Machine Learning Consensus Clustering of Hospitalized Patients with Admission Hyponatremia

Charat Thongprayoon; Panupong Hansrivijit; Michael A. Mao; Pradeep K. Vaitla; Andrea G. Kattah; Pattharawin Pattharanitima; Saraschandra Vallabhajosyula; Voravech Nissaisorakarn; Tananchai Petnak; Mira T. Keddis; Stephen B. Erickson; John J. Dillon; Vesna D. Garovic; Wisit Cheungpasitporn

doi:10.3390/diseases9030054

Machine Learning Consensus Clustering of Hospitalized Patients with Admission Hyponatremia

Charat Thongprayoon, Panupong Hansrivijit, Michael A. Mao, Pradeep K. Vaitla, Andrea G. Kattah, Pattharawin Pattharanitima, Saraschandra Vallabhajosyula, Voravech Nissaisorakarn, Tananchai Petnak, Mira T. Keddis, Stephen B. Erickson, John J. Dillon, Vesna D. Garovic, Wisit Cheungpasitporn

Research output: Contribution to journal › Article › peer-review

9 Scopus citations

Abstract

Background: The objective of this study was to characterize patients with hyponatremia at hospital admission into clusters using an unsupervised machine learning approach, and to evaluate the short- and long-term mortality risk among these distinct clusters. Methods: We performed consensus cluster analysis based on demographic information, principal diagnoses, comorbidities, and laboratory data among 11,099 hospitalized adult hyponatremia patients with an admission serum sodium below 135 mEq/L. The standardized mean difference was utilized to identify each cluster’s key features. We assessed the association of each hyponatremia cluster with hospital and one-year mortality using logistic and Cox proportional hazard analysis, respectively. Results: There were three distinct clusters of hyponatremia patients: 2033 (18%) in cluster 1, 3064 (28%) in cluster 2, and 6002 (54%) in cluster 3. Among these three distinct clusters, clusters 3 patients were the youngest, had lowest comorbidity burden, and highest kidney function. Cluster 1 patients were more likely to be admitted for genitourinary disease, and have diabetes and end-stage kidney disease. Cluster 1 patients had the lowest kidney function, serum bicarbonate, and hemoglobin, but highest serum potassium and prevalence of acute kidney injury. In contrast, cluster 2 patients were the oldest and were more likely to be admitted for respiratory disease, have coronary artery disease, congestive heart failure, stroke, and chronic obstructive pulmonary disease. Cluster 2 patients had lowest serum sodium and serum chloride, but highest serum bicarbonate. Cluster 1 patients had the highest hospital mortality and one-year mortality, followed by cluster 2 and cluster 3, respectively. Conclusion: We identified three clinically distinct phenotypes with differing mortality risks in a heterogeneous cohort of hospitalized hyponatremic patients using an unsupervised machine learning approach.

Original language	English
Article number	54
Journal	Diseases
Volume	9
Issue number	3
DOIs	https://doi.org/10.3390/diseases9030054
State	Published - Sep 2021
Externally published	Yes

Keywords

artificial intelligence
clustering
electrolytes
hospitalization
hyponatremia
machine learning
mortality
sodium

Access to Document

10.3390/diseases9030054

Cite this

Thongprayoon, C., Hansrivijit, P., Mao, M. A., Vaitla, P. K., Kattah, A. G., Pattharanitima, P., Vallabhajosyula, S., Nissaisorakarn, V., Petnak, T., Keddis, M. T., Erickson, S. B., Dillon, J. J., Garovic, V. D., & Cheungpasitporn, W. (2021). Machine Learning Consensus Clustering of Hospitalized Patients with Admission Hyponatremia. Diseases, 9(3), Article 54. https://doi.org/10.3390/diseases9030054

@article{8ff89bc3e01b41d2b7bad7ecbe80ab65,

title = "Machine Learning Consensus Clustering of Hospitalized Patients with Admission Hyponatremia",

abstract = "Background: The objective of this study was to characterize patients with hyponatremia at hospital admission into clusters using an unsupervised machine learning approach, and to evaluate the short- and long-term mortality risk among these distinct clusters. Methods: We performed consensus cluster analysis based on demographic information, principal diagnoses, comorbidities, and laboratory data among 11,099 hospitalized adult hyponatremia patients with an admission serum sodium below 135 mEq/L. The standardized mean difference was utilized to identify each cluster{\textquoteright}s key features. We assessed the association of each hyponatremia cluster with hospital and one-year mortality using logistic and Cox proportional hazard analysis, respectively. Results: There were three distinct clusters of hyponatremia patients: 2033 (18\%) in cluster 1, 3064 (28\%) in cluster 2, and 6002 (54\%) in cluster 3. Among these three distinct clusters, clusters 3 patients were the youngest, had lowest comorbidity burden, and highest kidney function. Cluster 1 patients were more likely to be admitted for genitourinary disease, and have diabetes and end-stage kidney disease. Cluster 1 patients had the lowest kidney function, serum bicarbonate, and hemoglobin, but highest serum potassium and prevalence of acute kidney injury. In contrast, cluster 2 patients were the oldest and were more likely to be admitted for respiratory disease, have coronary artery disease, congestive heart failure, stroke, and chronic obstructive pulmonary disease. Cluster 2 patients had lowest serum sodium and serum chloride, but highest serum bicarbonate. Cluster 1 patients had the highest hospital mortality and one-year mortality, followed by cluster 2 and cluster 3, respectively. Conclusion: We identified three clinically distinct phenotypes with differing mortality risks in a heterogeneous cohort of hospitalized hyponatremic patients using an unsupervised machine learning approach.",

keywords = "artificial intelligence, clustering, electrolytes, hospitalization, hyponatremia, machine learning, mortality, sodium",

author = "Charat Thongprayoon and Panupong Hansrivijit and Mao, \{Michael A.\} and Vaitla, \{Pradeep K.\} and Kattah, \{Andrea G.\} and Pattharawin Pattharanitima and Saraschandra Vallabhajosyula and Voravech Nissaisorakarn and Tananchai Petnak and Keddis, \{Mira T.\} and Erickson, \{Stephen B.\} and Dillon, \{John J.\} and Garovic, \{Vesna D.\} and Wisit Cheungpasitporn",

note = "Publisher Copyright: {\textcopyright} 2021 by the authors.",

year = "2021",

month = sep,

doi = "10.3390/diseases9030054",

language = "English",

volume = "9",

journal = "Diseases",

issn = "2079-9721",

publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",

number = "3",

}

Thongprayoon, C, Hansrivijit, P, Mao, MA, Vaitla, PK, Kattah, AG, Pattharanitima, P, Vallabhajosyula, S, Nissaisorakarn, V, Petnak, T, Keddis, MT, Erickson, SB, Dillon, JJ, Garovic, VD & Cheungpasitporn, W 2021, 'Machine Learning Consensus Clustering of Hospitalized Patients with Admission Hyponatremia', Diseases, vol. 9, no. 3, 54. https://doi.org/10.3390/diseases9030054

TY - JOUR

T1 - Machine Learning Consensus Clustering of Hospitalized Patients with Admission Hyponatremia

AU - Thongprayoon, Charat

AU - Hansrivijit, Panupong

AU - Mao, Michael A.

AU - Vaitla, Pradeep K.

AU - Kattah, Andrea G.

AU - Pattharanitima, Pattharawin

AU - Vallabhajosyula, Saraschandra

AU - Nissaisorakarn, Voravech

AU - Petnak, Tananchai

AU - Keddis, Mira T.

AU - Erickson, Stephen B.

AU - Dillon, John J.

AU - Garovic, Vesna D.

AU - Cheungpasitporn, Wisit

PY - 2021/9

Y1 - 2021/9

N2 - Background: The objective of this study was to characterize patients with hyponatremia at hospital admission into clusters using an unsupervised machine learning approach, and to evaluate the short- and long-term mortality risk among these distinct clusters. Methods: We performed consensus cluster analysis based on demographic information, principal diagnoses, comorbidities, and laboratory data among 11,099 hospitalized adult hyponatremia patients with an admission serum sodium below 135 mEq/L. The standardized mean difference was utilized to identify each cluster’s key features. We assessed the association of each hyponatremia cluster with hospital and one-year mortality using logistic and Cox proportional hazard analysis, respectively. Results: There were three distinct clusters of hyponatremia patients: 2033 (18%) in cluster 1, 3064 (28%) in cluster 2, and 6002 (54%) in cluster 3. Among these three distinct clusters, clusters 3 patients were the youngest, had lowest comorbidity burden, and highest kidney function. Cluster 1 patients were more likely to be admitted for genitourinary disease, and have diabetes and end-stage kidney disease. Cluster 1 patients had the lowest kidney function, serum bicarbonate, and hemoglobin, but highest serum potassium and prevalence of acute kidney injury. In contrast, cluster 2 patients were the oldest and were more likely to be admitted for respiratory disease, have coronary artery disease, congestive heart failure, stroke, and chronic obstructive pulmonary disease. Cluster 2 patients had lowest serum sodium and serum chloride, but highest serum bicarbonate. Cluster 1 patients had the highest hospital mortality and one-year mortality, followed by cluster 2 and cluster 3, respectively. Conclusion: We identified three clinically distinct phenotypes with differing mortality risks in a heterogeneous cohort of hospitalized hyponatremic patients using an unsupervised machine learning approach.

AB - Background: The objective of this study was to characterize patients with hyponatremia at hospital admission into clusters using an unsupervised machine learning approach, and to evaluate the short- and long-term mortality risk among these distinct clusters. Methods: We performed consensus cluster analysis based on demographic information, principal diagnoses, comorbidities, and laboratory data among 11,099 hospitalized adult hyponatremia patients with an admission serum sodium below 135 mEq/L. The standardized mean difference was utilized to identify each cluster’s key features. We assessed the association of each hyponatremia cluster with hospital and one-year mortality using logistic and Cox proportional hazard analysis, respectively. Results: There were three distinct clusters of hyponatremia patients: 2033 (18%) in cluster 1, 3064 (28%) in cluster 2, and 6002 (54%) in cluster 3. Among these three distinct clusters, clusters 3 patients were the youngest, had lowest comorbidity burden, and highest kidney function. Cluster 1 patients were more likely to be admitted for genitourinary disease, and have diabetes and end-stage kidney disease. Cluster 1 patients had the lowest kidney function, serum bicarbonate, and hemoglobin, but highest serum potassium and prevalence of acute kidney injury. In contrast, cluster 2 patients were the oldest and were more likely to be admitted for respiratory disease, have coronary artery disease, congestive heart failure, stroke, and chronic obstructive pulmonary disease. Cluster 2 patients had lowest serum sodium and serum chloride, but highest serum bicarbonate. Cluster 1 patients had the highest hospital mortality and one-year mortality, followed by cluster 2 and cluster 3, respectively. Conclusion: We identified three clinically distinct phenotypes with differing mortality risks in a heterogeneous cohort of hospitalized hyponatremic patients using an unsupervised machine learning approach.

KW - artificial intelligence

KW - clustering

KW - electrolytes

KW - hospitalization

KW - hyponatremia

KW - machine learning

KW - mortality

KW - sodium

UR - https://www.scopus.com/pages/publications/85115775715

U2 - 10.3390/diseases9030054

DO - 10.3390/diseases9030054

M3 - Article

AN - SCOPUS:85115775715

SN - 2079-9721

VL - 9

JO - Diseases

JF - Diseases

IS - 3

M1 - 54

ER -

Machine Learning Consensus Clustering of Hospitalized Patients with Admission Hyponatremia

Abstract

Keywords

Access to Document

Fingerprint

Cite this