2023
DOI: 10.1097/ccm.0000000000006100
|View full text |Cite
|
Sign up to set email alerts
|

Does Reinforcement Learning Improve Outcomes for Critically Ill Patients? A Systematic Review and Level-of-Readiness Assessment

Martijn Otten,
Ameet R. Jagesar,
Tariq A. Dam
et al.

Abstract: Objective: Reinforcement learning (RL) is a machine learning technique uniquely effective at sequential decision-making, which makes it potentially relevant to ICU treatment challenges. We set out to systematically review, assess level-of-readiness and meta-analyze the effect of RL on outcomes for critically ill patients. Data Sources: A systematic search was performed in PubMed, Embase.com, Clarivate Analytics/Web of Science Core Collection, Elsevier/S… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1

Citation Types

0
2
0

Year Published

2024
2024
2024
2024

Publication Types

Select...
2

Relationship

1
1

Authors

Journals

citations
Cited by 2 publications
(2 citation statements)
references
References 52 publications
0
2
0
Order By: Relevance
“…More specifically, these are algorithms that are capable of providing sequential suggestions for treatment within the ICU, such as ventilator settings or the optimal balance between vasopressors and fluids. While the results of reinforcement learning algorithms in retrospective databases are promising, there are no studies published that study performance in a prospective setting [ 23 ▪ ].…”
Section: Artificial Intelligence In the Icumentioning
confidence: 99%
“…More specifically, these are algorithms that are capable of providing sequential suggestions for treatment within the ICU, such as ventilator settings or the optimal balance between vasopressors and fluids. While the results of reinforcement learning algorithms in retrospective databases are promising, there are no studies published that study performance in a prospective setting [ 23 ▪ ].…”
Section: Artificial Intelligence In the Icumentioning
confidence: 99%
“…It is important to note that serial ABGs in patients are not statistically independent. Indeed, a future computerized CDS system could use modern machine learning methods like reinforcement learning (9) to adjust the system’s recommendation for subsequent ventilator changes based on whether a Pa co 2 after an index ventilator change resulted in an ABG that was higher or lower than expected, much as clinicians do at the bedside.…”
mentioning
confidence: 99%