Individual‐, household‐, and community‐level factors associated with pregnant married women's discriminatory attitude towards people living with HIV in sub‐Saharan Africa: A multicountry cross‐sectional study

Abstract Background and Aims Discriminatory attitude towards people living with human immunodeficiency virus (HIV) remains a major problem in the prevention and treatment of HIV in sub‐Sahara Africa (SSA). Understanding the multiple factors linked to discriminatory attitude towards people living with HIV/AIDS (PLWHA) in SSA is necessary for developing appropriate interventions. This study aimed at investigating the individual, household, and community‐level factors associated with pregnant married women's discriminatory attitude towards people living with HIV/AIDS. Methods We used data from the Demographic and Health Surveys of 12 sub‐Saharan African countries conducted between 2015 and 2019. Data on 17 065 pregnant married women were analyzed. Bivariate (chi‐squared test) and multivariable multilevel logistic regression analyses were applied to investigate the factors associated with discriminatory attitude towards PLWHA. The results were reported as adjusted odds ratio (aOR) at 95% confidence interval (CI). Results The mean age of participants was 31.2 ± 8.5. The prevalence of discriminatory attitude towards PLWHA was 36.2% (95% CI: 33.4%‐39.1%). Individual/household‐level factors associated with discriminatory attitude towards PLWHA were women's educational level (secondary school‐aOR = 0.49, 95% CI: 0.26‐0.93), husband's educational level (higher education‐aOR = 0.35, 95% CI: 0.16‐0.76), decision‐making power (yes‐aOR = 0.51, 95% CI: 0.38‐0.69), wife‐beating attitude (disagreement with wife beating‐aOR = 0.58, 95% CI: 0.43‐0.79), and religion (Muslim‐aOR = 1.92, 95% CI: 1.22‐3.04). Community socioeconomic status (medium‐aOR = 0.61, 95% CI: 0.41‐0.93) was the only community‐level factor associated with discriminatory attitude towards PLWHA. Conclusion More than one‐third of pregnant married women in SSA had discriminatory attitude towards PLWHA. Women's educational level, husband's educational level, decision‐making power, wife‐beating attitude, religion, and community socio‐economic status were associated with discriminatory attitude towards PLWHA. To lessen the prevalence of discriminatory attitude towards PLWHA, considering these significant factors is needed. Therefore, governments and other stakeholders in the respective countries need to increase education coverage. Moreover, empowering women through education and economy is crucial. Finally, working with religious leaders to increase awareness about HIV and discriminatory attitude towards PLWHA should also be a priority in SSA.

Conclusion: More than one-third of pregnant married women in SSA had discriminatory attitude towards PLWHA. Women's educational level, husband's educational level, decision-making power, wife-beating attitude, religion, and community socioeconomic status were associated with discriminatory attitude towards PLWHA. To lessen the prevalence of discriminatory attitude towards PLWHA, considering these significant factors is needed. Therefore, governments and other stakeholders in the respective countries need to increase education coverage. Moreover, empowering women through education and economy is crucial. Finally, working with religious leaders to increase awareness about HIV and discriminatory attitude towards PLWHA should also be a priority in SSA. globally, [1][2][3][4][5] particularly in sub-Sahara Africa (SSA). [1][2][3] Nearly 76 million people have been infected, and millions of people have died worldwide since the beginning of the epidemic. 1 It is estimated that about 37.9 million people are living with HIV/AIDS (PLWHA), and still about 1.7 million new infections were reported globally as of 2019. [1][2][3][4] HIV is the leading cause of death worldwide (690 000 people died in 2019) 1,5 and the leading cause of death globally among women of reproductive age (17.3 per 100 000 individuals in 2017). 6 Globally, approximately 5000 new HIV infections per day are reported, 4 and about 61% of these new infections occur in SSA. 4 Although significant progress has been made in recent decades to prevent and control HIV/AIDS, 1 many people, including women and children, still do not have access to treatment and care. 2,3 There is some evidence that HIV/AIDS-related discrimination threatens the effectiveness of prevention and care programmes 7 which may have a negative impact on victims 8 who may already be marginalized or stigmatized. 9 HIV-related discrimination not only may be directly or indirectly related to a person's perceived or actual HIV status, 10 but also includes acts or omissions aimed at other key populations and groups at intensified risk of HIV. 10 Discriminatory attitude is usually attributed to conformist cultural beliefs and practices, 11 which reflect inadequate knowledge and negative attitude towards PLWHA. 12 Features of stigma and discrimination towards PLWHA may vary from labeling and discriminatory behavior to negative treatment by family members, friends, healthcare professionals, and communities, 13,14 which may negatively affect their quality of life. [15][16][17] Prior studies have shown that fear of stigma and discrimination from families, communities, and health workers are some of the barriers towards the acceptance of HIV testing by pregnant women during antenatal care 18,19 and prevention of mother to child transmission (PMTCT) care. 20,21 Although there exist anti-retroviral treatments for PLWHA, 22,23 PLWHA are still faced with discrimination and isolation from their families, colleagues, and communities, 22,23 resulting in job losses and inadequate access and utilization of healthcare services. 23,24 Since HIV/AIDS-related stigma and discrimination occur at different levels, including among individuals, between family members, communities, and organizations, 10,16,25,26 systematic investigations of risk factors at multiple levels is key to designing appropriate policy interventions. 10,16,25,26 Previous studies in SSA showed that the magnitude of discriminatory attitude towards PLWHA varied from 50% in Nigeria 23 to about 64.5% in Ethiopia. 27 There are few studies in Botswana, 28 Zimbabwe, 29 Ethiopia, 27 Nigeria, 23 and three East African countries. 30 However, these studies do not reflect recent determinants, 28 and some are not nationally representative. 29 This study, therefore, aimed at examining the prevalence of pregnant married women's discriminatory attitude towards PLWHA and its associated factors in SSA using large nationally representative samples.

