Propensity score: help or hype?
Propensity scoring provides a useful way to analyze observational study data and are a necessary component of an epidemiologist’s methodological toolbox. This presentation will discuss how the propensity score can be derived and review their many applications, such as matching, adjustment, and weighting. A clinical example will be used to discuss how propensity score analyses compare to other methods to control for confounding and scenarios in which propensity scores are particularly suitable (and others in which they are not).

Tobias Kurth is a professor of public health and epidemiology and the director of the Institute of Public Health at the Charité – Universitätsmedizin Berlin, Germany. He is also the co-director of the Centre Virchow-Villermé for Public Health Paris – Berlin. He received his MD degree at the University of Tübingen, Germany, training in neurology at the University Hospital of Essen, Germany, and a doctorate in epidemiology from the Harvard T H Chan School of Public Health. He previously worked as a research director of the French National Institute of Health and Medical Research (Inserm) in Paris and Bordeaux. His areas of research interests and expertise are in neuroepidemiology, cardiovascular epidemiology, pharmacoepidemiology, and research methodology. He joined The BMJ in 2009 as editorial adviser and has served as consulting clinical epidemiology editor since 2010
Date: 28 March 2017, 10:00 (Tuesday, 11th week, Hilary 2017)
Venue: St Luke's Chapel, Oxford
Speaker: Professor Tobias Kurth (Professor of Public Health and Epidemiology, Universitätsmedizin Berlin)
Organising department: Nuffield Department of Primary Care Health Sciences
Organiser: Catia Nicodemo (University of Oxford, Nuffield Department of Primary Care Health Sciences. Research Fellow, Centre for Health Service Economics & Organisation)
Part of: Nuffield Department of Primary Care Health Sciences Seminars
Booking required?: Not required
Audience: Members of the University only
Editor: Dan Richards-Doran