Examining Overdose and Homelessness as Predictors of Willingness to Use Supervised Injection Facilities by Services Provided Among Persons Who Inject Drugs

Abstract

Background:

Internationally, supervised injection facilities (SIFs) have demonstrated efficacy in reducing rates of overdose and promoting entry into treatment among persons who inject drugs (PWID); however, they remain unavailable in the US. Early findings examining American PWID illustrate high overall willingness to use SIFs.

Objectives:

The current study expands upon this research by examining PWID’s likelihood to use SIFs based on services offered (e.g., provides clean needles, linkage to treatment programs) and whether known risk factors (prior overdose, homelessness) influence PWID’s willingness to use a SIF.

Methods:

Participants (n = 184) were patients entering short-term inpatient opioid withdrawal management in Massachusetts between May 2018-February 2019 who reported injection drug use in the prior 30 days. We examined PWID’s likelihood to use a SIF if eight unique services were available, and compared if this differed by overdose history and homelessness status using ordered logistic regression and Pearson χ² – tests of independence.

Results:

Participants (34.2 (± 8.3 sd) years of age, 68.5% male, 85.9% White, 8.2% Hispanic) reported being most likely to use SIFs that provided safety from police intervention (86.7%), entry into withdrawal management (85.9%), or clean needles (83.2%). Drug works disposal and safety from police were particularly important for PWID with a history of overdose.

Conclusions:

Overall, treatment-seeking PWIDs reported greater willingness to utilize SIFs if particular services were provided.

Scientific Significance:

These findings point to features of SIFs that may enhance treatment-seeking PWID’s amenability to utilizing these services if such sites open in the US.

1.0. INTRODUCTION

Drug overdose deaths in the US have doubled in the past decade ¹. Given that persons who inject drugs (PWID) face escalated risk for overdose ², there has been widespread interest in the promise of supervised injections facilities (SIFs)—sites supervised by trained medical professionals where individuals can inject pre-obtained drugs—to combat this public health crisis. Studies examining the efficacy of existing SIFs find they lower rates of overdose and overdose fatalities, and promote entry into treatment (e.g., methadone) ^3–6. Still, an overall lack of knowledge about SIFs and concerns about condoning drug use and exacerbating drug-related crime have hampered efforts to implement SIFs in the US e.g., ⁷. Although over 100 legally sanctioned SIFs operate in other countries, including Canada, Europe and Australia ⁸, no federal or state sponsored SIFs currently exist in the United States despite ongoing political debate. While Seattle and Philadelphia are on target to open the first SIFs nationally ^8,9, Massachusetts, where opioid-related overdose death rates are double the national average ¹⁰, has no plans to allow SIFs.

International ^11–14 and national ^15–17 studies demonstrate strong overall willingness among PWID to utilize SIFs. In US-based studies, 87.1% to 91.4% of PWID reported positive willingness to use SIFs ^16,17. Although risk factors, including heroin use, overdose history, homelessness or unstable housing status, and current desire for treatment were associated with greater willingness to use SIFs ^15–18, no study to date has examined PWID’s willingness to use SIFs based on the services provided. To gain insight into which services may draw PWID to utilize SIFs, we asked PWID in Massachusetts if they would be more likely to use a SIF if specific services were provided ^19,20. Our sample of persons entering an inpatient withdrawal management program for heroin use represents a sub-population of PWID who may garner particular benefit from SIFs. Moreover, because homeless PWID and those with a history of overdose appear most amenable and in need of SIFs, we examined if willingness differed by these risk factors ^15,17.

2.0. Methods

2.1. Recruitment

Between May 2018 and February 2019, persons seeking inpatient opioid withdrawal management at Stanley Street Treatment and Resources, Inc. (SSTAR) in Fall River, Massachusetts were asked to participate in a survey research study. SSTAR’s opioid withdrawal program provides withdrawal management using a methadone taper protocol, individual and group counseling, and aftercare planning and referral. Admitted patients stayed an average of 4.9 days.

Of consecutive patients admitted to SSTAR during the recruitment period for opioid misuse, 247 met the study’s eligibility criteria (18 years or older, English-speaking, and able to provide informed consent) as approved by the Butler Hospital Institutional Review Board. Thirty-three persons refused study participation, had participated previously, or were discharged before staff could interview them. The remaining 214 persons completed an incentivized, face-to-face interview administered by non-treating research staff over the course of approximately 15 minutes. Only those who had injected drugs in the past 30 days (n=184) were included in subsequent analyses.

2.2. Measures

In addition to age, sex, race, and ethnicity the following variables were assessed.

2.2.1. Overdose history.

Respondents reporting that they had overdosed since their first use of drugs were coded as having ever overdosed.

2.2.2. Recent homelessness.

Respondents reported where they had slept/spent their nights in the past three months using ten answer options (e.g., your own home, a friend’s home, a shelter, incarcerated). Those reporting spending any nights on the street or in a shelter were coded as recently homeless.

2.2.3. Use of Supervised Injection Facilities.

Participants were asked “If one were convenient for you to get to, how often do you think you would use it?” Responses were never, once in a while, most times I was ready to use, and always.

2.2.4. Interest in SIF Services.

Interviewers defined a SIF as “a place where you could bring your drugs and inject them with health care professionals around to keep you safe in case you have a bad reaction or an overdose. The goal is not to get drug users to stop using drugs, but to reduce disease and death.” All respondents, including those reporting that they would never use a SFI, were then provided a list of services (see Table 1) that SIFs could offer and asked which services would be “important to make you more likely to go?”

Table 1.

Willingness to use a SIF if Services are Provided by History of Overdose and Recent Homelessness.

		EVER OVERDOSE?			HOMELESSNESS
	Sample (n = 184)	No (n = 51)	Yes (n = 133)	p = a	No (n = 133)	Yes (n = 47)	p = a
Provides Clean Needles	153(83.2%)	38 (74.5%)	115 (86.5%)	.052	114 (83.2%)	39 (83.0%)	.971
Tests for Fentanyl	102 (55.4%)	28 (54.9%)	74 (55.6%)	.928	75 (54.7%)	27 (57.5%)	.748
Offers Medical Care	145 (78.8%)	41 (80.4%)	104 (78.2%)	.744	107 (78.1%)	38 (80.9%)	.691
Offers Drug Counseling	138 (75.0%)	39 (76.5%)	99 (74.4%)	.775	105 (76.6%)	33 (70.2%)	.380
Offers Entry to Detox	158 (85.9%)	42 (82.4%)	116 (87.2%)	.396	119 (86.9%)	39 (83.0%)	.510
Offers Start to Methadone	129 (70.1%)	32 (62.8%)	97 (72.9%)	.177	99 (72.3%)	30 (63.8%)	.276
Offers Safe Works Disposal	136 (73.9%)	32 (62.8%)	104 (78.2%)	.033	100 (73.0%)	36 (76.6%)	.627
Offers Safety from Policea	156 (86.7%)	40 (78.4%)	116 (89.9%)	.041	113 (85.0%)	43 (91.5%)	.258

^aP-value for Pearson χ2 – test of independence.

^bDue to item non-response valid n is 180; 4 persons with a lifetime history of overdose were missing on the item asking about safety from police.

2.3. Analytical Methods

We present descriptive statistics to summarize the characteristics of the sample. We used ordered logistic regression to compare persons with and without a lifetime history of drug overdose on an ordered categorical measure assessing how often they would use a SIF if it was available. We used the Pearson χ² – test of independence to compare persons with and without a lifetime history of drug overdose, and persons who had recently experienced homelessness, with respect to the likelihood of using SIFs if the sites offered specific services.

3.0. Results

Participants averaged 34.2 (± 8.3) years of age. About 68.5% were male, 85.9% were White, 4.4% were Black, and 9.8% identified mixed or other racial origins. Fifteen (8.2%) were Hispanic. Mean age of initiating heroin or other opioid use was 20.6 (± 6.2), and mean years of use was 13.6 (± 8.8). Only one person reported they had not used heroin in the past month and on average participants reported injecting drugs on 23.9 (± 9.7, Median = 30) of the past 30 days; 117 (64.0%) reported daily injection drug use. One hundred thirty-three (72.3%) reported a lifetime history of drug overdose. Participants were asked, “If one were convenient for you to get to, how often do you think you would use a Safe Injection Facility?” Sixty-four (35.0%) reported “never”, 55 (30.1%) “once in a while”, 29 (15.9%) “most times I was ready to use,” and 35 (19.1%) reported “always.”

Persons with a lifetime history of drug overdose were significantly more likely (OR = 1.98, z = 2.25, p = .025) to use a SIF that was conveniently located. Among persons with a lifetime history of drug overdose (n=133) the expected probabilities were 0.31, 0.30, 0.17, and 0.22 of saying they would “never,” “once in a while,” “most times,” and “always” use an available SIF, respectively. This compares to estimated probabilities of 0.47 (“never”), 0.29 (“once in a while”), 0.12 (“most times”, and 0.12 (“always”) for the 51 participants with no history of overdose.

Table 1 illustrates participant likelihood to use a SIF if the facility offered a range of services and compares SIF use by lifetime overdose history and homelessness in the past three months. The services associated with the greatest likelihood to use a SIF were “providing a place where police wouldn’t bother you” (86.7%), offering entry into managed withdrawal programs (85.9%), and providing clean needles (83.2%). About 78.8% reported they would be more likely to go to a SIF if it offered medical care, 75.0% if it offered drug counseling, 73.9% if it offered safe disposal of works (i.e., equipment for injecting drugs), and 70.1% if it offered initiation into methadone maintenance. Only 55.4% reported that a SIF offering testing for fentanyl would make them more likely to go to a SIF

Persons with a lifetime history of drug overdose were significantly (χ² = 4.17, p = .041) more likely to report they would go to a SIF if it offered safe disposal of works (78.2% vs 62.8%; χ² = 4.56, p = .033) and safety from the police (89.9% vs 78.4%; χ² = 4.18, p = .041) (Table 1). Also, a higher proportion of participants reported endorsing use of a SIF if it provided clean needles (86.5% vs 74.5%, χ² = 3.76, p = .052). History of overdose was not associated significantly with the likelihood of using an SIF if the other services were offered. The likelihood of using a SIF did not differ significantly by recent homelessness.

4.0. Discussion

Consistent with prior research, in the current sample of PWID, a majority reported a willingness to use SIFs. Although over one-third of participants were never willing to use a SIF, participants reported greater willingness to use an SIF if it offered specific services. For example, in all, more than eight in ten PWID reported greater endorsement of SIFs that offered safety from the police during drug injection activity, possible entry into treatment, or clean needles. Of the eight services, PWID with a history of overdose reported being more likely to endorse using a SIF offering works disposal and safety from police than those without an overdose history. Interest in the four services related to medical or addiction treatment did not differ by overdose status. Despite the known association between fentanyl use (which is often unintended) and overdose risk, PWID with prior overdose were no more interested in an SIF’s ability to test for fentanyl than PWID with no prior overdose, perhaps because many PWID in this region report intentional use of fentanyl ^21,22.

That more than eight in ten treatment-seeking PWID would be more likely to use a SIF that offered safety from police, entry into addiction treatment, and clean needles sheds important light on PWID’s needs. Public health initiatives that meet PWID’s desire to engage in safer IDU behaviors (e.g., use of clean needles) in one location hold promise to prevent infectious complications (e.g., HIV, hepatitis C, abscesses, sepsis, endocarditis). Second, PWID’s amenability to using SIFs that could offer entry into withdrawal management indicate that SIFs may act as a conduit to long-term treatment programs offering medications for opioid use disorder ⁴. However, more research is needed that examines the service needs and potential linkage among non-treatment seeking PWID. Third, consistent with qualitative research e.g., ^15,23 PWID in this sample exhibited concerns about arrest and risk of incarceration, consistent with prior research highlighting the negative effect that the criminalization of drug use has on risky injection use behaviors ^24–26. This finding may reflect the turmoil and risk associated with public injection behaviors among PWID, and the need for law enforcement and SIFs to work together to better balance community safety and overdose risk. In sum, that PWID’s willingness to use a SIF differed based on services offered points to the need to assess and advertise specific harm reduction services that appeal to PWID. Further, interest in certain services may differ by risk status (e.g., overdose history) while others may not (e.g., homelessness), providing further support for more comprehensive examination of services of greatest interest to subgroups of PWID.

Findings showing that PWID with a history of overdose, relative to those without, were more willing to use a SIF if it offered safety from police may indicate that those with a history of overdose have had more contact with police and emergency services in the context of an overdose. Despite some legal protections afforded by the Massachusetts Good Samaritans Law, many PWID remain unaware of these laws ^27,28. Additionally, PWID report being fearful of calling 911 due to factors such as past mistreatment and distrust of police ²⁹. That services associated with treatment seeking (i.e., medical care, drug counseling, entry into opioid withdrawal management, methadone treatment) did not differ by overdose status was surprising. However, given that respondents in the present sample were seeking treatment for heroin use suggests they acknowledge their risk for overdose and other adverse consequences of injection drug use whether they have experienced an overdose or not.

Surprisingly, PWID reporting homelessness in the past three months were no more willing to use a SIF for specific services, including hygienic injection equipment and protection from police, than those not experiencing recent homelessness. Relative to PWID with stable housing, homeless PWID are more likely to inject in secluded public places, hence escalating risks for violence, arrest, unsanitary injection conditions, and fatal overdose ¹⁵. Still, because many PWID with relatively stable housing may still inject in public settings we cannot assume that homelessness serves as a proxy for injecting publicly (as opposed to privately); research that accounts for public versus private injection practices is warranted ²².

4.1. Limitations

There are several study limitations to note. First, the current study surveyed PWID entering an opioid withdrawal management program. Relative to PWID not seeking withdrawal management, persons seeking withdrawal management may be more amenable to using SIFs and their accompanying services to change their drug use behavior. For instance, this sample’s amenability towards SIFs that offer entry into treatment may be inflated. Although studies that assess SIF service interest among a wider PWID sample may offer broader insight into interest in SIFs, convenience samples using persons seeking opioid treatment have commonly been used in literature assessing attitudes about novel harm reduction measures (e.g., naloxone administration, buprenorphine treatment formulation, pre-exposure prophylaxis (i.e., PrEP) ^30–32. Second, given the absence of US-based SIFs as an actual harm reduction option presently, it is important to note that this study assesses perceptions about SIFs conceptually rather than an actual SIF service model. Future qualitative research that explores PWID’s attitudes about SIF services used internationally may shed light on perceptions rooted in experience. Third, we assessed eight services represented in SIFs internationally, but this may not have captured all possible services that SIFs could offer and more research is needed. Finally, it is important to note that the timing and number of participants’ overdose history, and knowledge about fentanyl and prior fentanyl exposure is unknown; more recent overdoses may have played a role in amending attitudes and, in turn, may differentially impact interest in SIFs and provided services.

4.2. Conclusions

If SIFs become available in the United States, greater awareness about which SIF services might attract PWID could inform policy decisions. Replicating our findings in other locations and with other cohorts of PWIDs will provide deeper insights into the reasons that PWID are interested in specific services to aid service delivery at SIFs. Here, as abroad, SIFs could meet a demand for safer drug use practices and become entry points to treatment.