PMC

Treatment comparisons. Comparisons of long-acting anticholinergics, inhaled corticosteroids and the combination of long-acting beta-agonists + inhaled corticosteroids with long-acting beta-agonists alone stratified for trials including patients with an FEV₁ > or ≤ 40% predicted.

All comparisons among inhaled drug regimens. The forest plots show odds ratios (95% confidence intervals) indicating the odds of at least one exacerbation in patients with a drug treatment from the row as compared with treatment from the corresponding column. For example, the odds ratio of 0.91 (0.81 to 1.03) indicates that long-acting anticholinergics are more effective than long-acting beta-agonists, although not significantly so.

Study flow from database searches to inclusion oftrials. 40 trials were excluded after title and abstract screening because they obviously did not fulfill the inclusion criteria (not patients with COPD, patients with unstable COPD, short-acting bronchodilators, treatment duration < 4 weeks, no exacerbations ascertained). Reasons for exclusion for the 20 studies excluded after full text assessment are listed in Additional file .