Direct and Indirect Effects of Disasters on Drug-Related Outcomes

Project Summary Brought on by anthropogenic causes, disasters have increased in number, intensity, costliness, and human impact, including public health. Over the same period, drug-related morbidity, mortality, and dependence have increased dramatically. While research has shown that experiencing a disaster affects physical and mental

health, the literature on disasters’ effects on substance use outcomes is often limited to single disasters (e.g., Hurricanes Katrina and Sandy), based on convenience or opportunistic samples that are difficult to generalize across disasters, and with equivocal findings. This research study will use quasi-experimental methods and a

robust disaster database to examine the impact of disasters on aggregate-level mortality and individual-level substance use over an extended period, including examining intervening mechanisms through which disasters are hypothesized to operate using a Big Events conceptual model. For disaster data, we will employ a currently

underutilized dataset within public health research: the Federal Emergency Management Agency’s Disaster Declaration Database, a temporally-specific and geocoded dataset containing all disaster declarations in the U.S. dating back to 1964. We will combine this information with county-level mortality data from the Centers for

Disease Control and Prevention and drug arrest data from the FBI Uniform Crime Reports, individual-level data on use, distress, and treatment gaps from the National Survey on Drug Use and Health, and control variables from the U.S. Census Bureau. Taking advantage of the exogenous nature of disasters, often used as a

statistical instrument, we will employ quasi-experimental methods to estimate the main effects of disaster occurrence as well as the mechanisms though which disasters operate, including psychological distress, treatment disruption, and drug market disruption. In addition to county fixed-effects in aggregate analyses, both

levels of analysis will incorporate temporal and spatial variability in order to understand how disaster effects diffuse over both dimensions. Finally, we will conduct these analyses by key demographics of both the county and individual, including sex, race/ethnicity, education, and urbanicity, which will permit the identification of

fundamental causes of health disparities. Our proposed research provides an opportunity to test the effects of two phenomena that have greatly impacted public health. We will also contribute in additional ways. First, the examination of mechanisms through which disasters operate will identify which resources need immediate

redress in the face of disaster to curb adverse substance use outcomes. Second, the identification of possible health disparities in the effects of disasters on substance use outcomes can assist in knowing where and how resources can be deployed to facilitate more equitable outcomes. Finally, we will create and make publicly

available a county by month database of nearly 50-years of disaster data for future researchers to use to examine additional public health outcomes, which grows even more important given the anticipated continuation of global climate change.