Symptomatic chronic obstructive pulmonary disease in clinical trials and in a population-based study

Purpose: Evidence-based medicine promotes the current best evidence from clinical trials to guide decisions for individual patients. We assessed whether chronic obstructive pulmonary disease (COPD) patients included in exercise training studies and pharmacologic trials match those from a non-selected COPD target population sample. Methods: Exercise training studies were identified in a literature search. Towards a Revolution in COPD Health (TORCH) and Understanding Potential Long-Term Impacts on Function with Tiotropium (UPLIFT) were chosen to represent pharmacologic trials. Burden of Obstructive Lung Disease (BOLD) data were used to characterize target COPD population (BOLD target), defined as the presence of dyspnea (modified Medical Research Council ≥2) and non-reversible airway obstruction (post-bronchodilator FEV1/FVC ≤0.7 and FEV1% predicted ≤70 %). Results: Overall 240 exercise training studies with 13,901, TORCH and UPLIFT with 12,105, and BOLD with 16,218 participants were evaluated. Males were overrepresented in exercise training studies (67.5 %) and pharmacologic trials (TORCH 75.8 %; UPLIFT 74.6 %), whereas in BOLD target 55.8 % were males (p < 0.001). In exercise training studies, 7.2 % were never-smokers, 0.0 % in TORCH and UPLIFT, but 36.0 % in BOLD target (p < 0.001). Subjects with cardiac comorbidity were excluded from 75.4 % of exercise training studies, entirely from TORCH and UPLIFT, but comprised 24.5 % of BOLD target. Conclusions: COPD patients recruited in exercise training studies and in pharmacologic trials differ from target population of symptomatic COPD. Females, never-smokers, and patients with cardiac comorbidities are more likely excluded from the clinical trials.