Aberrant neurotransmitter function promotes cognitive deficits in schizophrenia. These abnormalities in functioning are seen as disruptions in attentional and information processing, as well as disruptions in the consolidation and retrieval of information. Tasks of attentional salience and memory that are used to model these disruptions include the latent inhibition (LI) task of attentional salience, prepulse inhibition (PPI) task of sensorimotor gating and an Episodic memory (EM) task, which is an index of memory for episodes at a particular point in time. Aberrant functioning of candidate genes that are associated with risk for schizophrenia may be seen as behavioural alterations in these tasks of schizophrenia relevant phenotypes. dopaminergic hyperactivity and hypofunction have been implicated in mediating disruptions on these cognitive tasks. Increased transmission in the dopamine system in the striatal region promotes schizophrenia symptoms, and indirect dopamine (DA) agonist Amphetamine worsens these symptoms in patients, and disrupts schizophrenia relevant behaviours on these cognitive tasks. We investigated the effects of deletion of two genes relevant to schizophrenia on cognitive tasks known to be disrupted in the disorder. The effect of deletion of the dopamine D2 receptor (D2R) and trans membrane (TM) domain Neuregulin-1 (Nrg-1) receptor were investigated in mediating disruptions in cognitive processes in an animal model of schizophrenia. The role of the D2R in an attentional model of sensorimotor gating was assessed. PPI was attenuated in D2R knock out (KO), in a one day sensorimotor gating task. In a one day PPI test protocol, amphetamine disruptions on PPI were spared in D2R WT and KO mice. Following on from previous reports of disrupted LI by a single low dose amphetamine injection, separated by 24h interval, we established a single vs. two low dose PPI protocol in order to facilitate a direct comparison of amphetamine induced disruption in LI with PPI. A one injection (prior to test only) vs. two injection (prior to habituation and prior to test) task was established. In the two day protocol, a single low dose of amphetamine disrupted PPI in D2R KO mice and reduced startle reactivity to the 120 dB pulse alone trials. Two low dose injections of amphetamine however, do not disrupt PPI in D2R KO or their WT littermates, and do not mimic low dose amphetamine disruptions in the LI task. These findings demonstrate that prior conclusions about the requirement of the D2R for amphetamine effects in PPI does not generalise to all dose regimens. Episodic memory was also investigated as a measure of cognitive impairment in schizophrenia. D2R KO mice show sex specific dissociations on an EM task. Male D2R WT and KO animals show equal exploration of old vs. recent objects on the what-when component of the EM task, and female KO animals show enhanced memory for old vs. recent objects. Both D2R WT and KO mice show intact memory for displaced objects. These deficits were also investigated in the TM domain Nrg-1 model. Nrg-1 has been implicated as a candidate gene for schizophrenia, and behavioural phenotypes assessing its role in cognitive impairment in schizophrenia were established. Intact LI is seen in both Nrg-1 WT and Het animals. Nrg-1 TM domain Het mutants also show deficits on the schizophrenia relevant PPI task. Nrg-1 Het mutants show attenuated % PPI compared to their WT littermates, which reflects interrupted sensorimotor gating in schizophrenia. Lastly, we found some evidence that reduced function of TM-domain of the Nrg-1 gene disrupted episodic-like memory (what- where-when recognition) in males and improved it in females.

Background: Epilepsy is a serious chronic neurological disorder that has a higher incidence in children and young people (CYP) than in adults. Epilepsy negatively impacts physical and psychosocial quality of life of CYP. Good outcomes of epilepsy are associated with optimal choice of drug treatment and adequate adherence to the prescribed medicines. Research on the patterns of medication use and adherence to prescribed medicines in CYP remains limited. The long-term clinical outcomes and costs of treating epilepsy have not been extensively studied in CYP in the UK. Aim of the study: This thesis aimed to investigate the pattern of antiepileptic drug (AED) prescribing and the dynamic of medication adherence in CYP with epilepsy. The long-term clinical outcomes and direct costs of treating epilepsy in CYP were estimated at population level. Methods: This study is an observational cohort study of CYP, age 0-17 years, identified from The Health Improvement Network (THIN) primary care database from the UK between January 1988 and December 2004. Four different analyses were carried out on this cohort. First, a cross-sectional design repeated annually was employed to estimate the incidence and prevalence of epilepsy and the pattern of AED prescribing in this population. Secondly, the long-term adherence to prescribed AEDs was calculated using the medication possession ratio (MPR) method. Applying panel data analysis and the Generalised Estimating Equation (GEE) multivariate regression, factors that may have been associated with adherence to the prescribed AEDs were examined. Thirdly, seizure outcomes in terms of seizure frequency and remission of seizures and potential associated factors were assessed using the method of multiple failure survival analysis. Finally, the direct costs of treating epilepsy in CYP in primary care were estimated and stratified by the number of years after the first recording of epilepsy in THIN data. Results: Of total 528,760 CYP born on or after 1st January 1988 and registered in general practices contributed to THIN until 31st December 2004, 2020 CYP were identified who had a diagnosis of epilepsy, from under 1 up to 16.3 years of age (mean=5.6; SD=4.1). The annual incidence of epilepsy in CYP stratified by calendar years ranged from 44.4 (95% CI=31.9-61.8) to 61.2 (95% CI=50.6 -74.1) per 100,000 person-years. Incidence of epilepsy was significantly higher in children with greater socioeconomic deprivation than those with lower deprivation. Around 60% of CYP with epilepsy were prescribed monotherapy each year. Old AEDs such as carbamazepine and sodium valproate were the most frequently prescribed drugs and often prescribed as monotherapy to control epilepsy throughout 1990-2003. Prescribing of lamotrigine, a new AED, increased from 0.07 per person-years in 1992 to 2 per person-years in 2003. The calculated annual adherence to AEDs showed that around 50% of CYP adhered to at least 80% of the prescribed medications each year. Demographic characteristics of CYP were of little significance to affect adherence levels. The incidence of seizures was 0.73 (95% CI=0.71-0.75) per person-years. Incidence of seizures was higher in younger children up to 2 years and decreased with increasing age. A proportion of 94% (95% CI=93%, 96%) of CYP achieved 1 year remission of seizures, 80% (95% CI= 78%, 83%) achieved 2 years and 47% (95% CI=43%, 50%) achieved 5 years remission of seizures. The mean total direct cost associated with treating epilepsy in CYP, according to information in the general practice records that also indicated specialist and hospital care, was estimated at £ 1,153 (SD=1,808) per child in the first year following epilepsy diagnosis and at £459 (SD=1,633) per child for subsequent years. The costs of hospital care and AEDs represented the highest contribution to the total direct costs of epilepsy. The annual direct cost was significantly higher in younger children up to 2 years old. No significant difference in the annual costs was observed between CYP who adhered to at least 80% of medications and those who adhered to less than 80%. Conclusions: The incidence of epilepsy was highest in young children and CYP of higher socioeconomic deprivation. Old AEDs were most often prescribed as first-line drugs and as monotherapy to control epilepsy. Of newer AEDs, there was an increasing trend of prescribing lamotrigine and topiramate as add-on therapy. Long-term adherence to prescribed AEDs was suboptimal in one-half of CYP and positively associated with higher seizure frequency. Inpatient hospital care and drugs were the major contributors to the direct costs of treating epilepsy in CYP. Non-adherence to prescribed medicines was associated with higher hospital care costs but not with total direct costs as the medicines themselves made large contribution to the direct costs

This study presents an evaluation of the FRIENDS for Life program (Barrett, 2010) with an autism spectrum (AS) population. FRIENDS for Life is an intervention program underpinned by the principles of cognitive behavioural therapy (CBT) with a primary aim of reducing participant anxiety levels (Barrett, 2010). Existing research suggests it is an effective intervention in reducing participant anxiety levels (Briesch, Hagermoser Sanetti and Briesch, 2010) and it has been recognised by the World Health Organisation (2004) as the only evidence based program effective in reducing anxiety as a universal and targeted intervention. In recent years an evidence base for the application of CBT with children with AS has emerged, though primarily this research has been conducted in a clinical setting. Therefore this study aims to contribute to both evidence bases through implementing the FRIENDS for Life program within a new population as well as contributing to the broader evidence base evaluating the effectiveness of CBT with children with AS. The study adopted a post positivist epistemology and used a single case experimental design (SCED) to evaluate the effectiveness of the intervention in reducing the anxiety of four participants, aged nine to eleven, accessing special school provision. Anxiety was measured during a baseline, intervention and follow up phase using two weekly measures: the Paediatric Index of Emotional Distress (PI-ED;O'Connor et al, 2010); a short pupil questionnaire, and a weekly observation of participant behaviour. These measures were also triangulated with pre and post measures of anxiety using the Spence Child Anxiety Scale, child (Spence, 1997) and parent (Spence, 1999) version, and the School Anxiety Scale- Teacher Form (Lyneham, Street, Abbott and Rapee, 2008). Outcomes from the SCED showed that for all four pupils there was a significant decrease in anxiety from baseline to follow up on at least one weekly measure of anxiety, indicating a delayed effect on anxiety. The parent, child and teacher report triangulation measures suggested there was no significant change in anxiety post intervention. When considering outcomes, several key limitations to the study's design and implementation were taken into account including threats to construct validity and missing data in the intervention phase for two participants. The study concludes with support for the positive impact on participant anxiety as a result of the FRIENDS for Life intervention and recommendations are made for further investigation of the use of CBT interventions in schools with an AS population.

Nightmares, a common sleep disturbance which provoke fearful awakening, have been found to be a significant predictor of suicidal thoughts and behaviours. The research presented in this thesis aims to firstly examine if nightmares are predictive of self-harm regardless of suicidal intent or motivation, and secondly to explore the psychological mechanism linking the occurrence of nightmares to self-harm. Chapter 2, an online survey, revealed that nightmares were a significant predictor of self-harm regardless of suicidal intent or motivation and that this relationship remained when controlling for the effects of depressive symptoms. High levels of nightmares were also associated with elevated levels of negative affect and defeat. Chapter 3 prospectively examined the direction of the predictive relationship between nightmares and self-injurious thoughts and behaviours (SITBs) through a 5-day diary study of undergraduate students. Nightmares unidirectionally predicted SITBs when controlling for depressive symptoms and negative affect. Mediation analysis revealed negative affect to be a partial mediator between nightmares and post-sleep SITBs. Chapter 4 explored differences in the linguistic content of nightmares in individuals with and without a history of self-harm, using nightmare reports prospectively obtained from participants taking part in the diary study. Contrary to the literature, participants with a history of self-harm did not report more words pertaining to death. Exploratory analysis investigating self-harm recency indicated a higher frequency of perceptual words such as ‘feel’ and body words such as ‘arm’ in participant with current self-harm (< 1 month) compared to those with a history of self-harm (> 1 month) and those without. Chapter 5 modelled the psychological mechanism linking nightmares to increased self-harm risk via structural equation modeling from survey data. This model incorporated negative affect, hyperarousal and a latent variable ‘self-harm cues’ building on our previous findings and the literature. Our retained model indicated that a 1 standard deviation increase in nightmare score increased the probability of participants having recently (< 1 month) engaged in self-harm. Chapter 6 tested the predictions of the model computed in the previous chapter using behavioural and psycho-physiological methodology. Psycho-physiological measures when exposed to negatively valenced stimuli did not reveal any differences between high and low nightmare participants, nor were differences observed in self-harm cue sensitivity. However, a medium effect was observed indicating the high nightmare group to be more sensitive to stressors. These findings are discussed in the context of the literature in Chapter 7. They provide novel insights into the relationship between nightmares and self-harm, and highlight the importance of negative affect and hyperarousal as reducing stress resilience in individuals at risk of deliberate self-injury.

Antisense oligonucleotide therapy is one of the most promising strategies for treatment of myotonic dystrophy type 1 (DM1), type 2 (DM2) and Duchene muscular dystrophy (DMD). These three diseases have nuclear retained mRNA, suitable for antisense therapy targeting. The delivery of oligonucleotides to their desired target has long been an obstacle in antisense therapy with a large number of delivery reagents or methods having adverse side effects. Promising work published detailing the successful delivery of various chemically modified oligonucleotides (CMOs) naked, via gymnosis, led to us investigating a number of these CMOs: deoxyribonucleic acids (DNA), Peptide nucleic acids (PNAs), 2’OMethyl (2’OMe), and Phosphorodiamidate morpholino (PMO) oligonucleotides. In DM1 expanded CUG repeat (CUGexp) mRNAs aggregate in the nucleus forming “foci”. Testing the CMOs effectiveness at disrupting nuclear foci in a cell based assay, using high content imaging to visualise the number, size and intensity of foci we initially discovered that PNA and 2’OMe, were seemingly taken up via gymnosis by DM1 cells, and removed nuclear foci at nanomolar concentrations. However further experimentation using live cell imaging indicated that although all CMOs could enter the cell, in all disease models tested, via gymnosis, the CMOs could not penetrate the nuclear membrane. In depth analysis led us to identify an artefact of the in-situ process used to identify these foci, explaining earlier positive results. As the target mRNA is trapped within the nuclear compartment we investigated several transfection reagents for their ability to deliver 2’OMe oligonucleotides to the nucleus using live cell fluorescent imaging and a modified northern blot based method. It was established that polyethylenimine could successfully deliver 2’OMe oligonucleotides to the cell, with a high abundance of the oligonucleotide residing within the nuclear compartment. It was observed that PEI degrades the expanded nuclear retained repeat in the DMPK transcript of a DM1 patient cell line alone, without the addition of an antisense agent, in a concentration dependent manner.

Reports of posttraumatic growth can sometimes be illusory. Several researchers have argued that reports of posttraumatic growth may incorporate two separate phenomena, namely real posttraumatic growth and illusory posttraumatic growth. However, it is not often made explicit which kinds of illusions indicate illusory posttraumatic growth. An explicit conceptualisation of illusory posttraumatic growth is, however, necessary in order to investigate the research questions within this thesis, namely (a) whether reports of posttraumatic growth are correlated with illusions and (b) whether real posttraumatic growth and illusory posttraumatic growth differ in their correlation with well-being. Within the thesis, it was, therefore, primarily suggested that defensiveness could be responsible for illusory posttraumatic growth. The assumption was that high levels of maladaptive defensiveness may be used when distress cannot be endured. Internal and external experiences could then be pushed away. This case could potentially indicate illusory posttraumatic growth. In contrast, people who use low levels of maladaptive defences or high levels of mature defences may be able to endure the distress following a trauma or adversity. Internal and external experiences may be processed and accommodated. This case could potentially indicate real posttraumatic growth. The assumptions about illusions were tested within the thesis. It was investigated which kind of illusions could be adaptive psychological operations and which illusions could be maladaptive psychological operations. The results supported the assumptions within this thesis concerning adaptive versus maladaptive illusions. Within three studies, self-reported posttraumatic growth was significantly correlated with a neurotic defence style. It was concluded that this correlation could indicate that sometimes reports of posttraumatic growth are not real. However, other interpretations were also discussed. Four studies investigated whether real posttraumatic growth and illusory posttraumatic growth differ in their correlation with well-being. Within chapter 5, real posttraumatic growth, indicated by low levels of a neurotic or an immature defence style, was correlated positively with negative change following adversity (non significant). In contrast, illusory posttraumatic growth, indicated by high levels of a neurotic or an immature defence style, was not correlated with negative change following adversity. Although the difference between illusory posttraumatic growth and real posttraumatic growth concerning negative change following adversity had a meaningful effect size, it was not significant. Within chapter 7.3, real posttraumatic growth, indicated by extremely low levels of neurotic defensiveness, was correlated significantly with posttraumatic stress. In contrast, illusory posttraumatic growth, indicated by extremely high levels of neurotic defensiveness, was not correlated with posttraumatic stress. The difference between real posttraumatic growth and illusory posttraumatic growth was significant. Within chapter 8, real posttraumatic growth (high levels of mature defensiveness) was correlated with decreases in hedonic well-being measured by positive and negative affect. In contrast, illusory posttraumatic growth (low levels of mature defensiveness) was correlated with increased levels of hedonic well-being. The difference between real posttraumatic growth and illusory posttraumatic growth was significant. In total, reports of posttraumatic growth may, in fact, sometimes be correlated with defensiveness. However, real posttraumatic growth and illusory posttraumatic growth do only slightly differ concerning their correlation with well-being.

Background: In April 2009, the NHS Institute for Innovation and Improvement released four Better Care Better Value (BCBV) prescribing indicators. One indicator targeted renin–angiotensin–aldosterone system (RAAS) drugs, i.e., angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs), to improve their prescribing efficiency. The indicator affects the proportion of items written for ACEIs as a percentage of the total number of prescriptions for RAAS drugs. A proportion of 80% ACEIs has been proposed and considered as an achievable target based on the small proportion (2–10%) of patients who cannot tolerate the ACEIs-related side effects, mostly dry cough. However, neither the impact of the BCBV indicator on the utilisation of ACEIs/ARBs nor its cost-saving implication is yet known. The BCBV indicator involves switching established ARBs users to ACEIs, where appropriate. However, no previous studies have specifically investigated the clinical and economic impact of the switch of ARBs and hence it is unclear whether the switch from ARBs to ACEIs, promoted by the BCBV indicator, would be associated with any clinical or economic issues. Without any financial incentives or legislation enforcement, uptake of the BCBV policy is predicted to be low, especially as evidence suggests that single measures are ineffective in altering GPs’ prescribing behaviour. However, the uptake and implementation of the BCBV indicator has not yet been explored. Aims: This thesis aimed to evaluate the impact of the BCBV indicator for ACEIs/ARBs on the utilisation of ACEIs/ARBs in treating hypertension (HT) and any associated cost savings. The effects of switching patients from ARBs to ACEIs on clinical outcomes (adherence, blood pressure [BP] level and HT-related complications), healthcare resources and costs were also investigated. Prior to this, patients’ adherence and persistence patterns (discontinuation and switching), and potential associated factors were studied. Finally, the uptake and implementation of the BCBV indicator in real practice was explored. Methods:This research used an explanatory sequential mixed-method approach. Firstly, an interrupted time series analysis was used on data extracted from the Clinical Practice Research Datalink (CPRD), from April 2006 to March 2012, to examine the impact of the BCBV indicator on the repeated cross-sectional monthly ACEIs/ARBs prescriptions and costs in patients with primary hypertension, which was validated using national dispensing datasets. Secondly, a retrospective cohort study was conducted using prescription records extracted from CPRD for patients with primary hypertension who were prescribed antihypertensive drugs from April 2006 to March 2013. Patients’ adherence and persistence patterns (including switching) to antihypertensive drug classes were quantified and their association with patients’ characteristics tested, using Generalised Linear Models and survival analysis. Thirdly, a retrospective cohort study was conducted on patients who switched their index drug class from ARBs to ACEIs to evaluate the clinical outcomes and resources used associated with ARBs switching, using data from CPRD in linkage with Hospital Episode Statistics, and applying multilevel regression for data analysis. Finally, semi-structured interviews were conducted with 16 general practitioners, to explore their views about ACEIs/ARBs prescribing and the BCBV policy to understand the reasons underpinning the effectiveness/ineffectiveness of the BCBV indicator. The interviews were recorded, transcribed verbatim and analysed using a thematic approach. Results: Impact of the BCBV indicator on ACEIs/ARBs utilisation and cost saving. The ACEIs prescription proportion declined from 71.2% in April 2006 to 70.7% in March 2012. The policy initially resulted in a sudden reduction of 0.3% (95%CI: –0.44, –0.16) in the level of ACEIs prescription proportion; however, it was thereafter associated with a sustained, significant increase of 0.013% (95%CI: 0.0007, 0.02). The failure to achieve the 80% target by the end of the study period resulted in missing a potential cost saving of 23.9% of the total ACEIs/ARBs cost. In June 2014, a potential cost saving of ~£1 million was predicted, had the 80% target been achieved; this substantial saving was several years after the availability of several low-cost generic ARBs. Adherence and persistence to antihypertensive drug classes: Among the 176,835 patients with primary hypertension included in this analysis, 38.4% and 20% were non-adherent (proportion of days covered [PDC]<80%) to their antihypertensive drug class and therapy, respectively. The discontinuation rate of antihypertensive drug class (56.1%) and therapy (35.5%) was relatively high by the end of six-year follow-up period. Using ARBs as the index drug class, prevalent hypertensive patients, prevalent antihypertensive drug users and increasing age were associated with higher PDC to antihypertensive drug class and therapy but a lower risk of drug discontinuation; whereas a higher deprivation level, comorbidity score and switching of antihypertensive drug class were associated with lower PDC to antihypertensive drug therapy but a higher risk of discontinuation. Overall, 26% of the included patients switched their antihypertensive drug class. Using ARBs, higher deprivation, higher comorbidity and prevalent antihypertensive drug users were protective factors against switching; whereas female gender, prevalent hypertensive patients, higher BP level and increasing age were associated with a higher risk of switching. Clinical and economic implications of switching patients from ARBs to ACEIs: Of the 46,193 patients who switched their antihypertensive drug therapy, only 470 patients switched from ARBs to ACEIs – they were included in this study. Based on whether the patient combined ACEIs with other antihypertensive drug classes in the post-switching period, 369 patients were classified into ACEIs-combined group and 101 patients into ACEIs-monotherapy group. Compared with the pre-switching period, the proportion of non-adherent patients (PDC<80%) in the post-switching period increased significantly (17% vs. 27%); however, this significant increase was observed only in the ACEIs-combined group (17.3% vs. 29%). The switching of ARBs was associated with a significant reduction in both systolic and diastolic BP (144.2 vs. 141.9 mmHg, p<0.001) and (84.6 vs. 82.6 mmHg, p<0.001), respectively; however, this effect was found only in the ACEIs-combined group. The incidence of all HT-related complications was comparable in both pre- and post-switching periods, except for the incidence of MI, which reduced significantly only in the ACEIs-combined group (OR=0.1, 95%CI: 0.04, 0.6). Importantly, the switching of ARBs to ACEIs was associated with a significant reduction in the total medical costs (mean cost difference: –£329.2, 95%CI: –534.6, –205.7). Uptake and implementation of the BCBV indicator: Potential barriers suggested for the poor uptake of the BCBV policy, included a lack of GP awareness of the policy, GPs’ negative attitudes toward the policy and a lack of financial incentives. Among other reasons, the current high proportion of ARBs users appeared to be related to an over-switching of ACEIs to ARBs in anticipation of an ACEIs-related dry cough. Suggested measures for increasing the policy uptake included effective dissemination strategies, linking the policy to financial incentives and providing guidance on performing the switching. Furthermore, strategies to address the over-switching of ACEIs to ARBs were considered essential for the BCBV policy to achieve its 80% ACEIs target. Conclusions:The BCBV indicator was ineffective and failed to achieve the 80% ACEIs target. The association of achieving the 80% target with a remarkable cost-saving opportunity and the lack of any negative clinical consequences following the switching of ARBs to ACEIs indicate the ongoing necessity to reconsider and reinforce this policy through multiple initiatives to increase its future effectiveness, building on the study’s identified barriers and suggested solutions. This thesis represents a case study of a failed and an ineffective prescribing policy, attributed primarily to inappropriate policy implementation. It provides key lessons for policy makers and healthcare authorities on the importance of effective implementation strategies as an integral component to any successful policy.

Young people with Autism Spectrum (AS) are a population at risk of experiencing Mental Health Problems (MHPs); with anxiety being the most commonly reported internalising MHP for this population (Skokauskas & Gallagher, 2012). Schools have been identified as well placed to support pupils experiencing MHPs (DoH & DfE, 2017). The primary aim of the present research was to evaluate the Homunculi Approach (Greig & MacKay, 2013b) as an intervention for secondary aged pupils with AS experiencing anxiety. The secondary aim of the research was to explore factors affecting implementation of the Homunculi Approach. An embedded mixed methods design was utilised where the primary research aim (Phase 1) was supplemented by the secondary research aim (Phase 2). Phase 1 consisted of an A-B single case experimental design (SCED) with four participants across four different secondary schools. Anxiety was measured through: repeated behavioural observations and a weekly anxiety questionnaire (PI-ED); pre and post intervention measures, completed by the young people, parents and teachers (SCAS, SAS-TF and SDQ); and measures which occur organically within the Homunculi Approach intervention. Phase 2 of the research explored the factors which affected implementation of the intervention in schools by interviewing the school staff who delivered the Homunculi Approach. The data was collected using an activity theory framework. The findings from the research indicated the intervention was effective in reducing anxiety for two participants, and ineffective in reducing anxiety for the other two participants. Several implementation factors were identified which may have impacted on the effectiveness of the Homunculi Approach when delivered by members of school staff in secondary schools. The findings are discussed in relation to the literature. Limitations of the research, such as the small sample size and possible impact of external factors, are acknowledged. Possible implications and ideas for future research are presented.

Background Many randomised controlled trials (RCTs) are available to support using different pharmacotherapy agents in the management of various neuropathic pain conditions. However, choosing these pharmacotherapy agents for neuropathic pain is challenging, due to the limited evidence-based knowledge to support the use of different pharmacotherapy agents in different neuropathic pain conditions. Aims The aim of this PhD is to evaluate the efficacy and safety of oral and topical pharmacotherapies for managing neuropathic pain by deriving placebo and active comparative efficacy and safety evidence from RCTs. Methods This research used three approaches to summarise and synthesise evidence from randomised controlled studies including: a systematic review of placebo and active control RCTs to summarise and criticise the current evidence in neuropathic pain; a meta-analysis and sequential analysis of eligible studies to provide a more precise estimate of the overall treatment effects; and a network meta-analysis to estimate the relative effectiveness of the most commonly used interventions in neuropathic pain. Results Systematic review Two hundred placebo and active-controlled trials met the inclusion criteria. A wide range of different treatments were studied in these trials, including anticonvulsants, antidepressants, opioids and topical capsaicin and lidocaine. Most of the included studies were parallel placebo-controlled trials and commonly lasted for 3 to 12 weeks. In addition, the vast majority of the included RCTs were conducted in participants with painful diabetic neuropathy and post-herpetic neuralgia, while only a few trials were conducted in participants with central neuropathic pain conditions. Pairwise meta-analysis Sixty seven trials were eligible for the pairwise meta analysis of efficacy outcomes. Of the anticonvulsants group pregabalin and gabapentin compared with placebo demonstrated efficacy for 50% and 30% pain reduction and global improvement in patients with neuropathic pain. The efficacy of anticonvulsants varied in different types of neuropathic pain. Gabapentin when compared against a placebo was better than a placebo in PHN and PDN, while pregabalin was significantly effective in patients with post-herpetic neuralgia (PHN) and painful diabetic neuropathy (PDN) but not in patients with HIV associated neuropathic pain. Others anticonvulsant agents, such as lamotrigine, valproic acid, topiramate, levetiracetam and oxcarbazepine, were tested in a small number of trials. These did not provide useful benefits compared with a placebo for a 50% and 30% pain reduction. Of the antidepressant group, duloxetine when compared to a placebo demonstrated efficacy for 50% and 30% pain reduction in diabetic neuropathic pain. A few active comparison trials failed to demonstrate superior efficacy of one drug over another for a 50% and 30% reduction in neuropathic pain. Trial sequential analysis To examine the reliability and conclusiveness of the available evidence, trialsequential analysis has been applied in this study. The results show convincing evidence of the efficacy of some interventions (e.g. pregabalin, gabapentin and duloxetine) to reduce pain by 50% in some neuropathic pain conditions (e.g. diabetic neuropathic pain and post-herpetic neuralgia). The continuation of RCTs of pregabalin and duloxetine in diabetic neuropathy and gabapentin in post-herpetic neuralgia is not necessary as there appears to be sufficient evidence of the efficacy of these treatments in the management diabetic neuropathic pain and post herpetic neuralgia. Further RCTs of duloxetine, pregabalin and gabapentin are however required for central neuropathic pain. In contrast, the analysis failed to provide evidence that opioids and high concentration capsaicin demonstrate a 50% pain reduction. Network meta-analysis Twenty-eight trials were eligible for the network meta-analysis. The results incorporating both direct-comparison and indirect-comparison evidence suggested that there is no superiority of duloxetine over amitriptyline, pregabalin and gabapentin in achieving at least a 30% and 50% pain reduction with a treatment duration of 7 to 12 weeks in patients with neuropathic pain conditions, such as diabetic neuropathic pain, postherpetic neuralgia and spinal cord injury. Conclusions In summary, this research has found that some good quality trials provide good evidence regarding the efficacy of duloxetine, pregabalin and gabapentin in a minority of patients with neuropathic pain. Until advancements in developing mechanism-based approaches and improved clinical trial design become available, the routine use of these medications is unlikely to be changed. This may support the hypothesis that traditional RCTs might not be a suitable method of choice to address provisional health questions in routine clinical practice.

Magnetic Resonance Imaging allows electromagnetic properties of the brain to be measured in vivo, providing new markers of structure at a microscopic level. Evaluation of the local complex signal evolution observed using a multi-echo gradient-echo (MEGE) sequence can allow the electromagnetic and relaxivity properties of individual tissue compartments to be accessed. The phase evolution carries valuable information about the different compartments, but is dominated by non-local, large-length-scale field variations which present the main challenge in processing complex MEGE data. In this work 2-compartment and 3-pool models are used to describe signal evolution from a mixed tissue and venous compartment and from white matter, respectively. A new method for removing the effects of non-local field variations without corrupting local non-mono-exponential phase evolution is proposed. The 2-compartment model of venous blood complex signal has been used to access the vascular dependence on orientation and oxygenation levels, and has been validated against existing methods on large distinguishable veins. A phantom consisting of capillary tubes filled with ferritin solution immersed in a water bath has been used to validate the 2-compartment model and the estimated frequency offsets between the ferritin and water compartments at multiple orientations to B⃗0 shown to agree with predictions of theory. White matter in the corpus callosum has been investigated using a saturation-recovery MEGE sequence with variable flip angles with the aim of revealing differences in the T1-relaxation properties of the myelin, the external and the intra-axonal water compartments. The results show that the relative saturation of the myelin water compartment decreases with increasing flip angle and is consistent with there being a short-T1 component associated with myelin water. However, the observed signal behaviour shows less contrast than expected based on the findings from the previous studies. This could be related to differences in exchange rates between compartments. Finally, diffusion-weighted, asymmetric spin-echo imaging was used to simultaneously investigate the diffusivity and electromagnetic properties of the external and intra-axonal compartments of white matter. This approach could provide additional information about white matter microstructure. Asymmetry of the spin echo sequence was achieved by delaying the acquisition by ∆t. The preliminary results show an increase in the scaled magnitude with echo delay at a constant b-value in some regions of the corpus callosum. The preliminary results are promising, but further investigation and method development are required. This thesis has investigated a number of novel methods of studying white matter structure and cerebral microvasculature.