Skip to main content

ORIGINAL RESEARCH article

Front. Cardiovasc. Med.
Sec. Intensive Care Cardiovascular Medicine
Volume 11 - 2024 | doi: 10.3389/fcvm.2024.1508766
This article is part of the Research Topic Critical Care Cardiology for Cardiovascular Emergencies View all articles

Extension of an ICU-based noninvasive model to predict latent shock in the emergency department: An exploratory study

Provisionally accepted
Mingzheng Wu Mingzheng Wu Shaoping Li Shaoping Li Haibo Yu Haibo Yu Cheng JIANG Cheng JIANG Dai Shuai Dai Shuai Jiang Shan Jiang Shan Yan Zhao Yan Zhao *
  • Zhongnan Hospital, Wuhan University, Wuhan, China

The final, formatted version of the article will be published soon.

    Background: Artificial intelligence (AI) has been widely adopted to predict the occurrence of latent shock in critical ills in the intensive care unit (ICU). However, usefulness of ICU-based model for predict latent shock risk in emergency department (ED) remains unclear. This study was aimed to develop an AI model to predict the latent shock risk in patients admitted to ED. Methods: Multiple regression analysis was used to compare the difference between Medical Information Mart for Intensive Care (MIMIC)-IV-ICU and MIMIC-IV-ED data sets. An adult noninvasive model was constructed based on MIMIC-IV-ICU v3.0 database, and externally validated in populations admitted to ED. Its efficiency was compared with the noninvasive systolic blood pressure (nSBP) and shock index. Results: A total of 50,636 patients of MIMIC-IV-ICU database was used for developing the model, and a total of 2,142 patients of the Philips IntelliSpace Critical Care and Anesthesia (ICCA)-ED and 425,087 patients of MIMIC-IV-ED was used for external validation. The modeling and validation data revealed similar non-invasive feature distributions. Multiple regression analysis of MIMIC-IV-ICU and MIMIC-IV-ED datasets showed mostly similar characteristics. Area under the receiver operating characteristic curve (AUROC) of noninvasive model on 10 minutes before the intervention was 0.90 (95% CI 0.84–0.96), and the diagnose accordance rate(DAR) is above 80%. More than 80% of latent shock patients could be identified more than 70 minutes earlier. Noninvasive model performed better than shock index and nSBP. Conclusion: The adult noninvasive model can effectively predict the latent shock occurrence in ED, which is better than shock index and nSBP.

    Keywords: Shock, emergency department, Intensive Care Unit, artificial intelligence, Model

    Received: 09 Oct 2024; Accepted: 13 Nov 2024.

    Copyright: © 2024 Wu, Li, Yu, JIANG, Shuai, Shan and Zhao. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

    * Correspondence: Yan Zhao, Zhongnan Hospital, Wuhan University, Wuhan, China

    Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.