Evidence-based monitoring and evaluation of the faith-based approach to HIV prevention among Christian and Muslim youth in Wakiso district in Uganda

Background: The Islamic Medical Association of Uganda, has been implementing the faith-based approach to HIV prevention without baseline data on expected positive outcomes. Objectives: To establish evidence-based baseline data on expected positive outcomes of the faith-based approach to HIV prevention. Methods: A cross-sectional study of 15-24 year-old youths was analyzed for significant associations between HIV infections, risky behaviors, and religiosity. Results: HIV prevalence was 3.6% among Christians and 2.4% among Muslims. Abstaining from sex among teenagers was at 54% for Christians and 58% for Muslims. Being faithful in marriage among males was at 41% for Christians and 34% for Muslims and among females it was 65% for Christians and 69% for Muslims. Praying privately was associated with lower HIV infections and was observed among 60% of Christians. Sujda, the hyperpigmented marker of regular prayers on the forehead of Muslims was associated with lower HIV infections and observed in 42% of them. Ever drank alcohol was associated with higher HIV prevalence and observed in 52% of Christians and 17% of Muslims. Male circumcision rates were 15% for Christians and 98% for Muslims. Conclusion: A sero-behavioral-religiosity survey can provide evidence-based data for monitoring and evaluation of the faith-based approach to HIV prevention.


Introduction
Uganda still has a high HIV prevalence of 6.4% and a high rate of new HIV infections which were estimated to be 135, 000 in 2005 1 . In Uganda's National HIV/AIDS Strategic Plan 2007/8 -2011/ 12, the goal of HIV prevention is to reduce the incidence rate of HIV by 40% by the year 2012. In order to achieve this, one of the strategic actions is to promote ABC+-which includes: Abstinence, Being Faithful, and Condom use with risky sexual encounters. Another strategic action is to utilize religious institutions for the delivery of HIV prevention messages 2 . Faith-based organizations are recognized promoters of abstinence and faithfulness in marriage as a means of HIV prevention. They do this by using their faith teachings that encourage these supportive behaviors 3,4 . These faith teachings are intended to increase religiosity within the target communities.
The Islamic Medical Association of Uganda (IMAU) is one of the faith-based organizations that has been promoting the faith-based approach to HIV prevention among Muslim and Christian communities. In this approach, Muslim and Christian religious leaders were trained to educate their communities using a curriculum with scientific information supported by religious teachings from the Holy Qur'an and Holy Bible 5

Relationship of the Islamic teaching to HIV prevention
People should not indulge in anything that stimulates their sexual desires to commit adultery. If one commits adultery it is a shameful act and an evil because it may result in getting HIV infection.
The believers must eventually win through; those who humble themselves in their prayers, who avoid vain talk, who are active in deeds of charity, who abstain from sex, except with those joined to them in the marriage bond ……(Holy Qur'an 23: 1-6. Translated by Abdullah Yusuf Ali).
The believers in Allah's guidance will eventually win in preventing HIV infection if they abstain from sex except to those married to them. This means that both men and women abstain from sex before marriage in which case they remain HIV negative and when they get married they only have sex with their HIV negative marital partners. This assumes that they do not get HIV infection from other routes such as HIV contaminated blood transfusion, getting pierced by HIV contaminated instruments, and through mother to child HIV transmission which may occur in a minority of cases who may have followed the Islamic teachings.

B. Holy Bible
You have heard that it was said, "Do not commit adultery". But now I tell you: anyone who looks at a woman and wants to possess her is guilty of committing adultery with her in his heart. So if your right eye causes you to sin, take it out and throw it away! It is much better for you to lose a part of your body than to have your whole body thrown into hell. If your right hand causes you to sin cut it off and throw it away! It is much better for you to lose one of your limbs than for your whole body to go to hell. It was also said, "Anyone who divorces his wife must give her a written notice of divorce". But

Relationship of Christian teaching to HIV prevention
People should not indulge in activities that stimulate their sexual desires to commit adultery such as looking lustfully at women. If they do so they would eventually commit adultery which may result in HIV infection. People should struggle to keep their marriages so that they minimize chances of getting and spreading infections such as HIV.
The IMAU has been using this faith-based approach to HIV/AIDS prevention for Muslim and Christian communities in Wakiso district since 2005. However, IMAU had no baseline data on expected positive outcomes upon which to base future monitoring and evaluation. It is recommended that surveys with both serological and behavioral components should be institutionalized and conducted every 2-3 years in order to improve data availability so that evidence-based planning and decision making can be made for HIV prevention interventions in the target communities 1 .
The objective of the study was to establish baseline data on expected positive outcomes of the faith-based approach to HIV prevention among Muslim and Christian communities in Wakiso district.

Study setting
The Islamic Medical Association of Uganda runs a hospital called Saidina Abubakar Islamic Hospital located at Wattuba, in Wakiso district, 14km on Bombo road. The IMAU has been promoting the faith-based approach to HIV prevention for communities surrounding this hospital. A convenience sample of 30 mosques and 30 churches around the hospital were identified to participate in the study.

Study design and measurement of religiosity
This was a cross-sectional study. Religiosity was defined as the degree of adherence to religious practices and getting associated spiritual experiences. It was measured using a modified Brief Multidimensional Measure of Religiousness/ spirituality 6 . The dimensions of religiosity that were most significantly associated with lower HIV infections were identified. These significant dimensions were included among indicators for which baseline percentages were to be determined from the cross-sectional data. Other dimensions of religiosity which were not statistically significantly associated with HIV infections but were known to be associated with several behaviors supportive of HIV prevention were also included as baseline indicators.

Assessment of socio-demographic characteristics and HIV-risk behaviors
Questions on socio-demographic characteristics and relevant HIV-risk behaviors were included in the study instrument. The socio-demographic and behavioral factors that appeared to be driving the epidemic in the target communities were identified in data analysis. These were also included among the baseline indicators. The indicators on Abstinence and Being faithful in marriage were included because they are a special target for the faith-based approach to HIV prevention.

Assessment of new HIV infections
The prevalence of HIV in the 15-24 year age-group is a reasonable proxy indicator of new HIV infections 7,8 . The IMAU hospital staff performed home-based HIV counseling and testing targeting clients 15-24 years of age. They used rapid HIV testing kits following the recommendation of the Uganda Ministry of Health 9 .

Results
Between July and December 2010, 4 268 clients aged 15-24 years were counseled and tested. There were 1 224 Muslims and 2 933 Christians. Because of different faith traditions, data analysis for Christians and Muslims was done separately. For example, Christianity does not advocate for excessive alcohol consumption but alcohol drinking is not altogether forbidden. In contrast, alcohol drinking is forbidden in Islam. Another example is that in accordance with Islamic teachings, Muslim males are circumcised as part of their religious obligation. On the other hand, Christian males are not generally circumcised according to their faith tradition. Both alcohol consumption and lack of circumcision have been associated with higher HIV infection rates 10 .
Male circumcision is a practice that has been associated with lower HIV infections. It may be initiated by parents for their children or by the individual himself later in life. In our study, among male Christians who were circumcised 3 (2%) were HIV positive and among the uncircumcised 42 (4%) were HIV positive. The difference was not statistically significant. Among the circumcised Muslim males, 10 (2%) were HIV positive and among the circumcised Christian males, 3 (2%) were HIV positive. The difference was not statistically significant (odds ratio 1.16, 95% CI 0.31-4.26, p=0.823). When all males were analyzed with Muslims and Christians combined, among those who were circumcised, 13 (2%) were HIV positive and among the uncircumcised, 42 (4%) were HIV positive. The difference was statistically significant (odds ratio 2.21, 95% CI 1.17-4.14, p=0.012). The majority of the circumcised HIV positive respondents (10 out of 13), were Muslims and all the 42 uncircumcised HIV positive respondents were Christians.

Religiosity dimensions associated with HIV infections among christians
The religiosity dimensions associated with HIV infections are shown in table 1. Religiosity dimensions were assessed as ordinal variables. For the purpose of comparison, respondents who reported the highest order of the dimension were categorized as having high religiosity for that dimension. Since there were very few respondents who reported the lowest orders of the religiosity dimensions, all other responses were categorized as moderate religiosity for the particular dimension under review. Behavioral factors and religiosity dimensions associated with age-group among christians Abstaining from sex was more common among teenagers as expected (odds ratio 12.80, 95% CI 10.13-16.19, p<0.001). Being faithful in marriage was more common among teenagers (odds ratio 1.91, 95% CI 1.30-2.80, p<0.001). Young adults were more likely to have multiple lifetime sexual partners (odds ratio 2.97, 95% CI 2.33-3.78, p<0.001). Sex during menstruation was also more common among young adults (odds ratio 1.35, 95% CI 1.02-1.79, p=0.037).

Baseline sero-behavioral-religiosity indicators among christians
The baseline serological, behavioral and religiosity indicators that were included in monitoring and evaluation of the faith-based approach to HIV prevention in the Christian communities are shown in table 2. They were derived from evidence obtained from the study. Baseline serological indicators These were derived from HIV prevalence segregated by gender. The overall HIV prevalence among Christian respondents was 3.6%.

Baseline behavioral indicators
These were derived from indicators that were significantly associated with HIV infection in the study and were most likely to be driving the HIV epidemic.
In addition, indicators of ABC+ were included among the baseline indicators, because this is one of the national strategic actions for HIV prevention. In the population studied, only abstinence from sex was significantly associated with lower HIV infections. Condom use is not usually significantly promoted in the faith-based approach because it is thought to encourage sex outside marriage which goes against both Muslim and Christian faith teachings. However it was included among baseline indicators because the majority of the respondents had ever used condoms anyway. Male circumcision was not significantly associated with lower HIV infection rates among the respondents studied. However, there were more HIV positive cases among the uncircumcised compared to the circumcised. Male circumcision was included among the baseline indicators because it is one of the national strategies for HIV prevention.

Baseline religiosity indicators
Baseline religiosity indicators were derived from dimensions of religiosity that were found to be significantly associated with HIV infection in the analysis of the data. Other religiosity dimensions that were significantly associated with behaviors supportive of HIV prevention such as church attendance and Bible reading were included although they were not significantly associated with HIV infections among the study respondents.

Muslim respondents Socio-demographic factors, behavioral factors and religiosity dimensions associated with HIV infections
The socio-demographic and behavioral factors associated with HIV infection among Muslims are shown in table 3. Muslim respondents who ever drank alcohol, had a higher HIV infection rate (odds ratio 2.28, 95% CI 1.02-5.10, p=0.038). Muslim respondents who ever had sex during menstruation also had a higher HIV infection rate (odds ratio 2.57, 95% CI 1.02-6.46, p=0.037). Muslims who had lost one or both parents also had higher HIV infections (odds ratio 3.23, 95% CI 1.44-7.23, p=0.003). Muslims who had no Sujda, the hyperpigmented spot on the forehead as result of prostration during regular prayers, had a higher HIV infection rate compared to those with Sujda (odds ratio 2.74, 95% CI 1.10-6.83, p=024). Muslims who fasted one month or less per year also had a higher HIV infection rate compared to those who fasted more frequently (odds ratio 2.46, 95% CI 1.07-5.67, p=0.028).

Baseline sero-behavioral-religiosity indicators for muslims
The baseline sero-behavioral-religiosity indicators for Muslims in the target community are shown in table 4. The overall HIV prevalence among Muslim respondents was 2.4%. The behavioral indicators include baseline percentages for ABC+ which is a national strategy for HIV prevention. Religiosity indicators were derived from religiosity dimensions that were significantly associated with lower HIV infection rates in the study. The main significant indicators were Sujda and fasting. Other indicators were added which were known to be significantly associated with behaviors supportive of HIV prevention although they were not associated with lower HIV infections among the study respondents. These included males wearing a Muslim cap and females wearing a long dress (Hijab). One limitation of our study was that the behavioral data were self-reported. It is possible that some respondents could have given socially desirable answers to some questions. This could result in underestimation or overestimation of the frequency of some behaviors depending on their social desirability. However, it is likely that this effect was minimal because the results were consist with what other researchers have noted. For example, the drivers of the epidemic in the community studied included multiple sexual partners, alcohol use, and narcotic drug use for recreation. This is consistent with what has been documented. 1 HIV infections were significantly associated with respondents who have ever had sex during menstruation. This is consistent with what other researchers have noted. Sex during menstruation may result in a five-fold increase in the risk of a woman becoming HIVinfected by an HIV positive sexual partner. Men who engage in sexual intercourse with an HIV positive woman during menstruation are more than 3 times as likely to contact HIV compared to men with HIV positive partners who do not have sex during menstruation. The increased risks for men getting HIV infections when sex involves blood are due to the higher concentration of HIV in blood compared to that in the non-bloody genital secretions 14,15 . Among Christian respondents, women were more religious and had fewer sexual partners outside marriage. This is consistent with what other researchers have observed 16 .
The baseline sero-behavioral and religiosity indicators were derived from the target community. They may be used in future sero-behavioral-religiosity surveys every 2-3 years to monitor and evaluate the status of the epidemic in the community. This is one of the recommendations in the Ugandan HIV prevention strategy 1 . Religious leaders may use the baseline indicators to plan their interventions and target the behavioral factors driving the epidemic in their communities. They may use faith teachings that discourage identified risky behaviors that are associated with HIV infection. They may also encourage religious behaviors and practices that were associated with lower HIV infections among the study respondents.
In future cross-sectional sero-behavioralreligiosity surveys to monitor and evaluate the status of the HIV epidemic in the target communities, it will be difficult to attribute any changes to any particular intervention. This is a well recognized problem 17 . If there are improvements in the indicators, the best that may be noted is that the faithbased approach to HIV prevention could have contributed to the positive outcomes. More rigorous study designs such as cohort studies or randomized community intervention trials would have to be done to make firm statements about the cause of any observed changes 18 . More studies also need to be done to assess how HIV affects religiosity. Some studies have shown that those with HIV report greater use of religion and spirituality when compared with similar HIV-negative individuals 19 . It is also important to find out if different Muslim and Christian thinking affects the relationship between HIV and religiosity. For example, it would be interesting to find out if "Conser vative" or "Fundamentalist" Christians or Muslims differ from the "Moderate" ones regarding the interaction between HIV and religiosity.

Conclusion
A sero-behavioral-religiosity survey can provide evidence-based data which can be used to monitor and evaluate the faith-based approach to HIV/AIDS prevention. However, there are limitations to attributing any observed future improvements in the baseline indicators to any one intervention.