Clinical phenotypes of chronic cough categorised by cluster analysis

Area of Science:

• Pulmonology
• Medical Diagnostics
• Data Science in Medicine

Background:

• Chronic cough is a complex condition with diverse and often unclear causes.
• Accurate phenotyping is crucial for effective management of chronic cough.
• Current diagnostic approaches may not fully capture the heterogeneity of chronic cough.

Purpose of the Study:

• To categorize the distinct phenotypes of chronic cough in adult patients.
• To apply cluster analysis for identifying homogenous subgroups within chronic cough patients.
• To establish a basis for stratified medical approaches in chronic cough management.

Main Methods:

• Adult patients with chronic cough were assessed using the COugh Assessment Test (COAT) and the Korean version of the Leicester Cough Questionnaire.
• A K-prototype cluster analysis was employed to group patients based on cough characteristics and severity.
• Correlation network analysis was used to determine the most relevant variables for clustering.

Main Results:

• Four distinct chronic cough phenotypes were identified among 255 participants.
• Phenotypes were characterized by age, cough severity (mild to severe), and gender predominance (male or female).
• Identified clusters included: elderly with mild cough, middle-aged with less severe cough, young males with severe cough, and elderly females with severe cough.

Conclusions:

• Cluster analysis successfully delineated four unique chronic cough phenotypes with differing etiologies.
• Subgrouping chronic cough patients into distinct phenotypes facilitates individualized diagnostic and therapeutic strategies.
• This phenotyping approach supports a more stratified and personalized medical approach to chronic cough.