Factors associated with self-reported HIV testing in the Dominican Republic

Abstract

Human immunodeficiency virus (HIV) testing is important for controlling the epidemic in low- and middle-income countries such as the Dominican Republic (DR) – a country in the Caribbean. This study aimed to examine factors associated with HIV testing in the DR. The study used HIV test data in the 2013 DR Demographic and Health Survey. HIV data were collected from 18,614 individuals aged 15–59 years. Wealth status, HIV-related stigma, and knowledge scales were constructed using factor analysis. Survey-weighted logistic regression was used to identify correlates of HIV testing in the DR. Results show that almost two-thirds (62.4%) of the participants reported ever having an HIV test. In the multivariable analysis, older age, higher education, female gender, richest wealth status, health insurance, married, higher HIV-knowledge, and lower HIV-related stigma increased the likelihood of getting an HIV test in the DR. People living in Norte/Cibao, which had a higher HIV prevalence, had a lower chance of having an HIV test than people living in the Este/Sureste region. Although HIV testing has increased in recent times, it is not yet satisfactory in the DR. Specific interventions building HIV awareness targeting specific sub-populations and regions will increase knowledge on HIV and reduce HIV-related stigma, and may increase HIV testing.

Introduction

The Caribbean has the highest human immunodeficiency virus (HIV) seroprevalence rate (1.2%) in the western hemisphere. Despite its small population size, the Caribbean has the second highest prevalence of HIV/AIDS in the world after Africa.¹ Almost three-quarters of the people living with HIV (PLHIV) are living in Hispaniola, which is an island containing two countries, Haiti and the Dominican Republic (DR).² Haiti has the highest HIV prevalence in the western hemisphere and the DR, being the bordering country of Haiti, also suffers from the HIV epidemic.²

New HIV incidence declined in the DR from 0.33 per 1000 population in 2010 to 0.26 in 2018. The DR also experienced a decline in the number of PLHIV from 72,000 in 2010 to 70,000 in 2018.³ However, the DR is lagging behind the 90–90–90 goal proposed by UNAIDS, which calls for 90% of PLHIV to know their status, 90% of PLHIV who know their status to be on treatment, and 90% of PLHIV on treatment to be virally suppressed.⁴ In 2018, 82% of PLHIV knew their status, of which 56% were on treatment, and of which only 37% were virally suppressed.³ Although over two-fifths (46.4%) of young women in the DR were knowledgeable about HIV prevention, almost half (49.3%) of the adult DR population reported discriminatory attitudes toward PLHIV.³

HIV testing is very important for early detection and treatment of the infection. Therefore, it plays a critical role in the HIV care continuum and can lead to decreasing the chance of secondary transmission. HIV testing uptake has increased all over the world, however, not uniformly among different demographic groups. Women are more likely to get tested in almost every country, mainly due to the increase in uptake of HIV test during antenatal care.^5,6 The fear of stigma associated with HIV is also a deterrent factor for HIV testing.^7–9 In a recent meta-analysis, the authors showed that people who had high levels of HIV-related stigma were two times more likely to present late for HIV care.¹⁰ Prior research suggests that knowledge on HIV increases uptake of HIV testing.^11,12

This study aims to describe the patterns of HIV testing across different sociodemographic characteristics and identify factors associated with HIV screening among populations aged 15–59 years in the DR. The study findings will add to the knowledge gap on HIV testing and associated factors in the DR and highlight populations that should be targeted for implementing HIV testing policies and interventions.

Materials and methods

Data source and study design

The study used the 2013 DR Demographic and Health Survey (DHS) HIV testing data.¹³ The DHS is a nationally representative survey conducted periodically since 1986. The DHS employs standardized, well-tested procedures to maintain data quality, including comprehensive interviewer training, field supervision, and data processing procedures. The survey follows a bi-phasic stratified cluster sampling method. In the first phase, 524 primary sampling units (clusters), stratified by province and area of residence (364 in urban areas and 160 in rural areas), were selected with probability proportional to size sampling strategy. In the second phase, 25 households were selected randomly from each cluster and from each household, all women aged 15–49 and all men aged 15–59 years, who lived in those homes regularly and those who slept in them the night before, were interviewed after taking their consent. All participants were asked questions regarding their HIV-related knowledge and behavior, among other questions. In this study, we analyzed the data collected from 18,614 individuals who responded to the HIV-related questionnaire in the 2013 DRDHS.

