Abstract: PO0183
Clinically Distinct Subtypes of AKI on Hospital Admission Identified by Machine Learning Consensus Clustering
Session Information
- AKI: Epidemiology, Risk Factors, and Prevention
November 04, 2021 | Location: On-Demand, Virtual Only
Abstract Time: 10:00 AM - 12:00 PM
Category: Acute Kidney Injury
- 101 AKI: Epidemiology, Risk Factors, and Prevention
Authors
- Mao, Michael A., Mayo Clinic's Campus in Florida, Jacksonville, Florida, United States
- Thongprayoon, Charat, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Zabala-Genovez, Jose Luis, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Kashani, Kianoush, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Dillon, John J., Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Garovic, Vesna D., Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Cheungpasitporn, Wisit, Mayo Clinic Minnesota, Rochester, Minnesota, United States
Background
Patients with acute kidney injury were clustered at hospital admission into clinically distinct subtypes using an unsupervised machine learning approach. Mortality risk was assessed among these distinct clusters.
Methods
Consensus clustering analysis was performed on demographics, principal diagnoses, comorbidities, and laboratory data on 4,289 hospitalized adult patients with acute kidney injury at admission. The standardized difference of each variable was calculated to identify each cluster’s key features. We assessed the association of each cluster with hospital and one-year mortality.
Results
Consensus clustering analysis identified four distinct clusters. There were 1,201 (28%) patients in cluster 1, 1,396 (33%) patients in cluster 2, 1,191 (28%) patients in cluster 3, and 501 (12%) patients in cluster 4. Figure 1 illustrates cluster differences. Figure 2 highlights associated increased mortality in clusters 2, 3, and 4.
Conclusion
Our study demonstrated the use of machine learning consensus clustering analysis to characterize a heterogenous cohort of patients with acute kidney injury on hospital admission into four clinically distinct clusters with different associated mortality risks.
Figure 1: Standardized differences across clusters
Figure 2: Hospital and One-Year Mortality among Different Clusters with Admission AKI