Promises and Challenges of Faith-Based AIDS Care and Support in Mozambique

Abstract

Objectives. We sought to examine the role of religious organizations in the provision of HIV/AIDS-related assistance in Africa.

Methods. We used data collected from Christian religious organizations in southern Mozambique. Bivariate comparisons and logistic regression analysis of survey data were performed. We conducted an analysis of the qualitative data to complement the quantitative results.

Results. Our analysis revealed little involvement of religious organizations in provision of assistance. Most assistance was decentralized and consisted of psychological support and some personal care and household help. Material or financial help was rare. Assistance to nonmembers of congregations was reported more often than to members. Members of larger and better–secularly connected congregations were more likely to report assistance than were members of smaller and less–secularly engaged ones. Assistance was reported more in cities than in rural areas. Women were more likely than men to report providing assistance to congregation members, and the reverse was true for assistance provided to nonmembers. The cooperation of religious organizations in provision of assistance was hindered by financial constraints and institutional rivalry.

Conclusions. Policy efforts to involve religious organizations in provision of HIV/AIDS-related assistance should take into account that organization’s resources, institutional goals, and social characteristics.

The potential of religious organizations to mitigate the societal impact of HIV/AIDS is often invoked in both policy and general public discourse.1^,2 Faith-based responses to the pandemic have found enthusiastic support at the highest political level,3 and funding agencies have been looking to channel their AIDS assistance through churches and other religious organizations.1^,2^,4 However, relatively little systematic and unbiased evidence exists to show how this potential has been realized in practice.5^,6

The literature on the role of religion and religious organizations in dealing with the HIV/AIDS crisis is scarce and primarily addresses the influence of religious beliefs and organizational structures on HIV/AIDS risks and prevention. The main thrust of this literature is that religion and religiosity may serve as a barrier to HIV/AIDS by promoting, and even enforcing, less risky behaviors.2^,5^,7 Much of the literature focuses on differences between more and less conservative denominations. Thus, Garner8 found that, because of doctrinal and organizational factors, Pentecostals in KwaZulu Natal, South Africa, were less likely to engage in extramarital sex than were members of other churches. Gregson et al.9 came to a similar conclusion with regard to adherents of “spirit-type” churches in rural Zimbabwe. Hill and Ali10 reported that in Brazil, evangelicals had less extramarital sex than did Catholics.

The institutional resources of religious organizations, it is argued, may also help national and community-level prevention efforts.2 It is acknowledged, however, that religious prevention messages, usually extolling premarital abstinence and marital fidelity, often clash with the secular emphasis on safer sex.2^,11 Some religious leaders throughout Africa and in other parts of the world continue to portray HIV/AIDS as punishment for sexual sins, or to express opposition to condom use on the basis that condoms encourage promiscuity.12^–14

As the issues of treatment and care increasingly move to the forefront of the fight against the AIDS pandemic, it is imperative to examine how, if at all, religious organizations are involved with these issues. However, it is also important to stress the differences between the factors that constrain prevention-related and care-related activities. The former are frequently predicated on religious organizations’ official stance on marriage, family, extramarital sex, and condoms, and, therefore, are subject to heated ideological debate. In contrast, the latter are less ideologically laden and, therefore, less controversial, and are primarily contingent on religious organizations’ material resources, organizational structure, embeddedness within the community, and relationships with one another. Religious organizations have a long and celebrated tradition of health and social care provision, especially in resource-poor settings, and HIV/AIDS-related care and support should not be an exception.6^,15 However, much of the information on the involvement of religious organizations in HIV/AIDS–related care and support comes from leaders and advocates of such organizations.12^,16^–18 This information is often selective and self-congratulatory and does not allow for an impartial assessment of the achievements and problems in the provision of HIV/AIDS-related assistance by religious organizations. The volume and frequency of HIV/AIDS-related assistance provided by religious organizations, as well as the relative distribution of that assistance between members and nonmembers of religious organizations, generally remain unknown.

We have addressed these knowledge gaps with quantitative and qualitative data from a study conducted in 2003 in southern Mozambique (population 19 million). Mozambique is an impoverished country in southeast Africa, where adult HIV prevalence was estimated at about 16% at the time of study.19 Although Mozambique has a sizeable Muslim minority, especially in the north, the study area is overwhelmingly Christian. Our analysis focused on Christian churches, leaving out both nongovernmental organizations whose activities are guided by Christian faith and non-Christian religious organizations. We examined the exposure of rank-and-file members of religious congregations to individuals with HIV/AIDS in their communities and the involvement of those members, both formal and informal, in the provision of care and support to AIDS patients (or sick people with AIDS-like symptoms) and their families. To highlight the complexity of religious organization involvement in HIV/AIDS-related care and support activities, we drew comparisons between city and rural congregations, between different types of religious denominations, and between female and male church members’ perspectives on HIV/AIDS-related matters.

METHODS

The fieldwork was carried out in 2 areas in southern Mozambique—Maputo, Mozambique’s sprawling capital, with some 1.5 million residents, and the predominantly rural district of Chibuto (Gaza province), about 200 km north of Maputo. The 2 areas had similar estimated adult HIV prevalence levels of around 20% (i.e., somewhat higher than the national average).19 The study was congregation-based, meaning that only individuals attending religious activities of their respective congregations were included, and was limited to the denominations that were most prominent in the 2 sites: Roman Catholic, Presbyterian, and Pentecostal-type denominations—Zionist, Apostolic, and the Assemblies of God. In Maputo, the study also included the Islamic community, but in Chibuto, the Islamic community was still very small, so Muslims were not included. Because 1 of our goals was to make comparisons between the city and the countryside, we omitted the small subsample of city Muslims from the analyses.

The fieldwork consisted of a representative survey and semistructured, in-depth interviews. For the survey, congregations within each denomination were selected randomly in each of the 2 sites, for a total of 31 congregations. Respondents were selected randomly from groups attending religious services in those congregations on a given day and were interviewed at the congregation or another location of their choice. This sampling procedure assured a response rate of nearly 100%. Excluding the city Muslims, the survey sample included 677 respondents, of whom 52% were from Maputo, 29% from rural Chibuto, and 19% from the semirural town of Chibuto and its surroundings. Because women typically predominate among church attendees, we stratified the sample by gender (the age brackets for men and women were 15–60 years and 15–50 years, respectively). In the resulting sample, women constituted 53% of the survey respondents and men constituted 47%. The survey instrument contained a variety of items dealing with HIV/AIDS, including questions on assistance provided by the respondent’s congregation to AIDS patients and their families either within or outside the congregation. Although the survey instrument was devised to fully protect confidentiality of both the respondents and the recipients of congregation help, it is possible that some respondents, restrained by the stigma of HIV/AIDS and concerned with the image of their churches, understated the presence of HIV/AIDS and the provision of the corresponding assistance in their congregations. The qualitative data, however, suggest that the rate of underreporting was modest.

About 8% of the survey respondents—57 persons, 27 in Maputo and 30 in Chibuto—also participated in semistructured interviews following the survey. The subjects for those interviews were selected purposefully to represent a broad range of demographic characteristics (gender, age, and marital status) and religious involvement. The semistructured interviews were designed to complement and expand upon the survey, especially in the areas of personal perceptions of HIV/AIDS risks and knowledge about the role of respective religious congregations in HIV/AIDS prevention and care.

In analyzing the survey data, we examined 2 types of outcomes: (1) personal encounters with HIV/AIDS, approximated by knowledge of someone who had HIV/AIDS or had died from it, and (2) assistance provided by the congregation to HIV/AIDS patients or their families. For each outcome type, we distinguished between church members and non-members and made comparisons along 3 dimensions. First, we compared city (Maputo) congregations to rural and semirural (Chibuto) congregations; aside from obvious socioeconomic and cultural differences between the 2 sites, access to testing and treatment was generally better in the city than in the countryside. Second, we examined differences across types of denominations. Although denominational nuances in the sub-Saharan context may be complex and elusive, we focused on a relatively simple dichotomy that encompasses both ideological and organizational dimensions critical to the role of churches in HIV/AIDS prevention and care.

We considered 2 groups of denominations: the first group was made up of “mainline” churches (Roman Catholic and Presbyterian), and the second group included the remaining churches (Zionist, Apostolic, and Assemblies of God), which we labeled “healing” churches because of the centrality of divine healing to their doctrines and practices. Mainline churches are generally more flexible doctrinally and more open socially; healing churches tend to have more conservative and rigid teachings and to be inward-oriented. This dichotomy has proven relevant to knowledge and views about HIV/AIDS prevention,20 and we now extend it to matters of care and support. Finally, we compared men’s and women’s perspectives on the outcomes of interest, because HIV/AIDS risks, views, and practices are constrained by gender ideology and vary by gender.20^–22 In order to ensure that area-, denominational-, and gender-associated variations in the provision of HIV/AIDS care were not confounded by other factors, we fit a multivariate logistic regression. The model controls for age, education, marital status index (defined on the basis of household possessions), marital status, and church attendance. We first present and interpret the patterns of association found in the survey data, and then elaborate on these patterns drawing from the semistructured interviews. Although not quantifiable, the insights from the interviews provide valuable information about the content and mechanisms of church-based HIV/AIDS care and support.

RESULTS

Encounters With HIV/AIDS and Exposure to Church-Based Assistance

Table 1 ▶ summarizes our results by area of residence, type of denomination, and respondent gender (associations with a χ² result statistically significant at the P < .05 level are noted as such). Despite high HIV prevalence in the region, more than 40% of the survey respondents did not personally know anyone who was HIV positive or had AIDS. The differences between rural and city residents and between men and women were not statistically significant. However, members of mainline churches demonstrated significantly greater familiarity with HIV/AIDS cases compared with respondents affiliated with healing denominations. This association was not owing to higher incidence of known HIV/AIDS cases among mainline denominations. In all, only about 10% of respondents reported knowing at least 1 person with confirmed or suspected HIV/AIDS within their respective congregations, and this percentage was about the same across the areas of residence, types of denomination, and gender.

TABLE 1—

Knowledge of HIV/AIDS Cases and Reporting of Church-Based Provision of HIV/AIDS Care and Support (%): Survey of Religious Congregations, Mozambique, 2003

		Location of Congregation		Type of Denomination		Reporting by Gender
	Total	Chibuto	Maputo	Healing	Mainline	Men	Women
No. respondents	677	325	352	383	294	317	360
Knows at least 1 person with HIV/AIDS	58.7	55.3	61.8	54.9*	63.7*	58.4	58.9
Knows a church member with HIV/AIDS	9.6	7.8	11.4	9.4	9.9	9.8	9.5
Church ever helped a member with HIV/AIDS	4.1	3.4	4.8	3.7	4.8	2.5*	5.6*
Church ever helped a nonmember with HIV/AIDS	17.1	8.0*	25.6*	12.8*	22.8*	20.2*	14.4*
Church ever helped any HIV/AIDS victim/family	20.3	10.6*	29.3*	15.2*	27.0*	22.5	18.4

*Associations significant at P < .05.

A very small percentage of respondents—only about 4%, on average—said that their churches had ever helped their fellow church members suffering from HIV/AIDS. Notably, the only significant difference was between men and women: the latter were more likely to report such activities in their congregations. By contrast, men were significantly more likely than women to report church assistance to HIV/AIDS patients outside the congregation. Overall, 17% of respondents said that their congregations had provided assistance to nonmembers affected by HIV/AIDS. In addition to the already noted gender differences, such assistance was much more commonly reported by members of city churches than by members of rural and semirural churches. Mainline church members were significantly more likely to report the involvement of their respective congregations in such activities than were members of healing churches.

The percentages in the last row of Table 1 ▶ refer to respondents who said that their congregations had offered HIV/AIDS care and support to any HIV/AIDS patients (church members and nonmembers) or their families. About one fifth of respondents reported the involvement of their congregation members, and the differences between Maputo and Chibuto, and between mainline and healing denominations, reflected the corresponding patterns detected in assistance provided to nonmembers. The gender differences, however, were not significant, largely because the opposite trends detected in assistance to church members and nonmembers nearly canceled each other out.

To ascertain the net effects of denomination type, location, and gender, we subjected these effects to a multivariate analysis. Table 2 ▶ presents the results of a logistic regression model that predicts the odds of reporting church-based provision of HIV/AIDS assistance to either members or nonmembers by location, type of denomination, and gender; the model controls for respondent’s age, education, marital and material status, and frequency of church attendance. The results confirmed the bivariate associations. In fact, the type of denomination and the congregation’s location were the only 2 predictors that significantly affected the outcome.

