The constant public demand for high-quality medical services drives an associated demand for professional doctors, and requires them to take high-stakes exams. MRCP(UK) is a major examination for physicians in the UK, which aims to assess their knowledge, skills and appropriate attitudes – all key aspects of being a medical professional. Although the existing literature provides extensive evidence supporting the quality of MRCP(UK), this research aimed to add to the existing body of knowledge by investigating the predictive validity of MRCP(UK), i.e. examining whether it truly selects candidates who possess the above-mentioned qualities. This research therefore investigated the relationships between MRCP(UK) scores and results of seventeen knowledge exams and two clinical skills assessments (including specialty exams and MRCGP), training performance assessment outcomes (ARCP), and cases of licence limitations and erasures. Operating with the hypothesis that MRCP(UK) would predict all of the above-mentioned criteria, a retrospective longitudinal approach was assumed. The main sample contained records of 50,311 MRCP(UK) candidates attempting MRCP(UK) between May 2003 and January 2011; however, the analyses were performed on smaller samples, from 8 to 33,359 cases, depending on the size of the criterial dataset. The results of univariate and multivariate analyses supported the hypothesis. MRCP(UK) scores were indeed predictive of results of all knowledge exams and clinical assessments (meta-analysed average effects: r=0.69 for Part I, r=0.70 for Part II, 0.48 for PACES), and of performance in specialty training and issues with the licence to practice (on average: r=0.24 for Part I, and r=0.22 for Part II and PACES). The magnitudes of these validity coefficients were consistent with the theoretical notions of psychometrics and concurred with the findings of published studies. In view of the evidence it was concluded that MRCP(UK) is a valid exam. The limitations of this study, directions for future research, and general implications were discussed.

Purpose: It has been reported that over 33 % of patients with OAB present with pyuria (≥10 wbc μl-1) on urine microscopy, but under a third of these have bacteriuria. To clarify this situation, an accomplished comparative scrutiny of the inflammatory state of the urothelium in OAB, was carried out. Materials and Methods: This was a prospective, blinded, observational study of idiopathic OAB patients compared with controls. CSU samples were obtained and submitted to fresh urine microscopy, routine culture and microscopic examination of spun sediment. Another sample of OAB patients and asymptomatic controls provided cystoscopic bladder biopsies for histopathology and electron microscopy. Results: 178 OAB patients and 21 controls provided spun sediment samples. 75 (42 %) of these OAB patients had microscopic pyuria (.1 wbc ƒÊl-1) with 25 of which (33 %) were culture-positive. None of the controls had pyuria, nor bacteriuria. In a 20 mm2 spun deposit there was an average of 48 wbc (95 % CI 26 to 589) in OAB with pyuria; 12 wbc (95 % CI 10 to 15) in OAB without pyuria; and 4 wbc (95 % CI 3 to 7) in controls. Biopsies from 79 OAB patients, showed chronic inflammation and hyperplasia in 69 (87 %; 95 % CI=78 to 92: only 20 % had pyuria) and none in 5 controls. EM of 22 OAB patients showed increased basement membrane thickness compared to 2 controls (H=48, df=2, p<0.001). Conclusions: The study shows evidence of chronic cystitis and urothelial hyperplasia associated with OAB irrespective of pyuria or bacteriuria. The phenomenon has been confirmed by 3 different methods.

In Emission Tomography the design of the Imaging System has a great influence on the quality of the output image. Optimisation of the system design is a difficult problem due to the computational complexity and to the challenges in its mathematical formulation. In order to compare different system designs, an efficient and effective method to calculate the Image Quality is needed. In this thesis the statistical and deterministic methods for the calculation of the uncertainty in the reconstruction are presented. In the deterministic case, the Fisher Information Matrix (FIM) formalism can be employed to characterize such uncertainty. Unfortunately, computing, storing and inverting the FIM is not feasible with 3D imaging systems. In order to tackle the problem of the computational load in calculating the inverse of the FIM a novel approximation, that relies on a sub-sampling of the FIM, is proposed. The FIM is calculated over a subset of voxels arranged in a grid that covers the whole volume. This formulation reduces the computational complexity in inverting the FIM but nevertheless accounts for the global interdependence between the variables, for the acquisition geometry and for the object dependency. Using this approach, the noise properties as a function of the system geometry parameterisation were investigated for three different cases. In the first study, the design of a parallel-hole collimator for SPECT is optimised. The new method can be applied to evaluating problems like trading-off collimator resolution and sensitivity. In the second study, the reconstructed image quality was evaluated in the case of truncated projection data; showing how the subsampling approach is very accurate for evaluating the effects of missing data. Finally, the noise properties of a D-SPECT system were studied for varying acquisition protocols; showing how the new method is well-suited to problems like optimising adaptive data sampling schemes.

The rapid emergence of Antimicrobial Resistance (AMR) over the past decades together with a lack of research into new drugs presents health care systems with serious challenges and threatens their ability to effectively treat serious bacterial infections. As a result, it is realistic to expect that effective treatment options for some infections will run out in the future. The thesis begins by outlining the mechanisms and consequences of AMR and argues that AMR differs from other distributive problems, due to the specific characteristics of antibiotics. It is suggested that for considerations of distributive justice, antibiotic effectiveness should be treated as a resource, which can be depleted and which must be fairly distributed between people and generations. The thesis then goes on to examine the distinctive moral challenge posed by AMR. It begins by considering a consequentialist account, which suggests that AMR is a moral problem due to the bad health outcomes it entails. However, this approach is subsequently dismissed because it struggles to account for some of the particular features of AMR. An alternative is to consider AMR as a morally wrongful harm to individuals, which requires not only that AMR has adverse effects, but also violates the victim’s rights. It is shown that the harm caused by AMR is morally wrongful and that people have a right to be protected from adverse health outcomes, which AMR violates. However, it is difficult to specify correlative duties that result from such a rights claim. As an alternative, the thesis suggests and defends a form of Scanlonian contractualism, which offers the best model to represent and address issues of distributive justice in the case of AMR. It is shown that a principle of antibiotic use, which rules out the use of antibiotics for infections that do not pose a serious risk of irreversible harm, offers a convincing contractualist argument. The thesis examines the concerns for intergenerational justice that arise as a consequence of AMR and shows that contractualism is capable of addressing them. The thesis concludes by suggesting a new way of framing AMR as a specific type of policy challenge, which better captures its complexity and advocates a reduction of future dependency on antibiotics.

Reducing socioeconomic-related health inequalities has been a pressing concern for many years, and the focus of much academic research. Until now, however, there have been few e orts to quantify the contribution that health care makes to health inequality. Our contention is that if health care increases health, the distribution of health care a ects health inequality, therefore changes in the distribution of health care can be used to reduced health inequalities. We begin by reviewing the literature on the e ects of health care on health and health inequality. Next, we show how a cross-sectional model of the e ect of health care on health can be used to examine the contribution of health care to health inequality. Through this we nd that for the area-level e ect of spending on mortality in England, cancer spending leads to a signi cant reduction (approximately 50%) in cancer mortality inequalities but Coronary Heart Disease (CHD) spending does not. We move on to consider the limitations of modelling health inequality over time when using area-level data and suggest a methodological innovation, using a country-level analysis to demonstrate its application. Returning to the setting of English cancer and CHD mortality we show that the results of an analogous panel data model support our earlier ndings. Finally, we use individual-level data to examine the e ect that health spending has on health care utilisation, the e ect utilisation has on health, and how these e ects contribute to health inequality. We nd that here is no e ect on health from General Practitioner visits. Inpatient days do increase health and, overall, reduce health inequalities by 40%. These ndings show the inherent di culty of measuring the e ect of health care on health, but do suggest that health care is currently making substantial contributions to the reduction of health inequalities.

Limiting the severity of inflammation and promoting its eventual resolution are vital for protecting host tissues both in autoimmunity and inflammation. In this thesis, I describe a potent CD8+FOXP3+ regulatory T cell (Treg) population in patients with rheumatoid and psoriatic arthritis, which can be induced from peripheral blood mononuclear cells after stimulation with different clones of anti-CD3 mAb, including OKT3 and the non Fc-receptor binding antibody otelixizumab. CD8+FOXP3+ Treg inhibited Th17 responses, thereby functioning to limit a wider range of inflammatory pathways compared to CD4+FOXP3+ Treg. They also regulated CD4+ T cell proliferation and Th1 responses. Investigation by microarray into the mechanism of CD8+FOXP3+ Treg mediated suppression revealed several candidate surface markers and cytokines. During investigation into the mechanism of CD8+FOXP3+ Treg induction, I demonstrate a ten-fold induction, not expansion, of these Treg by anti-CD3 monoclonal antibodies. This was supported both by p38 phosphorylation, intrinsic to CD8+ T cells, and by signals provided by monocytes via CD86 and membrane TNF-α. Artificially increasing monocyte membrane TNF-α or inhibiting CD8+ T cell p38 phosphorylation drove FOXP3 expression in a subset of initially unresponsive CD8+ T cells. These data define a previously unknown mechanism of CD8+FOXP3+ Treg induction by anti-CD3 antibodies. Furthermore, anti-CD3 mAb could be combined with a p38 inhibitor to improve the mAb’s therapeutic efficacy and resolve chronic inflammation via the restoration of tolerance. These data also provide a rationale for the repurposing of anti-CD3 mAb as a therapy for inflammatory arthropathies.

Incontinence is a common health problem among the human population, especially females. Although there have been many efforts to develop cures not all sufferers can be fully cured. A way to deal with this large portion of incontinence sufferers is pads. The continuous usage of pads creates some problems, though. The most common cause of these problems is friction between pads and skin. In order to describe friction, David Cottenden developed a mathematical model for describing friction between a conformable sheet and a curved surface. Previous work has already validated the model for strips of nonwoven fabric on rigid convex prisms and low – half angle cones . The aim of this project was to extend the validation to (i) large – half angle rigid cones (whose surfaces approximate to portions of the body); (ii) human volar forearms and (iii) highly compliant cylinders. In the first part of the project I validated the model for an example nonwoven fabric on rigid (Plaster of Paris) cones with half angles of 25°, 35° and 45°. As predicted by the model, the data for all fabric footprints on all cones fell on the same master curve, within experimental error. In the second part of the project, I used the volar forearms of young and older female participants. In this way I had the opportunity to test the model on real human skin (smooth and wrinkled) and different substrates (firm and flaccid tissues) as they varied between young and older subjects. Moreover, I observed the changing geometry of arms during experiments, especially the behaviour of – often wrinkled and flaccid – older arms and see how the model responded. I used strips of five different nonwoven fabrics investigating not only how the substrate affected the model, but also how behaviour varied between fabrics. The agreement between experimental data and model predictions was excellent for all fabrics. In the third part of the project, I used the same five fabrics on compliant cylinders made of soft silicone membrane “skins”. These cylinders helped me investigate how the model responded for extreme deformations (rucking) which were much greater than humans could have tolerated. Again, agreement between experiment and model was remarkably good. In summary, all three blocks of experimental work provided further validation of Cottenden’s model, increasing confidence that it can be used in future work to understand friction over the curved surfaces of the body and help develop products kinder to the skin.

Diffusion magnetic resonance imaging (MRI) allows the probing of tissue microstructure in the brain non-invasively. The standard method for quantitatively analysing microstructure is diffusion tensor imaging (DTI) with a typical b-value of 1000 s/mm2. However, high b-value imaging (b > 2000 s/mm2) can increase the contrast between tissue types. The disadvantage of high b is the detrimental effect of decreased signal-to-noise-ratio (SNR) on diffusion models. This work begins by investigating the efficacy of exploiting the fitting error of DTI at b = 3000 s/mm2 to provide further information of brain microstructure using bootstrapped principal eigenvectors, µ1. This study revealed binomial distributions of µ1 with potential to be exploited to provide further information of the underlying microstructure. Another factor affecting SNR in diffusion-weighted data is the magnetic field strength of the MRI scanner. Change in DTI measures with SNR by way of increasing b-value and at 1.5 versus 3 T was systematically assessed. Significant differences in DTI measures between field strengths were found suggesting careful comparison of DTI data when acquired at 1.5 and 3 T. The limitations of DTI in modelling the slow diffusion component (fitting between b = 2000 to 3000 s/mm2) were also demonstrated. Other algorithms exist for modelling diffusion profiles without the assumption of diffusion following a Gaussian distribution. One such method is the recently developed multi-compartment model NODDI (Neurite Orientation Dispersion and Density Imaging). The final aim of this work is to further understand how this technique behaves with change in magnetic field strength and its within-subject reproducibility in relation to DTI. This investigation found significant field strength effect on NODDI estimates as well as reasonable white matter diffusion measurements in multi-fibre regions where DTI fails. However, given the complexity of NODDI, greater within- subject variability was found in white matter compared to the simpler tensor model.

