Relative survival is an estimate of net-survival without the need for cause-of-death information. This is achieved by matching the study cohort to the general population by various covariates, including age, sex and year of hospitalisation, in order to obtain an expected mortality rate. In this thesis relative survival methodology will be applied in heart disease where the form of the excess hazard rate is known to be very different from cancer, where this methodology originates. The dataset presented is from the Leicester Royal Infirmary coronary care unit where all admissions to the unit were recorded between 1993 and 2006, which includes all patients in Leicestershire. Only patients who present with an ST-elevated acute myocardial infarction will be studied. Relative survival is a new methodology in heart disease and this thesis will describe some of the problems that are encountered including the increased prevalence of the disease in the population and the very high early excess mortality rate that is not present in most cancers. Also investigated are period analysis models, which are also new to heart disease and allow the estimation of up-to-date information. An analysis of admission blood glucose levels and diabetic status is performed to examine the potential impact on patient prognosis in the short and long term, which involves the use of relative survival. A new methodology is developed in this thesis for relative survival that fits spline based flexible parametric models on the log cumulative excess hazard scale. This methodology holds many advantages over current relative survival techniques due to the use of non-split- time data. This thesis demonstrates these advantages. This thesis details how current relative survival methods have been extended to heart disease. A new model is developed, which is suitable in heart disease and cancer that fits flexible parametric spline based models.

5-aminosalycilic acid (5-ASA) therapy is effective for maintaining remission in patients with ulcerative colitis (UC) and may also reduce colorectal cancer risk. However, medication non-adherence is a known barrier to the effectiveness of prescribed regimes and there is a lack of evidence about methods of improving adherence to 5-ASA treatment. This research programme addressed the hypothesis that adherence can be improved by a multi-faceted intervention tailored to individual patient needs. A qualitative study identified that important determinants of adherence to 5-ASA medication are: information provided, patient beliefs and the patient-clinician relationship. Adherence can change over time; the study highlighted the need for reinforcement and the fact that health care professionals have a crucial role to play in this dynamic. A cross-sectional study confirmed the difficulty of accurately assessing medication adherence. The two measures used (self-report and urine analysis) were not correlated, phi correlation 0.029 (p = 0.725). Logistic regression identified a significant association between self-reported non-adherence and: younger age [OR for increased age 0.954, 95% CI 0.932–0.976] and also doubts about personal need for medication (OR for BMQ – Specific Necessity scores 0.578, 95%CI 0.366–0.913). For non-adherence based on urine analysis, only South Asian ethnicity was independently associated with non-adherence (OR 2.940, 95%CI 1.303–6.638). A randomised controlled trial showed that a multi-faceted, tailored intervention (including an opportunity for patients to select reminder devices from a range offered) had a significant positive impact on maintaining adherence levels in the intervention group (p=0.001), with a 44% difference between adherence levels in the two groups at follow-up. Changes in questionnaire scores suggested a positive effect of the intervention on satisfaction with information (p<0.001). The intervention was feasible, and was acceptable to patients. The multi-faceted approach studied has potential for implementation in routine care for enhancing persistence with 5-ASA and thus improving patient outcomes.

Adverse events following the use of medical interventions are a major source of concern for patients, healthcare professionals and pharmaceutical companies. Therefore, evidence synthesis of potential adverse events are very important in determining whether an association exists, and the strength of such an association. It is also desirable to be able to quantitatively balance potential harms against the benefits of the intervention. However, standard statistical techniques for meta-analysis are often unsuitable when applied to datasets where the primary intervention is an adverse event. A review of standard meta-analysis methods, including Bayesian methods, is conducted. The specific challenges of meta-analysis in relation to adverse events datasets are described, with some of the main areas of contention being sparsity of events, subgroup analysis, class effects with regard to drug interventions, and issues related to time factors within the individual studies. Methods used in existing meta-analyses where the primary outcome is an adverse event have also been reviewed; this demonstrates the methods already used in this field and highlights some of their limitations, and where the methods could be extended. In the light of the reviews of methods and previous meta-analyses, four case-studies are performed. The first uses data from GlaxoSmithKline to investigate a potential relationship between paroxetine and suicidality, using many of the standard methods for comparison purposes. The second uses individual patient data for a time-to-event analysis of anti-TNF drugs used for rheumatoid arthritis. This clinical example is extended by the use of Mixed Treatment Comparisons for within-class comparisons, and to assess the effect of dose. Finally, a harm-benefits model is used to assess the interplay of risk of endometrial cancer against breast cancer recurrence for tamoxifen users. These models present novel ways of analysing adverse events data and demonstrate some of the difficulties in their use.

No description available.

No description available.

No description available.

One of the difficulties in teaching clinical medicine is the lack of opportunity a student has to acquire techniques for solving clinical problems. By using a computer to simulate a General Practice environment where patients with sets of symptoms are presented, a student can gain experience of diagnostic techniques and treatment management for any medical condition. Such an approach should enhance a student's development of properly structured clinical algorithms for interrogating a patient and arriving at an appropriate management plan. The intelligent tutoring system developed at the Department of Computer Science with the collaboration of the Department of General Practice aims not only to simulate this environment but also to formulate the basis for a general interactive learning environment for all subject domains with similar problem-solving model. In this system, a student may question, examine and provide treatment plans for a patient whilst constantly being monitored by the system:. Using Artificial Intelligence techniques, the tutor is able to assess the progress of a student throughout the tutorial session and produce tutoring interventions at appropriate stages, according to the student's ability. The system's knowledge base consists of disease profiles and population parameters which are created and updated by a separate system - the Medical Editor. The manipulation of this database allows tailoring of the system to simulate any clinical situation in Primary Care. This research considers in detail the current teaching/tutoring strategies adopted by all medical computer-assisted learning systems. It identifies the main areas of difficulty for using such systems in the Primary Care undergraduate course and discusses the consultation model used in this system with full comparison of the models used in Secondary Care. The research also discusses the main design issues which forms the framework for building learning environments based on intelligent tutoring systems.

