PhD on causal inference for competing risks data

Applications are invited for a 3-year PhD studentship from the ESRC UBEL DTP (UCL, Bloomsbury and East London Doctoral Training Partnership)

We are seeking applicants who would like to pursue PhD research on the project described below. This project is offered as part of the Longitudinal Analysis and Design topic under the Quantitative Social Science Pathway of the ESRC UBEL DTP. Successful applicants will based in the Department of Medical Statistics at LSHTM. Further information on the funding scheme can be found at

PhD Title:  Causal inference for competing risks data

Supervisors: Prof. Jonathan Bartlett and Prof. Ruth Keogh, Department of Medical Statistics, London School of Hygiene & Tropical Medicine


In medical studies investigating the effects of treatments or exposures, the outcome of interest is often time to a particular event, such as occurrence of Alzheimer’s disease. In such studies, some individuals may never experience the event of interest, because a so called competing event occurs first which precludes the event of interest from occurring. In the Alzheimer’s example, dying from a non-Alzheimer’s related cause without having a prior Alzhiemer’s disease diagnosis constitutes the competing event.

In recent years it has become apparent that the standard statistical methods for analysing such competing risks data rest on either implausible assumptions or are difficult to interpret. One approach, which involves fitting a so called ‘Cox’ regression model for the event of interest, censors individuals who fail from competing causes. For this analysis, the so called independent censoring assumption that it makes is likely violated when there exist individual level characteristics which affect both the risk of the event of interest and the competing events. In the case of Alzheimer’s disease, it is likely that cardiovascular risk factors exist which affect both the likelihood of developing Alzheimer’s disease and cardiovascular disease, such that the effect estimates are biased. An alternative approach, based on analysing effects on the so called cumulative incidence of the event of interest, does not make such assumptions, but is harder to interpret and arguably is not isolating the effect only on the cause of interest.

Project objectives

This PhD will investigate both existing and recently proposed statistical approaches for analysing competing risks data, including methods based on so called separable effects. The overarching objective is to better understand when the standard analysis approaches are likely to be misleading and whether newly developed approaches, such as that based on separable effects, can overcome the problems with existing standard approaches.

Techniques to be used

The project will involve a combination of statistical analysis of real world observational data, simulated data, and mathematical analytical work. The student will analyse data from the English Longitudinal Study of Ageing, to investigate the potential role of different variables in causing dementia. Death prior to dementia is the competing event.


The student will be based in the Department of Medical Statistics at London School of Hygiene & Tropical Medicine.

Project requirements

The project involves advanced quantitative methods: it is statistical and computational in nature, and applicants will need (or be expecting to get) an MSc in Medical Statistics or a related subject. Please also see the ESRC UBEL DTP’s general academic qualification requirements here.

Application process

Interested applicants are asked to express their interest by sending a CV and cover letter to Jonathan Bartlett ( by 12 noon UK time on 21st November 2022. Applicants will be contacted shortly afterwards and may be invited for an interview. One applicant will be selected to go forward to the second stage of the application process.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.