Rationale: There are many public health benefits of targeting adolescent for immunisation. However, and in many settings, adolescents do not get optimal benefits from immunisation. In the decade of vaccines (2011-2020), adolescent immunisation is a topical subject. An up-to-date and synthesized research on adolescent immunisation is lacking. Overall purpose: The purpose of the PhD thesis was to characterize adolescent immunisation in the decade of vaccines. Research methods: First, we conducted a comprehensive narrative review of the literature (chapter 2) on adolescent immunisation. Then, we conducted systematic reviews (chapters 3 and 4). One of the systematic reviews assessed the strategies to improve uptake of vaccines among adolescents. The other systematic review assessed the knowledge, attitudes and practices of adolescents and their parents and teachers towards immunisation. Finally (chapter 5), we conducted a cross-sectional study to describe the challenges experienced, and lessons learnt during the introduction of national human papillomavirus (HPV) vaccination programmes in Africa. Findings: Adolescents are an important group to target with primary, booster or catch up immunisation. Some global initiatives have advocated for adolescent immunisation. Multiple reasons, among them, lack of knowledge and access to immunisation services are barriers to adolescent immunisation. There exist multiple strategies to improve uptake of vaccines among adolescents. For example, health education, financial incentives, mandatory vaccination, and class-based school vaccine delivery. The evidence suggests that a combination of strategies may be more effective than one strategy alone in enhancing uptake of vaccines by adolescents. Knowledge of vaccines, immunisation and vaccine preventable diseases was found to be suboptimal among key stakeholders of adolescent immunisation in Africa. We found a disconnect between the level of knowledge on immunisation and the uptake of vaccines, an interesting finding that warrants further research in Africa. Six African countries shared the lessons learnt and experiences during the national introduction of HPV vaccination programmes that targeted adolescent girls. There were similarities in the results among the participating countries. The challenges included: logistical coordination, identification of the target population, obtaining political support, integration with other school programmes and stakeholder engagement. A lesson learnt was that schools are a convenient site to access and vaccinate adolescents. Conclusion: Adolescent immunisation is not routinely practiced in many countries. The introduction of HPV vaccines has created an ideal opportunity to build platforms for adolescent immunisation. Research on adolescent immunisation is limited, more so in low and middle-income countries. Existing research shows a combination of strategies can be used to enhance uptake of vaccines among adolescents. Strong advocacy programmes are required to drive the global agenda of adolescent immunisation, particularly in Africa.

Background: The objective of this study is to explore the determinants of decision space and to investigate how it functions within a central hospital in South Africa, where policy proposals regarding the decentralization of management authority are being considered in the context of concerns about poor hospital management, and in terms of planned implementation of the National Health Insurance policy. Methods: Using a mixed methods qualitative approach, the case study methodology involved the purposive selection of two central hospitals as case study sites in the South African province of Gauteng. Nested case studies that focused on the human resources and supply chain management departments involved exploratory and explanatory phases that included document review, work shadow with non-participant observation, and in-depth interviews. Thematic analysis was used to analyse the data and generate findings. Results: Emergent outcomes of the interactions between authority, organisational capacity and accountability were identified and contributed to the development of an analytical decision space framework. The framework outlines the way in which decision space influences the allocation of responsibilities, the responsiveness of decision-making and the use of resources within the organisation or system. Recommendations: In the context of national policy changes regarding the decentralisation of hospital management authorities, this framework can be used to understand how decision space functions in other settings, and to guide implementation of interventions for improved management.

Includes bibliographical references (p. 229-246). / The aim of this research was to provide an understanding of medical students' attitudes towards and perceptions of the PHC approach, and this was done using mainly qualitative methods, namely focus groups, interviews, and one questionnaire. This research also investigated students' views of the way in which the PHC approach was taught, their understanding of the PHC approach, what could influence students' views of the PHC approach, the appropriateness of the PHC approach in South Africa, their opinions of the fact that UCT has a PHC-driven MBChB curriculum, their views of the role of doctors in the PHC approach, and a number of other related issues. The main findings were that students enter their medical degree with an expectation of a biomedical emphasis and a lecture-based curriculum.

Includes bibliographical references. / This thesis plots a participatory action research (PAR) study that was done in collaboration with disabled women in Khayelitsha and Brown's Farm on the outskirts of Cape Town, South Africa, over a two and a half year period. The aim of the study was to explore how disabled women living in poor communities equalise opportunities for human development and social change. The PAR approach was used to encourage the participation of disabled women to produce new knowledge and consciousness-raising related to the barriers faced and strategies used for their development since acquiring their impairments. The research partners were the SACLA Health Project, the Disabled Women's Development Project of Disabled People South Africa and the Division of Occupational Therapy, University of Cape Town.

This thesis uses experience gained from a large implementation trial in two rural districts of the Western Cape, South Africa, to address the needs of patients with non-communicable diseases (NCDs) and depression, and to identify solutions to those needs. The Primary Care 101 intervention supports and expands nurses' role in integrated care, in particular for NCDs. It comprises a comprehensive clinical management tool implemented in primary care services using educational outreach training. It was evaluated using a pragmatic cluster randomised controlled trial: 38 clinics in the Eden and Overberg districts of the Western Cape were randomised to receive the intervention or to continue with usual care. 4393 Patients were enrolled and four cohorts identified: hypertension, diabetes, chronic respiratory disease and depression. Patients were re-interviewed once, 14 months later. Primary outcomes for the trial were treatment intensification for the hypertension, diabetes and chronic respiratory disease cohorts, and case detection for the depression cohort. Multimorbidity, NCD care and their socioeconomic associations were assessed on the whole trial cohort (combining intervention and control arms) at baseline and follow-up. The results are presented in published papers. Baseline data revealed considerable multimorbidity and unmet treatment needs (Paper 1). Socioeconomic indicators such as education, and modifiable clinic-level factors such as adequate staffing and communitybased chronic medication collection services were associated with blood pressure control (Paper 2) and depression management (Paper 3). The intervention was shown to be feasible and safe but none of the four primary outcomes showed significant improvement (Paper 4). The thesis addresses the public health challenge of providing integrated chronic disease primary care in South Africa by: • Providing original evidence for high levels of NCD multimorbidity and unmet treatment needs. • Identifying modifiable factors that could improve care for these diseases. • Providing new evidence from South Africa to support the bidirectional relationship between poverty and depression. • Reporting evidence of the effectiveness of a novel intervention aimed at improving NCD care. The findings point to the need for improved strategies for NCD care, including equipping primary health care providers to manage the complexities of multimorbidity.

Access to safe, legal abortion services is an important public health measure to address morbidity and mortality from unsafe abortion. To expand access and strengthen medical abortion provision in South Africa, evidence is needed on the safety, effectiveness, feasibility and acceptability of task sharing strategies and the implementation of evidence-based regimens. This research aims to: (a) evaluate the safety and acceptability of task sharing gestational age estimation for women seeking abortion, (b) determine the effectiveness and acceptability of text messaging on mobile phones to support women self-managing medical abortion, (c) evaluate the feasibility, safety and acceptability of self-assessment of medical abortion completion using mobile phones alone or in combination with a low-sensitivity pregnancy test, and (d) document clinical outcomes and women's experiences following the introduction of mifepristone into second trimester medical abortion services. Published or submitted papers included in this thesis are from four prospective studies evaluating interventions and interviewing women and health care workers in South African public sector and non-governmental clinics between 2011 and 2015. The first paper establishes that last menstrual period is sufficiently accurate to estimate gestational age in selected women (97%) and has potential to be task shared with community health workers or women themselves. The second paper reports reduced anxiety (p=0.013) and better preparedness (p=0.016) for self-managing abortion symptoms among women receiving automated text messages (compared to those receiving standard care). The third and fourth papers show that mobile phones are a feasible modality for self-assessment for most women (86%), but that clinical history needs to be combined with an appropriate pregnancy test to detect incomplete or failed procedures. Self-assessment using a low-sensitivity pregnancy test is preferred by most women (98%) to in-clinic follow-up, and providing a guided demonstration on the use of a low-sensitivity pregnancy test does not significantly impact on the accuracy of self-assessed abortion outcome compared to simple verbal instructions (88% vs. 85% accuracy; p=0.449). The fifth paper documents successful self-administration of mifepristone, a higher 24-hour abortion rate (93% vs 77%; p<0.001), and greater acceptability following the introduction of mifepristone into second trimester abortion care, compared to historic cohorts receiving misoprostol only. The thesis concludes that supported self-management and task sharing can strengthen medical abortion provision in South Africa. Research evaluating task sharing of medical abortion care has potential to inform similar approaches for other health care services.

Background: With 2.7 million women living with HIV, the burden of HIV remains high in South Africa but adherence to antiretroviral therapy remains a concern among pregnant and postpartum women. Disclosure, or the process of gradually revealing one’s HIV status to individuals in one’s social network, is regarded as an important factor in HIV care, with potential benefits that include improved psychological well-being and adherence to antiretroviral therapy. This thesis sought to provide insights into the patterns, predictors and impact of HIV-status disclosure among pregnant and postpartum women in the context of lifelong antiretroviral therapy in South Africa, including considerations of stigma, social support, depression and unintended pregnancy. Methods: This research included women living with HIV in Gugulethu, Cape Town. A total of 1554 pregnant women were enrolled; those who were initiating antiretroviral therapy were followed up to 18 months postpartum, with one further visit at 36-60 months postpartum. Data were collected using questionnaires and blood specimens for HIV viral load testing. Findings: Across analyses, women’s social and economic circumstances emerged as central to understandings of disclosure, mental health and viral load. At entry into antenatal care, 95% of women who were diagnosed HIV-positive before the pregnancy had disclosed to at least one person but disclosure events formed two separate dimensions: disclosure to (i) a male partner and (ii) family/community members. Among women diagnosed during the pregnancy and initiating antiretroviral therapy, 61% disclosed to a male partner and 71% to a family/community member by 12 months after diagnosis; relationship status modified the impact of each of pregnancy intentions and poverty on disclosure to a male partner. During pregnancy, 1 in 10 women reported elevated depressive symptoms and 60% of women who were subsequently followed during the postpartum period reported that their pregnancy was unintended. Stigma modified the association between social support and depression: when levels of stigma were high, higher levels of social support were not associated with decreased depressive symptoms. Pregnancy intention modified the impact of disclosure to a male partner on depression during pregnancy: disclosure was associated with higher depression scores among women who reported that their current pregnancy was unintended but was associated with lower depression scores among women who reported that the pregnancy was intended. Further, unintended pregnancy was a persistent predictor of elevated viral load up to 60 months postpartum. Finally, the effect of disclosure on elevated viral load at entry into antenatal care, delivery and 12 months postpartum was complex and modified by three factors: (i) timing of HIV diagnosis (before versus during the pregnancy); (ii) relationship to the person(s) to whom women disclose; and (iii) in the case of disclosure to a male partner, relationship status. Conclusions: These findings suggest that despite the widely-held view that disclosure has beneficial impacts on psychological well-being and adherence, the individual is central to our understanding of disclosure. In particular, both the prevalence of disclosure and its impact on depression and viral load are modified by women’s circumstances. Unintended pregnancy emerged as a critical factor that heightens women’s vulnerability. In this setting, HIV-status disclosure does not appear to be universally beneficial and counselling about disclosure may be most effective if tailored to individual women’s circumstances.

Includes bibliographical references (p. 253-265) / This study seeks to use the techniques of cost-effectiveness analysis to evaluate, within the context of effective vector control, the change to artemisinin-based combination therapies (ACTs) as first line malaria treatment and to evaluate the relevance of using definitive diagnosis (as opposed to clinical diagnosis) as the basis for initiating malaria treatment, especially when using ACTs for treatment. The cost-effectiveness of ACTs was evaluated in two study sites (i.e. In Kwazulu Natal which switched from SP monotherapy to AL in 2001 and in Mpumalanga which changed from SP monotherapy to AS+SP in 2003) in South Africa. The economic evaluation of use of routine definitive diagnosis as part of malaria case management, using rapid diagnostic tests (ROTs), was undertaken at two districts (Namaacha and Matutuine), in southern Mozambique, where routine use of ROTs and treating malaria patients with an ACT (using artesunate + SP) were implemented at pilot level in 2003.

Includes bibliographical references / Although South Africa has the largest antiretroviral therapy (ART) program worldwide, data on key outcomes like mortality and loss to follow-up (LTF) are limited. A few cohorts have published patient outcomes but there is no national reporting on ART scale-up and its impact on the health of HIV-infected individuals. Yet such monitoring of outcomes is vital to inform and improve service delivery. The International epidemiologic Databases to Evaluate AIDS Southern Africa collaboration (IeDEA-SA) was established in 2005 to collect and analyze individual level data from the larger cohorts of individuals on ART in Southern Africa. Using routine, anonymized data from the South African sites, this thesis aims to describe how the program has evolved over 10 years and to assess its effectiveness .Five quantitative analyses were performed using descriptive statistics and survival analysis methods. The studies used patient-level data on adult patients starting ART to describe characteristics and to explore outcomes and temporal changes in outcomes over time. Patient numbers ranged from 19,481 (limited to cohorts with civil identification numbers) to 83,576 adults, followed for up to 214,400 person years. The results are presented as four published papers and one submitted for publication. The thesis describes a rapid, massive scale-up of services. Despite improvements in baseline immunologic status, late diagnosis and ART initiation especially in men area challenge.

