¹û¶³Ó°Ôº

The course will cover a range of advanced statistical techniques and methodological approaches used in records research. In particular:

• More advanced regression techniques, including logistic regression and Cox modelling,
• Methods for handling missing data with a focus on multiple imputation.
• Introduction to causal inference: This part will cover:

• Association vs causation
• Potential outcome framework
• Causal diagrams
• G-methods
• Assumptions that underlie all causal inferences

By the end of the module, you should be able to:

• critically assess and use a variety of statistical methods and approaches relevant to electronic health records research;
• select appropriate methods to draw causal inferences from data sets;
• demonstrate an understanding of the principles and assumptions of the various statistical approaches and be able to critically evaluate these;
• design an appropriate analysis plan to address a specific research question;
• demonstrate knowledge and understanding to support the choice of appropriate statistical methods to answer a range of questions.

You will learn though a combination of lectures and computer-based practicals using Stata and R.

Kirkwood & Sterne (2010) Essential Medical Statistics. Malden, Mass. : Blackwell Science

Kenneth J. Rothman, Sander Greenland, Timothy L. Lash (2008) Modern Epidemiology. Philadelphia: Wolters Kluwer Health/Lippincott Williams & Wilkins

Kenward, M. G., & Carpenter, J. (2013). Multiple Imputation and its Application. Chichester: John Wiley & Sons

Kenward, M. G., & Carpenter, J. (2007). Multiple Imputation: current perspectives. Statistical Methods in Medical Research, 16(3), 199-218

Hernán MA, Robins JM (2019). Causal Inference. Boca Raton: Chapman & Hall/CRC, forthcoming