Association between blood eosinophils and acute exacerbation of COPD risk in patients with COPD in primary care

Objective

We examined the association between blood eosinophil levels and subsequent rate of acute exacerbation of chronic obstructive pulmonary disease (AECOPD) in a population-based cohort of COPD patients managed in primary care.

Methods

This retrospective cohort study included COPD patients from the UK Clinical Practice Research Datalink linked to the Hospital Episodes Statistics. Patients had ≥1 blood eosinophil count that was not performed within ±30 days of an AECOPD episode. Blood eosinophil counts were modelled as a continuous exposure variable using a fractional polynomial model; the best fitting model was plotted against the incidence rate (IR) of moderate or severe AECOPD per person-year (PY), stratified by AECOPD history and inhaled corticosteroid (ICS) use.

Results

A total of 17,495 patients were included. In the overall cohort, the adjusted IRs of moderate or severe AECOPD increased modestly with increasing eosinophil counts, from 0.80/PY for blood eosinophil levels <100 cells/μL to 0.93/PY for 700 + cells/μL. Adjusted IRs for moderate or severe AECOPD were higher for those with a history of moderate or severe AECOPD, and increased with increasing blood eosinophil levels compared with those with no prior history of AECOPD. Treatment with ICS was associated with higher IRs of moderate or severe AECOPD, which increased with increasing blood eosinophil levels compared with non-exposure to ICS.

Conclusion

History of AECOPD and ICS use appear associated with the impact of blood eosinophil levels on the rate of AECOPD, and are important considerations for future studies exploring the relationship between blood eosinophil levels and AECOPD risk.