Anaemia in pregnancy is a substantial public health problem, particularly in resource-limited settings where every second pregnant woman is estimated to be anaemic. This thesis aimed to address gaps in our knowledge of maternal anaemia in order to inform effective prevention strategies in resource-limited settings. Specifically, this thesis sought to determine: the drivers of haemoglobin level changes in pregnancy and the first 12 months postpartum; the interactions between key anaemia risk factors; and the impact of maternal anaemia on infant haemoglobin and iron stores. This was done using a cohort of 699 women and their infants living in the East New Britain province of Papua New Guinea. Firstly, I determined the relationship between red blood cell genetic polymorphisms, anaemia and the malaria-causing Plasmodium species (spp.) infection in pregnant women; and the subsequent effects on birthweight. Women with the homozygous alpha+-thalassemia genotype had significantly lower haemoglobin levels and babies with significantly lower birthweight compared to wildtype individuals, with the association between alpha+-thalassemia and birthweight mediated largely through mechanisms independent of maternal haemoglobin levels. There were no significant associations between genetic polymorphisms and maternal Plasmodium spp. infection. These findings contribute to our current understanding of how genetic polymorphisms influence pregnancy outcomes. Secondly, I determined the contribution of iron deficiency to maternal anaemia and assessed temporal changes in haemoglobin and ferritin levels from enrolment through to 12 months postpartum. Haemoglobin levels remained consistently low, with at least 69 percent anaemic throughout the study period (haemoglobin <110g/l in pregnancy; haemoglobin <120g/l postpartum). Ferritin levels were more dynamic with iron deficiency (ferritin <15 micrograms/l) highly prevalent in pregnancy (>80%), but less so in the postpartum period (~30%). Iron deficiency was identified as a key intervenable anaemia risk factor, associated with a significant 5-fold increased odds of anaemia over the entire study period, compared to iron replete. Plasmodium spp. infection was also an important intervenable risk factor for maternal anaemia throughout the study period. Thirdly, I investigated the relationship between host iron status and Plasmodium spp. infection during pregnancy and postpartum periods, in order to inform the safety of iron supplementation as an anaemia prevention strategy. Host iron status (replete/deficient) was not significantly associated with the prevalence of peripheral Plasmodium spp. infection at any evaluation time, suggesting iron supplement use is safe in this setting. Inflammation was shown to confound the relationship between host iron status and infection, biasing results towards a protective effect for iron deficiency; highlighting the need for future studies assessing this relationship to adjust for inflammation. Finally, I assessed the impact of maternal anaemia on infant haemoglobin and ferritin stores from birth through to 12 months old. Severe iron deficiency anaemia in pregnancy was associated with significantly lower infant haemoglobin and ferritin levels during the study period; demonstrating that maternal haemoglobin and ferritin levels in pregnancy impact newborn and infant stores. This suggests that infant anaemia could be prevented by targeting maternal anaemia during pregnancy. Overall, this thesis outlines the high burden of anaemia during pregnancy, as well as the first 12 months postpartum for women in the East New Britain province of Papua New Guinea, and provides a comprehensive assessment of maternal anaemia aetiology. The data suggests that effective provision of iron supplementation both during and between pregnancies, in conjunction with malaria prevention strategies, would significantly reduce maternal anaemia prevalence and improve infant haemoglobin and iron stores.

