Skip to main content
Wolfson Institute of Population Health

Current Students

 

We have an intake of 6 students to the programme each year. Browse the tabs to find out more about the current students and their projects.

One of our first cohort of students, Jing Hui Law, shares more about her PhD project in the video below.

 

 

The current cohort started their PhD projects in September 2021. See below project details and click on the links for their profiles.

Nick Branson

Supervisors - Prof Conrad Bessant, Prof Pedro Cutillas

"Using Artificial Intelligence to Leverage Proteomics Data for Stratified Medicine"

Stratified medicine, in which information from a patient’s genome is used to help direct their treatment, has revolutionised some areas of clinical practice. However, most diseases do not involve changes in an individual’s genome, so the power of insights derived from genotyping and whole genome sequencing is limited. 

 

Interrogation of a patient’s proteome (the collection of 20,000+ unique proteins that can be extracted from human tissue and body fluids) promises much more fine-grained insight into the state of a patient, because proteins are the key functional molecules of biology. The abundance, location, physical structure and activity of each protein can vary substantially during disease progression, providing a wealth of information for assessing a patient’s state and forming a treatment plan. But how to reliably turn proteomic data into clinical insights remains an open question. 

 

This project will investigate the extent to which artificial intelligence (AI) can be used to leverage proteomics data for stratified medicine. This will be achieved by applying cutting edge data science and AI techniques with the aim of producing reliable and explainable readouts of disease state from clinical proteomics data, overcoming inherent challenges in these data such as heterogeneity, missingness and lack of statistical power. 

"What drew me to the HDiP programme was the application of statistical methods to important problems in healthcare coupled with the human centred approach to data."

Anya Jacobs

Supervisors - Prof Borislava Mihaylova, Dr Florian Tomini, Dr Anna Di Simoni

"Developing a health economic framework for assessing asthma management strategies in UK primary care"

Asthma is a chronic inflammatory disease of the lungs with high morbidity that represents a major use of healthcare resources. This burden is unevenly distributed across the asthma population, with individuals suffering from more severe forms of the disease having higher morbidity and mortality risks and lower quality of life, accounting for a greater share of resources and costs. There is an urgent need to reduce this burden by appreciating the complexity of asthma and exploring ways to improve its management in view of individual demographic, lifestyle and clinical characteristics and asthma severity.

We aim to develop an evaluative framework for adult asthma that is able to track the long-term health outcomes, quality of life, and healthcare usage and costs of individuals with asthma according to their disease severity status and sociodemographic characteristics, thereby taking into account the diversity of the asthma population and the fact that a ‘one-size-fits-all’ approach to managing or treating asthma is ineffective. The asthma population will be derived from the OPCRD database, which contains comprehensive clinical data on over 18% of the UK population. The model should help to inform key healthcare decisions regarding the efficient management of asthma with focus on UK primary care.

"I was thrilled to join the Health Data in Practice PhD programme in 2020. I’m excited to undertake the MRes year and following PhD in the interdisciplinary and incredibly supportive environment that QMUL offers." 

Jing Hui Law

Supervisors - Dr Sarah Finer, Prof David van Heel, Dr Nina Fudge, Dr Magda Osman

"Applying a life course perspective to genetic risk perception for type 2 diabetes in British Bangladeshis
and British Pakistanis"

Research on the communication of genetic risk information for common conditions like type 2 diabetes (T2D) has largely focused on disease identification, prevention and management in older, White European populations. Much remains unknown about the use of genetic risk prediction tools in diverse ethnic and/or age groups—however, emerging work in South Asian populations suggest that such tools may have added benefits for the identification of younger people at risk of developing T2D. We are interested in applying a life course perspective to this research area to examine potential age differences in perceptions and beliefs surrounding T2D—including whether young people perceive unique benefits from acting on their genetic risk status. We will first synthesise currently available evidence on genetic risk communication and illness perceptions. Then, we will provide genetic risk information about T2D to a community-based cohort of British Bangladeshis and British Pakistanis in the UK—integrating the findings from our synthesis—to investigate any age differences in how people evaluate information and formulate ideas about their risk; and compare different formats of risk communication. This research will inform efforts addressing the potential of genetic risk communication for a population disproportionately affected by T2D—yet underrepresented in genetics research.

“Multi-disciplinarity research has always been the focus of my passion and commitment, and I feel very strongly that this aligns well with the framework of the programme. It places prominence in the different methodological issues and challenges surrounding the interdisciplinary nature of health data, and draws from the experience of a very attractive supervisory pool coming from various fields.”

Amaya Gallagher Syed

Supervisors - Prof Mike Barnes, Dr Myles Lewis

"Understanding the aetiology and stratification of Immune Mediated Inflammatory Diseases using AI"

Our goal is to use Artificial Intelligence (AI) to better understand the biology of immune-mediated inflammatory diseases (IMIDs) such as rheumatoid arthritis, psoriasis, lupus, and Sjögren’s syndrome. Despite being common conditions, the mechanisms which explain how IMIDs work are poorly understood. We consider there must be key cellular mechanisms underlying IMIDs and that studying the differences between them will provide clues as to their origin and development. This project will focus on studying interferon signalling, a central mechanism for control of inflammation, known to influence IMID development and drug response. Using genomic data from a range of IMIDs, we will build networks explaining the relation between the different components of interferon signalling. We will also integrate patient clinical information, genomic and imaging data to find sub-groups of patients whose disease and response to treatment progresses in a similar manner. Finally, we will use this research to identify pharmaceutical drugs, already on the market and in use for different conditions, which might be effective in the treatment of IMIDs because their mechanism of action indicates they would affect important areas of IMID biology. This could also lead to identifying areas of interest for future IMID drug development research.

"I was immediately drawn to the Wellcome Trust Health data in practice PhD programme because it emphasises the importance of developing a Human-centered data science, understanding how and why data is generated and how it is used by practitioners. I believe these points are key to developing meaningful models and insights, which add value both for the patient and the medical community, and that they are the cornerstone of future advances in AI and Machine Learning applied to health data."

Xuan (Charis) Xie

Supervisors - Dr Clare Relton, Dr Anna Di Simoni, Prof Hilary Pinnock

"Exploring the use of routine health data in improving the efficiency of randomised implementation
trials"

Trials that assess methods to help implement health interventions of proven effectiveness are known as implementation trials. However, randomised trials tend to be expensive and take many years to conduct. The use of patient health data collected from routine practice, such as GPs, hospitals, and secondary care services, may make these types of trials more efficient. However, there are many challenges such as gaining access to such data and the data might be inaccurate.

In this PhD I will explore how patient health data are used in implementation trials. I plan to start by reviewing reports of these types of trials and consulting the experts in the field. I will survey and interview those who have conducted or have been involved in such trials (e.g., trialists, IT staff, statisticians, researchers, patient participants etc.) and carry out some in-depth case studies. The involvement of patients and public representatives in all stages of the study will be vital. The key outcome will be a set of recommendations on the use of routine health data in studies implementing health interventions into practice. I hope the outputs of this PhD will increase the speed of adopting health interventions of proven effectiveness into real-world.

The current cohort started their PhD projects in September 2022. See below project details and click on the links for their profiles.

Laetitia De Abreu Nunes

Supervisors - Prof Richard Hooper, Dr Rachel Phillips

"Moving beyond frequency tables for the analysis and presentation of harm outcomes in
randomised controlled trials"

When conducting a randomised controlled trial to evaluate the impact of an intervention,
such as a new drug or vaccine, researchers also collect data on harms experienced by
patients. In clinical trials, these are called adverse events. Sometimes these events are
prespecified as events of interest. In addition, information on non-prespecified harms will
also be collected during a trial : these are referred to as emerging events. However, the
standards of practice for the analysis of emerging harm data are poor. Often practice relies
on the use of simple methods that limit the utility of such data or on inadequate use of
statistical techniques. To address this gap, I will aim to appraise existing specialised
methods to analyse harm outcomes using simulations and real-world clinical trial case
studies. I will engage and work with relevant stakeholders using the results of the
simulations to inform and develop recommendations for best practice. I will develop tutorials
and software to increase and facilitate the uptake of recommended approaches. The
desired impact is the adoption of these recommendations by the clinical trials community,
thus improving the analysis of harm outcomes in randomised controlled trials.

