This four-week course is about research methods for biomedical (pre-clinical and clinical) studies. The course aims at providing basic knowledge of frequently used designs and methods for biomedical studies. It is important that the student works with these designs and methods, and is able to discuss strengths and weaknesses. Three themes will be central in this course:
Choices between different study designs in animals or in humans, such as case-control, follow-up, randomized controlled trials will be discussed, as well as aspects specific to each of the study designs, for instance choice of controls, cross-over or parallel, randomisation, blinding.
Statistical methods to analyse data from these studies will be presented. The methods discussed include univariate and multivariate linear regression, logistic regression, survival analysis (Kaplan Meier survival curves and Cox regression), and linear mixed models.
Interpretation and validation.
Interpretation of the results involves the interpretation of relative risks, odds ratios and hazard ratios in univariate and multivariate analyses. Validation of the methods involves model checks, the reliability of measurements (inter- and intraobserver variation, kappa), the evaluation of diagnostic systems (ROC curves), and recognition of the principles of bias and confounding.
Other aspects discussed are literature search, database use and ethical aspects.
The student will be able to:
- perform database searches for literature (PubMed and Web of Science) and biological items and concepts (such as OMIM, LOVD, Anni) and account for the method of literature searching by following reproducible methodology
- perform a comprehensive literature search for the design of a biomedical study by following reproducible methodology
- translate a clinical problem into a research question
- explain methods of study design used in a clinical or population setting, such as a clinical trial, an observational follow-up study, and a case-control study, and knows how these studies are performed
- design a biomedical study in an efficient way
- choose, apply and interpret statistical analyses, such as linear or logistic regression analysis with and without repeated measurements and survival analysis (Kaplan Meier survival curves) in a clinical trial and an observational follow-up study
- evaluate the reliability and reproducibility of measurement systems
- recognize and discuss the possibility of bias and confounding in a study
- discuss ethical and legal aspects of clinical studies with human subjects
Mode of instruction
This course consists of a number of lectures, 5 working groups, 9 self-study assignments, and 7 computer practicals. Depending on the exercise, the students will work together with one student, or in small groups.
Students will be evaluated by preparing a protocol and presentation for an animal experiment (pass), a theme assignment of information literacy (10% of final grade), a written report on a clinical trial (pass), a written report of each of the computer practicals (pass), and an examination (tentamen, 90% of final grade).
The nature of the examination is a combination of multiple choice and “open questions”. The final grade will be published in uSis as soon as all assignments have been turned in. There is a re-examination during spring.
- J.P. Vandenbroucke, A. Hofman. Grondslagen der Epidemiologie.
- A. Petrie, C. Sabin. Medical Statistics at a Glance.