Perioperative medicine & outcomes
Postoperative Vasoactive Exposure Patterns and Outcomes After Non‐Cardiac Surgery: A Prespecified Analysis of the SQUEEZE Cohort
Authors' conclusion
Postoperative vasoactive exposure follows reproducible patterns characterised by persistence and overall treatment burden rather than binary classification alone. These patterns were associated with mortality, AKI, and hospital length of stay, and provided prognostic information beyond duration alone.
Study context
- Population
- Adults receiving postoperative vasoactive infusion within seven postoperative days after non-cardiac surgeryreceived postoperative vasoactive infusion (26.0%) among 25,672 patients
- Participants
- 6,676 adults
- Comparison
- Postoperative vasoactive exposure patterns (identified by k-means clustering) vs duration of postoperative vasoactive therapy for prognostic information
Methodological detailsunsupervised k-means clustering, multivariable regression, prespecified landmark analysis
- Unsupervised k-means clustering
- Multivariable regression
- Prespecified landmark analysis
Primary result
As reported in the abstract
Adjusted 30-day mortality across exposure patterns · 30 days
Low-burden pattern
9.7%
High-burden pattern
35.3%
Further findings
Adjusted acute kidney injury (AKI) risk across exposure patterns
- Increased across patterns from
- 28.0%
- Increased across patterns to
- 56.0%
- Adjusted
Exposure patterns identified
- Reproducible exposure patterns identified
- Four
Discrimination of covariate-adjusted models (same complete-case cohort)
- Model with exposure pattern, AUC
- 0.818
- Model with duration, AUC
- 0.795
- Pattern plus duration, AUC
- 0.819
- Adding duration to pattern classification resulted in little additional discrimination
Early postoperative landmark analysis (postoperative days 0-1)
- Findings consistent
- Prespecified
Original publication
Postoperative Vasoactive Exposure Patterns and Outcomes After Non‐Cardiac Surgery: A Prespecified Analysis of the SQUEEZE Cohort. Acta Anaesthesiol Scand. Published 11 October 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →