SHM Converge Scientific Abstract ePoster Gallery-A Novel Method to Improve the Identification of Time of Intubation for Retrospective Data Analysis.

Pdf Summary

Northwell Health describes itself as New York State’s largest health care provider and private employer, committed to improving community health through high-quality clinical care, education, biomedical research, health promotion, and care regardless of ability to pay. It operates 23 hospitals and more than 700 outpatient facilities, with over 13,600 affiliated physicians.

The document then presents a study titled <strong>“A Novel Method to Improve the Identification of Time of Intubation for Retrospective Data Analysis.”</strong> The study focuses on <strong>time of intubation (TOI)</strong>, an important marker of disease severity and a key point for analyzing patient outcomes before and after intubation in the electronic health record (EHR), especially in critically ill COVID-19 patients.

Researchers conducted a retrospective review of adults hospitalized with <strong>SARS-CoV-2</strong>, intubated with <strong>ARDS</strong>, across 12 hospitals between March 1 and April 30, 2020. They compared a prior search query based only on ventilator parameters (**Tv**) with a new, improved query (**Ti**) built from chart review and validated surrogates for TOI. These surrogates included procedure note timestamps, mechanical ventilation “start time,” first ventilation documentation, intubation medication administration, start of sedative infusions, and nursing documentation indicating ventilator use.

The study found that <strong>Ti identified TOI more accurately</strong> than Tv and improved analysis of peri-intubation patient data. Using Ti changed PaO2/FiO2-based ARDS severity classification for <strong>7.3% of patients</strong>, showing that better TOI identification can meaningfully affect research conclusions.

The authors conclude that their Ti logic can serve as a practical template for other health systems and may improve future retrospective analyses and machine learning models that predict intubation timing across emergency, ward, and ICU settings.

Asset Subtitle

Alex Makhnevich

Meta Tag

Author List Adey Tsegaye, Alex Makhnevich, Amir Gandomi, Daniel Jafari, Daniel Rolston, Michael Qiu, Negin Hajizadeh, Phil Wu

Category Research

Concept Time of Intubation

Concept Surrogate Data Element

Concept Intubation Query

Concept Electronic Health Record

Concept Respiratory Illness Severity

Distinguished Non-Finalist

Presenter Organization Northwell Health

Presenting Author Alex Makhnevich

Track Outcomes Research

Keywords

Northwell Health

time of intubation

retrospective data analysis

electronic health record

COVID-19

SARS-CoV-2

ARDS

mechanical ventilation

intubation timing

patient outcomes

Time of Intubation

Surrogate Data Element

Intubation Query

Electronic Health Record

Respiratory Illness Severity