The ARDS, Pneumonia, and Sepsis (APS) Consortium: Rationale, Design, and Feasibility of a National Platform for Phenotyping Critical Illness Syndromes.

Summary

A national, multicenter prospective cohort rapidly enrolled 1,000 critically ill adults and achieved high biospecimen capture across compartments, enabling biologically grounded phenotyping of ARDS, pneumonia, and sepsis. High illness severity and adjudicated syndromic classifications position the platform to support endotype-driven trials.

Key Findings

• Enrolled the first 1,000 participants in under 13 months, ahead of schedule.
• High biospecimen capture: blood 99%, upper respiratory 98%, lower respiratory 37%, urine 80%, gastrointestinal 65%.
• Expert adjudication classified 40% with ARDS, 52% with pneumonia, and 89% with sepsis; in-hospital 4-week mortality was 25%.

Clinical Implications

No immediate practice change; supports future phenotype-based enrollment, prognostication, and therapy selection once endotypes are validated.

Limitations

• Observational design limits causal inference
• Interim analysis focuses on the first 1,000 participants; lower respiratory sampling achievable in only 37%

Future Directions

Leverage the biospecimen-linked dataset to derive and validate biologically coherent ARDS/sepsis endotypes and to embed adaptive, phenotype-enriched interventional trials.