[MA 2024 06] Sustainable Data-Driven ICU Care
Quality of data, data standards, information model, terminologies
Proposed by: Sade Faneyte, MSc., EngD [s.faneyte@erasmusmc.nl]
At the Erasmus MC Datahub, we aim to make data-driven healthcare scalable and clinically applicable. The Datahub is the place where colleagues inspire each other and brainstorm meaningful clinical applications of data and AI. Together, we work towards the healthcare of the future. One of our research projects dives into the sustainable ICU supported by Zorginstituut Nederland. Engage in knowledge exchange about ontology and data stations, supported by experts from the Zorginstituut’s KIK-V program.1 We offer a dynamic work environment where collaboration and innovation are at the core. You also get the opportunity to be a pioneer in transforming healthcare. If you’re looking to make a significant impact on an innovative project, this is your chance!
Introduction
The healthcare sector faces an urgent need to adapt to the increasing health demands related to the ecological crisis while reducing its own environmental footprint. The healthcare sector is responsible for 6-7% of CO2 emissions and the Dutch healthcare sector's commitment, through the Green Deal 3.0, aims for a 50% reduction in primary raw material use by 2030 and fully circular healthcare by 20502. Achieving these goals necessitates a data-driven approach to make informed decisions towards sustainability. The Erasmus Medical Center already utilizes sustainability data and has previously shown the environmental effects of materials used in the intensive care unit (ICU).3 In this study, the mass, carbon footprint, agricultural land occupation and water usage of materials in the ICU were determined and so-called "environmental hotspots" were found. Based on this analysis the ICU-staff was able to take action to reduce the impact of the department and shift towards a circular economy.4 Also, the carbon footprint of the entire Erasmus Medical Center has been studied for every department in the hospital. However, the need for environmental, social and governance (ESG) considerations is important for making future-proof decisions and we need sustainability data to guide us.5 With the hospital’s leading position in the data-driven ICU and AI ethics through the Datahub and AI-ethical lab, the question also arises about sustainable use of AI-technologies.6,7 For example, there are concerns about the emissions from training and deploying AI models because of the large energy use, especially for the use of Large Language Models.8 On the other hand, AI could also reduce emissions through process optimization and changing models of care. Importantly, one of the World Health Organization’s (WHO) core principles of responsible use of AI is sustainable AI.9 This research seeks to develop a proof of concept (PoC) that enables effective sustainability monitoring within the ICU. By leveraging real hospital data, the project aims to create measurable sustainability indicators that can influence ICU staff behavior and reduce environmental impact. The challenge is to gain insights into the data availability to achieve this. The additional challenge lies in extracting relevant data from hospital information systems uniformly across individual hospitals and nationwide, ensuring comparability and actionable insights. This proposal aims to address the urgent need for sustainable practices in ICU care by leveraging data and technology to make a meaningful impact.
Description of the SRP Project/Problem
This Scientific Research Project aims to investigate a proof of concept for monitoring and enhancing sustainability in ICU care using data-driven approaches by:
· Conducting a data analysis of available sustainability data in the ICU (such as patient, consumption, and procurement information of healthcare materials)
· Identifying a uniform method related to data extraction, knowledge-modelling and information exchange on a nationwide level to monitor progress against overall sustainability goals
· Investigating the effectiveness of the ICU staff’s actions to improve sustainability
Research questions
1. How to monitor and enhance sustainability in ICU care using data-driven approaches?
· How can we effectively measure and monitor the sustainability impact of the ICU based on operational process information?
· What specific actions can be taken to make ICU care more sustainable and how can we measure the effectiveness of these actions?
· How can we (potentially) develop AI-models that supports ICU staff in making sustainable choices?
· How can we develop insights into the carbon footprint of AI models used in the ICU?
2. How can we improve the sustainability of ICU care based on data that works at the individual hospital level but is also scalable to a network and national level?
· How can we uniformly extract relevant sustainability data about the ICU process from the Hospital Information Systems (HIS)?
· Would a machine-readable knowledge model (ontology) be effective to make ICU sustainability data comparable related to uniform data availability across hospitals?
· How can we make data available in an effective way (for example, using a data station and standards) so that both information can be requested and predictive algorithms can be run?
· How can we realise a working PoC based on real hospital data from the HIS systems preferably between two or three hospitals? Including deployment of FAIR datastations.
Expected results
· Contribution to a (scientific) article and webinar to share findings with healthcare professionals and policymakers
· Open-source publication of results on GitHub (for example, a set of sustainability indicators, ontology, AI-models)
Are you the data enthusiast we’re looking for? Apply now and help shape the future of healthcare at Erasmus MC Datahub! We also offer internship compensation including travel compensation based on a full-time internship for 6-9 months.
Time period
o May - November
Completion of the research by December 31, 2024
Contact
Sade Faneyte, MSc., EngD | s.faneyte@erasmusmc.nl
References
1 https://kik-v-publicatieplatform.nl/
2 https://www.greendealduurzamezorg.nl/
3 Hunfeld, N., Diehl, J. C., Timmermann, M., van Exter, P., Bouwens, J., Browne-Wilkinson, S., de Planque, N., & Gommers, D. (2023). Circular material flow in the intensive care unit-environmental effects and identification of hotspots. Intensive care medicine, 49(1), 65–74.
4 Hunfeld, N., Diehl, J. C., de Zeeuw, S. R. G., Gommers, D., & Raaij, E. M. (2023). The Green Intensive Care: From Environmental Hotspot to Action. ICU Management & Practice, 23(3).
5 https://kpmg.com/xx/en/home/insights/2023/08/putting-esg-at-the-heart-of-healthcare-governance.html
6 https://datahub.health/
7 https://amazingerasmusmc.nl/algemeen-nieuws/erasmus-mc-en-tu-delft-openen-eerste-ai-ethieklab-voor-de-zorg/
8 Stanford, H. A. I. (2023). The AI Index Report: Measuring Trends in Artificial intelligence [Ebook]
9 Guidance, W. H. O. (2021). Ethics and governance of artificial intelligence for health. World Health Organization.