https://orcid.org/0000-0003-0312-3201
Abstract
To address high HIV prevalence rates in Haiti, disseminating information about HIV transmission has been emphasized. Yet, after several decades, we do not know how effective HIV information dissemination has been in reducing HIV misconceptions. Using the 2005-06, 2012, and 2016-17 Haiti Demographic and Health Surveys and applying logistic regression, we found nuanced gender dynamics in endorsing HIV misconceptions over time. Among females at the bivariate level, the odds of endorsement of HIV misconceptions in 2012 (OR = 0.87, p < 0.05) and 2016-17 (OR = 0.68, p < 0.001) had declined compared to 2005-06. At the multivariate level, however, we observed that demographic factors suppressed the difference between 2005-06 and 2012, although those in 2016-17 (OR = 0.71, p < 0.001) were still less likely to endorse HIV misconceptions. However, this relationship disappeared once we added behavioral factors (OR = 0.93, p > 0.05). Among males, after controlling for demographic, socioeconomic, and behavioral factors at the multivariate level, those in 2012 (OR = 1.55, p < 0.001) and 2016-17 (OR = 1.24, p < 0.01) were more likely to endorse HIV misconceptions compared to men in 2005-06. We recommend that while improving women’s access to HIV services, it is important to incorporate the HIV needs of males into the National HIV policy priority areas.
Introduction
Over three decades after the first human immunodeficiency virus (HIV) and acquired immunodeficiency syndrome (AIDS) was reported in Haiti, it is still overburdened with new cases of infection [Citation1,Citation2]. The country’s HIV prevalence dropped from as high as 6.2% in 1993 to 2.2% in 2012, with recent data from the Joint United Nations Programme on HIV/AIDS (UNAIDS), suggesting this may have further declined to 1.7% in 2022 [Citation1–3]. Notwithstanding this progress, Haiti has the highest HIV prevalence rate among countries in the Caribbean and the Western Hemisphere, with about 1,600 AIDS-related deaths in 2022 alone [Citation1]. AIDS-related illnesses and comorbidities are among Haiti’s 10 leading causes of death [Citation3].
Over the years, the Government of Haiti, together with its development partners and non-governmental organizations (NGOs), have responded to the country’s HIV epidemic through policy initiatives that are primarily focused on creating awareness about HIV transmission and prevention [Citation4]. The first response to HIV in the country was through the creation of the Haitian Study Group on Kaposi’s Sarcoma and Opportunistic Infections (GHESKIO) in 1982. This group studied the nature of the spread of the virus and recommended reducing HIV misconceptions by disseminating factual information about HIV transmission to citizens [Citation4,Citation5]. This first initiative was closely followed by a mass AIDS awareness campaign in 1986 where mass media platforms, including radio, television, billboards, and direct contact with high-risk populations, were used to disseminate HIV information [Citation6]. In the same year, Partners in Health (PIH), an NGO which has become a major stakeholder in HIV education and treatment in the country, started providing voluntary HIV testing and counselling for Haitians [Citation4,Citation7]. Other efforts by the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR) and the United Nations to increase Haitians exposure to information on HIV transmission has achieved some positive results as Haiti is reported to have made some progress towards achieving the targets for the UNAIDS agenda 95-95-95 [Citation2].
Regardless of these modest gains, however, the HIV prevalence in Haiti calls into question the effectiveness of prevention strategies and interventions in empowering Haitian females and males to reject the endorsement of HIV misconceptions. HIV misconceptions constitute all the beliefs about HIV transmission which have been proven to be clinically inaccurate [Citation8–11]. These HIV misconceptions may include the belief that HIV is spread by supernatural beings or through mosquito bites or that HIV can be cured by sleeping with a virgin or praying to a supernatural being, all of which may work to drive new infections [Citation12–14]. Given the suggestions from earlier studies that HIV misconceptions may still persist in Haiti and may be counterproductive to reducing new HIV infections despite the government’s effort to increase knowledge of HIV transmission over the years [Citation1,Citation15–17], our study seeks to contribute to the HIV literature and policy in Haiti and elsewhere by exploring the endorsement of HIV misconceptions over time among females and males.
