ICU sedation, delirium & recovery
Pharmacophenotype identification of intensive care unit medications using unsupervised cluster analysis of the ICURx common data model
Authors' conclusion
The results of this evaluation suggest that patterns among patient clusters and medication regimens may be observed using empiric methods of unsupervised machine learning in combination with a common data model.
Study context
- Population
- Critically ill adults30,550 medication orders
- Participants
- 991 adults
- Exposure
- Patient clusters derived from Intensive Care Unit (ICU) medication administration records in the first 24 h
- Outcome
- Pharmacophenotypes (medication clusters) correlated with ICU adverse events and patient-centred outcomes
- Follow-up
- First 24 h of ICU stay
Methodological detailsobservational, unsupervised machine learning (restricted Boltzmann machine, hierarchical clustering)
- Observational
- Unsupervised machine learning (restricted Boltzmann machine, hierarchical clustering)
Main finding
As reported in the abstract
Five unique patient clusters identified
5
Six unique pharmacophenotypes identified
6
Observational data: associations, not effects.
Original publication
Pharmacophenotype identification of intensive care unit medications using unsupervised cluster analysis of the ICURx common data model. Crit Care. Published 2 May 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →