In an analysis of the entire sample, multivariable regression models relating total and inhaler-specific out-of-pocket costs to CRN are shown in Table 3. In the base model, (model 1) membership in the Inhaler+/CPD+ group was strongly associated with CRN (OR, 1.43; 95% CI, 1.21-1.69). This effect was larger than that seen for the Inhaler−/CPD+ group, suggesting that the increased risk of CRN observed in the Inhaler+/CPD+ group is at least partly due to inhaler-specific, rather than pulmonary disease-specific factors. After adjusting for total out-of-pocket costs (model 2), this association remained statistically significant (OR, 1.25; 95% CI, 1.06-1.47). Similar results were obtained after adjusting for noninhaler out-of-pocket costs (model 3). However, as is seen in model 4, adjustment for out-of-pocket inhaler costs completely attenuates the association between Inhaler+/CPD+ status and CRN (OR, 0.95; 95% CI, 0.71-1.28), suggesting that inhaler cost is an important link between CPD and risk of CRN. Out-of-pocket inhaler cost categories of $21 to $50, $51 to $75, and > $75 had ORs for CRN of 1.82 (95% CI, 1.20-2.76), 2.06 (95% CI, 1.82-2.76), and 2.33 (95% CI, 1.46-3.71), respectively, compared with the reference group of $0. Not surprisingly, noninhaler out-of-pocket costs are also significantly associated with CRN (OR, 1.29; 95% CI, 1.23-1.35).