The literature review sought to identify whether there are any differences and risk factors associated with individuals who present to health services with 'medically unexplained symptoms'. A number of relevant factors were identified, although they were not consistently found across all the studies examined. Issues relating to methodology, classification, terminology, and social context are also considered, as are directions for future research. The empirical paper explored whether there might be differences between two groups of pi lots, both with a history of flying aircraft associated with 'contaminated air' or 'fume events'. (These can occur when unfiltered 'bleed air' used to pressurise the cabin becomes contaminated with heated (pyrolised) engine fuels, lubricants, and other toxic substances.) One pilot group, drawn from Mackenzie Ross' (2008) study, had sought out or had been referred to medical health professionals regarding symptoms of ill health, and attributed these to contaminated air, for which no causal link has yet been established. The other group ofpilots had not come forward reporting adverse health effects, nor had they been referred to medical experts, even though they continued to work in the same potentially hazardous environment. Differences between these two groups were explored by comparing pilots' cognitive function, symptoms of ill health, and symptom attribution. Although pilots not reporting symptoms of ill heath were found to perform significantly better on cognitive tasks than those who did, a closer inspection of the data revealed a number of interesting results that may be relevant for future research, such as a distinct and unusual similarity in the pattern of the cognitive profiles of both pilot groups. No significant differences were found with regard to symptom reporting and attribution, although there seemed to be a small emerging trend towards cognitive decline over time (independent of age). The critical appraisal discusses the methodological and other limitations of undertaking this study, draws attention to questions raised by the findings, and argues the case for more (larger scale) research to be undertaken in the future. It should be noted that this study was undertaken under the umbrella of a larger research project that was seeking to examine the effects of contaminated air on cognitive function in commercial airline pilots. As such, the study will to a degree cover some similar ground with regard to the measures used, data gathering, and testing procedures as Mackenzie Ross (2008), from which the data for the self-selected group is drawn, and Madeley (2008)1, who constructed the health questionnaire and clinical interview protocol (see Appendices).

The research employs the theoretical lens of human ecology of disease to examine the ecology of malaria in Lagos state, Nigeria. As a first step I examine the spatial and temporal trends in clinical malaria infection using a density-based algorithm to identify two locations (Ikeja and Kosofe LGAs) with one of the highest malaria infection rates and ecologically diverse terrain. They form the focus of this research. I gather data and derive measures on 26 theoretically relevant environment and socio-cultural risk variables in a cross-section of 208 households using mixed methods that comprise semi-structured interviews, a questionnaire, environmental observations, GIS and remote sensing data and GPS mapping. Through these efforts, I build a household spatial database. I assess the contributory influences of the risk variables through the development and assessment of ten ecologically relevant candidate models of urban malaria using statistical and GIS analysis. I also engage with the everyday lives of the households and qualify the quantitative relationships. Findings reveal that the most parsimonious candidate model is grounded on the human ecology of disease principle. While many of the variables are not statistically significant, some, such as travel history, animal presence and household size, are of public health importance. One important finding emerges. The risk variable “working at night without mosquito protection”, though it does not appear in this model, seems to be important across other models. I examine it further and note that its risk within households is higher than those associated with residential locations. In fact, households inhabit low-risk locations and have low vulnerability risk rates. This suggests that in urban areas, infection likely occurs outside homes and mostly from places of work or social gathering, and coincides with older household members rather than vulnerable children. This research suggests further insights for urban-like occupations and behaviours.

Background: Despite immense global malaria prevention efforts, the disease remains a major cause of morbidity and mortality in sub-Saharan Africa. An integrated approach to malaria prevention, which advocates the use of several malaria prevention measures holistically, is being explored to reduce the occurrence of the disease. Aim: The aim of the thesis was to assess experiences, perceptions and practices on the integrated approach to malaria prevention in Wakiso district, Uganda. Methods: Study I was a pilot project that promoted the integrated approach and involved 3 phases. Phase 1 (baseline) was cross-sectional, and determined knowledge and practices on malaria prevention. Phase 2 (intervention) implemented an intervention on the integrated approach. Phase 3 (evaluation) was a cross-sectional impact evaluation of using the integrated approach. Study II was a clustered cross-sectional survey that assessed perceptions, utilisation and barriers of integrated malaria prevention. Results: The majority of participants (64.6%) had low knowledge on malaria prevention methods, with mosquito nets (81.7%) most known. Insecticide-treated nets were the most used method with 45.5% of households owning at least one net (Study I phase 1). The pilot project trained 25 community volunteers, sensitised over 200 community members, and established 40 demonstration households (Study I phase 2). There was improvement in knowledge on removal of mosquito breeding sites (51% versus 7%) in the evaluation in comparison with the baseline respectively. Improvement in practices in the evaluation compared with the baseline included mosquito screening in windows and ventilators (χ2 = 62.3; p < 0.001). The benefits reported by the demonstration households included reduction in mosquito populations and occurrence of malaria (Study I phase 3). If trained, most participants (68.6%) would use all methods in the integrated approach. Only 33.0% households were using the integrated approach, which was associated with reading newspapers (AOR 0.34; 95% CI 0.22 – 0.53) (Study II). Conclusion: Stakeholders involved in malaria control should intensify efforts of promoting multiple malaria prevention methods.

The prevention of healthcare‐associated infections (HCAIs) is a major challenge currently being faced by hospitals in both the UK and worldwide. The hospital environment acts as a reservoir for nosocomial organisms contributing towards the transmission of bacteria and thus the colonisation and infection rates of the patient population. Therefore, it is desirable to implement measures to decrease the microbial load within the hospital environment as a whole and particularly on frequently touched surfaces. Antimicrobial coatings could be applied to these surfaces, and used as an adjunct to other infection control policies to reduce the incidence of HCAIs. Novel nitrogen‐doped, sulfur‐doped and silver‐coated titanium dioxide photocatalytic thin films were generated by sol‐gel or chemical vapour deposition. The materials exhibited antibacterial properties after exposure to a white light commonly used in UK hospitals. However, it was difficult to synthesise reproducible thin films using the CVD method of deposition. An additional antibacterial material was generated with the potential to be used in endotracheal tubes to reduce the incidence of HCAIs such as ventilator‐associated pneumonia. The novel polymer was impregnated with a photosensitiser using a swell encapsulation method, and activated with laser light; the antibacterial and anti‐adhesive properties were then assessed. Sampling the test surfaces by swabbing and subsequently performing viable counts was shown to provide an adequate estimate of concentration of bacteria on a test surface. The nitrogen‐ and sulfur‐doped titanium dioxide coatings displayed significant photocatalytic activity against Escherichia coli after exposure to a white light source, which demonstrate d a shift in the band gap from the UV to the visible region of the electromagnetic spectrum. Visible light photocatalysis was confirmed on the silver-coated titania thin films when a UV filter was used to block out the minimal UV component of the white light source, in the form of photo‐oxidation of stearic acid, a reduction in the water contact angle and photocatalytic activity against an epidemic strain of meticillin resistant Staphylococcus aureus (EMRSA‐16). This is the first example of unambiguous visible light photocatalysis and photo‐induced superhydrophilicity alongside a titanium dioxide control that shows no activation. A reduction in the viability of EMRSA‐16 adhered onto the surface of the irradiated silver‐coated titania thin films was also demonstrated. A significant reduction in the recovery of Pseudomonas aeruginosa, Stenotrophomonas maltophilia, Acinetobacter baumannii and Candida albicans was observed on TBO-impregnated polymers, after irradiation with a HeNe laser light. A recently isolated clinical strain of P. aeruginosa showed decreased susceptibility to the photo‐activity of the TBO‐impregnated polymers compared with a laboratory type strain. Finally, a significant reduction in the adhesion of P. aeruginosa on the TBO-impregnated polymers was demonstrated after a 3‐step irradiation schedule. A photo‐bleaching effect was noted after light exposure that reduced the antibacterial activity of the polymers, which demonstrates the requirement for further modification to retain the photosensitiser within the polyurethane matrix. These novel materials have the potential to be used as anti‐microbial surfaces in healthcare environments.

