Relationship between In-Hospital Sepsis Prediction Score and Prevalence of Community-Onset Sepsis: Triage for Sepsis Risk Management

Authors

  • Kyung Hyun LEE AITRICS. Inc, 218 Teheran-ro, Gangnam-gu, 06221 Seoul, Republic of Korea
  • Hyunwoo CHOO AITRICS. Inc, 218 Teheran-ro, Gangnam-gu, 06221 Seoul, Republic of Korea
  • Sungsoo HONG AITRICS. Inc, 218 Teheran-ro, Gangnam-gu, 06221 Seoul, Republic of Korea
  • Sungjun HONG Medical AI Research Center, Research Institute for Future Medicine, Samsung Medical Center, 81 Irwon-ro, Gangnam-gu, S06351 eoul, Republic of Korea
  • Ki-Byung LEE Division of Pulmonary, Allergy and Critical Care Medicine, Hallym University Chuncheon Sacred Heart Hospital, 77 Sakju-ro, 24253 Chuncheon, Republic of Korea
  • Hochan CHO Department of Internal Medicine, Keimyung University Dongsan Hospital, Keimyung University School of Medicine, 42601 Daegu, Republic of Korea

Keywords:

Sepsis, Prediction, Deep learning, Risk stratification, Regression analysis

Abstract

Early diagnosis of sepsis is crucial in clinical practice. Several studies have proposed sepsis prediction models to forecast the onset of sepsis in hospitals. However, validation of prediction models for community-onset sepsis, which is sepsis developed before admission to the hospital, is insufficient. This study investigates the relationship between the in-hospital prediction model scores and community-onset sepsis. We used hierarchical logistic regression analysis to explore the relationship between sepsis prevalence and AITRICS-VC SEPS tertile categories while adjusting for potential confounders. The low-SEPS group was used as the reference group. The odds ratio (ORs) of sepsis comparing the moderate versus low SEPS group are 1.198 (95%, 1.075-3.654), and the high versus low VC-SEPS group are 8.683 (95%, 4.995-15.095). Even though the sepsis prediction model was designed to predict in-hospital sepsis, high prediction scores are related to the prevalence of community-onset sepsis. This result implies that SEPS scores can stratify sepsis risks and be considered a patient assessment tool for triage.

Downloads

Published

21.11.2024

How to Cite

1.
LEE KH, CHOO H, HONG S, HONG S, LEE K-B, CHO H. Relationship between In-Hospital Sepsis Prediction Score and Prevalence of Community-Onset Sepsis: Triage for Sepsis Risk Management. Appl Med Inform [Internet]. 2024 Nov. 21 [cited 2024 Dec. 3];46(Suppl. 2):S13-S16. Available from: https://ami.info.umfcluj.ro/index.php/AMI/article/view/1076