Underpowered to show differences between antibiotics |
Heterogenous patient populations e.g. severity of exacerbation, comorbidity, evidence of infection |
No knowledge of pre-exacerbation condition by which to evaluate recovery |
Lack of long-term follow up |
No evaluation of co-medication likely to affect clinical outcome e.g. corticosteroids (oral, inhaled), long-acting β-agonists and anti-cholinergics |
Typically studies are designed as part of registration trials to show “noninferiority” (equivalence) i.e. new agent versus carefully chosen comparator |