Development of ePILOT, an Electronic Medical Record–Based Measure of Intracranial Pressure–Directed Therapy Intensity in Pediatric Traumatic Brain Injury

Background/Objective : Existing therapy-intensity scales for traumatic brain injury (TBI) require manual abstraction and are scored at coarse intervals, limiting temporal alignment with intracranial pressure (ICP). We aimed to develop an hourly, electronic medical record (EMR)-derived Pediatric Intensity Level of Therapy (PILOT) scale (ePILOT) and characterize its relationships with ICP and in-hospital mortality. Methods : We retrospectively analyzed hourly EMR data from the first 7 days of hospitalization in children with TBI who underwent ICP monitoring at a single center. Twelve components adapted from PILOT (cerebrospinal fluid diversion, hyperosmolar therapy, sedation/analgesia, barbiturates, neuromuscular blockade, temperature, PaCO 2 , and serum sodium) were derived from routinely stored data. A Mixed Graphical Model descriptively characterized conditional dependencies among components and ICP. Integer component weights were informed by associations with in-hospital mortality and adjusted to preserve guideline tier ordering, yielding an hourly score of 0–24. Patient-level relationships with ICP and mortality were assessed. Results : Among 135 ICP-monitored TBI patients (median age 4.8 [1.9–12.7] years; 65% male; admission Glasgow Coma Scale (GCS) score 4 [3–7]; mortality 16%), ePILOT increased progressively with ICP (patient-level Spearman ρ = 0.32, p < 0.001). Non-survivors had higher mean ICP (26 vs. 12 mmHg) and median hourly ePILOT (7.0 [5.0–10.4] vs. 3.5 [3.0–6.5]) than survivors (both p < 0.05). Peak hourly ePILOT discriminated mortality with an area under the receiver operating characteristic curve (AUROC) of 0.74 (95% CI 0.61–0.85), compared with 0.68–0.72 for ICP summaries and 0.61 for admission GCS. A median hourly ePILOT score of 5 was the Youden-optimal mortality cut-point (95% CI 4–8; sensitivity 0.77; specificity 0.61). Conclusions : ePILOT converts routinely stored EMR data into an automated, hourly measure of ICP-directed therapy intensity. Therapy intensity tracked ICP, lagged behind it, and discriminated mortality comparably to ICP, consistent with a reactive treatment response. External validation against manually scored PILOT should follow.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-18
DOI
https://doi.org/10.21203/rs.3.rs-11555789/v1
Primary Topic
Traumatic Brain Injury and Neurovascular Disturbances
Type
preprint

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Development of ePILOT, an Electronic Medical Record–Based Measure of Intracranial Pressure–Directed Therapy Intensity in Pediatric Traumatic Brain Injury

Jonathan H. Pelletier, Alicia K. Au, Mohammed A. Shaik, Robert S. B. Clark et al.
2 links
Research Square
Traumatic Brain Injury and Neurovascular Disturbances
preprint

Development of ePILOT, an Electronic Medical Record–Based Measure of Intracranial Pressure–Directed Therapy Intensity in Pediatric Traumatic Brain Injury

Jonathan H. Pelletier, Alicia K. Au, Mohammed A. Shaik, Robert S. B. Clark, Dennis W. Simon, Jaskaran Rakkar, Patrick Michael Kochanek, Gilles Clermont, Michael J. Bell, Christopher Horvat
preprint en

Abstract

Background/Objective : Existing therapy-intensity scales for traumatic brain injury (TBI) require manual abstraction and are scored at coarse intervals, limiting temporal alignment with intracranial pressure (ICP). We aimed to develop an hourly, electronic medical record (EMR)-derived Pediatric Intensity Level of Therapy (PILOT) scale (ePILOT) and characterize its relationships with ICP and in-hospital mortality. Methods : We retrospectively analyzed hourly EMR data from the first 7 days of hospitalization in children with TBI who underwent ICP monitoring at a single center. Twelve components adapted from PILOT (cerebrospinal fluid diversion, hyperosmolar therapy, sedation/analgesia, barbiturates, neuromuscular blockade, temperature, PaCO 2 , and serum sodium) were derived from routinely stored data. A Mixed Graphical Model descriptively characterized conditional dependencies among components and ICP. Integer component weights were informed by associations with in-hospital mortality and adjusted to preserve guideline tier ordering, yielding an hourly score of 0–24. Patient-level relationships with ICP and mortality were assessed. Results : Among 135 ICP-monitored TBI patients (median age 4.8 [1.9–12.7] years; 65% male; admission Glasgow Coma Scale (GCS) score 4 [3–7]; mortality 16%), ePILOT increased progressively with ICP (patient-level Spearman ρ = 0.32, p < 0.001). Non-survivors had higher mean ICP (26 vs. 12 mmHg) and median hourly ePILOT (7.0 [5.0–10.4] vs. 3.5 [3.0–6.5]) than survivors (both p < 0.05). Peak hourly ePILOT discriminated mortality with an area under the receiver operating characteristic curve (AUROC) of 0.74 (95% CI 0.61–0.85), compared with 0.68–0.72 for ICP summaries and 0.61 for admission GCS. A median hourly ePILOT score of 5 was the Youden-optimal mortality cut-point (95% CI 4–8; sensitivity 0.77; specificity 0.61). Conclusions : ePILOT converts routinely stored EMR data into an automated, hourly measure of ICP-directed therapy intensity. Therapy intensity tracked ICP, lagged behind it, and discriminated mortality comparably to ICP, consistent with a reactive treatment response. External validation against manually scored PILOT should follow.

Research Square
University of Pittsburgh (US), Akron Children's Hospital (US), Phoenix Children's Hospital (US), Children's Hospital of Pittsburgh (US), University of Pittsburgh Medical Center (US), The University of Texas Southwestern Medical Center (US)
National Institute of Child Health and Human Development, Eunice Kennedy Shriver National Institute of Child Health and Human Development
Good health and well-being
Traumatic Brain Injury and Neurovascular Disturbances
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