| Data source
We used data from the Demographic and Health Surveys (DHSs) of 12 sub-Saharan African countries conducted between 2015 and 2019. The DHS is a nationally representative survey that collects data from women of reproductive age (15-49 years) on several demographic and health indicators, including discriminatory attitude towards PLWHA. 31 Financial and technical supports for the surveys are usually from the United States Agency for International Development (USAID) and Inner-City Fund (ICF) international. 32 The DHSs in the selected countries usually adopt a two-stage stratified sampling procedure. 33 In the first stage, Enumeration Areas (EAs) are selected using Probability Proportional to Size (PPS). In the second stage, fixed numbers of households are selected from selected EAs using a systematic sampling technique. 33 In this study, the countries were selected if the survey was conducted between 2015 and study. A total of 17 065 currently pregnant married women from 12 countries were included in the final analysis. The individual recode (IR) files were used, and the datasets are available freely at https:// dhsprogram.com/data/available-datasets.cfm. We also followed the guidelines for Strengthening of Observational studies in Epidemiology (STROBE). 34 Table 1 provides detailed information about selected countries, year of survey, and samples.

| Outcome variable
The outcome variable was discriminatory attitude towards PLWHA.
Two questions were asked to assess discriminatory attitudes towards PLWHA among women who have heard of HIV or AIDS. In the survey, the questions asked were "Should children living with HIV be able to attend school with children who do not have HIV?" and "Would you buy fresh vegetables from a shopkeeper or vendor if you knew that this person had HIV?" Response in the affirmative to either question was considered discriminatory attitudes 35 and coded "1," and those who responded "No" were categorized as otherwise and coded "0."

| Individual-/household-level factors
The individual/household level factors were women's age in years,  women's educational status (no formal education, primary school, secondary school, higher), husband's educational status (no formal education, primary school, secondary school, higher), women's occupation (not working, professional/technical/managerial, agricultural, manual, others), economic status (poorest, poorer, middle, richer, richest), media exposure (no, yes), family size (<5, 5+), sex of household head (male, female), religion (Christian, Muslim, Others), antenatal care (ANC) follow-up (no, yes) decision-making capacity, and wife-beating attitude. Regarding decision-making, we used at least one of the three decision-making parameters: own health care, large household purchases, and visits to family or relatives, either alone or together with her husband. Responses were coded as "1" if decisions were either made alone or together with the husband on all three of the aforementioned decision-making parameters, otherwise coded as "0." Wife-beating attitude was coded as "0"-agreed with wife beating if women justified or accepted wife-beating norm on at least one of the five wife-beating parameters (ie, burning food, arguing with him, going out without telling him, neglecting the children and refusing to have sexual intercourse with him). Those who disagreed/not justified with all five parameters were coded as "1"disagreed with wife beating.

| Community-level factors
Place of residence was coded as urban vs rural. Community literacy level (low, medium, high) and community socioeconomic status (low, moderate, high) were calculated as below. The socioeconomic status was computed from occupation, wealth, and education of respondents. We applied principal component analysis to calculate women who were unemployed, uneducated, and poor. A standardized score was derived with a mean score (0) and SD. 1 The scores were then segregated into tertile 1 (least disadvantaged), tertile 2, and tertile 3 (most disadvantaged) where the least score (tertile 1) denoted greater socioeconomic status with the highest score (tertile 3) denoting lower socioeconomic status. Community literacy level was derived from women who could read and write or not read and write at all.

