There is a growing demand worldwide to increase the quality and productivity of healthcare services thereby increasing the value of the healthcare services delivered. To deal with these demands, increasingly importance is being placed on analysing and reducing unwarranted variations in healthcare services to achieve significant savings in healthcare expenditure. Unwarranted variations are defined as the variations in the utilisation of healthcare services that cannot be explained by variation in patient illness or patient preferences. Current modelling and simulation approaches for healthcare service efficiency and effectiveness improvements in hospitals do not utilise multiple types of heterogeneous service data such as qualitative information about hospital services and quantitative data such as historic system data, electronic patient records (EPR), and real time tracking data for analysing unwarranted variations in hospital. Consequently, due to the presence of large amount of unwarranted variations in the service delivery systems, service improvement efforts are often inadequate or ineffective. Therefore, there is urgent need to: (i) accurately and efficiently model complex care delivery services provided in hospital; (ii) develop integrated simulation model to analyse unwarranted variations on a care pathway of a hospitals; and, (iii) develop analytical and simulation models to analyse unwarranted variations from a care pathway. Current process modelling methods to represent healthcare services rely on simplified flowchart of patient flow obtained based on on-site observations and clinician workshops. However, gathering and documenting qualitative data from workshops is challenging. Furthermore, resulting models are insufficient in modelling important service interactions and hence the resulting models are often inaccurate. Therefore, a detailed and accurate process modelling methodology is proposed together with a systematic knowledge acquisition approach based on staff interviews. Traditional simulation models utilised simplified flow diagrams as an input together with the historic system data for analysing unwarranted variations on a care pathway. The resulting simulation models are often incomplete leading to oversimplified outputs from the conducted simulations. Therefore, an integrated simulation modelling approach is presented together with the capability to systematically use heterogeneous data to analyse unwarranted variations on service delivery process of a hospital. Maintaining and using care services pathway within hospitals to provide complex care to patients have challenges related to unwarranted variations from a care pathway. These variations from care pathway predominantly occur due ineffective decision making processes, unclear process steps, their interactions, conflicting performance measures for speciality units, and availability of resources. These variations from care pathway are largely unnecessary and lead to longer waiting times, delays, and lower productivity of care pathways. Therefore, methodologies for analysing unwarranted variations from a care pathway such as: (i) system variations (decision makers (roles) and decision making process); (ii) patient variations (patient diversion from care pathway); are discussed in this thesis. A system variations modelling methodology to model system variations in radiology based on real time tracking data is proposed. The methodology employs generalised concepts from graph theory to identify and represent system variations. In particular, edge coloured directed multi-graphs (ECDMs) are used to model system variations which are reflected in paths adopted by staff, i.e., sequence of rooms/areas traversed while delivering services. A pathway variations analysis (PVA) methodology is proposed which simulates patient diversions from the care pathway by modelling hospital operational parameters, assessing the accuracy of clinical decisions, and performance measures of speciality units involved in care pathway to suggest set-based solutions for reducing variations from care pathway. PVA employs the detailed service model of care pathway together with the electronic patient records (EPRs) and historic data. The main steps of the methodology are: (i) generate sample of patients for analysis; (ii) simulate patient diversions from care pathway; and, (iii) simulation analysis to suggest set-based solutions. The aforementioned unwarranted variations analysis approaches have been applied to Magnetic Resonance (MR) scanning process of radiology and stroke care pathway of a large UK hospital as a case study. Proposed improvement options contributed to achieve the performance target of stroke services.

Idiopathic pulmonary fibrosis (IPF) is a fatal form of fibrotic lung disease. The pathogenesis of IPF is unclear. An aberrant alveolar epithelial wound repair is likely to be involved in the disease process. Alveolar bronchiolisation, a process where bronchiolar Clara cells migrate into the affected alveoli, is a manifestation of abnormal alveolar wound repair. The role of Clara cells during alveolar injury repair in IPF is controversial. This study was undertaken to investigate the role of Clara cells in alveolar epithelial wound repair and pulmonary fibrosis. Currently, there is no curative treatment for IPF; therefore, stem-cell mediated regenerative therapy has been suggested. In this study, the paracrine role hMSC and hESC on pulmonary epithelial wound repair has also been evaluated. A direct-contact co-culture in vitro model was utilised to evaluate the role of Clara cells on alveolar epithelial cell wound repair. Immunohistochemistry was conducted on IPF lung tissue samples to replicate the in vitro findings ex vivo. The paracrine role of hMSC and hESC on pulmonary epithelial cells was evaluated by utilising the in vitro wound repair system. This study demonstrates that Clara cells induce apoptosis in AEC through a TRAILdependent mechanism, resulting in significant inhibition of wound repair. Furthermore in the IPF lungs, TRAIL-expressing Clara cells were detected within the fibrotic alveoli, together with widespread AEC apoptosis. This study also demonstrates that hMSC enhance AEC and SAEC wound repair via a paracrine mechanism through stimulation of cell migration; whereas, secretory factors of differentiated hESC promote AEC wound repair through stimulation of both cell proliferation and migration. Through this study I propose a novel hypothesis which implies that the extensive profibrotic remodelling associated with IPF could be driven by TRAIL-expressing Clara cells inducing AEC apoptosis through a TRAIL-dependent mechanism. My study also supports the notion of clinical application of hMSC and hESC or their secretory products as regenerative therapeutic modality for IPF.

Non-specific or medically unexplained physical symptoms (MUPS) are common among children, persist in considerable proportions of those affected, and can lead to primary care consultations. A systematic review in this thesis has provided limited evidence of an association between MUPS in parents and children. This thesis has investigated the association between GP consultation for MUPS in 5417 parent-child pairs registered with 12 GP practices, and examined whether this is related to persistent GP consultations for MUPS in children. One descriptive study, two case-control studies, and one prospective cohort study were conducted using GP electronic medical records. In children, the annual GP consultation prevalence for MUPS was 21%, and 12% of all consultations were for MUPS. A significant association was found between consultations for MUPS in mothers and children (adjusted OR 1.42, 95% CI 1.24, 1.63). No association was found between fathers and children, but the association was stronger when both parents consulted for MUPS (adjusted OR 1.52, 95% CI 1.19, 1.93). Significant dose-response relationships were found between numbers of consultations for MUPS and numbers of MUPS in mothers and children. These associations were clearest in maternal-child consultations for painful MUPS and MUPS in specific bodily systems including gastrointestinal, musculoskeletal and neurologic MUPS. Over a quarter (27%) of children who consulted for MUPS at baseline had persistent GP consultations for MUPS at one-year follow-up. Exposure to maternal consultations for MUPS was associated with persistent consultations for similar symptoms in children (adjusted RR 1.29, 95% CI 1.05, 1.58). Exposure to maternal consultations for painful, gastrointestinal, and neurologic MUPS was associated with persistence consultations for similar MUPS in the child. This thesis provides important information about the impact of parental health on child health and consulting behaviour. The implications for primary care and future research are highlighted.

Vascular access via the radial artery has recently been shown to reduce access site related vascular complications but is associated with a significant learning curve. Radial artery spasm, arterial puncture failure, vascular anomalies, failure to reach the ascending aorta and concern regarding higher radiation exposure with the transradial are some obstacles that impede widespread uptake of this technique. This study was performed to assess some of these learning curve issues and to explore the use of transradial access in high-risk patient subgroups. Six interlinked projects were setup for this study and a total of 3125 patients evaluated. Access site vascular complications remain unacceptably high in contemporary practice as discussed in Chapter 2. The transradial approach could minimise such complications. Radial artery anomalies are relatively common and are a common cause of transradial procedure failure as detailed in Chapter 3. Forearm arterial diameter variations and the effect of sublingual GTN were discussed in Chapter 4. The radial artery is bigger than the ulnar artery and GTN increases their diameters by an average of 15-22%. The issues with radiation exposure were studied as detailed in Chapter 5. With strict control of various variables and optimal radiation protection, we demonstrated that there is no difference in radiation exposure between transradial and transfemoral diagnostic angiography when performed by an experienced operator. The application of transradial technique in 2 high-risk patient subgroups was analysed as detailed in Chapter 6. Transradial rescue angioplasty for failed reperfusion and percutaneous right and left heart catheterisation via the arm approach without interruption to Warfarin therapy are found to be safe and effective. These findings have important clinical implication and may help shorten the learning curve and optimise procedure technique including high-risk patient subgroups, thereby help to further drive the adoption of transradial approach.

