Title: Advanced Epidemiological Methods
Keywords: Statistics (incl.. risk assessment)
Research (in general)
Quantitative methods
Country: Germany
Institution: Germany - Institute of Tropical Medicine and International Health, Berlin
Course coordinator: Dr. Matthias Borchert
Date start: 2018-04-09
Date end: 2018-04-27
About duration and dates: 3 weeks
Classification: advanced optional
Mode of delivery: Face to face
Course location: Charité – Universitätsmedizin Berlin
Campus Virchow-Klinikum
Augustenburger Platz 1 Berlin-Wedding
ECTS credit points: 4.5 ECTS credits
135 hours
90 contact hours (28.5h interactive lectures, 25.5h computer exercises, 6h exercises, 24h data analysis under supervision, 3h revision, 2h exam, 1h feedback)
45 self-study hours (independent)
Language: English
At the end of the module,students will be able to:
- Appraise the different alternative explanations to causality, and propose ways to address them.
- Evaluate the presence of effect modification, and propose ways to interpret and report it.
- Appraise the role of multivariable regression techniques to predict an outcome depending on several exposure variables.
- Appraise the role of multivariable regression techniques to assess interaction and control confounding.
- Appraise why data from cluster surveys require special analysis techniques, and demonstrate how to use them.
- Appraise why data from matched case control studies require special analysis techniques, and demonstrate how to use them.
- Propose an appropriate regression model for a set of epidemiological data.
- Propose an appropriate modelling strategy to select variables, identify interaction and linear trends.
- Judge the role of regression techniques in data analysis, and how they relate to table-based techniques.
- Appraise how results from regression analysis are presented and discussed in the scientific literature.
- Perform multivariable linear and (unconditional and conditional) logistic regression analyses using the statistical software package STATA, and interpret their results.
Assessment Procedures:
Three hour written closed-book examination. The overall pass mark is 60%.

Students who fail will be offered a re-sit examination, which should take place by the beginning of the following semester. Students who fail with a grade between “4.1” and “5” are offered to have a data analysis report evaluated instead of re-sitting the written closed-book examination. If the report achieves a “Pass” mark, the final grade will be a “4.0 – Sufficient”.
A second re-sit is allowed but may be linked to conditions set by the Committee of Admissions and Degrees.
- Review: Measures of disease frequency and strength of association; inference; study designs; causality and its alternatives: random error, bias, confounding (inflation and masking), reverse causality; interaction (synergistic and antagonistic); data management with STATA; stratified analysis with STATA.
- Analysis of cluster survey data
- Simple and multivariable linear regression
- Matching in case-control studies, analysis of matched data
- Unconditional and conditional logistic regression
- Model selection and variable selection
- Role of regression techniques in data analysis
- Role of regression techniques in scientific publications
- Outlook on further regression methods (regression models for count data, regression models for survival time data)
NB: The focus of the module is on linear, and even more on logistic regression.
The first and the second week are divided into 45% interactive lectures, where subject areas are introduced, and 55% exercises with our without computers, where students can practice their freshly gained knowledge and skills.
The third week devotes 80% to practicing data analysis, using datasets provided by ITMIH. A supervisor meets the students once a day to discuss obstacles and progress, and is available – within reason – for further support upon request.
The remaining 20% of week three are devoted to revision, examination, and feedback.
Note that week three aims at consolidating knowledge and skills, so that no new content is introduced in week three.
Participants are required to have basic knowledge in epidemiology and biostatistics (including confounding, interaction, and stratified analysis), and should be interested in theory and practice of epidemiology.

If not a native speaker: Internationally recognised English proficiency certificate equivalent to a TOEFL score of 550 paper/213 internet/80 online, or IELTS score 6, or DAAD (A or B in all categories). tropEd students need to provide proof of registration as tropEd student at their home institution only. For further exceptions refer to http://internationalhealth.charite.de/en/admission/application/
Max 24 students,
Students and participants must attend 80% of the teaching time.
Places are allocated on a “first-come, first-served” basis.
Deadline for application: 6 weeks before module start (25 Feb 2018).
We shall confirm the module 4 weeks before module start latest subject to a sufficient number of applications. Late applications will be considered as long as places are available.
1,125.00 EUR for tropEd MScIH students and alumni
1.687,50 EUR for others
Possibly 50 to 200 EUR for STATA licence or laptop rent
Major changes since initial accreditation:
Given the complexity of the issues addressed in this module we continue to devote a substantial amount of time (2.5 days) to the revision of content covered for a first time in ITMIH’s core course. This revision has been found to be helpful, even necessary for many students.
The content of the module has been revised with the view to emphasise content that many students tend to need for epidemiological MScIH thesis projects, and to deemphasize other advanced content. Therefore the analysis techniques for cluster survey data and data from matched case control studies have been added, while standardisation has been eliminated. Linear and logistic regression continue to be taught, the latter quite extensively given its importance in epidemiological research, while Poisson regression has been eliminated, as few students will analyse data from open cohort studies as MScIH thesis project.
Overall, the emphasis has moved away from mathematical background and formulae, in favour of understanding when to use which technique, how to use it, and how to interpret its result. This shift means to take into account the background and interests of the majority of students.
In response to student feedback about the high density of the two-week module and the lack of time to practice new skills, we have decided to add a third week, during which students analyse data sets under supervision. No new content is planned to be offered in the third week, which is devoted to consolidate what has been learned in week one and two.
Dr Matthias Borchert (ITMIH) took over as course coordinator in 2012. Since then, co-lecturers have come from the Antwerp Institute of Tropical Medicine.
Student evaluation:
Overall student evaluation has been very good to excellent. Students were especially pleased with the combination of theory and extensive practical exercises. Students demanded more time to practice data analysis under supervision, and a revision session at the end of the module.
Lessons learned:
The course has in most years experienced high demand and interest. Most students have limited interest in statistical theory but are keen to acquire knowledge and skills they can use to analyse epidemiological data and to critically read scientific literature where results from multivariable analyses are presented. In the context of ITMIH’s MScIH this advanced module should prepare students to conduct an epidemiological thesis project with competence and confidence.
tropEd accreditation:
Accredited in Marseille in 2004, re-accredited in Mexico in May 2010 and in Uppsala in September 2012. Accreditation is valid until September 2017.
Students are expected to bring a calculator without internet connectivity (i.e. not a smartphone etc.) to the teaching sessions and the written exam. We recommend the following calculator: Casio fx-115MS or equivalent.

Recommended textbooks:
For this advanced module: Campbell MJ, 2006. Statistics at Square Two. 2nd edition. London: BMJ Books. ISBN-13: 978-1405134903

For reviewing relevant core course content: Campbell MJ & Swinscow TDV, 2009. Statistics at Square One. 11th edition. London: BMJ Books. ISBN-13: 978-1405191005

The module is run in collaboration with the Antwerp Institute of Tropical Medicine.
Course application form
Email Address: mscih-student@charite.de
Date Of Record Creation: 2011-11-16 04:14:39 (W3C-DTF)
Date Of Record Release: 2011-11-16 05:20:47 (W3C-DTF)
Date Record Checked: 2017-11-19 (W3C-DTF)
Date Last Modified: 2017-11-19 15:58:10 (W3C-DTF)

Report a Problem with this Resource Record

Log In: