This thesis concerns investigations into adverse health outcomes among survivors of childhood cancer using the British Childhood Cancer Survivor Study (BCCSS). The BCCSS is a large-scale population-based cohort of 17,981 survivors of childhood cancer who were diagnosed with childhood cancer (age 0-14 years) between 1940 and 1991, in Britain, and had survived for at least five years. The specific aims were to investigate, within the BCCSS cohort; (1) the psychometric properties of the SF-36 health-status questionnaire, (2) the self-reported health-status by using the SF-36, (3) the effect of therapeutic radiation on the offspring sex ratio, (4) the risks of adverse pregnancy outcomes, and (5) the risks of second primary breast cancer. This thesis demonstrates that the SF-36 questionnaire exhibits good validity and reliability when used in long-term survivors of childhood cancer. Survivors rate their physical and mental health similarly to those in the general population, apart from bone and central nervous system tumour survivors who rate their physical health below population norms. Therapeutic irradiation does not alter the sex ratio of offspring. Female survivors exposed to abdominal irradiation are at a three-fold risk of delivering premature and two-fold risk of producing low birth-weight offspring. Lastly, the risk of breast cancer among female survivors is two-fold that of the general population, but is not sustained into ages at which the risk of breast cancer in the general population becomes substantial.

Vitamin D deficiency (VDD) is widespread in England and Wales, affecting both children and adults by increasing their risk of poor bone health. In order to inform public decisions on the prevention of VDD, a decision analytic model was developed to compare four alternatives to prevent population VDD: wheat flour fortification, supplementation of at-risk groups, a combination of wheat flour and supplementation, and no intervention. Methods used to inform the model development stage included literature reviews, expert consultation, and preference-based questionnaires to elicit proxy utility values. An individual-level simulation Markov model was chosen to simulate the costs and benefits of each alternative. A cost-effectiveness and cost-utility analysis were undertaken for the base case analysis, using a 90-year time horizon. A societal perspective was chosen to account for private and public costs of implementing each of the alternatives. Parameter uncertainty was tested through deterministic and probabilistic sensitivity analyses. Moreover, in order to inform decisions under uncertainty, a value of information analysis was undertaken. The model results support the decision to implement interventions to prevent VDD, namely a combined strategy of at-risk groups supplementation and wheat flour fortification. Furthermore, this thesis contributes evidence on methodological considerations for modelling micronutrient interventions.

Background Eliciting health-state utility values (HSUVs) for use in cost-utility analysis (CUA) for some diseases is limited by the variation in the duration of the health-states such as it includes both temporary and chronic health-states. There is no 'gold standard' approach for valuing temporary health states (THS). Conventional time trade-off (TTO) is deemed inappropriate because it presents an unrealistic scenario to respondents and the chained TTO has been proposed to solve this problem. This thesis uses a case study, Chlamydia trachomatis (a sexually transmitted infection) to explore this challenge. The main burden of chlamydia is typically believed to be borne by women, with a paucity of studies describing the psychosocial outcomes for men, hence making it difficult to value outcomes associated with men. Objectives This thesis aimed to: develop approaches for valuing the health-states associated with chlamydia and derive utility values for the chronic and temporary health-states associated with chlamydia. An additional objective was to define the psychosocial impact of chlamydia in men via an exploration of the asymmetric nature of the disease burden. Methods The thesis elicited utilities for seven health-states (five THS and two chronic health-state) depicting the symptoms of chlamydia. The health-states were developed using evidence from the literature and interviews with clinical experts. Chained time trade-off (TTO) was applied to THSs and conventional TTO to CHSs. Ectopic pregnancy was used as a lower anchor for chained TTO. The VAS technique was also employed. The study sampled from three different population groups and the survey was administered face-to-face. A qualitative synthesis was conducted using meta-ethnography, to identify the psychosocial impact of chlamydia in men. Discussion The thesis identified appropriate approaches for valuing chlamydia health-states and showed that chained TTO is feasible. Methodological challenges arising from this study include the development of health-state description, the selection of appropriate anchor states and the duration of the anchor state. The HSUV s could potentially be used in cost-utility analyses examining the cost-effectiveness of screening. The meta-ethnography highlighted the need to broaden the focus of future evaluations.

Objectives: This study linked Scottish education data to a number of administrative health datasets to explore associations between childhood chronic ill health and subsequent educational and health outcomes. Chronic conditions investigated were diabetes, asthma, epilepsy, attention deficit hyperactivity disorder (ADHD) and depression. Educational outcomes were number of days absent from school, number of school exclusions, special educational need (SEN), academic attainment and unemployment. Health outcomes were all-cause and cause-specific hospital admission, total number of hospital admissions, total length of hospital admission and all-cause mortality. Approach: Pupil census data and associated education records for all children attending primary and secondary schools in Scotland between 2009 and 2013 were linked to national prescribing data, hospital admissions, death records and retrospective maternity records enabling outcomes to be studied whilst controlling for socioeconomic, demographic and obstetric factors including birth outcomes and maternal antecedents. Specific medications are prescribed for some particular chronic conditions; therefore, children identified as receiving these medications whilst at school were assumed to have these conditions. Results: Children treated for each of the five conditions had more frequent absenteeism from school and were more likely than their peers to have SEN. However, only children treated for depression, epilepsy or ADHD experienced poorer academic attainment and increased odds of unemployment. Furthermore, children treated for depression or ADHD were significantly more likely to be excluded from school. Children treated for asthma experienced poorer academic attainment but no increased odds of unemployment and the association with attainment disappeared after adjusting for their increased absenteeism. Children treated for each of the five conditions had an increased risk of hospital admission and children treated for depression or epilepsy also had an increased risk of recurrent hospitalisation and longer stays in hospital. All of the chronic conditions, with the exception of ADHD, were associated with increased mortality. Conclusion: All five of the chronic conditions investigated in this thesis were associated with adverse educational and health outcomes. The number of outcomes affected varied by condition. Treated depression, epilepsy and ADHD were associated with the most wide-ranging impacts. Children treated for depression fared worse than their peers across all nine outcomes, and children treated for epilepsy and ADHD across eight and six respectively. In contrast, children treated for asthma and diabetes fared worse than their peers in respect of around half the outcomes investigated. Children with these chronic conditions at school appear to experience significant educational and health disadvantage; therefore further work is required to understand the underlying mechanisms and to develop effective interventions to reduce their risk.

No description available.

Management of long term conditions is seen as one of the greatest challenges facing the English National Health Service currently (House of Commons Select Committee, 2014). One of the proposed solutions for this growing crisis is to support patients to manage their conditions themselves, thereby reducing hospital visits and lowering costs to the NHS overall (NHS England, 2013). Self-management necessarily involves the patient having some level of knowledge regarding their condition. In addition, there is a commitment within healthcare for patients to be involved in decision making regarding their healthcare needs with the mantra "no decision about me without me" (Department of Health, 2010) as well as development of the 'expert patient' agenda within healthcare (NHS Choices, 2014). In order for knowledge to be shared between health care professionals, patients and carers, knowledge boundaries must be managed within the process of knowledge sharing. This ethnographic study takes as it starting point, literature concerned with knowledge boundaries specifically, but highlights how current understanding of these boundaries is limited in some ways. In particular, issues of power and choice within the management of knowledge boundaries are inadequately explored within this body of literature which limits its applicability to contexts such as healthcare, in which power dynamics are endemic. In addition, the concepts of identity and expertise are implicated in issues of power and therefore this study seeks to explore the relationships between power, identity, expertise and knowledge in order to better understand the nature of knowledge boundaries and the dynamics of their management. This study found that issues of power were fundamentally implicated in knowledge sharing and the management of knowledge boundaries. Knowledge boundaries were found to be not only sites of struggle over knowledge, but also sites of struggle for control more generally. Whilst knowledge and power have long been associated, this study found that this relationship was complex and greater knowledge did not necessarily confer greater power on the social actor. Despite being evident in the rhetoric of healthcare, as a basis for power, expertise was redundant as more embedded social discourses remained prevalent. It is proposed that bases of social power are hierarchical in some respects as power based on one foundation can be rendered ineffective if faced with opposing power exerted by another social actor based on an alternative and more efficacious basis. This study emphasises the pervasive nature of embedded power dynamics within healthcare which concurrently facilitate conventional knowledge sharing processes and impede the development of new processes which would encourage the increased involvement of patients.