<I>Hypothesis</I>: The work examined the effect of rate and extent of WL on body composition, physiological function (e.g. respiratory function), psychological function (e.g. mood), appetite and behaviour. <I>Methods</I>: The PhD was designed as a series of structured, longitudinal studies examining specific physiological and psychological parameters, in the three groups of six healthy, but obese (BMI 30-35) men. Subjects were assigned to either a WL group: (1) total fast to achieve 5% WL in 6 days; (2) VLCD (2.5 MJ/d) to achieve 10% WL in 3 weeks; (3) LCD (5.0 MJ/d) to achieve 10% WL in 6 weeks. <I>Results</I>: The main findings of this thesis are, (i) There was little evidence of energy balance regulation in response to an acute total fast (36 h) in human subjects; (ii) Rate of WL is important in determining tissue loss. A slowest rate of WL (with LCD) induced the greatest loss of fat mass, relative to lean tissue. The VLCD only had a protein sparing effect, relative to fasting, after initial loss of LBM; (iii) Rate of WL had a pronounced effect on subjectively-rated fatigue. This, in turn, influenced physical activity and hence total daily energy expenditure. The faster the rate of weight loss, the more fatigued the subject felt. WL had no effect on muscle function nor central nervous fatigue; (iv) Fasting is a stressful means of weight loss leading to a negative mood, which was reversed upon re-feeding. Extent of WL positively influenced mood; (v) All groups cognitively restrained <I>ad libitum</I> intake, post weight loss. There was a negative correlation between increased restraint score and body weight maintenance (-0.519; p=0.027). <I>Conclusion</I>: Rate and extent of WL are both important in determining the success of WL in terms of tissue loss, health, well-being and quality of life. Starvation cannot be recommended as a means of WL. The most effective WL strategy, should be a LCD, conducted over a period of weeks, resulting in a slow rate of WL to maximise fat loss and health benefits.

The incidence and epidemiology of hip fracture in Shenyang, a large city in the north-east of PR China was investigated by collecting data from the admission and medical records from the 36 hospitals in Shenyang. Females and males in the age ranges 25-35 years and 65-75 years (48 to 50 in each group, 194 in total) were recruited for a cross-sectional comparison of bone mineral content (BMC), diet, anthropometry and bone metabolism. BMC was measured using dual energy X-ray absorptiometry at the lumbar spine and hip. Nutrient intakes of subjects were assessed by 5-day weighed food records. The biochemical markers of bone metabolism measured in plasma and urine were 25(OH)D, 1,25(OH)2D, parathyroid hormone, calcitonin, plasma and urinary calcium, phosphorus, total and bone specific alkaline phosphatase, osteocalcin, urinary sodium, potassium, deoxypridinoline and urinary titratable acidity. In addition, biochemical markers of bone metabolism of the young and old Chinese females in the cross-sectional study were compared with that of 32 British (19 old and 14 young) and 33 Gambian females (22 old and 11 young). The results demonstrate that the incidence of hip fracture in Shenyang was low, especially in women compared with more affluent countries; males had a higher overall incidence. The low incidence cannot be explained in terms of superior bone mineral status because old women and men both had low bone mineral status compared with young people, and bone mass was much better in old men than in old women. The pronounced low bone mass in old women was associated with increased bone turnover, but the reason for the low bone mineral status in old men is not clear. Poor vitamin D status, a low calcium and a high sodium intake appear to be the major risk factors for bone health in this population.

This thesis explores, develops and implements modelling strategies for studying relationships between childhood growth and later health, focusing primarily on the relationship between the development of body mass index (BMI) in childhood and later obesity. Existing growth models are explored, though found to be inflexible and potentially inadequate. Alternative approaches using parametric and nonparametric modelling are investigated. A distinction between balanced and unbalanced data structure is made because of the ways in which missing data can be addressed. A dataset of each type is used for illustration: the Stockholm Weight Development Study (SWEDES) and the Uppsala Family Study (UFS). The focus in each application is obesity, with the first examining how the adiposity rebound (AR), and the second how the adiposity peak (AP) in infancy, relate to later adiposity. In each case a two-stage approach is used. Subject-specific cubic smoothing splines are used in SWEDES to model childhood BMI and estimate the AR for each subject. As childhood BMI data are balanced, missingness can be dealt with via mUltiple imputation. The relationship between the AR and late-adolescent adiposity is then explored via linear and logistic regression, with both the age and BMI at AR found to be strongly and independently associated with late-adolescent adiposity. In the UFS, where childhood BMI data are unbalanced, penalised regression splines are used within a mixed model framework to model childhood BMI and estimate the AP for each subject. The data correlations induced by the family structure of the observations are addressed by fitting multilevel models in the second stage. Both age and BMI at AP are found to be positively associated with later adiposity. The two nonparametric modelling approaches are found to be effective and flexible. Whilst the thesis concentrates on BMI development in childhood and later adiposity, the techniques employed, both in terms the modelling of growth and the relating of the derived features to the outcomes, are far more widely applicable.