Validation of prediction models in epidemiology and medicine

Welcome to Validation of prediction models in epidemiology and medicine 

Program: Epidemiology and Biostatistics (EB)

Description: 

Clinical prediction models constitute a valuable resource in a complex health care organism with resource scarcity and can be applied in various context including capacity planning, decision support, diagnostics, screening, and exploratory analyses. Depending on the context these models are required to pass through several pre- and post-clinical implementation phases including pre-implementation validation, monitoring, and de-implementation.

The course covers the following topics:

1. The basic differences between explanatory and predictive studies.
2. The phases and contexts of prediction models for clinical and population medicine.
3. Variable selection, shrinkage and SHapley Additive exPlanations (SHAP).
4. N-fold cross validation, bootstrap, geographical and temporal validation.
5. Net-benefit, calibration, receiver operating characteristic curve, and Bland-Altman plots.
6. Decision curve analysis and model assessment with a clinical justified threshold probability.
7. Sample size estimation for development as well as for validation.
8. Examples of specific use cases from the scientific literature and live working applications.

Upon completion of the course, the student will be able to distinguish between predictive and explanatory data analyses, understand the different phases and contexts of model development, implementation, and validation, as well as understand the basic statistical tools used in predictive studies.

The course form is a mixture of lectures, assignments, and a mini project. Some home assignments are to be anticipated. Students will be evaluated based on the mini project and an oral presentation in groups of 2-4 participants.

Literature/Requirements:  

Basic statistics and basic programming abilities will be assumed. The course is taught in R, however, some support and example code will be provided for Python and Stata upon request. All participants must bring a laptop with either R, Stata or Python installed.

Suggested reading ordered by priority:

Wynants L, Collins GS, Van Calster B. Key steps and common pitfalls in developing and validating risk models. BJOG. 2017 Feb;124(3):423-432. doi: 10.1111/1471-0528.14170. Epub 2016 Jun 30. PMID: 27362778.

Ewout W. Steyerberg, Yvonne Vergouwe; Towards better clinical prediction models: seven steps for development and an ABCD for validation, European Heart Journal, Volume 35, Issue 29, 1 August 2014, Pages 1925–1931, https://doi.org/10.1093/eurheartj/ehu207

Organizer: Martin Bøgsted

Lecturers: Heidi Søgaard Christensen, Rasmus Froberg Brøndum, Matin Bøgsted, and external presenter(s)

Important information concerning PhD courses: 

There is a no-show fee of DKK 3,000 for each course where the student does not show up. Cancellations are accepted no later than 2 weeks before the start of the course. Registered illness is of course an acceptable reason for not showing up on those days. Furthermore, all courses open for registration approximately four months before start of the course.

We cannot ensure any seats before the deadline for enrolment, all participants will be informed after the deadline, approximately 3 weeks before the start of the course.

To attend courses at the Doctoral School in Medicine, Biomedical Science and Technology you must be enrolled as a PhD student.

For inquiries regarding registration, cancellation or waiting list, please contact the PhD administration at phdcourses@adm.aau.dk When contacting us please state the course title and course period. Thank you.