Course content

Study design, measures of disease occurrence, measures of effect, intern validity, Directed acyclic graphs (DAGs), interpretation of multivariable models, causal inference, causal interaction and effect measure modification, screening, instrumental variable estimation, family designs, case only designs.

Learning outcome

After completing SMED8002:

Knowledge

• The candidate is at the forefront of knowledge regarding various study designs most commonly used in population-based and clinical research (cross-sectional studies, cohort studies, case-control studies, and randomized controlled trials).
• The candidate can assess the appropriateness of different measures of disease occurrence (prevalence, incidence rate, and incidence proportion) and screening.

Skills

• The candidate can handle and interpret various complex measures of the relationship between exposure/treatment and disease based on absolute and relative risk (risk difference, incidence rate difference, numbers needed to treat versus risk ratio, incidence rate ratio, odds ratio, hazard ratio).
• The candidate can distinguish between and understand the concepts of ‘causal relationship’ (causality) and ‘statistical association’.
• The candidate can differentiate between causal interaction and effect measure modification.

General proficiency

• The candidate has in-depth knowledge of common challenges in observational studies (selection bias, information bias, and confounding).
• The candidate has in-depth knowledge of interpreting Directed Acyclic Graphs (DAGs).
• The candidate has in-depth knowledge of assessing precision in epidemiological studies with an emphasis on the interpretation and use of p-values and confidence intervals.
• The candidate has a thorough understanding of the principles of multivariable analyses and the ability to interpret such analyses.
• The candidate has in-depth knowledge of the principles of instrumental variable analysis, various family designs, and case-only studies.

Learning methods and activities

Teaching modalities: Lectures and problem solving. Participation is mandatory. The course coordinator may approve up to 20% absence from mandatory lectures.

Compulsory assignments

• Practice tasks
• Mandatory attendance at lectures

Recommended previous knowledge

KLMED8004 Medical statistics 1, KLMED8009 Clinical Research, introductory course in epidemiology (e.g. master course in epidemiology)

Required previous knowledge

Master's degree. Medical students at The Student Research Programme.

Course materials

Books:

Hernan M, Robins J: Part I, Part II, chapter 16, https://www.hsph.harvard.edu/miguel-hernan/causal-inference-book/

Rothman K, Greenland S and Lash (2021): Chapter 2, 3, 5-10 in Modern epidemiology

Papers:

Hernán MA, Hernández-Díaz S, Werler MM, Mitchell AA. Causal knowledge as a prerequisite for confounding evaluation: an application to birth defects epidemiology. Am J Epidemiol 2002;155:176-184

Hernán MA, Hernández-Díaz S, Robins JM (2004): A structural approach to selection bias. Epidemiology. 2004 Sep;15(5):615-25.

JanszkyI, Ahlbom A, Svensson AC (2010): The Janus face of statistical adjustment: confounders versus colliders. Eur J Epidemiol. 2010 Jun;25(6):361-3. Epub 2010 May 7.

Glymour M, Weuve J, Berkman L, Kawachi I, Robins J. When is baseline adjustement usuful in analysis of change? An example with education and cognitive change. Am J Epid 2005; 162:267-78

Krieger N, Davey Smith G. The tale wagged by the DAG: broadening the scope of causal inference and explanation for epidemiology. Int J Epidemiol 2016; 45: 1787-808.