Abstract
BACKGROUND: Overdose deaths in the United States rose substantially during the COVID-19 pandemic. Disruptions to the drug supply and service provision introduced significant instability into the lives of people who use drugs (PWUD), including volatility in their drug use behaviors. METHODS: Using data from a multistate survey of PWUD, we examined sociodemographic and drug use correlates of volatile drug use during COVID-19 using multivariable linear regression. In a multivariable logistic regression model, we assessed the association between volatile drug use and past month overdose adjusting for sociodemographic and other drug use characteristics. RESULTS: Among participants, 52% were male, 50% were white, 29% had less than a high school education, and 25% were experiencing homelessness. Indicators of volatile drug use were prevalent: 53% wanted to use more drugs; 45% used more drugs; 43% reported different triggers for drug use, and 23% used drugs that they did not typically use. 14% experienced a past-month overdose. In adjusted models, hunger (β=0.47, 95% CI: 0.21-0.72), transactional sex (β=0.50, 95% CI: 0.06-0.94), and the number of drugs used (β=0.16, 95% CI: 0.07-0.26) were associated with increased volatile drug use. Volatile drug use was associated with increased overdose risk (aOR=1.42, 95% CI: 1.17-1.71) in the adjusted model. CONCLUSIONS: Volatile drug use during the COVID-19 pandemic was common, appeared to be driven by structural vulnerability, and was associated with increased overdose risk. Addressing volatile drug use through interventions that ensure structural stability for PWUD and a safer drug supply is essential for mitigating the ongoing overdose crisis.