I was drawn to the interdisciplinary nature of the HDiP programme and the opportunity to address meaningful problems in health research through the use of statistical and computational methods. I've loved the supporting community and environment the programme provides, in particular through the multiple cohorts of students and the great management team, and I'm very excited to embark on my PhD project this year!

Rebecca Muir

Supervisors - Dr Nina Fudge, Dr Manuella Perrotta, Dr Meredith Hawking

"An interpretative policy analysis of BMI lifestyle criteria for IVF access"

 Due to financial struggles and increasing demands, local commissioners within the NHS must make difficult decisions about how healthcare funds are allocated. lVF is a particularly controversial area of healthcare rationing, and there is a well-documented ‘postcode lottery’ in treatment. For several years, a variety of criteria including BMI has been used to restrict care by different local commissioning groups, soon to be Integrated Care boards (ICBs).

These criteria are based on NICE recommendations and commissioning group deliberation. However, previous work has shown that healthcare rationing is not coldly evidence-based: ‘intuitive rationality’ is inherent to the health rationing process: gut-feelings, compassion and relatability all play a role in how healthcare is rationed (Russell, 2017).

However, it has not been explored how political/social value judgements and moral discourses influence the introduction of lifestyle criteria. BMI as a criterion is a good case study as obesity is a highly stigmatised condition and BMI has been criticised as a blunt tool for understanding individual health.

This project will use interpretive policy analysis (IPA) to understand how BMI is used as a lifestyle rationing criteria, to understand how different levels of stakeholders, from NICE, to local commissioning, to (potential) patients, interpret BMI restrictions. The social, political and cultural values underpinning perspectives on the criteria will be investigated.

The MRes element of the programme really appealed to me due to the breadth of modules, as I wanted to gain experience qualitative methods and understand health data from different perspectives.

Due to the training and research expenses offered to you as a Wellcome-funded student, you have the opportunity to be ambitious with your PhD project and invest in your skills throughout the PhD.

Being part of a cohort of students appealed to me, it is great to be going through the experience with other students.

 

Binur Orazumbekova

Supervisors - Prof Rohini Mathur, Dr Sarah Finer, Dr Moneeza Siddiqui (University of Dundee)

"Ethnic variations in risk trajectories for type 2 diabetes: an observational research study combining genetic, cohort, and electronic health record data in the UK"

Type 2 diabetes (T2D) is a chronic disease with increasing incidence and prevalence worldwide. For a long time, people living with pre-diabetes were considered as the only high-risk population for T2D, which led most of current interventions to target this group. However, recent studies demonstrate that only one-third of people with pre-diabetes eventually develop disease. Moreover, most of those studies were conducted on the older and white populations and did not take into account the heterogeneous nature of disease. Therefore, current evidence may not be generalizable to the whole world population, as it is known that some ethnic groups are at higher risk of developing diabetes at a younger age and at lower thresholds of some modifiable and non-modifiable risk factors compared to people with white ethnic background, but may never be clinically identified as having pre-diabetes. Therefore, in this study we will systematically review scientific literature to elucidate global differences in factors associated with high-risk of developing T2D; Model the transition to T2D utilising representative population from electronic health records to understand the role of modifiable and non-modifiable risk factors and potential interactions among them; and identify ethnic variations of risk trajectories for T2D.

I was very excited to join Wellcome Trust-funded Health Data in Practice program, which gives an opportunity to gain necessary research skills during the MRes year and to shape your own human-centred interdisciplinary PhD project. I believe with this support and resources I will contribute to the diabetes research field and positively impact the prevention of this global pandemic.

Lizzie Remfry

Supervisors - Prof Mike Barnes, Dr Sarah Finer, Prof Rohini Mathur

"Using artificial intelligence to examine the relationships between multiple long-term conditions and care pathways through the lens of inequalities"

Background

Living with multiple long-term conditions (MLTC) presents a burden to individuals and the healthcare system, leading to poorer health outcomes and a shorter life expectancy. The presence of MLTCs is worsened by inequalities, driven amongst other things by structural biases, differential access to healthcare, socio-economic status (SES), ethnicity and gender1. The rise of AI in healthcare threatens to exacerbate this situation as machine learning techniques can amplify existing inequalities2.

