A machine learning-based risk score for prediction of infective endocarditis among patients with Staphylococcus aureus bacteraemia - The SABIER score

Christopher Koon-Chi Lai; Eman Leung; Yinan He; Ching-Chun Cheung; Mui Oi Yat Oliver; Qinze Yu; Timothy Chun-Man Li; Alfred Lok-Hang Lee; Li Yu; Grace Chung-Yan Lui

doi:10.1093/infdis/jiae080

A machine learning-based risk score for prediction of infective endocarditis among patients with Staphylococcus aureus bacteraemia - The SABIER score

J Infect Dis. 2024 Feb 29:jiae080. doi: 10.1093/infdis/jiae080. Online ahead of print.

Authors

Affiliations

¹ Department of Microbiology, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China.
² School of Public Health and Primary Care, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China.
³ Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China.
⁴ Department of Computer Science and Engineering, Faculty of Engineering, The Chinese University of Hong Kong, Hong Kong SAR, China.
⁵ Department of Medicine and Therapeutics, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China.
⁶ Department of Microbiology, Prince of Wales Hospital, Hospital Authority, Hong Kong, Hong Kong SAR, China.

PMID: 38420871
DOI: 10.1093/infdis/jiae080

Abstract

Background: Early risk assessment is needed to stratify Staphylococcus aureus infective endocarditis (SA-IE) risk among Staphylococcus aureus bacteraemia (SAB) patients to guide clinical management. The objective of this study is to develop a novel risk score independent of subjective clinical judgment and can be used early at the time of blood culture positivity.

Methods: We conducted a retrospective big data analysis from territory-wide electronic data and included hospitalized patients with SAB between 2009 and 2019. We applied a random forest risk scoring model to select variables from an array of parameters, according to the statistical importance of each feature in predicting SA-IE outcome. The data was divided into derivation and validation cohorts. The areas under the curve of the receiver operating characteristic (AUCROC) were determined.

Results: We identified 15,741 SAB patients, among them 4.18% had SA-IE. The AUCROC was 0.74 (95%CI 0.70-0.76), with a negative predictive value of 0.980 (95%CI 0.977-0.983). The four most discriminatory features were age, history of infective endocarditis, valvular heart disease, and being community-onset.

Conclusion: We developed a novel risk score with good performance as compared to existing scores and can be used at the time of SAB and prior to subjective clinical judgment.

Keywords: Staphylococcus aureus; Artificial intelligence; Bloodstream infections; Machine learning; Prediction Model; Sepsis; infective endocarditis.