PROJECT SUMMARY
Opioid and benzodiazepine misuse and abuse has increased dramatically in recent years. Despite
precipitous increases in overdoses and rising misuse, the role that physicians play in
inappropriate prescribing remains understudied. However, 34% of individuals who misused
prescription medications obtained them directly from a single doctor. Network-based models,
which analyze how the structure of social relations shape norms and behaviors, have successfully
been used to understand physician behavior and change physician practices. Applying network
methods to inappropriate prescribing will move surveillance beyond patient-based algorithms to
physician prescribing patterns and norms. This reorientation has the potential to help identify
physicians who are at risk of inappropriate prescribing, understand how these tendencies
correlate with their network positions, and design network-based interventions to reduce the
spread of prescription drug abuse. In this study, we aim to: (1) Examine the association between
social network position, physician characteristics, and different forms of inappropriate
prescribing and co-prescribing; (2) Use network and behavior co-evolution models to analyze the
role of social influence in the diffusion of inappropriate prescribing practices; (3) Compare the
relative effectiveness of insurance network optimization to eliminate inappropriate prescribers,
Prescription Drug Monitoring Programs, and insurance formulary restrictions. We will draw on
prescriptions of benzodiazepines and opioids from IMS Health's LRx database between January
1, 2005 and December 31, 2016. The 2009 data covered 224,140,604 unique patients, 916,338
prescribers, encompassing 135 million opioid prescriptions and 46 million benzodiazepine
prescriptions. We will augment the LRx data with Medicaid Managed Care claims from three
states. Using IMS and Medicaid data, we will construct longitudinal physician referral networks.
The proposed analyses using this data will provide greater insight into how social network
position and social influence can be leveraged to combat inappropriate prescribing of controlled
substances. Physicians are uniquely positioned to help end the prescription drug epidemic. Past
research indicates that changing prescribing behavior will require interventions targeted at
groups of physicians, rather than individuals, since prescribing norms are reinforced in
communities of practice. Our study will strengthen efforts to better understand prescription drug
abuse and lead to actionable recommendations that public and private stakeholders, as well as
payers can take to reduce the rate of inappropriate prescribing through improved detection
algorithms, more efficient targeting of policies and educational efforts, and new policies.