The Association of Industry Payments to Physicians with Prescription of Brand-Name Intranasal Corticosteroids

Abstract

Objectives To examine the association of industry payments for brand-name intranasal corticosteroids with prescribing patterns. Study Design Cross-sectional retrospective analysis. Setting Nationwide. Subjects and Methods We identified physicians prescribing intranasal corticosteroids to Medicare beneficiaries 2014-2015 and physicians receiving payment for the brand-name intranasal corticosteroids Dymista and Nasonex. Prescription and payment data were linked by physician, and we compared the proportion of prescriptions written for brand-name intranasal corticosteroids in industry-compensated vs non-industry-compensated physicians. We associated the number and dollar amount of industry payments with the relative frequency of brand-name prescriptions. Results In total, 164,587 physicians prescribing intranasal corticosteroids were identified, including 7937 (5%) otolaryngologists; 10,800 and 3886 physicians received industry compensation for Dymista and Nasonex, respectively. Physicians receiving industry payment for Dymista prescribed more Dymista as a proportion of total intranasal corticosteroid prescriptions than noncompensated physicians (3.1% [SD = 9.6%] vs 0.2% [SD = 2.5%], respectively, P < .001). Similar trends were seen for Nasonex (12.0% [SD = 16.8%] vs 4.8% [SD = 13.6%], P < .001). The number and dollar amount of payment were significantly correlated to the relative frequency of Dymista (ρ = 0.26, P < .001 and ρ = 0.20, P < .001, respectively) and Nasonex prescriptions (ρ = 0.09, P < .001 and ρ = 0.15, P < .001, respectively). For Dymista, this association was stronger in otolaryngologists than general practitioners ( P < .001). There was a stronger correlation between the percentage of prescriptions and the number and dollar amount of payments for Dymista than for Nasonex ( P = .014 and P < .001). Conclusions Industry compensation for brand-name intranasal corticosteroids is significantly associated with prescribing patterns. The magnitude of association may depend on physician specialty and the drug’s time on the market.