PhD Studentships in Health Analytics and Modelling

Supervisory team

Imperial:

Dr Erik Volz

LSHTM:

Dr Stephane Hue

UKHSA:

Meera Chand
Andre Charlett

Brief description of project / theme

Efficiency of pathogen genomic surveillance systems (GSS) depends on intelligent design based on best practices and principles of sample survey methodology. Large-scale genomic surveillance of SARS-CoV-2 has provided a vivid demonstration of the promise of GSS and revealed new challenges for the analysis of epidemiological big data. Optimal design of GSS must account for different types of analyses and the expected outputs that are provided to stakeholders. The implementation of new GSS’s will be suboptimal or yield uninterpretable data if plans for data collection are divorced from analytical pipelines. This PhD project concerns the development of bespoke sample designs for GSS, including sample size calculations and power analysis leveraging clinical, genomic and public health surveillance data. We will conduct research into optimal design of GSS with a view towards developing guidelines for design of GSS to meet a wide variety of public health needs across a wide variety of pathogens.

This project will involve the design of a GSS which begins with the anticipated outputs of the surveillance system and involves the optimisation of data collection under resource and logistical constraints. This project will require the Ph.D. candidate to gain skills in pathogen genomic analysis, phylogenetics, population genetics, epidemiological modelling, and samplesurvey methodology.

• Objective 1. We will carry out research into GSS design bestpractices, including methods for sample size calculation and the utility of different genomic data streams, including clinical data, community sampling, random household surveys, and environmental sampling including wastewater surveillance. This will involve the optimisation of sampling effort across data streams and the investigation of surge-sampling strategies.
• Objective 2. Design choices are more difficult for surveillance based on contact tracing or convenience sampling. It is common for convenience sampling from clinical sources to be used by necessity in GSS, and we will provide methods and statistical modelling to debias such samples using patient-level covariates. We will develop methods tailored for clustered data, contact tracing studies, and household surveys, providing optimal genomic sequencing choices within contact pairs or other highly correlated samples.
• Objective 3. We will develop easy-to-use software libraries and dashboard for sample design choices. This will facilitate uptake of these procedures to a wider community and translation to real-world applications.

The role of the different institutions in this collaborative project

Imperial:

• Tutorial elements related to infectious disease modelling
• Advice on infectious disease models tailored to various sample design tasks
• Primary supervision on development of methods, models and software for sample design

LSHTM:

• Tutorial elements related to pathogen evolution and genetics
• Advice on genetic clustering methods and applications to infectious disease surveillance

UKHSA:

• Contribution of data and expertise related to current and planned genomic surveillance systems
• Provide experience to student of working within a public health agency

Particular prior educational requirements and skills for a student undertaking this project

It is expected that the student will have a Master-level degree or equivalent experience in a computational field (bioinformatics, computational biology, computer science) and/or a quantitative (mathematics, statistics) field.

Skills we expect a student to develop/acquire whilst pursuing this project

• Principles of sample survey methodology and statistics
• Infectious disease modelling
• Phylogenetic analysis

Supervisory team

Imperial:

Dr Anne Cori

LSHTM:

Professor Sebastian Funk

UKHSA:

Dr Edwin van Leeuwen

Brief description of project / theme

The reproduction number (R) and generation time are fundamental parameters in infectious disease epidemiology, guiding understanding of disease spread and informing outbreak response strategies. However, their estimation is complex due to the interdependence between R and generation times, and the challenges introduced by temporal changes in these parameters. These temporal changes can have multiple causes including changes in pathogen biology (e.g. new variants), changes in population-level behaviours (possibly prompted by interventions or public health messaging) and changes in the immune landscape in the population. Existing methods often fail to fully capture these dynamics, limiting their reliability for real-time decision-making during outbreaks.

This project aims to address these challenges by developing robust statistical methods for the joint estimation of R and generation times, with a focus on their dynamic interplay. This will include using household-level data from COVID-19 and other diseases in order to determine their utility for real-time decision-making when analysed jointly with population-level data. It will also explore how temporal changes in one parameter affect the estimation of the other and what the implications are for outbreak modelling. The research will build on existing software tools for real-time estimation of reproduction numbers such as the EpiEstim and EpiNow2 R packages and ultimately aim to improve situational awareness in infectious disease outbreaks.

Gostic KM, McGough L, Baskerville EB, Abbott S, Joshi K, Tedijanto C, et al. (2020) Practical considerations for measuring the effective reproductive number, Rt. PLoS Comput Biol 16(12): e1008409. https://doi.org/10.1371/journal.pcbi.1008409

Nash RK, Nouvellet P, Cori A (2022) Real-time estimation of the epidemic reproduction number: Scoping review of the applications and challenges. PLOS Digit Health 1(6): e0000052. https://doi.org/10.1371/journal.pdig.0000052

The role of the different institutions in this collaborative project

The student will be supervised by a team of supervisors representing all participating institutions and part of a team of researchers that spans the institutions. The Imperial College and LSHTM supervisors will bring the substantial methodological expertise and breadth of their institution as well as experience in bringing these to bear on public health problems. UKHSA will bring the UK-specific public health focus and ensure the project is firmly grounded in the application to data and problems as they are present in the UK.

Particular prior educational requirements and skills for a student undertaking this project

A postgraduate degree, ideally in a quantitative subject (e.g. Biostatistics, Bioinformatics, Mathematics, Statistics, Computer Science or Physics) or a related discipline (e.g. Epidemiology or Biology) with a strong quantitative element either awarded or imminent or equivalent training. Also some coding experience, ideally in R.

Skills we expect a student to develop/acquire whilst pursuing this project

Insights into the application of quantitative techniques in public health contexts, specifically advanced analytics applied in epidemiological contexts; inference with mathematical models applied to infectious disease data sets.

Supervisory team

UKHSA:

Professor Peter White (primary)

Imperial:

Professor Helen Ward

LSHTM:

Dr Nicholas Davies

Wider supporting team:

• REACT team members: Dr Christina Atchison (Consultant in Public Health Medicine), Dr Matt Whitaker (Research Associate in Computational Epidemiology), Professor Marc Chadeau, Dr Elizaveta Semenova (Lecturer)
• STI modelling: Dr Lilith Whittles (Lecturer)
• UKHSA Immunisation: Professor Shamez Ladhani

Brief description of project / theme

Sexually-transmitted infections present a growing threat to global public health, with increasing incidence, increasing antimicrobial resistance leading to multidrug-resistant infection, and the emergence of mpox - with men who have sex with men (MSM) disproportionately affected in the UK. Fortunately, vaccines are available for gonorrhoea and mpox, and are in development for other STIs. For gonorrhoea the Bexsero (4CMenB) meningitis B vaccine offers partial protection, and gonorrhoea-specific vaccines are in development, with results from trials expected during the project.

Key questions are how to target vaccination to maximise health gains and address inequalities within a limited health budget, which requires identifying groups most at risk, and understanding their views on vaccination (“vaccine sentiment”) to inform promotion campaigns and provision of services. We published in Lancet Infectious Diseases (2022) the first transmission-dynamic model of gonorrhoea vaccination used for cost-effectiveness analysis, which underpinned the decision by the UK’s Joint Committee on Vaccination and Immunisation (JCVI) to advise the UK to become the first country to implement a national vaccination programme, targeting high-risk persons – primarily MSM.

In further modelling work in Journal of Infectious Diseases (2024) we showed the importance of understanding variation in vaccine sentiment (i.e. the rates of uptake of vaccination in different population subgroups). We now have an exciting opportunity to collect the most representative data set on sexual behaviour and vaccine sentiment (for multiple vaccines) in MSM, and use it to parameterise a transmission-dynamic model of STIs more robustly than has been previously possible.

The student will be involved in analysis of both parts of the project:

1. The novel data set will come from targeted surveys of participants in the REACT cohort, which comprises 2.5 million people across England who were recruited through population-based sampling and who consented to contact for further study. Previously, population-based studies of MSM have been small (so providing limited information), and large studies of MSM have been convenience samples (with unquantifiable biases).
2. Transmission-dynamic modelling will synthesise data from the behavioural survey with surveillance data on rates of testing and diagnoses, estimates from literature on natural history of infection, and estimates of vaccine effectiveness from observational studies and trials using Bayesian methods. This model will represent in detail heterogeneity in sexual behaviour and mixing patterns in the population, linked to vaccine uptake and healthcare utilisation. It will then be used to optimise vaccination strategies.

The intended priority of the work will be gonorrhoea, as this is a complex problem with rising incidence, drug resistance being a growing concern, and novel vaccines in development. However, mpox might become a priority if there are changes in its transmission patterns or emergence of new variants of concern. There will be scope to tailor to work to the student's particular interests as the project evolves. For example, there can be further data collection, including qualitative (via focus groups) or quantitative (further surveys), use of machine learning in data analysis, or development of methods for dynamic representation of behaviour in transmission-dynamic models (linking with other work of the HPRU).

Related publications:

• Nikitin D, Whittles LK, Imai-Eaton JW, White PJ. Cost-effectiveness of 4CMenB vaccination against gonorrhea: importance of dosing schedule, vaccine sentiment, targeting strategy, and duration of protection. Journal of Infectious Diseases 2024: jiae123. doi.org/10.1093/infdis/jiae123.
• Whittles LK, Didelot X, White PJ. Public health impact and cost-effectiveness of gonorrhoea vaccination: an integrated transmissiondynamic health-economic modelling analysis. Lancet Infectious Diseases 2022; 22: 1030–41. doi.org/10.1016/S1473-3099(21)00744-1.
• Whittles LK, White PJ, Didelot X. Assessment of the Potential of Vaccination to Combat Antibiotic Resistance in Gonorrhea: A Modeling Analysis to Determine Preferred Product Characteristics. Clinical Infectious Diseases 2020; 71(8): 1912–1919. doi.org/10.1093/cid/ciz1241.
• Whittles LK, White PJ, Didelot X. A dynamic power-law sexual network model of gonorrhoea outbreaks. PLOS Computational Biology 2019; 15(3): e1006748.
• Whittles LK, White PJ, Didelot X. Estimating the fitness cost and benefit of cefixime resistance in Neisseria gonorrhoeae to inform prescription policy: a modelling study. PLOS Medicine 2017; 14(10): e1002416.

The role of the different institutions in this collaborative project

UKHSA’s Immunisation Division provides advice to the Department of Health & Social Care and JCVI on vaccination policy, and to the health service in implementation. UKHSA’s Blood Safety, Hepatitis, STIs and HIV (BSHSH) Division collects STI surveillance data and provides advice on STI control. The student will gain experience of how this advice is generated and communicated as part of the policy-making process.

Imperial College developed the gonovax model in R (https://rdrr.io/github/mrc-ide/gonovax/), which will be the basis of the modelling component of this project. As this model has evolved it has been used for the papers in Clinical Infectious Diseases 2020, Lancet Infectious Diseases 2022, and Journal of Infectious Diseases 2024, and is used in other papers in preparation.

• The work in Lancet Infectious Diseases 2022 provided the essential cost-effectiveness evidence which underpinned JCVI’s decision to advise a national targeted vaccination programme.
• The work in Journal of Infectious Diseases 2024 on the importance of understanding variation in vaccine sentiment is what inspired this project.

Imperial College is also modelling other STIs including mpox. Imperial hosts the Vaccine Impact Modelling Consortium, which has many international members.

The REACT study is run by Imperial College and provided vital population-based information on COVID. It was also been used for other studies.

Imperial’s Patient Experience Research Centre promotes participatory approaches to improving healthcare and biomedical research.

