Overdose mortality has been on the rise globally in rural areas of the United States, Canada, Australia, and elsewhere (European Monitoring Centre for Drugs and Drug Addiction, 2018, Han et al., 2019, Martins et al., 2015, Monnat and Rigg, 2018, Peterson et al., 2007, Scamvougeras et al., 2020). For example, from 2011 to 2018 the rate of unintentional drug-induced deaths in rural and regional Australia increased by 15.9%, while the rate in capital cities increased by only 3.6% (Penington Institute, 2020). Compared to other high-income countries, the United States has the greatest rate of drug overdose deaths (Ciccarone, 2017; Ciccarone, 2021; Preuss et al., 2020; Snowdon, 2022; Gumas and Baumgartner, 2023). In 2021, synthetic opioids contributed to about 66% of all overdose deaths in the US, compared to 18% in Sweden and Norway, and 11% in Australia (Gumas and Baumgartner, 2023). Rural areas in the United States (US) have been greatly impacted (Monnat and Rigg, 2018); between 1999 and 2016, in the US opioid mortality rates increased most outside of urban areas, upwards of 740 percent in nonmetro counties (Monnat and Rigg, 2018) though recent reports suggest that rural/urban disparities are now waning (Blair et al., 2023, Hedegaard and Spencer, 2021).

To date, studies of overdoses in US rural areas have almost exclusively analyzed administrative data (e.g., hospitalization data, death records, Emergency Medical Services (EMS) data to explore risk factors for opioid overdose (Blair et al., 2023, Dilokthornsakul et al., 2016, Dunn et al., 2010, Fix et al., 2021, Jones et al., 2012). Major risk factors identified in these administrative analyses include higher daily mean morphine doses (Dunn et al., 2010), psychiatric illness (Dilokthornsakul et al., 2016), polysubstance use (Jones et al., 2012, Kerr et al., 2007), Medicaid enrollment (Nam et al., 2020), and employment in physically demanding occupations like construction (Hawkins et al., 2019). While administrative data can sometimes support analyses of a census of events (e.g., all hospitalizations for an overdose, all deaths ascribed to overdose), they suffer multiple limitations. They typically exclude a wide array of crucial factors that may shape vulnerability and resilience to overdose, such as participating in substance use disorder (SUD) treatment; patterns of illegalized drug use; or criminal-legal interactions.

Moreover, hospital data on overdoses only capture the fraction of overdose events in which individuals sought and received medical care, and analyses of mortality data only capture the fraction of overdoses that resulted in death. Finally, analyses of overdose correlates in hospital, mortality, or EMS data generally fail to pinpoint risk factors for overdose among people who use drugs (PWUD) because these data sources rarely contain information needed to identify PWUD who did not overdose, a key comparison group. Analyses of these factors in these administrative data thus meld determinants of drug use among the general population with determinants of overdose among PWUD.

Surveys with PWUD samples, in contrast, can comprehensively identify a wide range of possible factors that create vulnerability and resilience to overdose in this population specifically. These surveys have long been conducted in cities, and this urban research has revealed multiple correlates of nonfatal overdose, including witnessing an overdose, recent release from jail or prison, economic insecurity, houselessness, Hepatitis C infection (HCV), co-use of opioids and benzodiazepines, heroin injection, fentanyl injection, and the number of drugs injected (Armoon et al., 2021, Baggett et al., 2013, Darke and Hall, 2003, Galea et al., 2006, Hunter et al., 2018, Kerr et al., 2007, Liebling et al., 2018, Mateu-Gelabert et al., 2020, Wallace et al., 2019).

Despite surging overdose rates, only a handful of studies have surveyed correlates of non-fatal overdose among rural PWUD (Ahmad et al., 2021, Havens et al., 2011, Otachi et al., 2020, Schalkoff et al., 2020, Schalkoff et al., 2021). These vital studies have found that injecting fentanyl or prescription opioids, witnessing an overdose, ever having been in substance use disorder treatment or recently attempting to stop using, and post-traumatic stress disorder elevate nonfatal overdose risk (Ahmad et al., 2021, Havens et al., 2011, Otachi et al., 2020). The present analysis builds on this vital but nascent research by analyzing survey data on nonfatal overdoses and their correlates among a large sample of rural PWUD who live in eight states at the heart of the US opioid epidemic.