Development and usability testing of a machine learning clinical decision support tool for suggesting etiologies and treatment of clinical deterioration
Authors' conclusion
Our controlled usability tests suggest that incorporating these requirements may improve overall usability.
Study context
- Population
- Critical care or hospital medicine clinicians
- Participants
- 23 usability tests
- Comparison
- Experimental vs Control
- Outcome
- Usability (diagnostic accuracy, perceived workload, satisfaction)
Methodological detailsfocus groups, usability testing, within-participant comparison with washout period, controlled lab setting
- Focus groups
- Usability testing
- Within-participant comparison with washout period
- Controlled lab setting
Main finding
As reported in the abstract
- Experimental interface associated with improvements in diagnostic accuracy, perceived workload and satisfaction vs control
- Experimental vs Control: P < 0.05
- Controlled lab setting
Further findings
Design requirements (focus groups)
- Understanding rationale for ML-CDS recommendations; assessing how EHR data quality issues affect accuracy; key information on a single screen; graphical visualization; integration into pre-existing care pathways
Original publication
Development and usability testing of a machine learning clinical decision support tool for suggesting etiologies and treatment of clinical deterioration. J Crit Care. Published 18 September 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →