23 Feb 2023 09:00am

Estimation and validation of counterfactual risk prediction models

Event Location
Ruth Keogh
London School of Hygiene & Tropical Medicine

Clinical risk prediction models enable predictions of a person’s risk of an outcome (e.g. mortality) given their observed characteristics. It is often of interest to use risk predictions to inform whether a person should initiate a particular treatment. However, when standard clinical prediction models are developed in a population in which patients follow a mix of treatment strategies, they are unsuitable for informing treatment decisions. Counterfactual risk predictions (CRPs) aim to address this problem. CRPs are estimates of what a person’s risk would be if they were to follow a particular treatment strategy, given their individual characteristics that are also predictive of the outcome. CRPs do not depend on the observed mix of treatments and are therefore suitable for informing treatment decisions.

In this talk I will discuss causal inference methods for estimating CRP models using longitudinal observational data on treatment use, patient characteristics and a time-to-event outcome. An essential step in development and reporting of prediction models is to validate their performance. However, methods for validating standard clinical risk prediction models do not apply to CRP models. I will describe some new methods for assessing the predictive performance of CRPs.

In a motivating example, we are interested in CRPs for mortality in patients awaiting a liver transplant under the strategies of receiving or not receiving a transplant. I will illustrate the methods for estimation and validation of CRPs using data from the US Scientific Registry of Transplant Patients.


Ruth Keogh is a Professor of Biostatistics and Epidemiology in the Medical Statistics Department at London School of Hygiene & Tropical Medicine. Her research focuses on statistical methods for the analysis of observational data, and she works especially in causal inference methodology and methods for the analysis of time-to-event data. Other areas of interest include methods for handling measurement error and missing data, prediction modelling, and design and analysis of case-control studies. Ruth is involved in a number of areas of application in health research, with a particular focus on cystic fibrosis, and also including organ transplantation, Covid-19 and cancer. Her research is funded by a UK Research and Innovation Future Leaders Fellowship.


Thursday 23rd February, 9-10am AEDT

This event will be streamed via Zoom and is being recorded.
Please click here to join

Or, go to https://monash.zoom.us/join and enter

Meeting ID: 87519891685
Passcode: 939195