Approach

Using machine learning techniques, we will use healthcare data to map the relationships between MLTCs and their treatments. We will characterise the relationship between the health trajectories and recommended treatment guideline such as NICE. By using diverse datasets, we can also look at how different drivers of inequalities impact these pathways.
This project will use explainable machine learning methods to ensure AI is used in an ethical manner. In order to communicate methods and results effectively, we will conduct focus groups with clinicians and non-experts on how to explain the findings.

Impact

We hope to understand how inequalities impact the different health journeys of patients with MLTCs and how this progression interacts with care pathways and treatment. This knowledge can help inform guidelines and develop new strategies to manage MLTCs which recognise and where possible tackle health inequalities.

The 2022 cohort have recently embarked on their PhD projects, read more about their research below.

Gracia Andriamiadana

Supervisors - Dr Zahra Raisi-Estabragh, Prof Steffen E. Petersen, Dr Liliana Szabo

"The burden and mechanisms of multi-organ morbidity in people with past cancer"

Improvements in treatments have led to an increase in the number of people surviving cancer. Cancer treatment places significant strain on physical and mental health of patients, which can impact well-being for many years after treatment completion. Recent studies show that as people survive for longer after their initial cancer diagnosis, non-cancer illnesses become more important as causes of ill health or premature death. In the UK, the number of people surviving cancer has doubled in the last 40 years, with more than half of the people diagnosed with cancer surviving at least 10 years after their diagnosis.

Optimising the long-term health of cancer survivors is an important public health priority. The UK Biobank is a very large research resource with extensive linkages to clinical records and highly detailed characterisation of health across multiple organ systems. We aim to comprehensively characterise multi-organ health of UK Biobank participants with past cancer across the following themes: heart health, brain health, mental well-being, cardio-metabolomic profile. We will define the excess burden of disease across these areas in cancer survivors (compared to participants without cancer). Disease mechanisms will be explored using detailed medical scans and bloodwork using state-of-the-art statistical and machine learning methods.

Jack Brown

Supervisor - Prof Jianhua Wu

"The relationship between neighbourhood determinants and cardiovascular disease risk in England and Wales"

Cardiovascular disease (CVD) is a significant cause of mortality and morbidity, accounting for a quarter of all deaths in the UK. Traditional CVD risk models such as QRISK3 and SCORE utilise clinical measures and some limited demographic information (i.e. sex and age); however, systematic reviews, focussing on determinants of CVD have concluded a need to take a more holistic view of what constitutes ‘CVD risk’, calling for inclusions of environmental and social factors to enhance risk prediction models. Utilising the vast data contained within the CPRD and linking this at the geographic LSOA level with other social and environmental datasets, this project aims to take a holistic view of what impacts CVD risk and develop a machine-learning-based risk model that accounts for environmental and social (neighbourhood) impacts on CVD risk.

The opportunity the MRes year gave me to expand my knowledge and study modules that weren’t offered to me in my BSc/MSc was a big draw for me to the HDiP programme. Additionally, the programme allows you to pursue a health-related PhD question of your choosing, this freedom gives you the opportunity to incorporate the methods and topics you’re most passionate about.

Tooba Hamdani

Supervisor - Prof Borislava Mihaylova

"Develop a framework for economic assessment of wearable medical technology: interdisciplinary research"

Wearable medical technology and associated software applications is a new and rapidly evolving field which holds the promise of revolutionizing the way health is monitored and managed, offering a new paradigm of personalized healthcare by continuously collecting and analysing vital health data in real time. Novel ways of using this technology for screening, diagnosis, delivery of treatment as well as monitoring are being developed which necessitates comprehensive assessment of this technology.

In this study, I aim to develop a broad framework which enables assessment of the economic value of this technology with further consideration of sociological and ethical viewpoints.

The study will be conducted in three steps: first a systematic review of literature will be performed to identify and apprehend current frameworks and relevant literature. Next, substantive case studies of wearable medical technologies and medical mobile applications for cardiovascular diseases and diabetes will be conducted in parallel to consultations with stakeholders. The final step will involve formulation of a broad health economic assessment framework for wearable medical devices and medical mobile applications.

I was drawn to the interdisciplinary and data-driven aspects of the HDiP program. The experience so far has been enriching and broadened my vision. I was interested in digital health from the beginning, however, the modules I took during the MRes part helped me further build and refine my ideas before beginning the PhD. Importantly, the program has a good support system as well!

