Related Experiment Video
Updated: Feb 18, 2026
11:02
Identification and Quantification of Deranged Metabolites in Critically Ill Patients Using NMR-Based Metabolomics
Published on: November 29, 2024
1.3K
Association between borderline dysnatremia and mortality insight into a new data mining approach
Yannick Girardeau1,2,3, Anne-Sophie Jannot4,5, Gilles Chatellier4,6,7
1Biomedical Informatics and Public Health Department, Hôpital Européen G. Pompidou, Assistance Publique-Hôpitaux de Paris, Paris, France. yannick.girardeau@aphp.fr.
BMC Medical Informatics and Decision Making
|November 24, 2017
Summary
Even borderline serum sodium levels at hospital admission significantly increase in-hospital mortality risk. This study confirms the association using electronic health records and phenome-wide association analysis (PheWAS).
Area of Science:
- Clinical Medicine
- Medical Informatics
- Biostatistics
Background:
- Serum sodium concentration variations are linked to patient mortality.
- Investigating borderline dysnatremia's impact on in-hospital mortality is crucial.
- Electronic health records (EHR) offer valuable real-world data for such studies.
Purpose of the Study:
- To confirm the association between borderline dysnatremia and in-hospital mortality.
- To utilize real-world data from EHR and phenome-wide association analysis (PheWAS).
- To identify potential confounding factors influencing dysnatremia and mortality.
Main Methods:
- Retrospective observational study of 45,834 patients admitted between 2008 and 2014.
- Analysis of serum sodium levels on admission and in-hospital mortality.
- Multivariate logistic regression and PheWAS to adjust for confounders.
Main Results:
- 12.0% of patients had hyponatremia and 1.0% had hypernatremia.
- Borderline hyponatremia (ORa=1.57) and hypernatremia (ORa=3.47) were independently associated with mortality.
- PheWAS identified new confounding factors associated with mortality and dysnatremia.
Conclusions:
- Borderline dysnatremia on hospital admission independently elevates in-hospital mortality risk.
- EHR data and novel data mining approaches can uncover significant clinical associations.
- Identifying and adjusting for confounding factors is key to understanding dysnatremia's true impact.
