Comparison of heroin and fentanyl use in US nationally representative surveys.

Background: Given the opioid overdose crisis, surveillance of evolving opioid use patterns is critical to the effective deployment of mitigation strategies. The National Survey on Drug Use and Health (NSDUH) provided the first annual US estimate of illicitly manufactured fentanyl (IMF) use in 2022. However, as a household survey, NSDUH may not capture the full extent of population heroin and IMF use. We compare estimates of past-year heroin and IMF use and correlates of use in NSDUH and the Survey of Non-Medical Use of Prescription Drugs (NMURx) survey which employ an alternate sampling strategy.

Methods: We conducted a cross-sectional analysis of the 2022 NSDUH and NMURx. NSDUH samples respondents using a probability-based approach targeting community-dwelling individuals, while NMURx samples respondents using an opt-in, online survey panel. US adults ages 18 years and older were included. The main outcomes were differences in the weighted percentage of population reporting past-year use of heroin, IMF, and either heroin or IMF between the surveys. Secondary outcomes were the patterns of association of past-year heroin or IMF use with comorbid substance use, treatment utilization, and demographic characteristics between the surveys. Data were analyzed March to June 2024.

Results: NSDUH (n = 47,100 respondents) had a lower proportion of respondents who identified as non-Hispanic White and graduated college, and a higher proportion with past week employment than NMURx (N = 59,041 respondents). Past-year use of heroin, IMF, and either heroin or IMF were lower in the NSDUH than the NMURx. NSDUH estimated 0.52% (95% CI: 0.40%, 0.69%) %) of the US population used either heroin or IMF in the past year compared to 1.05% (95% CI: 0.97%, 1.14 0) in NMURx. In regression models, stimulant and benzodiazepine use were consistently associated with increased heroin or IMF use across both surveys.

Conclusions And Relevance: The estimated prevalence of heroin or IMF use was nearly 50% higher in the NMURx compared to NSDUH. These results highlight the importance of using complementary surveillance approaches to obtain accurate estimates of the prevalence and patterns of heroin or IMF use.