Interview by Michelle Slattery, 3/31/2019
Diversion of prescription (Rx) drugs for non-medical purposes is a serious public health concern, with over 50 million individuals (20 percent of the population) having misused Rx drugs in their lifetime. Among illicit drug users, Rx drug abuse is second in prevalence only to marijuana, and among new initiates of illicit drugs, more than 25 percent have initiated with non-medical use of Rx drugs, fueling the concern that Rx drugs may be a gateway to harder illicit drug use, especially for youth.
To combat these problems, a popular initiative has been to implement Prescription Drug Monitoring Programs (PDMP), a state level electronic database that tracks the prescribing and dispensing of controlled Rx drugs. PDMPs can provide critical information on the patient’s prescription history to physicians and dispensers, helping identify patients who may be doctor-shopping, misusing Rx substances, and are at high risk of an overdose and would benefit from timely interventions.
This study, Mandatory Access Prescription Drug Monitoring Programs and Prescription Drug Abuse
, exploits variation across states in the timing of PDMP adoption and the strengthening of PDMP requirements to assess whether, and the extent to which, PDMPs have been effective in reducing Rx drug abuse, based on objective measures including substance abuse treatment admissions and mortality rates. In the study, authors Dr. Anca M. Grecu, Dr. Dhaval M. Dave and Dr. Henry Saffer report robust evidence that mandatory access provisions, which require providers to query the PDMP prior to prescribing controlled drugs and have raised PDMP utilization rates, have significantly reduced Rx drug abuse.
Non-mandatory PDMP laws, however, are not found to significantly impact treatment admissions or mortality, expectedly because provider query rates in states that do not mandate PDMP access have been historically low. The effects are driven primarily by a reduction in opioid abuse, generally strongest among young adults (ages 18 to 24) and underscore important dynamics in the policy response. The results also show some complementary reduction in treatment admissions related to cocaine and marijuana abuse, suggesting that the effects of mandatory PDMP provisions may also spillover to the use of other illicit drugs.
My interview with the authors, Dr. Anca M. Grecu (top left), Dr. Henry Saffer (top center) and Dr. Dhaval M. Dave (top right), follows:
1. What inspired you to research the prescription drug monitoring programs (PDMP)?
We came to think of it in a rather indirect way. It is hard not to notice how much and how fast our world is changing due to changes in technology. Of course, health care delivery is affected by these changes. Patients can meet with their providers virtually and obtain medical advice, services referred to as telecare. We welcome technology when it offers easier access to health care to people in rural areas for instance, but we shy away from it upon realizing that it can be abused. It is easier to lie and obtain prescriptions drugs illicitly when you do not meet with a physician in person. So we ask patients to meet with the doctor in person, which can be costly for some people, so it is a barrier to care for some. By providing easy access to the prescription history of a patient, on-line PDMPs help doctors better monitor their patients. Implementing PDMPs, and enhancing these programs, has also been a population statewide initiative (and a component of broader recommendations by the Obama administration, President Trump’s Opioid Commission, the GAO, and others). Hence, we started thinking that if PDMPs work well, these tools could lead the way to other ways we can use technology to help physicians monitor their patients so we can also relax other constraints we now impose on the delivery of care. We were very excited to see that well thought out PDMPs can help reduce opioid abuse.
2. The paper points out that mandatory PDMP is effective in order to reduce prescription drug abuse. Can you talk about the main findings and how it is different from the previous studies?
This study recognizes that people respond to incentives and so if checking the PDMP takes time, perhaps some physicians, not recognizing the full importance of checking the PDMP, would not access the system. Policy makers also respond to incentives so it makes sense that some states could be more likely to be early adopters of PDMPs. However, each adopting state would like to have an effective PDMP. So conditional on recognizing the importance of having a PDMP, the introduction of the mandatory clause that requires physicians to check the system is the better way to investigate whether PDMPs are an effective way to curb doctor shopping and identify other at-risk patients prior to prescribing opioids. We found that mandatory PDMPs are in fact associated with lower treatment admissions for opioid misuse and with lower opioid-related mortality rates. Much of the older literature had found little consistent evidence on the effectiveness of PDMPs, which can be explained by the failure to distinguish mandatory access vs. voluntary PDMPs. Most of the older state PDMPs were voluntary, and in these cases physician query rates were quite low; thus, if physicians are not consulting the PDMPs when they should prior to prescribing opioids or other controlled substances, then perhaps the lack of a consistent finding on their effectiveness is to be expected.
3. Were any challenges encountered with conducting this research and if so how did you overcome them?
We would like to think that at the end of our research we have THE answer. A more honest way to think about research is that the end result should help the reader understand the correct interpretation of the results of the analysis, understand the importance of differentiating gradations in the policy, and assess the reliability and generalizability of reported estimates. In this research we had to conclude that an analysis of PDMPs in general was not very reliable because of differences in their design before 2003 when the National Alliance for Model State Drug Laws published the Model Prescription Monitoring Program Act, which subsequently was followed by most states. Furthermore, as we were going through the literature and some of the prior work, it became apparent that most of the earlier effects were conflating the voluntary programs with the very few strict mandatory provisions and thus not finding much of an effect. This offered little guidance for policymakers and states continuing to enact and enhance their PDMPs. So we focused on the programs we could study and offer a reliable, generalizable estimate of their impact.
4. Is further research needed and if so what would you recommend?
Markets are interconnected so every time there is an intervention on a market, it is important to also examine spillovers as well as interactions between the policies. There is some research into the impact of PDMPs on illegal drug markets. But it is also important to recognize that physicians also react to revealed information. For instance, the CDC updated their recommendations for opioid prescribing. The question at this point is whether physicians changed their preferred prescribing practices and, if so, what was the impact of this change on patients suffering from pain; in other words what was the impact on both appropriate and inappropriate prescribing.
5. U.S. has been going through an opioid crisis and this paper is very timely to respond this matter. What are the policy implications and what recommendations do you have from this research?
Our research suggests that the design of PDMP matters for their efficacy and that must-access provisions are very important predictors of their impact. The timing of the policy response, that there is a lag in the effect of these programs, also points to the importance of learning and practice diffusion. As with all new technologies and policies, it may take time for physicians to shift their practice patterns.
• ANCA M. GRECU is an Associate Professor in the Department of Economics and Legal Studies in the Stillman School of Business at Seton Hall University (e‐mail: email@example.com).
• DHAVAL M. DAVE is a Stanton Research Professor in the Department of Economics at Bentley University (e‐mail: firstname.lastname@example.org). He is also a Research Associate at the National Bureau of Economic Research (NBER) and the Institute of Labor Economics (IZA).
• HENRY SAFFER is a Research Associate at the National Bureau of Economic Research (NBER) (e‐mail: email@example.com).
View the authors' JPAM article by clicking the link above.