Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Nov 1.
Published in final edited form as: Drug Alcohol Depend. 2021 Aug 28;228:109006. doi: 10.1016/j.drugalcdep.2021.109006

BUPRENORPHINE/NALOXONE ASSOCIATED WITH A REDUCED ODDS OF FENTANYL EXPOSURE AMONG A COHORT OF PEOPLE WHO USE DRUGS IN VANCOUVER, CANADA

Samantha Young a,b, Kanna Hayashi a,c, Cameron Grant a, MJ Milloy a,d, Kora DeBeck a,e, Evan Wood a,d, Nadia Fairbairn a,d
PMCID: PMC8812726  NIHMSID: NIHMS1773789  PMID: 34509737

Abstract

Background

Little is known about the relationship between opioid agonist therapy (OAT) and fentanyl use, specifically. This study aimed to estimate the association between current use of different forms of OAT, including methadone, buprenorphine/naloxone (BUP/NX), slow release oral morphine (SROM), or injectable opioid agonist treatment (iOAT), and the likelihood of a fentanyl-positive urine drug test (UDT) as compared to no OAT.

Methods

Data were obtained from three community-recruited prospective cohort studies of people who use drugs in Vancouver, Canada from December 2016 through November 2018. Using multivariable Generalized Estimating Equations (GEE), we examined the association between current use of each form of OAT, as compared to no OAT, and fentanyl-positive UDT among participants who use opioids.

Results

The 915 participants contributed 2112 UDTs over a median of two follow-up visits. The majority of UDTs (74.9 %) were positive for fentanyl. After adjustment for a priori defined confounding factors, compared to no OAT, current use of BUP/NX was associated with lower odds of fentanyl-positive UDT (odds ratio [OR] = 0.36, 95 % confidence interval [CI]: 0.22–0.58) while current use of methadone (OR = 0.84, 95 % CI: 0.65–1.07), iOAT (OR = 1.30, 95 % CI: 0.75–2.28), and SROM (OR = 1.34, 95 % CI: 0.74–2.43) were not.

Conclusions

In this cohort of people who use opioids in Vancouver, only use of BUP/NX was associated with lower odds of fentanyl-positive UDT. Our findings highlight high rates of ongoing fentanyl use despite the use of OAT and support the expansion of BUP/NX for the treatment of people who use fentanyl.

Keywords: Opiate substitution treatment, Fentanyl, Opioid addiction, Urine drug test

1. INTRODUCTION

Illicitly manufactured fentanyl is the primary driver of the overdose crisis, which continues to prematurely end the lives of over 70,000 people each year in North America (Crabtree et al., 2020; Jones et al., 2018; O’Donnell et al., 2017; Scholl et al., 2018). Fentanyl differs from heroin and other short-acting opioids in several ways, including a shorter duration of action leading to higher frequency of use, and potency estimated at 50 times higher than that of heroin (Han et al., 2019; Lambdin et al., 2019). Fentanyl is also highly lipophilic and stored in adipocytes, which can lead to protracted renal clearance (Huhn et al., 2020).

Opioid agonist therapy (OAT) is effective in reducing all-cause and overdose-related mortality, and in reducing non-prescribed opioid use (Larochelle et al., 2018; Mattick et al., 2009, 2014; Sordo et al., 2017; Wakeman et al., 2020). First-line OAT includes methadone and buprenorphine/naloxone (BUP/NX) (American Society of Addiction Medicine, 2020; British Columbia Centre on Substance Use, 2017). In some places in Canada including Vancouver, British Columbia, other forms of OAT are part of the continuum of care for opioid use disorder including injectable opioid agonist treatment (iOAT) and slow release oral morphine (SROM), (British Columbia Centre on Substance Use, 2017; Canadian Research Initiative in Substance Misuse, 2019).

Despite the ubiquity of fentanyl in the illicit drug supply in many regions for the past several years, most of our knowledge of OAT comes from studies of people who use drugs (PWUD) using heroin or prescription opioids, which predate the emergence of illicitly manufactured fentanyl. There is some emerging evidence from within treatment settings indicating that OAT remains effective in reducing non-prescribed opioid use among people with opioid use disorder (OUD) using fentanyl compared to those who are not (Stone et al., 2020; Wakeman et al., 2019). It is also known, however, that there can be high rates of concomitant non-prescribed opioid use while on OAT – up to 40% in some cases – documented outside of treatment settings (Karamouzian et al., 2020; Krause et al., 2017; Zhu et al., 2018). Our understanding of how fentanyl impacts OUD treatment is still relatively limited, and there remains a large gap in the literature as to how various forms of OAT compare to the use of no OAT in terms of exposure to fentanyl. This study therefore sought to examine the association between use of various forms of OAT and the likelihood of a fentanyl-positive urine drug test (UDT) compared to no use of OAT among a large cohort of PWUD at risk for fentanyl exposure in Vancouver, Canada. Understanding how various forms of OAT impact fentanyl exposure risk is critical to inform patient-provider decision-making around choice of treatment to reduce overdose risk in the modern drug supply landscape.

2. METHODS

2.1. Study design and data sources

The study design is a retrospective cohort study with serial cross-sectional measurements. Data were obtained from the Vancouver Injection Drug Users Study (VIDUS), AIDS Care Cohort to evaluate Exposure to Survival Services (ACCESS), and At-Risk Youth Study (ARYS), three open and ongoing community-recruited prospective cohort studies of PWUD in Vancouver, Canada. Details of the study designs have been previously described (Strathdee et al., 1998; Wood et al., 2009). In brief, since May 1996, street-based outreach and snowball sampling have been used to recruit participants located predominantly in Vancouver’s Downtown Eastside neighbourhood, where there is a high prevalence of substance use and related disorders (Linden et al., 2013). VIDUS enrols participants ≥ 18 years of age who are HIV-seronegative at baseline and report injection of any drug within the month prior to enrollment. ACCESS participants must be ≥ 18 years of age, HIV-seropositive at baseline, and report use of an illicit drug other than or in addiction to cannabis within the month prior to enrollment. The At-Risk Youth Study (ARYS) participants must be 14–26 years of age, street-involved, and report using an illicit drug other than or in addition to cannabis in the month prior to enrollment. The cohorts use a harmonized study protocol whereby participants complete an interviewer-administered questionnaire and submit a UDT at baseline and semi-annually thereafter. The questionnaire covers a wide range of personal and health-related domains including demographic data and data related to substance use patterns, risk behaviors, and addiction treatment services. Participants receive a $40 CAD honorarium per visit. The studies have annual approval from the University of British Columbia/Providence Healthcare Research Ethics Board.

2.2. Causal framework

Due to the cross-sectional study design whereby the exposure and outcome were measured at the same timepoint, temporality cannot be meaningfully established. This precludes any causal conclusions beyond whether an association exists (Rothman and Greenland, 2005). However, the basis for this study is a hypothesized causal effect of OAT on fentanyl exposure, represented in the directed acyclic graph (DAG) in Fig. 1. This proposed relationship guided our decisions on which factors required conditioning upon in our primary analytic model to mitigate confounding bias.

Fig. 1.

Fig. 1.

Open in a new tab

Directed acyclic graph (DAG) representing the proposed causal relationship between opioid agonist therapy and fentanyl exposure.

*Unmeasured outcome of interest; urine drug test is used as a surrogate measure.

2.3. Study population

Data for this study was restricted to participants with a study visit between December 2016 through November 2018, as measures for UDT are not available prior to December 2016. Participants were eligible for inclusion if they had at least one baseline or follow-up visit during this time.

The sample was further restricted to participants who reported frequent use of non-prescribed opioids in the preceding six months at least once during the study period. For the purpose of this study, the term “non-prescribed opioid” is used to reflect any opioid – including prescription pills, OAT, heroin, or fentanyl – that is not prescribed for the individual, or taken only for the experience or feeling they caused while being prescribed to the individual. Frequent use was defined as at least weekly use, which has been used in other studies as a proxy for regular exposure (Carroll et al., 2017). This was done in an effort to generate a representative sample with OUD at risk for the outcome; the parent cohorts include participants that use illicit drugs other than opioids. Participants who are in sustained remission from OUD or do not use opioids are not at risk for exposure to fentanyl. The first visit during the study period at which a participant reported at least weekly use of non-prescribed opioids served as their baseline visit for this study. All subsequent visits were included, generating additional cross-sectional measurements at which a participant may or may not have continued to use non-prescribed opioids. This is intended to enhance clinical applicability in an attempt to represent a sample of people with active OUD followed forward over time.

2.4. Measures

Based on the duration of the study period, each participant could contribute a maximum of five biannual study visits. The primary exposure and outcome variables were measured at each included study visit and treated as a distinct observation clustered at the participant level. Additional covariates of sociodemographic factors, comorbidities, drug use patterns, and dose characteristics, were also all time-updated for each corresponding follow-up visit. Ethnicity remained fixed throughout the study and was taken from the baseline visit.

2.4.1. Fentanyl UDT

The primary outcome was fentanyl-positive UDT, used as a surrogate marker for fentanyl exposure. This was measured using a multi-panel qualitative UDT, which uses BTNX Rapid Response™ Multi-Drug Test Panel (Markham, ON, Canada). This is a rapid, chromatographic immunoassay able to rapidly qualitatively and simultaneously detect numerous substances in urine; for the purpose of this study, only fentanyl was included, which detects fentanyl and norfentanyl at a cut-off of 100 ng/ml and 20 ng/mL, respectively. According to the product monograph, specificity exceeds 95% when compared to the gold standard of gas chromatograph-mass spectrometry. While detection times vary depending on a variety of pharmacokinetic factors, the fentanyl test panel is commonly accepted to detect exposure within three days (Silverstein et al., 1993).

2.4.2. Type of current OAT

The primary explanatory variable of interest was type of current OAT, categorized as a five-level variable: (1) no OAT (treated as the reference category), (2) methadone, (3) BUP/NX, (4) iOAT, and (5) SROM. This was defined based on self-report of current use and type of OAT for the treatment of opioid use from the questionnaire, and “current use” corresponded to the same time point that fentanyl UDT was measured. UDT was not used to confirm adherence as iOAT, SROM, and non-prescribed opioids could all generate a positive opiate screen.

2.4.3. Additional covariates

Additional variables included in the model were sociodemographic factors: age (per ten years older), ethnicity (white versus not white), gender (female versus not female), homelessness (yes versus no), and receipt of income assistance (yes versus no); HIV seroprevalence (positive versus negative); and substance use factors: heavy alcohol use, defined as more than 14 drinks per week or more than 4 drinks on one occasion for men or more than 7 drinks per week or 3 drinks on one occasion for women, (yes versus no) and daily stimulant use (yes versus no). All behavioral variables were treated as time-updated covariates and referred to the previous six months.

2.5. Statistical analyses

Baseline characteristics, stratified by the five-level variable of type of current OAT, are reported as counts and percentages for binary variables and median and interquartile range for the continuous variable of age.

