Course organizers
Jonas Björk and Helena Jernström

Target group
This course is intended for doctoral students at the Medical Faculty at Lund University, who conduct clinical or health sciences research based on individual level data. The course is also open for other applicants if there are available slots. Course requirements are Applied statistics I. 

Time
Weeks 15-16 2018 (April 9--April 20, 2018)

The course is held in English.

Credits
3 hp

Aim
The overall aim of the course is to contribute to increased quality in clinical and health sciences research at Lund University by introduction of epidemiological methods in all stages of such research. The course will provide both theoretical and practical knowledge regarding planning, execution and evaluation of common study designs in epidemiological studies.

Learning outcomes 

After the course the student should be able to:
Design, apply and critically evaluate commonly occurring epidemiological study designs in both clinical medicine and health sciences
 
Describe common data sources and collection methods for exposure and outcome data (clinical data, registry data, biobanks, and surveys/questionnaires), including strengths and weaknesses
 
Conduct basic statistical analysis of epidemiological data
 
Explain common systematic errors (selection bias, confounding, and information bias) in epidemiological studies, and evaluate the extent of these with simpler sensitivity analyzes

Contents

There are four themes:

1) Introduction to Epidemiology

• The foundation of Epidemiology. Applications, internationally, nationally and regionally.
• Data sources and methods to collect exposure and outcome data
• Principles of causal inference
• Introduction to Direct Acyclic Graph (DAG)

 2) Cohort studies

• Design aspects. Fixed cohorts and dynamic populations.
• Incidence calculations. Absolute and relative risks
• Kaplan-Meier analysis and Cox regression for basic data analyses
• Systematic errors (common reasons for wrong conclusion)

 3) Case control studies

• Design aspects. Principles for selection of cases and controls. Matching.
• Odds ratios. Logistic regression for basic data analyses.
• Systematic errors

 4) Introduction to more advanced methodology

• Effect modification
• Mediation, i.e. separation of direct and indirect effects on disease outcome
• Bias quantification, i.e. evaluation of the magnitude of systematic errors
• Scoring systems (propensity & disease risk scores)

Course literature 
The course book is ”Rothman KJ. Epidemiology - an introduction. Oxford University Press 2nd Edition 2012”.It can take two to three weeks to get the book so please order on time!We will send a list of the relevant chapters in the book and articles that will be used later on.