The research, led by Jamie L. Humphrey of RTI International together with Jason Williams, Sofia A. Oviedo, Bradley R. Ray, and Jessica D. Cance, addresses a question that has lingered as the synthetic opioid crisis deepened. Fentanyl and its analogs have driven unprecedented overdose mortality in the United States, but their effects extend beyond fatal poisoning. The drug’s extreme potency and rapid onset have altered injection frequency and syringe-sharing behavior in ways that heighten the risk of HIV transmission, and clusters of new infections in several states have underscored the urgency of embedding HIV testing, prevention, and treatment within settings where people who use drugs already seek care.

Integrated HIV services within substance use disorder treatment facilities have long been viewed as a critical intervention point for this syndemic of addiction, overdose, and infectious disease. Whether such services actually proliferated as fentanyl spread geographically, however, remained unknown. To find out, the research team assembled a longitudinal dataset of United States counties from 2015 through 2023, relying on geocoded facility listings from the National Survey of Substance Abuse Treatment Services and its successor, the National Substance Use and Mental Health Services Survey. The outcome of interest was the annual county-level proportion of substance use disorder treatment facilities reporting that they provided HIV services.

A central methodological challenge was defining when the fentanyl era began for any given county. Rather than relying on a single national date, the investigators derived state-specific onset points for rapid illicit fentanyl market penetration using data from the National Forensic Laboratory Information System, which tracks drug submissions to forensic laboratories across the country. This allowed the team to model each county’s trajectory of HIV service availability before and after the moment fentanyl effectively took over the local illicit drug supply, capturing the staggered diffusion of the synthetic opioid wave as it moved across the country.

The statistical approach was equally refined. The researchers employed longitudinal growth models with a beta distribution, a technique suited to modeling proportions bounded between zero and one, to estimate rates of change in HIV service availability across the study period. The models incorporated a set of time-varying covariates designed to isolate the effect of fentanyl’s arrival from other concurrent forces: county-level drug overdose mortality and HIV incidence rates, each lagged by two years to reflect plausible causal timing, as well as state-level policy factors including Medicaid expansion and prohibitions on prior authorization requirements for medications for opioid use disorder.

The headline result is striking in its simplicity. Between 2015 and 2023, the average proportion of substance use disorder treatment facilities providing HIV services doubled among counties that had treatment infrastructure in place. This was not a uniform trend, however. The geographic heterogeneity was considerable: following the introduction of fentanyl, HIV service availability increased in 38.8 percent of counties, remained stable in 14.0 percent, and decreased in 24.5 percent. Perhaps most sobering, 22.7 percent of counties had no measurable HIV service availability at all during the period because they lacked substance use disorder treatment facilities entirely, leaving residents of those areas without any local point of access to integrated care.

The adjusted models sharpened the temporal story. Before fentanyl’s arrival, the increase in HIV service provision within counties was modest and not statistically significant, with an estimated slope of 0.021 and a standard error of 0.013. After fentanyl was detected in a state’s drug market, integration accelerated markedly, with a post-fentanyl slope of 0.072 and a standard error of 0.009, a difference that was highly significant at P less than .001. A formal test of the interaction between the slope and fentanyl’s introduction confirmed that the acceleration was genuinely tied to the synthetic opioid’s arrival, with an interaction coefficient of 0.051 and a standard error of 0.014, also significant at P less than .001.

What these coefficients imply, in practical terms, is that the arrival of fentanyl appears to have functioned as a shock to the addiction treatment system, prompting facilities to adapt their service portfolios in response to a visibly escalating threat. The authors interpret this pattern as evidence of adaptive integration of HIV prevention and care within addiction treatment systems during the fentanyl era. Treatment organizations, confronting a drug supply that dramatically raised both overdose risk and injection-related HIV vulnerability, appear to have recognized that testing for HIV, linking patients to antiretroviral care, and offering preventive services were no longer optional adjuncts but core components of responsible care for a population facing compounded dangers.

Yet the study’s darker findings demand equal attention. Nearly one-quarter of United States counties had no in-person substance use disorder treatment infrastructure at all during the study window, a persistent service desert that no amount of integration within existing facilities can remedy. Moreover, even among counties with treatment capacity, the post-fentanyl period saw service availability decline in roughly one in four, indicating that the adaptive response was far from universal. Geographic disparities in access to integrated HIV and addiction services remain substantial, and the counties most vulnerable to overlapping overdose and HIV epidemics are often precisely those least equipped to respond.

The authors conclude that these patterns underscore the need for policies that expand substance use disorder treatment capacity and actively incentivize HIV service integration in underserved communities. As fentanyl, and increasingly its more potent analogs, continues to dominate the illicit drug supply, the window for preventive action narrows in the places least served. The research, which was supported by the National Institute on Drug Abuse under Award Number U24DA057611 and reviewed by the RTI International Institutional Review Board under protocol STUDY00022322, provides both a benchmark and a warning: the treatment system can respond to crisis, but only where treatment exists, and only when policy deliberately directs resources toward the communities where the syndemic burns hottest.

Subject of Research: Longitudinal county-level analysis of HIV service availability in US substance use disorder treatment facilities before and after the state-specific arrival of illicit fentanyl, 2015 to 2023.

Article Title: HIV service integration in substance use treatment facilities in the fentanyl era, 2015–2023

Article References: HIV service integration in substance use treatment facilities in the fentanyl era, 2015–2023. (n.d.). https://doi.org/10.1186/s12889-026-29460-0

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29460-0

Keywords: HIV services, fentanyl, substance use disorder treatment, integrated care, overdose crisis, health services accessibility, geospatial analysis, HIV prevention, Medicaid expansion, opioid epidemic, public health, service integration