For the primary analysis, bivariate and multivariable Generalized Estimating Equations (GEE) with a logit link function and exchangeable correlation structure were used to examine the relationship, as unadjusted and adjusted odds ratios, between type of current OAT and fentanyl-positive UDT. GEE was chosen for the analysis given that factors affecting the relationship are time-dependent measures, and it accounted for correlated data with multiple observations per individual (Hardin and Hilbe, 2013). Patient characteristics to be included in the multivariable model were selected a priori based on previously published literature and clinical expertise (Arfken et al., 2017; Greenland, 1989; Krause et al., 2017; Macmadu et al., 2017; Morales et al., 2019; Pinto et al., 2010). As represented in Fig. 1, the patient characteristics that required conditioning on due to the potential for confounding bias were adjusted for in the multivariable model; these variables are listed above in section 2.4.3. Variables in the multivariable model were tested for multicollinearity, which was not present.

As can be seen in the DAG, frequency of opioid use was hypothesized to be a mediator of the relationship between use of OAT and fentanyl exposure; therefore, it was not adjusted for in the primary multivariable analysis. However, due to the potential for other unmeasured confounding associated with frequency of opioid use and to assess whether this is the driving factor in the relationship (whereby adjustment may ameliorate a significant association), a sensitivity analysis was performed including daily non-prescribed intravenous opioid use (yes versus no) and daily non-prescribed non-intravenous opioid use (yes versus no) in the adjusted model, both of which were time-updated covariates and pertained to the previous six months.

As a secondary analysis, the correlation between self-reported known or suspected fentanyl exposure within the previous 3 days and fentanyl UDT was assessed. This was done to examine the relationship between known versus unknown fentanyl exposure.

All reported p-values are two-sided and considered significant at p < 0.05. Analyses were performed using SAS version 9.4.

3. RESULTS

In total, 915 participants were included in the study and followed for a median of two follow-up visits (interquartile range [IQR]: 1–3). The median age of participants was 38 years old with an IQR of 30–50. Of 2112 UDTs collected during the study period, 1667 (78.9%) were positive for fentanyl. Baseline participant characteristics stratified by type of current OAT are shown in Table 1. A total of 409 (45%) of participants were not on OAT at baseline while 509 (55%) were on some form of OAT. Of those on OAT at baseline, 418 (82.1%) participants were on methadone, 38 (7.5%) were on BUP/NX, 30 (5.9%) were on iOAT, and 20 (3.9%) were on SROM. Out of the total 2112 observations, 874 (41.5%) came from participants not on OAT while 979 (46.6%) were on methadone, 85 (4.0%) were on BUP/NX, 72 (3.4%) were on iOAT, and 92 (4.4%) were on SROM, with 10 (0.5%) missing data. Baseline rates of fentanyl-positive UDT were high overall, averaging 75%, although only 53% self-reported known fentanyl use in the preceding 3 days.

Table 1.

Baseline characteristics of participants, stratified by current type of opioid agonist therapy (OAT)

Characteristic Total n (%)
n = 915		 Current Type of OAT
None
n (%)
409 (44.7)		 Methadone
n (%)
418 (45.7)		 BUP/NX
n (%)
38 (4.2)		 iOAT
n (%)
30 (3.3)		 SROM
n (%)
20 (2.2)
Age
 Median (IQR) 38.2 (29.6–50.4) 33.9 (26.4–47.5) 42.1 (33.3–51.8) 29.8 (24.7–45.3) 48.9 (42.3–54.6) 39.6 (31.3–51.4)
Gender
 Female 362 (39.6) 171 (41.8) 159 (38.0) 12 (31.6) 12 (40.0) 8 (40.0)
 Other 541 (59.1) 231 (56.5) 255 (61.0) 25 (65.8) 18 (60.0) 12 (60.0)
Ethnicity
 White 429 (46.9) 184 (45.0) 193 (46.2) 19 (50) 20 (66.7) 13 (65.0)
 Non-white 478 (52.2) 220 (53.8) 222 (53.1) 19 (50) 10 (33.3) 7 (35.0)
Homelessa
 Yes 310 (33.9) 162 (39.6) 119 (28.5) 17 (44.7) 6 (20.0) 6 (30.0)
 No 599 (65.5) 244 (59.7) 297 (71.1) 20 (52.6) 24 (80.0) 14 (70.0)
Income assistancea
 Yes 864 (94.4) 377 (92.2) 403 (96.4) 34 (89.5) 30 (100.0) 20 (100.0)
 No 50 (5.5) 32 (7.8) 15 (3.6) 3 (7.9) 0 (0.0) 0 (0.0)
Opioid injectiona
 ≥ Daily 472 (51.6) 237 (57.9) 196 (46.9) 9 (23.7) 16 (53.3) 14 (70.0)
 < Daily 443 (48.4) 172 (42.1) 222 (53.1) 29 (76.3) 14 (46.7) 6 (30.0)
Opioid non-injectiona
 ≥ Daily 199 (21.7) 109 (26.7) 73 (17.5) 10 (26.3) 3 (10.0) 4 (20.0)
 < Daily 716 (78.3) 300 (73.3) 345 (82.5) 28 (73.7) 27 (90.0) 16 (80.0)
Stimulant usea
 ≥ Daily 323 (35.3) 165 (40.3) 131 (31.3) 12 (31.6) 9 (30.0) 6 (30.0)
 < Daily 592 (64.7) 244 (59.7) 287 (68.7) 26 (68.4) 21 (70.0) 14 (70.0)
Heavy alcohol usea
 ≥ Daily 128 (14.0) 65 (15.9) 55 (13.2) 4 (10.5) 2 (6.7) 2 (10.0)
 < Daily 787 (86.0) 344 (84.1) 363 (86.8) 34 (89.5) 28 (93.3) 18 (90.0)
HIV seropositive
 Yes 225 (24.6) 81 (19.8) 123 (29.4) 8 (21.1) 7 (23.3) 6 (30.0)
 No 687 (75.1) 326 (79.7) 294 (70.3) 30 (78.9) 23 (76.7) 14 (70.0)
Self-reported fentanyl useb
 Yes 488 (53.3) 235 (57.5) 203 (48.6) 16 (42.1) 18 (60.0) 16 (80.0)
 No 426 (46.6) 173 (42.3) 215 (51.4) 22 (57.9) 12 (40.0) 4 (20.0)
Fentanyl-positive UDT
 Yes 685 (74.9) 317 (77.5) 310 (74.2) 19 (50.0) 23 (76.7) 16 (80.0)
 No 230 (25.1) 92 (22.5) 108 (25.8) 19 (50.0) 7 (23.3) 4 (20.0)
Open in a new tab

OAT = opioid agonist therapy; BUP/NX = buprenorphine/naloxone; iOAT = intravenous opioid agonist treatment; SROM = slow release oral morphine; IQR = interquartile range; UDT = urine drug test.

a

Behavioural variables refer to the 6 months prior to the follow-up questionnaire.

b

Pertains to the preceding 3-day period.

A total of 596 participants were on OAT at some point during the study period. At their first visit on OAT, 435 (73.0%) had UDTs positive for fentanyl. Subsequent UDTs can be seen in Table 2, stratified by whether or not they remained on OAT at future study visits. Participants who subsequently discontinued OAT had a higher percentage of UDTs positive for fentanyl, although this was not statistically compared due to low numbers. Rates of fentanyl-positive UDTs remained high – above 75% – throughout the study for both those who continued and discontinued OAT.

Table 2.

Chronologic fentanyl-positive urine drug tests after first study visit on opioid agonist therapy during the study period (n = 596)

Study visit On OAT Not on OAT
UDT positive
n (%)		 UDT negative
n (%)		 Number of participants
n		 UDT positive
n (%)		 UDT negative
n (%)		 Number of participants
1 435 (73.0) 161 (27.0) 596 – – 0
2 290 (84.3) 54 (15.7) 344 95 (93.1) 7 (6.9) 102
3 157 (80.5) 38 (19.5) 195 32 (91.4) 3 (8.6) 35
4 74 (79.6) 19 (20.4) 93 5 (83.3) 1 (16.7) 6
Open in a new tab

OAT = opioid agonist therapy; UDT = urine drug test.

While all participants reported the use of non-prescribed opioids at least weekly at baseline, 93 (10.2%) participants subsequently reported no non-prescribed opioid use in the previous six months at a later follow-up visit. Of those participants who reported six-month abstinence from non-prescribed opioids, 39 reported no current OAT use, 39 were on methadone, 7 were on BUP/NX, 1 was on SROM, and 7 were on iOAT at baseline.

Bivariate and multivariable GEE analyses are shown in Table 3. Compared to no current OAT use, current use of methadone (odds ratio [OR] = 0.84, 95% confidence interval [CI]: 0.65–1.07), iOAT (OR = 1.30, 95% CI: 0.75–2.28), or SROM (OR = 1.34, 95% CI: 0.74–2.43) were not significantly associated with fentanyl-positive UDT in the adjusted model. However, use of BUP/NX was associated with significantly lower odds of fentanyl-positive UDT (OR = 0.36, 95% CI: 0.22–0.58) even after adjustment for confounding. In the sensitivity analysis adding daily injection opioid use and daily non-injection opioid use in the adjusted model, current use of BUP/NX remained significantly associated with fentanyl-positive UDT (OR = 0.45, 95% CI: 0.27–0.75) compared to no current OAT use, while methadone (OR = 1.10, 95% CI: 0.84–1.45), iOAT (OR = 1.84, 95 % CI: 0.96–3.55), and SROM (OR = 1.80, 95% CI: 0.90–3.62) were not.

Table 3.

Type of opioid agonist therapy and related characteristics associated with fentanyl-positive urine drug test - bivariate and multivariable generalized estimating equations.

Characteristic Unadjusted Odds Ratio (95 % CI) p-value Adjusted Odds Ratio (95 % CI) p-value
Current use of methadonea 0.76 (0.60–0.97) 0.025 0.84 (0.65–1.07) 0.16
Current use of BUP/NXa 0.38 (0.25–0.60) <0.001 0.34 (0.22–0.58) <0.001
Current use of iOATa 1.15 (0.67–1.95) 0.62 1.30 (0.75–2.28) 0.35
Current use of SROMa 1.20 (0.67–2.16) 0.54 1.34 (0.74–2.43) 0.33
Ageb 0.98 (0.97–0.99) 0.001 0.98 (0.97–0.99) 0.002
Female gender (yes vs no) 1.48 (1.11–1.98) 0.008 1.34 (0.99–1.82) 0.058
Ethnicity (white vs other) 1.09 (0.83–1.43) 0.52 1.03 (0.78–1.36) 0.82
Homelessc (yes vs no) 1.20 (0.93–1.55) 0.16 1.06 (0.81–1.39) 0.67
Income assistancec (yes vs no) 1.62 (0.99–2.64) 0.053 1.78 (1.06–2.98) 0.028
Daily stimulant use (yes vs no) 1.26 (1.01–1.58) 0.040 1.19 (0.95–1.51) 0.14
Heavy alcohol used (yes vs no) 0.73 (0.55–0.98) 0.039 0.70 (0.52–0.96) 0.027
HIV seropositive (yes vs no) 1.05 (0.78–1.42) 0.74 1.17 (0.85–1.61) 0.35
Open in a new tab

CI = confidence interval; OAT = medication for opioid use disorder; BUP/NX = buprenorphine/naloxone

a

Compared to reference category of no current OAT.

b

Per ten years older.

c

Variables refer to the 6 months prior to the follow-up questionnaire.

d

Heavy alcohol use is defined more than 14 drinks per week or more than 4 drinks on one occasion for men or more than 7 drinks per week or 3 drinks on one occasion for women.

Of the 1667 UDTs that were positive for fentanyl, 1145 (68.7%) came from participants who reported known or suspected fentanyl exposure within the previous three days, while 520 (31.1%) had unknown exposure; two were missing data on self-reported fentanyl use. Only 79 (3.7%) of UDTs came from participants who reported fentanyl exposure within the previous three days but were negative for fentanyl. The correlation between self-reported fentanyl exposure and UDT was 0.42.

4. DISCUSSION

In this population of people with frequent non-prescribed opioid use at baseline, BUP/NX was found to be associated with lower odds of a fentanyl-positive UDT compared to no OAT, while methadone, iOAT, and SROM – all of which are full opioid agonists – were not. This finding is surprising given the extensive literature on the efficacy of OAT to reduce opioid-related morbidity and mortality, even in the ‘fentanyl era’ (Larochelle et al., 2018; Wakeman et al., 2020). However, it is important to note that our study is not meant to nor designed to comment on the efficacy of OAT in terms of abstinence, overdose, or mortality, and can only comment on the relationship to fentanyl exposure. This result may imply that the known protective nature of full opioid agonists against overdose and mortality are mediated by factors other than cessation of illicit opioid use. Our finding that rates of fentanyl-positive UDTs remained high even among participants who remained on OAT at subsequent visits seems to support this notion. Participants who remain on OAT but continue to use fentanyl are likely deriving some benefit beyond complete abstinence to remain engaged in care.

Our findings suggest that use of BUP/NX is associated with factors that are protective against exposure to fentanyl, which could be related to its pharmacologic properties or patient-related factors. Buprenorphine is a partial agonist with a high binding affinity for the mu-opioid receptor (Lewis, 1985). Therefore, patients on buprenorphine-containing medications have higher relative blockade at the mu-opioid receptor compared to those on methadone (Volpe et al., 2011), which may disincentivize ongoing fentanyl use. At baseline, 24% of participants on BUP/NX were using injection opioids at least daily compared to 52% of those on methadone and 58% on no OAT. The association remained significant in sensitivity analyses after adjusting for daily opioid use, indicating that factors other than frequency of opioid use are likely driving this relationship, although it would not account for frequency of use beyond daily. Additionally, fentanyl is highly lipophilic which which is hypothesized to have made the buprenorphine induction process more challenging due to a higher incidence of precipitated withdrawal (Brico, 2020; Huhn et al., 2020). Therefore, people who use fentanyl may be more likely to self-select full agonist medications to avoid the potential for precipitated withdrawal. It is also possible that patients who select for BUP/NX as OAT have other factors that portend a lower odds of fentanyl exposure. Prior literature indicates that patients choosing methadone over BUP/NX may have more severe substance use and a higher incidence of psychiatric and physical comorbidities (Pinto et al., 2010). Patients choosing BUP/NX over other forms of OAT may also be more inclined toward a goal of abstinence rather than aiming to reduce their use without full cessation.

Two separate cohort studies recruited from OAT treatment programs – one among patients on methadone and one among patients on BUP/NX – have shown no difference in rates of achieving abstinence between patients who were fentanyl-exposed and non-exposed upon intake (Stone et al., 2020; Wakeman et al., 2019). However, in the study of patients in methadone maintenance, less than 50% achieved sustained remission during one-year follow-up and 69% of those who tested positive for fentanyl at intake returned to fentanyl use at some point during follow-up (Stone et al., 2020). This highlights that ongoing non-prescribed opioid use is common during OAT treatment as was seen in our study, where only 10% of participants reported a six-month period of abstinence during study follow-up, although this is subject to bias from differential loss to follow-up with uncertain directionality.

During the study, just under 60% of participants were on OAT. Methadone was by far the most commonly used medication, with less than 5% of participants on BUP/NX. Numbers of participants on SROM and iOAT were similarly small although confidence intervals were considerably wider, making conclusions less reliable. The number of participants on BUP/NX is fewer than expected compared to regional data from British Columbia which estimate approximately 18% of OAT was BUP/NX in 2016 (Office of the Provincial Health Officer, 2017). Studies in the United States have shown that access to BUP/NX is more common in higher socioeconomic areas, while nearly all participants in this study were on income assistance and one third were experiencing homelessness (Hansen et al., 2013; Lagisetty et al., 2019). In our setting in Canada, where all forms of OAT except iOAT are available for dispensation at pharmacies and both BUP/NX and methadone are considered first-line OAT, this may be more reflective of local preferences of PWUD; a recent study in this population found that less than 20% of people who use opioids in Vancouver were willing to try BUP/NX (Weicker et al., 2019). Despite a potential benefit in reducing fentanyl use as seen in this study and the favorable safety profile of BUP/NX compared to other full-agonist forms of OAT (American Society of Addiction Medicine, 2020; British Columbia Centre on Substance Use, 2017), our study indicates that uptake of BUP/NX was extremely low among this cohort of PWUD, many of whom face a high degree of structural marginalization. While our findings support the expansion of BUP/NX, they also indicate that many people who use opioids may prefer a full agonist medication, and support the need for expansion of harm reduction services and alternatives to the illicit drug supply to reduce the harms of fentanyl use that may continue despite the provision of OAT. Additionally, our study indicates that a large portion (40%) of people who use opioids remain disengaged with OAT, which may reflect barriers to access that require further attention.

Our findings of 75% fentanyl-positive UDTs among participants using opioids at baseline is consistent with rates of fentanyl contamination seen within Vancouver’s illicit opioid supply at the time of the study, which approached 80 % (Karamouzian et al., 2018). The correlation between self-reported exposure and UDT for fentanyl was moderate, with over 30% of participants testing positive for fentanyl despite no known fentanyl use within the prior three days (Schober et al., 2018). Some of this may reflect known fentanyl ingestion more than three days prior with protracted renal excretion of fentanyl (Huhn et al., 2020). However, some is likely due to unknown exposure of fentanyl, underscoring risk for overdose and other harms of a contaminated street drug supply.

4.1. Limitations

There are important limitations to this study. First, the cohorts are not random samples and may not be representative of all PWUD in Vancouver or elsewhere. Additionally, by restricting our sample to participants with regular non-medical opioid use at baseline in an attempt to mitigate the potential bias related to more stable patients, it limits the generalizability to all people on OAT. All data aside from UDT are based on self-report and therefore subject to recall bias. UDTs also have the potential for false positives and negatives, although based on the high prevalence of fentanyl in the opioid supply in Vancouver, the high pre-test probability of a true positive within our sample, and the detection limits of the test we anticipate this to be minimal (Karamouzian et al., 2018; Sherman et al., 2018). Further, due to the lipophilic nature of fentanyl leading to protracted renal clearance in people using regularly (Huhn et al., 2020), UDTs may overestimate recent exposure among those with frequent use and under-estimate exposure among those using less frequently. The use of biannual UDTs also leaves us unable to comment on fentanyl use between study visits, although our multivariable model did control for several substance use-related variables that corresponded to the preceding six months. Despite our attempt to control confounding factors, it is possible that some confounders have not been accounted for. This may include some confounding factors that affect the determination of which form of OAT was selected for a patient, although we have attempted to interpret the results in the context of these differences that may influence medication choice. Future work is required to compare the relative effectiveness of BUP/NX versus methadone in key clinical outcomes for individuals with OUD using fentanyl. We did not adjust for dose of OAT, and it is possible that being on a subtherapeutic dose of methadone or other forms of OAT biased our findings toward the null (Faggiano et al., 2003). Finally, as mentioned, causality cannot be inferred although we hypothesize that there are protective factors associated with use of BUP/NX that may lead to decreased fentanyl exposure.

4.2. Conclusions

Among people with regular use of non-prescribed opioids within a highly fentanyl-contaminated drug supply in Vancouver, Canada, BUP/NX was associated with lower odds of a fentanyl-positive UDT while other forms of OAT were not. The results of our study support the use of BUP/NX as OAT in patients with potential exposure to fentanyl, but due to the observational nature of the study do not imply superiority to methadone or other forms of OAT (Gueyffier and Cucherat, 2019). Our study also highlights that many people who use opioids will continue to use non-prescribed opioids despite receiving OAT. Further research is needed to elucidate the potentially protective role of buprenorphine-based OAT and relative efficacy versus full-agonist forms of OAT among patients who use opioids at risk for fentanyl exposure.

Highlights.

  • We examined the association of opioid agonist therapy (OAT) and fentanyl in urine.

  • OAT included methadone, buprenorphine, slow release morphine, and injectable OAT.

  • Three quarters of urines in this study were positive for fentanyl.

  • Only buprenorphine was associated with lower odds of fentanyl-positive urine.

  • Buprenorphine may reduce fentanyl use via pharmacologic or patient characteristics.

Acknowledgements

We gratefully acknowledge this work was conducted on the traditional, ancestral, unceded and continually occupied territory of the Coast Salish Peoples, including the unceded homelands of the xʷməθkwəýəm (Musqueam), Sḵwx̱wú7mesh (Squamish), and səĺílwətaɬ (Tsleil-Waututh) Nations. The authors thank the study participants for their contributions to the research, as well as current and past researchers and staff. We also thank Ekaterina Nosova for her input.

Funding

The study was supported by the US National Institutes of Health (NIH) (U01DA0251525, U01DA038886). This research was undertaken, in part, thanks to funding from the Canada Research Chairs program through a Tier 1 Canada Research Chair in Inner City Medicine which supports Dr. Evan Wood, as well as the Canadian Institutes of Health Research (CIHR) Canadian Research Initiative on Substance Misuse (SMN-139148). Dr. Evan Wood is also the Chief Medical Officer of Numinus Wellness, a mental health company focussed on psychedelic medicines. Dr. Nadia Fairbairn is supported by a Michael Smith Foundation for Health Research (MSFHR)/St. Paul’s Foundation Scholar Award. Dr. M-J Milloy is supported by the United States National Institutes of Health (U01DA0251525), a New Investigator award from the Canadian Institutes of Health Research, and a Scholar Award from MSFHR. He is the Canopy Growth professor of cannabis science at the University of British Columbia, a position established by arms’ length gifts to the university from Canopy Growth, a licensed producer of cannabis, and the Government of British Columbia’s Ministry of Mental Health and Addictions. Dr. Kanna Hayashi holds the St. Paul’s Hospital Chair in Substance Use Research and is supported in part by a NIH grant (U01DA038886), a CIHR New Investigator Award (MSH141971), a MSFHR Scholar Award, and the St. Paul’s Foundation. Dr. Kora DeBeck is supported by a MSFHR/St. Paul’s Hospital Foundation-Providence Health Care Career Scholar Award and a CIHR New Investigator Award. Dr. Samantha Young participates in the Research in Addiction Medicine Scholars (RAMS) program funded by the National Institute on Drug Abuse (NIDA) grant R25DA033211 and is supported by the University of British Columbia Clinician Investigator Program.

Footnotes

Declaration of Competing Interest

No conflict declared.

REFERENCES

  1. American Society of Addiction Medicine, 2020. The ASAM National Practice Guideline For the Treatment of Opioid Use Disorder: 2020 Focused Update. [DOI] [PubMed]
  2. Arfken CL, Suchanek J, Greenwald MK, 2017. Characterizing fentanyl use in methadone-maintained clients. J. Subst. Abuse Treat 75, 17–21. [DOI] [PubMed] [Google Scholar]
  3. Brico E, 2020. Starting Bupe From Fentanyl Can Be a Nightmare. Microdosing Methods Help. Filter [Google Scholar]
  4. British Columbia Centre on Substance Use, 2017. A Guideline for the Clinical Management of Opioid Use Disorder. https://www.bccsu.ca/wp-content/uploads/2017/06/BC-OUD-Guidelines_June2017.pdf (Accessed 9 April 2020).
  5. Canadian Research Initiative in Substance Misuse, 2019. National Injectable Opioid Agonist Treatment for Opioid Use Disorder Clinical Guideline. https://crism.ca/projects/ioat-guideline/ (Accessed 9 April 2020).
  6. Carroll JJ, Marshall BDL, Rich JD, Green TC, 2017. Exposure to fentanyl-contaminated heroin and overdose risk among illicit opioid users in Rhode Island: A mixed methods study. Int. J. Drug Policy 46, 136–145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Crabtree A, Lostchuck E, Chong M, Shapiro A, Slaunwhite A, 2020. Toxicology and prescribed medication histories among people experiencing fatal illicit drug overdose in British Columbia, Canada. Can. Med. Assoc. J 192 (34), E967–E972. [Google Scholar]
  8. Faggiano F, Vigna-Taglianti F, Versino E, Lemma P, 2003. Methadone maintenance at different dosages for opioid dependence. Cochrane Database Syst. Rev (3), CD002208. [DOI] [PubMed] [Google Scholar]
  9. Greenland S, 1989. Modeling and variable selection in epidemiologic analysis. Am. J. Public Health 79 (3), 340–349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gueyffier F, Cucherat M, 2019. The limitations of observation studies for decision making regarding drugs efficacy and safety. Therapie 74 (2), 181–185. [DOI] [PubMed] [Google Scholar]
  11. Han Y, Yan W, Zheng Y, Khan MZ, Yuan K, Lu L, 2019. The rising crisis of illicit fentanyl use, overdose, and potential therapeutic strategies. Transl. Psychiatry 9 (1), 282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hansen HB, Siegel CE, Case BG, Bertollo DN, DiRocco D, Galanter M, 2013. Variation in use of buprenorphine and methadone treatment by racial, ethnic, and income characteristics of residential social areas in New York City. J. Behav. Health Serv. Res 40 (3), 367–377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hardin JW, Hilbe JM, 2013. Generalized Estimating Equations, second edition. CRC Press, Boca Raton, FL. [Google Scholar]
  14. Huhn AS, Hobelmann JG, Oyler GA, Strain EC, 2020. Protracted renal clearance of fentanyl in persons with opioid use disorder. Drug Alcohol Depend. 214, 108147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. 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] [PMC free article] [PubMed] [Google Scholar]
  16. Karamouzian M, Dohoo C, Forsting S, McNeil R, Kerr T, Lysyshyn M, 2018. Evaluation of a fentanyl drug checking service for clients of a supervised injection facility, Vancouver, Canada. Harm. Reduct. J 15 (1), 46. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Karamouzian M, Papamihali K, Graham B, Crabtree A, Mill C, Kuo M, Young S, Buxton JA, 2020. Known fentanyl use among clients of harm reduction sites in British Columbia, Canada. Int. J. Drug Policy 77, 102665. [DOI] [PubMed] [Google Scholar]
  18. Krause D, Plorer D, Koller G, Martin G, Winter C, Adam R, Canolli M, Al-Iassin J, Musselmann R, Walcher S, Schafer F, Pogarell O, 2017. High concomitant misuse of fentanyl in subjects on opioid maintenance treatment. Subst. Use Misuse 52 (5), 639–645. [DOI] [PubMed] [Google Scholar]
  19. Lagisetty PA, Ross R, Bohnert A, Clay M, Maust DT, 2019. Buprenorphine treatment divide by Race/Ethnicity and payment. JAMA Psychiatry. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lambdin BH, Bluthenthal RN, Zibbell JE, Wenger L, Simpson K, Kral AH, 2019. Associations between perceived illicit fentanyl use and infectious disease risks among people who inject drugs. Int. J. Drug Policy 74, 299–304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Larochelle MR, Bernson D, Land T, Stopka TJ, Wang N, Xuan Z, Bagley SM, Liebschutz JM, Walley AY, 2018. Medication for opioid use disorder after nonfatal opioid overdose and association with mortality: a cohort study. Ann. Intern. Med 169 (3), 137–145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lewis JW, 1985. Buprenorphine. Drug Alcohol Depend. 14 (3–4), 363–372. [DOI] [PubMed] [Google Scholar]
  23. Linden IA, Mar MY, Weker GR, Jang K, Krausz M, 2013. Research on a vulnerable neighborhood-the vancouver downtown eastside from 2001 to 2011. J. Urban Health 90 (3), 559–573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Macmadu A, Carroll JJ, Hadland SE, Green TC, Marshall BD, 2017. Prevalence and correlates of fentanyl-contaminated heroin exposure among young adults who use prescription opioids non-medically. Addict. Behav 68, 35–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Mattick RP, Breen C, Kimber J, Davoli M, 2009. Methadone maintenance therapy versus no opioid replacement therapy for opioid dependence. Cochrane Database Syst. Rev 3, CD002209. [DOI] [PubMed] [Google Scholar]
  26. Mattick RP, Breen C, Kimber J, Davoli M, 2014. Buprenorphine maintenance versus placebo or methadone maintenance for opioid dependence. Cochrane Database Syst. Rev 2, CD002207. [DOI] [PubMed] [Google Scholar]
  27. 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 Alcohol Depend. 204, 107547. [DOI] [PubMed] [Google Scholar]
  28. O’Donnell JK, Halpin J, Mattson CL, Goldberger BA, Gladden RM, 2017. Deaths involving fentanyl, fentanyl analogs, and U-47700–10 states, july-december 2016. MMWR Morb. Mortal. Wkly. Rep 66 (43), 1197–1202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pinto H, Maskrey V, Swift L, Rumball D, Wagle A, Holland R, 2010. The SUMMIT trial: a field comparison of buprenorphine versus methadone maintenance treatment. J. Subst. Abuse Treat 39 (4), 340–352. [DOI] [PubMed] [Google Scholar]
  30. Rothman KJ, Greenland S, 2005. Causation and causal inference in epidemiology. Am. J. Public Health 95 (Suppl 1), S144–150. [DOI] [PubMed] [Google Scholar]
  31. Schober P, Boer C, Schwarte LA, 2018. Correlation coefficients: appropriate use and interpretation. Anesth. Analg 126 (5), 1763–1768. [DOI] [PubMed] [Google Scholar]
  32. Scholl L, Seth P, Kariisa M, Wilson N, Baldwin G, 2018. Drug and opioid-involved overdose deaths - United States, 2013–2017. MMWR Morb. Mortal. Wkly. Rep 67 (5152), 1419–1427. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Sherman SGP, Glick J, McKenzie M, Morales K, Christensen T, 2018. FORECASE Study Summary Report. https://americanhealth.jhu.edu/sites/default/files/inline-files/Fentanyl_Executive_Summary_032018.pdf (Accessed 6 April 2020).
  34. Silverstein JH, Rieders MF, McMullin M, Schulman S, Zahl K, 1993. An analysis of the duration of fentanyl and its metabolites in urine and saliva. Anesth. Analg 76 (3), 618–621. [DOI] [PubMed] [Google Scholar]
  35. Sordo L, Barrio G, Bravo MJ, Indave BI, Degenhardt L, Wiessing L, Ferri M, Pastor-Barriuso R, 2017. Mortality risk during and after opioid substitution treatment: systematic review and meta-analysis of cohort studies. BMJ 357, j1550. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Stone ACC, Rich JD, Green TC, 2020. One year of methadone maintenance treatment in a fentanyl endeic area: safety, repeated exposure, retention, and remission. J. Subst. Abuse Treat [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Strathdee SA, Palepu A, Cornelisse PG, Yip B, O’Shaughnessy MV, Montaner JS, Schechter MT, Hogg RS, 1998. Barriers to use of free antiretroviral therapy in injection drug users. JAMA 280 (6), 547–549. [DOI] [PubMed] [Google Scholar]
  38. Volpe DA, McMahon Tobin GA, Mellon RD, Katki AG, Parker RJ, Colatsky T, Kropp TJ, Verbois SL, 2011. Uniform assessment and ranking of opioid mu receptor binding constants for selected opioid drugs. Regul. Toxicol. Pharmacol 59 (3), 385–390. [DOI] [PubMed] [Google Scholar]
  39. Wakeman SE, Chang Y, Regan S, Yu L, Flood J, Metlay J, Rigotti N, 2019. Impact of fentanyl use on buprenorphine treatment retention and opioid abstinence. J. Addict. Med 13 (4), 253–257. [DOI] [PubMed] [Google Scholar]
  40. Wakeman SE, Larochelle MR, Ameli O, Chaisson CE, McPheeters JT, Crown WH, Azocar F, Sanghavi DM, 2020. Comparative effectiveness of different treatment pathways for opioid use disorder. JAMA Netw Open 3 (2), e1920622. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Weicker SA, Hayashi K, Grant C, Milloy MJ, Wood E, Kerr T, 2019. Willingness to take buprenorphine/naloxone among people who use opioids in Vancouver, Canada. Drug Alcohol Depend. 205, 107672. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Wood E, Kerr T, Marshall BD, Li K, Zhang R, Hodd RS, Harrigan PR, Montaner JS, 2009. Longitudinal community plasma HIV-1 RNA concentrations and incidence of HIV-1 among injecting drug users: prospective cohort study. BMJ 338, b1649. 10.1136/bmj.b1649. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Zhu Y, Evans EA, Mooney LJ, Saxon AJ, Kelleghan A, Yoo C, Hser YI, 2018. Correlates of long-term opioid abstinence after randomization to methadone versus buprenorphine/naloxone in a multi-site trial. J. Neuroimmune Pharmacol 13 (4), 488–497. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES