Overdoses with Xylazine and Fentanyl Recorded in Pennsylvania’s Overdose Information Network

Introduction

Xylazine, a veterinary non-opiate sedative and analgesic, is increasingly detected in overdose deaths in the United States (US), usually accompanied by fentanyl (Spencer et al., 2023). Philadelphia, Pennsylvania is considered the earliest epicenter of xylazine-adulterated drugs in the mainland US (Friedman et al., 2022; Zhu, 2023), and xylazine-related overdose deaths have primarily been concentrated in the northeastern US (Spencer et al., 2023).

Nonetheless, recent studies suggest that xylazine’s presence in illicit drug supplies may be spreading throughout the US as well as other nations (Bradford et al., 2024; Cano et al., 2024; Copeland et al., 2024; Delcher et al., 2024; Friedman et al., 2022, 2024; Thomas et al., 2024; Wu and Austin, 2024).

Although the potential impacts of xylazine on overdose outcomes are unclear (D’Orazio et al., 2023; Hoffman, 2023), polysubstance overdoses co-involving xylazine and fentanyl present additional challenges for clinical management and public health strategies (Gupta et al., 2022). Since xylazine is not an opioid, administration of the opioid overdose reversal agent naloxone does not directly address xylazine’s effects (Gupta et al., 2023); however, naloxone is still recommended as a critical tool to save lives in cases of xylazine-related overdose (Gupta et al., 2023; Zagorski et al., 2023; Zhu, 2023) because opioids such as fentanyl are nearly always co-involved (reported in 99% of xylazine-related fatalities nationwide; Spencer et al., 2023).

Most studies on xylazine-involved overdoses focus exclusively on fatal (Johnson et al., 2021; Kariisa et al., 2023; Sibbesen et al., 2023; Spencer et al., 2023) or non-fatal overdoses (German et al., 2024) or clinical samples (Love et al., 2023). Considering that a better understanding of both fatal and non-fatal overdoses involving xylazine could strengthen prevention and harm reduction efforts, the present study leverages data on both fatal and non-fatal overdoses recorded by law-enforcement/first-responders in Pennsylvania (excluding Philadelphia). This descriptive study examines co-involved drugs, naloxone administration, and survival in fentanyl overdoses with and without xylazine involvement suspected.

Methods

Results

As reflected in Table 1, ODIN-recorded fentanyl overdoses, both with (n=95) and without (n=10,531) xylazine reported, were concentrated among men, in ages 30-39, and among Non-Hispanic (NH) White individuals, with no statistically significant demographic differences observed. As detailed in Supplemental Table S2, the highest numbers of both fentanyl-xylazine and fentanyl-no-xylazine ODIN overdose events were recorded in Allegheny and Montgomery counties. All of the suspected xylazine-fentanyl overdoses occurred during or after 2021, with the majority occurring after 2021 (Supplemental Table S2).

As depicted in Figure 1 (and detailed in Supplemental Table S3), several drugs were reported in significantly larger proportions of overdoses in the suspected xylazine group than the no-xylazine group: carfentanil (11.6% vs. 2.2%; p<0.001); other synthetic opioids/fentanyl analogs (20.0% vs. 7.6%; p<0.001); methamphetamine (21.1% vs. 6.9%; p<0.001); and benzodiazepines (6.3% vs. 2.3%; p=0.022). Conversely, heroin was reported in less than half (44.2%) of the overdoses in the xylazine group versus more than two thirds (67.6%) in the no-xylazine group (p<0.001). Naloxone was administered in 41.1% of the cases reportedly involving xylazine, versus 67.5% of the no-xylazine cases (p<0.001), and death was the outcome of 62.1% of the cases with xylazine recorded, compared to 27.7% of the no-xylazine cases (p<0.001).

• Download figure
• Open in new tab

Figure 1.

Characteristics of suspected overdose events involving fentanyl and xylazine versus fentanyl and no xylazine, as recorded in the Pennsylvania Overdose Information Network,^¶ 2018-2024.

Notes. Total number of fentanyl-involved overdoses = 10,626 (95 with xylazine also recorded and 10,531 without xylazine recorded).

^¶Data are from January 2018 through July 15, 2024 and do not include Philadelphia county. P values calculated based on Fisher’s Exact tests for associations between fentanyl/xylazine versus fentanyl/no xylazine and each characteristic examined: Heroin, p<0.001; Carfentanil, p<0.001; Other synthetic opioid/fentanyl analog, p<0.001; Pharmaceutical opioid, p=1.000; Methamphetamine, p<0.001; Cocaine/crack, p<0.364; Benzodiazepines, p=0.022; Alcohol, p=0.094; Naloxone administered, p<0.001; survived overdose, p<0.001. Abbreviations. Oth. Syn. Opi., other synthetic opioids/fentanyl analogs; Pharm. Opi., pharmaceutical opioid; Methamphet., methamphetamine; Benzo., Benzodiazepine; Unk., unknown.

In multivariable regression models (Table 2), even after adjusting for age, gender, race/ethnicity, year, county, and suspected involvement of other drugs, suspected xylazine involvement was associated with 65% lower odds of naloxone administration (AOR 0.35; 95% CI, 0.23-0.56; or 21 percent lower probability) and 79% lower odds of survival (AOR 0.21; 95% CI, 0.13-0.35; or 28 percent lower probability) among the suspected fentanyl overdoses recorded in ODIN. Regression results from sensitivity analyses (Table 3) were similar when broadening the analytic sample to overdoses involving any type of opioid (instead of fentanyl specifically) and when restricting the analysis time frame to the years xylazine was most frequently reported in ODIN, 2022-2024.

Discussion

Study findings highlight several differences between suspected fentanyl overdoses with, versus without, xylazine in the Pennsylvania ODIN (excluding Philadelphia County). First, compared to fentanyl-involved overdoses without xylazine suspected, significantly higher proportions of overdoses with xylazine suspected also reportedly involved fentanyl analogs/other synthetic opioids, methamphetamine, and benzodiazepines. This finding aligns with results of studies documenting stimulant or benzodiazepine co-involvement in xylazine-fentanyl deaths and larger numbers of co-involved drugs among deaths involving xylazine than those not involving xylazine (Bradford et al., 2024; Sibbesen et al., 2023).

Second, among ODIN-recorded fentanyl overdose deaths, suspected xylazine co-involvement was associated with lower probabilities of naloxone administration and survival, even after adjusting for other drug co-involvement, demographics, county, and year. In an analysis of fatal overdoses in 20 US states and the District of Columbia, naloxone administration was reported in similar proportions of fentanyl-related overdose deaths with and without xylazine detected (Kariisa et al., 2023). The contrast between this prior study’s findings and the results of the present study may possibly reflect differences in study settings as well as the inclusion of both fatal and nonfatal overdoses in the present study.

Although not examined in the present study, differences in naloxone administration rates based on suspected xylazine co-involvement may plausibly reflect factors such as: uncertainty regarding how to respond in light of distinctions between xylazine-associated sedation versus opioid-induced respiratory depression; elevated stigma surrounding xylazine (Bowles et al., 2024); and varying levels of access to naloxone or overdose education based on local area rurality or resource levels. All of the cases included in the present study reportedly involved fentanyl, yet it is not clear in what quantity and whether all cases evidenced opioid-induced respiratory depression that would specifically indicate a need for naloxone. Moreover, in some portion of ODIN cases, naloxone might not have been administered because the person was already deceased by the time first responders arrived.

Limitations

ODIN data are not inclusive of, nor representative of, all law-enforcement-attended overdoses in Pennsylvania. Moreover, overdoses in Philadelphia (the largest county, one with a particularly diverse population, and often regarded the epicenter of xylazine; Friedman et al., 2022) were not included in the study’s analysis, as the Philadelphia Police Department was not an ODIN reporting agency at the time of the study. All ODIN measures, including drug involvement, are based on first-responder assessment/report. ODIN reporters use a variety of methods to identify suspected drugs involved (including drug field testing, interviews with those at the scene, and investigation of drugs and paraphernalia present; Barboza and Angulski, 2020), and it is plausible that fatal overdoses may receive more investigative resources than non-fatal events; it is also possible that law enforcement may be more likely to suspect and report more drugs in fatal, relative to non-fatal, cases.

ODIN data may also possibly over-report historically common drugs such as heroin (as drug checking data indicate that what is sold as heroin has largely been replaced by fentanyl; The Center for Forensic Science Research & Education, 2023) while under-identifying emerging drugs such as xylazine. It is unclear to what extent this potential underreporting of xylazine compares to the level of xylazine underreporting in mortality data and nonfatal overdose surveillance data due to limited testing (Korona-Bailey et al., 2023; Spencer et al., 2023) or the level of discrepancy between residue xylazine test strip results and full toxicology testing (Thompson et al., 2024). Researchers have also noted a lack of standardized guidelines specifying quantities or signs that distinguish between an overdose in which xylazine was detected versus an overdose in which xylazine specifically contributed to the drug toxicity (Kariisa et al., 2023). Finally, many unmeasured factors (e.g., underlying health status and opioid tolerance levels, quantity of drugs consumed, and resources and procedures of each local area’s emergency system) may contribute to the outcomes examined in the study; some of these factors (e.g., local first responder procedures and resources) may also impact the measurement of the predictor (that is, the processes used to assess whether or not xylazine was involved in the overdose).

Conclusions

The present study did not test causal relationships, but rather provided preliminary evidence of lower naloxone administration and less survival in suspected xylazine-fentanyl-involved overdoses recorded by ODIN-participating law enforcement/first-responders in Pennsylvania (outside of Philadelphia). The potential role of xylazine in overdose outcomes is unclear (D’Orazio et al., 2023; Hoffman, 2023); research with rodents suggests that xylazine exposure may worsen risk of fatality from opioid overdose (Choi et al., 2023), while an analysis of opioid overdoses treated in nine US emergency departments documented less severe outcomes (less cardiac arrest and coma) in xylazine-positive, versus xylazine-negative, opioid overdose patients (Love et al., 2023).

Although further research is needed to better understand xylazine-fentanyl-involved overdoses, findings of the present study underscore the importance of expanding access to lifesaving medications such as naloxone. While naloxone does not directly reverse the effects of xylazine, it remains effective in reversing the effects of the fentanyl with which xylazine is often mixed (Kariisa et al., 2023). Public health and harm reduction programs play a vital role in disseminating evidence-based information on the local drug supply, naloxone, and rescue breathing and other lifesaving care for individuals with suspected overdoses involving fentanyl and xylazine. Study findings also support the importance of making low-threshold harm reduction and drug checking services widely available to people who use drugs and equipping first responders with naloxone and training regarding recognizing and managing the distinct challenges associated with xylazine exposure and polysubstance overdoses.