Use of Artificial Intelligence in Early Warning Score in Critical ill Patients: Scoping Review

Abstract

Early Warning Score (EWS) systems can identify critical patients through the application of artificial intelligence (AI). Physiological parameters like blood pressure, body temperature, heart rate, and respiration rate are encompassed in the EWS. One of AI's advantages is its capacity to recognize high-risk individuals who need emergency medical attention because they are at risk of organ failure, heart attack, or even death. The objective of this study is to review the body of research on the use of AI in EWS to accurately predict patients who will become critical. The analysis model of Arksey and O'Malley is employed in this study. Electronic databases such as ScienceDirect, Scopus, PubMed, and SpringerLink were utilized in a methodical search. Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA SR) guidelines were utilized in the creation and selection of the literature. This analysis included a total of 14 articles. This article summarizes the findings on several aspects: the usefulness of AI algorithms in EWS for critical patients, types of AI algorithm models, and the accuracy of AI in predicting the quality of life of patients in EWS. The results of this review show that the integration of AI into EWS can increase accuracy in predicting patients in critical condition, including cardiac arrest, sepsis, and ARDS events that cause inhalation until the patient dies. The AI models that are often used are machine learning and deep learning models because they are considered to perform better and achieve high accuracy. The importance of further research is to identify the application of AI with EWS in critical care patients by adding laboratory result parameters and pain scales to increase prediction accuracy to obtain optimal results.