Panel Paper:
Pharmaceutical Promotion, Shared Patient Networks and Second Generation Antipsychotic Prescribing in Medicare Part D
*Names in bold indicate Presenter
Study Design: Cross-sectional exploratory analysis linking publicly available prescriber data, pharmaceutical promotion data, and shared patient networks, in the United States in 2015. We use Care set labs root NPI graph to create shared patient networks for the population of physicians billing to Medicare. We link these networks to individual prescribing and pharmaceutical promotion data using National Provider Identifiers (NPI). Local shared patient networks are identified using the Louvain community detection algorithm. We use hierarchical binomial regressions to examine effects of promotion on prescribing at the physician and network level.
Population Studied: Physicians who prescribe one or more SGA in 2015
Principal Findings: Physicians who took 4+ payments for Abilify in 2015 had 13% higher odds of prescribing it than those with no payments, after adjusting for physician level variables. At the network level the odds of a physician prescribing Abilify were 44% higher for a physician in a network that was in the highest quartile of receiving Abilify payments compared to one in the lowest quartile.
Conclusions: Receiving a promotional payment for an SGA is strongly associated with higher odds of prescribing the promoted drug. Practicing in a network of providers who receive more promotional payments on average is also strongly associated with higher odds of a given physician in that network prescribing the promoted drug.
Implications for Policy or Practice: Until now, research informing policy aimed at mitigating conflicts of interest related to pharmaceutical promotional payments and prescribing has only considered individual level effects. This work shows additional effects at the physician network level that should also be considered when forming policy.