Introduction: A portfolio of studies is presented aimed at understanding the clinical and pathophysiological significance of cerebral microbleeds (CMBs) in stroke patients. CMBs are the radiological marker of microscopic haemosiderin deposits on iron-sensitiveMRI sequences (mainly gradient-recalled echo [GRE] T2* MRI). They are common in patients with cerebrovascular disease and are hypothesised to be a biomarker for brain small vessel diseases, including hypertensive arteriopathy and cerebral amyloid angiopathy (CAA). Important questions relating to CMBs include their use as a prognostic marker for antithrombotic-related intracerebral haemorrhage (ICH) and cognitive impairment. Our aims were to address the pathophysiological and clinical relevance of CMBs using longitudinal, case-control and cross-sectional studies. Methods: Patients were ascertained from prospective databases of admissions to the stroke service at the National Hospital for Neurology and Neurosurgery and at University College London Hospital’s (UCLH) NHS Trust. Magnetic resonance imaging data was collected and analysed for markers of small vessel disease including CMBs. Clinical and radiological associations of CMBs were determined using appropriate statistical tests. Objectives: First, ways of improving microbleed detection and reporting were explored through the development of a visual rating scale (theMicrobleed Anatomical Rating Scale, MARS) aimed at reliably rating CMBs. Second, the prognostic relevance of CMBs was investigated for antiplatelet-related ICH in a case-comparison study. Third, the detection of new CMBs over time and the factors that influence this were explored. Fourth, the impact of CMBs on cognitive impairment was studied in a cross-sectional study. Finally, the association between CMBs and acute silent ischaemia on diffusion-weighted MRI was investigated via a multi-centre cross-sectional MRI study of patients with ICH. Main findings: 1. MARS is a reliable scale with good intra- and inter-rater agreement for rating CMBs presence and number in any brain location. 2. Lobar CMBs, especially if numerous, are a risk factor for antiplatelet-related ICH independent of the extent of white matter changes. 3. CMBs accumulate over time in stroke patients, and the risk is related to baseline systolic blood pressure. 4. Lobar CMBs are an independent predictor of frontal executive impairment; this suggests that CAA is a potential underlying contributor to cognitive impairment. 5. Silent acute infarcts are frequent in patients within 3 months of ICH, especially in those with probable CAA, and are associated with markers of small vessel disease severity, including CMBs. Conclusion: These studies provide new information on detection, clinical impact and associations of CMBs in stroke patients. They suggest that CMBs have useful roles in understanding pathophysiology, diagnosis and prognosis in patients with small vessel diseases. Further studies are required to determine the direct therapeutic consequences of CMBs, but the present work suggests several promising areas for future research.

Adoptive immunotherapy for cancer is a rapidly expanding field. Along with new techniques and technologies for cell engineering comes a pressing clinical need to discover the location of these cells after injection and to quantify the number of cells in a particular location. The use of agents to enhance tissue contrast is crucial to improve the ability of imaging techniques to distinguish cells from background signal, in order to track cells in vivo. Human donor T cells were transduced to express a tumour antigen specific T cell receptor. Transduced T cells were labeled with novel and conventional radiolabels for in vivo tracking using a combined single photon emission computed tomography and computed tomography (SPECT/CT) scanner. Transduced T cells were used in combination with TRAIL-expressing MSCs to produce a novel dual cell anti-cancer therapy in an in vivo lung metastases cancer model. The data in this thesis demonstrates that a novel tri-functional probe designed for long-term cell tracking, whilst resulting in superior cell labeling and retained activity over time compared with conventional methods, causes significant cell toxicity. It also demonstrates, for the first time, that a tumour antigen specific T cell therapy can be effective against lung metastases, leading to a significant reduction in tumour burden. These engineered T cells combined with MSC TRAIL also significantly reduce metastatic tumour burden, although there was no significant benefit to using the dual cell therapy.