| Dozent/in |
PD Dr. med. Patrick Beeler |
| Veranstaltungsart |
Vorlesung |
| Code |
FS231145 |
| Semester |
Frühjahrssemester 2023 |
| Durchführender Fachbereich |
Gesundheitswissenschaften |
| Studienstufe |
Master |
| Termin/e |
Mi, 22.02.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 01.03.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 08.03.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 15.03.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 22.03.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 29.03.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 05.04.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 19.04.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 26.04.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 03.05.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 10.05.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 17.05.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 24.05.2023, 08:30 - 10:00 Uhr, 4.B55
Mi, 31.05.2023, 16:15 - 18:00 Uhr, 4.B55 |
| Umfang |
2 Semesterwochenstunden |
| Inhalt |
In healthcare, more and more data are routinely collected and stored, driven bydigitalization. Such data are called real-world data which the U.S. Food and Drug Administration (FDA) defines as “[…] data relating to patient health status and/or the delivery of health care routinely collected from a variety of sources”. In the research context, real-world evidence results from the analysis of real-world data.
Electronic health records constitute an important real-world data source that collects data during routine clinical practice for patient management and documentation purposes. Electronic health record data can be used to address novel research questions with minimal risks for patients. According to the “Framework for FDA’s Real-World Evidence Program”, real-world evidence may help expand indications for drugs only approved for specific conditions.
Curiosity is an asset in the seminar Analysis of routinely collected health care data. The students will get the opportunity to exploratively work on anonymized but real patient data routinely collected in electronic health records in intensive care units (the MIMIC patient datasets). The students will get to know scientific articles based on MIMIC data, will practice the handling of large patient datasets, will learn how to process and analyze data and how to apply appropriate statistical methods and machine learning for research purposes. Over the course of this seminar, the students will develop their own real-world evidence project. |
| E-Learning |
To become a credentialed user by following the instructions on https://mimic.mit.edu/docs/gettingstarted/ is a prerequisite. This prerequisite includes an e-learning training course before a student gets access to the MIMIC patient datasets.
In this seminar, the students will work on their own devices (tutorials, exercises, MIMIC patient data analysis). |
| Lernziele |
After having completed this seminar, you will
- be able to deal with large datasets of real patient data routinely collected in electronic health records
- know how to explore, understand and describe such real-world data, be aware of the advantages and disadvantages of real-world data
- know what techniques are used to process, transform, aggregate and present patient data
- be able to apply the most important statistical methods to generate real-world evidence
- and you will have understood the basic principles and methods of machine learning |
| Voraussetzungen |
• Become a credentialed user by following the instructions on https://mimic.mit.edu/docs/gettingstarted/
before the start of the seminar.
• It’s a hands-on seminar: Bring your own device.
Overall grade of 4.0 or better. |
| Sprache |
Englisch |
| Begrenzung |
This a core course in the major "Health Data Science" |
| Anmeldung |
https://elearning.hsm-unilu.ch/course/view.php?id=601 |
| Prüfung |
1. First oral presentation with slides of a scientific article, on April 5 during seminar (not graded)
2. Written abstract on student’s own project developed over the course of the seminar, submission before oral presentation on May 31 (not graded*)
3. Second oral presentation with slides of student’s own project developed over the course of the seminar, on May 31 during seminar (mean grade of the two examiners’ grades; *abstract may be considered in cases of disagreement between examiners)
Co-examinators:
Prof. Dr. med. Balthasar Hug
Dr. med. Dr. sc. nat. Michael Havranek |
| Abschlussform / Credits |
First oral presentation with slides, Written abstract,Second oral presentation with slides / 3 Credits
|
| Hinweise |
Teaching methods:
Longitudinal seminar with blended learning, including lectures, tutorials, hands-on exercises and class discussions. |
| Hörer-/innen |
Ja |
| Kontakt |
patrick.beeler@unilu.ch |
| Material |
The teaching material is based on slides, hands-on exercises in class, selected scientific articles, and online resources. Offline material will be provided via moodle. |
| Literatur |
While slides and selected scientific articles will be presented and discussed,in this seminar it will be more important for the students- to learn and practice working on data, - to be curious and to explore data, techniques and methods, - to get to know essential online resources, and - to learn resolving issues/overcoming obstacles with the help of online research. |