Background: New malaria vector control tools hold the promise of sustaining gains in malaria control achieved to date and achieving the goal of elimination set for 2030. However, insecticide resistance has the potential to derail these malaria control achievements. Access to innovative vector control tools is key to surmounting the threat of insecticide resistance and will play a major role if malaria elimination is to be achieved. The aim of this thesis is to gather new evidence and provide insight into strategies for accelerating access to new malaria vector control tools. This is done by examining access to new malaria vector control tools in two national settings (Nigeria and Burkina Faso) as well as at the global level. Methods: Three retrospective policy analyses were carried out using an analytical framework to guide the selection of key informants (KI), data collection and analysis. Semi-structured interviews were carried out with KIs in Nigeria (2013), Burkina Faso (2014) and at the global level (2014). Interviews were conducted in English (French in Burkina Faso) audio recorded, transcribed and entered into NVivo10 for data management and analysis. Data were coded according to the framework themes and then analysed to provide a description of the key points and explain patterns in the data. Results: A total of 40 interviews were conducted with policymakers, researchers, donors, multilaterals, Non-governmental organizations and private sector. The synthesized findings of the three case studies show that, in the context of insecticide resistance, the evidence required to facilitate policy change is nuanced and context specific; national policymaking may be well defined and appear to be evidence based, but can be open to being circumvented and hindered by inefficiencies in global policymaking and lack of donor funding; price rather than cost-effectiveness is the key financial variable at the national level; and no readily identifiable policy champions exist to facilitate global and national adoption of new vector control tools. Conclusions: This thesis has identified five areas that need to be strengthened in order to facilitate access to new malaria vector control tools by fostering their global and national adoption. The thesis demonstrates that, without a well-coordinated architecture to: facilitate the development of robust and appropriate evidence; support a transparent and timely global policymaking process; diversify the available funding base, and facilitate price reductions without stifling innovation, accelerating access to new vector control tools and achieving malaria elimination goals is unlikely.

The prevention and control of infectious disease continues to be a central priority for human public health. Two infectious agents, the bacterium Yersinia pestis and the filarial nematode Onchocerca volvulus have been responsible for significant public health burden in many populations, and challenges remain in the control and prevention of them both. With known environmental and climatic drivers, vector based transmission and in the case of Y. pestis, zoonotic reservoirs, fully characterising the ecological context of both diseases is vital to properly understanding their epidemiology. In the present study, ecological field data is utilised to investigate this, towards improving predictions and designing control strategies. Despite being contrasting infectious agents, the role of spatial scale and resolution are identified to be vital aspects of both investigations. Ultimately, the importance of a holistic approach to infectious disease epidemiology is highlighted across both systems. In humans and some other animals, infection with Yersina pestis causes the disease plague. Although human infections are now limited to the hundreds annually, they disproportionately affect individuals in lower income settings and as such are still of public health concern. A number of wild rodent species tolerant to infection with Y. pestis act as reservoirs for the bacterium, from which it can transmit into less tolerant susceptible populations. To this end, much research has focused on predicting periods of high infection prevalence (epizootic outbreaks) in the reservoir species in order to predict and prevent transmission events into less tolerant populations. One such reservoir species is the great gerbil, Rhombomys opimus, residing across the pre Balkhash desert in Kazakhstan. Y. pestis prevalence in this population is known to be strongly associated with the gerbil population size. The present study aims to resolve current ability to predict the gerbil population. Burrow level environmental properties are identified to be predictors of gerbil presence at, and long term occupancy of burrows. Gerbil occupancy of burrows could not be replicated using standard metapopulation approaches, nor related to the connectivity of the burrow. Atypical strains of Y. pestis lacking the fraction 1 (F1) surface antigen are isolated and found to be common in the gerbil population. This antigen is highly immunogenic and has a role in virulence determinance. However, no significant hetereogeneities in their distribution or associations with epidemiological outputs could be identified. In several African countries with endemic levels of infection with Onchocerca volvulus, an unusually high prevalence of epilepsy has been observed in the population. Here cross sectional data from a population in the Democratic Republic of Congo is used to build a picture of epilepsy prevalence and to explore a possible functional relationship between epilepsy and O. volvulus. The relationship between O. volvulus and epilepsy is strengthened, and spatial and ecological data lead to the suggestion that responsible drivers have local, rather than global relationships with epilepsy.

Lymphatic filariasis (LF) and onchocerciasis are highly endemic in Sierra Leone. Using World Health Organization (WHO) guidelines for monitoring national programmes where both infections are co-endemic, this study aimed to determine the impact of preventive chemotherapy on transmission intensity by measuring changes in human infection status using standard epidemiological indicators. Separate longitudinal studies designed to deliver WHO outcomes for programmes targeting the elimination of both diseases were conducted. Onchocerciasis mapping surveys from 1988-2005 revealed that twelve of fourteen health districts were endemic. The baseline average mf prevalence was 53.1%, and mf densities in positive-only or entire populations were 28.87 and 15.33 mf/snip, respectively. Mf prevalence and density increased with age and was higher in males than females. Baseline prevalence and intensity surveys showed that LF was endemic in all 14 districts (Wuchereria bancrofti antigenaemia prevalence > 1%). Mean LF prevalence by ICT cards was 21% (males 28%; females 15%) with higher prevalence in the northeast (Bombali 52%; Koinadugu 46%; Tonkolili 37%; Kono 30%) and lower in the southwest (Bonthe 3%; Pujehun 4%). Mf prevalence was also relatively higher in the northeast (Bombali 6.7%; Koinadugu 5.7%; Port Loko 4.4%; Kono 2.4%). Mf prevalence was higher in males (males 2.9%; females 1.8%) and infection rate was higher in the over 20 years age-group (2.5%) than younger (1.7%). Arithmetic mean mf density was 50.30 mf/ml among mf-positive individuals and 1.19 mf/ml in the population examined. Nationwide mass drug administration (MDA) using ivermectin plus albendazole was applied to eliminate both diseases. In 2010, after five rounds of MDA (2005-2009) with effective treatment coverage for onchocerciasis during 4/5 years, overall onchocerciasis mf prevalence was reduced by 60.26% (from 53.10% to 21.10%), overall mf density among positive-only individuals was reduced by 71.29% (28.87 to 8.29 mf/snip) and overall mf density among the entire population studied was reduced by 88.58% (15.33 to 1.75 mf/snip). Mf prevalence and density were higher in males, lowest in the 1-9 and highest in the 40-49 year age groups. Mf prevalence was reduced by >50% in 10/12 districts, and reduction in skin mf density was ≥50% among positives-only in 11/12 districts. After MDAs with effective treatment coverage in 2008-2010, LF mf prevalence decreased to less than 1% in 11/12 districts. Mf prevalence fell by 88.5% to 0.3%, with decreases of 70-95% in seven and 100% (0 prevalence) in four districts, respectively. Overall arithmetic mean mf density after three MDAs was 17.59 mf/ml among mf positive individuals and 0.05 mf/ml for the entire population examined. After five MDAs, the overall mf prevalence was 0.54% and was higher in males (0.7%) than females (0.36%). Eight of twelve districts with < 1% mf prevalence passed the pre-transmission assessment survey (TAS) and therefore qualified for a TAS to determine whether MDA could be stopped. Four districts failed the pre-TAS: Koinadugu (0.98% i.e. close to 1%), Bombali (2.67%), Kailahun (1.56%) and Kenema (0%). Following WHO recommendations, Kenema and Kailahun districts were paired to form a unit of approximately one million. Kenema, the spot check site, was considered to have failed the pre-TAS even though the mf prevalence was 0% because Kailahun, the sentinel site, failed. A qualitative study examining the impact of the Ebola virus disease (EVD) outbreak on the NTD programme found that despite a one-year absence of interventions, two rounds of MDA had been completed, including one during the ongoing outbreak in May/June 2015. Although it compromised the likelihood of achieving the 2020 targets of LF elimination and Onchocerciasis control, the EVD outbreak has enhanced awareness about the important role of community volunteers in ensuring its success. While it may be the ‘endgame’ for LF, the NTD community and collaborating research institutions must address additional challenges if Onchocerciasis is to be eliminated from Sierra Leone.

Background: This thesis presents and discusses the findings of a study that assessed health care seeking pathways and experiences in accessing antiretroviral treatment and continuing with treatment. The thesis explores how patients' experiences in health care were mediated by socioeconomic factors such as gender, poverty and rural/urban residence. The thesis aims to answer four main research questions: • What is the health care seeking behaviour of patients leading to the time they are eligible for antiretroviral therapy (ART) and begin the antiretroviral drug initiation process? • What are the associated direct and indirect costs of care seeking and are there inequalities in how the costs impact on different social economic groups? • Are there patients who do not initiate treatment despite knowing that they are eligible for treatment and if so, why do these patients not start treatment? • Among the patients who are lost to follow up or die whilst on ART, what are the factors that contribute to attrition? Study Methods: The study employed a mixed quantitative and qualitative research design. A survey was conducted among males and females identified as eligible for ART in seven health facilities in urban and rural Lilongwe. Patients were then followed up to identify those who started treatment or did not and those who were continuing with treatment or dropped off. Follow up qualitative interviews were conducted with different groups of patients and health workers. Findings: Patients make multiple visits to different health providers ranging from selftreating, to visiting private clinics and public hospitals and health centres which result in a delay to HIV testing and ultimately delay in initiating ART. Inequities exist in how patients experience care seeking for HIV and ART services. The health care seeking pathway is influenced by socio-economic factors which in turn affect patient treatment outcomes. The type of health facility and the direct costs of care seeking in particular have an influence on whether patients start treatment or not. Patients' perceived severity of illness plays a role in determining whether patients start or remain in HIV care. Male sex and poverty were associated with patients' loss to follow up and deaths while on ART. From the qualitative interviews, long distance to health facilities, psychosocial problems and the challenges of disclosure interplayed with other factors to cause patients to interrupt treatment. Discussion: The pathway to care seeking is influenced by the interactions at three levels - the community, the patients and their households and the health system. These interactions are mediated by gender, poverty, residence, illness meaning and stigma to determine a complex and long health care seeking behaviour, delays in HIV diagnosis, missed opportunities for HIV testing and pre-ART care, drop out from the health care seeking pathway, early mortality and survival or attrition from ART. Conclusion and recommendations: To address existing inequities in ART access, there is a need to increase the availability of services to the rural poor and strengthen health systems. Supporting community based systems have the potential to bring services closer to poor communities. However, Malawi needs to consider revisiting the ART equity policy in a way that will create new financing opportunities for HIV treatment.

Due to an increased risk of anal cancer and genital warts, the UK Joint Committee on Vaccination and Immunisation recommended targeted HPV vaccination programme for men who have sex with men (MSM). The research in this thesis aimed to examine the feasibility and acceptability of HPV vaccination for MSM in the United Kingdom. Accordingly, two systematic reviews and three empirical studies were conducted to address the thesis objective. Two reviews synthesised evidence on factors associated with hepatitis A-B and HPV vaccine acceptability and uptake amongst MSM. The first study explored knowledge and beliefs about HPV and HPV-related diseases as well as attitudes towards the HPV vaccine amongst MSM in the UK, using qualitative methods. The second study examined psychological and behavioural factors associated with HPV vaccine acceptability amongst MSM, using quantitative methods. The third study investigated attitudes of sexual healthcare professionals (HCPs) towards a targeted HPV vaccination programme for MSM in the UK, using mixed methods. The majority of MSM did not know about HPV and their risk of contracting genital warts and anal cancers. HPV infection was perceived as a female problem and HPV vaccination was thought to be against cervical cancer. While only 55% of MSM would be willing to ask for the HPV vaccine, 89% would accept it if offered by an HCP. Access to sexual health clinics, the disclosure of sexual orientation to an HCP and HIV-positive status were positively associated with HPV vaccine acceptability. Perceptions of HPV risk, HPV infection severity, HPV vaccination benefits, HPV vaccine effectiveness, and the lack of perceived barriers to HPV vaccination were also associated with acceptability. Around 14% of HCPs were already vaccinating men against HPV, 83% recommended gender-neutral HPV vaccination and 65% recommended targeting MSM. Although nearly half of MSM would not actively pursue HPV vaccination, the vast majority would accept the vaccine if recommended by HCPs. MSM need to be informed about HPV to appraise the benefits of HPV vaccination for their health. In order to achieve optimal uptake, vaccine promotion campaigns need to focus on MSM that do not access sexual health clinics and those unwilling to disclose their sexual orientation. Clear advice and guidelines on HPV vaccine use for men at sexual health clinic are required to ensure equitable opportunities for vaccination.

In developing countries, malaria treatment is often inadequate, notably in retail shops where the majority of people seek care. Shopkeepers are the last link in a chain of wholesalers who have an influence on treatment availability, price and quality. Evidence on competition in retail and wholesale markets is scarce, partly due to the methodological challenges of studying healthcare markets in poor countries. The thesis investigates how market structure, provider conduct, customer demand and regulation affect malaria treatment outcomes in Cambodia. In addition the thesis contributes to the development of methods for studying private drug markets. Cross-sectional surveys and semi-structured interviews of representative samples of antimalarial retailers and wholesalers were conducted to collect data on provider practices and perceptions. The contribution of different empirical methods for identifying and sampling wholesalers and measuring sales volumes was also assessed. Private commercial providers supplied the majority of antimalarial drugs, reflecting the relative proximity, long opening hours, reliable drug stock and friendliness of private retailers. Retail and wholesale competition increased accessibility to malaria treatment but did not lead to optimal supply of affordable quality treatment. Several market failures were evident: intense product differentiation, high concentration, and imperfect consumer information on treatment quality. These provided opportunities for higher mark-ups, although not in all market segments. With high market heterogeneity, higher retail mark-ups did not necessarily translate into higher consumer prices, highlighting the influence of distribution chain structure and wholesaler's price setting decisions. Government failures were also frequent, with poor public sector treatment accessibility and ineffective regulation'. Recommendations include widening distribution networks for artemisinin combination therapy and rapid diagnostic tests; improving product stock reliability; decreasing wholesale and retail product prices; intensifying providers' training; diffusing information to consumers on what constitutes appropriate management of malaria fever; and strengthening regulation and the potential to extend its supportive role.