Chronic obstructive pulmonary disease (COPD) is a progressive lung disease requiring timely diagnosis and assessment of the severity as well as intensive treatment and health care support to prevent disease progression. According to a recent National Institute for Health and Care Excellence (NICE) report, there are approximately 1 million people in the UK with diagnosed COPD and another 2 million with undiagnosed COPD. It is currently the fourth leading cause of death worldwide and fifth in the UK. The British Lung Foundation recently estimated the annual cost of managing COPD was £1.9 billion suggesting that COPD has a large burden on healthcare resources. This is expected to rise further as current data show an increasing trend of COPD emergency admissions.
While smoking is the most important risk factor for COPD, several comorbid and pathophysiological conditions along with environmental factors are linked to COPD-associated exacerbation which is the most common cause of hospital admissions in the UK. For clinicians, assessment of the frequency and length of hospital admissions, the time to severe exacerbations and identifying strategies for the management of disease and preventing further progression are all vital.
However, COPD is clinically heterogeneous and therefore standard clinical measures and phenotyping may not assess the risk fully. The electronic health record (EHR) data are increasingly being used for several decision-making processes and are considered a potential resource for predictive modelling and risk assessment. The National Health Service (NHS) routinely manages multiple sources of linked EHR of patients, for example, demography, GP visits, pathophysiology and biochemical parameters, pharmacy, hospital and emergency admissions. More importantly, it is possible to identify longitudinal cohorts of patients with the exact date of exacerbation events which provides valuable information on disease dynamics; a critical element for a heterogeneous disease like COPD.
Harnessing these multiple sources of primary health care information could potentially contribute to the diagnosis, monitoring and evaluating the prognosis of COPD patients. Adopting appropriate strategies tailored to the individual patient would deliver improved healthcare leading to a personalised medicine-based approach to the management of COPD and reduction of burden on healthcare resources. There are, however, several methodological and technical challenges as these would involve the integration of high-volume and high-dimensional complex multi-source data along with the application of diverse mathematical and statistical methodologies and computational tools. In collaboration with an industry partner, we are currently investigating the association of baseline patient-level demography, physical, clinical and other pathophysiological predictors with the frequency and rate of exacerbations among COPD patients at subsequent time points and assessing the relevant health care resource allocation costs. The proposed PhD project will extend the current research work further to developing predictive analytical tools to identify the risk of severe exacerbations integrating multiple sources of EHR data with advanced statistical modelling framework.
The project will use novel methodological techniques to establish the complex relationship between multiple sources of EHR COPD-related data. It will evaluate the distributional properties of the data, explore novel predictive modelling strategies in conventional as well as diverse machine learning and deep learning modelling frameworks and examine their accuracies. An overall aim is the production of an optimised protocol which harmonises the use of EHR data and maximises the reproducibility of outcomes and ultimately improves clinical practice. The project will, therefore, target several key gaps in the current statistical methodologies with implications in wider disease contexts.
The project will provide an excellent opportunity for a prospective student to work in one of the most active and exciting areas of medical research: the student will develop key knowledge and skills of handling high-volume and high-dimensional complex EHR data, learn novel and advanced mathematical and statistical modelling approaches particularly those related to high-dimensional data analysis and apply these advanced tools and techniques to real clinical datasets.
This project is advertised in relation to the research areas of APPLIED HEALTH SCIENCE. Formal applications can be completed online: https://www.abdn.ac.uk/pgap/login.php
. You should apply for Degree of Doctor of Philosophy in Applied Health Science, to ensure that your application is passed to the correct person for processing.
NOTE CLEARLY THE NAME OF THE SUPERVISOR AND EXACT PROJECT TITLE ON THE APPLICATION FORM.
Candidates should contact the lead supervisor to discuss the project in advance of submitting an application, as supervisors will be expected to provide a letter of support for suitable applicants. Candidates will be informed after the application deadline if they have been shortlisted for interview.