Measures

The outcome variable for this study is a dichotomous response (Yes/No) item ‘ever had HIV test.’ Demographic variables included age (15–19, 20–29, 30–39, 40–49, 50–59 years), gender (male, female), marital status (never in union, married, free union, other including widowed, separated, and divorced); and socioeconomic variables, i.e. education (no education, primary, secondary, higher), religion (Catholic, Evangelic, Adventist, none), residence (rural, urban), insurance status (no, yes), wealth index (poorest, poorer, middle, richer, richest), and region (Este/Sureste, Sur, Norte/Cibao).

Wealth index

DHS uses principal component analysis to estimate the wealth index for each household in the survey. The wealth index shows the household’s cumulative living standard and is calculated using data on a household’s ownership of selected assets including materials used for housing construction, types of water access, and sanitation facilities.¹⁴

HIV knowledge

Knowledge of HIV transmission and prevention was assessed using nine items included in the standard DHS AIDS Indicator Survey (AIS).¹⁵ Survey items explored knowledge about 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, 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. Correct responses were coded as 1 and incorrect or uncertain responses were coded as 0. Items were summed to create an HIV knowledge score (mean = 6.76, SD = 1.82, range = 0–9) with higher scores indicating more knowledge about HIV transmission and prevention. Cronbach’s alpha was 0.48 for HIV knowledge scale.

HIV-related stigma

HIV-related stigma was assessed using five items included in the standard DHS AIS. These items included garnering responses from participants on whether he/she would buy fresh vegetables from a shopkeeper who has the HIV virus, would allow a female teacher with the virus but not sick to continue teaching, would not want to keep a secret that a family member got infected with the virus, would be willing to care for a family member with the virus in the respondent’s home, and whether children should be taught about condoms to avoid HIV infection. Positive responses to these questions were coded as 0. Negative responses were coded as 1, and those who were not sure/did not know were coded as 0.5. Items were summarized to create the HIV stigma score (mean = 1.41, SD = 1.06, range = 0–5) with higher scores indicating greater HIV-related stigma. Cronbach’s alpha was 0.38 for HIV-related stigma scale.

Data analysis

Survey-weighted descriptive statistics, and bivariate and multivariable analyses using logistic regression models were performed to examine the association between HIV testing in the DR and sociodemographic factors. The sociodemographic variables included in the multivariable model were age, gender, education, religion, marital status, place of residence, health insurance, wealth index, and regions of the DR based on the statistical significance observed in the bivariate analysis. All analyses were performed using Stata 14.2.¹⁶

Results

Table 1 shows the demographic characteristics of participants in the final study sample (N = 18,614). Most of the sample were 20–29 years old (31.1%) and male (52.9%). Approximately 45% reported being Catholic and 21% were in the middle wealth index. Most participants were in free union (38.3%), resided in urban areas (73.9%), and had health insurance (56.6%). Almost half (49.9%) of the participants were from the Este/Sureste region of the country.

Table 1.

Demographic characteristics of participants in 2013 Dominican Republic demographic health survey HIV section (N = 18,614).

Variables	Frequency	Percent
Age group (years)
15–19	3518	18.9
20–29	5784	31.1
30–39	4450	23.9
40–49	3663	19.7
50–59	1199	6.4
Gender
Male	9851	52.9
Female	8763	47.1
Education
No education	562	3.0
Primary	6779	36.4
Secondary	7270	39.1
Higher	4003	21.5
Religion
Catholic	8426	45.3
Evangelic	3684	19.8
Adventist	572	3.1
None	5911	31.8
Marital status
Never in union	5556	29.9
Married	2446	13.1
Free union	7126	38.3
Other (widowed/separated/ divorced)	3487	18.7
Residence
Rural	4848	26.1
Urban	13,766	73.9
Health insurance
No	8076	43.4
Yes	10,527	56.6
Wealth index
Poorest	3660	19.7
Poorer	3810	20.5
Middle	3867	20.8
Richer	3756	20.2
Richest	3521	18.9
Region (DR)
Este/Sureste	9291	49.9
Sur	3096	16.6
Norte/Cibao	6226	33.4

DR: Dominican Republic.

