Professional Documents
Culture Documents
Mpoma Health Profile 2011
Mpoma Health Profile 2011
11!
Mpoma Community HIV/AIDS Initiative Lukojjo Village, Mukono District, Uganda ! +256 775364300 office ashlaurenrogers@gmail.com
Table of Contents
1.0 2.0 2.1 2.2 2.3 2.4 3.0 3.1 3.2 3.2.1 3.3.1 3.3.2 3.3.3 3.4.1 3.4.2 3.5 3.5.1 3.5.2 3.5.3 3.6 3.6.1 3.6.2 3.7 3.7.1 3.7.2 3.8 3.8.1 3.8.2 4.0 5.0 EXECUTIVE SUMMARY METHODOLOGY OBJECTIVES ASSESSMENT DESIGN HOUSEHOLD SURVEY PARTICIPATORY ASSESSMENT ANALYSIS PROXY INDICATOR OF ECONOMIC STATUS ENVIRONMENTAL HEALTH HYGIENE INFRASTRUCTURE TYPE OF WATER SOURCE NUMBER OF WATER SOURCES WATER TREATMENT FOOD SOURCE ENERGY SOURCE HEALTH FACILITIES TYPE OF HEALTH FACILITY DIVERSITY OF HEALTH SERVICES SATISFACTION WITH HEALTH SERVICE FAMILY PLANNING PREVALENCE OF FAMILY PLANNING FAMILY PLANNING METHODS MALARIA MALARIA INCIDENCE PREVALENCE OF SLEEPING UNDER NETS HIV/AIDS HIV STATUS AWARENESS HIV PREVALENCE CONCLUSION APPENDIX 3 4 4 4 4 5 6 6 7 7 8 9 10 11 11 12 12 13 13 14 14 15 16 16 17 19 19 20 20 22
1.0
Executive Summary
The villages that form Mpoma Community have a unique health profile that is driven by socioeconomic, environmental, and behavioral factors. This report sketches the health status of this community as a rough, but informative, guide to local policy makers, project implementers, and health workers. Using a survey of 247 households (covering 1200 individuals) and a participatory assessment engaging community members and Village Health Teams (VHT), this study finds:
Water-borne Illnesses
Poorer households are at greater risk of water-borne illnesses because they are 1) 13% more likely to have access to only one water source and 2) 7% less likely to treat drinking water. Half of all households in Mpoma and Lutengo villages rely on contaminated water sources. Households of lower economic status are more vulnerable to crop failure as they are 13% more likely to only source food from their gardens. Conversely, households of higher economic status are more vulnerable to food price fluctuations as they are 11% more likely to only source food through purchase. Fifty-seven percent of households rely solely on firewood, while another 25% use both firewood and charcoal. Households with lower economic status are 30% more likely to use only firewood than households with higher economic status. Households with lower economic status are 7% more likely to self-medicate. Proximity is the most significant determinant of the type of health services households access. Households from Mpoma and Lwanyonyi villages are particularly vulnerable because they have the lowest economic status and face considerable obstacles to reach health centers. Due mostly to drug shortages in government facilities, households that only use government health services are 30% less likely to be satisfied with the care they receive. Only 40% of households practice a method of family planning. Forty-five percent of households that arent using a method of family planning are interested in receiving family planning services. Sixty-nine percent of households that practice family planning are interested in further services. Households with lower economic status are 23% less likely to use a method of family planning. Over one-third of individuals report suffering from malaria in the last two months. Fifty-seven percent of individuals report sleeping under a bed net, however bed net use is 6% higher amongst pregnant women and 8% higher amongst children under five. Though 82% of households name cost as the major limitation to using a bed net, net use is driven by community-wide socio-economic factors, rather than household economic status. Only one-third of the overall population is aware of their HIV status. The rate of awareness varies significantly by village and is connected to access to free testing. Wakiso is especially in need of HIV testing, as only 20% of residents are aware of their status. 3
Food Security
Energy Consumption
Malaria Prevention
2.0
2.1
Methodology
Objectives
In October 2011 Mpoma Community HIV/AIDS Initiative (Mpoma) conducted a health assessment of Nama subcounty. The objectives of this assessment were to: 1. Create baseline data of community health 2. Understand socio-economic and environmental factors that contribute to health outcomes 3. Ascertain levels of sensitization in areas of: family planning, malaria, and HIV/AIDS 4. Estimate prevalence of illness and prevention with special emphasis on malaria
2.2
Assessment Design
Mpoma collected quantitative and qualitative data through an in-person household survey of 247 participant households. Additionally, Mpoma collected qualitative data through a participatory assessment process conducted with community members and Village Health Teams (VHTs). Community Integrated Development Initiative (CIDI) commissioned the participatory assessment with a special interest in malaria prevention. This analysis combines the results of both studies. Mpoma decided to use a mixed-method design because such an approach: allowed for greater validity and credibility, reducing the threat of methodological error by accessing data from groups and from individuals in two different settings, created a more robust portrait of health status by allowing local leaders to inform the analysis of the quantitative data, and increased the depth of inquiry by opening up the range of questions enumerators could ask.i
2.3
Household Survey
In October 2011 Mpoma enumerators surveyed 245 households from 3 parishes within Nama subcounty. Mpoma Community HIV/AIDS Initiative chose Mpoma, Bulika, and Namubiru parishes to represent variation across the following characteristics: rural/urban, number of villages, population, income levels, and distance from health centers. Mpoma then selected two villages from each parish. Again, selection was based on representing the greatest possible variance on the aforementioned characteristics. At the village level, the enumerators first met with the Local Council 1 (LC1) to map the households in the community. From that mapping, approximately forty households were randomly selected, as seen in Figure 1. Figure 1: Number of respondents by village
Parish Bulika Bulika Mpoma Mpoma Namubiru Namubiru Village Lutengo Wakiso Mpoma Nsanvu Lwanyonyi Takajunge # of Respondents 49 38 40 39 41 39
Enumerators targeted heads of households as the key respondents and asked them to provide information concerning the entire household. Thus, as seen in Figure 2, a total of 1200 people are represented in the survey. Figure 2: Respondent profile
Respondent information Number of households Average age of respondent Average household size Average number of children Average number of children under 5 Total population covered in survey 246 40.4 4.9 3.5 1.34 1200
The survey itself includes recall data collected from the participants to answer questions covering household environment, knowledge of household members illness prevention methods, frequency of household members illnesses, and household members health practices. There are inherent reliability issues with head of household reporting because participants may forget, overestimate, underestimate, or intentionally skew data. Additionally, the respondent may have inaccurate information concerning other members of the household. Thus, the survey also employed observational data via an observation guide. This was a useful data collection tool to collect benchmark and descriptive data of the household environment. The method was adapted from examples found in Stake (1995), University of Wisconsin (1996), and Yin (2003).ii 2.4 Participatory Assessment Mpoma also conducted a participatory assessment with funding from and in coordination with CIDI. This assessment occurred in two stages. The first was through a series of focus groups meetings held with Village Health Teams (VHTs). The second stage consisted of meetings with community members who had been involved in a maternal malaria prevention program implemented through an Mpoma-CIDI partnership. According to the World Bank (2009), The group process tends to elicit more information than individual interviews, because people express different views and engage in dialogue with one another. iii Accordingly, this assessment uses data from the participatory assessment to corroborate and explain survey data.
3.0
3.1
Analysis
Proxy Indicator of Economic Status
Poverty is often strongly correlated with health outcomes. However, measuring incomes levels or other quantitative indicators of poverty is extremely technical and costly. Self-reported income is not reliable within this population because most people do not receive a set salary nor keep track of inflows and outflows of financial resources. Additionally, much of the wealth of a household is held in assets like farm animals, agriculture inputs, housing and land, rather than currency. Thus, this study employs floor type (either brick or mud) as an indicator of economic status. The survey also recorded occupation, wall type, tenancy, and house permanence. However, after reviewing the data with community members through our participatory assessment, we found that of all these variables, floor type is most consistent with local delineation of economic status. Even more, the population is split almost exactly in half by floor type, providing a simple way to bifurcate the population by wealth. Though this is a rough measurement, it offers insight to the role of poverty in health outcomes. Figure 3 illustrates floor type by village, and accordingly shows a rough idea of the relative wealth of each village. Ttakajjunge is the most economically secure village with 77% of households having cemented floors. Mpoma and Lwanyonyi villages are the least, with only 35% and 38% respectively. Lutengo, Wakiso, and Nsanvu all hover around the average of 49%. Figure 3: Floor type as proxy for economic status
63% Households
60%
35%
38%
Mpoma
Lwanyonyi
Nsanvu
Wakiso
Lutengo Ttakajjunge
All
3.2
Environmental Health
Figure 5: Wall type disaggregated by village 3.2.2 Wall Type Nearly, three quarters of households have brick walls. Ttakajjunge has the highest incidents of bricks walls, at 94% of households, while Lutengos has the lowest, with only 65% of households. The remaining villages cluster around the average.
75%
80%
78%
82%
94% 65%
79%
Mud Brick
3.3
Water Access
3.3.1 Type of water source Figure 6: Types of water sources (all villages) As seen in Figure 6, the most common types of water sources are rainwater, Types of Water Sources 41%; spring water, 30%; and boreholes 18%. There is no statistically significant difference in type of water sources between those with lower Spring economic status and those with water 30% Rainwater higher. However, water source does Tapped 41% vary significantly by village (See water Chart 1, Appendix). 1% Figure 7 shows that Mpoma Village is especially reliant upon well water, with 51% of the population using well water. Qualitative data reveals that this well is open and contaminated. Even more, enumerators and residents report that the water flows under trees where birds and monkeys drop waste. Lutengo Village has a high proportion of its population (53%) using borehole water. However, the water from the borehole comes out brown, leaving residents to believe the pipes have rusted. Figure 7: Types of water sources disaggregated by village
Well water 10% Borehole 18%
Households
3.3.2 Number of water sources As seen in Figure 8, 68% of households use exactly two sources of water; 27% use only one. There is a statistically significant relationship between economic status and access to more than one source of water.iv Households with higher economic status are 13% more likely to have access to two or more sources of water than households with lower economic status (See Figure 9). VHTs explain that wealthier families are more likely to buy catchment tanks to supplement local sources of water. Figure 8: Number of water sources households are able to access
Households
73% 68%
Overall
3.3.3 Water treatment As seen in Figure 10, 77% of households report treating their drinking water. There is a statistically significant difference in rates of water treatment between economic groups.v Figure 11 illustrates that households with higher economic status are 7% more likely to treat their drinking water than households with lower economic status. Of those who treat their water, 93% do so through boiling. Only 7% use a product like Aqua Safe or Water Guard. Figure 10: Percent of households and drinking water treatment
NO 23%
Yes 77%
77% Households
74%
All
10
3.4
3.4.1 Food source More than half of all households source food both through buying and gardening. This rate remains constant across economic status. However, for those with only one source of food, households with lower economic status are 13% more likely to only source food from their gardens. Conversely, households with higher economic status are 11% more likely to only source food through purchase (See Figure 12). vi
51% Households
50%
51%
18% 32%
Figure 13: Energy source for cooking and economic status 3.4.2 Energy source Firewood is the most common source of energy used for cooking, with 57% of the population relying solely upon firewood and another 25% using a combination of firewood and charcoal. Households with lower economic status are significantly more reliant on firewood with 71% using only firewood compared to 44% of households with higher economic status. vii
25%
21%
28%
Households
57%
71%
44%
18% All
11
3.5
Health Facilities
3.5.1 Type of health facility When asked to report the types of health facilities they use, 44% of households reported using government health centers, 40% reported using private health centers, 12% reported self-medicating and buying drugs from a shop, and 5% reported using herbal or traditional remedies. Households with lower economic status are 7% more likely to self-medicate. However, economic status does not affect use of other types of health facilities (See Chart 2, Appendix). Rather, the type of health facility a household accesses varies significantly by village. For example, Lutengo households reported a much higher rate of use of government health centers at 52%, while Mpoma, Lwanyonyo, Nsanvu, and Ttakajjunge range between 33% and 37%. Ttakajjunge has the highest use of private health centers, at 57%, while Mpoma has the highest use of self-medication at 21% (See Chart 3, Appendix). Data from the participatory assessment reveals that much of these discrepancies can be explained by proximity to health centers. Lutengo is very near a government health center III that is currently being upgraded to a health center IV. On the other hand, Ttakkajjunge is very near Good Samaritan, a private clinic. Mpoma is far from a health center and also has the lowest economic status, thus, residents self-medicate to circumvent consultation costs. Households in Lwanyonyi underuse government services because they have to cross the Metha sugar cane field to reach the nearest government health center. Many Lwanyonyi residents fear for their personal safety when passing through the fields and are thus reluctant to take the risk to visit the health center. Figure 14: Household use of varying types of health facilities disaggregated by village
Households
HH Overal using Mpom Lwany Nsanv l only 1 a onyi u facility 5% 12% 40% 44% 2% 2% 41% 56% 2% 21% 45% 33% 8% 17% 38% 37% 8% 11% 44% 36%
Wakis Luteng Ttakajj o o unge 5% 11% 31% 52% 2% 3% 27% 68% 2% 6% 57% 34%
Herbal or Traditional Healing Self Medication/ Shops Private Health Center Government Health Center
12
3.5.2 Diversity of health services Most households (55%) use only one type of health facility. Economic status does not affect the diversity of health facilities a household accesses. Households that only use one type of health service are more likely to use government centers and less likely to use self-medication or herbal/traditional remedies. (Compare Figures 14 & 15). ix
If respondents reported transparently, this data suggests that most households use selfmedication and traditional remedies only as supplementary forms of healthcare. However, through the participatory assessment, VHTs raised doubts about this data. They noted that respondents likely felt stigmatized and that use of these services is likely much higher than the survey data suggests.
