Interpretable Machine-Learning Model for Real-Time, Clustered Risk Factor Analysis of Sepsis and Septic Death in Critical Care

Jiang, Zhengyu; Bo, Lulong; Xie, Yuan; Wang, Lei; Cao, Jianping; Ying, Yao; Lu, Wenbin; Deng, Xiaoming; Yang, Tao; Bian, Jinjun

doi:10.2139/ssrn.4260034

Cited by 1 publication

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We identified 12 blood metabolites that were causally related to sepsis and these metabolites may influence sepsis development through pathways such as glycolysis/gluconeogenesis, and pyruvate metabolism. Among the 12 blood metabolites, in addition to unknown metabolites and reported glucose 42 and propionylcarnitine 43 as risk factors for sepsis, we also discovered that elevated levels of salicylate, metoprolol acid metabolite, and androsterone sulfate were associated with an increased risk of sepsis. While, elevated levels of heptanoate (7:0), 1-oleoylglycerophosphoeth-anolamine, alpha-glutamyltyrosine, and saccharin could reduce the risk of sepsis.…”

Section: Discussionmentioning

confidence: 83%

“…Through MR research, we establish causal relationships between 9 metabolites and sepsis, of which glucose has been intensively studied in sepsis. Recently, Jiang et al established a predictive model for 47,185 septic patients in a retrospective observational cohort study, and they identified glucose levels as a significant risk factor for sepsis 42 . We further added genetic evidence for the role of glucose in sepsis.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Mendelian randomization analysis reveals causal associations of serum metabolites with sepsis and 28-day mortality

Jing,

Zuo,

Liu

et al. 2024

Sci Rep

View full text Add to dashboard Cite

Metabolic disorder has been found to be an important factor in the pathogenesis and progression of sepsis. However, the causation of such an association between serum metabolites and sepsis has not been established. We conducted a two-sample Mendelian randomization (MR) study. A genome-wide association study of 486 human serum metabolites was used as the exposure, whereas sepsis and sepsis mortality within 28 days were set as the outcomes. In MR analysis, 6 serum metabolites were identified to be associated with an increased risk of sepsis, and 6 serum metabolites were found to be related to a reduced risk of sepsis. Furthermore, there were 9 metabolites positively associated with sepsis-related mortality, and 8 metabolites were negatively correlated with sepsis mortality. In addition, “glycolysis/gluconeogenesis” (p = 0.001), and “pyruvate metabolism” (p = 0.042) two metabolic pathways were associated with the incidence of sepsis. This MR study suggested that serum metabolites played significant roles in the pathogenesis of sepsis, which may provide helpful biomarkers for early disease diagnosis, therapeutic interventions, and prognostic assessments for sepsis.

show abstract

Section: Discussionmentioning

confidence: 83%

Section: Discussionmentioning

confidence: 99%