Methods
We used the 2005-06, 2012, and 2016-17 Haiti Demographic and Health Surveys (HDHS), which include a nationally representative sample of Haitian women aged 15-49 and men aged 15-59. The HDHS employed a multistage sampling design in which a stratified probability proportional to size sampling methodology was applied. These surveys were conducted by the Haitian Institute for Children in collaboration with the Haitian Bureau of Statistics. Ethics approval and permission to use the DHS data was obtained from the Haitian Bureau of Statistics and Macro, a US-based organization that collaborates with and provides technical assistance to DHS program. For this study, we analyzed the socio-demographic characteristics of the population, limiting the sample to women and men who answered the questions on HIV transmission beliefs.
Measures
The dependent variable for this study was ‘HIV misconception endorsement’. To construct this variable, we used four questions about HIV transmission beliefs, including whether respondents believe that (1) a healthy-looking person cannot be infected with HIV and whether HIV can be spread through (2) witchcraft and other supernatural means, (3) sharing food with a person with HIV, and (4) mosquito bites. We created a binary variable where respondents were coded as ‘1′ if they answered ‘yes’ to at least one of these questions and coded as ‘0′ otherwise (0 = no; 1 = yes). The independent variable is ‘the year of survey’ (0 = 2005-06; 1 = 2012; 2 = 2016-17). To account for possible confounding factors that may further impact the relationship between HIV misconception endorsement and the year of survey, we controlled for demographic (i.e. age, marital status, and place of residence), socioeconomic (i.e. household wealth, education, and employment), and behavioural (i.e. ever tested for HIV and adequate HIV knowledge) factors.
Statistical analysis
We employed two separate analyses: cross-classification analysis and logistic regression analysis. Cross-classification analysis was performed to examine whether the distribution of demographic, socioeconomic, and behavioural characteristics differs across 2005-06, 2012, and 2016-17. For logistic regression analysis, Model 1 explored the bivariate relationship between the relationship between HIV misconception endorsement and the year of survey while Models 2, 3, and 4 further accounted for demographic, socioeconomic, and behavioural factors, respectively. Results were reported with odds ratio. All analyses were carried out using STATA 17 (State Corp, College Station, TX, USA). The ‘svy’ function was applied in statistical analysis to adjust for the cluster sampling design as well as sampling weights.
Results
shows findings from cross-classification analysis. We found that the proportion of women who endorse HIV misconceptions has decreased over time (i.e. 66%, 63%, and 57% for 2005-06, 2012, and 2016-17, respectively). By contrast, for men, the proportion has increased in 2012 (67%) from 2005-06 (61%) although decreased again in 2016-17 (60%). It is also interesting that the proportion of women and men who have secondary education has increased over time (i.e. 35%, 43%, and 50% for women; 37%, 42%, and 48% for men). The proportion of employed men has also increased in 2012 (70%) and 2016-17 (69%), compared to 2005-06 (65%). In terms of behavioural characteristics, the proportion of those who have been tested for HIV has drastically increased over time (19%, 50%, and 62% for women; 11%, 31%, and 42% or men). We also found that the proportion of women who have adequate HIV knowledge has increased in 2016-17 (85%) compared to 2005-06 (81%) and 2012 (81%). For men, this trend was reversed, indicating that the proportion has decreased in 2012 (86%) and 2016-17 (86%) compared to 2005-06 (89%).
Table 1. Cross-classification analysis of selected variables by survey years among women and men in Haiti.
shows findings from logistic regression analysis for women. In Model 1, we found at the bivariate level that women in 2012 (OR = 0.87, p < 0.05) and 2016-17 (OR = 0.68, p < 0.001) were less likely to endorse HIV misconceptions compared to those in 2005-06. In Model 2, once we controlled for demographic factors, the significance for 2012 became no longer robust (OR = 0.90, p > 0.05), although this relationship remained significant for 2016-17 (OR = 0.71, p < 0.001). The significance of 2016-17 was still robust in Model 3 after accounting for socioeconomic factors (OR = 0.81, p < 0.001); however, this relationship disappeared once we added behavioural factors in Model 4 (OR = 0.93, p > 0.05). In addition to the year of the survey, we further found several control variables to be significantly associated with HIV misconception endorsement (see Model 4). For example, women aged 25-29 (OR = 0.82, p < 0.001), 30-34 (OR = 0.73, p < 0.001), 35-39 (OR = 0.72, p < 0.001), 40-44 (OR = 0.76, p < 0.001), and 45-49 (OR = 0.79, p < 0.001) were less likely to endorse HIV misconceptions compared to those aged 15-19. Also, currently (OR = 1.22, p < 0.001) and formerly (OR = 1.28, p < 0.001) married women were more likely to endorse HIV misconceptions compared to never-married women. In terms of socioeconomic factors, women who belong to the ‘richest’ (OR = 0.53, p < 0.001), ‘richer’ (OR = 0.68, p < 0.001), ‘middle’ (OR = 0.79, p < 0.001), and ‘poorer’ (OR = 0.85, p < 0.01) household wealth category were less likely to endorse HIV misconceptions compared to their ‘poorest’ counterparts. Women with primary (OR = 0.82, p < 0.001), secondary (OR-0.42, p < 0.001), and higher (OR = 0.18, p < 0.001) education were also less likely to endorse HIV misconceptions compared to those without any formal education. For behavioural characteristics, women who tested for HIV (OR = 0.75, p < 0.0010 and had adequate HIV knowledge (OR = 0.70, p < 0.001) were less likely to endorse HIV misconceptions compared to their counterparts who did not.
Table 2. Logit models predicting ‘HIV misconception endorsement’ among women in Haiti.
shows findings from logistic regression analysis for men. In Model 1, we found at the bivariate level that men in 2012 (OR = 1.33, p < 0.01) were more likely to endorse in HIV misconceptions compared to those in 2005-06 although we did not observe any significant difference for 2016-17 (OR = 0.99, p > 0.05). Once we controlled for demographic factors in Model 2, the results were largely consistent with Model 1, indicating that men in 2012 (OR = 1.39, p < 0.001) were more likely to endorse in HIV misconceptions compared to those in 2005-06. We further accounted for socioeconomic factors in Model 3, observing that the non-significant result of 2016-17 became robust (OR = 1.17, p < 0.05) while the difference between 2012 and 2005-06 remained robust (OR = 1.50, p < 0.001). Finally, when we included behavioural factors in Model 4, men in 2012 (OR = 1.55, p < 0.001) and 2016-17 (OR = 1.24, p < 0.01) were more likely to endorse HIV misconceptions compared to those in 2005-06. In addition to the year of the survey, we further found several control variables to be significantly associated with HIV misconception endorsement (see Model 4). For example, men aged 20-24 (OR = 0.86, p < 0.01), 25-29 (OR = 0.82, p < 0.01), 35-39 (OR = 0.75, p < 0.001), 40-44 (OR = 0.70, p < 0.001), 45-49 (OR = 0.80, p < 0.05), 50-54 (OR = 0.69, p < 0.001), and 55-59 (OR = 0.74, p < 0.01) were less likely to endorse HIV misconceptions compared to those aged 15-19. In terms of socioeconomic factors, men who belong to the ‘richest’ (OR = 0.57, p < 0.001), ‘richer’ (OR = 0.65, p < 0.001), and ‘middle’ (OR = 0.81, p < 0.05) household wealth category were less likely to endorse HIV misconceptions compared to their ‘poorest’ counterparts. Men with primary (OR = 0.82, p < 0.01), secondary (OR-0.44, p < 0.001), and higher (OR = 0.26, p < 0.001) education were also less likely to endorse HIV misconceptions compared to those without any formal education. For behavioural characteristics, men who tested for HIV (OR = 0.83, p < 0.001) were less likely to endorse HIV misconceptions compared to their counterparts who did not.
Table 3. Logit models predicting ‘HIV misconception endorsement’ among men in Haiti.
Discussion
Earlier studies in Haiti have suggested that despite the government’s continuous efforts to increase knowledge of HIV transmission among the population, new HIV infections in Haiti may still be driven by HIV misconceptions. Our study, therefore, sought to examine HIV misconceptions over time among females and males in the country. Our findings point to nuanced gender dynamics in the endorsement of HIV misconceptions over time in Haiti, where for females, endorsing HIV misconceptions seems to have declined over time, although this was the reverse for their male counterparts whose endorsement of HIV misconceptions seems to be worsening over time.