Includes bibliographical references. / Migration is a social determinant of health. The relationships between migration and health are not well understood for the large numbers of migrants in low- and middle-income countries, including South Africa. In particular, nutrition during pregnancy and the first two years of a child’s life impact infant morbidity (e.g. diarrhoea, chronic disease) and mortality, yet little is known about maternal and infant nutrition in relation to migration. Positing that migration alters the family structures that traditionally provide social support and advice for mothers, this study was framed in terms of migrants’ loss of these social supports in Cape Town. This framing provided context for the broader investigation of migrants’ nutrition during pregnancy and the first two years of their infants’ lives. This qualitative study with migrants from the Democratic Republic of Congo (DRC), Somalia, and Zimbabwe included 23 in-depth interviews with recently (<2 years) postpartum women, and nine focus groups with adult men (N=3; n=21) and women (N=6; n=27). While in-depth interviews provided insights into individual nutritional motivations, focus groups provided insights into the social context of nutrition. Language interpretation was used in cases where participants did not speak English. Interviews and focus groups were recorded and transcribed verbatim; bilingual research assistants checked the quality of language interpretation and transcription. Rooted in notions of social constructivism, thematic analysis guided the development of a codebook of themes and subthemes. These analytic themes were grouped in relation to (1) maternal nutrition, (2) infant nutrition, and (3) past and present social support. Thereafter, a “thick description” involved interpreting key themes and producing the narrative that integrated focus group data and in-depth interview data. Analysis of maternal nutrition involved documenting foods consumed during pregnancy, as well as investigating the motivations that undergird nutrition during this period. The findings related to migrant maternal nutrition affirmed and built on previous research, which suggested that pregnant women generally maintained their previous non-pregnant eating habits. While cravings were dominated by self-perceived “traditional” foods of home, that were expensive and hard to find in Cape Town, women also commonly described consumption of fast foods and junk foods during pregnancy. Participants did not mention food scarcity, despite the fact that some migrants appeared to be food insecure. These findings illuminated the role of the nutrition transition in Cape Town, that is, migrants were at risk of consuming energy-dense, nutrient-poor diets, particularly during pregnancy. Secondly, analysis of participants’ experience of breastfeeding, formula feeding, and complementary feeding took place in the context of high rates of breastfeeding initiation but low rates of exclusive breastfeeding in many parts of Africa, including migrants’ countries of origin. In this study, migrants presented the common decision to introduce formula in light of their experiences of Cape Town as a work environment. Participants framed the introduction of formula and complementary food early in an infant’s life as primarily a pragmatic and intuitive decision in response to their infant’s cues. Whereas past studies conducted in LMIC tend to present breastfeeding as an important intervention to improve child “survival”, participants in this study were not primarily concerned with child survival. Rather, they were concerned with their family’s tenuous circumstances in Cape Town. As such, efforts by the health system to promote breastfeeding amongst migrants should emphasize benefits to long-term health as well as the short-term financial costs of formula feeding. However, I argue that this shift cannot occur without recognition of, and attention to, migrants’ pressing short-term needs, including housing, legal work status, and safety. Another important factor influencing maternal and infant nutrition revolved around the loss of social support, particularly the loss of the elder generation. Elder women played a central role in providing physical, social and informational support to new mothers in migrants’ countries of origin. As such, the absence of grandmothers in migrant communities in Cape Town was central to understanding participants’ maternal and infant nutrition decisions. Yet participants focused on the loss of household help, including cooking, rather than on grandmothers’ traditional authority or nutritional advice. Given the relative absence of the older, nonworking, generation in Cape Town, community support was limited by the pressures of work and survival. These pressures seemed to make healthy nutrition during pregnancy, or exclusive breastfeeding very difficult. Given this lack of support, medical providers presented one avenue of additional support. However, migrants were frequently unable to communicate with health care providers, and felt discriminated against and unwelcome in the health care system. The three primary findings relating migration and maternal and infant nutrition in Cape Town suggest several avenues for intervention and further study. Firstly, migrants’ descriptions of energy-dense, nutrient-poor diets suggest a role of policy-makers to improve the overall accessibility, availability, and affordability of more nutritious food to the poor in Cape Town. Recognising that foods from migrants’ countries of origin were of particular cultural and nutritional value, a smaller scale intervention might involve creating space and time for the preparation of “traditional” foods. Secondly, improving infant feeding involves re-orientating migrants towards the long-term benefits of breastfeeding and complementary feeding, and engaging spouses and male partners as integral to this process. Further research is needed to create a strong evidence base for the increasing rates of breastfeeding, both in Cape Town and in other urban centres in LMIC. Thirdly, given self-described social isolation and poor experiences in healthcare settings, free-to-patient medical interpretation may play an important role in connecting migrants to both healthcare services as well as broader social services. The improved communication facilitated by medical interpretation may also play a role in combatting the xenophobia that migrants face, both in the healthcare setting as well as in daily life. Indeed, my recommendations must be part of a broader public health research effort to explicate the negative health consequences of xenophobia. To design appropriate research and interventions for migrants, it is important to acknowledge the overarching roles played by xenophobia, legal status, and the broader socio-economic context in shaping maternal and infant nutrition