Abstract
Background: In the past, β-blockers were contra-indicated in COPD. Recent evidence, including a cohort study by Rutten et al. (Arch Intern Med 2010), suggests that β-blockers produce good outcomes in COPD. This degree of protection might be biased.
Objectives: To study the effect of β-blockers on the risk of severe COPD exacerbations.
Methods: We conducted several analyses to study the effect of β-blockers on COPD exacerbations by: 1) mimicking the cohort design of Rutten 2) avoiding immortal time and exposure bias through a nested case-control study in a COPD cohort and 3) reducing confounding by indication by restricting the case-control analyses to users of β-blockers during follow-up. Data from the Dutch IPCI GP database (2000 to 2007) were used. Cases were all COPD patients with a first COPD exacerbation requiring hospitalization. To each case, controls were matched on age, sex and indexdate. Cox proportional hazard and conditional logistic regression analyses were used.
Results: Within the cohort of 6788 COPD patients, 619 patients had a COPD exacerbation requiring hospitalization. Current use of β-blockers significantly reduced the risk of severe COPD exacerbation in the cohort analysis (HRadj 0.73, 95%CI 0.60-0.90). In the case-control analysis, use of β-blockers reduced the risk of exacerbations by 40% (ORadj 0.60, 95% CI 0.44-0.82). When controlling for confounding by indication, the protective effect was clearly attenuated (ORadj 0.87, 95% CI 0.52-1.45). In patients with co-existing heart failure, a significant protective effect was observed (ORadj 0.15, 95% CI 0.03-0.80).
Conclusions: The choice of studydesign is crucial when assessing the effect of β-blockers on COPD outcomes.
- © 2011 ERS