Table 2 shows the distribution of HIV testing across demographic and socioeconomic characteristics in the DR. Almost two-thirds (62.4%) of the participants reported that they ever had an HIV test. HIV testing was more frequent among individuals who were middle aged (30–39 years, 81.6%) than other age groups (p < 0.001), and of Catholic belief than other religious groups and those who reported not having any religious beliefs (66.5%, p < 0.001). Individuals who had health insurance (67.1%, p < 0.001) and were from the Este/Sureste region (64.9%, p < 0.0001) reported more HIV testing than those who had no insurance and lived in the Sur and Norte/Cibao regions, respectively. Testing rates were higher among female than male participants (75.6% versus 50.7%, p < 0.001) and urban residents than rural residents (64.0% versus 57.9% %, p < 0.001). HIV testing rates also increased with increasing wealth level (p < 0.001).

Table 2.

Survey weighted bivariate analysis: ever had HIV tests done by demographic and socioeconomic characteristics in the Dominican Republic (2013) (N = 18,614).

Variables	Ever had HIV test
	No (%)	Yes (%)	p-value
All	37.6	62.4
Age group (years)			<0.001
15–19	81.8	18.2
20–29	37.0	63.0
30–39	18.4	81.6
40–49	20.5	79.5
50–59	34.1	65.9
Gender			<0.001
Male	49.3	50.7
Female	24.4	75.6
Education			<0.001
No education	50.9	49.1
Primary	37.2	62.8
Secondary	43.1	56.9
Higher	26.3	73.7
Religion			<0.001
Catholic	33.5	66.5
Evangelic	37.9	62.1
Adventist	35.2	64.8
None	43.4	56.6
Residence			<0.001
Rural	42.1	57.9
Urban	36.0	64.0
Marital status			<0.001
Never in union	73.8	26.2
Married	17.4	82.6
Free union	24.0	76.0
Other (widowed/separated/divorced)	21.7	78.3
Health insurance			<0.001
No	43.7	56.3
Yes	32.9	67.1
Wealth index			<0.001
Poorest	44.4	55.6
Poorer	38.9	61.1
Middle	36.9	63.1
Richer	34.1	65.9
Richest	33.6	66.4
Region (DR)			<0.001
Este/Sureste	35.1	64.9
Sur	43.7	56.3
Norte/Cibao	38.2	61.8
HIV-related stigma			<0.001
Low	33.4	66.6
Medium	40.8	59.2
High	46.2	53.8
Knowledge on HIV			<0.001
Low	46.9	53.1
Medium	34.2	65.8
High	25.9	74.1

DR: Dominican Republic.

Multivariable logistic regression shows that in the DR, predisposing factors, such as age, gender, education, marital status, health insurance, wealth index, region, HIV-related stigma, and knowledge were associated with HIV testing in the multivariable analysis (Table 3). Participants aged 15–19 years were 90% less likely to get HIV testing (odds ratio [OR]: 0.10, 95% confidence interval [CI: 0.08, 0.12]) than those who were 30–39 years old. Females were 3.5 times more likely to get an HIV test (OR: 3.50, 95% CI: 3.03, 4.05) than males. Those who had higher education were almost three times more likely to ever be tested for HIV (OR: 2.72, 95% CI: 2.02, 3.67) than those who had no education. People who were married, in free union, or of other marital status (widowed/separated/divorced) were approximately five times as likely to have an HIV test (OR: 4.74, 95% CI: 3.86, 5.81; OR: 4.89, 95% CI: 4.10, 5.83; OR: 5.33, 95% CI: 4.53, 6.26, respectively) as those who were never married. Participants in the richest quintile had 45% higher odds of ever having an HIV test (OR: 1.45, 95% CI: 1.19, 1.76) than those who were in poorest quintile. Participants with a high level of HIV-related stigma were 19% less likely to get an HIV test (OR: 0.81, 95% CI: 0.70, 0.93) than those who had a low level of stigma. Medium and high levels of HIV-related knowledge were associated with higher odds of having an HIV test (OR: 1.36, 95% CI: 1.22, 1.52; OR: 1.64, 95% CI: 1.39, 1.95, respectively) than having a low level of HIV-related knowledge. HIV test was less likely among participants who resided in the Sur (OR: 0.73; 95% CI: 0.64, 0.83) and Norte/Cibao (OR: 0.86, 95% CI: 0.76, 0.97) regions than those who resided in the Este/Sureste region. There was no statistically significant association between religion and HIV testing.

Table 3.

Survey weighted multiple logistic regression of ever had HIV test in Dominican Republic, 2013.

