COPD profiles and treatable traits using minimal resources: identification, decision tree and stability over time

Area of Science:

• Pulmonary Medicine
• Clinical Research
• Patient Stratification

Background:

• Chronic Obstructive Pulmonary Disease (COPD) patient profiles often lack validation and information on treatable traits and temporal stability.
• Existing COPD phenotyping lacks focus on actionable, treatable characteristics and their dynamic nature.

Purpose of the Study:

• To identify distinct COPD patient profiles and their associated treatable traits using simple, meaningful measures.
• To develop and validate a decision tree for classifying patients into these profiles.
• To assess the stability of these profiles over a six-month period.

Main Methods:

• Prospective observational study involving 352 COPD patients.
• Cross-sectional and longitudinal data collection including clinical characteristics, lung function (FEV1), symptoms (CAT), and quality of life.
• Principal component analysis and k-medoids clustering for profile identification; decision tree validation; six-month follow-up for stability analysis.

Main Results:

• Four distinct COPD profiles with identifiable treatable traits were identified.
• A validated decision tree utilizing CAT score, age, and FEV1 achieved 71.7% agreement with identified profiles.
• Patient profiles demonstrated moderate stability over six months, with 48.9% remaining in their initial profile.

Conclusions:

• Four COPD profiles with treatable traits were identified using accessible measures, suitable for low-resource settings.
• A practical decision tree is available for routine clinical use in COPD patient profiling.
• The dynamic nature of COPD profiles necessitates regular reassessment for effective, personalized treatment strategies.