HAC Analysis to Explore Clusters within Chronic Comorbid Inpatient Visits

TL;DR

This study uses hierarchical clustering to identify ten distinct, clinically relevant multimorbidity groups within Ohio's inpatient population, aiding healthcare analysis.

Contribution

It introduces an empirical HAC-based approach to explore and identify meaningful multimorbidity clusters in a specific geographic population.

Findings

Identified ten distinct multimorbidity clusters

Clusters are clinically pertinent and geographically specific

Provides a new exploratory method for healthcare data analysis

Abstract

Multimorbidities are associated with significant burden on the healthcare system and the lack of accurate and pertinent statistical exploratory techniques have often limited their analysis. Here we employ exploratory hierarchal agglomerative clustering (HAC) of multimorbidities in the inpatient population in the state of Ohio. The examination exposed the presence of ten discrete, clinically pertinent groups of multimorbidities within the Ohio inpatient population. This method offers an assessable empirical exploration of the multimorbidities present in a specific geographic populace.