What do we mean when we say we are clustering multimorbidity?

2025-05-04

AI Summary

In multimorbidity clustering research, ambiguity surrounding core concepts—such as “clustering units” (diseases vs. individuals) and “clustering objectives” (clinical phenotyping vs. public health intervention)—leads to ad hoc method selection and poor comparability and applicability of results. To address this, we propose a novel paradigm: *downstream analytical purpose–driven clustering design*. We introduce a four-dimensional mapping framework linking *objective → data representation → distance metric → evaluation criterion*. Through conceptual analysis and methodological reflection, we systematically delineate six prototypical clustering purposes and their corresponding methodological requirements. Our framework is model-agnostic—neither prescribing specific statistical nor machine learning models—yet explicitly integrates epidemiological reasoning with clustering principles. It enhances transparency, reproducibility, and practical utility of multimorbidity clustering studies, thereby providing a rigorous methodological foundation for evidence-informed multimorbidity classification guidelines. (149 words)

Technology Category

• Machine Learning: Clustering
• Data Mining & Knowledge Management: Mining of Visual, Multimedia & Multimodal Data
• Search and Optimization: Metareasoning and Metaheuristics

Application Category

• Web Mining and Content Analysis: Normalization, clustering, classification, and summarization of Web text
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Clustering multimorbidity has been a global research priority in recent years. Existing studies usually identify these clusters using one of several popular clustering methods and then explore various characteristics of these clusters, e.g., their genetic underpinning or their sociodemographic drivers, as downstream analysis. These studies make several choices during clustering that are often not explicitly acknowledged in the literature, e.g., whether they are clustering conditions or clustering individuals, and thus, they lead to different clustering solutions. We observe that, in general, clustering multimorbidity might mean different things in different studies, and argue that making these choices more explicit and, more importantly, letting the downstream analysis, or the purpose of identifying multimorbidity clusters, guide these choices, might lead to more transparent and operationalizable multimorbidity clusters. In this study, we discuss various purposes of identifying multimorbidity clusters and build a case for how different purposes can justify the different choices in data and methods.

Problem

• Clarifying ambiguous definitions in multimorbidity clustering studies
• Linking clustering choices to downstream analysis purposes
• Improving transparency in multimorbidity cluster operationalization

Innovation

• Clustering choices guided by downstream analysis
• Explicitly acknowledging clustering methodology differences
• Purpose-driven data and method selection