3.5.3 Satisfaction with health service When asked how often they receive the services they need from a given health facility, 61% of households selected the option always, with the remainder selecting sometimes. Only one household reported that they never receive the services they need. Figure 16: Satisfaction with health center disaggregated by facility type and village
When you go to a health center, how often do you get the services you need?
100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Households
Sometimes Always
13
As seen in Figure 16, satisfaction with health services varies by facility type and village. More than 80% of households that only use private clinics reported that they always receive the services they need, compared to just 50% of those who use only government clinics.viii Over three quarters of all households that express dissatisfaction with government health services mentioned drug shortages as a major complaint. When comparing Figures 14 & 16 Mpoma, Lwanyonyo, Nsanvu, and Ttakajjunge use government facilities at a similar rate, yet their levels of satisfaction vary greatly. The participatory assessment suggests that the difference may be related to the timing of visits to the government health center, which could affect the stock of drugs.
3.6
Family Planning
3.6.1 Prevalence of family planning Just less than 40% of all households report using a method of family planning. As seen in Figure 17, this rate varies significantly with economic status. Households with lower economic status are 23% less likely to report using family planning than households with higher economic status.ix The participatory assessment points out that those with higher economic status are more likely to have jobs in town where family planning information is more available and workplace sensitization programs are more common. Figure 17: Family planning use disaggregated by economic status
39% Households
27%
All
14
3.6.2 Family planning methods Of households that practice family planning, the Depo-Provera injection is by far the most popular, with 55% of the market. Qualitative data indicates that this is because the Depo-Provera injection is the most convenient. Pills and condoms are the next most popular at 17% and 16%, respectively. Rhythm method, Norplant, and IUDs cover only a small portion of the population (See Figure 18). The participatory assessment suggests that people perceive norplants and IUDs as more invasive and are more fearful of their side-effects.
Norplant 3%
Condoms 16%
IUD 4%
DepoProvera 55%
As seen in Figure 19, family planning methods vary greatly with economic status. Households with higher economic status are 25% more likely to use the Depo-Provera injection than poorer households. Conversely, households with lower economic status rely more heavily on pills and condoms. Type of health service is less of an influence on family planning method as methods vary only slightly from households that use only government health services to households that use only private services (See Chart 4, Appendix). Figure 19: FP method disaggregated by economic status and health service type
Households
55%
38%
63%
44%
45%
16% All
15
Of households that already practice a method of family planning, 69% say that they are interested in additional services. Of households that are not practicing family planning, 45% say they are interested in receiving family planning services. Many of these individuals asked about specific family planning methods they would like to access. Specifically, fifteen individuals asked to know more about the norplant and another ten asked to learn more about the injection plan. Others reported wanting to receive education that compared the different family planning options available. Figure 20: Interest in additional family planning services
Households
45%
3.7
Malaria
3.7.1 Malaria incidence According to the survey, 1.81 people per household or 32.5% of all household members, suffered from malaria in the last two months. Mpoma has the highest reported rate of malaria at 37%, while Lutengo has the lowest at 26.6% (See Figure 21). However, variation by village and the 4% difference between higher economic status and lower economic status are not statistically significant (Chart 5, Appendix). Therefore based on survey data alone it is not possible to make reliable comparisons between groups.
16
However the participatory assessment suggests that differential access to health facilities may contribute to variation in malaria rates by village. VHTs explain that when patients visit government health facilities they are often only given a portion of the malaria treatment and told to return later to receive the remainder of the supply. Patients from distant villages are less likely to return for the balance of the treatment, increasing the likelihood and severity of malaria. Figure 21: Malaria rates disaggregated by economic status and village
3.7.2 Prevalence of sleeping under nets As illustrated in Figure 21, households report that 57% of individuals always sleep under a bed net. The rate is higher for individuals who are pregnant (63%) and under five years of age (65%). VHTs attribute this to government programs that distributed free bed nets to these two vulnerable groups. In the qualitative portion of the survey 11% of people who dont use bed nets explained the reason was because the government didnt provide free nets to their demographic. Similarly, another 71% reported that cost was the limiting factor in using a bed net. There is not a statistically significant difference in the rate of always sleeping under a bed net by economic status. x This suggests that economic access is not the only limiting factor to net use. For example, data from the participatory assessment points out that some pregnant women are reluctant
17
to use nets because of the strong smell they give off. Through the household survey, 18% of individuals not using nets reported one of the limiting factors below: nets are too hot (4%) prefer slashing bush, closing doors early, or boiling water (4%) nets spoil easily (3%) lack of space (2%) nets smell bad (1%) have heard stories about negative side-effects (1%) Also, the participatory assessment revealed that many people could not tell the difference between treated and untreated nets. Therefore, it is not possible to distinguish how many individuals represented in Figure 21 are sleeping under treated nets. Figure 21: Bed net use disaggregated by age group, economic status, and village
18
Bed net use also varies by village. Figure 22 shows that villages with a larger percentage of households in the higher economic group also have larger rates of bed net use. When coupled with the aforementioned weak relationship between household-level economic status and bed net use, this data suggests that bed net use is motivated by community-wide socio-economic factors, rather than household economic status. This hypothesis is supported by the participatory assessment, in which members explained that as a critical mass of village residents use and value nets, households, regardless of resources, are more likely to seek out and value nets. Figure 22: Bed net use and economic status by village
Individuals
48.1% 46%
50%
53%
35%
% of individuals "always" sleeping under a bed net % of households in higher economic group
3.8
HIV/AIDS
3.8.1 HIV status awareness Only one-third of the overall population is aware of their HIV status. The rate of awareness varies significantly by village. As seen in Figure 23, there is a 49-percentage point gap between Ttakajjunge and Wakiso (See Chart 5, Appendix).
19
Participatory assessment data suggests that theses differences are connected to access to free testing services. Good Samaritans free weekly testing likely contributes to Takajjunges high awareness rates. Nnsanvu residents report that an NGO (though they cant remember the name) went house to house offering free testing. Also, Lutengo hosted a short-term HIV testing outreach program that was open to surrounding villages. Wakisos low awareness rate may be connected to the fact that it is distant from clinic-based free services and from outreach programs like the one held in Lutengo. Figure 23: HIV status awareness disaggregated by economic status and village
3.8.2 HIV prevalence Of the 1200 individuals covered by the household survey only 18 were reported to be HIV positive. This number is extremely low compared to the 6.5% regional HIV prevalence rates released in the 2009 UNGASS Country Progress Report.xi This suggests that Mpomas household survey did not accurately capture HIV rates. This is not surprising considering the stigma and vulnerability connected to HIV/AIDS. A much more methodologically targeted study is needed to accurately assess HIV/AIDS prevalence.
4.0
Conclusion
Building from this assessment, Mpoma Community HIV/AIDS Initiative looks forward to collaborating with local government, partner organizations, and outside funders to implement health and social programs that address the specific need of our community. We welcome our stakeholders to use and share this data for the promotion, education and general advancement of the Nama subcounty. 20
Johnson, Burke and Anthony J. Omwuegbuzie (2004) Mixed Methods Research: A Research Paradigm Whose Time Has Come, Educational Researcher Vol 33 No. 7, 14-26
ii
Stake, Robert E. (1995). The Art of Case Study Research. Thousand Oaks, CA: Sage Publications.
University of Wisconsin, Cooperative Extension (1996). Program Development and Evaluation, Collecting Evaluation Data: Direct Observation. http://learningstore.uwex.edu/pdf/G3658-5.pdf Colorado State University. Writing Guide: Ethnography, Observational Research, and Narrative Inquiry. http://writing.colostate.edu/guides/research/observe/
iii
Morra, L. G., & Rist, R. C. (2009). The Road to Results: Designing and Conducting Effective Development Evaluations. Washington, D.C: World Bank. Chapter 8.
iv
This is a statistically significant difference based on a two-tailed t-test comparing the mean number of households with access to more than one water source in the lower economic group to the mean in the higher economic group. Analysis used a .05 significance level, with a p-value of .0209.
v
This is a statistically significant difference based on a two-tailed t-test comparing the mean number of households that treat their water in the lower economic group to the mean that does so in the higher economic group. Analysis used a .05 significance level, with a t-value of 1.9899 and a p-value of .0478
vi
This is a statistically significant difference based on a two-tailed t-test comparing the mean number of households that rely on gardening as a food source in the lower economic group to the mean that do so in the higher economic group. Analysis used a .05 significance level, with a t-value of 3.1467 and a p-value of less than .0021
vii
This is a statistically significant difference based on a two-tailed t-test comparing the mean number of households that rely on firewood in the lower economic group to the mean that does so in the higher economic group. Analysis used a .05 significance level, with a t-value of 4.5821 and a p-value of less than .0001
viii
This is a statistically significant difference based on a two-tailed t-test comparing the mean number households satisfied with government facilities to the mean satisfied with private facilities. Analysis used at a .05 significance level, with a t-value of 4.5941 and a p-value of less than .0001
ix
This is a statistically significant difference based on a two-tailed t-test comparing the mean number of households that practice family planning in the lower economic group to the mean that does so in the higher economic group. Analysis used a .05 significance level, with a t-value of 3.7362 and a p-value of less than .0002
x
This is a statistically significant difference based on a two-tailed t-test comparing the mean number of households that reported always sleep under a bed net in the lower economic group to the mean that reported this in the higher economic group. Based on a two-tailed t-test at a .05 significance level, with a t-value of 1.2965 and a pvalue of 0.1966
xi
21
5.0
Appendix
Difference in water source by village
Mpoma Lutengo 47% 53%*** 0%*** 0% 1%*** Lwanyonyi 47% 2%** 1.5%** 2% 48%** Nsanvu 39% 20% 2% 0% 39% Ttakajjunge 50% 4%*** 0%*** 0% 46.4%* Wakiso 51% 20% 0%*** 0% 29% All 41% 18% 10% 1% 30%
Chart 1:
* p-value less than .05 ** p-value less than .005 *** p-value less than .0001 Notes: Based on two-tailed t-tests comparing village means in each category to the average mean (Re: Figure 7)
Chart 2:
Government Health Center Private Health Center Self Medication/ Shops Herbal or Traditional Healing
* p-value less than .05 ** p-value less than .005 *** p-value less than .0001 Notes: Based on two-tailed t-tests comparing usage means of the lower economic group to the higher economic group for each type facility
Chart 3:
Overall Government Health Center Private Health Center Self Medication/ Shops Herbal or Traditional Healing
Mpoma
Lwanyonyi
Nsanvu
Wakiso
Lutengo
Ttakajjunge
44%
33%
37%
36%
52%
68%***
34%***
40%
41%
45%
38%
44%
31%
27%
57%***
12%
2%*
21%
17%
11%
11%
3%
6%
5%
2%*
2%
8%
8%
5%
2%
2%
Notes: Based on two-tailed t-tests comparing village means in each category to the average mean (Re: Figure 14)
Chart 4:
* p-value less than .05 ** p-value less than .005 *** p-value less than .0001 Notes: Based on two-tailed t-tests comparing usage means of the lower economic group to the higher economic group and only government to only private health services (Re: Figure 19)
Chart 5:
Notes: Based on two-tailed t-tests comparing malaria rates of the lower economic group to the higher economic group and then each village malaria rate by the average malaria rate (Re: Figure 21)
Chart 5:
Mpoma
Nsanvu
Wakiso
Lutengo
31%
35%
47%
30%
43%
59%**
20%*
45%*
69%***
* p-value less than .05 ** p-value less than .005 *** p-value less than .0001 Notes: Based on two-tailed t-tests comparing HIV awareness rates of the lower economic group to the higher economic group and each village awareness rates by the average awareness rate (Re: Figure 21)