Alfred Kayira

Supervisors - Dr Garth Funston, Prof Claude Chelala, Prof Fiona Walter

"Using Natural Language Processing to identify symptoms and risk factors associated with lung cancer in primary care records"

Background: Lung cancer is the third most common cancer in England, but the leading cause of cancer death. Most patients with lung cancer are not diagnosed until the disease is advanced, which contributes to the poor patient outcomes. Patients with lung cancer usually visit their GP with symptoms of the disease before diagnosis, and symptom based electronic tools are used to help GPs make decisions about urgent cancer referrals. However, such tools often rely on codes entered by clinical staff, ignoring the vast amount of information entered as free text.

My approach: I will use an approach called Natural Language Processing (NLP) to automatically pull-out key information on risk factors (e.g. smoking) and symptoms of lung cancer from the GP records of around 150,000 patients who were risk assessed for cancer. I will compare this with coded information to determine whether there are differences in the frequency or type of information recorded. I will compare how the accuracy of commonly used lung cancer risk assessment tools changes when using coded and NLP information vs coded information alone.

Impact: This research will help determine whether NLP could improve the accuracy of lung cancer risk assessment and cancer detection in primary care.

What attracted me to HDiP is the programme’s emphasis on an in-practice approach to health data science research and its endeavour to have every patient encounter with the health system count. I am therefore thrilled to undertake a project that seeks to explore the potential of free text data in EHR, which hasn’t been exploited,  to improve the accuracy of lung cancer risk assessment and cancer detection in primary care

Tommy Maltby

Supervisors - Dr Maria Turri, Professor Claudia Cooper, Dr Victoria Tzortziou Brown

"Accessing arts based intervention through social prescribing, for whom does it work and why?"

Research has demonstrated that arts-based interventions can improve health and well-being, help prevent a variety of mental and physical illnesses as well as support the treatment or management of acute and chronic illnesses. Such interventions are typically low-risk, cost-effective and can help provide holistic treatment options for complex health challenges with no current solutions. Social prescribing provides one of the main referral pathways to social and arts programs in primary care. However, collaborations between the arts and health sectors are often not properly developed with limited research investigating barriers to access and effective referral pathways, and mostly focusing on other interventions rather than arts programmes. I aim to address this research gap by first conducting a scoping review to gather evidence on arts-based social prescribing and patient referral pathways within arts in health. Then, I will use data collected by North-East London primary care providers about social prescribing referrals to explore the demographic make-up of those who are prescribed arts interventions, reasons arts are prescribed, and outcomes. I will then conduct a qualitative study building upon our findings to explore barriers and challenges to access to arts-based interventions for diverse populations.

 

Rowan Morris

Supervisors - Dr Silvia Liverani, Dr Leonardo Soares Bastos

"Probabilistic joint seasonal dengue and chikungunya forecasting in Brazil"

Dengue and chikungunya are mosquito-borne viruses, which have been spreading into new
parts of the world in recent years. They are both transmitted by the same two species of
day-biting mosquitoes, Aedes aegypti and Aedes albopictus.
With enough advanced notice, dengue and chikungunya outbreaks can be mitigated. As a
climate-sensitive disease, environmental conditions and past patterns of dengue and
chikungunya can be used to make predictions about future outbreak risk. These predictions
improve public health planning and decision-making to ultimately reduce the burden of
disease.
For this project we partnered with Dr Leonardo Soares Bastos to develop a new method,
inspired by the work in Colón-González et al (2021) and Pavani et al. (2023).
Rowan will extend the method to model bivariate data and apply the model to the
forecasting of dengue and chikungunya in Brazil. These two diseases are closely related,
they share the same vector and the initial symptoms are very similar. Therefore the
development of a joint model ensures borrowing strength over the two outcomes and will
lead to more accurate predictions for both diseases.

I chose the Health Data in Practice program because, whilst I am primarily a statistician, it is vitally important whilst working in the health field that we remain in touch with multidisciplinary approaches with a focus on an “in-practice” context. I chose my particular project to hone my statistical skills, applied to a critical issue in the age of climate crisis.

Back to top