All the events of egg activation in mammalian eggs are triggered physiologically by transient increases in cytosolic free Ca²⁺ referred to as Ca²⁺ioscillations. These oscillations are initiated by the sperm derived PLC isoform, PLCζ. PLCζ releases Ca²⁺ by hydrolysing its substrate PI(4,5)P₂ to produce IP₃, however, many of the mechanisms by which PLCζ elicits Ca²⁺release in eggs are poorly understood. The results of this thesis confirm that whilstPLCζ cRNA and recombinant protein is able to cause Ca²⁺ioscillations in mouse eggs the sperm derived protein PAWP does not cause any Ca²⁺ release in any circumstances. It is shown that EF hand domain and XY linker of PLCζ are important in determining its Ca²⁺i releasing ability by enabling PLCζ binding to its substrate PI(4,5)P₂ through electrostatic interactions. The C2 domain of PLCζ was also found to play a crucial role in the Ca²⁺ releasing ability of PLCζ, possibly by binding to lipids or proteins in the target membrane. The Ca²⁺releasing ability of eggs is acquired during oocyte maturation and a dramatic increase in PLCζ sensitivity of oocytes occurs after germinal vesicle breakdown. A variety of markers for PLCζ’s substrate PI(4,5)P₂ including fluorescent PI(4,5)P₂ and gelsolin based fluorescent probes suggests that this PI(4,5)P₂ is localised to intracellular vesicles that could derive from Golgi apparatus. Attempts are made to measure PI turnover in these intracellular compartments of eggs during PLCζ induced Ca²⁺i oscillations using several probes. The results of this thesis suggest that PLCζ releases Ca²⁺ by a novel IP₃ based signalling pathway that involves an intracellular source of PI(4,5)P₂.

Chronic neuropathic pain (CNP) occurs as a consequence of injury to the nervous system. Despite recent advances, CNP lacks objective diagnostic criteria, is often unrelenting and refractory to treatment. The primary aims of this thesis are twofold; the identification of CNP biomarkers using both human cohorts and an animal model (spinal nerve ligation; SNL) of neuropathic pain, and to provide clarity on the role of GTP cylcohydrolase I (GCH1) in CNP. Analysis of GCH1 and related genes and metabolites was conducted. As biomarkers, nitrite/nitrate and neopterin did not differentiate controls from CNP patients. However, significant differences were observed with biopterins, whilst correlations were observed between GCH1, nitrite/nitrate and neopterin, which were notably stronger in patients than controls. Analysis in human cohorts and in the SNL model also inferred that downregulation of GCHFR may contribute to BH4 synthesis. In order to provide clarity on the role of the GCH1 pain protective haplotype, reporter gene assays were used. This demonstrated a potential regulatory role for the GCH1 5’ SNP (rs8007267). In silico prediction of transcription factor binding sites suggested that this may be mediated by the aryl hydrocarbon nuclear translocator. The use of electrophoretic mobility shift assays showed strong specific binding with probe pertaining to the major allele. Further analysis is required to elucidate transcription factor binding, potentially facilitated by 2D-PAGE and mass spectrometry. In order to further elucidate potential CNP biomarkers, microarray analysis and qRT-PCR were performed using blood obtained from CNP patients. Data refinement led to the isolation of 27 potential CNP biomarkers, of which several cross-validated between cohorts. Microarray data, literature evidence, and correlations with previous microarrays provided evidence suggestive of a role for TIMP1. Multiple other genes, including CASP5, TLR4, TLR5, MC1R and CX3CR1, were differentially regulated in CNP. Genes surviving microarray data refinement were subsequently analysed in the dorsal horn of Sprague Dawley and Wistar Kyoto rats after SNL. Several genes, including Dpp3, Mc1r and Timp1, were similarly differentially expressed in the rodent SNL model, which suggests that these genes may be involved in the pathophysiological mechanisms of CNP, and may also function as potential translational biomarkers of CNP. This work provides multiple avenues for expansion and further investigation. Clearly, the challenges associated with biomarker discovery in CNP states are considerable, though it is hoped that this thesis provides valuable insight and the necessary foundation for future work.

Depression is a common condition, affecting around one in 20 people worldwide. It is challenging conceptually and clinically, with treatment being ineffective for many, and significant consequences for individuals and societies alike. Depression is particularly problematic during pregnancy, where it is no less common, but poses additional difficulties. Both depression and its pharmacological treatments are associated with a range of short- and longer- term sequelae for offspring, and current data is insufficient to allow fully informed decisions to be made by mothers, midwives, or doctors. Research is affected by practical, ethical, and methodological issues, and a myriad of confounding factors, which combine to increase uncertainties over the risks and benefits of prescribing (or not). Retrospective and prospective observational studies accompany epidemiological data linkage and meta- analyses involving millions of subjects, in contributing to both current knowledge and testable hypotheses to inform future directions for research, while clinical and preclinical studies with smaller sample sizes provide invaluable and complementary details. However, significant gaps remain, not least in delivering optimal care to each individual mother and baby. While the overall emerging picture appears reassuring to some, others acknowledge that we do not even possess all the pieces of the puzzle yet. There remains an urgent need for more comprehensive and relevant data. This thesis presents the findings from a series of pilot studies on evaluating the characteristics and consequences of antenatal exposure to selective serotonin reuptake inhibitors. Up to one in 10 women in the general Scottish population may be exposed to an antidepressant at some point during pregnancy, but adverse outcomes may be related more to underlying maternal depression, rather than its pharmacological treatment. We highlight areas of both intelligence and ignorance, and make proposals for future research.

Background: Musculoskeletal pain is a major burden on society. Research in adults has identified risk factors associated with musculoskeletal pain onset, however at present evidence for risk factors in children and adolescents is limited. Aims: Identify potential risk factors for musculoskeletal pain onset in children and adolescents from current literature, and generate specific hypotheses to be tested using existing cohort data. Methods: A systematic review was conducted to summarize existing evidence of risk factors for musculoskeletal pain onset in children and adolescents. Two child and adolescent prospective cohort datasets and a local primary care consultation database were used to test hypotheses using logistic and survival regression analysis. Results: The systematic review found evidence that sleep problems and psychological symptoms (internalizing and externalizing) were associated with musculoskeletal pain onset with added evidence of potential effect modifiers. For sleep problems, analysis within a prospective cohort showed higher odds (OR 1.35, 95%CI 0.84, 2.16) for musculoskeletal pain onset, but this association was significant only for chronic pain onset (OR 2.22, 95%CI 1.43, 3.44), with evidence of effect modification by gender (association was stronger in boys). Testing within a primary care cohort showed a 49% increased hazard of sleep consultations with musculoskeletal consultations. In a cohort of adolescents musculoskeletal pain was not significantly associated with internalizing symptoms (OR 1.43, 95%CI 0.96, 2.12), however a significant association was found for externalizing symptoms (adjusted OR 1.99, 95% CI 1.28, 3.10), with evidence of effect modification by pubertal status and screen time use. Testing in a primary care cohort revealed a 39% increased hazard for musculoskeletal consultations. Conclusions: Potential risk factors (sleep and psychological symptoms) and effect modifiers were identified for (chronic) musculoskeletal pain onset within child and adolescent population and primary care samples. Future work is required to explore mechanisms explaining these associations, and develop appropriate interventions.

Introduction: Patients surviving a severe stroke are at risk of developing contractures. Evidence suggests that spasticity may be a cause of contractures, particularly in patients who have not recovered functional movement. The relationship and the time course of spasticity and contractures remain unclear. This thesis aims to identify when spasticity can be identified and investigate whether treating spasticity at onset using botulinum-toxin, might slow contracture development. Methods: A double blind randomised placebo-controlled trial with an initial six-week screening phase was conducted in an acute NHS hospital. Patients with no arm function (Action Research Arm Test grasp-score < 2) within six-weeks of stroke were eligible for screening. Screening for spasticity was carried out using a neurophysiological method. Patients who developed spasticity were randomly assigned to receive intra-muscular injections of 0.9%sodium chloride solution or onabotulinumtoxinA. Measures of spasticity and contracture development (reduced passive range of motion (PROM) and increased stiffness) were taken at the wrist and elbow at baseline, weeks-two, four, six and twelve post injection and six-months post stroke. Results: Over a 23-month period, 1143 patients were admitted with stroke and 120 consented to study participation. Of these, 100 developed spasticity without functional recovery 84%(95% confidence interval(95%CI):76%-89%). Mean time of spasticity onset was 13.5-days(SD:8.5). Of the 100 eligible for randomisation 93 were included in intention to treat analysis. At six-weeks, treatment results in a reduction in wrist spasticity (mean difference(MD):4.8μV;95%CI:1.2to8.4;p=0.009), stiffness (MD=4.2mN/deg;95%CI:0.7to7.7;p=0.02) and PROM (MD=13.8o;95%CI:6.1to21.6;p=0.01). At the elbow; four-weeks spasticity (MD=9.8μV;95%CI:4.3to15.4;p=0.001), four-week stiffness (MD=4.8mN/deg;95%CI:-0.1to9.6;p=0.056) and twelve-weeks PROM–(MD=6.5o;95%CI:0.6to12.3;p=0.03). These changes were not maintained at the six-month follow-up assessment. Conclusion: Spasticity occurs earlier and is more common than previously reported. Treating spasticity early with onabotulinumtoxinA can reduce the rate of contracture formation. Further work is required to elucidate who is at greatest risk of contractures and to explore if these treatment effects can be sustained with adjunct therapies.

This thesis presents a critical edition of John Hall’s casebook (composed around 1634-1635) and commentaries on aspects of it. My research involved close reading of Hall’s Latin, and its translation into English. In the process it became apparent that Hall had made considerable use of unattributed borrowings from Latin medical books, making up between thirty and forty per cent of his text. These were mostly identified by detailed word searches of on-line databases. This is a use of medical texts not previously noted, and makes a clear connexion between Latin medical texts and routine medical practice. The thesis is presented in four sections, plus introduction and conclusion. The first section, the Background, gives the history of Hall’s manuscript from its composition in 1634-35 to its acquisition by the British Library. It sets out the reasons for producing a new translation, the editorial principles and practice followed, and some medical themes running through Hall’s case reports. Section two contains the critical edition itself, with parallel Latin and English texts. Footnotes to the Latin text give the sources of all of Hall’s borrowings from and references to medical and other texts. The third section (Chapter 1) analyses the process and results of identifying Hall’s working library, of forty-three authors and sixty titles, from his borrowings. It puts his library in the early modern medical context in terms of its contents and categories of composition. I show that there were changes in the books Hall acquired over time, from those suitable for a student through to his later interests in chymical practice and the diagnosis of scurvy. Despite these changes, he continued to rely on old familiar texts for most of his remedies throughout his life. The fourth section (Chapter 2) examines Hall’s manuscript in the context of casebooks generally. It differs from the majority of casebooks, the differences being explained by its composition as the draft for a book to be published. It shows that a casebook can have an internal structure related to the chronologies of its composition and the cases it draws on. This thesis demonstrates the importance of Latin sources in at least one medical casebook of the early seventeenth century. I show that borrowings such as Hall’s were not unique even if rarer in other texts. The possibility of a Latin textual source should be considered for any Latin text in a casebook of that period.