Determinants and Phenotypes of Poorly Controlled COPD Using the RADAR Score: A Cohort in Real-World Primary Care

Area of Science:

• Pulmonary Medicine
• Clinical Pharmacy
• Health Services Research

Background:

• Poor clinical control is common in Chronic Obstructive Pulmonary Disease (COPD) patients on maintenance inhaled therapy.
• The interplay between disease severity, modifiable factors, and clinician perception in COPD control is not well understood.
• Existing research lacks comprehensive characterization of poorly controlled COPD populations in primary care.

Purpose of the Study:

• To determine the frequency of poor clinical control in COPD patients receiving maintenance inhaled therapy.
• To identify independent determinants of poor COPD control, including disease severity and modifiable factors.
• To characterize the heterogeneity of poorly controlled COPD patients using cluster analysis.

Main Methods:

• A multicenter, cross-sectional study (Study SIMPLIFY) involving 988 COPD patients.
• Objective classification of COPD control using the RADAR score.
• Multivariable logistic regression and Machine Learning (Random Forest with SHAP) for determinant identification; k-medoids clustering for phenotype characterization.

Main Results:

• 45.7% of patients exhibited poor clinical control (RADAR score ≥ 4).
• Significant discrepancies noted between physician assessment and objective RADAR classification.
• Exacerbator phenotypes, active smoking, low adherence, high dosing frequency, and high inhaler burden were key determinants of poor control.

Conclusions:

• Poor COPD control is prevalent and under-recognized in primary care, linked to a clinician perception gap.
• Addressing modifiable factors like adherence, regimen complexity, and dosing frequency is essential for improving COPD management.
• Distinct clinical phenotypes identified offer a pathway toward personalized, evidence-based COPD care.