2026 Critical Care Congress Skip to main content
2026 Critical Care Congress Main banner
Final Program


Advance Program
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
Session Icons
Thought Leader Session
Thought Leader Session
Guidelines Session
Guidelines Session
Aging Adult Track
Aging Adult Track
Additional Registration Fee
Additional Registration Fee

Research Snapshot Theater: Infectious Disease, Adult II

Impact of an Information-Seeking Sepsis Model on Lactate Ordering

Sunday, March 22, 2026
11:00 AM - 12:00 PM Central Time
Location: Connections Central - RST 05

  • Woojung Kim, MS (he/him/his)

    California

    Disclosure(s): No relevant financial relationship(s) to disclose.

  • GW

    Gabriel Wardi, MD, MPH

    Associate Professor of Clinical, Emergency Medical Services
    UC San Diego Health

    Disclosure information not submitted.

  • AM

    Atul Malhotra, MD

    Professor
    University of California San Diego Thornton Hospital

    Disclosure information not submitted.

  • KS

    Karandeep Singh, MD, MMSc

    Associate Professor
    University of California San Diego, School of Medicine

    Disclosure information not submitted.

  • SN

    Shamim Nemati, PhD

    Associate Professor
    University of California San Diego, School of Medicine

    Disclosure information not submitted.

  • AB

    Aaron Boussina, PhD

    Assistant Adjunct Professor
    University of California San Diego, School of Medicine

    Disclosure information not submitted.

Introduction: Early identification and treatment of sepsis are critical for reducing mortality. To support these, predictive models aim to identify high risk patients using Electronic Health Record data but are limited by the availability of timely data and often determine risk based on delayed diagnostic tests that reflect preexisting clinician suspicion. To better support differential diagnosis, there is potential in models that can suggest data collection as part of a sequential decision-making process. We previously showed that our COMPOSER model led to a 17% relative reduction in in-hospital sepsis mortality. In this work, we modified the COMPOSER-based best practice advisory (BPA) to actively recommend lactate ordering when no lactate measurement is available. We hypothesized this would increase timely lactate ordering.


Methods: We conducted a prospective quasi-experimental study to evaluate whether BPA modification increased timely lactate ordering. On 2025-02-11, the system BPA was updated to recommend lactate for patients without a lactate result in the prior six hours. We included BPA alerts from 2024-01-01 to 2025-05-01 for patients without a lactate order, those with one >6 hours prior, or those ordered after the alert. The primary outcome was whether a lactate level was ordered within six hours of the BPA. Multivariable logistic regression was used, adjusting for intervention period, acknowledgment reason, gender, age, Elixhauser comorbidity scores, and baseline sequential organ failure assessment (SOFA).


Results: Among 4,328 alerts, post-intervention BPAs were associated with a 17.9% greater likelihood of timely lactate ordering (95% CI; 0.9%-37.7%). Users who selected ‘Sepsis Treatment/Workup in Progress’ or ‘Will Notify MD Immediately’ were 48.8% (8.2%-104.4%) and 42.5% (4.2%-94.8%) more likely, respectively, to order lactate promptly compared to those who gave no acknowledgment. Higher baseline SOFA and higher Elixhauser were associated with 4.5% lower likelihood (0.1%-8.7%), and 0.9% greater likelihood (0.1%-1.7%), respectively.


Conclusions: These findings suggest that the updated BPA system may have influenced user behavior toward more timely lactate ordering, offering a promising direction for refining sequential decision support systems.