Hierarchical Clustering Analysis for Predicting 1-Year Mortality After Starting Hemodialysis

Abstract

Introduction: For patients with end-stage renal disease (ESRD), due to the heterogeneity of the population, appropriate risk assessment approaches and strategies for further follow-up remain scarce. We aimed to conduct a pilot study for better risk stratification, applying machine learning-based classification to patients with ESRD who newly started maintenance hemodialysis. Methods: We prospectively studied 101 patients with ESRD, who were new to maintenance hemodialysis therapy, between August 2016 and March 2018. Baseline values of variables such as blood and urine tests were obtained before the initiation of hemodialysis. Agglomerative hierarchical clustering was conducted with the collected continuous data. The resulting clusters were followed up for the primary outcome of 1year mortality, as analyzed by the Kaplan-Meier survival curve with log-rank test and the Cox proportional hazard model. Results: The participants were divided into 3 clusters (cluster 1, n ¼ 62; cluster 2, n ¼ 15; cluster 3, n ¼ 24) by hierarchical clustering, using 46 clinical variables. Patients in cluster 3 showed lower systolic blood pressures, and lower serum creatinine and urinary liver-type fatty acid-binding protein levels, before the initiation of hemodialysis. Consequently, cluster 3 was associated with the highest 1-year mortality in the study cohort (P < 0.001), and the difference was significant after adjustment for age and sex (hazard ratio: 10.2; 95% confidence interval: 2.94-46.8, cluster 1 as reference). Conclusion: In this proof-of-concept study, hierarchical clustering discovered a subgroup with a higher 1year mortality at the initiation of hemodialysis. Applying machine learning-derived classification to patients with ESRD may contribute to better risk stratification.

Exploring and Identifying Prognostic Phenotypes of Patients with Heart Failure Guided by Explainable Machine Learning

“…Because the phenotype of new patients in the validation dataset was unknown, the above evaluation indexes were inapplicable. In this case, Kaplan–Meier (KM) estimator and log-rank test were employed to confirm the survival curves of phenotypes [ 25 , 28 , 29 , 30 ] classified by the classifier on the validation dataset. Simultaneously, a univariate Cox proportional hazards (CPH) model was performed on each phenotype to explore the phenotype- or risk-level-specific significant risk and protective factors.…”

Rhythms Synchronization Effects on Cognition during Listening to Quranic Recitation

“…AHC is an unsupervised learning classification technique that is commonly used (Massart & Kaufman, 1983), which sequentially joins nodes of data with the shortest distance (Komaru et al, 2020), where the data were then stratified into clusters or groups with high homogeneity level within the classes and high heterogeneity level between classes concerning a predetermined selection criterion (McKenna, 2003). AHC used Ward's methods with Euclidean distances as a similarity measure (Otto 1998; Usman et al 2014;Azhar et al 2015).…”

Analisis Perbandingan Pengelompokan Indeks Pembangunan Manusia Indonesia Tahun 2019 dengan Metode Partitioning dan Hierarchical Clustering

“…Agglomerative Clustering melakukan pengelompokan hirarki dengan pendekatan bottom-up, yaitu setiap observasi dimulai di klusternya sendiri dan kluster secara berturut-turut digabungkan. Kriteria keterkaitan ward akan meminimalkan perbedaan jumlah kuadrat dalam setiap kluster [7] [8].…”