Estimating cancer treatment effects in patient audit data

   School of Medicine, Dentistry & Biomedical Sciences

  Dr F Bannon  Applications accepted all year round  Self-Funded PhD Students Only

About the Project

Cancer treatment effects on patient survival can be estimated in population-based observational cancer audit datasets. For certain patient groups, these estimates could be their only treatment evidence-base. The PhD will explore the data conditions and the causal inference methods required for unbiased estimation. 

Cancer patient survival in Northern Ireland lags behind other central and northern European countries. To improve patient outcomes, cancer services in NI are audited regularly by the Northern Ireland Cancer Registry providing service metrics, including the provision of timely and optimal treatment (chemotherapy, radiotherapy, surgery) . The audits captures a wealth of clinical information that, when coupled with causal inference (CI) statistical methods, have the potential to estimate unbiased treatment effects in the population, or in sub-groups.  

The aim of the PhD study is to increase the treatment evidence-base for patients not included in conventional randomised control trials (RCTs). Such patients may have a rare cancer, or due to inclusion and exclusion criteria, are under-represented in RCTs, e.g. elderly or patients with dementia.  

The objectives are to assess, in the audit data, if 1) treatments are well-defined, 2) there is sufficient information to adjust for confounding (exchangeability), 3) sufficient variety in clinical practice around treatment use (positivity). The study will also explore how best to apply CI methods. 

In situations where the conditions are met, the study will estimate treatment effects, thus contributing to the evidence base. Otherwise, the study will recommend if and how audits can be improved. 

Start Date: October 2022

Email Now

Search Suggestions
Search suggestions

Based on your current searches we recommend the following search filters.

PhD saved successfully
View saved PhDs