TABLE 2—

Reporting of Church-Based HIV/AIDS Care and Support: Survey of Religious Congregations, Mozambique, 2003

Predictor	Odds Ratio (95% CI)
Maputo (Chibuto)	2.73 (1.71, 4.38)
Mainline church (healing church)	2.18 (1.37, 3.47)
Women (Men)	0.91 (0.61, 1.36)
Age ≥ 30 y (age ≤ 29 y)	0.89 (0.56, 1.41)
≥ 6 y of education (< 6 y)	1.65 (0.97, 2.79)
Material status index	1.02 (0.87, 1.20)
Currently in marital union (currently not in union)	0.78 (0.51, 1.21)
Attends church ≥ 3 times/wk (attends < 3 times/wk)	1.45 (0.91, 2.30)
Model likelihood ratio χ2	63.1*
No. respondents	670

Note. CI = Confidence interval. Reference categories are in parentheses in first column.

* χ² test significant at P < .05.

Types of Care and Support Provided Through Religious Organizations

Both the survey and semistructured interviews supplied information on the types of assistance offered through church channels. Table 3 ▶ groups the reported types of assistance ever provided (regardless of volume and frequency) into 2 broad categories: psychological support (prayer, advice, encouragement, etc.) and tangible support (help with money, food, transportation, and other tangible items, as well as personal physical care). Psychological support was reported by a slightly higher number of respondents than was tangible support for both church members and non-members. The gaps between urban and rural churches and between the 2 types of denominations remained large but, as in the overall comparisons (Table 1 ▶), were only statistically significant for assistance provided to non-members. Interestingly, when assistance was broken down by type, the gender differences were no longer statistically significant.

TABLE 3—

Types and Recipients (%) of Reported Church-Based HIV/AIDS Care and Support: Survey of Religious Congregations, Mozambique, 2003

		Location of Congregation		Type of Denomination		Gender
	Total	Chibuto	Maputo	Healing	Mainline	Men	Women
No. respondents	677	325	352	383	294	317	360
Church ever provided psychological support to members with HIV/AIDS	14.5	8.0*	20.5*	10.2*	20.1*	15.8	13.3
Church ever provided psychological support to nonmembers with HIV/AIDS	3.7	2.8	4.6	3.1	4.4	2.5	4.7
Church ever provided psychological support to any HIV/AIDS victim/family	12.3	6.5*	17.6*	8.9*	16.7*	14.2	10.6
Church ever provided tangible support to members with HIV/AIDS	11.0	5.9*	15.6*	7.8*	15.0*	12.0	10.0
Church ever provided tangible support to nonmembers with HIV/AIDS	3.1	2.5	3.7	2.9	3.4	2.9	3.3
Church ever provided tangible support to any HIV/AIDS victim/family	8.1	3.4*	12.5*	5.2*	11.9*	9.5	6.9

*Associations significant at P < .05.

The data presented in the tables cannot show important details about the volume, frequency, nature, and mechanisms of assistance to patients with HIV/AIDS and their families. The qualitative data help shed some light on these nuances, and we summarized the most valuable insights from the in-depth interviews. They suggest that prayer and related forms of psychological support by far constituted the most frequent and sizeable form of assistance. Less frequently, church members, usually women, provided help with basic personal care and with house care—cooking, cleaning, occasional child care, and so on. Financial and other material help remained very limited, with most interviewees citing the lack of resources to share. When offered, this help was usually reported to have gone toward transportation fares to the clinic or hospital, medical fees, drug costs, and, especially, funeral expenses.

Much of the assistance, especially psychological and spiritual support, was neither centralized nor organized. Depending on the size of the congregation, these activities may have involved only a fraction of congregation members, and, in cases where no material or financial transfers occurred, may not even have had a specific sanction of the church leadership. For example, individual church women or groups of women could have visited a sick member or a nonmember living in the church neighborhood and prayed for his or her recovery without the church superior leaders being aware of it. These activities were sometimes part of the church routine, aimed at supporting its members in distress, as well as similar regular or occasional activities outside the congregation (especially home prayer visits), which often also carry an implicit proselytizing impetus.

Congregation leaders usually coordinate more formal assistance, especially assistance involving money or material resources. The data suggest that such centralized assistance tends to be more common in collectivist rural settings than in the individualized city environment. Yet congregation leadership involvement in the provision of assistance to the sick outside the congregation may be problematic, because of church rivalries. Religious leaders are particularly cautious while sanctioning such activities in rural areas, where church boundaries are clearly demarcated and well known; overstepping these boundaries, even for the purpose of caring for the sick, may be interpreted as unsolicited proselytizing.

