Biomarker-Based Assessment Model for Detecting Sepsis: A Retrospective Cohort Study

Area of Science:

• Critical Care Medicine
• Biomarker Discovery
• Sepsis Pathophysiology

Background:

• The quick sequential organ failure assessment (qSOFA) and SOFA scores are used for sepsis detection but may not fully capture infection suspicion.
• There is a need for improved, simple, and rapid sepsis detection tools in intensive care units (ICUs).

Purpose of the Study:

• To develop and validate a novel biomarker-based assessment model for detecting sepsis (BADS).
• To compare the predictive performance of BADS against existing sepsis scoring systems like qSOFA, SOFA, and APACHE II.

Main Methods:

• Retrospective analysis of electronic health records from 989 ICU patients in a South Korean university hospital.
• Development of the BADS model using logistic regression, incorporating variables such as mean arterial pressure, shock index, lactate, and procalcitonin.
• Model validation using a training and validation set (80:20 ratio).

Main Results:

• The developed BADS model demonstrated an area under the curve (AUC) of 0.754 for sepsis prediction.
• BADS performance was comparable to the SOFA score (AUC 0.763) and superior to qSOFA (AUC 0.615) and APACHE II (AUC 0.668).
• The model identified mean arterial pressure, shock index, lactate, and procalcitonin as key predictive variables.

Conclusions:

• The biomarker-based assessment model for detecting sepsis (BADS) is a simple and effective tool for rapid sepsis detection in critically ill patients.
• BADS shows non-inferiority to the SOFA score in predicting sepsis, suggesting its potential utility at the bedside.
• Further validation in diverse clinical settings is warranted to confirm BADS's broad applicability.