Lectures

The lectures in 2012 will be available soon.

Beyond the Case-Control Study
Monday August 29, 2011, 16.00 – 17.00 hrs

Professor Norman Breslow
Department of Biostatistics
University of Washington, USA

In a “nested” case-control or case-cohort study, cases and controls are sampled from a defined finite population such as a health maintenance organization. The case-control design is used to collect additional data, for example from biological assay of stored tissue, for an informative sample. Considerable existing information about the members of this finite population not sampled as cases or controls, however, is often ignored both in the design of the study and in the analysis of the resulting data. By viewing the problem as a survey statistician would, this talk explores ways of recovering information often neglected by epidemiologists. Statistical tools for survey statisticians now implemented in the freely available R statistical system facilitate this process.
The iron laws of mortality
Tuesday August 30, 2011, 16.00 – 17.00 hrs

Professor Johan Mackenbach
Erasmus MC, Rotterdam, The Netherlands

The idea of a “law of mortality” which governs the increase with age of the risk of dying was first proposed by Benjamin Gompertz (1779-1865), who observed that among humans the mortality rate doubles with every 8 years lived. Since then, mortality rates in western populations have declined spectacularly, but the exponential increase of mortality with age has remained, at least until the age of 80 years. I will review recent progress in the human struggle with mortality, and discuss evidence that beyond the age of 80 the iron laws of mortality no longer hold, suggesting that achieving immortality is less unlikely than it seems.
Negative Controls: A Tool to Detect Confounding and other Sources of Bias in Epidemiologic Studies
Wednesday August 31, 2011, 16.00 – 17.00 hrs

Marc Lipsitch
Professor of Epidemiology
Director, Center for Communicable Disease Dynamics
Harvard School of Public Health, Boston, USA

Noncausal associations between exposures and outcomes threaten the validity of causal inference in observational studies. Many techniques have been developed for study design and analysis to identify and eliminate such errors. Many scientists imagine that such problems are minor or nonexistent in experimental studies, where careful standardization of conditions (for laboratory work) and randomization (for population studies) should eliminate most such noncausal associations. More information
Confounding and effect modification one more time
Thursday September 1, 2011, 16.00 – 17.00 hrs

Anders Ahlbom
Professor of Environmental Medicine
Karolinska Institutet, Stockholm, Sweden

Confounding and effect modification are basic concepts to epidemiologists and are on one level easy to grasp. Yet, they have created confusion and occasionally heated debate for decades. From some perspectives they have rather little in common but on the other hand they are both addressed by multivariable modeling and thus appear in tandem in many analyses. On top of this is the sometimes also heated discussion of biological interaction versus statistical interaction. The purpose of this lecture is to provide a historical perspective to the seemingly never ending discussion about these fundamental concepts and also to provide as clear as possible an explanation of what confounding and effect modification and some related concepts actually are. While the talk will be mainly on concepts and ideas there will be some discussion about methodological approaches as well.