Kennedy-Hendricks et al. Respond

Measuring the unique impacts of particular components of Florida’s multifaceted response to the prescription opioid overdose epidemic is complicated by the implementation of these interventions in close temporal proximity. In a previous study,¹ Delcher et al. found significant declines in oxycodone-caused overdose mortality associated with Florida’s implementation of a prescription drug monitoring program (PDMP). In our study, we used a flexible modeling approach to discern significant changes in trends in Florida’s prescription opioid overdose mortality rates and to distinguish these aberrations from changes in trends in a comparison state. Our models did not identify any change in these trends corresponding with the time that Florida implemented its PDMP. Rather, we found that the only point at which Florida’s trends shifted—and dramatically so—was in early 2010 when pain clinics had to register with the state and law enforcement arrested a number of individuals running the state’s largest pill mills. This was the beginning of a long, steep decline in opioid overdose mortality that continued through a period in which the state enacted two laws curtailing and then eliminating pain clinic and physician dispensing of opioids, conducted additional law enforcement operations targeting pill mills, and implemented the PDMP.

Although the study by Delcher et al. suggests that the Florida PDMP reduced oxycodone-caused overdose mortality rates,¹ there is limited national evidence that PDMPs, as implemented to-date, are associated with reductions in prescription opioid overdoses.³ State PDMP laws vary substantially, which complicates national studies’ ability to assess their effectiveness as a standardized policy intervention.⁴ The protective effect of Florida’s PDMP estimated in the study by Delcher et al. also may be partly attributable to interventions directed at pill mills in the months prior to and soon after the PDMP was implemented.

Delcher et al. also raise concerns about potential unintended consequences of these policies. However, analyses with no comparison group that attribute increases in heroin overdose mortality to prescription opioid policies should be interpreted with caution. Heroin overdose deaths are rising nationally.⁵ A recent review of the relationship between nonmedical prescription opioid use and heroin use found little evidence to-date that implementation of prescription opioid policies has led directly to increases in heroin use although the authors note that the literature on this relationship remains sparse.⁶ It is clear that more research is imperative to determining the effectiveness of various policy interventions and to understanding the interconnections between these policies and heroin use.