Interchurch Coordination of HIV/AIDS-Related Activities

Denominational and organizational differences and competition notwithstanding, churches do coordinate some of their HIV/AIDS-related assistance. The activities of Tshembeka (“to be trustworthy” in Shangana, the main local language), a nongovernmental organization operating in Chibuto, illustrate both the potential and challenges inherent in the involvement of religious organizations in interfaith HIV/AIDS care and support. Tshembeka was established with a small grant from the United States Agency for International Development awarded to World Relief, an international nongovernmental organization promoting community-based social projects. Tshembeka is made up of a dozen churches whose volunteer activists identify patients with AIDS-like symptoms in their neighborhoods and visit them at home to encourage them to get tested for HIV and to offer moral support and counseling if the patients’ seropositivity was confirmed (HIV-negative individuals were also said to receive similar support).

According to Tshembeka’s coordinator, a local mainline church leader, the organization had virtually no financial resources beyond those necessary to provide basic training to volunteer activists, to obtain pamphlets, and to obtain a small quantity of disposable gloves. As a result, earlier attempts to offer more complex home-based care and nutrition supplements had been abandoned or drastically reduced, and dropout and turnover rates among activists were high. High transportation costs prevented Tshembeka activists from taking patients to specialized HIV/AIDS clinics located in the province’s 2 main cities.

Transportation costs were also a major barrier to the expansion of Tshembeka’s activities beyond the limits of the town of Chibuto into more distant rural areas of the district. However, even in the town, only a small fraction of churches was involved; the rest chose to stay out of the project. Although no church leaders would directly admit that his congregation’s nonparticipation was a result of inter-church friction, our fieldwork showed that such friction does play an important role. Even for many leaders whose churches were involved in Tshembeka’s activities, the humanitarian and charitable aspects of these activities were intertwined with the quest to boost the prestige and, eventually, the membership of their congregations.

DISCUSSION

Our study shows that despite high HIV prevalence in the region, public encounters with individuals with HIV/AIDS remained infrequent, as did the involvement of religious organizations in the provision of HIV/AIDS-related care and support. The degree of involvement, however, varied between urban and rural areas and between different types of denominations. Within churches, the type of involvement in HIV/AIDS care and support also varied by gender.

The assistance offered by religious organizations included both tangible and intangible components. The latter clearly predominated, whereas the former was usually reserved for extreme situations—terminal stages of AIDS-related illness or for funerals. Much of the church-provided assistance was small in scale and episodic; it was neither organized nor controlled by the church leadership. Congregation leadership was involved mainly in infrequent, larger-scale actions that required pooling of meager congregational resources. Whatever the level and scope of assistance, it often targeted people outside of congregations. Church leaders’ involvement in HIV/AIDS assistance to non-members of theirs churches could be perceived by outsiders as attempts to spread their influence beyond the ranks of their congregations. These perceptions, and the resulting tensions among church leaders and members, hinder interfaith congregational cooperation initiatives, such as Tshembeka, which are already crippled by a chronic shortage of financial resources.

Our analysis suggests, therefore, that the potential of religious organizations in the fight against HIV/AIDS in southern Mozambique (which we perceive to be as a typical sub-Saharan setting) exists, but this potential is underutilized. This underutilization stems, in part, from limited practical exposure to HIV/AIDS cases, in part from church financial constraints and from interchurch ideological tensions and organizational rivalries. As the health and social burden of HIV/AIDS mounts, it becomes increasingly critical to remove the barriers to greater utilization of religious organizations. Because the material and financial resources of most religious organizations in sub-Saharan Africa are unlikely to improve substantially, the efforts of policymakers should focus on the sociopsychological and organizational obstacles faced by these organizations.

Critical areas for the involvement of religious organizations in HIV/AIDS-related assistance could include encouraging both members and nonmembers to use voluntary counseling and testing services, providing support for access to antiretroviral therapy, and supporting antiretroviral therapy adherence. Because the success of voluntary counseling and testing and antiretroviral therapy is obstructed not only by individuals’ material and financial constraints but also by the stigma and similar social barriers, the emphasis of religious organizations on compassion and solidarity may prove an effective mobilization tool.23^–25 At the same time, while seeking to enlist the help of religious organizations in providing this and other types of HIV/AIDS-related assistance, policymakers should take into account doctrinal and organizational differences among religious denominations, as well as their perpetual (and often bitter) ideological and organizational competition.