scholar]
- 16.Rodriquez B, Steel-Duncan JC, Pierre R, et al. Socio-demographic characteristics of HIV-exposed and HIV-infected Jamaican children. West Indian Med J. 2004;53:303–307. [PubMed] [Google Scholar]
- 17.Ohene SA, Ireland M, Blum RW. The clustering of risk behaviors among Caribbean youth. Matern Child Health J. 2005;9:91–100. doi: 10.1007/s10995-005-2452-6. [DOI] [PubMed] [Google Scholar]
- 18.Blum RW, Halcon L, Beuhring T, et al. Adolescent health in the Caribbean: risk and protective factors. Am J Public Health. 2003;93:456–460. doi: 10.2105/ajph.93.3.456. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Smith Fawzi MC, Lambert W, Singler JM, et al. Factors associated with forced sex among women accessing health services in rural Haiti: implications for the prevention of HIV infection and other sexually transmitted diseases. SocSci Med. 2005;60:679–689. doi: 10.1016/j.socscimed.2004.06.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Fitzgerald DW, Behets F, Caliendo A, et al. Economic hardship and sexually transmitted diseases in Haiti's rural Artibonite Valley. Am J Trop Med Hyg. 2000;62:496–501. doi: 10.4269/ajtmh.2000.62.496. [DOI] [PubMed] [Google Scholar]
- 21.Malow RM, Cassagnol T, McMahon R, et al. Relationship of psychosocial factors to HIV risk among Haitian women. AIDS Educ Prev. 2000;12:79–92. [PubMed] [Google Scholar]
- 22.Boulos ML, Boulos R, Nichols DJ. Perceptions and practices relating to condom use among urban men in Haiti. Stud Fam Plann. 1991;22:318–325. [PubMed] [Google Scholar]
- 23.Rowe R, Maughan B, Worthman CM, et al. Testosterone, antisocial behavior, and social dominance in boys: pubertal development and biosocial interaction. Biol Psychiatry. 2004;55:546–552. doi: 10.1016/j.biopsych.2003.10.010. [DOI] [PubMed] [Google Scholar]
- 24.Singer SM, Willms DG, Adrien A, et al. Many voices--sociocultural results of the ethnocultural communities facing AIDS study in Canada. Can J Public Health. 1996;87 1:S26–S35. [PubMed] [Google Scholar]
- 25.Lammers C, Ireland M, Resnick M, et al. Influences on adolescents' decision to postpone onset of sexual intercourse: a survival analysis of virginity among youths aged 13 to 18 years. J Adolesc Health. 2000;26:42–48. doi: 10.1016/s1054-139x(99)00041-5. [DOI] [PubMed] [Google Scholar]
- 26.Walrond E, Jones F, Hoyos M, et al. An AIDS-related knowledge, attitudes, beliefs, and practices survey among schoolchildren in Barbados. Bull Pan Am Health Organ. 1992;26:208–219. [PubMed] [Google Scholar]
- 27.Robillard HH. The Jamaican adolescent: an assessment of knowledge and attitudes regarding HIV/AIDS. Pediatr Nurs. 2001;27:176–179. [PubMed] [Google Scholar]
- 28.Adrien A, Cayemittes M, Bergevin Y. AIDS-related knowledge, attitudes, beliefs, and practices in Haiti. Bull Pan Am Health Organ. 1993;27:234–243. [PubMed] [Google Scholar]
- 29.O'Toole BJ, McConkey R, Casson K, et al. Knowledge and attitudes of young people in Guyana to HIV/AIDS. Int J STD AIDS. 2007;18:193–197. doi: 10.1258/095646207780132398. [DOI] [PubMed] [Google Scholar]
- 30.Stringer EM, Sinkala M, Kumwenda R, et al. Personal risk perception, HIV knowledge and risk avoidance behavior, and their relationships to actual HIV serostatus in an urban African obstetric population. J Acquir Immune Defic Syndr. 2004;35:60–66. doi: 10.1097/00126334-200401010-00009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Farmer P. Political violence and public health in Haiti. N Engl J Med. 2004;350:1483–1486. doi: 10.1056/NEJMp048081. [DOI] [PubMed] [Google Scholar]
- 32.Markham CM, Tortolero SR, Escobar-Chaves SL, et al. Family connectedness and sexual risk-taking among urban youth attending alternative high schools. Perspect Sex Reprod Health. 2003;35:174–179. doi: 10.1363/psrh.35.174.03. [DOI] [PubMed] [Google Scholar]
- 33.Norman LR, Figueroa JP, Wedderburn M, et al. Trends in HIV risk perception, condom use and sexual history among Jamaican youth, 1996-2004. Int J Adolesc Med Health. 2007;19:199–207. doi: 10.1515/ijamh.2007.19.2.199. [DOI] [PubMed] [Google Scholar]
- 34.Connell P, McKevitt C, Low N. Investigating ethnic differences in sexual health: focus groups with young people. Sex Transm Infect. 2004;80:300–305. doi: 10.1136/sti.2003.005181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Pettifor AE, Kleinschmidt I, Levin J, et al. A community-based study to examine the effect of a youth HIV prevention intervention on young people aged 15-24 in South Africa: results of the baseline survey. Trop Med Int Health. 2005;10:971–980. doi: 10.1111/j.1365-3156.2005.01483.x. [DOI] [PubMed] [Google Scholar]
- 36.Curtis SL, Sutherland EG. Measuring sexual behaviour in the era of HIV/AIDS: the experience of Demographic and Health Surveys and similar enquiries. Sex Transm Infect. 2004;80 2:ii22–ii27. doi: 10.1136/sti.2004.011650. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Miller KS, Clark LF, Moore JS. Sexual initiation with older male partners and subsequent HIV risk behavior among female adolescents. Fam Plann Perspect. 1997;29:212–214. [PubMed] [Google Scholar]
- 38.Fernandez DM, Gomez ML, Figueroa W, et al. A comparison of the sociodemographic, risk-behavior, and substance-abuse profile of young vs older HIV-infected Puerto Rican patients. Ethn Dis. 2005;15:S5–S9. [PubMed] [Google Scholar]
- 39.Justesen M, Verner D. Factors Impacting Youth Development in Haiti. World bank Policy Research Working Paper No 4110. 2007 [Google Scholar]
- 40.Ross DA, Changalucha J, Obasi AI, et al. Biological and behavioural impact of an adolescent sexual health intervention in Tanzania: a community-randomized trial. AIDS. 2007;21:1943–1955. doi: 10.1097/QAD.0b013e3282ed3cf5. [DOI] [PubMed] [Google Scholar]
- 41.Coates TJ, Richter L, Caceres C. Behavioural strategies to reduce HIV transmission: how to make them work better. Lancet. 2008;372:669–684. doi: 10.1016/S0140-6736(08)60886-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Jewkes R, Nduna M, Levin J, et al. Impact of stepping stones on incidence of HIV and HSV-2 and sexual behaviour in rural South Africa: cluster randomised controlled trial. BMJ. 2008;337:a506. doi: 10.1136/bmj.a506. [DOI] [PMC free article] [PubMed] [Google Scholar]