Pediatric anesthesia & critical care
Cytokine-based strategies to improve prognostic enrichment of pediatric acute respiratory distress syndrome (ARDS)
Authors' conclusion
Cytokine-based prognostic enrichment strategies show promise for children with ARDS. In addition to models based primarily on IL-6 and TNFR1, models including CCL7 and IL-18 may be of utility for assessing the full spectrum of inflammation.
Study context
- Population
- Children with ARDS; plasma within 24 h of ARDS, cytokines by microfluidic immunoassay
- Participants
- 500 children
- Comparison
- Traditional 3-term phenotype model vs parsimonious cytokine models
- Trainingn = 300
- Validationn = 200
- Primary endpoint
- 90-day mortality discrimination
- Follow-up
- 90 days
Methodological detailsprognostic model development and validation
- Prognostic model development and validation
Primary result
As reported in the abstract
| Group | AUC | AUROC, training set (n = 300) | AUROC, validation set (n = 200) | Sensitivity | PPV |
|---|---|---|---|---|---|
| Traditional 3-term (IL-6, TNFR1, bicarbonate) | 0.62 | — not stated in the abstract | — not stated in the abstract | 26% | 24% |
| Two-term (CCL7, IL-18) and three-term (CCL7, IL-18, TNFR1) | — not stated in the abstract | 0.72-0.73 | 0.66 | — not stated in the abstract | — not stated in the abstract |
AUC
Traditional 3-term (IL-6, TNFR1, bicarbonate)0.62
Two-term (CCL7, IL-18) and three-term (CCL7, IL-18, TNFR1)— not stated in the abstract
AUROC, training set (n = 300)
Traditional 3-term (IL-6, TNFR1, bicarbonate)— not stated in the abstract
Two-term (CCL7, IL-18) and three-term (CCL7, IL-18, TNFR1)0.72-0.73
AUROC, validation set (n = 200)
Traditional 3-term (IL-6, TNFR1, bicarbonate)— not stated in the abstract
Two-term (CCL7, IL-18) and three-term (CCL7, IL-18, TNFR1)0.66
Sensitivity
Traditional 3-term (IL-6, TNFR1, bicarbonate)26%
Two-term (CCL7, IL-18) and three-term (CCL7, IL-18, TNFR1)— not stated in the abstract
PPV
Traditional 3-term (IL-6, TNFR1, bicarbonate)24%
Two-term (CCL7, IL-18) and three-term (CCL7, IL-18, TNFR1)— not stated in the abstract
— not stated in the abstract
- Improved net reclassification with CCL7/IL-18 models
- ≈40%
Further findings
Cohort characteristics · 90 days
- Hyper-inflammatory
- 14%
- 90-day mortality
- 13.4%
CCL7 (MCP-3) association with 90-day mortality
- Most significant association among additional cytokines
- P < 0.0001
Immune compromised status and 90-day mortality
- Independently associated; optimal model may vary by immunocompromised status
- P < 0.0001
Original publication
Cytokine-based strategies to improve prognostic enrichment of pediatric ARDS. Crit Care. Published 1 July 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →