Implementing and Interacting with Artificial Intelligence at the Intensive Care

Human-Machine Interaction and Pre-Implementation Process of Pacmed Critical at the ICU. HUMAN-IMPACT

Artificial Intelligence (AI) has the potential to revolutionise the healthcare sector, but hardly any AI tools have been implemented into clinical care yet. This project will produce guidance to bring AI into clinical practice and secure a successful pre-implementation framework for future projects. Pacmed Critical, an AI decision support tool for the ICU developed by Pacmed will be used to study the process leading up to implementation.

Pacmed Critical intends to address the capacity strain on Intensive Care Units (ICU) by supporting intensivists in determining the optimal moment for discharge. Premature discharge can lead to readmission or even death. Successful implementation of the AI decision support tool could lead to an increase in general longevity and quality of life of ICU patients. Also, by providing guidance on the facilitators and barriers to AI implementation, this project can inform other implementation trajectories and promote the large-scale adoption of AI in clinical practice. This will unlock the major potential of AI to improve the efficiency and quality of our healthcare systems.

A multidisciplinary team will investigate the steps leading up to successful implementation of Pacmed Critical in simulated and operational environments at the LUMC through interviews, focus groups, job shadowing and a prospective experiment on decision making. The aim is to obtain the following outcomes:

  1. Optimise the user interface design, functionality and integration in the clinical workflow of an AI decision support tool, in a way that is user-friendly, interpretable and safe. 
  2. Identify critical factors, i.e. related to clinician’s view and product functionality, to develop the most effective pre-implementation strategy for an AI decision support tool.
  3. Distill these findings into best practices that will be shared with the medical AI community.
Summary
Artificial Intelligence (AI) has the potential to revolutionise the healthcare sector, but hardly any AI tools have been implemented into clinical care yet. This project will produce the missing guidance to bring AI into clinical practice and secure a successful pre-implementation framework for future projects.
Technology Readiness Level (TRL)
6 - 7
Time period
26 months
Partners
Logo
Logo
Logo