Methodological research

The main focus of my research has been in the design, analysis and reporting of cluster randomized trials. Cluster randomized trials are trials in which intact groups such as medical practices, hospitals, schools, or communities — rather than individuals themselves — are randomly allocated to different interventions. This design is increasingly being used in health research, for example, to evaluate different models of health services delivery, health promotion campaigns, or interventions designed to increase the uptake of evidence-based methods in medical practice. Cluster randomization requires special consideration in the design and analysis of the trial due to the usual presence of positive intracluster correlation.

Since 2007 and in collaboration with Dr. Jeremy Grimshaw at the OHRI, and Dr. Charles Weijer at the Rotman Institute of Philosophy, my work has also included the analysis of ethical and policy challenges posed by cluster randomized trials. Cluster randomized trials pose unique ethical challenges that complicate the interpretation of standard research ethics guidelines. For example, there may be differing units of allocation (e.g., hospitals), intervention (e.g., health professionals), and outcome measurement (e.g., patients) within a single study. This complicates the identification of research participants and informed consent procedures within the trial. Funded by the Canadian Institutes of Health Research, we embarked on a five-year mixed-methods research study, that included a series of empirical studies, an in-depth ethical analysis, and a consensus process. The main product of this project: The Ottawa Statement on the Ethical Design and Conduct of Cluster Randomized Trials, was published in Plos Medicine and is the first international guidelines targeted specifically at this increasingly important research design.

My other research interests include applications in longitudinal data analysis, multilevel models, missing data methods, quasi-experimental designs, and derivation of clinical prediction rules.

Statistical consulting work

As a biostatistician, I regularly collaborate with clinicians and scientists in their research studies. Through the Ottawa Methods Center, I provide biostatistical support to researchers at the OHRI, The Ottawa Hospital, the University of Ottawa, and elsewhere, including in the design, statistical analysis, sample size determination, and publication of studies. I am usually involved from the initial protocol development and grant application stages, through to final publication and dissemination of results. Potential student projects often become available through this collaborative work.