The Effects of Prescription Drug Monitoring Programs on Opioid Related Hospitalizations

Between 1999 and 2015, overdose deaths involving opioids have increased nearly six-fold from 5,441 in 1999 to 31,546 in 2015, and drug poisonings have become the leading cause of external injury deaths. In response to this epidemic, several states have implemented Prescription Drug Monitoring Programs (PDMPs). PDMPs are intended to facilitate and increase available information to prescribers, government agencies, and medical licensing boards, among others. Such a program may address market failures related to imperfect information, which could affect the behavior of patients, prescribers, pharmacists, and insurance companies. Using administrative hospitalization discharge records from the Healthcare Cost and Utilization Project (HCUP) National Inpatient Sample (NIS) from 1998 to 2011, I exploit variation within states and over time in the implementation of PDMPs using difference-in-differences models and event studies to identify the effect of PDMPs on monthly hospital aggregates of nonfatal opioid related inpatient stays.

This paper contributes to the literature in two ways. First, this paper identifies the impact of PDMPs on nonfatal opioid events rather than fatal overdoses. Nonfatal opioid related adverse events are economically and socially costly; the Council of Economic Advisers estimates that the cost of nonfatal opioid misuse was $72.3 billion in 2015. Additionally, nonfatal opioid events are more common than fatal opioid related events; it is estimated that for every fatal opioid overdose, there are nearly 12 nonfatal opioid overdoses, and nonfatal adverse opioid events occur 23 times more frequently than death. Consequently, nonfatal opioid overdose and adverse events are a larger proportion of total opioid related adverse events than fatal events. Additionally, nonfatal opioid related hospitalizations are a public health concern because of potential negative externalities, such as the increased risk and transmission of methicillin-resistant Staphylococcus aureus. Second, this paper is the first to use HCUP data to measure the impact of PDMPs on inpatient hospital visits involving opioid related events for a broad population (not limited to the disabled and Medicare enrollees) in certain states.

I find no evidence that PDMPs reduce the number of nonfatal inpatient visits related to opioid dependence, abuse, or poisoning. Additionally, I examine the effect of PDMPs on various subgroups, including non-Hispanic white men ages 45 to 54, a group that has recently experienced increases in morbidity and mortality (Case & Deaton, 2015). Although non-Hispanic white men ages 45 to 54 have higher rates of nonfatal prescription opioid poisoning inpatient stays, on average, PDMPs are not associated with significant decreases in these events for this demographic group or any demographic group.