I am a medical statistician and a research fellow in the Department of Medical Statistics at the LSHTM, since April 2023. In addition to my research, I also act as an early career researcher reprentative (ECR rep), admissions tutor for the Medical Statistics MSc, and member of the Research Staff Forum (for the Concordat monitoring group), and additionally co-organise the IRM monthly meetings, at which members of the MedStats department present recent and ongoing research. Prior to my appointment at LSHTM, I held a short-term fellowship at the University of Bristol. There, at the MRC Integrative Epidemiology Unit, is also where I completed my PhD (2022), on the subject of identifying and adjusting for bias resulting from missing outcomes in clinical trials. I completed my 2-year research MSc in Statistics for the Life and Behavioural Sciences in 2018, at the University of Leiden (The Netherlands), with a focus on Cox-model extensions and machine learning alternatives for survival outcomes, with applications in osteosarcoma.
My current research involves the development and application of methods for recurrent events in a survival framework, with a particular focus on heart failure hospitalisations and cardiovascular disease. Within this, key themes are the use of recurrent-event versus time-to-first event analyses, the impact of the violation of estimator assumptions, and the win ratio method for composite outcomes. Previous areas of research include bias resulting from missing-not-at-random dropout in clinical trials, multi-state and dynamic prediction modelling for time-to-event outcomes in osteosarcoma, and machine-learning methods for survival outcomes (specifically, neural networks and random survival forests). More generally, I am interested in the various biases that may arise from protocol and design deviations in clinical trials and in developing methods with estimator assumptions that are plausible in a clinical setting.
Affiliations
Teaching
I am a co-module organsier for the Inference submodule of the Foundations module of the Medical Statistics MSc. I additionally teach on several additional modules of the the same MSc (Clinical Trials; Statistical Models for Discrete Outcomes), and aditionally act as a personal student tutor and MSc project supervisor. Recently I completed my Postgraduate Certificate in Learning and Teaching (PGCILT) at the associate fellowship level (AFHEA).
Research
My current research involves the development and application of methods for recurrent events in a survival framework, with a particular focus on heart failure hospitalisations and cardiovascular disease. Within this, key themes are the use of recurrent-event versus time-to-first event analyses, the impact of the violation of estimator assumptions, and the win ratio method for composite outcomes. Previous areas of research include bias resulting from missing-not-at-random dropout in clinical trials, multi-state and dynamic prediction modelling for time-to-event outcomes in osteosarcoma, and machine-learning methods for survival outcomes (specifically, neural networks and random survival forests). More generally, I am interested in the various biases that may arise from protocol and design deviations in clinical trials and in developing methods with estimator assumptions that are plausible in a clinical setting.
Selected Publications
Implications, and Analytical Improvements