See related Editorial

Introduction Clinical decisions in chronic obstructive pulmonary disease (COPD) treatment often utilize serially assessed physiologic parameters and biomarkers. To better understand the reliability of these tests, we evaluated changes in commonly assessed biomarkers over 3 months in patients with clinically stable COPD. Methods We performed an observational prospective cohort study of 89 individuals with clinically stable COPD, defined as no exacerbation history within 3 months of enrollment. Biomarkers included lung function and functional performance status, patient-reported outcomes of symptoms and health status, and blood markers of inflammation. The correlation between testing at baseline and at 3-month follow-up was reported as the intraclass correlation coefficient (ICC). “Outliers” had significant variability between tests, defined as >1.645 standard deviations between the two measurements. Differences in clinical features between outliers and others were compared. Results Participants with COPD (n = 89) were 70.5 ± 6.7 years old, 54 (61%) male, had a 40 pack-year smoking history with 24.7% being current smokers, and postbronchodilator forced expiratory volume in one second (FEV 1 ) 62.3 ± 22.7% predicted. The biomarkers with excellent agreement between the initial and the follow-up measurements were FEV 1 (ICC = 0.96), Saint George’s Respiratory Questionnaire (SGRQ) (ICC = 0.98), COPD Assessment Test (CAT) (ICC = 0.93) and C-reactive protein (CRP) (ICC = 0.90). By contrast, parameters showing less robust agreement were 6-minute walking distance (ICC = 0.75), eosinophil count (ICC = 0.77), erythrocyte sedimentation rate (ICC = 0.75) and white blood cell count (ICC = 0.48). Individuals with greater variability in biomarkers reported chronic bronchitis more often and had higher baseline SGRQ and CAT scores. Conclusion Our study evaluated the stability of commonly assessed biomarkers in clinically stable COPD and showed excellent agreement between baseline and three-month follow-up values for FEV 1 , SGRQ, CAT and CRP. Individuals with chronic bronchitis and more symptomatic disease at baseline demonstrated greater variability in 3-month interval biomarkers.