Electronic Children’s Hospital Early Warning Score: Development and Validation of a Novel Machine Learning Algorithm Predicting Critical Deterioration Events in a Retrospective, Single-Center Cohort

OBJECTIVE: To develop and validate a novel, machine learning prediction model for critical deterioration events. DESIGN: Retrospective cohort study. SETTING: Single quaternary care freestanding pediatric hospital. PATIENTS: All patients up to 21 years old admitted to acute care wards (excluding cardiac and psychiatry units) between January 2012 and June 2023. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The primary outcome was a critical deterioration, defined as utilization of advanced respiratory support or continuous IV medication infusions not administered on acute care units. Secondary outcome was ICU transfer. Each outcome was evaluated between 6 and 30 hours after each prediction score. Among 179,074 unique hospitalizations with 2,803,017 windows of scoring data, 10,656 (0.4%) had a subsequent critical deterioration event. Compared with the legacy Children's Hospital Early Warning Score (CHEWS), the model, which we call electronic CHEWS (eCHEWS), had a higher area under the precision-recall curve (0.157; 95% CI [0.150-0.164]) vs. 0.017; 95% CI [0.013-0.027]). At sensitivity approximately 44%, CHEWS had positive predictive value (PPV) 1.8% (number needed to alert [NNA] 55), whereas eCHEWS had PPV 11.4% (NNA 9). CONCLUSIONS: eCHEWS outperforms traditional early warning scores for predicting critical deterioration among acute care inpatients and can detect critical deterioration at least 6 hours before occurrence. If prospectively validated, eCHEWS could serve as an early warning trigger integrated into the electronic health record to support proactive rounding, escalation decisions, and timely transfer to ICUs.

Authors

Institutions

Publication Details

Journal
Pediatric Critical Care Medicine
Published
2026-11-13
DOI
https://doi.org/10.1097/pcc.0000000000004112
Primary Topic
Machine Learning in Healthcare
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Electronic Children’s Hospital Early Warning Score: Development and Validation of a Novel Machine Learning Algorithm Predicting Critical Deterioration Events in a Retrospective, Single-Center Cohort

Alireza Akhondi‐Asl, Alon Geva, Monica E. Kleinman, Lena Saleh
4 links
Pediatric Critical Care Medicine
Machine Learning in Healthcare
article

Electronic Children’s Hospital Early Warning Score: Development and Validation of a Novel Machine Learning Algorithm Predicting Critical Deterioration Events in a Retrospective, Single-Center Cohort

Alireza Akhondi‐Asl, Alon Geva, Monica E. Kleinman, Lena Saleh
article en

Abstract

OBJECTIVE: To develop and validate a novel, machine learning prediction model for critical deterioration events. DESIGN: Retrospective cohort study. SETTING: Single quaternary care freestanding pediatric hospital. PATIENTS: All patients up to 21 years old admitted to acute care wards (excluding cardiac and psychiatry units) between January 2012 and June 2023. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The primary outcome was a critical deterioration, defined as utilization of advanced respiratory support or continuous IV medication infusions not administered on acute care units. Secondary outcome was ICU transfer. Each outcome was evaluated between 6 and 30 hours after each prediction score. Among 179,074 unique hospitalizations with 2,803,017 windows of scoring data, 10,656 (0.4%) had a subsequent critical deterioration event. Compared with the legacy Children's Hospital Early Warning Score (CHEWS), the model, which we call electronic CHEWS (eCHEWS), had a higher area under the precision-recall curve (0.157; 95% CI [0.150-0.164]) vs. 0.017; 95% CI [0.013-0.027]). At sensitivity approximately 44%, CHEWS had positive predictive value (PPV) 1.8% (number needed to alert [NNA] 55), whereas eCHEWS had PPV 11.4% (NNA 9). CONCLUSIONS: eCHEWS outperforms traditional early warning scores for predicting critical deterioration among acute care inpatients and can detect critical deterioration at least 6 hours before occurrence. If prospectively validated, eCHEWS could serve as an early warning trigger integrated into the electronic health record to support proactive rounding, escalation decisions, and timely transfer to ICUs.

Pediatric Critical Care Medicine
Boston Children's Hospital (US), Harvard University (US)
Good health and well-being
Machine Learning in Healthcare
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon