Abstract
Haiti has the greatest burden of HIV in the Caribbean. In 2018, the country’s HIV prevalence was 2% with an estimated 2,200 AIDS-related deaths. Using 2016–2017 Demographic and Health Survey (DHS) data, Haitian men’s self-reported ever-having HIV testing was analyzed with a focus on their regions of residence. Only 34% of the men reported ever-having HIV testing. Men who lived in Northern region (aOR:1.59, 95%CI:1.23–2.05), and Southern region (aOR:1.26, 95%CI:1.04–1.53) had higher odds of ever-having HIV testing compared to men residing in Central region. Further research should prioritize targeted health promotion for engaging Haitian men who are younger, poorer, with low-level of education and single as well as those who reside in regions where HIV testing is not easily accessible. Haitian men who have not had an HIV test must first be identified, and demographic-specific interventions and programming should be used to increase HIV testing among this population.
Keywords: Haiti, HIV, Men, Regional, Testing
Introduction
Although the Caribbean region has experienced a decline in HIV incidence rates (Garcia et al., 2014), HIV infections remain a public health concern for several Caribbean island nations, including Haiti. Haiti is the second most populated country in the Caribbean with an estimated population of 10.9 million persons (The World Bank, 2020). In 2018, Haiti reported an estimated 160,000 people living with HIV (PLWH) (UNAIDS, 2020), which accounted for an HIV prevalence rate of 2% with an estimated 2,200 AIDS-related deaths in 2018 (CDC, 2019). In 2018, only two-thirds (67%) of all PLWH knew their HIV positive status and only 58% of all PLWH were on treatment (UNAIDS, 2020). In order to attain the UNAIDS first 95 of the 95–95-95 goals, which calls for 95% of all PLWH knowing their status, 95% of those who know their status being on treatment, and 95% of those on treatment achieving virally suppression by 2030 (UNAIDS, 2014), Haiti must ramp up its HIV testing initiatives. However, in Haiti, there remains unmet infrastructure and intuitional challenges resulting from the 2010 earthquake (World Bank, 2019) which may impede upon HIV-testing efforts; consequently, HIV testing has faced challenges in the country. A previous study found that younger, single men were less likely to have tested for HIV (Bristow et al., 2017). Another study found that rural men aged 15–24 had lower odds of HIV testing in three Caribbean countries, including Haiti (Andrews, 2011). Conserve et al. (2017) found that older men were more likely to have been tested than younger men and the likelihood of testing increased with educational attainment in Haiti. However, few studies have specifically investigated the geographic variations in HIV testing for men in Haiti. Although, previous research examined community-based HIV and health testing in seven slums neighborhoods in Port-au-Prince (Reif et al., 2016) and provider-initiated HIV testing in rural Haiti (Ivers et al., 2007), there are significant gaps in the literature regarding the regional differences of HIV testing among Haitian men and the regions which require more targeted focus. This paper aims to investigate the association of geographic regions and HIV testing uptake by Haitian men after controlling for other covariates.
Methods
Participants
This study used the 2016–17 Haiti Demographic and Health Survey (HDHS) which included two-stage stratified sampling techniques (USAID, n.d.). This paper used data collected via the HDHS men’s questionnaire. A total of 11,818 men aged 15–64 were interviewed for HDHS, but due to incomplete data, the analysis was restricted to 9,754 men.
Variables
Men’s self-reported HIV testing was the dependent variable with a dichotomous response (yes/no). Independent variables were selected based on the existing literature on self-reported HIV testing (Conserve et al., 2017; Haider et al., 2020). Sociodemographic variables included age (15–19, 20–29, 30–39, 40–49, 50–59 years), education (no education, primary, secondary, higher), religion (no religion, Catholic, Protestant, Vaudousant and other), residence (rural, urban), marital status (never-in-union, married, free union, widowed/separated/divorced), health insurance (no, yes), and wealth index (poorest, poorer, middle, richer, richest). Region was categorized as: Metro (metropolitan area of Port-au-Prince), North (Nord, Nord-est, and Nord-ouest departments), Central (Artibonite, Centre, and Rest-ouest), and South (Sud, Sud-est, Grand’Anse, and Nippes).
HIV-related Knowledge and Stigma
Knowledge of HIV transmission and prevention was assessed using nine items (e.g. whether people knew they could reduce the risk of getting HIV by using condoms every time they have sexual intercourse, or by having one sexual partner who is not infected and has no other partners; whether a healthy-looking person can have HIV, etc.) Correct responses were coded as 1 and incorrect or uncertain responses were coded as zero.
HIV-related perceived stigma was assessed using five items (e.g. whether people hesitate to take HIV tests because of possible negative reactions from others if a positive result was returned; whether they would purchase fresh vegetables from a shopkeeper who has HIV; believe children with HIV should be allowed to attend school with children without HIV, etc.) Positive responses were coded as 1, while negative responses were coded as 0, and not sure/did not know were coded as 0.5.
Items were summed to create scale score for both scales with higher scores indicating greater HIV-related knowledge and stigma. Based on the scores, respondents were categorized into three tertiles (high, medium, low) for both scales. Cronbach’s alpha for HIV-related knowledge and stigma scales was 0.40 and 0.47.
Statistical Analysis
Descriptive statistics (frequencies and percentages) were calculated for all variables. Bivariate (Chi-square tests) analyses were performed to assess the association between dependent and independent variables. Multilevel multivariable logistic regression with region as the random effect variable and all other variables with significant bivariate relationship with HIV testing as fixed effect variables was used to determine the statistically significant covariates for self-reported HIV testing among Haitian men. All analyses were survey-weighted and performed using Stata 14.2 (StataCorp, 2015).
Results
Table 1 presents descriptive statistics of the respondent’s sociodemographic characteristics. About one-third (34.0%) of men reported ever-having HIV testing. Most men (51.3%) reported medium-levels of perceived stigma, while 36.5% reported low and 38.6% reported medium-levels of HIV knowledge.
Table 1.
Demographic Characteristics of Participants in 2016-2-17 Haiti Men’s Demographic Health Survey, (N=11,818)
| Variables | Survey-Weighted Frequency | % |
|---|---|---|
| Ever Had HIV Test | ||
| No | 7,795 | 66.0 |
| Yes | 4,023 | 34.0 |
| Age Group (Years) | ||
| 15–19 | 2,080 | 17.6 |
| 20–29 | 2,759 | 23.3 |
| 30–39 | 2,404 | 20.3 |
| 40–49 | 2,109 | 17.9 |
| 50–64 | 2,465 | 20.9 |
| Education | ||
| No education | 2,104 | 17.8 |
| Primary | 3,557 | 30.1 |
| Secondary | 5,109 | 43.2 |
| Higher | 1,048 | 8.9 |
| Religion | ||
| No Religion | 1,722 | 14.6 |
| Catholic | 4,631 | 39.2 |
| Protestant | 5,126 | 43.4 |
| Vaudousant and Other | 338 | 2.9 |
| Residence | ||
| Rural | 6,972 | 59.0 |
| Urban | 4,846 | 41.0 |
| Marital Status | ||
| Never-in-union | 4,646 | 47.4 |
| Married | 4,292 | 43.8 |
| Free union | 249 | 2.5 |
| Other (Widowed/Separated/ Divorced) | 608 | 6.2 |
| Health Insurance | ||
| No | 11,243 | 95.1 |
| Yes | 575 | 4.9 |
| Wealth Index | ||
| Poorest | 2,162 | 18.3 |
| Poorer | 2,180 | 18.5 |
| Middle | 2,427 | 20.5 |
| Richer | 2,396 | 20.3 |
| Richest | 2,652 | 22.4 |
| HIV related Stigma * | ||
| Low | 4,029 | 34.1 |
| Medium | 6,068 | 51.3 |
| High | 1,721 | 14.6 |
| Knowledge on HIV ** | ||
| Low | 4,316 | 36.5 |
| Medium | 4,567 | 38.6 |
| High | 2,935 | 24.8 |
| Justified for a woman to ask her husband to use a condom | ||
| No | 732 | 7.5 |
| Yes | 9,023 | 92.5 |
| Justified for a husband to hit or beat his wife if she refuses to have sex with him | ||
| No | 9,557 | 97.6 |
| Yes | 236 | 2.4 |
| Region | ||
| Metro | 2,683 | 22.7 |
| North | 2,204 | 18.6 |
| Central | 4,321 | 36.6 |
| South | 2,610 | 22.1 |
HIV-related stigma: Five items included in the scale were whether they would purchase fresh vegetables from a shopkeeper with HIV, believe children with HIV should be allowed to attend school with other children, perceived that people hesitate to take HIV test because of possible negative reactions from others if a positive result was returned, people talk negatively about persons who have or are believed to have HIV, and whether persons who have or are believed to have HIV lose respect from other people.
HIV knowledge: Nine items included in the scale were: whether people knew they could reduce the risk of getting HIV by using condoms every time they have sexual intercourse, or by having one sexual partner who is not infected and has no other partners; whether a healthy-looking person can have HIV; whether HIV cannot be transmitted through mosquito bite, saliva of HIV-infected person, or food, but can be transmitted during pregnancy, during delivery, by breastfeeding; and whether medications are available to avoid HIV transmission to babies during pregnancy.
Table 2 shows the results of the bivariate associations between ever-been tested for HIV and other covariates. Statistically significant associations were observed among ever-been tested for HIV and age, education, religion, residence, marital status, health insurance, wealth index, HIV-related stigma, HIV knowledge, and region.
Table 2.
Survey-weighted Bivariate Analysis: Ever Had HIV Tests Done by Demographic and Socio-Economic Characteristics in Haiti (2016–2017). (N=11,818)
| Variables | Ever had HIV test (N=11,818) | ||
|---|---|---|---|
| No% (n) | Yes% (n) | p-Value | |
| Total | 66.0 (7,795) | 34.0 (4,023) | |
| Age Group (Years) | <0.0001 | ||
| 15–19 | 88.5 (1,841) | 11.5 (239) | |
| 20–29 | 53.5 (1,476) | 46.5 (1,283) | |
| 30–39 | 48.4 (1,163) | 51.6 (1,241) | |
| 40–49 | 67.0 (1,413) | 33.0 (696) | |
| 50–64 | 77.2 (1,902) | 22.9 (563) | |
| Education | <0.0001 | ||
| No education | 86.2 (1,814) | 13.8 (290) | |
| Primary | 76.2 (2,712) | 23.8 (845) | |
| Secondary | 58.1 (2,966) | 41.9 (2,142) | |
| Higher | 28.9 (303) | 71.1 (745) | |
| Religion | <0.0001 | ||
| No Religion | 72.0 (1,241) | 28.0 (482) | |
| Catholic | 67.9 (3,145) | 32.1 (1,486) | |
| Protestant | 62.5 (3,204) | 37.5 (1,921) | |
| Vaudousant and Other | 60.6 (205) | 39.4 (133) | |
| Residence | <0.0001 | ||
| Rural | 73.9 (5,154) | 26.1 (1,817) | |
| Urban | 54.5 (2,641) | 45.5 (2,205) | |
| Marital Status | <0.0001 | ||
| Never in union | 69.0 (3,204) | 31.0 (1,442) | |
| Married | 49.6 (2,129) | 50.4 (2,163) | |
| Free union | 42.3 (105) | 57.7 (144) | |
| Other (Widowed/Separated/ Divorced) | 54.9 (334) | 45.1 (274) | |
| Health Insurance | <0.0001 | ||
| No | 67.5 (7,589) | 32.5 (3,654) | |
| Yes | 35.9 (206) | 64.1 (368) | |
| Wealth Index | <0.0001 | ||
| Poorest | 85.7 (1,852) | 14.3 (310) | |
| Poorer | 76.7 (1,671) | 23.3 (509) | |
| Middle | 66.8 (1,622) | 33.2 (805) | |
| Richer | 57.8 (1,384) | 42.2 (1,012) | |
| Richest | 47.7 (1,266) | 52.3 (1,386) | |
| HIV related Stigma | <0.0001 | ||
| Low | 72.3 (2,914) | 27.7 (1,115) | |
| Medium | 60.1 (3,644) | 39.9 (2,424) | |
| High | 71.9 (1,237) | 28.1 (484) | |
| Knowledge on HIV | <0.0001 | ||
| Low | 84.3 (3,638) | 15.7 (678) | |
| Medium | 59.5 (2,719) | 40.5 (1,848) | |
| High | 49.0 (1,438) | 51.0 (1,497) | |
| Region | <0.0001 | ||
| Metro | 52.1 (1,398) | 47.9 (1,285) | |
| Central | 72.2 (3,120) | 27.8 (1,201) | |
| North | 64.1 (1,412) | 35.9 (792) | |
| South | 71.5 (1,865) | 28.5 (744) | |
Table 3 presents the results for the multilevel multivariable logistic regression model. Men who lived in Northern region (aOR:1.59, 95%CI:1.23–2.05), and Southern region (aOR:1.26, 95%CI:1.04–1.53) had higher odds of ever-having HIV testing compared to men residing in Central region. Men who reported medium levels of HIV-related stigma (aOR:0.71, 95%CI:0.57–0.88) and high levels of stigma (aOR:0.56, 95% CI:0.42–0.76) reported lower chances of ever-having HIV testing compared to men who reported low levels of HIV-related stigma. Men who reported high HIV knowledge (aOR:1.34, 95%CI:1.10–1.65) compared to those who reported low HIV knowledge had higher odds of ever-having HIV testing. Men aged 20–29 years (aOR: 5.22, 95%CI:2.85–9.54), 30–39 years (aOR:8.42, 95% CI:3.88–18.28), 40–49 years (aOR:6.67, 95%CI:3.31–13.44), and 50–64 years (aOR:4.51, 95%CI:2.57–7.91) reported higher odds of ever-having HIV testing compared to men of 15–19 years. Men who reported completing primary education (aOR:2.09, 95%CI:1.49–2.92), secondary education (aOR:4.44, 95%CI:2.35–8.39), and higher education (aOR:11.92, 95%CI:4.43–32.08) were more likely to ever-having HIV testing compared to those who reported no education. Men who were married (aOR:2.32, 95%CI:1.59–3.38), in a free union (aOR:2.58, 95%CI:1.50–4.42), or other marital status (widowed, separated, and divorced) (aOR:2.01, 95%CI:1.30–3.11) compared to those who were never in union were more likely to report ever-having HIV testing.
Table 3.
Survey-weighted Multilevel Logistic Regression of Ever Had HIV Test in Haiti, 2016–2017, (N=9,795)
| Variables | Ever had HIV test (N=9,754) | |||
|---|---|---|---|---|
| Adjusted Odds Ratio | 95% CI | p-Value | ||
| Lower Limit | Upper Limit | |||
| Age Group (Years) | ||||
| 15–19 | 1.00 | |||
| 20–29 | 5.26 | 3.13 | 8.83 | <0.001 |
| 30–39 | 8.44 | 4.33 | 16.46 | <0.001 |
| 40–49 | 6.64 | 3.65 | 12.10 | <0.001 |
| 50–64 | 4.45 | 2.72 | 7.30 | <0.001 |
| Education | ||||
| No education | 1.00 | |||
| Primary | 2.07 | 1.54 | 2.78 | <0.001 |
| Secondary | 4.36 | 2.52 | 7.56 | <0.001 |
| Higher | 11.96 | 5.13 | 27.89 | <0.001 |
| Religion | ||||
| No Religion | 1.00 | |||
| Catholic | 1.17 | 0.96 | 1.44 | 0.128 |
| Protestant | 1.32 | 1.07 | 1.62 | 0.010 |
| Vaudousant and Other | 1.54 | 0.97 | 2.44 | 0.070 |
| Residence | ||||
| Rural | 1.00 | |||
| Urban | 1.10 | 0.89 | 1.35 | 0.370 |
| Marital Status | ||||
| Never in union | 1.00 | |||
| Married | 2.32 | 1.64 | 3.28 | <0.001 |
| Free union | 2.56 | 1.54 | 4.25 | <0.001 |
| Other (Widowed/Separated/ Divorced) | 2.02 | 1.33 | 3.08 | 0.001 |
| Health Insurance | ||||
| No | 1.00 | |||
| Yes | 1.95 | 1.30 | 2.93 | 0.001 |
| Wealth Index | ||||
| Poorest | 1.00 | |||
| Poorer | 1.59 | 1.21 | 2.09 | 0.001 |
| Middle | 2.28 | 1.66 | 3.14 | <0.001 |
| Richer | 3.00 | 1.99 | 4.51 | <0.001 |
| Richest | 3.63 | 2.35 | 5.61 | <0.001 |
| HIV related Stigma | ||||
| Low | 1.00 | |||
| Medium | 0.71 | 0.58 | 0.87 | 0.001 |
| High | 0.56 | 0.43 | 0.74 | <0.001 |
| Knowledge on HIV | ||||
| Low | 1.00 | |||
| Medium | 1.15 | 0.96 | 1.38 | 0.128 |
| High | 1.37 | 1.12 | 1.67 | 0.002 |
| Region | ||||
| Metro | 1.00 | |||
| Central | 1.16 | 0.94 | 1.43 | 0.158 |
| North | 1.39 | 1.11 | 1.75 | 0.004 |
| South | 1.55 | 1.22 | 1.97 | <0.001 |
| Latent Variable (M) | 2.72 | |||
Discussion
The data in the current study indicate that men who resided in the Northern and Southern region of the country had higher odds of HIV testing than those who lived in Central region, even though some areas in the Central region reported a high prevalence of PLWH (PEPFAR, 2019). Thus, a lower likelihood of HIV testing among men may result in lower detection of PLWH and higher chance of HIV transmission. Results also showed that high HIV-related stigma and low HIV transmission knowledge were associated with lower odds of HIV testing; therefore, health promotion to reduce stigma and increase knowledge should be prioritized among Haitian men.
Our study results showed that only one-third of Haitian men reported previous HIV testing; this is consistent with the 35% of men reported ever-having HIV testing in another study (Conserve et al.,2017). In past research, only 25% of young adolescent and adult men had been tested for HIV in Haiti (Dorjgochoo et al., 2009), supporting our finding of older age groups reporting higher odds of ever-been tested compared to younger men.
With such high proportions of men having never been tested, these results indicate that young men in Haiti should be targeted for HIV testing programming and uptake. Results showed that men with higher educational attainment and income were more likely to ever-having HIV testing. This finding is supported by both education and income being found to be significant predictors of HIV testing for men in earlier research (Ogunmola et al., 2014). Men who were married, in a free union, or who were ever married were more likely to ever-having testing compared to never-married men. Consequently, unmarried men should also be prioritized for HIV testing programing, specifically, those who fall into the poorest and/or poorer wealth index. Still, in order to increase access to and uptake of HIV testing in all Haitian regions, better linkage to HIV testing and care are required, as well as improved education and intervention offerings.
This study is subject to limitations. Due to the cross-sectional nature of the HDHS, it is impossible to determine causation. All data in the HDHS is self-reported and there is no mechanism in place to verify the credibility of responses.
Conclusion
There is a geographic variation in HIV testing among Haitian men. Haitian men who have not had an HIV test must first be identified; demographic-specific interventions and programming should then be used to increase HIV testing among this population.
Acknowledgments
The authors would like to thank USAID for their work on the DHS data and for making the data available to the public, as well as granting permission to use the data.
Funding
A.H. thanks the Southern Regional Education Board (SREB) and the Grace Jordan McFadden Professor’s Program at the University of South Carolina for their continued financial support. D.F.C was supported by a grant from the National Institute of Mental Health (Grant #R00MH110343: PI: D.F.C.). M.J.B. is also supported by a grant from the National Institute of Mental Health (Grant #K01MH115794).
Footnotes
Ethics
The authors do not have any ethical issues to report.
Conflicts of Interest
The authors do not have any conflicts of interests to disclose.
References
- Andrews BE (2011). Prevalence and correlates of HIV testing among Caribbean youth. International Journal of STD & AIDS, 22(12); 722–726. doi: 10.1258/ijsa.2011.011088. [DOI] [PubMed] [Google Scholar]
- Bristow CC, Lee SJ, Severe L, William Pape J, Javanbakht M, Scott Comulada W, Klausner JD (2017). Attributes of diagnostic tests to increase uptake of dual testing for syphilis and HIV in Port-au-Prince, Haiti. International Journal of STD & AIDS, 28(3); 259–264. doi: 10.1177/0956462416642340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Centers for Disease Control and Prevention (CDC). (2019). Haiti country profile. Retrieved from: https://www.cdc.gov/globalhivtb/where-we-work/haiti/haiti.html.
- Conserve DF, Iwelunmor J, Whembolua GL, Sofolahan-Oladeinde Y, Teti M, Surkan PJ (2017). Factors associated with HIV testing among men in Haiti: results from the 2012 Demographic and Health Survey. American Journal of Men’s Health, 11(5); 1322–1330. doi: 10.1177/1557988316635247 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dorjgochoo T, Noel F, Deschamps MM, Theodore H, Charles S, Dupont W, Wright PF… (2009). Risk factors for HIV infection among Haitian adolescents and young adults seeking counseling and testing in Port-au-Prince. Journal of Acquired Immune Deficiency Syndrome, 52(4); 498–508. doi: 10.1097/QAI.0b013e3181ac12a8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Garcia PJ, Bayer A & Caramo CP (2014). The changing face of HIV in Latin America and the Caribbean. Current HIV/AIDS Reports, 11(2); 146–157. doi: 10.1007/s11904-014-0204-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Haider MR, Brown MJ, Karim S, Siddiqi KA, Olatosi B, & Li X. (2020). Factors associated with self-reported HIV testing in the Dominican Republic. International Journal of STD & AIDS, 31(6), 560–567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ivers LC, Freedberg KA, Mukherjee JS (2007). Provider-initiated HIV testing in rural Haiti: low rate of missed opportunities for diagnosis of HIV in a primary care clinic AIDS Research Therapy, 4; 28. doi: 10.1186/1742-6405-4-28 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ogunmola OJ, Oladosu YO, Olamoyegun MA (2014). Relationship between Socioeconomic status and HIV infection in a rural tertiary health center. HIV AIDS (Auckland), 6; 61–67. doi: 10.2147/HIV.S59061 [DOI] [PMC free article] [PubMed] [Google Scholar]
- President’s Plan for Emergency AIDS Relief (PEPFAR). (2019). Haiti country operational plan (COP) 2019 strategic direction summary. Retrieved from: https://www.state.gov/wp-content/uploads/2019/09/Haiti_COP19-Strategic-Directional-Summary_public.pdf. [Google Scholar]
- Reif LK, Rivera V, Louis B, Bertand R, Peck M, Anglade B, Seo G, Abrams EJ, Pape JW, Fitzgerald DW, McNairy ML Community-based HIV and health testing for high-risk adolescents and youth. AIDS Patient Care STDS, 30(8); 371–378. doi: 10.1089/apc.2016.0102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- StataCorp. 2015. Stata Statistical Software: Release 14. College Station, TX: StataCorp LLC. [Google Scholar]
- UNAIDS. (2014). 90–90-90: an ambitious treatment target to help end the AIDS epidemic. Retrieved from: https://www.unaids.org/sites/default/files/media_asset/90-90-90_en.pdf.
- UNAIDS. (2020). Haiti. Retrieved from: https://www.unaids.org/en/regionscountries/countries/haiti.
- USAID. (n.d.). The DHS and Health Surveys. Retrieved from: https://dhsprogram.com/what-we-do/survey/survey-display-503.cfm.
- World Bank. (2019). Rebuilding Haitian Infrastructure and Institutions. Retrieved from: https://www.worldbank.org/en/results/2019/05/03/rebuilding-haitian-infrastructure-and-institutions.
- World Bank. (2020). The World Bank in Haiti. Retrieved from: https://www.worldbank.org/en/country/haiti.