Phenotyping Multiple Organ Dysfunction Syndrome Using Temporal Trends in Critically Ill Children

Area of Science:

• Pediatric critical care medicine
• Computational biology
• Systems medicine

Background:

• Multiple organ dysfunction syndrome (MODS) is a leading cause of mortality in critically ill children.
• Current management strategies and targeted therapies for pediatric MODS have shown limited success in improving outcomes.
• Understanding the heterogeneity of MODS is crucial for developing effective treatment approaches.

Purpose of the Study:

• To identify distinct subgroups within pediatric MODS patients.
• To characterize the clinical heterogeneity and temporal patterns of organ dysfunction in pediatric MODS.
• To determine if identified subgroups are predictive of clinical outcomes.

Main Methods:

• Analysis of a large cohort (5,297 children) with MODS from two children's hospitals.
• Application of subgraph-augmented non-negative matrix factorization (SANMF) to identify temporal patterns.
• Characterization of patient subgroups based on clinical features and outcomes.

Main Results:

• Identification of four novel subgroups within the pediatric MODS cohort.
• Demonstration that these subgroups exhibit distinct clinical characteristics.
• Validation that the identified subgroups are independently predictive of clinical outcomes.

Conclusions:

• The four identified subgroups represent distinct phenotypes of pediatric MODS.
• These phenotypes can aid in a better understanding of MODS heterogeneity.
• The findings suggest potential for developing novel, targeted management strategies for pediatric MODS.