Table 2.

Bradford–Hill criteria for causality applied to the association between COPD, asthma and bronchiectasis

Bradford Hill criteria	COPD	Asthma
Strength of association/effect size	Some studies report strong associations but may have methodological limitations	Strong association demonstrated
Consistency/reproducibility	Reported by several studies with most reporting a high prevalence	Reported by several studies with most reporting a high prevalence
Specificity	Lacking, as there are often other possible causes present	Lacking, as there are often other possible causes present
Temporality	Rarely demonstrated	Rarely demonstrated
Biological gradient	Clearly demonstrated across multiple studies that as COPD becomes more severe bronchiectasis is more prevalent	Some evidence that bronchiectasis is more common in more severe asthma
Plausibility	Both COPD and bronchiectasis are characterised by neutrophilic inflammation and share common features and therefore probably some common pathophysiology	Strong association between bronchiectasis and ABPA proves an association with eosinophilic inflammation, therefore association with asthma is highly plausible
Coherence	Not fully testable	Not fully testable
Experiment	Experimental models of COPD are generally poor and conclusions cannot be drawn	Experimental models of asthma do not show features of bronchiectasis
Analogy	COPD fits the model of the vicious cycle of bronchiectasis, being characterised by airway inflammation, infection and airway damage	Asthma fits the model of the vicious cycle of bronchiectasis, being characterised by airway inflammation, infection and airway damage