The revelation that the endorsement of HIV misconceptions among men over time in Haiti may be getting worse contrasts with the findings from Antabe et al. [Citation18], who found earlier in Malawi that the endorsement of HIV misconceptions over time declined for both females and males. We offer some plausible explanations for our findings in the context of Haiti. First, it is possible that females in Haiti may experience more interactions with healthcare spaces through maternal healthcare utilization programs. As part of the UNAIDS strategies to end mother-to-child HIV transmission and improve the maternal health outcomes of women, provider-initiated HIV testing is recommended where during prenatal sessions, women are screened for and educated on HIV transmission by healthcare workers [Citation19–22]. For instance, while at the general population level, Joseph et al. [Citation23] estimate that only 21.3% of sexually active women in Haiti had ever tested for HIV, this is observed to be relatively higher among women using formal antenatal care (ANC) services. In a study examining the quality and uptake of antenatal and postnatal care in 10 large health facilities in the Nord and Nord-East department and communes of St. Marc, Verrettes, and Petite Rivière in Haiti, Mirkovic et al. [Citation19] found that as much as 96% of women using ANC services in these facilities were tested for HIV. These interactions may have become a useful medium for females in Haiti to become factually informed about HIV transmission, leading to their rejection of HIV misconceptions.
Secondly, women in Haiti being more vulnerable to HIV has resulted in increased policy attention from the Haitian government and NGOs who seek to reduce this vulnerability through various initiatives. For example, the PEPFAR initiative dubbed ‘Determine, Resilient, Empowered, AIDS-free, Mentored and Safe’ (DREAMS) program and other community-based initiatives duplicated across the country are exclusively designed to reduce the HIV vulnerability of adolescent girls and young women by improving their access to HIV information, preventing community-based gender violence, extending sexual and reproductive health education and counselling [Citation2,Citation23–25]. As such opportunities are limited for males, it may be leading to their continuous endorsement of HIV misconceptions. Earlier, Magee et al. [Citation26] recommended increasing the knowledge base of males in Haiti by working to reduce their endorsement of HIV misconception as part of a broader national strategy on HIV prevention.
Although not the focus of our study, we observed that the demographic characteristics of females and males were associated with their endorsement of HIV misconceptions. For instance, compared to the youngest age cohort, all other age groups were less likely to endorse HIV misconceptions. It is possible that members of this age cohort in Haiti do not have frequent interaction with spaces where they can be informed about HIV transmission or that the approaches to HIV information over the years may have been missing them. This finding may justify the position of PEPFAR to target younger people with HIV information and testing services in Haiti [Citation2]. This finding is consistent with Etowa et al. [Citation27], who found in Ontario, Canada, that the youngest age cohort of 15-19 years was more correlated with having an HIV misconception compared to the older cohorts. We also find that currently married or formerly married women were more likely to endorse HIV misconceptions compared to their never-married counterparts, although this relationship was not significant for males. This finding underscores the contextual dynamics of power imbalance within marriage in developing countries, including Haiti, which works to increase married women’s exposure to HIV by reducing their interaction with factual HIV information [Citation28,Citation29]. Indeed, earlier in Haiti, Simon and Colleagues [Citation30] also noted that among women of reproductive age, those who reported their most recent relationship to be with a spouse were less likely to have used a condom during sexual intercourse, which may further underscore the limited opportunity for married women to interact with HIV information sources in the country.
We also found that socioeconomic factors influenced the endorsement of HIV misconceptions. Among females and males, the poorest and those with no formal education were found to be more likely to endorse HIV misconceptions. This is not surprising given the positive influence of higher socioeconomic status on health outcomes [Citation31,Citation32]. Consistent with the literature, having adequate knowledge of HIV transmission and ever testing for HIV were associated with the rejection of HIV endorsement among both females and males, underscoring the need to intensify HIV information dissemination and opportunities for testing in Haiti [Citation10,Citation33,Citation34].