Variables	Ever had HIV test (N = 18,590)
	Odds ratio	95% CI		p-Value
		Lower limit	Upper limit
Age group (years)
30–39	1.00
15–19	0.10	0.08	0.12	<0.001
20–29	0.51	0.42	0.61	<0.001
40–49	0.77	0.66	0.90	0.001
50–59	0.66	0.54	0.80	0.030
Gender
Male	1.00
Female	3.50	3.03	4.05	<0.001
Education
No education	1.00
Primary	1.85	1.46	2.35	<0.001
Secondary	2.14	1.67	2.75	<0.001
Higher	2.72	2.02	3.67	<0.001
Religion
Catholic	1.00
Evangelic	1.09	0.97	1.23	0.131
Adventist	0.98	0.83	1.16	0.795
None	1.31	0.86	2.00	0.202
Residence
Rural	1.00
Urban	1.08	0.95	1.23	0.241
Marital status
Never in union	1.00
Married	4.74	3.86	5.81	<0.001
Free union	4.89	4.10	5.83	<0.001
Other (widowed/separated/divorced)	5.33	4.53	6.26	<0.001
Health insurance
No	1.00
Yes	1.33	1.20	1.48	<0.001
Wealth index
Poorest	1.00
Poorer	1.03	0.87	1.22	0.705
Middle	1.15	0.95	1.39	0.143
Richer	1.18	0.99	1.42	0.069
Richest	1.45	1.19	1.76	<0.001
Region
Este/Sureste	1.00
Sur	0.73	0.64	0.83	<0.001
Norte/Cibao	0.86	0.76	0.97	0.017
HIV-related stigma
Low	1.00
Medium	0.92	0.80	1.05	0.220
High	0.81	0.70	0.93	0.003
Knowledge on HIV
Low	1.00
Medium	1.36	1.22	1.52	<0.001
High	1.64	1.39	1.95	<0.001

CI: confidence interval.

Discussion

Although it is well-accepted that HIV testing is the first step for getting diagnosed and receiving HIV care, very little research has been done on HIV testing in the DR. Almost one-third of the study participants reported not having HIV test done in their lifetime, which indicates a need for understanding the contextual factors that may contribute to improving the rates of HIV testing in the DR.

Results suggest that almost two-thirds of the participants had an HIV test in the DR with the prevalence of HIV testing increasing from 20% in 2007 to 62.4% in 2013.¹⁷ The HIV testing rate in the DR is higher than neighboring Haiti, which had an HIV testing prevalence of 39.6% among study participants of the DHS in 2012.¹⁸

The results indicated that men and younger populations had lower rates of ever getting an HIV test. This finding is similar to findings in other countries, such as South Africa,¹⁹ USA,²⁰ Zimbabwe,²¹ and 28 other countries of sub-Saharan Africa.⁵ In a qualitative study conducted in eastern African countries Kenya and Uganda, the researchers showed that men were more mobile for their jobs and, therefore, had more challenges in accessing resources for HIV testing. This study also found that the men believed that their wives’ HIV test results could serve as a proxy for their HIV status.²² Women’s HIV testing rates had increased due to the inclusion of HIV testing for pregnant women in both countries. In addition, while vertical transmission of HIV from mother to child may occur at the time of pregnancy and delivery, this time also provided an opportunity for counseling the mothers and providing opt-out options in HIV testing, which helped to increase the testing rate as evident among Zimbabwean children.²³ However, challenges remained in getting younger populations access to healthcare and HIV testing, and as a result, they were often left untested and untreated.

Late diagnosis continues to be a major problem in Caribbean countries, especially among men and key populations, such as who are men who have sex with men (MSM), transgender individuals, sex workers, and people who use drugs (PWUD), where almost one-third of the PLHIV are diagnosed with a CD4 cell count of >200 cells/mm³.¹ Our study also shows that men had a lower chance of getting tested for HIV and at higher risk of getting diagnosed at advanced stage of disease.

HIV-related stigma plays a crucial role in HIV prevention and treatment.²⁴ Research has shown a significant negative association between HIV-related stigma and use of voluntary counseling and testing.^8,25 In the current study, higher levels of stigma were associated with lower odds of getting tested as expected. Research has shown that stigma related to HIV deterred persons from getting HIV testing for fear of having their HIV-positive status disclosed or of suffering further stigma and discrimination based on their HIV status.²⁶ In a research study conducted in the USA, Fortenberry et al.²⁷ opined that stigma was more powerful than shame in lowering the voluntary uptake of HIV testing. Beyond HIV testing, HIV-related stigma also affects poor access to care, not having regular source of care, and poor ART adherence.²⁸ Moreover, PLHIV who are MSM, transgender individuals, sex workers, and PWUD, often referred to as ‘key populations,’ are experiencing intersectional stigma because of their identities as well as living with HIV.²⁹ Therefore, key populations, which constitute the majority of the DR PLHIV, are at a higher risk of not getting HIV testing and also have limited access to care.

