Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 May 22.
Published in final edited form as: Subst Use Misuse. 2020 May 22;55(11):1774–1780. doi: 10.1080/10826084.2020.1762653

Fentanyl Preference among People Who Inject Drugs in West Virginia

Alyona Mazhnaya a, Allison O’Rourke b, Rebecca Hamilton White a, Ju Nyeong Park a, Michael E Kilkenny c, Susan G Sherman a, Sean T Allen a
PMCID: PMC7487277  NIHMSID: NIHMS1618352  PMID: 32441202

Abstract

Background:

Overdose fatality rates in rural areas surpass those in urban areas with the state of West Virginia (WV) reporting the highest drug overdose death rate in 2017. There is a gap in understanding fentanyl preference among rural people who inject drugs (PWID). The aim of this study is to investigate factors associated with fentanyl preference among rural PWID in WV.

Methods:

This analysis uses data from a PWID population estimation study conducted in Cabell County, WV in June-July 2018. Factors associated with fentanyl preference were assessed using multivariable Poisson regression with a robust variance estimate.

Results:

Among PWID who reported having ever used fentanyl (n=311), 43.4% reported preferring drugs containing fentanyl. Participants reported high levels of socioeconomic vulnerability, including homelessness (57.9%) and food insecurity (66.9%). Recent increases in drug use and injecting more than one drug in the past 6 months were reported by 27.0% and 84.2% of participants, respectively. In adjusted analyses, fentanyl preference was associated with being younger (PrR:0.98, 95% CI: 0.97–1.00), being female (PrR:1.45, 95% CI:1.14–1.83), being a Cabell county resident (PrR:0.60, 95% CI: 0.45–0.81), increased drug use in the past 6 months (PrR:1.28, 95% CI: 1.01–1.63), and injecting fentanyl in the past 6 months (PrR:1.89, 95% CI: 1.29–2.75).

Conclusion:

Fentanyl preference is highly prevalent among rural PWID in WV and associated with factors that may exacerbate overdose risks. There is an urgent need for increased access to tailored harm reduction services that address risks associated with fentanyl preference.

Keywords: fentanyl preference, PWID, rural, overdose

1. INTRODUCTION

According to the Centers for Disease Control and Prevention (CDC), opioids are the primary driver of overdose fatalities in the United States.(CDC, 2019) The current wave of opioid overdose fatalities is primarily driven by synthetic opioids, such as fentanyl (DEA, 2018; Gladden, 2016; Jones, Einstein, & Compton, 2018; Kuczynska, Grzonkowski, Kacprzak, & Zawilska, 2018). Shifts in global drug markets combined with lower production and transportation costs due to the potency of fentanyl led to the widespread penetration of clandestinely-manufactured fentanyl and its analogues in the US illicit drug market (Ciccarone, 2017; Jones et al., 2018; Mars, Rosenblum, & Ciccarone, 2018). Fentanyl entered illicit drug markets predominantly as adulterants to heroin (DEA, 2018; NFLIS, 2018). Forensic evidence suggests that states with the most reports of illicit fentanyl in drug supplies were in the South, Northeast, and Midwest of the US (NFLIS, 2018). Paired with reports of dramatic increases in opioid mortality, these data suggest that people who use drugs (PWUD) are routinely exposed to fentanyl (Ciccarone, 2019). For example, among people who used heroin that were recruited from an inpatient detoxification program in New York City, 97% tested positive for fentanyl (Palamar et al., 2019). Similarly, a study from Massachusetts found that 87% of people seeking inpatient withdrawal management tested positive for fentanyl (Kenney, Anderson, Conti, Bailey, & Stein, 2018). Rapid onset of intoxication during drug injection makes people who inject drugs (PWID) more vulnerable to overdose compared to people who use other routes of drug administration (Degenhardt et al., 2011). Thus, understanding fentanyl preference and its connection to fentanyl use among PWID may support the development of high impact overdose prevention strategies.

The extent to which PWID prefer, actively seek, and can identify fentanyl is an evolving area of scientific investigation. In two recent studies from the Eastern US, prevalence estimates for fentanyl preference varied between 26% among people who used drugs (including via injection) (Morales et al., 2019; Sherman et al., 2019) and 31% among people who use drugs via injection (Peiper et al., 2019). Further, during a brief ethnographic assessment among PWID, Ciccarone et al. (2017) found that proponents of fentanyl cited its intense rush and ability to overcome opioid tolerance as key benefits of using fentanyl. A study by McLean et al. (2019) illustrated that PWID may choose to use fentanyl over or alongside heroin and that drug market inconsistencies decrease their agency in controlling the amount of fentanyl consumed. Mars et al. (2018) further discussed the demand-led vs. supply-led nature of increased fentanyl consumption and concluded that people who use drugs identify fentanyl in their drugs with low reliability.

The current wave of opioid overdose fatalities has affected nearly every community in the United States with rural areas more deeply impacted. For example, in 2017, West Virginia (WV) had the highest rate of drug overdose deaths in the US (Mattson et al., 2017). From 2016–2017, there was a 42% increase in overdose deaths associated with synthetic opioids in WV (Mattson et al., 2017; NIDA, 2018). Across the state, rural counties have experienced a range of adverse consequences linked to fentanyl and its analogs; for example, in 2016, there were 20 fentanyl-associated overdoses in a mere 5-hour period in Cabell County (Massey et al., 2017). In response to the escalating overdose crisis, Cabell County implemented several initiatives to prevent overdose, including expanding naloxone access and implementing quick response teams to link persons who overdose to health services. Ensuring overdose prevention initiatives are tailored to meet population-level needs in rural communities also requires a better understanding of fentanyl preference among PWID.

Little epidemiological research has examined fentanyl preference in rural communities. While overdose decedent data may be used as a proxy to understand high-risk fentanyl exposure, they do not afford insights into the multiplicity of factors (e.g., socio-demographic characteristics, substance use profiles, drug use preferences, high-risk injection practices) that may be associated with fentanyl preference. This is an important gap in the literature as high prevalence of fentanyl injection among rural PWID has been demonstrated; for example, a 2018 study estimated that there were 1,857 PWID in Cabell County, WV, 56% of whom had recently injected fentanyl (Allen, O’Rourke, White, Schneider, Kilkenny, et al., 2019). Unfortunately, these data do not speak to the degree to which rural PWID may prefer fentanyl. Because fentanyl preference may lead to active fentanyl seeking and given limited agency to control the amount of fentanyl in the drug supply, PWID in rural areas who prefer fentanyl may be at increased risk of overdose. Therefore, the aim of this study is to investigate fentanyl preference and factors associated with it among a rural sample of PWID in West Virginia.

2. METHODS

2.1. Setting and sample

Data for the current study come from a PWID population estimation study conducted in Cabell County, WV in June-July 2018. Details of the parent study can be found in previous publications (Allen, Grieb, et al., 2019; Allen, O’Rourke, White, Schneider, Hazelnett, et al., 2019; Allen, O’Rourke, White, Schneider, Kilkenny, et al., 2019; Allen, White, et al., 2019; BAHI, 2019). Briefly, eligibility criteria included being at least 18 years old and having ever used drugs by any route of administration. Data were collected anonymously through audio computer-assisted self-interview (ACASI), and participants received either a $10 grocery gift card or a small snack bag for their participation. Participants were recruited at the harm reduction program at the Cabell-Huntington Health Department and in community locations where PWID were known to congregate. The survey instrument included measures for sociodemographic characteristics, substance use, access to harm reduction and drug treatment services, and HIV/HCV risk behaviors. In total, 797 surveys were completed as part of the parent study; this study uses a sub-sample (n=311) of participants who reported injecting drugs in the past 6 months and having used fentanyl in their lifetime.

This study was reviewed and approved by the Institutional Review Board at the Johns Hopkins Bloomberg School of Public Health.

2.2. Measures

Our primary outcome was fentanyl preference, defined as answering ‘yes’ to the question, “Do you prefer drugs that contain fentanyl?” which was asked among those who reported having ever used fentanyl.

Demographic variables included: age defined continuously, gender defined dichotomously as female vs. male, and Cabell County residence (yes/no). Race (White, Black or African-American, Asian, Native-American or other Pacific Islander, American Indian or Alaskan Native, Multiracial, and other) and ethnicity (Hispanic vs. non-Hispanic) were asked of all participants, but due to low variability in responses, we collapsed these two items to a single binary measure: White, non-Hispanic vs. Other. Education was dichotomized into those who reported having at least a high school education (including GED) or higher-level education vs. those with less than high school education. Relationship status was defined as reporting being single vs. being in a relationship or married. Sexual orientation was dichotomized into sexual minority (reporting being gay, lesbian, bisexual, or other) vs. heterosexual or “straight.”

Socio-economic vulnerability characteristics included: homelessness (considered self homeless vs. not), unemployment status (reported being unemployed vs. not), engagement in transactional sex work in the past 6 months (yes vs. no), and being arrested in the past 6 months (yes vs. no). We also constructed a binary measure for food insecurity that reflected if persons reported going to sleep at night hungry at least once a week vs. from a few times a month to never. Service access measures included binary (yes vs. no) items for both getting naloxone from any sources in the past 6 months and if persons ever accessed services at the Cabell-Huntington Harm Reduction Program.

The survey included multiple substance use measures, including dichotomized measures (yes vs. no) for individual drugs used and by routes of administration. Individual drugs injected in the past 6 months included: fentanyl, heroin, buprenorphine or Suboxone, prescription pain relievers, crystal methamphetamine, speedball [co-injection of cocaine and heroin], and cocaine. We created a composite variable that indicated the total number of drug types each participant reported injecting in the past 6 months (1, 2, 3, and 4 or more). Four measures of non-injection drug use in the past six months were included in the analysis: having smoked heroin and having swallowed fentanyl, painkillers, and/or buprenorphine/Suboxone. Additional substance use measures included: whether participants’ self-reported drug use level increased in the past 6 months (increased vs. decreased or stayed the same), and injection initiation as a minor (being less than 18 when first injected any drug – yes vs. no). Persons were asked to report the number of times they injected on a typical day; we then collapsed these data to four daily injection count categories (0, 1–2, 3–5, and more than 5). Attempting to quit using drugs and access to treatment were measured dichotomously (yes vs. no) via separate questions which asked whether participants tried to quit using drugs in the past 6 months and wanted treatment but were unable to get services in the past 6 months, respectively. We included two recent (past 6 months) overdose measures, witnessed non-fatal overdoses and experienced overdoses. Both questions were captured as continuous measures and collapsed into 4 categories: 0, 1–2 times, 3–5 times, and more than 5 times.

2.3. Statistical analyses

Chi-square tests for categorical variables and Kolmogorov-Smirnov tests for continuous variables were used to test for associations with fentanyl preference. Multivariable Poisson regression with robust variance estimation was conducted to examine adjusted associations with fentanyl preference (McNutt, Wu, Xue, & Hafner, 2003). Candidate variables for the multivariable analysis were those that were significant in bivariate analyses (p<.05) and those that had hypothesized conceptual associations with the outcome. In cases of a strong association between independent variables, we selected those with the strongest conceptual and statistical association with fentanyl preference. Potential interaction effects were tested by fitting separate models that contained main effects and interaction terms and testing nested models without interaction terms using Log-likelihood ratio tests. The final model was selected based on scientific merit and the Akaike Information Criterion (AIC). Final model fit was checked using the goodness-of-fit Chi-square test. Statistical analyses were performed using STATA version 14 (2014).

3. RESULTS

In total, 797 surveys were completed as part of the parent study with n=311 reflecting persons who reported having injected drugs in the last 6 months and used fentanyl in their lifetime. Among these 311 participants, 43.3% reported preferring drugs containing fentanyl (Table 1). Most of the sample was in their thirties (median: 36, interquartile range: 30–41), had at least a high school diploma or GED (71.9%), and reported being single (51.9%). The majority were Cabell County residents (88.1%), White, non-Hispanic (89.7%) and male (50.2%). Participants reported high levels of socioeconomic vulnerability, such as considering self homeless (57.9%), being unemployed (66.6%), and experiencing food insecurity (66.9%). Approximately one-fifth of our sample (20.9%) reported engaging in transactional sex in the past 6 months.

Table 1.

Fentanyl Preference among PWID in West Virginia

Characteristic Total (N=311), n (%) Prefer drugs containing fentanyl (N=135), n (%) Do not prefer drugs containing fentanyl (N=177), n (%) p-value*
Demographics
 Age, median (IQR) 36 (30–41) 35 (28–40) 37 (31–42) 0.074§
 Female 127 (40.8) 69 (51.1) 58 (32.9) 0.001
 White, non-Hispanic 279 (89.7) 114 (84.4) 165 (93.8) 0.007
 High school graduate or GED, or higher 223 (71.9) 96 (71.6) 127 (72.2) 0.920
 Single 161 (51.9) 74(55.2) 87 (49.4) 0.312
 Sexual Minority 46 (14.8) 23 (17.2) 23 (13.1) 0.315
 Cabell county resident 274 (88.1) 112 (83.0) 162 (92.1) 0.014
Socio-economic vulnerability
 Consider self-homeless 180 (57.9) 81 (60.0) 99 (56.3) 0.507
 Unemployed 207 (66.6) 97 (71.9) 110 (62.5) 0.083
 Food insecurity 208 (66.9) 93 (68.9) 115 (65.3) 0.510
 Transactional sex work, past 6 months 65 (20.9) 45 (33.3) 20 (11.4) <0.001
 Arrested in past 6 months 114 (36.7) 51 (37.8) 63 (35.8) 0.719
Access to services
 Got Naloxone, past 6 months 164 (52.9) 69 (51.5) 95 (54.0) 0.664
 Ever accessed services at the Cabell-Huntington Harm Reduction Program 224 (73.4) 93 (70.5) 131 (75.7) 0.302
Substance use
 Injection drug use, past 6 months
  Fentanyl 214 (68.8) 113 (83.7) 101 (57.4) <0.001
  Heroin 285 (91.6) 132 (97.8) 153 (86.9) 0.001
  Buprenorphine or suboxone 86 (27.7) 34 (25.2) 52 (29.6) 0.394
  Painkillers 73 (23.5) 34 (25.2) 39 (22.2) 0.533
  Crystal methamphetamine 233 (75.2) 103 (76.3) 130 (74.3) 0.685
  Heroin and cocaine together (speedball) 137 (44.1) 75 (55.6) 62 (35.2) <0.001
  Cocaine 121 (38.9) 67 (49.6) 54 (30.7) 0.001
 Number of types of drugs injected past 6 months 0.005
  1 49 (15.8) 12 (8.9) 37 (21.1)
  2 86 (27.7) 38 (28.2) 48 (27.4)
  3 62 (20.0) 24 (17.8) 38 (21.7)
  4 or more 113 (36.5) 61 (45.2) 52 (29.7)
 Number of injections per day 0.056
  0 25 (8.2) 7 (5.4) 18 (10.3)
  1–2 48 (15.7) 14 (10.8) 34 (19.4)
  3–5 142 (46.6) 67 (51.5) 75 (42.9)
  5+ 90 (29.5) 42 (32.3) 48 (27.4)
 Smoked heroin, past 6 months 55 (17.7) 32 (23.7) 23 (13.1) 0.015
 Swallowed drugs, past 6 months
  Fentanyl 34 (10.9) 24 (17.8) 10 (5.7) 0.001
  Painkillers 97 (31.2) 47 (34.8) 50 (28.4) 0.227
  Buprenorphine or suboxone 96 (30.9) 41 (30.4) 55 (31.3) 0.868
 Drug use level increased, past 6 months 84 (27.0) 51 (37.8) 33(18.8) <0.001
 Injection initiation as minor 60 (19.3) 27 (20.0) 33 (18.8) 0.782
 Tried to quit using drugs in past 6 months 231 (74.3) 100 (74.1) 131 (74.4) 0.943
 Wanted treatment, unable to get services in past 6 months 123 (39.8) 54 (40.3) 69 (39.4) 0.877
 Number of non-fatal overdoses witnessed in past 6 months 0.120
  0 75 (24.7) 28 (21.2) 47 (27.3)
  1–2 58 (19.1) 25 (18.9) 33 (19.2)
  3–5 77 (25.3) 29 (22.0) 48 (27.9)
  More than 5 94 (30.9) 50 (37.9) 44 (25.6)
 Number of overdoses experienced in past 6 months 0.180
  0 150 (48.2) 56 (41.5) 94 (53.4)
  1–2 85 (27.3) 40 (29.6) 45 (25.6)
  3–5 43 (13.8) 21 (15.6) 22 (12.5)
  More than 5 33 (10.6) 18 (13.3) 15 (8.5)
Open in a new tab
*

chi squared

§

Kolmogorov-Smirnov test

The majority of our sample reported injecting heroin or fentanyl in the past 6 months (91.6% and 68.8%, respectively), while 10.9% also reported swallowing fentanyl during the same period. Most injected more than one type of drug in the past 6 months (84.2%) and three-quarters (76.1%) of the sample reported 3 or more injections per day. Additionally, 27.0% reported increased drug use in the past 6 months. One-third of the sample witnessed five or more recent (past 6 months) non-fatal overdoses, and one in ten participants reported experiencing five or more recent non-fatal overdoses.

In binary analysis, participants that reported preferring drugs containing fentanyl were significantly more likely to be female (51.1% vs. 32.9%, p=0.001) and report recent engagement in transactional sex work (33.3% vs.11.4%, p-value < 0.001) while significantly less likely to be White, non-Hispanic (84.4% vs. 93.8%, p=0.007) and a Cabell County resident (83.0% vs. 92.1%, p=0.014). Individuals who reported preferring drugs containing fentanyl were also significantly more likely to report having recently swallowed or injected fentanyl (17.8% vs. 5.7%, p=0.001; 83.7% and 57.4%, p<0.001, respectively). Participants who reported fentanyl preference were similarly more likely to report having recently smoked or injected heroin (23.7% vs. 13.1%, p=0.015; 97.8% and 86.9%, p=0.001, respectively). Individuals who preferred drugs that contain fentanyl were more likely to report recent injection of speedball (cocaine and heroin) and cocaine (55.6% vs. 35.2%, p<0.001; 49.6% vs. 30.7%, p=0.001, respectively). Injection of more than one drug in the past 6 months was higher among participants who reported a preference for fentanyl-containing drugs (p=0.005), as well as the recent increase in drug use was more prevalent among those preferring fentanyl (37.8% vs. 18.8%, p<0.001).

Recent engagement in transactional sex work, recent cocaine and speedball injection, and the number of types of drugs recently injected were not included in the multivariate analyses due to collinearity. In adjusted analyses, fentanyl preference was associated with being younger (PrR: 0.98, 95% CI: 0.97–1.00), female (PrR: 1.45, 95% CI:1.14–1.83), not being a Cabell County resident (PrR: 0.60, 95% CI: 0.45–0.81), increased drug use in the past 6 months (PrR: 1.28, 95% CI: 1.01–1.63), and injecting fentanyl in the past 6 months (PrR: 1.89, 95% CI: 1.29–2.75). We tested whether race or age modified the association between gender and fentanyl preference, as well as whether age modified the association between race and fentanyl preferences and found no evidence of such effect.

4. DISCUSSION

Our results indicate that fentanyl preference is associated with younger age, being female, reporting recent increases in drug use, having recently injected fentanyl, and not being a Cabell County resident. These data suggest that subpopulations of rural PWID exist that prefer fentanyl and, by extension, may be at increased risk for overdose. Previously, it has been shown that in the US, women may be more likely to use nonmedical prescription opioids than heroin (Marsh, Park, Lin, & Bersamira, 2018). Research also suggests women are more likely to report pain and be prescribed opioids for pain management with higher doses (Campbell et al., 2010). In context, our findings may indicate that women prefer fentanyl due to its increased potency and increased ability to overcome opioid tolerance. However, research that examines the complex interactions between sex and environmental factors, and how they may drive substance use preferences is warranted (Siciliano, 2019). Though not statistically significant, our findings also suggest that PWID who prefer fentanyl more commonly experience non-fatal overdose than their counterparts who do not prefer fentanyl, extending previous work from Baltimore which showed a positive association between suspecting fentanyl exposure and non-fatal overdose (Park, Weir, Allen, Chaulk, & Sherman, 2018). Our finding that 43% of rural PWID reported fentanyl preference was higher than in studies among PWID from more urban areas (26% - 31%) (Peiper et al., 2019; Sherman et al., 2019). The introduction of fentanyl to drug markets dramatically changed the risk environment for PWID in the United States (Ciccarone, 2019; Mars, Bourgois, Karandinos, Montero, & Ciccarone, 2016; Mars, Ondocsin, & Ciccarone, 2018; McLean et al., 2019), yet, the extent to which PWID prefer this drug is an understudied area, particularly among rural PWID populations. More research is needed to understand and compare fentanyl preferences between rural and urban areas and ethnographic research is needed to better understand the prevalence of fentanyl preference among rural PWID as it may be a strong underlying driver of the overdose epidemic.

While the mechanisms by which fentanyl preference was associated with both self-reported increased drug use and recent injection of fentanyl cannot be disentangled using these cross-sectional data, they do suggest a complex nexus of interacting factors may be at play among rural PWID that could exacerbate overdose risks. These factors operate at the macro- and micro-levels. At the macro-level, it is plausible that rural PWID in remote areas of Appalachia are exposed to illicit drug markets that are saturated with fentanyl, resulting in a greater likelihood of fentanyl use and preference as opposed to other illicit opioids, such as heroin (DEA, 2018). At the micro-level, it is important to consider our findings relative to fentanyl potency. Fentanyl preference may be partially derived from increased tolerance to other opioids. The Appalachian region has been disproportionately affected by the opioid crisis and many areas lack access to comprehensive harm reduction services (Allen, Grieb, et al., 2019), thus, setting the stage for an on-going crisis of overdose fatalities. In addition to scaling up harm reduction services and access to naloxone, public health initiatives should be implemented that educate PWID on safer injection practices, such as performing tester shots, pushing the plunger more slowly, changing the modality of drug use, and not using alone (Ciccarone, 2017; McKnight & Des Jarlais, 2018; Peiper et al., 2019).

In addition to educating PWID on safer injection practices, the modern fentanyl era also necessitates innovation in overdose prevention. Fentanyl test strips are one such innovation that carries a great deal of promise for PWID. A study conducted by Sherman and colleagues (2019) found that fentanyl test strips not only had high sensitivity and specificity for fentanyl detection but also that 86% of PWID would use them to test for fentanyl. Further, 70% reported that they would engage in safer injection behaviors if they knew fentanyl was present in their drugs (Rouhani, Park, Morales, Green, & Sherman, 2019). Those who actively seek fentanyl may also benefit from fentanyl test strips by using them to confirm fentanyl presence in their drug and applying similar safer injection practices. Fentanyl test strips are a viable harm reduction strategy that may precipitate decreases in overdose fatalities as PWID are often unaware fentanyl is in their drugs. The fact that nearly half of PWID in our sample reported fentanyl preference and that almost seventy percent reported having injected fentanyl in the past 6 months underscores the urgent need for PWID to have access to legally sanctioned fentanyl test strips.

While we did not find significant differences in naloxone access based on fentanyl preference, it is important to note that only slightly more than half of our participants reported having gotten naloxone in the past 6 months. Further, a majority of participants reported having recently witnessed at least one non-fatal overdose and approximately half reported having recently witnessed a fatal overdose. These data reinforce the importance of increasing access to overdose prevention resources and naloxone among rural PWID.

This study is subject to several limitations. First, the cross-sectional nature of our data prevents investigation into the mechanisms underlying the associations between fentanyl preference and use of drugs containing fentanyl. However, fentanyl use (injection and swallowing) was recalled over a period of 6 months while fentanyl preference was measured on the day of the survey which provided some operationalization of the relationship between fentanyl use and preference. Directionality and reciprocity of the associations identified in our analyses should be further investigated within longitudinal epidemiological and ethnographic studies. Second, some of the subgroups within our sample have a small number of observations (e.g., racial and ethnic minorities) and limit our ability to conduct more in-depth examinations among these subgroups. Specific to the relative racial homogeneity among our sample, these data reflect the underlying population distribution of the study setting; 90.1% of the population in Cabell county is White, non-Hispanic (USCB, 2018). Future work should be conducted that oversamples minority PWID populations in rural Appalachia. Third, our outcome measure and factors associated with it are self-reported, thus subject to reporting bias. However, data were collected anonymously through ACASI, which might have mitigated risks for social desirability bias. Fourth, our data provide limited insights on whether PWID who prefer fentanyl have increased exposure to the drug compared to their counterpart PWID who do not prefer fentanyl. However, if there is misclassification in reports of fentanyl use, we expect it to be non-differential. Additionally, our outcome measure did not allow us to garner insights into what drug(s) PWID preferred who indicated they did not prefer drugs containing fentanyl.

Our study also had several strengths. We were able to explore factors associated with fentanyl preference in a setting where overdose rates are extremely high, thereby enabling us to garner insights among a population deeply affected by the opioid crisis. We also had data from a large sample of rural PWID that were sampled from areas throughout the target county. The volume of data for this study also enabled us to explore multiple factors and potential interactions.

In conclusion, our study demonstrates that fentanyl preference is highly prevalent among rural PWID in Appalachia. Fentanyl preference was associated with other factors that may exacerbate risks for overdose. While existing overdose response efforts are commendable, there is an urgent need for increased access to harm reduction services that are tailored to the needs of this population and address risks associated with fentanyl preference. Reducing overdose risks among rural PWID that prefer fentanyl will require innovation in intervention design to ensure those at elevated risk have access to harm reduction services.

Table 2.

Factors Associated with Fentanyl Preference among PWID in West Virginia, (N=311)

Crude Adjusted
Prevalence ratio for fentanyl preference 95% CI P-value Prevalence ratio for fentanyl preference 95% CI P-value
Age 0.98 0.96 – 0.99 0.007 0.98 0.97 – 1.00 0.022
Female 1.52 1.19 – 1.96 0.001 1.45 1.14–1.83 0.002
Cabell county resident 0.66 0.49 – 0.88 0.004 0.60 0.45 – 0.81 0.001
White, non-Hispanic 0.64 0.48 – 0.86 0.003 0.79 0.58 – 1.05 0.106
Drug use level increased, past 6 months 1.65 1.29 – 2.10 0.000 1.28 1.01–1.63 0.042
Injection drug use, past 6 months
 Heroin 4.00 1.37 – 11.70 0.011 2.54 0.85 – 7.60 0.096
 Fentanyl 2.35 1.59 – 3.47 <0.000 1.89 1.29 – 2.75 0.001
Open in a new tab

Acknowledgements

We are grateful to the collaboration of the Cabell Huntington Health Department, without whom, this project would not have been possible. We are especially grateful to Tim Hazelett, Thommy Hill, Tyler Deering, Kathleen Napier, Jeff Keatley, Michelle Perdue, Chad Helig, and Charles “CK” Babcock for all their support throughout the study implementation. We are also grateful for the hard work of the West Virginia COUNTS! research team: Megan Keith, Anne Maynard, Aspen McCorkle, Terrance Purnell, Ronaldo Ramirez, Kayla Rodriguez, Lauren Shappell, Kristin Schneider, Brad Silberzahn, Dominic Thomas, Kevin Williams, and Hayat Yusuf. We gratefully acknowledge the West Virginia Department of Health and Human Resources. We also wish to acknowledge Josh Sharfstein, Michelle Spencer, Dori Henry, and Akola Francis for their support throughout each phase of the study. Most importantly, we are grateful to our study participants.

Role of Funding Sources

This research was supported by a grant from the Bloomberg American Health Initiative at the Johns Hopkins Bloomberg School of Public Health to Dr. Sean T. Allen. This research has been facilitated by the infrastructure and resources provided by the Johns Hopkins University Center for AIDS Research (P30AI094189) and the District of Columbia Center for AIDS Research (AI117970). STA is also supported by the National Institutes of Health (K01DA046234). The funders had no role in study design, data collection, or in analysis and interpretation of the results, and this paper does not necessarily reflect views or opinions of the funders.

Footnotes

Conflict of Interest

SGS is an expert witness for plaintiffs in opioid litigation. The other authors declare that they have no competing interests.

References

  1. Allen ST, Grieb SM, O’Rourke A, Yoder R, Planchet E, White RH, & Sherman SG (2019). Understanding the public health consequences of suspending a rural syringe services program: a qualitative study of the experiences of people who inject drugs. Harm Reduct J, 16(1), 33. doi: 10.1186/s12954-019-0305-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Allen ST, O’Rourke A, White RH, Schneider KE, Hazelnett T, Kilkenny M, & Sherman SG (2019). Applying Population Estimation Methods in Rural America. Retrieved from [Google Scholar]
  3. Allen ST, O’Rourke A, White RH, Schneider KE, Kilkenny M, & Sherman SG (2019). Estimating the Number of People Who Inject Drugs in A Rural County in Appalachia. Am J Public Health, 109(3), 445–450. doi: 10.2105/AJPH.2018.304873 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Allen ST, White RH, O’Rourke A, Ahmad NJ, Hazelett T, Kilkenny ME, & Sherman SG (2019). Correlates of Transactional Sex Among a Rural Population of People Who Inject Drugs. AIDS Behav. doi: 10.1007/s10461-019-02612-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. BAHI. (2019). Rural Opioids Count. Estimate Services Needs in Your Rural Community Retrieved from https://americanhealth.jhu.edu/Calculator [Google Scholar]
  6. Campbell CI, Weisner C, LeResche L, Ray GT, Saunders K, Sullivan MD, … Von Korff M (2010). Age and Gender Trends in Long-Term Opioid Analgesic Use for Noncancer Pain. American Journal of Public Health, 100(12), 2541–2547. doi: 10.2105/AJPH.2009.180646 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. CDC. (2019, June 27). Drug Overdose Deaths. CDC: Opioid Overdose: Data. Retrieved February 2020, from https://www.cdc.gov/drugoverdose/data/statedeaths.html [Google Scholar]
  8. Ciccarone D (2017). Fentanyl in the US heroin supply: A rapidly changing risk environment. Int J Drug Policy, 46, 107–111. doi: 10.1016/j.drugpo.2017.06.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Ciccarone D (2019). The triple wave epidemic: Supply and demand drivers of the US opioid overdose crisis. Int J Drug Policy. doi: 10.1016/j.drugpo.2019.01.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ciccarone D, Ondocsin J, & Mars SG (2017). Heroin uncertainties: Exploring users’ perceptions of fentanyl-adulterated and -substituted ‘heroin’. Int J Drug Policy, 46, 146–155. doi: 10.1016/j.drugpo.2017.06.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. DEA. (2018). Fentanyl. from Drug Enforcement Administration https://www.deadiversion.usdoj.gov/drug_chem_info/fentanyl.pdf#search=fentanyl [Google Scholar]
  12. Degenhardt L, Bucello C, Mathers B, Briegleb C, Ali H, Hickman M, & McLaren J (2011). Mortality among regular or dependent users of heroin and other opioids: a systematic review and meta-analysis of cohort studies. Addiction, 106(1), 32–51. [DOI] [PubMed] [Google Scholar]
  13. Gladden RM (2016). Fentanyl law enforcement submissions and increases in synthetic opioid–involved overdose deaths—27 states, 2013–2014. MMWR. Morbidity and mortality weekly report, 65. [DOI] [PubMed] [Google Scholar]
  14. Jones CM, Einstein EB, & Compton WM (2018). Changes in Synthetic Opioid Involvement in Drug Overdose Deaths in the United States, 2010–2016. JAMA, 319(17), 1819–1821. doi: 10.1001/jama.2018.2844 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kenney SR, Anderson BJ, Conti MT, Bailey GL, & Stein MD (2018). Expected and actual fentanyl exposure among persons seeking opioid withdrawal management. J Subst Abuse Treat, 86, 65–69. doi: 10.1016/j.jsat.2018.01.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kuczynska K, Grzonkowski P, Kacprzak L, & Zawilska JB (2018). Abuse of fentanyl: An emerging problem to face. Forensic Sci Int, 289, 207–214. doi: 10.1016/j.forsciint.2018.05.042 [DOI] [PubMed] [Google Scholar]
  17. Mars SG, Bourgois P, Karandinos G, Montero F, & Ciccarone D (2016). The Textures of Heroin: User Perspectives on “Black Tar” and Powder Heroin in Two U.S. Cities. J Psychoactive Drugs, 48(4), 270–278. doi: 10.1080/02791072.2016.1207826 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Mars SG, Ondocsin J, & Ciccarone D (2018). Sold as heroin: perceptions and use of an evolving drug in Baltimore, MD: Journal of Psychoactive Drugs, 50(2), 167–176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Mars SG, Rosenblum D, & Ciccarone D (2018). Illicit fentanyls in the opioid street market: desired or imposed? Addiction. doi: 10.1111/add.14474 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Marsh JC, Park K, Lin Y-A, & Bersamira C (2018). Gender differences in trends for heroin use and nonmedical prescription opioid use, 2007–2014. Journal of Substance Abuse Treatment, 87, 79–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Massey J, Kilkenny M, Batdorf S, Sanders SK, Ellison D, Halpin J, … Gupta R (2017). Opioid overdose outbreak—West Virginia, August 2016. MMWR. Morbidity and mortality weekly report, 66(37), 975. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Mattson CL, Schieber L, Scholl L, Rudd RA, Seth P, Xu L, … Paulozzi LJ (2017). Annual surveillance report of drug-related risks and outcomes--United States, 2017. [Google Scholar]