We acknowledge some limitations of our study. First, the data were collected contemporaneously, limiting our study findings to statistical association. Second, our construct of HIV misconceptions may not be exhaustive of all forms of HIV misconceptions. It will be useful for a follow-up qualitative study that captures the depth of HIV misconceptions in Haiti. Finally, our results may be biased due to the self-reported nature of the HDHS and the sensitive nature of HIV, which may lead to participant underreporting. Despite these limitations, however, our study makes an important contribution to HIV policy in Haiti.
Based on our findings, we offer some policy recommendations. First, there is an urgent need for policymakers and stakeholders in Haiti to pay attention to how best to disseminate information on HIV transmission to males. Specifically, the Ministry of Health, PEPFAR and others such as GHESKIO can design gender-specific HIV interventions that target females and males. In this design, it would be useful to understand the context of males' continuous endorsement of HIV misconceptions and targeting the underlying contributory factors. Furthermore, it would be useful to combine HIV intervention strategies that target men with additional economic improvement programs. In line with the call by Magee et al. [Citation26], it is important to think of males as major stakeholders in HIV prevention whose HIV risk needs to be considered as part of the HIV policy priority areas for Haiti. It is equally prudent to continue targeting females with HIV information and increasing their opportunities to test for HIV. Overall, improving the socioeconomic status of females and males should be made an integral part of HIV intervention strategies in Haiti.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Additional information
Funding
References
- UNAIDS. Country factsheets: Haiti 2022. UNAIDS. 2021. Available from: https://aidsinfo.unaids.org/%250D
- PEPFAR. Haiti PEPFAR country operational plan (COP). 2020 strategic direction summary. 2020. Available from: https://www.state.gov/wp-content/uploads/2019/09/Namibia_COP19-Strategic-Directional-Summary_public.pdf
- Delcher C, Robin EG, Pierre DM. Haiti⇔s HIV surveillance system: past, present, and future. Am J Trop Med Hyg. 2020;103(4):1372–1375.
- Koenig S, Ivers LC, Pace S, et al. Successes and challenges of HIV treatment programs in Haiti: Aftermath of the earthquake. HIV Ther. 2010;4(2):145–160.
- Pape JW, Liautaud B, Thomas F, et al. Characteristics of the acquired immunodeficiency syndrome (aids) in Haiti. N Engl J Med. 1983;309(16):945–950.
- Pape J, Johnson WD. AIDS in Haiti: 1982–1992. Clin Infect Dis. 1993;17 Suppl 2(Supplement_2):S341–S345.
- Nierengarten M-B. Haiti’s HIV equity initiative. Lancet Infect Dis. 2003;3(5):266.
- Gourlay A, Birdthistle I, Mburu G, et al. Barriers and facilitating factors to the uptake of antiretroviral drugs for prevention of mother-to-child transmission of HIV in Sub-Saharan africa: a systematic review. J Int AIDS Soc. 2013;16(1):18588.
- Sano Y, Antabe R, Atuoye KN, et al. Persistent misconceptions about HIV transmission among males and females in Malawi. BMC Int Health Hum Rights. 2016;16(1):16.
- Antabe R, Sano Y, Luginaah I. HIV misconceptions among married women in Malawi : the role of household decision-making autonomy. J Public Heal. 2020;1:1–8.
- Letamo G. Misconceptions about HIV transmission among adolescents: levels, trends and correlates from the Botswana AIDS impact surveys, 2001–2013: a short report. AIDS Care. 2019;31(1):48–52.
- Dévieux JG, Deschamps MM, Malow RM, et al. Knowledge, attitudes, and behaviors among a sample of HIV-positive and HIV-negative females visiting an urban VCT center in Haiti. J Health Care Poor Underserved. 2009;20(2):554–568.
- Bogart LM, Skinner D, Weinhardt LS, et al. HIV/AIDS misconceptions may be associated with condom use among black South africans: an exploratory analysis. Afr J AIDS Res. 2011;10(2):181–187.
- Kang’ethe SM, Xabendlini S. Exploring myths associated with HIV/AIDS with examples from a few countries of the world. J Hum Ecol. 2014;46(1):55–61.
