Decoding candidemia in critically ill patients: unsupervised clustering identifies three unique phenotypes
Authors' conclusion
Unsupervised clustering identified three clinically different candidemia phenotypes with different outcomes. Cirrhosis, higher age and illness severity were associated with mortality, whereas a catheter-related source of infection was protective.
Study context
- Population
- Intensive care unit (ICU) patients with candidemia from 16 French ICUs (2015–2023)
- Participants
- 492 patients · 16 French ICUs
- Exposure
- Three clinical phenotypes identified by unsupervised clustering
- Phenotype 1n = 70
- Phenotype 2n = 223
- Phenotype 3n = 199
- Outcome
- Mortality
- Follow-up
- 90 days
Methodological detailsunsupervised clustering (factor analysis of mixed data (FAMD) followed by hierarchical clustering on principal components (HCPC))
- Unsupervised clustering (FAMD followed by HCPC)
Main finding
As reported in the abstract
90-day mortality by phenotype
P < 0.001
- Overall 90-day mortality
- 62.6%
Observational data: associations, not effects.
Further findings
Independent predictors of mortality (Cox model)
| aHR (95% CI) | P value | ||
|---|---|---|---|
| Age | 1.01 (1.00–1.02) | P = 0.003 | |
| Cirrhosis | 1.90 (1.39–2.60) | P < 0.001 | |
| SAPS II | 1.01 (1.01–1.02) | P < 0.001 | |
| Echinocandin use | 0.49 (0.39–0.63) | P < 0.001 | |
| Proven catheter-related candidemiaProtective | 0.48 (0.33–0.69) | P < 0.001 | |
| 0.20.5125Adjusted hazard ratio (log scale) |
- Immunodepression not associated with mortality
Original publication
Decoding candidemia in critically ill patients: unsupervised clustering identifies three unique phenotypes. Crit Care. Published 28 June 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →