**2. Context**

Uganda had an estimated population of 33.8 million people as of mid 2010 with the majority (87%) living in rural areas and nearly half (49%) being below 15 years of age (Population Reference Bureau, 2010). The first AIDS case was diagnosed in the country in 1982 and by 1986, it had reached the level of a generalized epidemic with the predominant mode of transmission being heterosexual contact (Ministry of Health & ORC Macro, 2006; Serwadda et al., 1985). The sentinel surveillance system from which national HIV prevalence estimates were initially derived was established in 1989 starting with six clinics in urban areas but later expanded to include clinics from peri-urban and rural areas (Garbus & Marseille, 2003). HIV prevalence was estimated at 15% among all adults (15-49 years) in 1991-1992 which was considered to be the peak of the epidemic in the country (Kirungi et al. 2006; Stoneburner & Low-Beer, 2004). This, however, declined to 7% in 2001 and was still at this level in 2009 (Joint United Nations Programme on HIV/AIDS [UNAIDS], 2010; Kirungi et al. 2006; Stoneburner & Low-Beer, 2004). As in other sub-Saharan Africa countries affected by the epidemic, HIV prevalence is higher among women than among men and in the urban compared to rural areas. In 2004-2005, for example, 8% of women aged 15-49 years were HIV-positive compared to 5% of men of similar age groups while prevalence was nearly twice as high in the urban compared to rural areas (10% versus 6%; Ministry of Health & ORC Macro, 2006). Similarly in 2009, prevalence among young people aged 15-24 years was 5% for women and 2% for men (UNAIDS, 2010).

The first AIDS control program was set up in the country in 1987 with emphasis on abstinence, being faithful to one partner, and condom use (Ministry of Health & ORC Macro, 2006). This prevention strategy—referred to as ABC—has since been expanded to include voluntary counselling and testing (VCT), prevention of mother-to-child transmission (PMTCT) of the virus, antiretroviral treatment (ART), and HIV care and support services (Ministry of Health & ORC Macro, 2006). The Ministry of Health started a voluntary door-to-door HIV screening programme in 1999 and has also begun implementing provider-initiated testing and counselling (PITC) at health facilities (Menzies et al., 2009; Wanyeze et al., 2008). By 2009, HIV testing and counselling services were available in 1,215 health facilities representing an increase from 812 facilities in 2008 and 554 in 2007 (World Health Organization [WHO] et al., 2009, 2010). Free ART has been available since 2004 and by 2009, 39% of adults in need of treatment were receiving it (WHO et al., 2010). The Ministry of Health began offering free PMTCT services in 2000; the proportion of HIV-positive mothers receiving PMTCT services increased from 12% in 2005 to 53% in 2009 (WHO et al., 2009, 2010). However, challenges still remain with respect to reaching all those in need of treatment due to limited human resource capacity to provide the services, and lack of efficient management of funds and supplies (Onyango and Magoni, 2002).

With respect to education, Uganda implemented the Universal Primary Education (UPE) programme in 1997 which removed fees for enrolment in primary schools and resulted in substantial increases in enrolments (Deininger, 2003; Murphy, 2003; Nishimura et al., 2008; UBOS and Macro International, 2007). This was followed ten years later (in 2007) with the introduction of Universal Secondary Education (USE) whose impact is yet to be systematically evaluated (Chapman et al., 2010). Estimates of HIV prevalence among members of the school community (students and teachers) are unavailable. However, realizing the challenges posed by HIV to the education sector, the Ministry of Education and Sports issued the *Education and Sports Sector National Policy Guidelines on HIV/AIDS* in 2006 to provide a framework for responding to the epidemic within the sector. The objectives of the policy are to: (1) raise knowledge of students, education managers and other sector employees on HIV/AIDS; (2) ensure that students, education managers, and educators access prevention, treatment, care and support services; (3) eliminate all forms of stigma and discrimination; (4) mitigate the impact of HIV/AIDS that impede access to and provision of quality education; (5) strengthen the education sector's capacity to effectively respond to HIV/AIDS; and (6) contribute to the knowledge base on HIV/AIDS through research (Ministry of Education and Sports- Uganda, 2006).

#### **3. Data**

74 Social and Psychological Aspects of HIV/AIDS and Their Ramifications

ensuring an inclusive education system that adequately responds to the challenges in-school HIV-positive young people may face. This chapter therefore responds to the need for relevant evidence by exploring the experiences of HIV-positive adolescent boys and girls in primary and secondary schools in Uganda from the perspectives of school officials and teachers, the general student body, as well as adolescents perinatally infected with HIV. It specifically focuses on: (1) school attendance and experiences with class repetition; (2) experiences of stigma and discrimination within the school environment; and (3) availability of school-based health and psychosocial support programs and services for HIV-positive students. It ends with a discussion of the implications of these experiences for addressing the needs of in-school HIV-positive young people by the education sector not only in

Uganda had an estimated population of 33.8 million people as of mid 2010 with the majority (87%) living in rural areas and nearly half (49%) being below 15 years of age (Population Reference Bureau, 2010). The first AIDS case was diagnosed in the country in 1982 and by 1986, it had reached the level of a generalized epidemic with the predominant mode of transmission being heterosexual contact (Ministry of Health & ORC Macro, 2006; Serwadda et al., 1985). The sentinel surveillance system from which national HIV prevalence estimates were initially derived was established in 1989 starting with six clinics in urban areas but later expanded to include clinics from peri-urban and rural areas (Garbus & Marseille, 2003). HIV prevalence was estimated at 15% among all adults (15-49 years) in 1991-1992 which was considered to be the peak of the epidemic in the country (Kirungi et al. 2006; Stoneburner & Low-Beer, 2004). This, however, declined to 7% in 2001 and was still at this level in 2009 (Joint United Nations Programme on HIV/AIDS [UNAIDS], 2010; Kirungi et al. 2006; Stoneburner & Low-Beer, 2004). As in other sub-Saharan Africa countries affected by the epidemic, HIV prevalence is higher among women than among men and in the urban compared to rural areas. In 2004-2005, for example, 8% of women aged 15-49 years were HIV-positive compared to 5% of men of similar age groups while prevalence was nearly twice as high in the urban compared to rural areas (10% versus 6%; Ministry of Health & ORC Macro, 2006). Similarly in 2009, prevalence among young people aged 15-24 years was

The first AIDS control program was set up in the country in 1987 with emphasis on abstinence, being faithful to one partner, and condom use (Ministry of Health & ORC Macro, 2006). This prevention strategy—referred to as ABC—has since been expanded to include voluntary counselling and testing (VCT), prevention of mother-to-child transmission (PMTCT) of the virus, antiretroviral treatment (ART), and HIV care and support services (Ministry of Health & ORC Macro, 2006). The Ministry of Health started a voluntary door-to-door HIV screening programme in 1999 and has also begun implementing provider-initiated testing and counselling (PITC) at health facilities (Menzies et al., 2009; Wanyeze et al., 2008). By 2009, HIV testing and counselling services were available in 1,215 health facilities representing an increase from 812 facilities in 2008 and 554 in 2007 (World Health Organization [WHO] et al., 2009, 2010). Free ART has been available since 2004 and by 2009, 39% of adults in need of treatment were receiving it (WHO et al., 2010). The Ministry of Health began offering free PMTCT services in 2000; the proportion of HIV-positive mothers receiving PMTCT services increased from 12% in 2005 to 53% in 2009

Uganda but in SSA countries affected by the pandemic.

5% for women and 2% for men (UNAIDS, 2010).

**2. Context** 

The data are from a study conducted in Uganda in 2009 whose objective was to understand the needs of in-school HIV-positive young people. The study involved two major components. The first component was a survey among 718 young people aged 12-19 years (of school-going age) who were perinatally infected with HIV (had been living with the virus since infancy) and who knew their sero-status. The sample members were identified and recruited through existing HIV/AIDS treatment, care and support programs/centres selected by The AIDS Support Organization (TASO)-Uganda in four districts (Kampala, Wakiso, Masaka and Jinja). Thirteen such centres participated in the study. All adolescents who received services from the centres and satisfied the eligibility criteria in terms age, perinatal infection, and awareness of sero-status were targeted for inclusion in the study. TASO counsellors assisted with the identification and mobilization of the eligible respondents. The process involved obtaining clearance from the management of the centres, identifying the target sample from the existing records, and making calls to their parents to request them to come to the centres or targeting days when they visit the centres for routine reviews or drug re-fills.

Information was collected using a structured questionnaire which was translated into *Luganda* and *Lusoga*, the two dominant local languages. Interviews were partially completed with 6 of the participants. Information was gathered on the respondents' background characteristics, educational attainment, school attendance (absenteeism, repetition, changing of schools, and drop-out), motivations for being in school or dropping out, disclosure of HIV status within the school environment and the reactions of others to the disclosure, availability of support programs for HIV-positive young people within schools, psychosocial feelings in school and whether these affected school attendance, and experiences of physical or verbal abuse, discrimination and stigma in school. For adolescents aged 12-17 years, written consent to

HIV Infection and Schooling Experiences of Adolescents in Uganda 77

The second component of the study involved in-depth interviews with school officials to assess the school environment and their preparedness to support in-school HIV-positive young people. A total of eight schools (four primary and four secondary) in five districts (Kampala, Jinja, Wakiso, Mukono and Iganga) were included in the assessment. Two of the primary and two of the secondary schools were mixed day while one school in each category was a boys' only and the other a girls' only boarding institution. The schools were purposively selected in consultation with the Ministry of Education and Sports. The Ministry granted the research team permission to visit the schools and talk to the officials. The research team obtained oral consent for participation in the study from the school officials. Information was collected through in-depth interviews to determine the operationalization of the HIV/AIDS policies in schools, perceptions and practices of teachers and school management towards in-school HIV-positive young people, the existence of support programs, and possible responses by the education sector to the needs of infected students. A total of 52 in-depth interviews were conducted with head teachers (7), deputy head teachers (4), director of studies (1), deans of students (2), senior teachers (12), Presidential Initiative on AIDS Strategy for Communication to Youth (PIASCY) teachers (3), school nurses/clinical officers (8), school matron (1), peer counsellor (1), health

In addition, a total 1,012 students in Senior Three and Five from the four secondary schools wrote essays on specified themes. These included the perceived and actual attitudes and practices of students towards peers who are HIV-positive as well as possible responses by fellow students and the school administration to the needs of HIV-positive students. The essays were anonymous-- students were asked to indicate only their age, sex, and class but not their names-- and were administered to the students as a class exercise. It was explained to them that the exercise was voluntary and that they had the freedom not to participate in it if they did not wish to. Twenty nine (3%) of these essays were, however, discarded because it was apparent the students did not understand the nature of information required. Participants whose essays were analyzed were aged between 11 and 25 years, 71% of them were females (1% of the students did not indicate their sex), and about two-thirds (63%)

The first part of analysis involves cross-tabulations with Chi-square tests and significance tests of proportions to examine differences in schooling, experiences of stigma and discrimination, and availability of school-based support programs and services by various background characteristics of the respondents including age, sex, district, and whether the respondent lived with a biological parent. In the second part of the analysis, random-effects logistic regession models are estimated to predict the likelihood of HIV-positive adolescents being in school at the time of the survey, experiencing stigma and discrimination in school, and receiving any form of support from school. The choice of the analysis technique is guided by the need to account for unobserved characteristics of individuals identified from the same HIV/AIDS treatment, care and support program/centre. The model is of the form:

log ( ) *ij* = + *ij j it X*

where *πij* is the probability of a given outcome for individual *i* identified from facility *j*; *Xij* is

 βμ (1)

μ*j* is the

π

the vector of covariates; *β* is the associated vector of parameter estimates; and

prefects (8), and club patrons and members (5) from the selected schools.

were in Senior Three.

**4. Analysis** 

participate in the study was sought from parents/guardians and assent was sought from the adolescents themselves. However, for adolescents aged 18-19 years, the study obtained individual written consent only. The TASO Internal Review Board, the Uganda National Council for Science and Technology (UNCST), and the Population Council's Institutional Review Board granted ethical clearance for the study.

Female respondents comprised more than half (59%) of the survey participants. There was, however, no significant difference in the distribution by sex according to most of the background characteristics such as age, district of residence, whether the respondent lived with a biological parent, and the living arrangements of the biological parents (Table 1). The majority (65%) of the respondents were aged below 18 years, hence considered minors while slightly more than one-third (37%) were from Kampala district. Besides, 80% of the respondents reported that one or both parents had died, which suggests that the majority of them might lack proper support not only in school but also at home.


*Notes:* Percentages may not add up to exactly 100 in some cases due to rounding; Differences between male and female proportions are statistically significant at: \*\*p<0.01; \* p<0.05.

Table 1. Percent distribution of survey participants by various background characteristics

The second component of the study involved in-depth interviews with school officials to assess the school environment and their preparedness to support in-school HIV-positive young people. A total of eight schools (four primary and four secondary) in five districts (Kampala, Jinja, Wakiso, Mukono and Iganga) were included in the assessment. Two of the primary and two of the secondary schools were mixed day while one school in each category was a boys' only and the other a girls' only boarding institution. The schools were purposively selected in consultation with the Ministry of Education and Sports. The Ministry granted the research team permission to visit the schools and talk to the officials. The research team obtained oral consent for participation in the study from the school officials. Information was collected through in-depth interviews to determine the operationalization of the HIV/AIDS policies in schools, perceptions and practices of teachers and school management towards in-school HIV-positive young people, the existence of support programs, and possible responses by the education sector to the needs of infected students. A total of 52 in-depth interviews were conducted with head teachers (7), deputy head teachers (4), director of studies (1), deans of students (2), senior teachers (12), Presidential Initiative on AIDS Strategy for Communication to Youth (PIASCY) teachers (3), school nurses/clinical officers (8), school matron (1), peer counsellor (1), health prefects (8), and club patrons and members (5) from the selected schools.

In addition, a total 1,012 students in Senior Three and Five from the four secondary schools wrote essays on specified themes. These included the perceived and actual attitudes and practices of students towards peers who are HIV-positive as well as possible responses by fellow students and the school administration to the needs of HIV-positive students. The essays were anonymous-- students were asked to indicate only their age, sex, and class but not their names-- and were administered to the students as a class exercise. It was explained to them that the exercise was voluntary and that they had the freedom not to participate in it if they did not wish to. Twenty nine (3%) of these essays were, however, discarded because it was apparent the students did not understand the nature of information required. Participants whose essays were analyzed were aged between 11 and 25 years, 71% of them were females (1% of the students did not indicate their sex), and about two-thirds (63%) were in Senior Three.

#### **4. Analysis**

76 Social and Psychological Aspects of HIV/AIDS and Their Ramifications

participate in the study was sought from parents/guardians and assent was sought from the adolescents themselves. However, for adolescents aged 18-19 years, the study obtained individual written consent only. The TASO Internal Review Board, the Uganda National Council for Science and Technology (UNCST), and the Population Council's Institutional

Female respondents comprised more than half (59%) of the survey participants. There was, however, no significant difference in the distribution by sex according to most of the background characteristics such as age, district of residence, whether the respondent lived with a biological parent, and the living arrangements of the biological parents (Table 1). The majority (65%) of the respondents were aged below 18 years, hence considered minors while slightly more than one-third (37%) were from Kampala district. Besides, 80% of the respondents reported that one or both parents had died, which suggests that the majority of

 12-14 31.0 38.4 35.4 15-17 31.0 28.1 29.3 18-19 37.4 33.3 35.0 Don't know 0.7 0.2\* 0.4

 Jinja 21.1 27.6\* 24.9 Kampala 40.1 34.0 36.5 Wakiso 7.1 5.2 6.0 Masaka 31.6 33.3 32.6

 Yes 39.8 39.4 39.6 No 58.8 58.0 58.4 Missing/no answer 1.4 2.6 2.1

 Married/living together 11.2 10.9 11.0 Divorced/separated 3.7 6.6 5.4 Mother dead 16.3 10.9 13.1 Father dead 27.2 24.3 25.5 Both parents dead 39.1 44.1 42.1 Don't know/missing 2.4 3.3 2.9 *Notes:* Percentages may not add up to exactly 100 in some cases due to rounding; Differences between

Table 1. Percent distribution of survey participants by various background characteristics

p<0.05.

male and female proportions are statistically significant at: \*\*p<0.01; \*

Male (N=294) %

Female (N=424) %

Both sexes (N=718) %

Review Board granted ethical clearance for the study.

Characteristics

Age group

District

Lives with a biological parent

Parents' living arrangements

them might lack proper support not only in school but also at home.

The first part of analysis involves cross-tabulations with Chi-square tests and significance tests of proportions to examine differences in schooling, experiences of stigma and discrimination, and availability of school-based support programs and services by various background characteristics of the respondents including age, sex, district, and whether the respondent lived with a biological parent. In the second part of the analysis, random-effects logistic regession models are estimated to predict the likelihood of HIV-positive adolescents being in school at the time of the survey, experiencing stigma and discrimination in school, and receiving any form of support from school. The choice of the analysis technique is guided by the need to account for unobserved characteristics of individuals identified from the same HIV/AIDS treatment, care and support program/centre. The model is of the form:

$$\log it(\pi\_{ij}) \quad = \quad X\_{ij}\boldsymbol{\beta} + \boldsymbol{\mu}\_{j} \tag{1}$$

where *πij* is the probability of a given outcome for individual *i* identified from facility *j*; *Xij* is the vector of covariates; *β* is the associated vector of parameter estimates; and μ*j* is the

HIV Infection and Schooling Experiences of Adolescents in Uganda 79

Slightly more than half (52%) of the respondents attending school at the time of the survey missed going to school the previous term (Table 3) with no significant variations by age, sex or whether the respondent lived with a biological parent. Nonetheless, the proportion that missed school the previous term was significantly lower in Jinja compared to other districts. Further analysis shows that illness was the major reason for missing school (cited by 57% of those who missed school) followed by lack of school fees or education materials (23%), and going for treatment/ medication (12%). There were also significant variations in the major reasons for missing school by sex and district of residence (p<0.05 in each case). For instance, 60% of female respondents cited illness as the major reason for missing school the previous term compared to 51% of male respondents. In contrast, slightly more than twice as many male as female respondents cited going for treatment/medication as the major reason for missing school (18% versus 8%). Similarly, the proportion mentioning illness was about 10 percentage points higher in Jinja and Masaka (61% in each case) compared to Kampala and Wakiso districts (52% and 50% respectively) while the proportion citing treatment/medication

was nine times higher in Kampala compared to Jinja district (19% versus 2%).

Missed school previous term (%)

Age group p=0.24 p=0.42 p=0.46 12-14 55.3 56.1 96.2 15-17 48.5 59.1 93.6 18-19 50.3 50.3 95.4 Sex p=0.32 p=0.66 p=0.39 Male 49.2 56.3 94.1 Female 53.3 54.5 95.7 District p<0.01 p<0.01 p<0.01 Jinja 37.2 64.1 98.6 Kampala 58.7 45.8 95.6 Wakiso 52.9 52.9 79.4 Masaka 54.1 60.2 94.5 Lives with a biological parent p=0.75 p=0.13 p=0.18 Yes 52.4 51.6 96.4 No 51.1 58.0 94.0 All respondents 51.6 55.2 95.2 Number of respondents 585 585 585

Table 3. Percentage of survey participants who missed school the previous term, percentage that ever repeated a class and percentage that considered schooling very important by

More than half (55%) of the respondents who were still in school at the time of the survey reported ever repeating a class (Table 3). The proportion having repeated a class did not

Ever repeated a class (%)

Considers schooling very important (%)

Background characteristics

Notes: p-values are from Chi-square tests.

background characteristics

disturbance term due to unmeasured characteristics that may also affect the outcome for individuals identified from facility *j*.

The first dependent variable is measured by whether the respondent was still in school at the time of the survey conditional on having ever attended school. Stigma and discrimination, on the other hand, refer to whether those attending school had ever been teased, called nasty names, or suspected that rumours were spreading about them because of their HIV status. The third dependent variable is measured by whether the respondent received any support from groups, clubs or the school. The models control for age (single years), sex (male or female), district of residence (Kampala, Jinja, Masaka, Wakiso), and whether the respondent lived with a biological parent (yes or no).
