phoenix: an R package and Python module for calculating the Phoenix pediatric sepsis score and criteria

Peter E DeWitt1, Seth Russell1, Margaret N Rebull1

  • 1Department of Biomedical Informatics, University of Colorado School of Medicine, University of Colorado Anschutz Medical Campus, Aurora, CO 80045, United States.

JAMIA Open
|July 5, 2024
PubMed

Insights

New R package and Python module help researchers consistently apply the Phoenix criteria for pediatric sepsis and septic shock using electronic health records (EHRs). These tools ensure accurate data analysis for improved clinical care and research.

Area of Science:

  • Pediatric critical care medicine
  • Clinical informatics
  • Biostatistics

Background:

  • The Phoenix criteria represent a significant advancement in defining pediatric sepsis and septic shock.
  • Consistent application of these criteria is crucial for robust research and clinical practice.
  • Electronic health records (EHRs) are vital data sources for sepsis research.

Purpose of the Study:

  • To develop user-friendly tools for applying the Phoenix criteria to EHR data.
  • To facilitate consistent and accurate analysis of pediatric sepsis data.
  • To support a unified body of research across different institutions.

Main Methods:

  • Development of the "phoenix" R package.
  • Development of a "phoenix" Python module.
  • Functions designed for intuitive application to EHR datasets.

Main Results:

  • The R package and Python module enable derivation of key indicators, total scores, and sub-scores from EHR data.
  • These tools ensure consistent and accurate implementation of the Phoenix criteria.
  • The software is readily available for research use.

Conclusions:

  • The "phoenix" R package and Python module are essential for researchers utilizing the Phoenix criteria.
  • These tools promote standardized data analysis, enhancing the reliability of pediatric sepsis research.
  • The availability of these free tools supports improved clinical care and research globally.
Abstract