Epi Special Seminar
UW Epidemiology 2010 Distinguished Visiting Epidemiologist
|Malcolm Maclure, ScD
British Columbia Chair in Patient Safety and Professor, Department of Anesthesiology, Pharmacology
Tue, Feb 2nd, 3:30-4:50 pm, Health Sciences room T-733.
Refreshments at 3:15 pm.
A review of ideas about confounding, information bias, and selection bias that underscores the need for routinely analyzing the sensitivity of study findings to multiple hypothesized biases.
Causation of Bias:The Episcope, by Dr. Maclure and Dr. Sebastian Schneeweiss.
Journal Club: ‘Why me?’ vs. ‘Why Now?’
Wed, Feb 3rd, 12:00-12:50 pm, Health Sciences room F-348
Lunch will be provided – please RSVP to email@example.com by Mon, Feb 1st.
- ‘Why Me?’ vs. ‘Why Now?’ – Differences Between Operational Hypotheses in Case-Control vs. Case-Crossover Studies, by Dr. Maclure
- Benzodiazepines and elderly drivers: a comparison of pharmaco-epidemiological study designs, coauthored by Dr. Chris Delaney of UW
Discussion of the articles will be led jointly by Drs. Maclure and Delaney.
Individual Mtgs: Wed, Feb 3rd, various times are available, Health Sciences Building.
Please contact firstname.lastname@example.org to reserve a time to talk with Dr. Maclure one on one or with a small group. Students, faculty, and all others are welcome.
Panel discussion: Inventing New Methods in Epidemiology and Health Services
Co-sponsored by the UW Department of Epidemiology and the Grand Rounds-Moments in Leadership series.
Wed, Feb 3rd, 4:30-6:00 pm, Health Sciences room T-739
Reception at 4:00 pm.
Dr. Maclure, ScD
Dr. Thomas D Koepsell, MD, MPH, Professor, Departments of Epidemiology and Health Services, UW
Dr. David Grembowski, PhD, Professor, Department of Health Services, UW
Dr. Scott S Emerson, MD, PhD, Professor, Department of Biostatistics, UW
Motivation for the panel discussion:
Epidemiologists learn classic methodological approaches for good reasons: A study is more understandable and credible if it uses a standard, established design and proven methods of statistical analysis. On the other hand, epidemiologists may also develop new methodologies for good reasons: Breakthroughs in our understanding of the scientific method or in technology can yield methods that produce more valid results.
According to UW epidemiologist, Dr. Tom Koepsell, when faced with a research situation where our tried and true methods are inadequate, epidemiologists earnestly, “retreat through a hierarchy of stronger to weaker (but usually well-known) alternatives until we find an option that fits the constraints. Then we have to decide whether that option is good enough: how likely and serious the biases would be, what we can do to probe for artifact and try to get around it, and what the consequences of error would be."
Dr. Maclure writes, “Some of the worst and most influential epidemiology is put to use in the domain of health services research. In circumstances where epidemiological methods are used to address health services questions, we are often forced to innovate for practical reasons because databases are large and almost free, but were designed for administration, not research. Yet the risks of innovating are often less than the risks of knowing almost nothing about expensive and controversial problems. To bend or not to bend – that is the question.”
Updated on January 25, 2010