  23. McKnight C, & Des Jarlais DC (2018). Being “hooked up” during a sharp increase in the availability of illicitly manufactured fentanyl: Adaptations of drug using practices among people who use drugs (PWUD) in New York City: Int J Drug Policy, 60, 82–88. doi: 10.1016/j.drugpo.2018.08.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. McLean K, Monnat SM, Rigg K, Sterner III GE, & Verdery A (2019). “You Never Know What You’re Getting”: Opioid Users’ Perceptions of Fentanyl in Southwest Pennsylvania. Substance Use & Misuse, 54(6), 955–966. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. McNutt LA, Wu C, Xue X, & Hafner JP (2003). Estimating the relative risk in cohort studies and clinical trials of common outcomes. Am J Epidemiol, 157(10), 940–943. doi: 10.1093/aje/kwg074 [DOI] [PubMed] [Google Scholar]
  26. Morales KB, Park JN, Glick JL, Rouhani S, Green TC, & Sherman SG (2019). Preference for drugs containing fentanyl from a cross-sectional survey of people who use illicit opioids in three United States cities. Drug and Alcohol Dependence, 204, 107547. doi: 10.1016/j.drugalcdep.2019.107547 [DOI] [PubMed] [Google Scholar]
  27. NFLIS. (2018). NFLIS Brief: Fentanyl and Fentanyl-Related Substances Reported in NFLIS. 2015–2016. NFLIS Brief. Retrieved from https://www.nflis.deadiversion.usdoj.gov/DesktopModules/ReportDownloads/Reports/11350_R2_NFLIS_Research_Brief_Fentanyl.pdf [Google Scholar]
  28. NIDA. (2018). West Virginia Opioid Summary. Opioid Summaries by State. Retrieved from https://www.drugabuse.gov/drugs-abuse/opioids/opioid-summaries-by-state/west-virginia-opioid-summary [Google Scholar]
  29. Palamar JJ, Salomone A, Bigiarini R, Vincenti M, Acosta P, & Tofighi B (2019). Testing hair for fentanyl exposure: a method to inform harm reduction behavior among individuals who use heroin. Am J Drug Alcohol Abuse, 45(1), 90–96. doi: 10.1080/00952990.2018.1550652 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Park JN, Weir BW, Allen ST, Chaulk P, & Sherman SG (2018). Fentanyl-contaminated drugs and non-fatal overdose among people who inject drugs in Baltimore, MD: Harm Reduction Journal, 15(1), 34. doi: 10.1186/s12954-018-0240-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Peiper NC, Clarke SD, Vincent LB, Ciccarone D, Kral AH, & Zibbell JE (2019). Fentanyl test strips as an opioid overdose prevention strategy: findings from a syringe services program in the Southeastern United States. International Journal of Drug Policy, 63, 122–128. [DOI] [PubMed] [Google Scholar]
  32. Rouhani S, Park JN, Morales KB, Green TC, & Sherman SG (2019). Harm reduction measures employed by people using opioids with suspected fentanyl exposure in Boston, Baltimore, and Providence. Harm Reduct J, 16(1), 39. doi: 10.1186/s12954-019-0311-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Scholl L, Seth P, Kariisa M, Wilson N, & Baldwin G (2019). Drug and opioid-involved overdose deaths—United States, 2013–2017. Morbidity and Mortality Weekly Report, 67(5152), 1419. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Sherman SG, Morales KB, Park JN, McKenzie M, Marshall BDL, & Green TC (2019). Acceptability of implementing community-based drug checking services for people who use drugs in three United States cities: Baltimore, Boston and Providence. Int J Drug Policy, 68, 46–53. doi: 10.1016/j.drugpo.2019.03.003 [DOI] [PubMed] [Google Scholar]
  35. Siciliano CA (2019). Capturing the complexity of sex differences requires multidimensional behavioral models. Neuropsychopharmacology, 1.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. StataCorp. (2014). Stata: Release 14 In Statistical Software. College Station, TX:: StataCorp LP. [Google Scholar]
  37. USCB. (2018). QuickFacts: Cabell County, West Virginia. Retrieved March 21st, 2019, from United States Census Bureau; https://www.census.gov/quickfacts/cabellcountywestvirginia [Google Scholar]

RESOURCES