Phenotyping Clusters of Patient Trajectories suffering from Chronic Complex Disease

TL;DR

This paper evaluates clustering methods for patient trajectories with chronic disease, proposing novel modifications to handle uneven sampling and class imbalance, aiming to improve clinically relevant phenotyping.

Contribution

It introduces new modifications to clustering models for better phenotyping of patient trajectories with chronic disease, addressing data sampling and class imbalance issues.

Findings

Improved cluster separation with proposed modifications

Enhanced interpretability of patient phenotypes

Potential for better personalized treatment strategies

Abstract

Recent years have seen an increased focus into the tasks of predicting hospital inpatient risk of deterioration and trajectory evolution due to the availability of electronic patient data. A common approach to these problems involves clustering patients time-series information such as vital sign observations) to determine dissimilar subgroups of the patient population. Most clustering methods assume time-invariance of vital-signs and are unable to provide interpretability in clusters that is clinically relevant, for instance, event or outcome information. In this work, we evaluate three different clustering models on a large hospital dataset of vital-sign observations from patients suffering from Chronic Obstructive Pulmonary Disease. We further propose novel modifications to deal with unevenly sampled time-series data and unbalanced class distribution to improve phenotype separation. Lastly, we discuss further avenues of investigation for models to learn patient subgroups with distinct behaviour and phenotype.