The impact of moral distress on health professionals, patient care, healthcare teams, organisations and the healthcare sector itself, is at this time significant and increasing. Most research into moral distress has focussed on nurses in acute care settings, and largely in the United States. This combined ethnographic and interpretive phenomenological research investigated gaps in the moral distress literature including the experiences of health professionals’ moral distress (1) in the Australian context, (2) in a multidisciplinary specialist palliative care team, and (3) in a community setting. My research began with the question: “how do community specialist palliative care health professionals experience and respond to moral distress?” This study makes three significant contributions to the contemporary exploration of moral distress. The first is the concept of moral equilibrium-disequilibrium that offers a new way of understanding and responding to moral challenges in palliative care. This new concept emerged from the iterative analysis of the experiences of participants in this setting. Moral disequilibrium results when health professionals, faced with challenges to their values and moral integrity, are unable to adapt to maintain their moral integrity. Moral disequilibrium is part of the everyday experience of being a health professional, and like all experiences will vary from person to person and in intensity, from mild (e.g. moral discomfort) to severe (e.g. moral distress, moral injury). The moral equilibrium-disequilibrium concept provides a positive perspective to moral challenges, and also considers the psychological, spiritual and physical well-being of health professionals. The concept is of practical value to clinicians as it names the dynamic involved in their experiences without the associations of varied current definitions of moral distress. If moral disequilibrium can be identified and addressed early, it may mitigate the progression into further disequilibrium. The second significant contribution is the development of the novel Moral Equilibrium Framework (MEF). The framework comprises three dynamic phases: identification of moral disequilibrium and the moral values involved; its resolution in attaining a “good enough” moral equilibrium; and the evolution phase of growth in moral understanding including the building of moral resilience. Whilst the framework was derived from the experiences of health professionals who were part of this study, I argue that its phases and processes are consistent with various recommendations in the moral distress literature, adding to its strength. The moral distress literature also emphasised the importance of organisational responsibilities in mitigating moral distress. Similarly, whilst the MEF focusses on individual and team processes, its use does not negate the organisation’s role in helping to minimise moral disequilibrium. The third important finding of this research is that the team as an entity can experience moral disequilibrium. This is in contrast to previous studies that name moral distress as individual experiences occurring within the team. This revelation of team moral disequilibrium highlights the importance of being attentive to and addressing the state of moral disequilibrium not only at the individual level but also at the team level. In answer to my research question, I conclude that experiencing daily moral challenges, community specialist palliative care health professionals often choose to respond in ways that help to regain a “good enough” moral equilibrium, both as individuals and members of a team. The thesis will be of interest to health professionals who experience daily moral challenges in an ever growing, stress-filled environment of caring for patients, and who seek to find ways of regaining and sustaining “good enough” moral equilibrium and integrity.

Significant progress has been made in reducing maternal and child mortality at the global level. However, the burden of maternal and child mortality remains high. Indonesia, home to over 270 million population, has one of the highest burden of maternal and child mortality in the world. One of the main challenges in reducing maternal and child health burden in Indonesia, and other LMICs, is inequitable access to essential maternal and child health (MCH) interventions, particularly among the poorest population and those living in remote areas with poor access to health services. It is crucial to understand the influence of health systems on MCH to inform the design and implementation of health interventions. Furthermore, there is also limited evidence on the impact of health insurance on effective MCH service coverage, disparities of health care, and health outcomes. The four objectives in this thesis addressed these issues. First, I described the geographical inequalities in skilled birth attendance (SBA) coverage across Indonesia regions and quantified the contributions of socioeconomic and health supply-side factors towards the geographical variations. Geographical inequality in SBA coverage was mainly driven by disparities in the availability of health supply-side, i.e. the number of skilled birth attendants and hospitals. Interestingly, the number of primary health centres did not influence the uptake of SBA. Second, I assessed the trends and inequalities in the RMCH effective coverage from 2002-2017 to examine gaps in the RMCH care cascade, and investigate variations of these cascade indicators among regions and socio-demographic groups over time. The study revealed that while significant improvement has been made in the levels of crude coverage for all MCH services, the quality of care is still low. The user adherence in following the standards of antenatal care and childhood immunisation was also found to be a challenge. Antenatal care and maternal delivery services were the least equitable interventions compared to family planning and childhood immunisation services. Provinces with similar levels of overall coverage had varying degrees of inequity. This indicates different pathways towards high coverage in relation to the distribution of the services. Next, I evaluated the association of the Indonesia national health insurance (JKN program), the largest single-payer health insurance program in the world, on access to maternal health service and child mortality. Findings from my thesis clearly indicated that JKN was associated with increased coverage of antenatal care, SBA and facility-based delivery. However, JKN was not associated with the reduction in neonatal or infant mortality. This study supports WHO call for universal health coverage through the expansion of social health insurance in LMICs. Quality improvement initiatives must place greater emphasis on the vulnerable population, less developed regions, and geographically remote areas. Finally, the JKN influenced women’s choice of the type of birth attendant and place of delivery. JKN could reduce unsafe deliveries by increasing the SBA use and facility-based deliveries. However, JKN also promoted the use of high-level maternal care, particularly among the richer population and women living in more developed regions in Indonesia. Hence, the main lesson from this study for the healthcare system in Indonesia and other countries is that health insurance coverage by itself may not be sufficient to address challenges in the quality and appropriateness of care. Consequently, future health system reform must place greater emphasis on strengthening integrated primary health care to achieve universal health coverage in Indonesia.

Participatory research with children aims to actively collaborate with child co-researchers to co-construct knowledge and social action about their worlds. Recently, a growing number of studies have increased the size and scope of participatory research with children. However, there is scarce evidence on how collaborating across study sites and engaging with data that reflects the lives of children from different backgrounds and contexts influences child co-researchers’ views about themselves, their communities, and the world, as well as their experiences during the research process. Additionally, although scaled-up participatory studies tend to show overall high levels of child agency, little is known about the process of increasing child co-researchers’ agency within and beyond the research setting. My PhD research focused on answering the question of how does scaling-up participatory research with children influence child co-researchers’ critical consciousness and their agency in the research? To answer this question, I carried out a scoping review of the literature and two qualitative participatory studies. The Children’s Araucaria study focused on exploring children’s experiences in two disaster risk reduction programs, one in Chile, the other one in Australia. My second study, Kids Contribute, was part of a larger initiative led by the Child and Community Wellbeing Unit at the University of Melbourne and carried out in collaboration with a children’s television news program. Child co-researchers from different backgrounds and places in Australia helped create questions and interpret results for a poll about how children contribute at home, school and elsewhere in the community. Then, a group of child co-researchers presented their findings and recommendations at a national forum, organized with the Australian Human Rights Commission. My work in this project focused on examining the qualitative and participatory component of the research. My findings showed that scaling-up participatory research with children can help promote child co-researchers’ critical consciousness at the individual, social and global levels of awareness. As the research process increased their exposure to other children’s perspectives, child co-researchers began to explore the connections between their personal experiences and those of other children in their communities and across Australia. Child co-researchers were also motivated to learn about the realities of children in different parts of the world and to collaborate across geographic frontiers to address global issues like climate change and racism. My research also found that, if the goal is to scale-up participatory research with children, child co-researchers should have opportunities to engage in research activities beyond their local community across the different stages of the research process. This task can be complex, since the context of each study location can facilitate or constrain child agency in different ways. Through their dialogue with child co-researchers, adult researchers can promote critical consciousness at different levels of awareness by using research activities as prompts and integrating discussions about the different geographic and ecological levels involved in the study. Child co-researchers can also benefit from collaborating with each other across study sites. Finally, to increase the impact of the research and child co-researchers’ empowerment in their relationship with adults, researchers can help build increasingly wider audiences that are willing to listen and act upon children’s ideas.

Health economic evaluation is a fundamental component in helping inform health care providers and policy makers in making decisions on priorities in health care. This is particularly relevant in today’s tight budgetary climate and in response to continued calls for sustainable health care systems. Economic evaluation evidence has influenced and contributed to many areas of health policy making, at all levels of the health care system – from shaping guidelines that guide clinical care to informing decision for subsidy of pharmaceuticals and medical services. For these evaluations to be useful to decisions makers, they need to provide useful and reliable information and to achieve this, methodological guidelines should be followed, and robust evidence of effectiveness and cost is paramount. With recent advances in information technology, data and statistical methods and implementation of electronic health records, health decision makers are increasingly seeking real-world, generalisable evidences to complement and support policy and clinical decisions. This thesis aims to demonstrate the usefulness and practicality of applying real-world longitudinal data in health economics research and applications. It features six individual health economics studies which explore longitudinal data and show their value and contribution towards advancing economic evaluation methodologies and better decision making. Each of the studies answer specific research questions and contribute to the research literature through methodological research to improve consistency in extrapolating costs, utility inputs and modelling long-term outcomes, generating robust evidence for resource allocation decisions, promoting a better understanding of real-world heterogeneity and approaches to optimise patient outcomes. Collectively, these studies highlight important variations in the cost and outcomes of health care delivery in real-world settings, provide useful insights into the implications of such variations and demonstration of translating research findings to implementation.

The working environment is central in the lives of individuals in employment, influencing health outcomes including psychological and physical well-being. Psychosocial work stressors are common exposures in the workplace and are important and modifiable determinants of health and health behaviours. There is broad agreement in the literature that exposure to adverse psychosocial work stressors, such as high job demands, low job control, low job security and high effort-reward imbalance are associated with poor health outcomes. All of these exposures are, in turn, associated with unhealthy behaviours such as smoking, alcohol consumption, poor diet and inactivity, and to the development of hypertension, cardiovascular disease, diabetes, musculoskeletal disorders and depressive disorders. Psychosocial work stressors have been identified as significant emerging risks linked to global changes in the structure, organisation and management of work during previous decades, presenting pressing challenges to occupational health and safety and workplace health, in general. Most of the literature in this area assesses disease incidence or other morbidity outcomes, with a growing number of studies focusing on mortality. However, whether exposure to these psychosocial work stressors associated with adverse health outcomes translates into increased mortality remains unclear, and barriers to making causal interpretations about the relationship between psychosocial work stressors and health persist, mostly due to inherent biases in the methodology across studies. In this thesis, the effect of exposures to psychosocial work stressors in the working environment such as job control, job demands, job strain, long working hours, job insecurity and shift work on health and mortality were investigated. A comprehensive systematic review including meta-analyses of the epidemiological research was conducted. Panel data from the Household, Income and Labour Dynamics in Australia survey was used to capture natural experiments of psychosocial work stressors associated changes in health and well-being, and mortality to investigate whether and how the effect of psychosocial work stressors on mortality differ in relation to demographic and socioeconomic characteristics. The initial focus of the research included establishing and quantifying the risks associated with adverse psychosocial work stressors, which workers are commonly exposed to in the workplace, on mortality. The final study in the PhD focused on understanding the exposure-outcome dynamics. The dynamics of the connexions between psychosocial work stressor perceived job control and general health, a strong predictor of future morbidity and mortality, were investigated for evidence of a causal relationship. The first study showed that improving the quality of work might contribute to better health and well-being and decrease the effect of psychosocial work stressors on all-cause mortality and cardiovascular disease mortality. If this observed association is causal, then policy and practice interventions to improve job control could contribute to reductions in mortality. Study II adds to the growing body of evidence showing an effect of adverse psychosocial work stressors on mortality. Study III showed that long-term exposure to low job control and job security was associated with increased risk of all-cause mortality. In Study IV, the dynamics of job control and general health were explored for evidence of a causal relationship using three complementary longitudinal modelling approaches. This study, using improved causal inference methods than previous research, showed that increased job control was strongly associated with increasing general health. Although the estimates for the single measures suggest modest increases in the risk of death and poor health, this translates to large population impacts because the exposures are relatively common across the working population. Thus, reducing exposure to psychosocial work stressors associated with mortality and poor health could have considerable public health benefit. Awareness by stakeholders, government and organisations of the implications of the adverse effects of psychosocial work stressors on health and mortality in workplaces; and the availability appropriate, evidence-based work stress interventions to reduce the exposure to work stressors might contribute to better health and well-being, reduce sickness absence and presenteeism to the benefit of workers, workplaces and society.

Introduction Non-Communicable Diseases (NCDs) are the leading cause of death and morbidity throughout the world, with the greatest burden in low- and middle-income countries (LMICs). Of the estimated 17.9 million CVD deaths in 2016, more than 10 million occurred in LMIC countries. Coronary heart disease and stroke were the two major causes contributing to these deaths. Papua New Guinea (PNG) is categorized as a lower-middle-income country according to World Bank criteria and it is experiencing rapid economic growth as a result of large-scale mineral and gas resource developments. These economic changes are contributing to epidemiological transitions associated with rapid lifestyle changes that, in turn, are leading to increases in cardiovascular diseases and diabetes. NCDs and associated risk factors have not been well investigated in PNG, however, several small studies conducted over the past 40 years have been suggestive of increasing prevalence of NCDs and their associated risk factors (Chapter 3). The aim of this research is to establish an up-to-date NCD risk factor prevalence data enabling a better understanding of the differences in these risk factors in relation to socio-economic status in three locations in PNG. The present study was undertaken during the construction phase of a large-scale gas development which was projected to more than double the gross domestic product (GDP) of PNG. The study was designed to provide baseline prevalence data on NCD risk factors in the initial years of a gas project impact site (West Hiri) and in two non-project impact sites (Asaro and Karkar Island). It was also anticipated that the study findings would provide up-to-date NCD risk factor prevalence data to help the national government plan services and develop cost-effective interventions. This thesis describes the methods used and presents the initial findings for NCD risk factors in a survey of three different socio-demographic populations of PNG. Methods The analysis of the data presented in the results sections of this thesis was based on the NCD Risk factor survey. The survey was a cross-sectional study of the prevalence of NCD risk factors across three different sites namely West Hiri in Central province, Asaro in Eastern Highlands province and Karkar Island in Madang province. A modified questionnaire based on the WHO STEPwise instrument was used for data collection. In addition, physical measurement and biochemical samples were collected from participants (Chapter 4). Results A total of 785 participants participated in the survey. The prevalence of NCD risk factors varied markedly different across the three sites. The metabolic risk factors such as obesity, elevated blood pressure, increased total cholesterol and HbA1c levels were higher in West Hiri compared to the other two sites (Chapter 5). Further analysis of the data was done to investigate the association between socio-economic status (SES), and the CVD risk factors. Findings show that elevated CVD risk factors were common among all SES groups, but metabolic risk factors were more prevalent among homemakers, peri-urban West Hiri and Asaro and among the highest quintile groups. Adults in the peri-urban West Hiri had higher risk of obesity, hypertension, abnormal lipids and elevated HbA1c. Similarly, Asaro adults had increased risk of central obesity, hypertension, elevated triglycerides and MetS compared to the residents of rural Karkar Island (Chapter 6). Conclusion The research reported in this thesis adds to our understanding of the associations between SES and CVD risk factors in a LMIC like PNG. Data from high income countries show a negative correlation with socio-economic status, however findings from our study showed the association between CVD risk factors and SES varied greatly depending on the type of SES measure used. Understanding these associations is important to inform the government to develop appropriate and effective prevention and control strategies. With this information at hand, the government would be able to make informed decisions and prioritize its prevention and control strategies targeting high risk populations or settings in the country.

Male suicide continues to be a major public health concern both internationally and in Australia. While a number of factors have been found to contribute to high suicide rates, little is known about what drives the much higher rates in men compared to women. Mostly, risk factors for suicide are considered in isolation, with little regard for the mechanisms that may underpin them. In more recent discussions on male health, masculinity has been highlighted as a potential driving force underpinning unhealthy male behaviors. In Australia, the “dominant masculinity” is one that endorses the norms of stoicism, independence, invulnerability and avoidance of negative emotions. Conformity to these masculine norms has been associated with suicidal thinking, poor mental health and reduced and delayed help-seeking. Men also describe experiencing stigma and how the fear of being alienated prevents them from seeking help for mental health issues. For these reasons, men are often described as a “hard-to reach” population group when it comes to mental health messaging. Suicide prevention interventions are needed that adopt a gendered approach and attend to the influence of masculine norms on mental health and suicide. The Man Up intervention was an innovative media-based male health promotion and suicide prevention intervention that featured a three part television documentary and digital campaign delivered via a website and five social media platforms (Facebook, Twitter, Instagram, YouTube and tumblr). It was funded by the Movember Foundation in Australia and its aim was to address the high suicide rate among men by promoting help-seeking for personal or emotional problems via an exploration of Australian masculinity. The documentary and digital campaign examined how society shapes the way men and boys see themselves and explored how this might affect mental health and, potentially, lead to thoughts of suicide. It was aired by the Australian Broadcasting Corporation (ABC), Australia’s national public free-to-air broadcaster, in October 2016. This thesis explores the potential for social media to be used in health promotion with the view to generate and increase engagement and influence conversations on the issue of male suicide and its link to masculinity. As such, this thesis focuses on the evaluation of the social media components of the digital campaign. The thesis is based on the idea that masculinity, the rules prescribed by society about how men should live their lives, is a fundamental determinant influencing how men negotiate their health throughout life. Following social constructionist theory, this thesis proposes that by challenging dominant masculinity social learning processes can take place that will support redefining restrictive behavior patterns and opening up alternative behaviors which in turn will help reducing stigma for men facing mental health problems, increasing male help-seeking and ultimately reducing male suicide rates. The above aim was addressed through two empirical studies that analyzed various data sources collected through two social media platforms: Facebook and Twitter. In Study 1, Twitter Insights data were used to assess reach and engagement with the campaign and to determine highest and lowest performing tweets. We also analyzed the volume of conversations over time by tracking the use of common campaign hashtags and conducted thematic analysis on a sub-set of tweets to determine most engaging campaign content themes. Study 2 was a qualitative study of comments published on the Man Up Facebook page with the aim to provide further insight into the conversations that were instigated by the campaign. Both studies have been published as peer-reviewed journal articles and together they demonstrated that the Man Up social media campaign was able to generate an impressive reach and exposure to campaign messages in the Australian population and beyond. Overall, the studies showed that the Man Up campaign messages were perceived positively across both Twitter and Facebook. Campaign content was widely shared on social media and positive feedback showed endorsement for the campaign. It highlighted the need to openly talk about male suicide and the stigma introduced by gendered practices. The increased use of the hashtag keywords of the Man Up campaign in line with the campaign phases and the fact that the hashtag MANUP was trending during the campaign signaled a strong uptake and sharing of campaign messages. Despite this, the hashtag analysis could not find a lasting increase of engagement. The analysis of Facebook comments confirmed that the social media campaign triggered conversations about masculinity and suicide that might otherwise not have happened. For some, this may have led to shifting attitudes towards expressing emotions and reaching out to others for help, however, this could not be formally investigated based on social media data alone. This thesis concludes that social media interventions can be used successfully in engaging men in discussions about male mental health, suicide and gender norms and therefore demonstrate potential for suicide protective properties of social media. However, more research is needed to better understand how and where on the internet to best engage men, particularly men at high risk of suicide. While gender-transformative campaigns such as Man Up can help to further destigmatize help-seeking for men, the longevity of changes needs further consideration. Addressing these issues will require high quality interventions and evaluation designs for social media campaigns which follow a systematic framework for measuring effectiveness.

Asthma is the most common chronic disease in children, and often persists into and adulthood. Further, despite its known health benefits, it is estimated that more than 80% of the world’s adolescents are insufficiently physically active. Since both asthma and physical inactivity are significant problems in child and adolescent populations, it is essential to focus research on these populations. The relationship between asthma and physical activity is complex and controversial, and the temporality of these associations remains unclear. My doctoral work attempted to tease out these relationships. This work was divided into sections with distinct research questions, which together aimed to advance our current understanding of the relationship between asthma and physical activity in children and adolescents. Hence, this thesis contains two systematic reviews which were performed in order to understand the existing literature on the nature of the relationship between physical activity and asthma in children. These reviews are followed by three data analyses which employed a variety of statistical techniques to analyse data from questionnaires, accelerometry and clinical assessments collected by two well-established cohorts of Australian children: The Longitudinal Study of Australian Children (LSAC) and the HealthNuts study. The first of the two systematic reviews found no evidence that children and adolescents with and without asthma participated in different amounts of objectively measured physical activity. The second review, which aimed to investigate the effect of physical activity on subsequent asthma outcomes, found insufficient evidence to determine the effect of physical activity on subsequent asthma and lung function outcomes. Hence, in the first data analysis, I aimed to elucidate the direction of the association by investigating the possibility of bidirectionality. However, there was no evidence of longitudinal associations in either direction. Next, I compared the amount of time spent in moderate to vigorous physical activity (MVPA) between children and adolescents with and without asthma at 5 different ages and again found no evidence that youth with and without asthma engaged in differential amounts of time in physical activity. Finally, I investigated the effect of early wheeze and asthma on objectively measured physical activity in early childhood and again found no evidence that children who experience early life asthma and wheeze were less active at a later age. Thus, my thesis presents sustained and compelling evidence of a lack of association between asthma and physical activity in youth – in either direction. A number of implications can be drawn from these findings. First, physical activity campaigns targeted at children need not focus specifically on children with asthma. Second, the effect of physical activity on future asthma outcomes remains unclear. However, even without evidence that physical activity is beneficial for future asthma outcomes, physical activity should be encouraged from a young age for its many other health benefits.

Background Various caries risk assessment tools (CRATs) are currently used in the management of early childhood caries. The evidence to inform the selection of a specific tool is unclear, as CRATs differ in their content and development methodology, mostly due to the absence of a standard reporting and testing guideline. Risk-based and outcomes focused (the shift from volume to value) models of caries management have been gaining popularity in recent times as they attempt to prevent and manage the disease itself rather than its sequela. Risk-based models of oral health care attempt to stratify populations, with those most at-risk prioritised for care, while concurrently reducing unnecessary care for low-risk people. This becomes an important consideration when allocating resources in limited resource settings such as public dental services. If new models of caries management depend on an accurate and precise risk assignment, then evidence is needed to inform each step of the development and selection of CRATs. Without this essential body of work, risk-based models of caries management run the risk of not delivering the intended positive outcomes, for either the patient or the health care organisation. The aim of this PhD was to generate new knowledge on caries risk assessment tools for young children, to demonstrate their utility in models of caries management. Methods The philosophical underpinnings of my PhD research were informed by the value-based health care (VBHC) agenda. The theoretical foundations, including the intent, motivation and expectations for the research were guided by two socio-ecological conceptual frameworks – National Institute for Health and Clinical Excellence’s (NICE) conceptual framework for public health and the Fisher-Owens conceptual model of influences on child oral health. The PhD research was conducted as three separate research strands: Strand 1. The systematic review – provided evidence for the current state of knowledge including knowledge gaps for CRATs for young children, relating to content, development process, and testing methods, risk categorisation and use in practice; Strand 2. The calibration study – provided evidence for examiner inter- and intra-examiner reliability for the VicGen data and also generated new knowledge related to the process for calibrating clinical examiners in caries diagnosis when conducting research with young children; and Strand 3: The CRAT testing – provided the evidence, based on an assessment of the reliability and validity required to inform the selection of a CRAT (CAMBRA CRAT used as the case study) for young children in the State of Victoria. Results The CRATs reported in the literature for use among children less than 6 years were: Caries Management By Risk Assessment (CAMBRA), Cariogram, National University of Singapore CRAT (NUS-CRAT), MySmileBuddy, Dundee Caries Risk Assessment Model, University of North Carolina Risk Assessment Models, University of Michigan pediatric dental clinic caries risk assessment sheet and American Academy of Pediatric Dentistry (AAPD) CRAT. For examiner reliability in the calibration study, Kappa scores for inter-examiner reliability ranged from 0.56 to 0.70 (moderate to good). In the in vivo community examination, using ICDAS caries codes only, weighted Kappa scores for inter-examiner reliability ranged from 0.50 to 0.64 (moderate to good). In the CRAT testing study, good reliability (inter and intra) was observed when using the CAMBRA CRAT. For the classic CAMBRA three-level risk categorisation (low, moderate and high), at 36 and 48 months of age, respectively, the area under the receiver operating curve was close to 0.50, which is indicative of a low ability for the CRAT to predict future caries development. Conclusion To conclude, this PhD study identified a major knowledge gap in the current state of evidence on the subject of caries risk assessment tools (CRATs); which was the lack of methodologically sound studies to provide the evidence to inform decision making on tool selection, for use in practice. The PhD study addressed this knowledge gap through methodologically rigorous research to demonstrate the process required to generate the evidence to inform the selection of a CRAT. The CAMBRA CRAT which was used as the case study in the PhD research was shown to be ineffective in predicting a caries outcome among young Victorian children.

Background Irresponsible media reporting of suicide can lead to suicidal acts, particularly if the subject of the reporting is a celebrity. When Robin Williams took his own life on 11 August 2014, media reporting in the United States was less than optimal, with many reports detailing the suicide method Williams used, romanticising his suicide, and failing to provide information on sources of help. The reporting was followed by an increase in suicides in the United States, as well as an increase in helpline calls. In Australia, the Mindframe guidelines provide media professionals with advice on ways to safely report on suicide. This thesis set out to determine whether the Australian reporting of Williams’ suicide adhered to the recommendations in the Mindframe guidelines and whether there were increases in suicides or calls to helplines in Australia following the media reporting of Williams’ suicide. Method The thesis involved three studies. In Study 1, relevant newspaper articles were identified through a systematic search process, and trained coders rated the articles for quality against criteria in the Mindframe guidelines. In Study 2, suicide data were extracted from the National Coronial Information System for the period 2001 to 2016 and interrupted time series regression analyses were conducted to determine whether there were changes in the number of suicides in the five month period immediately following Williams’ suicide. In Study 3, weekly calls data for the period 2013 to 2015 were provided by Australia’s two most prominent helplines, Lifeline and Beyond Blue. Again, interrupted time series regression analyses were conducted, this time to determine whether there was an increase in the average weekly number of calls received by each helpline one week and one month after the news of Williams’ suicide broke. Results Study 1 identified 303 newspaper articles, 67% of which adhered to at least eight of the 10 Mindframe guidelines. Study 2 found an 11% increase in suicides in the five month period following Williams’ death, largely accounted for by men aged 30-64 years and by people who died by hanging (the method Williams used). Study 3 provided strong evidence of an increase in calls to Lifeline and Beyond Blue in the week after Williams’ suicide was reported. This levelled out over the ensuing month. Conclusion Australian newspaper reporting of Williams’ suicide was largely consistent with the Mindframe guidelines. Despite this, there were increases in suicides in the immediate aftermath of his death, which is obviously a negative outcome. There were also increases in calls to helplines, which may be interpreted either as negative (i.e., suggesting that people’s levels of distress and feelings of suicidality were heightened) or positive (i.e., suggesting that people who might have otherwise been influenced to harm themselves called a helpline instead). It may be that Australians were exposed to reports that contravened safe reporting recommendations, particularly via overseas media and social media or that some Australian reports may have had unhelpful overarching narratives, despite largely adhering to the Mindframe guidelines. The Mindframe guidelines constitute international best practice but consideration should be given to whether certain recommendations within them should be further reinforced and whether more nuanced information about how stories should be framed could be provided.

Studies of relationships between dietary factors, lung function and bronchial responsiveness (BR) are limited and findings inconsistent. My aim was to investigate these relationships in middle-aged and older adults. A secondary aim was to identify a more suitable statistical method to assess factors associated with BR and compare its findings to those from the common regression model of the log-transformed dose-response slope (logDRS). I used data from two cross-sectional studies – the Tasmanian Longitudinal Health Study (TAHS) 2010 follow-up and the Chronic Obstructive Pulmonary Disease (COPD) Study. The TAHS is a respiratory study of Tasmanian school children born in 1961. In 2010-2012, an asthma and bronchitis enriched subsample completed spirometry, a methacholine challenge and a questionnaire. I used a linear mixed model (LMM) to examine “known” predictors of BR and compared the findings to those from regression of the logDRS. I used multivariable linear regression to investigate associations between fruit and vegetable intakes and lung function and LMMs to examine associations with BR. The COPD study is a population-based cross-sectional study of adults aged 45-69 years living in inner south-east Melbourne. A random subsample completed spirometry, a methacholine challenge, and questionnaires including a semi-quantitative food frequency questionnaire. I derived dietary patterns from nutrient intakes using principal component analysis and calculated an energy-adjusted dietary inflammatory index (E-DII) as a measure of the inflammatory potential of the diet. I examined associations between these dietary factors and lung function and BR using linear regression and LMMs, respectively. I explored sex, BMI, smoking, asthma status and atopy as effect modifiers of these associations. Results from the LMM differed to those from regression of the logDRS. In particular, sex predicted BR in the regression model but not the LMM. I found relationships between several dietary factors and lung function in those with current asthma only. In this group, higher vegetable intake, higher intakes of a “high potassium & magnesium” dietary pattern, indicating a diet high in fruits, vegetables and wholegrains, and higher intakes of a “low calcium & sugars” dietary pattern, indicating a diet high in vegetables and low in sugar and dairy products, were associated with better lung function. A higher E-DII, indicating a more proinflammatory diet high in animal products and low in fruits and vegetables, was associated with poorer lung function. I also found higher fruit intake was associated with increased BR. Conversely, in those with current asthma, higher scores for several dietary patterns were associated with less BR. In conclusion, I demonstrated results from an LMM can differ to those from regression of the logDRS, and recommend using the LMM to investigate factors associated with BR. My findings suggest a diet low in animal products and high in fruit, vegetables and wholegrains may be beneficial for lung function in adults with asthma. Therefore, a dietary modification program in this group may improve lung function and reduce the prevalence and severity of asthma and COPD. However, further studies are needed to establish causality of the diet-lung function associations and clarify relationships with BR.