BMJ Open Respir Res. 2025 Oct 10;12(1):e003615. doi: 10.1136/bmjresp-2025-003615.
ABSTRACT
BACKGROUND: Chronic obstructive pulmonary disease (COPD) patients are encouraged to manage exacerbations (acute exacerbation of COPD (AECOPD)) through self-management (SM) plans. Since only around half of AECOPD are bacterial, and sputum colour correlates with bacterial load, it may help guide antibiotic use. This pragmatic randomised controlled trial (RCT) assessed the safety and effectiveness of using a sputum colour chart in UK primary care.
METHODS: The multicentre RCT, Colour COPD randomised COPD adults who had ≥2 AECOPD or ≥1 AECOPD hospital admission in the preceding year. The primary objective was to assess the non-inferiority of the Bronkotest sputum colour chart compared with usual care, with hospital admission for AECOPD at 12 months as the primary outcome. Secondary outcomes included second courses of treatment requirement and quality of life (CAT score). Nested substudies examined daily symptoms via e-diaries and sputum culture.
RESULTS: 115 severe COPD patients (global obstructive lung disease(GOLD) D, 54% Medical Research Council (MRC) 4 or 5, CAT score 24) were randomised. A trend towards more hospital admissions (32% vs 16%, relative risk (RR) 1.95 (0.92-4.18)) and increased antibiotic use within 14 days (34% vs 18%, adjusted relative risk (aRR) 1.80 (0.85-3.79)) was seen in the colour chart group. From 38 sputum substudy patients, 57 samples were received (42 stable, 15 during AECOPD), with 30% containing potentially pathogenic bacterium (PPB). Purulent sputum was more frequent in bronchiectasis, independent of disease state (stable vs exacerbation) or PPB presence, suggesting sputum colour alone does not reliably guide antibiotic use.
CONCLUSION: Under-recruitment precluded definitive conclusions. However, sputum colour is unlikely to be a useful addition to COPD SM in primary care.
TRIAL REGISTRATION NUMBER: The UK’s Clinical Study Registry: ISRCTN14955629 (https://doi.org/10.1186/ISRCTN14955629; registration date: 11 Number 2020).
PMID:41073133 | DOI:10.1136/bmjresp-2025-003615