LSHTM hosts the NIHR Health Protection Research Unit in Immunisation, a partnership based at LSHTM and with additional researchers at UKHSA, UCL, and the University of Cambridge. The HPRU in Immunisation performs modelling and costeffectiveness analysis of changes to the UK’s immunisation policy for JCVI. In addition to modelling, the HPRU in Immunisation also has strong programmes in social science and electronic health records research pertaining to vaccines, which will help the student gain experience in critical considerations of vaccine acceptance and real-world data to ensure that their work is grounded and robust.

Particular prior educational requirements and skills for a student undertaking this project

• Masters degree (or equivalent) in a relevant quantitative subject, ideally with experience in transmission-dynamic modelling.
• Experience of analysis of survey data would be beneficial but not essential.
• Strong skills in coding, ideally in R, but skills developed in other languages are acceptable if willing to use R for this project.

Skills we expect a student to develop/acquire whilst pursuing this project

This project will train the student in epidemiology, modern survey data analysis methods, advanced coding techniques, model design and implementation, infectious disease transmissiondynamic modelling, and Bayesian parameter inference. It will provide an introduction to health-economic analysis integrated with transmissiondynamic modelling.

There will be opportunities to engage with affected communities through Patient and Public Involvement and Engagement (PPIE) activity, and potentially through running focus groups to use qualitative research to help understand quantitative patterns detected in the survey, or to design follow-up surveys to collect additional data. The student will build skills in critical appraisal of literature, and communications skills including, visualisation of data and results, oral presentation skills, and scientific writing as well as. The study will provide experience in project management, collaboration across disciplines and institutions, and translation of research to inform policy and practice.

Supervisory team

Imperial:

Professor Katharina Hauck (lead supervisor)

LSHTM:

Professor John Edmunds

UKHSA:

Michael Borowitz (Michael.Borowitz@ukhsa.gov.uk)

Brief description of project / theme

The world needs to get prepared to prevent a catastrophic pandemic from ever happening again, and effectively respond to emerging threats. Imperial has launched the ‘The Jameel Institute – Kenneth C Griffin Initiative for the Economics of Pandemic Preparedness’ (EPPI), an ambitious international program of research on the health, economic and educational impacts of epidemic threats. Aims are to project the societal losses associated with hypothetical future pandemics under alternative scenarios, develop modelling tools that can provide real-time evidence on the health and economic impacts of mitigation measures during outbreaks, and evaluate the broader returns to investments into pandemic preparedness.

EPPI is novel and innovative in its integrated approach to epidemiological, economic, and social research by drawing upon advanced methods of data science and analytics including modelling of infectious transmission dynamics, micro- and macro-economic analyses, econometrics, behavioural science, and health economics. EPPI has developed an integrated economic-epidemiological model DAEDALUS for the COVID-19 pandemic in the UK, and DAEDALUS Explore, an online dashboard that projects the health, economic and educational losses associated with 7 hypothetical future respiratory pandemics for 67 countries in the world.

The PhD project will be based within EPPI. The specific aims of the PhD will be to build a UK version of DAEDALUS that can be used for real-time modelling to support policy makers in emergency response to an outbreak. The model will be powered to project health, economic and education impacts of alternative mitigation strategies into the future, relying on automated data streams from other projects within the HPRU. A particular focus will be to build in fiscal impacts considering alternative fiscal relief interventions, including government transfers such as furlough payments, and subsidies to businesses.

A second objective of the PhD project is to use the model to estimate the societal value and return-oninvestment to UK specific pandemic prevention and preparedness investments including investments into vaccine R&D and manufacturing capacity, surveillance and test-and-trace capacity.

The PhD project will generate urgently needed modelling tools to assess the complex and varied impacts of severe epidemics on population health, economic welfare, and the social fabric of the UK society.

References

‘The Jameel Institute – Kenneth C Griffin Initiative for the Economics of Pandemic Preparedness’: DAEDALUS Explore (2024). A pandemic simulation tool. https://daedalus.jameelinstitute.org/scenarios/new

‘The Jameel Institute – Kenneth C Griffin Initiative for the Economics of Pandemic Preparedness’: DAEDALUS (2024). Software. https://github.com/jameelinstitute/daedalus

Haw, David J., Giovanni Forchini, Patrick Doohan, Paula Christen, Matteo Pianella, Robert Johnson, Sumali Bajaj, et al. 2022. Optimizing Social and Economic Activity While Containing SARS-CoV-2 Transmission Using DAEDALUS. Nature Computational Science 2 (4): 223– 33.

The role of the different institutions in this collaborative project

Katharina Hauck will integrate the student in a team of established researchers and software engineers that work in closely related fields. John Edmunds will contribute his unique experience in real-time modelling and epidemiological-economic analysis to the supervisory team. Michael Borowitz will make sure that the direction of the PhD remains aligned with UKHSA priorities on pandemic preparedness.

Particular prior educational requirements and skills for a student undertaking this project

We are looking for a student with very strong quantitative skills, with a background in mathematical epidemiology, quantitative economics, or related subjects. The student should already have a Masters level qualification in these subjects.

Skills we expect a student to develop/acquire whilst pursuing this project

The student will most likely come from either an epidemiological/modelling background, or economic background, and will need to acquire skills from the respective other discipline.

Supervisory team

Imperial:

Dr Elizaveta Semenova
Professor Marc Chadeau-Hyam

UKHSA:

Professor Andrew Hayward

LSHTM:

An LSHTM supervisor will be identified based on the focus of the project.

Brief description of project / theme

This PhD project will investigate the impact of deprivation-induced environmental, viral, behavioural, phyco-social exposures on human health with focus on chronic conditions, including cancers and cardio-respiratory, and cognitive outcomes.

Taking an exposome approach, we will investigate social gradients in (co-occurring and correlated) exposures including air pollution, noise, living environment (including green and blue space proximity, urban density, food environments, and area-level deprivation), behaviours along with history of viral infection to characterise (complex and multifacetted) exposure profiles associated with social adversity, and will explore the geographical distribution of these profiles. This research will examine how these profiles affect health outcomes, including incident chronic conditions and multimorbidity. The definition of deprivation will not only rely on established metrics but will also leverage the data to come up with a reproducible and interpretable definition of ‘socio-environmental deprivation’ affecting health. The data will also offer the possibility to investigate the biological signatures of exposure profiles, through multiple omics data available in the REACT study and their joint and marginal effects on health.

The REACT study includes data on more than 2mio individuals with detailed socio-demographic information and exposure to SARS CoV-2. Multiomics data have been generated in over 10K participants including genome sequencing, proteomics and metabolomics. Data on environmental exposures will be obtained through linkage to existing surfaces generated by partners, and already accessible. REACT data will be made available, subject to submission of access request.

The project will use modern computational methodologies, including deep learning-based clustering to identify patterns of comorbidities and spatiotemporal models to assess the impacts of exposures. Mediation analysis will further disentangle causal pathways, revealing how environmental and social determinants collectively shape health disparities. By linking data from the REACT study with existing exposure surfaces, this research aims to advance our understanding of the biological and geographical dimensions of health inequities along the social gradient, providing evidence for targeted public health interventions.

The role of the different institutions in this collaborative project

The successful applicant will be part of the Department of Epidemiology and Biostatistics (EBS) within Imperial College London's School of Public Health as well as the Health Equity and Inclusion Unit at UKHSA. EBS provides a vibrant environment with an opportunity to interact with other researchers focusing on exposome.

Particular prior educational requirements and skills for a student undertaking this project

• MSc (or equivalent training) in a quantitative subject, such as statistics, data analysis, mathematics, computer science or a closely for a student undertaking this project relate discipline. Ideally with some experience of computational statistics.
• Good coding skills in R and/or Python

Skills we expect a student to develop/acquire whilst pursuing this project

This project will train the student in computational epidemiology, exposome research, statistical inference, and, potentially, deep learning.

They will also build skills in scientific writing, as well as critical reading of the literature. The student will also improve their skills in research coding and management and learn to give clear scientific presentations communicating their work.