ANZICS Intensive Care Datathon
A Datathon is a great opportunity to explore large intensive care health datasets motivated by real-world questions with clinicians and data scientists working side-by-side.
Learn new skills and be mentored by the best in the field.
What is a Datathon why is it important?
A Datathon is an interactive event where participants work together in teams to solve a clinical question through applying data science tools and techniques to analyse large dataset(s). Clinicians and data scientists (including trainees) work together to generate insights and potential solutions to the clinical question.
Since 2017, ANZICS has hosted Datathons to provide opportunities for intensive care clinicians and trainees, data scientists, biostatisticians and researchers to explore the large repository of data held in the ANZICS Clinical Registry to answer practical questions that may inform or improve patient care.
The ANZICS Intensive Care Datathon brings together clinical and data science enthusiasts and experts to explore the depths of the ANZICS Registry datasets which hold over 3 million episodes of care.
We welcome students and trainees alike to get involved. Datathons are a great way to enhance your skills, refine a formal project idea and collaborate with diverse professionals, whilst being mentored along the way by the best experts in health and data science.
Dr Matt Maiden_Why a Critical Care Datathon is important
Why attend a Datathon? Ask the experts.
Datathon_How to ask answerable questions for publication_Prof Rinaldo Bellomo
Listen to our mentors as they explain how to prepare for the Datathon: Pre-Datathon Webinar
About the Data:
The data provided is extracted from the ANZICS Registry and curated to preserve privacy, while providing enough information for meaningful analyses. Participants are required to sign a Non-Disclosure Agreement, and if the work results in a publication, the ANZICS Publication Policy must be adhered to. If there is scientific merit, we support and encourage teams to continue to work on their field of enquiry towards a formal publication beyond the day of the Datathon.
The datasets frequently available at the Datathon included curated sets from the ANZICS CORE Adult Database, the ANZICS CORE Paediatric Intensive Care Registry, and the Critical Care Resources Dataset.
Who can attend the Datathon?
It is helpful if you are a:
– Clinician (or trainee) with real experience in the critical care environment
– Data Scientist (or trainee) with experience in
- R/Python/Matlab or SQL queries
- machine learning
- statistical modelling or data visualisation.
How does it work?
Teams are each assigned a clinical question to interrogate one or more datasets with the aim of generating an insight or solution to the question. At the close of the day, each team will present their finding(s), providing the opportunity for participants to develop skills in showcasing the meaning and value behind the data and its application to the real world.
*Please note, some preparatory work may be required by participants prior to start of the Datathon.
What if we get stuck on a problem?
There is plenty of support for teams to navigate the challenge of the day, with Datathon mentors and experts in the field of Intensive Care Medicine and Data Science available to assist and mentor you every step of the way.
What do I need to bring?
Bring your own laptop (don’t forget the charger), be open to working with others and ready to jump in for the challenge. A rewarding and fun learning opportunity awaits!
What datasets will be available?
Teams may choose to use one or all the datasets on offer to answer their assigned clinical question.
The ANZICS Intensive Care Registry Datasets:
• The Adult Patient Database (FY 2018 – 2023)
• The Critical Care Resources Dataset (FY 2018 – 2023)
• The Socio-economic data by postcode
How are the Datathon presentations judged?
Abstract presentations are reviewed by the scientific mentors. Judging is based on clinical relevance and impact – i.e. how your solution may influence or improve current practice, the novelty and impact of the proposed problem statement, the result presentation, and the Team’s plans for the study moving forward.
Find out more about the ANZICS CORE datasets:
APD and CCR Databases by David Pilcher_ANZICS CORE