Faculty Member
Anna Heath
- Email Address(es)
- anna.heath(at)sickkids.ca
- Division(s)/Institute(s)
- Biostatistics Division
- Position
- Associate Professor
- SGS Status
- Full Member
- Appointment Status
- Status Only
- Currently Accepting Doctoral Students?
- Yes
Research Interests
Value of Information, Clinical Trial Design, Bayesian Analysis, Decision Modelling
Research Interests
I lead the IMPaCT (Innovative Methods in Paediatric Clinical Trials) lab at the Hospital for Sick Children. Overall, the IMPaCT lab aims to improve the conception, design and conduct of clinical research through the development of novel statistical methodology, software, expertise and guidance. We have a specific focus on paediatric clinical trials but work across a range of disease areas including critical care and oncology. Primarily, we focus on two interconnected methodologies:
1. Value of Information
This method was first defined in 1959 and is built on principles from statistical decision theory. More recently, researchers have combined these ideas with health economic modelling to allow us to maximize the impact of clinical research by prioritizing research that supports clinical decision making. The IMPaCT lab works on theoretical and applied projects to improve the use of Value of Information implement it to improve research designs in paediatrics and ensure efficient access to novel healthcare interventions. Value of Information methods require a broad range of research interests including Bayesian decision analysis, health economic modelling, Bayesian expert elicitation, utility theory, meta analysis, Bayesian computation methods and clinical trial design.
2. Bayesian Statistics
This approach to statistical inference has been gaining popularity in clinical trials over the last few decades, helped by increases in available computational power. They are a particularly powerful tool for innovative trial designs and have been used for adaptive and platform trials, novel trial designs that aim to increase the efficiency of trials. Bayesian methods can also be useful in rare diseases, common in paediatrics, as expert judgements can formally be incorporated into the analysis where data are limited.
Education & Training History
PhD, Statistical Science, University College London
MMath, Mathematics with French Language, University of Sheffield
Other Affiliations
Scientist, The Hospital for Sick Children
Honorary Research Fellow, University College London