Adjunct Professor of Epidemiology
Adjunct Professor, School of Health Information Science, University of Victoria, BC, Canada
Michael Smith Foundation for Health Research, Distinguished Scholar, 2002-2006
Kenneth Rothman Epidemiology Prize, 2002
BA Biochemistry, 1975, Oxford University
SM Epidemiology, 1979, Harvard School of Public Health
SD Epidemiology, 1984, Harvard School of Public Health
I am a health services epidemiologist specializing in methodology. My career has moved gradually from ‘micro’ to ‘macro’ sciences.
After studying biochemistry at Oxford, I spent two years in Ottawa with the Canadian federal government’s health and welfare department. As a Project Officer with the National Health Research and Development Programme, I first learned about public health research. Initially interested in mental health research, I was attracted to Harvard School of Public Health for a master degree in epidemiology. I soon learned how difficult psychiatric illnesses are to study using epidemiologic methods. For my doctoral training, I shifted to cancer epidemiology and conducted a case-control study of renal adenocarcinoma. I joined the faculty to conduct biochemical epidemiology of carcinogen exposures but delays in processing blood samples caused me to seek other paths.
I developed an interest in improving epidemiologic methods by applying Karl Popper’s philosophy of science. That led to papers on misuse and misinterpretation of certain statistical methods, including the Kappa statistic and trend tests, as well as papers on the causation and measurement of bias in epidemiologic studies.
In the mid 1980s, concerns about the risk of nuclear war were palpable. A frequent topic of lectures, workshops and activism was the lack of any possible medical response to a small nuclear war. From 1986 through 1988, I volunteered part-time for Dr. Bernard Lown, co-founder of the International Physicians for Prevention of Nuclear War, winner of the 1986 Nobel Peace Prize. The result of my ‘peace research’ was Dr. Lown’s establishment of SatelLife, a non-profit that promoted low-cost e-mail communication to Africa by low-orbit satellite at a time when telecommunication to Africa was prohibitably expensive.
In the late 1980s, I began working with James Muller on identifying triggers of myocardial infarction. This led to my discovery (or invention) of the case-crossover study design.
Having enjoyed the opportunity to discover or invent new methods, I was attracted to health services epidemiology in 1990. Health policy and management had grown to prominence, health administrative databases were proliferating and computer power was accelerating. In 1991-2, I took leave from Harvard to join the research branch of the British Columbia Ministry of Health. That turned into a permanent position in Victoria BC, and I continued to visit HSPH to teach episodically. I gravitated to pharmaceutical policy impact evaluations using administrative databases. Methods I have developed include camouflaged contacting (a way of using administrative databases to contact patients while preserving data privacy) and developing low-cost randomized policy trials (the use of designed delays in implementation to enable rigorous evaluation with central administrative databases.)
In the 2000s, when knowledge translation research began to be vigorously promoted, I discovered there was a scientific name for my work with pharmaceutical policymakers on integrating research and policy. I joined a Cochrane Collaboration systematic review team focusing on drug policy impact studies. My collaboration with Sebastian Schneeweiss, which began in 1995, led to validations of, and innovations in, pharmaceutical policy in BC. One of these is an ongoing randomized policy trial of educational feedback to community physicians by means of prescribing portraits, called Education for Quality Improvement of Patient care (EQIP). Another is a strategy of ‘coverage with evidence development’ applied to medications for Alzheimer’s Disease.