Learning outcomes

At the end of this course, the student should be able to:

-        Organize, document and prepare biomedical and health data for analysis;

-        Use, under supervision, programming environments such as R to carry out the main steps of a health data management and analysis workflow;

-        Select and apply certain biostatistical methods to answer a biomedical question;

-        Identify limitations that may affect the validity, reproducibility and generalizability of health data analysis results;

-        Critically interpret analytical results from biomedical and health perspectives.

These learning outcomes may be adjusted depending on the schedule, available resources, the number of enrolled students, and identified teaching needs. If necessary, they will be specified at the beginning of the course and on the Webcampus platform.

Goals

This course aims to develop applied skills in analyzing biomedical and health data, including when these data are complex, multimodal, or high-dimensional.

It should enable students to apply health data science principles to produce robust, reproducible, and relevant results for research, clinical practice, and public health.

The objective is also to develop a critical perspective on the performance, applications and limitations of health data models and results.

Content

The course may cover the following topics:

-        Types and sources of biomedical and health data;

-        Management and preparation of biomedical and health data in environments such as R;

-        Principles for linking multimodal databases, particularly biological, clinical and imaging data, and data organized at the molecular, clinical and population levels;

-        Concepts of data harmonization and standardization, with an introduction to international initiatives (OMOP and the European Health Data Space);

-        Main biostatistical and data science methods used to answer a biomedical or health question, and issues related to the analysis of high-dimensional health data in environments such as R;

-        Main methodological limitations of health data analysis results;

-        Interpretation and communication of scientific results from health data analysis.

The content is indicative and may change depending on the schedule, available resources, accessible data and identified teaching needs. Any adaptations will be specified at the beginning of the course and on the Webcampus platform.

Table of contents

1. Biomedical and health data: types, structures and quality

2. Data management

3. Biostatistical models for the analysis of biomedical health data

4. Biases and limitations of health data analysis models and results

5. Interpretation and communication of results

6. Practical sessions

 

This table of contents is indicative and may change depending on the schedule, available resources, the datasets used and identified teaching needs. If necessary, it will be specified at the beginning of the course and on the Webcampus platform.

Exercices

The practical sessions may follow a common thread from data preparation to communication of results. Starting from a biomedical or health problem, students could be asked to:

-        Explore and prepare a health dataset;

-        Formulate an analysis strategy for a biomedical question and apply one or more biostatistical models;

-        Interpret biostatistical results and examine biases and limitations;

-        Structure the approach, results and discussion in a report, following the main sections of a scientific article.

These activities should form the basis of the assessed work. Their organization, the datasets used, group composition and follow-up arrangements will be specified at the beginning of the course and on the Webcampus platform.

Teaching methods

Teaching should combine theoretical sessions (30 h) and practical sessions (15 h). Expectations and practical organization will be specified at the beginning of the course.

Activities could include:

-        Theoretical lectures intended to present methodological concepts and their conditions of application;

-        Demonstrations by the instructor using existing databases and study materials in programming environments such as R;

-        Practical sessions organized as problem-based learning groups, during which students would progress from data preparation to modeling, then to interpretation and communication.

These arrangements may be adjusted depending on the schedule, available computing resources, the number of enrolled students and the group’s progress. If necessary, they will be specified at the beginning of the course and on the Webcampus platform.

Assessment method

The assessment may comprise two parts: a written examination and a report analyzing a biomedical or health problem.

The written examination may include short-answer open questions and/or multiple-choice questions. It may also include an application exercise drawing on the concepts and exercises covered during the practical sessions.

Participation in all practical sessions is expected and does not receive an attendance grade. These sessions will focus on exercises and the progressive resolution of a biomedical or health problem requiring data analysis. This work will result in a written report, to be submitted at the end of the course, which will contribute to the final grade. The report will normally be completed in groups, subject to adaptations depending on the number of enrolled students.

The respective weighting of the written examination and the report will be specified at the beginning of the course and published on WebCampus.

In the second examination session, the assessment arrangements will depend on the component that was not passed:

-        if the written examination is failed, the student must sit a new written examination in the same format;

-        if the report is failed, the student must revise it individually, independently and outside the practical sessions. The improved version must be submitted by a date set before the oral examination, during which the student will present and defend the work;

-        if both components are failed, the student must retake the written examination and individually defend an improved version of the report during the oral examination.

These arrangements may be adapted in accordance with institutional rules and will be specified on the Webcampus platform.

Sources, references and any support material

Reference books

 

-        Boland M.R. Health Analytics with R: Learning Data Science Using Examples from Healthcare and Direct-to-Consumer Genetics. Springer.

-        Etzioni R., Mandel M., Gulati R. Statistics for Health Data Science: An Organic Approach. Springer.

Teaching materials

-        Course slides made available on Webcampus;

-        Potential introductory documents, online resources or articles selected by the instructor and made available on Webcampus;

-        Health datasets, visualizations, demonstrations and case studies presented during the course and practical sessions (R environment) and made available on Webcampus.

Language of instruction

English