Introduction
Artificial intelligence–based clinical decision support (AI-CDS) tools to enhance personalized medicine is on the rise. In the ICU, complex and high-stakes decisions are made that might benefit from data-driven decision support. However, implementation of these AI-C DS tools in clinical ICU practice is lacking due to several levels. These difficulties include patient privacy, regulatory aspects, and a lack of demonstrations of their value. There is a need to study human factors for the safe and effective implementation of AI-CDS tools, as high predictive performance does not ensure acceptance. In terms of the complexity of decisions in ICU and the pressure under which decisions have to be made, knowledge focused on this clinical domain is highly relevant. Pacmed Critical is a machine learning–based AI-CDS tool that predicts a patient’s combined readmission and mortality risk within 7 days of ICU discharge. The software is intended for use as a complementary tool by qualified ICU medical professionals. It will be accessed on hospital premises; it will not be used on mobile devices. This study aimed to investigate physicians’ perspectives and their current decision-making behavior before implementing a discharge AI-CDS tool for predicting readmission and mortality risk after ICU discharge.
Methods
We conducted a survey of physicians involved in decision-making on discharge of patients at two Dutch academic ICUs between July and November 2021. Questions were divided into four domains:
(1) physicians’ current decision-making behavior with respect to discharging ICU patients,
(2) perspectives on the use of AI-CDS tools in general,
(3) willingness to incorporate a discharge AI-CDS tool into daily clinical practice, and
(4) preferences for using a discharge AI-CDS tool in daily workflows.
Results
Most of the 64 respondents (of 93 contacted, 69%) were familiar with AI (62/64, 97%) and had positive expectations of AI, with 55 of 64 (86%) believing that AI could support them in their work as a physician. The respondents disagreed on whether the decision to discharge a patient was complex (23/64, 36% agreed and 22/64, 34% disagreed); nonetheless, most (59/64, 92%) agreed that a discharge AI-CDS tool could be of value. Significant differences were observed between physicians from the 2 academic sites, which may be related to different levels of involvement in the development of the discharge AI-CDS tool.
Conclusions
To sum up, this study offers insightful perspectives on the usage of AI-CDS tools that may be used in the implementation process and the training of end users, as well as useful insights into current decision-making behaviour. The use of AI-CDS tools in the ICU environment in general and a tool that forecasts a patient’s probability of readmission and mortality within seven days of discharge in particular was well received by ICU doctors. The observed variations between the two research locations, each of which contributed differently to the creation of the AI-CDS tool, highlight the necessity of education and training for departments with minimal prior exposure to AI-CDS. It is recommended to involve end-users in AI-CDS tool development and conduct surveys for acceptance, improve design and facilitate clinical adoption.
Source: van der Meijden SL, de Hond AAH, Thoral PJ, Steyerberg EW, Kant IMJ, Cinà G, Arbous MS Intensive Care Unit Physicians’ Perspectives on Artificial Intelligence–Based Clinical Decision Support Tools: Preimplementation Survey Study, JMIR Hum Factors 2023;10:e39114, doi: 10.2196/39114.