Respiratory failure & ventilation

An Electronic Health Record-Based Classifier for Moderate-to-Severe Acute Respiratory Distress Syndrome With Persistent Hypoxemia

Crit Care MedPublished Retrospective observational study

Authors' conclusion

An EHR-based classifier using readily available data had acceptable performance for identifying patients with persistent moderate-severe ARDS.

Study context

Population
Adults receiving ≥24 hours of invasive mechanical ventilation with PaO2/FiO2 ≤150 mm Hg on FiO2 ≥0.6 in the first 72 hours
Participants
924 vs 90
Exposure
Development vs Validation
  • Developmentn = 924
  • Validationn = 90
Outcome
Primary electronic health record (EHR)-based classifier
Methodological detailstwo health systems, six ICUs, temporally and geographically distinct development and validation cohorts
  • Two health systems
  • Six ICUs
  • Temporally and geographically distinct development and validation cohorts

Main finding

As reported in the abstract

GroupValuePPV
Development382 events71%; 95% CI, 66-75%
Validation38 events66%; 95% CI, 50-81%

Value

Development382 events
Validation38 events

PPV

Development71%; 95% CI, 66-75%
Validation66%; 95% CI, 50-81%

Observational data: associations, not effects.

Further findings

Clinician-adjudicated ARDS

Classifier using SpO2/FiO2 ≤162 as second hypoxemia marker

PPV

Classifier using SpO2/FiO2 ≤162 as second hypoxemia marker

More patients classified as acute respiratory distress syndrome (ARDS) in both cohorts but with lower PPVs

Parts of this Brief are withheld: the automatic checks could not confirm them against the abstract.

Original publication

An Electronic Health Record-Based Classifier for Moderate-to-Severe Acute Respiratory Distress Syndrome With Persistent Hypoxemia. Crit Care Med. Published 30 December 2025. doi:10.1097/ccm.0000000000007007

Composed by AI from the published abstract and checked against it. How Briefs are made →

Get them every week

New studies on this topic in My Week and in your Monday e-mail.

We e-mail you a sign-in link and a code. No password, no spam.