| Statistical analyses
The analysis was carried out as follows. First, descriptive analysis including frequencies and percentages of discriminatory attitude towards PLWHA was calculated and presented using frequency tables and bar charts for the pooled data, and for each country. Thereafter, the Pearson chi-squared test was conducted to select the candidate explanatory variables using a P-value less than .05 as a cut-off point.
Next, a multicollinearity test was conducted using the variance infla- The multilevel logistic regression model included both fixed and random effects. 36,37 The fixed effects (measures of association) show the association between the explanatory variables and discriminatory attitude towards PLWHA, whereas the random effects (measures of variations) were assessed using intra-cluster correlation (ICC). 38 The likelihood ratio (LR) test was used to check model adequacy, whereas Akaike's information criterion (AIC) was used to measure how best the different models fitted the data. The "svyset" command was used to account for the complex survey design, including weight, cluster, and strata. The analyses were performed using Stata version-14 software (Stata Corp, College Station, Texas, USA).

| Distribution of discriminatory attitude across explanatory/control variables
The prevalence of discriminatory attitude towards PLWHA by explanatory variables and subgroups is shown in Table 2. The prevalence varied across the explanatory variables. For instance, discriminatory attitude was found to be higher among pregnant married women with no formal education (65.9%) than those with higher education (13.6%). We further observed a higher prevalence (52.5%) among those with no decision-making power than those with decision-making power (37.0%). A higher prevalence of discriminatory attitude was also found among those who accepted or justified wife beating (54.2%).

| Prevalence of discriminatory attitude
Overall, more than one-third (36.2%, 95% CI: 33.4%-39.1%) of respondents in the selected countries had discriminatory attitude towards PLWHA. We however found cross-country differences in the prevalence of discriminatory attitude towards PLWHA. Figure 1 shows the prevalence of discriminatory attitude towards PLWHA among married pregnant women across the studied countries.
Sierra Leone reported the highest prevalence (83.6%) followed by

| Random effect results (measures of variation)
The random effects models of the individual-/household-and community-level factors associated with married pregnant women's discriminatory attitude towards PLWHA are shown in

| DISCUSSION
We investigated the prevalence of discriminatory attitude towards PLWHA among pregnant married women in 12 countries in SSA and examined associated individual-, household-, and community-level factors. The study shows that 36.2% (95% CI: 33.4%-39.1%) of pregnant married women had discriminatory attitude towards PLWHA.
Both individual-/household-and community-level factors were found to be linked with discriminatory attitude.
Regarding the individual/household factors, we found that pregnant married women who had higher educational levels were less likely to have discriminatory attitude compared to those with no formal education. 11,19,21 Prior studies have indicated that educated women have better knowledge about HIV transmission and F I G U R E 1 Prevalence of discriminatory attitude towards PLWHA among married pregnant women across studied countries: Evidence from 12 SSA countries DHSs T A B L E 3 Multilevel multivariable logistic regression results of the individual-/household-and community-level factors associated with discriminatory attitude towards PLWHA among married pregnant women (N = 17 065): Evidence from 12 sub-Saharan African countries DHSs Abbreviations: AIC, Akaike information criterion; ICC, intra-class correlation coefficient; LR, likelihood ratio; N, total observation. PLWHA. 27 Similarly, lower odds of discriminatory attitude were observed among educated husbands as found in previous studies in Kuwait 46 and Bangladesh. 47 The findings revealed a strong association between religion and discriminatory attitude towards PLWHA. 48  Consistent with prior studies in Zimbabwe, 29 Ghana, 55 and other two African counties 56 and three East African countries, 30

| Strengths and limitations of the study
This study has some strengths and limitations. One of the strengths is that we used nationally representative data to investigate individual-, household-, and community-level factors associated with discriminatory attitude towards PLWHA. Nonetheless, the study has the following limitations. First, the findings may not represent all SSA countries since our analysis was based on only 12 countries. Second, recall bias may affect the findings, due to our reliance on self-reported data. 33 Finally, the cross-sectional nature of the study design may not permit concluding causal-effect relationship.

| CONCLUSION
More than one-third of pregnant married women in SSA had discriminatory attitude towards PLWHA. While women's educational level, husband educational level, decision-making, women's attitude towards domestic violence, and religion were pragmatic as important individual-/household-level factors associated with discriminatory attitude towards PLWHA, community socioeconomic status was the only community-level factors shown to be significantly associated with discriminatory attitude. To lessen the prevalence of discriminatory attitude towards PLWHA, considering these significant factors are needed. Therefore, governments and other stakeholders in the respective countries essential to increase education coverage. Moreover, empowering women through education and economy is crucial.
Finally, working with religious leaders to increase awareness about HIV and discriminatory attitude towards PLWHA should also be a priority in SSA.

TRANSPARENCY STATEMENT
Sanni Yaya affirms that the manuscript is an honest, accurate, and transparent account of the study being reported, that no important aspects of the study have been omitted, and that any discrepancies from the study as planned have been explained.

DATA AVAILABILITY STATEMENT
All analyzed data are freely available to the public through www. measuredhs.com.