Drug overdose among female participants of a harm reduction program in Philadelphia

Introduction: In the past two decades, women's rates of fatal drug overdose tripled and nonfatal overdose has been common among women who use illicit drugs. While research has focused on epidemiological aspects of overdose, particularly opioid misuse, little evidence is available on behavioral and social determinants of overdose, especially those driven by gender. This is important because women often use substances for different reasons than men and have less control over their drug use practices than men. The present study addresses this gap through a mixed methods investigation of women's overdose experiences, with a focus on how gender shapes overdose risks and responses. Methods: Quantitative data were collected in Philadelphia in 2016-2017, from a sample of 220 largely street-involved, opioid-injecting adult women who approached harm reduction services. Quantitative methods assessed whether adulthood violence influences overdose frequency. A purposive subsample of 42 women completed a qualitative interview to understand the context of women overdose experiences using the "drug, set, and setting" framework. Finally, data were triangulated in a mixed-methods analysis to assess whether training in overdose prevention improved women's response to a witnessed overdose. Results: Overall, participants were largely unstably housed, unemployed, and had limited sources of social support. Quantitative analysis of personal overdose revealed that more than two-thirds experienced an overdose in their lifetime with the median of 3. Higher overdose frequency was associated with a higher number of violence types (ranging 0 to 16), higher score on a polyvictimization scale (i.e. reporting 9-16 violence types), being a victim predominantly of sexual violence or predominantly intimate partner violence. Findings from the qualitative analysis of personal overdose experiences identified three types of circumstances preceding the most recent overdose: "drug" type for overdoses attributed to the unpredictable quality of street opioids, polysubstance use, or chasing heroin's euphoric effect; "set" type, when women's emotional states affected by a "good" or "bad" day led them to unsafe drug use to celebrate or cope; and "setting" type where women's overdose was preceded by a recent change in setting, such as release from prison, which prompted unsafe drug use to address cravings or withdrawal. Findings from the mixed-methods analysis of experiences witnessing an overdose revealed that compared to untrained participants, trained women were more likely to administer naloxone and use two recommended responses, however the utilization of recommended overdose responses was impeded by fear of police, inability to carry naloxone or a phone, as well as a perceived lack of safety on the streets and when interacting with strangers. Conclusion: The study demonstrated the limitations of a conventional understanding of current overdose crisis as caused by the accessibility of harmful substances. Women's gender, intersecting with social exclusion, poverty, and vulnerability to multiple forms and perpetrators of violence, made women susceptible to personal overdose, and impeded an appropriate response to a witnessed overdose. Individual-level interventions emphasizing overdose avoidance through safer drug use may not be a long-term solution for women experiencing chronic stress due to violence, homelessness, and isolation. In addition to overdose trainings, findings call for upstream measures, such as assistance with recovering from victimization, securing housing and employment, and re-connecting with families. Those measures may not explicitly address overdose, but may prove highly effective for the overall normalization of women's lives, reduction of unsafe drug use, and ultimately, reversing the surge of overdose deaths.