The course will present the basic precepts and the principles underlying the primary methods of epidemiologic data analysis. The aim of the course is for the participant to arrive at a coherent conceptualization of the core principles of epidemiologic data analysis. This is not a statistics course; although examples of analytic calculations are given and there are lab exercises assigned for homework, there is no emphasis on proficiency in the execution and calculation of results nor how to build mathematical models.
The course begins with a discussion of the principles of epidemiologic data analysis, and then progresses to a discussion of precision and validity, placing a strong emphasis on a quantitative approach to analysis, using estimation, rather than a qualitative approach based on statistical significance testing. After covering the analysis of crude data, the focus shifts to the control of confounding using stratified analysis and multivariate models. Other topics that are covered include the analysis of matched data, the evaluation of interaction, the use of multivariate summary confounder scores (including propensity scores), marginal structural models, imputation of missing data, sensitivity analysis, and the estimation of trends in effect.
The class presentations will be supplemented with discussion of selected published papers and computer assignments using the Episheet spreadsheet to illustrate key analytic concepts.
Course objectives
- Students will become conversant with the principles of estimation of epidemiologic measures from basic epidemiologic data.
- Students will be able to explain and demonstrate the advantages of stratified analysis as a primary approach to epidemiologic data analysis, and to use a spreadsheet program to conduct basic epidemiologic analysis of stratified data, and to interpret the results.
- Students will be able to describe a strategy for using regression models in epidemiologic data analysis, either using an outcome model or a model that serves as a confounder summary score.
Recommended book: Epidemiology, An Introduction by KJ Rothman, Oxford University Press, ISBN-13:978-0195135541.
Books are for sale during the Erasmus Winter Programme.