Background (Kenya) Sub-Saharan Africa (SSA) is in the midst of experiencing an unprecedented increase in non-communicable diseases (NCD), specifically diabetes and hypertension. This shift has required public sector health systems, which have historically focused on managing acute diseases, to redesign their services to appropriately serve chronic disease needs. Issue Addressed In order to provide a description of our efforts to bring up comprehensive services for NCDs in rural Kenya within this thesis, I have specifically selected publications which target different aspects of the healthcare system. This includes our efforts related to clinical training for pharmacists, screening for NCDs, medication supply chains, remote phone-based care services, and care delivery based in the community. Prior to the implementation of the programs mentioned in these domains, access to these services was largely not available in western Kenya. Furthermore, the publication of our research from this western Kenyan cohort is designed to supplant the relatively limited research which emanates from rural sub-Saharan Africa. Research Questions For each of these selected publications, we defined a set of primary and, in some cases, secondary research questions focused on identifying the contextualized attributes of service delivery in this setting while also assessing the impact. For the first publication on training for clinical pharmacists, we assessed the impact of Kenyan Bachelor of Pharmacy interns and North American Doctor of Pharmacy interns while providing clinical care in an inpatient setting in Kenya. Our primary research question assessed whether there was a significant difference in the number of clinical interventions documented by interns from the two countries. In the second paper, we shifted our focus to outpatient care and wanted to address the uptake of different strategies of screening for diabetes and hypertension. Our primary research question assessed whether there were any significant differences in follow-up at the public sector clinic after screening positive via home-based screening (community health volunteer provides screening at your home) versus community-based screening (a community wide event is established where people voluntarily show up to receive screening) in a rural setting. In the third paper, we sought to continue to improve aspects of outpatient care by describing our model for improving access to medications. Our primary research question focused on descriptively assessing the change in availability of essential medications before and after implementation of this model. In the fourth paper, we described and assessed our model for providing intensive diabetes follow-up remotely. Our primary research question focused on whether patients experienced statistically significant improvements in blood glucose control after participating in this service for six months. In the fifth paper, we brought together various elements of our prior activities to design and evaluate the community-based model of care called BIGPIC - Bridging Income Generation through grouP Integrated Care. The primary research question for this investigation was to identify the frequency with which patients who screened positive for diabetes or hypertension linked to care. Secondary research questions compared the linkage frequency observed with this model compared to a historical control, along with a descriptive assessment of the loss to follow up, and an assessment of whether this model led to statistically significant reductions in blood pressure after 1 year of implementation. Short Summary of the Individual Papers with Results Linking Them Together Within our assessment of pharmacy training, we found that the Kenyan pharmacy interns provided statistically significantly more clinical interventions per day than their North American counterparts. This result highlighted the potential for Kenyan pharmacy providers to provide clinical services which were largely unavailable in western Kenya prior to this research. Despite the lack of the clinically focused Doctor of Pharmacy curriculum in Kenya, Kenyan pharmacy interns within the Bachelors in Pharmacy program were able to make an average of 16.7 consultations per day with the medical team compared to 12.0 per day for the North Americans. In the second paper we shifted our focus to the outpatient setting and were surprised to find that there weren’t any statistically significant differences in follow-up between home-based versus community-based screening for NCDs. This highlighted the reluctance of rural patients to travel to public sector facilities for care regardless of the screening method utilized. This realization led us to simultaneously focus on improving the reliability of services available in public sector while also trying to implement solutions to facilitate the provision of remote services for care. Within our efforts to improve medication access in paper 3, we were able to demonstrate how our revolving fund pharmacy model was able to improve access to medications from < 40% to > 90%. In paper 4, we were able to implement a self-monitored blood glucose program and demonstrate a dramatic improvement in the blood sugars of patients enrolled in the self-monitored blood glucose program with a statistically significant 31.6% absolute decline in HbA1c. The culmination of these efforts and learnings is described in paper 5, where we implemented the BIGPIC care delivery model which resulted in a statistically significant improvement in linkage to care for screened patients, a retention in care frequency of 70.3%, and a statistically significant mean decline in the systolic blood pressure of 21mmHg (95% CI 13.9-28.4, P < 0.01).

Background: Although there is a growing body of literature surrounding new psychoactive substances (NPS), and reasons for general use have been described, there is little understanding as to why certain NPS spread through user populations and become popular. This research used Rogers’ 1962 diffusion of innovations theory (DOI) to help better understand the NPS market and how it is shaped and characterised. Objective The aim of this research was to explore the diffusion of NPS in the UK and why different NPS diffuse and others fail to do so, to identify appropriate public health interventions to reduce harm. Methodology: A mixed methods approach was undertaken, which comprised four studies. The first study involved a critical analysis of the appropriateness of Rogers’ DOI to explain the diffusion of NPS. This study was followed by two sets of interviews. The first interview study was conducted with NPS online retailers based in the UK. The second interview study involved interviews with NPS professionals including law enforcement professionals, drug policy organisations and NPS early warning system representatives from the UK, wider Europe, America and Australasia. These findings were analysed using thematic analysis. The final study was an online questionnaire and choice-based conjoint analysis with UK pre-existing recreational drug users aged between 18 and 35. These findings were analysed using Latent Class Analysis. Results: The DOI was found to be applicable for the diffusion of an NPS product. However, the theory should be used in application to different individual NPS; NPS should not be classed as a homogenous group of substances and NPS users should not be classed as a homogenous group either. It was found that the theory should be updated in relation to NPS to include the influence of the internet. The key reason for the diffusion of an NPS was found to be the psychopharmacological effects of a product. However, there should also be an acknowledgment of the importance of friendship networks, and increasingly online forums. Even if a product had the desired psychopharmacological effects, if these are not communicated then it is unlikely to diffuse at a fast rate. Conversely, unless a product had the psychopharmacological effects desired by an individual, despite positive feedback from friends and online forums, it would be unlikely to diffuse. The emergence of NPS did not have a transformative effect for all drug-using groups; instead, it affected different user groups in different contexts. Similarly, it is likely that the introduction of the UK Psychoactive Substances Act will not have a transformative effect on the use of NPS by all drug-using groups. Nevertheless, the changes in health and social harms associated with individuals accessing NPS through the underground market or choosing to use traditional illegal drugs should be recognised. Finally, the need to conduct research with a range of stakeholders, to gain a greater understanding of motivations for drug use to assist with future public health interventions, was an important finding of the thesis.

Background: There is strong evidence linking obesity to health issues and long term conditions such as high blood pressure, type II diabetes, heart disease and some cancers (NICE, 2006). Despite this growing evidence base, the prevalence of obesity continues to rise and rates have more than doubled in England in the last 25 years (Public Health England, 2014). Currently 26% of adults are obese (Health Survey for England, 2014), and the proportion of women that are classified as overweight and obese has risen to 57% (Health Survey for England, 2014). Weight loss can reduce the risk of an individual developing these conditions and can increase their healthy life expectancy. It is estimated that approximately one in every two adults in England are actively trying to lose weight, the majority of which are over-weight or obese women (Piernas, Aveyard and Jebb, 2016). This indicates that over-weight and obese women are motivated to lose weight but are struggling to achieve or maintain a healthy weight. To-date potentially effective weight management interventions have been identified as long term multi-component interventions including diet and exercise components along with behavioural strategies. However, weight changes have been small and weight regain has been found to be very common (Loveman, Frampton, Shepherd, Picot, Cooper et al, 2011; Dombrowski, Knittle, Avenell, Araújo-Soare & Sniehotta, 2014). Despite the primary focus on weight loss being to improve health, research suggest that people’s prime motivation to lose weight is unrelated to health (Piernas, Aveyard and Jebb, 2016). Aim: This study aims to improve the understanding of the challenge of achieving and maintaining a healthy weight in overweight and obese young women. In particular, it aims to i) better understand the barriers and facilitators to achieving and maintaining a healthy weight as experienced by these young women, ii) further understand the relationships and influences of these factors, to iii) inform and develop a new theoretical framework in which to capture this social phenomena and societal challenge. Recruitment: This study recruited 14 female participants aged 18- 35 years, with a BMI over 30 (or 28 with co-morbidity), actively seeking support to lose weight. Participants were recruited through purposive sampling in two primary care practices in East London as part of the ‘Peer Support Weight Action Programme’ (SWAP). This was a Randomised Controlled Trial run by Barts Health NHS Trust and Queen Mary’s School of Medicine and Dentistry, funded by the National Institute for Health Research, Health Technology Assessment fund. Design: The research is qualitative in design utilising in-depth semi-structured interviews. Interviews took place with women recruited to take part in a weight loss programme before commencing the intervention, and follow up interviews took place approximately six months after completion of the weight management programme. Grounded Theory Analysis was used to analyse the data. Results and Findings: An overarching theoretical framework is presented from the findings of the data analysis of the pre and post weight loss attempt interviews. A new ‘Emotion and Mindset’ model is presented to explain the challenge of achieving and maintaining a healthier weight in young women. It includes the core categories of sense of self, emotion and mindset, self-efficacy, and stress and conflicting priorities and has been theoretically framed around the concept of Finding the Health Enhancing Equilibrium - maintaining a positive sense of self whilst generating action to achieve and maintain a healthy weight. It describes the balancing act required between these key contributing elements to engage in positive health behaviour which contributes to achieving and maintaining a healthy weight. Recommendations: Based on the findings from this study, and supported by previous findings (Cochrane, 2008), weight management interventions targeting young women need to build in coping strategies to support individuals cognitively, behaviourally and emotionally. These should include building self-efficacy (NOO, 2011; Ashford, Edmunds, French, 2010), sense of self and re-aligning identity (West & Brown, 2013). Individuals need to be taught how to identify, address and re-orient dysfunctional thoughts, to identify potential stressors such as triggers and environmental cues to prevent relapse. Consideration needs to be given to weight loss maintenance and ongoing tailored support. Further research is needed to identify what type and method of support is most effective and for whom.

Perinatal mental illness is a global health issue with detrimental outcomes for women and their families if left untreated. Unfortunately, many women do not get the treatment they need for many reasons, one often acknowledged reason is that the stigma some women experience inhibits disclosure of their needs to healthcare providers. This thesis looked at Internet resources for women with perinatal mental illness, in particular online forums. It examined how forums might affect stigma and thus disclosure behaviour. This thesis is by prospective publication. Article 1 aimed to describe and interpret qualitative studies regarding forum use and perinatal mental illness stigma. A metasynthesis of five studies identified four key themes: a safe place to talk; virtual support; stigma and identity; and repair of the mother identity (Moore, Ayers & Drey, under review). Article 2 aimed to identify what websites about postnatal mental illness were available and assess them for content and quality. A systematic review of 114 websites evaluated accuracy of information, resources and website quality. Results showed information was largely incomplete and difficult to read; resources were limited and website quality was variable (Moore & Ayers, 2011). Article 3 aimed to determine how women with perinatal mental illness use web based resources. A qualitative interview study (n= 15) found that the anonymity and non-judgemental social support on forums may have made it an acceptable way to challenge internal stigma and that most women described forums as providing a space to discuss stigma and test out disclosing about their illness to others (Moore & Ayers, 2016). Article 4 therefore aimed to identify whether forums for perinatal mental illness reduce stigma and facilitate disclosure. Thematic analysis of 1546 posts over six months on a forum for postnatal mental illness suggested that forum discourse reconstructed ideology of motherhood as compatible with perinatal mental illness. Many women overcame stigma and posted that they had taken advice and disclosed to a healthcare provider (Moore, Ayers, & Drey, 2016). The final article developed and tested a hypothesised model of the relationship between stigma and disclosure about perinatal mental illness. Study 5 developed a questionnaire measure of stigma for perinatal mental illness in order to test the model. Questionnaire items were completed online by women with perinatal mental illness (n=279). Psychometric testing suggested it was a valid scale with three subscales: external, internal and disclosure stigma (Moore, Ayers, & Drey, 2017). Study 6 tested a hypothesised model that stigma would mediate the relationship between forum use and disclosure to healthcare providers. An online survey of women with perinatal mental illness (n=200) who had used forums provided partial support for this hypothesis, with internal stigma mediating the relationship between length of forum use and disclosure (Moore, Drey, & Ayers, 2017). This research highlights the importance of considering the stigma associated with perinatal mental illness and its role in online forum use and disclosure. Overall, findings suggest that forums may facilitate recognition of stigma, which may in turn lead to greater disclosure of symptoms to healthcare providers. However, the relationship between forum use, stigma and disclosure may be more complex than our initial model proposed. Similarly, most participants in these studies were white, well-educated and actively participated in the forums. Future studies would benefit from testing these relationships using longitudinal designs with more representative samples.