Abstract
Background:
In 2018, there were over 67,000 drug overdose deaths in the United States, with almost half involving illicit fentanyl and other synthetic opioids. While overall age-adjusted drug overdose deaths decreased by 4.6% from 2017 to 2018, synthetic opioid deaths increased 10.0%. This pilot study evaluates the impact of a brief fentanyl test strip (FTS) intervention to increase fentanyl awareness and reduce overdose risk.
Methods:
Female sex workers (FSW) reporting past month illicit opioid use were recruited between April 2018 through February 2019 in Baltimore City, Maryland. At baseline, they completed a baseline survey, and received tailored harm reduction messaging, 5 FTS and training, and a naloxone kit, then completed a survey after one month. McNemar's test was used to compare repeated measures.
Results:
Among N = 103, 54% were <40 years, 59% were white, and 24% had overdosed in the past year. Among 68 who completed follow-up, most (84%) used ≥1 FTS to test their drugs, 86% had ≥1 fentanyl-positive result, 57% were surprised by the result, and 69% engaged in harm reduction behaviors following the result (e.g., asked someone to check on them, did a tester shot, used a smaller amount). Significant pretest–posttest reductions in daily illicit opioid use (77% to 56%; p = 0.003), injection frequency (40% to 25%; p = 0.004), benzodiazepine use (22% to 7%; p = 0.008), and solitary drug use (96% vs. 68%; p < 0.001) were observed. No change in preferring drugs containing fentanyl was found. Some (18%) gave their FTS to others. All but three (96%) reported being likely to use FTS in the future.
Conclusions:
We found high FTS acceptability and reductions in drug use frequency and solitary drug use following FTS use among FSW who use drugs in Baltimore. These findings demonstrate that FTS-based interventions hold potential in reducing overdose risk.
Keywords: Fentanyl, Opioid overdose, Substance use
1. Introduction
Drug overdoses contributed to a two-year decline in U.S. life expectancy in 2016 and 2017 (Wonder, 2018). The emergence and rise of illicitly manufactured fentanyl and related analogs (IMFA) in the current crisis beginning in 2013 profoundly shaped national overdose mortality trends. In 2016, IMFA and other synthetic opioids surpassed heroin and prescription opioids as the leading cause of drug overdose mortality (Hedegaard et al., 2018). Of the 67,367 drug overdose deaths recorded in 2018, almost half involved synthetic opioids excluding methadone. Although drug overdose deaths decreased by 4.6% from 2017 to 2018, synthetic opioid deaths (excluding methadone) increased by 10.0% (Hedegaard et al., 2020). IMFA are substantially more potent by weight than heroin, are rapidly absorbed, and are appearing in a growing number of illicit drugs, including heroin, cocaine, and counterfeit opioid pills (Mars et al., 2018; DEA, 2018). The unregulated manufacturing, distribution and marketing involved in the illicit drug economy gives rise to products of unpredictable purity and dosage (Allen et al., 2019; Suzuki and El-Haddad, 2017), posing high mortality risk to a wide range of drug-using communities (Ciccarone, 2017).
Checking drugs for the presence of fentanyl prior to use could help to reduce overdose risk among people who use drugs (PWUD). Drug checking programs are well-established in Europe and Canada (Barratt et al., 2018; Harper et al., 2017; Tupper et al., 2018), and have been gaining traction in the U.S., particularly at outdoor dance festivals and at harm reduction organizations, including syringe services programs (SSP). Fentanyl test strips (FTS) are a rapid, low-cost ($1/strip), and easy-to-use immunoassay-based screening tool originally designed for the detection of fentanyl and related analogs in human urine. However, in response to fentanyl supply shocks, harm reduction organizations in a growing number of locations began distributing FTS as an off-label drug checking tool to PWUD to provide a direct method of testing drug samples (McGowan et al., 2018; Peiper et al., 2019; Krieger et al., 2018). One study has conducted a preliminary validation of the use of FTS for this purpose, demonstrating high sensitivity (98%) and specificity (94%) at detecting IMFA in street opioids compared to Gas Chromatograph/Mass Spectrometry (GC/MS) (Sherman et al., 2018).
FTS programs appear to be acceptable among PWUD and service providers and have been shown to be associated with protective behaviors in cross-sectional evaluations (Glick et al., 2019; Goldman et al., 2019; Green et al., 2020; Krieger et al., 2018; Schneider, Park, Allen, Weir, & Sherman, 2019; Peiper et al., 2019). A cross-sectional study of people who inject drugs (PWID) demonstrated that receiving a fentanyl-positive FTS result was associated with a five-fold higher odds of engaging in harm reduction behaviors such as using less than usual and using drugs more slowly (Peiper et al., 2019). Another pilot study conducted in Rhode Island documented high FTS uptake (77%) and acceptability (95%) among young adult PWUD (n = 87); at one-month follow-up, those who received a fentanyl-positive result (n = 31) were more likely to engage in harm reduction behaviors (68% vs. 32%) (Krieger et al., 2018). One study in Canada showed low uptake among clients of a supervised injection facility (Karamouzian et al., 2018).
The impact of FTS distribution as an overdose risk reduction strategy requires additional investigation, particularly over time and among subpopulations-at-risk. Urban female sex workers (FSW) who use drugs are at high risk of a host of conditions including overdose, homelessness, and trauma, and are dually criminalized in most settings (Park et al., 2019). Studies have shown high levels of risky drug use and overdose, as well as the direct exchange of sex for drugs among this population (Goldenberg et al., 2020; Seal et al., 2001; Inciardi and Surratt, 2001; Lorvick et al., 2018). We have documented high levels of drug use and overdose among street-based cis-gender FSW in Baltimore City, Maryland with two-thirds reporting heroin injection, 84% reporting smoking crack cocaine and 55% having a history of overdose (Sherman et al., 2019; Park et al., 2019; Rouhani et al., 2020). We have also showed that FSW in this study often engaged in sex work to financially support their dependence on drugs (Sherman et al., 2019).
The current study evaluates the impact of a brief FTS intervention among FSW who use opioids (N = 103) in Baltimore City, which was nested within a larger ongoing cohort study (N = 387) of street-based FSW. We hypothesized that FTS-based interventions would reduce the occurrence of non-fatal overdose and overdose risk behaviors (i.e., frequency of substance use, using drugs alone) after one month of follow-up. We also explored changes in service access, including mental health services, SSP and medication-assisted treatment (MAT).
2. Methods
The Fentanyl Innovative Testing (FIT) study is a pilot study nested within the Enabling Mobilization Empowerment, Risk reduction And Lasting Dignity (EMERALD) study, an ongoing prospective cohort examining the impact of a community level intervention on the HIV/STI risk of FSW (Tomko et al., 2020). Women were recruited through targeted sampling (Allen et al., 2019) using a mobile van in several neighborhoods throughout Baltimore, Maryland with high concentrations of drug and sex markets, between April 2018 and February 2019. Eligibility criteria for the EMERALD study were: 1) aged 18 or older; 2) self-identifying as a cis-gender woman; 3) sold or traded oral, vaginal, or anal sex worded in the screener as “for money or things like food, drugs, or favors in the past 6 months;” 4) picked up clients 3 or more times in the past three months; and 5) willing to provide contact information for follow up visits. Eligible participants who provided informed consent completed a survey, which included sections on demographic characteristics, current sex work, drug use, overdose history, and MAT use, as well as HIV/STI testing for chlamydia and gonorrhea, and referrals to health and social services (e.g., medical care, MAT, SSP, housing). Follow-up visits are ongoing and occur at 6-month intervals in a variety of settings (e.g., mobile van, study office, fast food restaurants, and home visits). This study was approved by the Johns Hopkins Bloomberg School of Public Health Institutional Review Board. The primary prespecified outcome was the number of non-fatal overdoses.
2.1. FIT study recruitment and data collection
All EMERALD participants enrolled between April 2018 and February 2019 were invited to participate in the FIT study following completion of their EMERALD visit. If interested, study staff briefly described the FIT study and screened all interested women. Eligibility criteria were: (1) currently enrolled in the EMERALD study; and (2) using opioids obtained illegally (heroin, fentanyl, and/or prescription opioid pills purchased on the street) in the past month. Eligible participants who provided informed consent completed a supplementary 10-minute interviewer-administered computer-assisted personal interview (CAPI) survey consisting of an overdose risk assessment, fentanyl-related attitudes, prior FTS use, harm reduction behaviors, and fentanyl knowledge that were not captured through the EMERALD survey.
Following completion of the baseline FIT survey, participants underwent a brief (5–10 min) training on use of the FTS (collecting a sample, completing the test, and interpreting results) as well as harm reduction micro-counseling. Harm reduction micro-counseling was tailored individually to behaviors participants had reported throughout the risk assessment portion of the FIT survey, (e.g., if the participant reported drinking alcohol, we informed them that mixing alcohol and opioids increases the chances of overdose; if the participant used drugs alone, we helped brainstorm ways of increasing safety in case of an overdose). Staff also emphasized the potential risks of fentanyl (e.g., high strength relative to heroin, what types of street drugs it has been found in), and safe drug use practices such as doing a test dose of drugs before using the full amount, having someone nearby to administer naloxone, and polysubstance use. An insert card with FTS instructions and key harm reduction concepts covered in the training was also provided (Fig. 1) along with 5 FTS (Rapid Response Single drug test strip, BTNX Inc), safe injection equipment (e.g., cookers, gauze, clean water, and an alcohol pad) and two doses of intramuscular naloxone hydrochloride (Hospira, Lake Forest, IL, USA).
Fig. 1.
FIT intervention insert card (double-sided print).
Participants were located for their one-month follow-up visit using several methods including repeated visits to recruitment neighborhoods and homes, phone calls and text messages, outreach on social media accounts, and outreach to stable contacts such as relatives, peers, and romantic partners. This follow-up period was chosen for comparability to a previous pilot study in Rhode Island among young PWUD. At one-month follow up, participants completed a brief 10-minute CAPI survey and were offered more FTS, safe injection equipment and naloxone. Participants were compensated with a $10 Visa gift card for completing the baseline visit, and a $20 Visa gift card for completing follow up. We also provided additional FTS and naloxone to any participant who wanted these at subsequent EMERALD study visits.
2.2. Measures
2.2.1. Socio-demographic characteristics
Age, race, the highest level of education completed, current engagement in sex work (“sold or traded sex for things like money, drugs, or other resources”), and recent (past 6 months) homelessness and arrest were collected at the baseline EMERALD visit.
2.2.2. Substance use
Recent (past 6 months at pretest, past 1 month at posttest) injection and non-injection drug use was measured separately. Women were asked if they injected (yes/no): heroin, cocaine, speedball (heroin/cocaine together), fentanyl, and crushed opioid pills. Women were also asked if they snorted, smoked, or swallowed (yes/no): heroin, fentanyl, opioid pills, benzodiazepine pills, powdered cocaine, and crack cocaine. For each drug and method (i.e. injecting heroin, swallowing pills), women were asked the frequency of use, which was dichotomized as daily/less than daily for analysis. An “illicit opioid use” variable was constructed, comprised of heroin, fentanyl, speedball and opioid pill misuse. Women who reported injecting any drug were asked if they had reused injection equipment in the past 6 months, including needles/syringes, cookers, or cotton. Alcohol use and frequency were measured by a single question from the Alcohol Use Disorders Identification Test-Concise (AUDIT-c) (Bush et al., 1998): “how often did you have a drink containing alcohol in the past month?” Alcohol use frequency was dichotomized into “at least weekly” or “less than weekly.” Preference for drugs with fentanyl and concern about fentanyl in drugs was also assessed using the following statements on a 5-point Likert scale: “I prefer drugs that have fentanyl in them” and “I am concerned about my drugs having fentanyl in them,” respectively. Responses were categorized as “yes” if the participant strongly/somewhat agreed with the statement.
2.2.3. FTS use
At baseline, women were asked if they had ever used FTS. At follow-up, women were asked how many of the five FTS were used, unused but still in their possession, lost/thrown away/stolen, or given to someone else. FTS uptake was defined as any posttest use. Women were also asked if they used their FTS before using drugs, after using drugs, or both. They were then asked about their FTS results, if the results surprised them (yes/no) and why (open-ended response). Women who reported at least one positive or negative result were asked what actions they took after the result (select all that apply); answer choices included: engaging in harm reduction behaviors (e.g., tester shot, injecting slower); giving away or selling the drugs; and continuing to use the drugs they just tested. Likeliness of future FTS use (i.e., FTS acceptability) was measured on a 5-point Likert scale from extremely likely to extremely unlikely.
2.2.4. Risk behaviors and overdose
At both the first study visit and follow-up, women were asked about the frequency of solitary drug use in the past month, whether they had asked someone to check on them when using drugs or if they were around someone who had naloxone when using drugs. Answers were coded as “yes” if response was any frequency greater than “never.” Women were asked how many times in the past month they overdosed “to the point of passing out;” women were considered to have overdosed if they reported at least one overdose.
2.2.5. Service access
The following services were assessed dichotomously: reported receiving services from a “counselor, psychologist, or receiving mental health care” in the past month; used the Baltimore City Needle Exchange Program; and enrolled in MAT (i.e., currently in a methadone or buprenorphine/suboxone program).
2.3. Statistical analysis
Frequencies of baseline characteristics were generated from the entire sample; z-tests showed differences in proportions, confidence intervals, and statistical significance between those who did and did not complete follow-up. Frequencies of FTS use variables were generated from women completing both surveys. McNemar’s test was used to measure changes from baseline to follow-up for all outcome variables, producing proportions, odds ratios, and confidence intervals. This analysis was performed on all participants completing follow-up, and stratified by FTS use (i.e., among women who used at least one FTS, and women who didn’t use any FTS). Statistical significance was set at p < 0.05. All analyses were conducted in Stata/SE 15.1 (College Station, TX).
3. Results
Fig. 2 shows results for screening, recruitment and retention. At baseline 22.3% of the sample were between 18 and 29 years old, 59.2% were non-Hispanic White, and 87.4% were actively selling sex (Table 1). Nearly half the sample (48.5%) injected heroin and 30.1% reported injecting fentanyl. Smoking crack cocaine was the most prevalent form of illicit non-injection drug use reported (73.8%) followed by heroin (62.1%). Eighty-nine percent had never used a FTS prior to the study. Three-quarters of women (75.7%) said they were concerned about fentanyl in their drugs.
Fig. 2.
Participant flow diagram for the FIT study: Baltimore, Maryland 2018–2019. Note: all 20 participants were ineligible because they did not use opioids in the past month.
Table 1.
Baseline pre-intervention characteristics of FSW who use opioids (n = 103) participating in the FIT study and baseline comparison of FIT study participants who did and did not complete follow-up: Baltimore, Maryland.
| Total (n = 103) N (%) |
% completed follow-up (n = 68) |
% did not complete follow-up (n = 35) |
P (z-test) | |
|---|---|---|---|---|
| Age | ||||
| 18–29 | 23 (22.3) | 16.2 (7.4, 24.9) | 34.3 (18.6, 50.0) | 0.04 |
| 30–39 | 33 (32.0) | 32.4 (21.2, 43.5) | 31.4 (16.1, 46.8) | 0.92 |
| 40–49 | 26 (25.2) | 27.9 (17.3, 38.6) | 20.0 (6.7, 33.3) | 0.38 |
| 50+ | 21 (20.4) | 23.5 (13.4, 33.6) | 14.3 (2.7, 25.9) | 0.27 |
| Race/ethnicity | ||||
| Non-Hispanic White | 61 (59.2) | 58.8 (47.1, 70.5) | 60.0 (43.8, 76.2) | 0.91 |
| Non-Hispanic Black | 35 (34.0) | 33.8 (22.6, 45.1) | 34.3 (18.6, 50.0) | 0.96 |
| Hispanic | 7 (6.8) | 7.4 (1.2, 13.6) | 5.7 (−2.0, 13.4) | 0.75 |
| Did not complete high school (n = 100) | 51 (51.0) | 51.5 (39.5, 63.6) | 50.0 (33.2, 66.8) | 0.89 |
| Active sex worker | 90 (87.4) | 85.3 (76.9, 93.7) | 91.4 (82.2, 100.0) | 0.38 |
| Homeless, past 6 months | 67 (65.1) | 58.8 (47.1, 70.5) | 77.1 (63.2, 91.1) | 0.07 |
| Arrested, past 6 months | 21 (20.4) | 17.7 (8.6, 26.7) | 25.7 (11.2, 40.2) | 0.34 |
| Alcohol, past month | 45 (43.7) | 47.1 (35.2, 58.9) | 37.1 (21.1, 53.2) | 0.34 |
| IDU, past 6 months | 52 (50.5) | 50.0 (38.1, 61.9) | 51.4 (34.9, 68.0) | 0.89 |
| Injected heroin | 50 (48.5) | 48.5 (36.7, 60.4) | 48.6 (32.0, 65.1) | 0.99 |
| Injected cocaine | 18 (17.5) | 22.1 (12.2, 31.9) | 8.6 (−0.7, 17.9) | 0.09 |
| Injected speedball | 24 (23.3) | 26.5 (16.0, 37.0) | 17.1 (4.7, 29.6) | 0.29 |
| Injected fentanyl | 31 (30.1) | 26.5 (16.0, 37.0) | 37.1 (21.1, 53.2) | 0.26 |
| Injected crushed opioid pills | 5 (4.9) | 1.5 (−1.4, 4.3) | 11.4 (0.9, 22.0) | 0.03 |
| Reused injection equipment (needles/syringes, cookers, or cotton, of those who injected) n = 1 missing | 50 (96.2) | 97.1 (91.4, 100.0) | 94.4 (83.9, 100.0) | 0.64 |
| Non-IDU, past 6 months | 89 (86.4) | 88.2 (80.6, 95.9) | 82.9 (70.4, 95.3) | 0.45 |
| Snorted/smoked heroin | 64 (62.1) | 58.8 (47.1, 70.5) | 68.6 (53.2, 84.0) | 0.33 |
| Snorted/smoked fentanyl | 11 (10.7) | 10.3 (3.1, 17.5) | 11.4 (0.9, 22.0) | 0.86 |
| Snorted/swallowed opioid pills | 24 (23.3) | 25.0 (14.7, 35.3) | 20.0 (6.7, 33.3) | 0.57 |
| Snort/swallowed benzodiazepine pills | 22 (21.4) | 22.1 (12.2, 31.9) | 20.0 (6.8, 33.3) | 0.81 |
| Snorted powdered cocaine | 35 (34.0) | 35.3 (23.9, 46.7) | 31.4 (16.1, 46.8) | 0.69 |
| Smoke crack cocaine | 76 (73.8) | 75.0 (64.7, 85.3) | 71.4 (56.5, 83.4) | 0.70 |
| Smoked marijuana | 41 (39.8) | 35.3 (23.9, 46.7) | 48.6 (32.0, 65.1) | 0.19 |
| Smoked or snorted methamphetamine | 2 (1.9) | 1.5 (−1.4, 4.3) | 2.9 (−2.7, 8.4) | 0.63 |
| Ever used fentanyl test strip | 11 (10.7) | 11.8 (4.1, 19.4) | 8.6 (−0.7, 17.9) | 0.62 |
| Concerned with fentanyl in drugs (agree) | 78 (75.7) | 80.9 (71.5, 90.2) | 65.7 (50.0, 81.4) | 0.09 |
| Prefer drugs containing fentanyl (agree) | 44 (42.7) | 42.7 (30.9, 54.4) | 42.9 (26.7, 59.3) | 0.98 |
| Accessed mental health care | 29 (28.2) | 29.4 (18.6, 40.2) | 25.7 (11.2, 40.2) | 0.69 |
| Accessed Baltimore City Needle Exchange Program | 44 (42.7) | 42.7 (30.9, 54.4) | 42.9 (26.7, 59.3) | 0.98 |
| Accessed medication assisted treatment | 40 (38.8) | 39.7 (28.1, 51.3) | 37.1 (21.1, 53.2) | 0.80 |
| Solitary drug use | 99 (96.1) | 95.6 (90.7, 100.0) | 97.1 (91.6, 100.0) | 0.70 |
| Asking someone to check on you | 68 (66.0) | 63.2 (51.8, 74.7) | 71.4 (56.5, 86.4) | 0.41 |
| Non-fatal overdose | 10 (9.7) | 8.8 (2.1, 15.6) | 11.4 (0.9, 22.0) | 0.67 |
Women completed follow-up in an average of 56 days (22.1% within the one-month goal). A significantly higher percentage of participants aged 18–29 were lost to follow-up (16.2% completed follow-up, 34.3% did not; p = 0.04), participants who injected opioid pills were also more likely to be lost to follow-up (1.5% completed follow-up, 11.4% did not; p = 0.03), though the numerators are very small. There were no other significant differences between completers and women lost to follow-up on any of the baseline demographic characteristics, drug use behaviors, or harm reduction behaviors that were measured (Table 1).
At follow-up, 83.8% used at least one FTS, with over half (57.9%) using FTS before using drugs (Table 2). Half the sample (50.8%) had at least one unused FTS at the follow-up visit. Of those who used at least one FTS, 86.0% had at least one positive result, 50.9% had at least one negative result, and 14.0% had an unclear result.
Table 2.
Utilization of fentanyl test strips at one-month follow-up among FSW who use opioids (n = 68) in Baltimore, Maryland.
| n (%) | |
|---|---|
| Used FTS in past month (1–5 strips) | 57 (83.8) |
| When used FTS | |
| Both before and after using drugs | 17 (29.8) |
| Before using drugs only | 33 (57.9) |
| After using drugs only | 7 (12.3) |
| ≥1 FTS tested positive | 49 (86.0) |
| What did you do after you received a positive result? Select all that apply.* | |
| Engaged in a harm reduction behavior‡ | 33 (68.8) |
| Used the tested drugs | 30 (62.5) |
| Did not use drugs† | 6 (12.6) |
| Told others about result | 6 (12.5) |
| Stopped going to that dealer/supplier | 2 (4.2) |
| Other | 1 (2.1) |
| Cessation§ | 0 (0.0) |
| ≥1 FTS tested negative | 29 (50.9) |
| What did you do after you received a negative result? Select all that apply.** | |
| Used the tested drugs | 21 (87.5) |
| Used harm reduction behavior‡ | 2 (8.4) |
| Did not use drugs† | 2 (8.3) |
| Other | 1 (4.2) |
| Told others about result | 0 (0.0) |
| Stopped going to that dealer/supplier | 0 (0.0) |
| Cessation§ | 0 (0.0) |
| Surprised by results†† | 31 (57.4) |
| Among ≥ 1 FTS positive* | 26 (54.2) |
| Among ≥ 1 FTS negative** | 17 (70.8) |
| ≥1 FTS showed unclear result | 8 (14.0) |
| Received mixed positive, negative, and/or indeterminate FTS results | 20 (38.5) |
| Has unused FTS†† | 34 (50.8) |
| FTS were lost/thrown away/stolen | 9 (13.2) |
| Gave some/all FTS to someone else | 12 (17.7) |
| How likely are you to use FTS in the future? | |
| Extremely likely | 49 (72.1) |
| Likely | 16 (23.5) |
| Neither likely nor unlikely | 1 (1.5) |
| Unlikely | 1 (1.5) |
| Extremely unlikely | 1 (1.5) |
among n = 48 who used at least 1 FTS and had a positive result (n = 1 missing data for these questions).
includes using less than intended, going slower, doing tester shot, asking someone to check on them when using.
includes threw drugs out, sold to someone else, gave to someone else for free.
includes stopped using drugs, entered MAT.
among n = 24 who used at least 1 FTS and had a negative result (n = 5 missing data for these questions).
among n = 52 who used more than 1 FTS.
n = 3 missing.
Over half (57.1%) of the 57 women who used one FTS reported that they were surprised by the results; 54.2% of the 49 participants who received at least one positive were surprised by the result, and 70.8% of the 25 participants who received at least one negative were surprised. Of the 28 women who received only positive results, 15 were surprised at how widespread fentanyl was or had become. One woman stated she “knew [fentanyl] was in dope but wasn’t expecting it to be in all 5 fentanyl test strips.” Of the 5 women who received all negative results, each stated they were surprised that every test was negative because they had assumed fentanyl was present in all drugs. Women who received a mix of positive, negative and unclear results echoed these thoughts and were also surprised at the ubiquity of fentanyl, even in drugs other than heroin. Several participants who received at least one positive test indicated that they were also told by a dealer or others that there was no fentanyl. One woman who received mixed results expressed she was surprised how well the FTS worked, while another said that it’s a “50/50 chance around here” of getting fentanyl in her drugs.
Forty-eight women received at least one positive result and answered an open-ended survey question about their reactions to results. Thirty women (30/48, 62.5%) reported that they used their drugs as originally intended (i.e., without behavioral modifications) after receiving at least one positive result. The plurality of these women (12/30, 40.0%) stated that they used their drugs as originally intended to alleviate withdrawal symptoms (i.e. sickness or pain); others reported that they had missed or had an insufficient methadone dose. Five women (5/12, 41.7%) said they used as originally intended despite a positive result because they enjoyed fentanyl or “wanted to” use drugs with fentanyl, and 2 women (2/12, 16.7%) said they have developed a “fentanyl habit.” Four women (4/12, 33.3%) expressed skepticism of results, either saying they believed there was “not much” fentanyl in the drugs, or because the drugs in question were the same as typically purchased and have been previously used without negative side effects. After receiving a positive result, 15 (15/48, 31.3%) used a smaller amount of drugs than originally intended, 10 (10/48, 20.8%) went slower when injecting, 5 (5/48, 10.4%) did a tester shot before fully injecting, and 3 (3/48, 6.3%) asked someone to check on them when they were using drugs. Six (6/48, 12.5%) women told others about the positive result, 3 (3/48, 6.3%) threw the drugs out, 3 (3/48, 6.3%) sold them to someone else, and 2 (2/48, 4.2%) stopped going to that dealer/supplier.
Twenty-four women received at least one negative result and answered an open-ended survey question about their reactions to results. Of the 25 women who received at least one negative test result, 21 (87.5%) reported using their drugs as originally intended. Primary reasons for using as originally intended due to withdrawal symptoms (5/25, 20.0%) or because of a negative test result (5/25, 20.0%). Two women (2/25, 8.3%) reported selling the drugs to someone else after receiving a negative result, one (1/24, 4.2%) used a smaller amount of drugs than originally intended, and one (1/24, 4.2%) went slower when injecting. Lastly, among those who completed the follow-up visit, virtually all (24/25, 96%) women said they were “extremely likely” or “likely” to use FTS in the future.
3.1. Pretest-posttest differences
Changes in substance use and frequency were assessed comparing baseline and follow-up behaviors (Table 3). Among the sample who completed follow-up and used ≥ 1 FTS (n = 57), we observed reductions in past month illicit opioid use (100% to 89.5%, p = 0.01), frequency of illicit opioid use (daily: 75.4% to 59.7%, p = 0.03) and frequency of injection (daily: 40.4% to 26.3%, p = 0.01). Past month benzodiazepine use (22.8% to 7.40%, p = 0.007) and frequency (daily: 14.0% to 1.8%, p = 0.02) also saw significant declines. Nearly the entire sample (94.9%) reported using drugs alone at baseline and this was reduced to 67.8% at follow-up (p < 0.001). Changes in asking someone to check on you and non-fatal overdose were in the expected direction but did not reach significance (Table 2). Notably, there were no significant shifts in preferring drugs containing fentanyl (42.1% to 38.6%, p = 0.41). Odds ratios and confidence intervals show precision estimates. Results among FTS nonusers trended in the direction, though results did not reach statistical significance likely because of the small sample size (n = 11).
Table 3.
Changes in drug use and overdose risk and harm reduction behaviors among FSW who use opioids (n = 68) in Baltimore, Maryland: results of McNemar’s pretest–posttest.
| Total (n = 68) |
Used at least 1 FTS (n = 57) |
Did not use any FTS (n = 11) |
||||
|---|---|---|---|---|---|---|
| Baseline | Follow-up | Baseline | Follow-up | Baseline | Follow-up | |
| Substance use | ||||||
| Frequency of illicit opioid use‡ | ||||||
| At least daily | 52 (76.5) | 38 (55.9)** | 43 (75.4) | 34 (59.7)* | 9 (81.8) | 4 (36.4) |
| Less than daily | 16 (23.5) | 30 (44.1)** | 14 (24.6) | 23 (40.4)* | 2 (18.2) | 7 (63.6) |
| Any injection drug use‡ | 34 (50.0) | 29 (42.7) | 29 (50.9) | 26 (45.6) | 5 (45.5) | 3 (27.3) |
| Frequency of injection drug use‡ | ||||||
| At least daily | 27 (39.7) | 17 (25.0)** | 23 (40.4) | 15 (26.3)** | 4 (36.4) | 2 (18.2) |
| Less than daily | 41 (60.3) | 52 (75.0)** | 34 (59.7) | 42 (73.7)** | 7 (63.6) | 9 (81.8) |
| Prefer drugs containing fentanyl (agree) | 29 (42.7) | 26 (38.2) | 24 (42.1) | 22 (38.6) | 5 (45.5) | 4 (36.4) |
| Concerned with fentanyl in drugs (agree) | 57 (83.8) | 51 (75.0) | 47 (82.5) | 43 (75.4) | 10 (90.9) | 8 (72.7) |
| Benzodiazepine use‡ | 15 (22.1) | 5 (7.4)** | 13 (22.8) | 4 (7.0)** | 2 (18.2) | 1 (9.1) |
| Frequency of benzodiazepine use‡ | ||||||
| At least daily | 9 (13.2) | 2 (2.9)* | 8 (14.0) | 1 (1.8)* | 1 (9.1) | 1 (9.1) |
| Less than daily | 59 (86.8) | 66 (97.1)* | 49 (86.0) | 56 (98.3)* | 10 (90.9) | 10 (90.9) |
| Alcohol use‡ | 32 (47.1) | 29 (42.7) | 27 (47.4) | 24 (42.1) | 5 (45.5) | 5 (45.5) |
| Frequency of alcohol use‡ | ||||||
| At least daily | 21 (30.9) | 15 (22.1) | 17 (29.8) | 13 (22.8) | 4 (36.4) | 2 (18.2) |
| Less than daily | 47 (69.1) | 53 (77.9) | 40 (70.2) | 44 (77.2) | 7 (63.6) | 9 (81.8) |
| Changes in risk behaviors and overdose† | ||||||
| Non-fatal overdose‡ | 6 (8.8) | 1 (1.5) | 4 (7.0) | 1 (1.8) | 2 (18.2) | 0 (0.0) |
| Solitary drug use‡ | 56 (94.9) | 40 (67.8)*** | 48 (94.1) | 33 (64.8)*** | 8 (100.0) | 7 (87.5) |
| Asking someone to check on you‡ | 36 (61.0) | 43 (72.8) | 32 (62.8) | 36 (70.6) | 4 (50.0) | 7 (87.5) |
| Service access | ||||||
| Mental health care‡ | 20 (29.4) | 22 (32.4) | 16 (28.1) | 18 (31.6) | 4 (36.4) | 4 (36.4) |
| Baltimore City Needle exchange Program‡ | 29 (42.7) | 22 (32.4) | 17 (77.3) | 13 (59.1) | 2 (66.7) | 1 (33.3) |
| Medication-assisted treatment(e.g., methadone, suboxone) ‡ | 27 (39.7) | 31 (45.6) | 21 (36.8) | 25 (43.9) | 6 (54.6) | 6 (54.6) |
p < 0.05
p < 0.01
p < 0.001.
past month.
among n = 59 who reported opioid use at follow-up.
4. Discussion
The FIT study is one of the first to document the short-term effectiveness of a brief, FTS-based intervention among a high-risk group of women who use drugs. The study was conducted among urban street-based FSW who used illicit opioids in the era of fentanyl. At follow-up, the majority of women who received the FTS reported using them (84%), and virtually all (96%) reported being likely to use FTS in the future. We observed pretest–posttest reductions in the frequency of illicit opioid use (including heroin and fentanyl), injection drug use and benzodiazepine use as well as solitary drug use. This pilot was conducted against a backdrop of high fentanyl-related concern and mortality rates. Our data suggest that the FIT intervention—consisting of a brief overdose risk assessment, tailored harm reduction messaging, 5 FTS and training, and naloxone—could be a promising model for reducing overdose risk among high-risk populations.
In our experience, clinicians and public health officials often question the utility of implementing FTS programs in settings with high fentanyl market penetration, such as Baltimore. Our data highlight the potential role of FTS programs even in such settings. Despite the pervasiveness of fentanyl, which accounted for 573 deaths and over 75% of drug overdose deaths in Baltimore the year prior to the study (Maryland Department of Health, 2017), FTS were highly acceptable and provided many participants with novel information about their drugs and risk of overdose. In addition, we found significant pretest–posttest reductions in overdose risk behaviors, namely the frequency of opioid use, benzodiazepine use, and injection drug use. We also observed large reductions in solitary drug use, which was common among this cohort at baseline. The occurrence of non-fatal overdose was six times lower at follow-up though this finding did not reach significance likely due to the short follow-up duration and small sample size. All of these positive findings were found even as women continued to use drugs throughout the study, and notably, there were no changes in preferring drugs containing fentanyl. Interestingly, many did not use all five FTS provided. We hypothesize that the FTS may be perceived as less useful after the first positive result, especially when drugs are obtained from the same dealer. These findings demonstrate that harm reduction interventions involving FTS distribution and education have a role in promoting the mitigation of drug-related morbidities. A larger longitudinal study will be required to evaluate the effectiveness and sustainability of this promising intervention on reducing overdose risk.
Our findings corroborate and expand the existing literature that together demonstrate high levels of acceptability of drug checking programs among PWUD and service providers (Krieger et al., 2018a, 2018b; Glick et al., 2019; Schneider et al., 2019). The majority of women in our study tested their drugs prior to use and engaged in harm reduction behaviors upon seeing a fentanyl-positive result, which corroborates findings from a recent study conducted in Greensboro, North Carolina (Peiper et al., 2019), and demonstrates the potential for FTS as a preventative strategy against overdose. We observed a clear demarcation in the types of behaviors that followed a fentanyl-positive versus negative result; while 69% of women with fentanyl-positive results engaged in harm reduction behaviors during drug use (e.g., doing a tester shot, using less), only 8% engaged in harm reduction behaviors following a fentanyl-negative result. These trends are consistent with recent findings from Rhode Island (a high fentanyl-burden area) that documented a strong association between receiving a fentanyl-positive result and harm reduction behaviors to mitigate the risk of overdose (Krieger et al., 2018). One criticism of FTS programs is that they are not useful in settings where fentanyl dominates the opioid market and fentanyl awareness is high. Findings from the Rhode Island study and the current study suggest that the testing of drugs at the individual-level through FTS and educational messaging may encourage risk reduction among PWUD even in locations where fentanyl contamination of the drug supply is an established issue. However, programs must emphasize to PWUD that a fentanyl-negative result does not guarantee safety of the drug being tested. FTS are not 100% accurate (with estimated sensitivity of 98% and specificity of 94% for off-label use from an initial validation study (Krieger et al., 2018) and do not have the ability to detect other potentially lethal contaminants (e.g., novel fentanyl analogs, U-47700) present in street drugs. The implementation of comprehensive drug checking programs that test for a variety of drugs could also be used to alert the community of other potentially lethal drugs (Harper et al., 2017). Another powerful community-based intervention to rapidly reduce fatal overdoses would be to implement overdose prevention sites (OPS), which enjoy broad support among urban and rural PWUD in the U.S., and could provide increased access to social support, harm reduction services including naloxone and FTS and other needed health services (Rouhani et al., 2020; Park et al., 2019; O’Rourke et al., 2019; Kral and Davidson, 2017; Potier et al., 2014).
The FIT intervention was designed to be brief and easy to implement, yet personalized and as comprehensive as possible. For example, we assessed an individual’s overdose risk through our pretest survey, which allowed us to provide tailored harm reduction messages and referrals to participants; we also provided naloxone kits given the likelihood of ongoing drug use and risk. FIT could be incorporated into clinical settings such as emergency rooms that see high rates of overdose among PWUD, as well as at existing SSP and naloxone outreach programs that engage with high-risk PWUD.
When designing messaging for FTS programs, special consideration will be required around the risks associated with polysubstance use, including concomitant opioid and alcohol use, which was prevalent among our population but did not change over time. Concomitant benzodiazepine and opioid use is common among PWUD globally and also associated with elevated overdose risk and mortality (Jones et al., 2012, 2018); encouragingly we observed a significant change in self-reported benzodiazepine use from 22% to 7% in the current study. The increasing rates of polysubstance use, an increasingly important area of research in the substance use field, requires attention in future FTS intervention work especially given that these individuals may be at elevated risk of overdose yet less likely to receive overdose trainings (Schneider et al., 2019). Given that 14% of respondents had an unclear result, research will need to be conducted in order to evaluate whether this is due to lack of training around FTS use, difficulties reading the result, or issues around comprehension.
These findings are subject to limitations. A relatively low proportion of women who were approached agreed to screening mostly due to lack of time, childcare responsibilities or drug withdrawal. This was not unexpected as the pilot study was embedded within a larger study visit. The eligibility criteria and intervention were masked thus we do not anticipate significant selection bias affecting FTS uptake and acceptability, however, the relatively low participation rate may have resulted in selection bias. Some women only used FTS after drug use, which may reduce the effectiveness of FTS in overdose prevention. However, given the high frequency of drug use in this population, even among individuals who report testing their drugs after use, we argue that the information is not lost and could influence their future behavior. The current study characterized changes in risk behaviors using dichotomous exposures (≥1 fentanyl-positive result; ≥1 fentanyl-negative result). However, future studies with a larger sample should investigate the cumulative impact of multiple positive and negative results on risk behaviors over time.
Furthermore, our study was only conducted among one population recruited from one city, which may affect generalizability, and utilized a pretest–posttest design with a single follow-up visit. We did not include a control group, although we were able to explore differences between FTS users and nonusers. Additionally, many difficulties were incurred in attempting to find participants for their follow-up surveys, such as phone numbers for participants and/or stable contacts being inoperable or participants no longer living at the same address at the time their follow-up survey was due. This was especially true for women experiencing homelessness or other transient situations that we may have had little or no direct contact information for other than chance random encounters in the areas they were recruited. The sub-optimal retention rates combined with the small sample size may have resulted in an underpowered analysis. Lastly, survey data may be subject to social desirability bias and recall bias. To address these limitations, a larger and longer-term cohort of FSW who use opioids and other PWUD recruited from multiple locations will be required to determine the short-term and long-term effectiveness of brief FTS-based interventions on risk behaviors and the prevention of non-fatal and fatal overdose.
Despite these limitations, to our knowledge this study is among the first to develop an FTS-based intervention that could serve as a model for reducing risk among high-risk populations. Given the high levels of solitary drug use observed here and elsewhere, it is evident that naloxone alone will not be sufficient in curbing overdose mortality among PWUD. Given the current overdose crisis, there is an urgent need to develop and deploy interventions that can reduce risk; large-scale FTS distribution programs are currently being piloted at health agencies in California, Maryland and Delaware among other states. The early experiences of harm reduction organizations and PWUD have been instrumental in conceptualizing and designing these programs and will continue to be key in informing the development of drug checking tools and guidelines. Further monitoring and evaluation will help elucidate the long-term impact of FTS and other drug checking programs.
HIGHLIGHTS.
Fentanyl test strips allow people to test drugs for fentanyl prior to consumption.
We found high fentanyl test strip acceptability among female sex workers.
We also observed pre/posttest reductions in drug use frequency and solitary use.
Fentanyl test strip interventions may reduce overdose risk in female sex workers.
Acknowledgements
We are grateful to the FIT and EMERALD study teams, collaborators, and study participants. We also especially thank Jess Tiley, Van Asher, and William Matthews for sharing their early experiences with fentanyl test strip distribution.
5. Role of funding source
This work was supported in part by the National Institute on Drug Abuse (5R01DA041243-04) and the Johns Hopkins University Center for AIDS Research (1P30AI094189). Dr. Ju Nyeong Park is partially supported by a Faculty Development Grant from the Johns Hopkins University Center for AIDS Research. Catherine Tomko is supported in part by the National Institute of Mental Health (F31MH118817). Dr. Brandon Marshall is supported in part by the National Institute of General Medical Sciences (P20GM125507). 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
Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: SGS is an expert witness for plaintiffs in opioid litigation.
References
- Allen ST, Footer KHA, Galai N, Park JN, Silberzahn B, & Sherman SG (2019). Implementing targeted sampling: Lessons learned from recruiting female sex workers in Baltimore, MD. Journal of Urban Health, 96(3), 442–451. 10.1007/s11524-018-0292-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Barratt MJ, Kowalski M, Maier LJ, & Ritter A (2018). Global review of drug checking services operating in 2017. Sydney, Australia: National Drug and Alcohol Research Center, UNSW. [Google Scholar]
- Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA (1998). The AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for problem drinking. Ambulatory Care Quality Improvement Project (ACQUIP). Alcohol Use Disorders Identification Test. JAMA Internal Medicine, 158(16), 1789–95. Epub 1998/09/17. PubMed PMID: 9738608. [DOI] [PubMed] [Google Scholar]
- Ciccarone D (2017). Fentanyl in the US heroin supply: A rapidly changing risk environment. International Journal of Drug Policy, 46, 107–111. 10.1016/j.drugpo.2017.06.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- DEA. (2018). National drug threat assessment. Drug Enforcement Administration, 2018 Contract No.: DEA-DCT-DIR-032-18. [Google Scholar]
- Glick JL, Christensen T, Park JN, McKenzie M, Green TC, & Sherman SG (2019). Stakeholder perspectives on implementing fentanyl drug checking: Results from a multi-site study. Drug and Alcohol Dependence, 194, 527–532 Epub 2018/12/15. 10.1016/j.drugalcdep.2018.10.017 PubMed PMID: 30551090. [DOI] [PubMed] [Google Scholar]
- Goldenberg S, Watt S, Braschel M, Hayashi K, Moreheart S, & Shannon K (2020). Police-related barriers to harm reduction linked to non-fatal overdose amongst sex workers who use drugs: Results of a community-based cohort in Metro Vancouver, Canada. International Journal of Drug Policy, 76,102618 Epub 2019/12/16. 10.1016/j.drugpo.2019.102618 PubMed PMID: 31838244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Goldman JE, Waye KM, Periera KA, Krieger MS, Yedinak JL, & Marshall BDL (2019). Perspectives on rapid fentanyl test strips as a harm reduction practice among young adults who use drugs: A qualitative study. Harm Reduction Journal, 16(1), 10.1186/s12954-018-0276-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Green TC, Sherman SG, Park JN, Gilbert M, McKenzie M, Struth E, et al. (2020). A multi-site assessment of the limits of detection, sensitivity and specificity of three devices for public health-based drug checking of fentanyl in street-acquired samples. International Journal of Drug Policy, 77(102661), 10.1016/j.drugpo.2020.102661 31951925. [DOI] [PubMed] [Google Scholar]
- Harper L, Powell J, & Pijl EM (2017). An overview of forensic drug testing methods and their suitability for harm reduction point-of-care services. Harm Reduction Journal 14(1), 10.1186/s12954-017-0179-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hedegaard H, Minino AM, & Warner M (2018). Drug overdose deaths in the United States, 1999–2017. NCHS Data Brief, 329, 1–8 Epub 2018/12/01 PubMed PMID: 30500323. [PubMed] [Google Scholar]
- Hedegaard H, Minino AM, & Warner M (2020). Drug overdose deaths in the United States, 1999–2018. NCHS Data Brief, 356, 1–8. [PubMed] [Google Scholar]
- Inciardi JA, & Surratt HL (2001). Drug use, street crime, and sex-trading among cocaine-dependent women: Implications for public health and criminal justice policy. Journal of Psychoactive Drugs, 33(4), 379–389 Epub 2002/02/05. 10.1080/02791072.2001.10399923 PubMed PMID: 11824697. [DOI] [PubMed] [Google Scholar]
- 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. 10.1001/jama.2018.2844. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jones JD, Mogali S, & Comer SD (2012). Polydrug abuse: A review of opioid and benzodiazepine combination use. Drug and Alcohol Dependence, 125(1–2), 8–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 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 Reduction Journal, 15(1), 10.1186/s12954-018-0252-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kral AH, & Davidson PJ (2017). Addressing the Nation’s opioid epidemic: Lessons from an unsanctioned supervised injection site in the U.S. American Journal of Preventive Medicine, 53(6), 919–922. 10.1016/j.amepre.2017.06.010. [DOI] [PubMed] [Google Scholar]
- Krieger MS, Goedel WC, Buxton JA, Lysyshyn M, Bernstein E, Sherman SG, et al. (2018). Use of rapid fentanyl test strips among young adults who use drugs. International Journal of Drug Policy; 61, 52–8. Epub 2018/10/23. doi: 10.1016/j.drugpo.2018.09.009. PubMed PMID: 30344005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Krieger MS, Yedinak JL, Buxton JA, Lysyshyn M, Bernstein E, Rich JD, Green TC, Hadland SE, & Marshall BDL (2018). High willingness to use rapid fentanyl test strips among young adults who use drugs. Harm Reduction Journal, 15(1), 10.1186/s12954-018-0213-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lorvick J, Browne EN, Lambdin BH, & Comfort M (2018). Polydrug use patterns, risk behavior and unmet healthcare need in a community-based sample of women who use cocaine, heroin or methamphetamine. Addictive Behaviors, 85, 94–99 Epub 2018/06/09. 10.1016/j.addbeh.2018.05.013 PubMed PMID: 29883856. [DOI] [PubMed] [Google Scholar]
- 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. 10.1080/02791072.2017.1394508. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Maryland Department of Health. (207). Unintentional drug- and alcohol-related intoxication deaths in Maryland annual report 2017. Annapolis, Maryland. [Google Scholar]
- McGowan CR, Harris M, Platt L, Hope V, Rhodes T (2018). Fentanyl self-testing outside supervised injection settings to prevent opioid overdose: Do we know enough to promote it? International Journal of Drug Policy, 58, 31–6. Epub 2018/05/15. doi: 10.1016/j.drugpo.2018.04.017. PubMed PMID: 29758542. [DOI] [PubMed] [Google Scholar]
- O’Rourke A, White RH, Park JN, Rodriguez K, Kilkenny ME, Sherman SG, & Allen ST (2019). Acceptability of safe drug consumption spaces among people who inject drugs in rural West Virginia. Harm Reduction Journal, 16(1), 10.1186/s12954-019-0320-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Park JN, Decker MR, Bass JK, Galai N, Tomko C, Jain KM, et al. Cumulative violence and PTSD symptom severity among urban street-based female sex workers. Journal of Interpersonal Violence 2019:886260519884694. Epub 2019/11/05. doi: 10.1177/0886260519884694. PubMed PMID: 31679445; PubMed Central PMCID: PMCPMC7195245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Park JN, Footer KHA, Decker MR, Tomko C, Allen ST, Galai N, & Sherman SG (2019). Interpersonal and structural factors associated with receptive syringe-sharing among a prospective cohort of female sex workers who inject drugs. Addiction, 114(7), 1204–1213. 10.1111/add.v114.710.1111/add.14567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Park JN, Sherman SG, Rouhani S, Morales KB, McKenzie M, Allen ST, et al. (2019). Willingness to use safe consumption spaces among opioid users at high risk of fentanyl overdose in Baltimore, Providence, and Boston. Journal of Urban Health, 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 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, 122–8. Epub 2018/10/08. doi: 10.1016/j.drugpo.2018.08.007. PubMed PMID: 30292493. [DOI] [PubMed] [Google Scholar]
- Potier C, Laprévote V, Dubois-Arber F, Cottencin O, & Rolland B (2014). Supervised injection services: What has been demonstrated? A systematic literature review. Drug and Alcohol Dependence, 145, 48–68. [DOI] [PubMed] [Google Scholar]
- Rouhani S, White RH, Park JN, & Sherman SG (2020). High willingness to use overdose prevention sites among female sex workers in Baltimore, Maryland. Drug and Alcohol Dependence, 212, 108042. 10.1016/j.drugalcdep.2020.108042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schneider KE, Park JN, Allen ST, Weir BW, & Sherman SG (2019). Patterns of polysubstance use and overdose among people who inject drugs in Baltimore, Maryland: A latent class analysis. Drug and Alcohol Dependence. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Seal KH, Kral AH, Gee L, Moore LD, Bluthenthal RN, Lorvick J, et al. (2001). Predictors and prevention of nonfatal overdose among street-recruited injection heroin users in the San Francisco Bay Area, 1998–1999. American Journal of Public Health, 91(11), 1842–1846. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sherman SG, Park JN, Glick J, Mackenzie M, Morales K, Christensen T, et al. (2018). FORECAST Study Summary Report: Johns Hopkins Bloomberg School of Public Health. Available from: https://americanhealth.jhu.edu/fentanyl. [Google Scholar]
- Sherman SG, Park JN, Galai N, Allen ST, Huettner SS, Silberzahn BE, Decker MR, Poteat TC, & Footer KHA (2019). Drivers of HIV infection among cisgender and transgender female sex worker populations in Baltimore city: Results from the SAPPHIRE study. JAIDS Journal of Acquired Immune Deficiency Syndromes, 80(5), 513–521. 10.1097/QAI.0000000000001959. [DOI] [PubMed] [Google Scholar]
- Suzuki J, El-Haddad S (2017). A review: Fentanyl and non-pharmaceutical fentanyls. Drug and Alcohol Dependence, 171, 107–16. Epub 2017/01/10. doi: 10.1016/j.drugalcdep.2016.11.033. PubMed PMID: 28068563. [DOI] [PubMed] [Google Scholar]
- Tomko C, Nestadt DF, Rouhani S, Silberzahn BE, Haney K, Park JN, et al. (2020). Confirmatory factor analysis and construct validity of the internalized sex work stigma scale among a cohort of cisgender female sex workers in Baltimore, Maryland, United States. Journal of Sex Research, 1–11 Epub 2020/05/14. 10.1080/00224499.2020.1755821 PubMed PMID: 32401075. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tupper KW, McCrae K, Garber I, Lysyshyn M, Wood E (2018). Initial results of a drug checking pilot program to detect fentanyl adulteration in a Canadian setting. Drug and Alcohol Dependence, 190, 242–5. Epub 2018/08/01. doi: 10.1016/j.drugalcdep.2018.06.020. PubMed PMID: 30064061. [DOI] [PubMed] [Google Scholar]
- CDC WONDER. Overdose Death Rates 2018 [September/30/18]. Available from: https://www.drugabuse.gov/related-topics/trends-statistics/overdose-death-rates.