In the current study, possessing HIV-related knowledge also affected HIV testing uptake. A study conducted in South Africa also found that increased knowledge on HIV transmission was associated with HIV testing.³⁰ In another study conducted among undergraduate students of a privately owned university in Nigeria, older students with higher knowledge were more willing to take an HIV test.¹¹ HIV-related knowledge coupled with the synergistic effect of education has a very important policy implication. As the current findings show that education has a positive effect on HIV testing in DR and disseminating proper knowledge on HIV transmission and prevention would be of paramount importance for reaching the first 90 of the UNAIDS’ 90–90–90 goal, the design and implementation of mass media campaigns may help to increase knowledge and reduce HIV-related stigma and, thereby, increase the prevalence of HIV testing.

Participants in the highest wealth index category were more likely to have ever being tested for HIV. This finding is supported by other studies conducted in other African countries.^31,32 Research conducted in Uganda and Zimbabwe found that socioeconomic status was associated with seeking HIV testing.^31,33 This relationship may be due to the increased awareness building effort among urban people undertaken by the DR in recent times, which has been partially funded by international donor agencies.³⁴

Having health insurance and being married were positively associated with HIV testing. A study conducted in Mozambique also found a significant association between health insurance and HIV testing.³⁵ Health insurance policies should include HIV testing as a benefit for each beneficiary to ensure HIV testing for all. The association between being married and HIV testing may be explained by the availability of HIV tests as part of the reproductive healthcare services, which may have a higher percentage of users among those in marriage or cohabiting relationships.³⁶ However, no statistically significant effect was found between religion and HIV testing, which was supported by a study conducted in Tanzania.³⁷

Regional variation in HIV testing was also evident. In the DR, residents of Sur and Norte/Cibao were less likely to have an HIV test compared to residents of the Este/Sureste. Norte/Cibao is the bordering region of Haiti and HIV prevalence in this region is higher. Paradoxically, HIV testing probabilities are low in this region. Regional variation and demographic distribution of HIV is crucial to understand the epidemiology of HIV/AIDS in the DR. Poor Haitian migrants living in the Norte/Cibao region have higher rates of HIV compared to other sub-populations or residents of other regions.² Therefore, special attention is required for this group of people living in the endemic areas.

Limitations

Although this study used a nationally representative sample of the DR population, the analysis has known limitations. First, since DHS data is cross-sectional in nature, we can only examine the association and cannot determine causality between different factors and HIV testing. Second, we used the latest available data sets for the DR, which are dated back to 2013. Therefore, analysis of the upcoming DHS survey data sets would provide the most current trends. Third, the lower Cronbach’s alpha level for both scales may be due to the HIV knowledge and HIV-related stigma scales were created using available variables in the DHS data set and differ from validated scales.^38,39 More comprehensive stigma measurement tools like PLHIV Stigma Index may be used in later rounds of the DHS surveys for better understanding the level of HIV-related and intersecting stigma.²⁹ Fourth, we were unable to determine the distribution of key populations and groups that tend to be disproportionately affected by HIV (e.g. female sex workers, MSM, or transgender populations) due to these questions not being asked in the survey. Future surveys should include questions garnering information on gender/sexual preferences to obtain data on these underrepresented and vulnerable populations.

Conclusions

While HIV-related stigma has a negative effect on HIV testing, knowledge on HIV transmission has a positive association with HIV testing. Women were more likely to get tested compared to men because of inclusion of HIV testing as a standard practice during antenatal care. There were also regional variations in HIV testing in the DR. To achieve the first 90 of 90–90–90 goal (i.e. 90% of PLHIV are diagnosed), policy makers should consider these findings, and implement and target HIV testing programs to different segments of the DR population, including younger and older populations, males, populations with lower educational attainment, those who were never married, residents of Norte/Cibao and Sur regions, those with lower HIV knowledge and high levels of HIV-related stigma. Further research is warranted to include key populations to find out the extent of HIV-related and intersecting stigma and how that affects HIV testing and the HIV care continuum.