Presentation Authors: Kevin Chua*, Brooklyn, NY, Gen Li, Peter Stahl, Elias Hyams, New York, NY
Introduction: Prior studies have shown that pharmaceutical industry payments may be associated with prescribing habits among physicians. The relationship between payments and prescription of phosphodiesterase type 5 inhibitors (PDE5i) has not been explored. In this study, we evaluated whether industry payments for tadalafil were associated with prescribing habits among urologists (URO) and primary care physicians (PCPs).
Methods: Medicare Part D Public Use File and Open Payments Program database (OPP) were linked to identify URO and PCPs who prescribed and received a payment for tadalafil in 2015. PCPs were defined as internal medicine and family medicine physicians. We determined whether presence of and extent of payment were associated with prescription of tadalafil as well as the number of claims. Statistical tests included chi-squared test, univariable logistic regression and Pearson correlation coefficient.
Results: Within Medicare Part D, 2,602 URO and 3,095 PCPs prescribed tadalafil. Within OPP, 2,304 urologists and 12,465 PCPs received a payment from Eli Lilly pertaining to tadalafil. Range of payments were $10.21-$15,478.88 (median $25.16) for URO and $1.39-$21,681 (median $20.11) for PCPs. Payments were associated with tadalafil prescription among PCPs, but not among URO (Table 1). For URO, increased payment amount was not associated with prescribing (Table 2), but claim count was very weakly correlated with payment amount (p=0.042, r=0.063) and frequency (p=0.006, r=0.089). For PCPs, increased payment amount was associated with prescribing (Table 2), and claim count was very weakly correlated with payment amount (p=0.01, r=0.1), but not frequency (p=0.474, r=0.032).
Conclusions: There does not appear to be a strong relationship between payments and prescribing habits, which is reassuring regarding the ethics of physician-industry interaction. However, given the presence of a weak association, further study in other samples (e.g. private insurance) and with other PDE5i may be worthwhile.