- Zhou B, Carrillo-Larco RM, Danaei G, Riley LM, Paciorek CJ, Stevens GA, et al. Worldwide trends in hypertension prevalence and progress in treatment and control from 1990 to 2019: a pooled analysis of 1201 population-representative studies with 104 million participants. Lancet, 10304. 2021;398: 957–980.
- Joshi M, Rahill GJ, Rice C, et al. Back to basics : Correlates of HIV risk in a community sample of haiti back to basics : Correlates of HIV in a community sample of urban haiti. medRxiv. 2020; 1–34.
- Perrin G. A Comparative Analysis of the Attitudes towards People Living with HIV/AIDS between Haiti and the Dominican Republic. Georgia State University. Georgia State University. 2010. Available from: http://digitalarchive.gsu.edu/iph_theses/87/
- Antabe R, Sano Y, Anfaara FW, et al. Reducing HIV misconceptions among females and males in Malawi: are we making progress. AIDS Care. 2020;33(3):408–412.
- Mirkovic KR, Lathrop E, Hulland EN, et al. Quality and uptake of antenatal and postnatal care in Haiti. BMC Pregnancy Childbirth. 2017;17(1):52.
- Deschamps MM, Noel F, Bonhomme J, et al. Prevention of mother-to-child transmission of HIV in Haiti. Rev Panam Salud Publica. 2009;25(1):24–30.
- Bassett IV, Walensky RP. Integrating HIV screening into routine health care in resource‐limited settings. Clin Infect Dis. 2010;50 Suppl 3(Suppl 3):S77–S84.
- Dhillon S, Amoak D, Sano Y, et al. The association between mother-to-child HIV transmission knowledge and antenatal care utilisation in Cameroon. Int J Health Plann Manage. 2023;38(6):1877–1888.
- Joseph F, Simon DJ, Claude V, et al. Trends and factors associated with recent HIV testing among women in Haiti : a cross ‑ sectional study using data from nationally representative surveys. BMC Infect Dis. 2024;24(1):74.
- Seo G, Joseph JMB, Confident N, et al. The FANMI (“my FAMILY” in creole) study to evaluate community-based cohort care for adolescent and young women living with HIV in Haiti: Protocol for a randomized controlled trial. BMC Public Health. 2019;19(1):1749.
- Logie CH, Daniel CA. ‘My body is mine’: Qualitatively exploring agency among internally displaced women participants in a small-group intervention in leogane, Haiti. Glob Public Health. 2016;11(1-2):122–134.
- Magee EM, Small M, Frederic R, et al. Determinants of HIV/AIDS risk behaviors in expectant fathers in Haiti. J Urban Health. 2006;83(4):625–636.
- Etowa EB, Wong JP, Omorodion F, et al. Addressing HIV misconceptions among heterosexual black men and communities in Ontario. Healthcare. 2023;11(7):997. Available from: https://www.mdpi.com/2227-9032/11/7/997
- Letamo G. Misconceptions about HIV prevention and transmission in Botswana. Afr J AIDS Res. 2007;6(2):193–198.
- Gupta J, Small MJ, Kershaw T. Gender and HIV/AIDS in Haiti: Women’s lack of power as an overarching vulnerability. Gender HIV/AIDS. London and New York: Routledge; 2016. pp. 85–100.
- Jean Simon D, Paul B, Kiragu A, et al. Prevalence and factors associated with condom use among sexually active young women in Haiti: evidence from the 2016/17 Haiti demographic and health survey. BMC Womens Health. 2023;23(1):146.
- Kansanga MM, Braimah JA, Antabe R, et al. Examining the association between exposure to mass media and health insurance enrolment in Ghana. Int J Health Plann Manage. 2018;33(2):e531–e540.
- Anfaara FW, Antabe R, Atuoye KN, et al. Hepatitis B knowledge among women and men in the Upper West region of Ghana : What sources of health information matter ? Int J Health Plann Manage. 2020;35(5):1098–1110.
- Tenkorang EY. Myths and misconceptions about HIV transmission in Ghana: What are the drivers? Cult Health Sex. 2013;15(3):296–310.
- Tenkorang EY, Owusu AY. Examining HIV-related stigma and discrimination in Ghana: what are the major contributors? Sex Health. 2013;10(3):253–262.