CAST Seminar: Richard Cook

Join us at the CANSSI Ontario STatistics Seminars (CAST) with

Richard Cook

Professor
Mathematics Faculty Research Chair
Department of Statistics and Actuarial Science
University of Waterloo

Free Event | Registration Required

Talk Title

Assessing the Effects of Interventions in  Life History Processes Using Observational Data

Abstract

Real-world observational studies and administrative data are viewed as a rich source of information on disease processes and associated intervention effects. The validity of findings from the analysis of such data depends critically on addressing sources of observation bias and confounding related to treatment by indication. Furthermore, disease and related features are dynamic processes where changes in disease states, markers, and treatment occur at random times that may not be fully observed. We consider challenges arising from the analysis of data on such processes, motivated in part by a cohort of persons with psoriatic arthritis whose clinic visits occur at random times associated with the  disease and treatment processes. Through the formulation of joint multistate models for disease, biomarker and treatment processes we address confounding by indication, while an expanded model addresses possible biases arising from a disease-related visit process. We use this framework to specify and estimate average treatment effects in a corresponding conceptualized experimental regime through trial emulation.  We illustrate our methodology on the psoriatic arthritis cohort and estimate an average effect of biologic versus conventional therapy on the progression of joint damage over time. This is joint work with Jerry Lawless and Lily Zou.

Speaker Profile

Richard Cook is University Professor in the Department of Statistics and Actuarial Science and Faculty of  Mathematics Research Chair at the University of Waterloo.  He holds a cross-appointment in the School of Public Health (UW) and is an affiliate scientist at the McMaster Centre for Transfusion Research. Richard’s research interests include the analysis of life history data, efficient study design for clinical and population health research, and statistical methods for the analysis of incomplete data.  He has published extensively in these areas and written two books with Jerald Lawless (The Statistical  Analysis of Recurrent Events, Springer, 2007; Multistate Models for the Analysis of Life History Data, Taylor and Francis, 2018).   He has designed and was founding director of three graduate training programs in biostatistics to build capacity for  pharmaceutical research, cardiovascular research and cancer research. His research has been funded from diverse sources including NSERC, CIHR, NIH, OICR and the Public Health Agency of Canada and he consults widely for government and industry.