Safety in solitude? Competing risks and drivers of solitary drug use among women who inject drugs and implications for overdose detection

Abstract

Background and aims:

Solitary drug use (SDU) can amplify risks of fatal overdose. We examined competing risks and drivers of SDU, as well as harm reduction strategies implemented during SDU episodes, among women who inject drugs (WWID).

Design:

A cross-sectional qualitative study, including telephone and face-to-face in-depth interviews.

Setting:

Baltimore City, MD, USA.

Participants:

Twenty-seven WWID (mean age = 39 years, 67% white, 74% injected drugs daily) recruited via outreach and street intercept (April–September 2021).

Measurements:

Interviews explored the physical (i.e. indoor/private, outdoor/public) and social (i.e. alone, accompanied) risk environments in which drug use occurred. Guided by the principles of emergent design, we used thematic analysis to interrogate textual data, illuminating women’s preferences/motivations for SDU and strategies for minimizing overdose risks when using alone.

Findings:

Many participants reported experiences with SDU, despite expressed preferences for accompanied drug use. SDU motivations clustered around three primary drivers: (1) avoiding opioid withdrawal, (2) preferences for privacy when using drugs and (3) safety concerns, including threats of violence. Participants nevertheless acknowledged the dangers of SDU and, at times, took steps to mitigate overdose risk, including naloxone possession, communicating to peers when using alone (‘spotting’) and using drugs in public spaces.

Conclusions:

WWID appear to engage frequently in SDU due to constraints of the physical and social environments in which they use drugs. They express a preference for accompanied drug use in most cases and report implementing strategies to mitigate their overdose risk, especially when using drugs alone.

INTRODUCTION

Drug overdose is a protracted public health crisis, which has recently worsened during the COVID-19 pandemic. Overdose fatalities in the United States reached record-shattering levels, claiming more than 100 000 lives between May 2020 and April 2021; this represents a 29% increase in fatal overdoses from the prior reporting period (78 056 deaths) [1]. In a large study of 11 884 opioid overdose deaths examined in 11 US states, most (> 56%) deceased individuals were found alone, without a bystander present [2], suggesting that most fatal overdoses probably occurred during solitary drug use (SDU) episodes.

SDU is a well-established behavior among people who use drugs (PWUD) that involves using drugs in private or public spaces without others present. SDU poses a challenge for overdose detection and response efforts, as it minimizes opportunities for bystander intervention with naloxone and/or emergency medical services [3]. Studies have demonstrated the ubiquity of SDU among PWUD; for example, two US studies among PWUD recruited in clinical and community settings found that PWUD used drugs alone for approximately half of the time [4, 5]. Another study reported that among women who inject drugs (WWID) in New York City, 59% of injections were solitary events [6]. These survey findings corroborate broader trends detected during investigations of overdose fatalities [2, 7].

SDU may be more common among specific subgroups of PWUD and in specific drug use contexts. Previous studies have demonstrated heightened propensities for SDU among older women, individuals who use at home and in the context of polysubstance use [6, 8, 9]. Qualitative studies have found that SDU can optimize the drug use experience by minimizing external distractions and help PWUD to control the settings in which they use drugs [10–12]. Drug use stigma and criminalization also marginalize PWUD and drive social isolation, precipitating SDU and, consequently, elevating solitary overdose risk [13–15].

Nevertheless, social or accompanied drug use, defined as using drugs in the presence of others, also carries its own set of risks, from physical or sexual violence to infectious disease transmission (i.e. HIV and hepatitis B and C). Evidence suggests these risks may be amplified for women. A study of WWID in New York City found that the risk of injection equipment sharing was highest when women used drugs with intimate partners [6]; women’s motivations for SDU could, therefore, stem from intentions to avoid higher-risk drug use encounters. Ethnographic studies have also shown that women may prefer using drugs in private spaces or concealed settings, allowing them to avoid coercion and violence while using drugs in the presence of men [11, 16]. While these SDU events can minimize specific health risks to women and allow them to re-exert agency over the drug use experience, SDU can also heighten the risk of fatal overdose. As COVID-19-related control measures and service disruptions continue to drive increases in SDU among PWUD [17, 18], examining SDU drivers, specifically for women, and identifying strategies for minimizing harms to WWID using drugs alone, have never been more urgent.

In response, we qualitatively examined WWID’s motivations for SDU, as well as harm reduction strategies implemented by WWID during these episodes. We conducted this study in Baltimore City, which consistently ranks among the most impacted municipalities in Maryland and the United States by drug overdose deaths, with an age-adjusted overdose mortality rate of 96 per 100 000 population [19]. In 2020, 1028 substance overdose deaths occurred in Baltimore City, 94% of which were opioid-related [19]. Our findings help to illuminate the risk environments underpinning SDU as well as their implications for women-centered overdose prevention interventions, which traditionally rely upon the presence of bystanders and their ability/willingness to intervene.

METHODS

Setting and procedures

We derived data from the Optimizing PrEP engagement Among women Living in Baltimore City (OPAL) study–a formative, mixed-methods research study aiming to develop and pilot a work-force development intervention capacitating frontline harm reduction workers to promote HIV pre-exposure prophylaxis (PrEP) to PWUD [20]. In the first phase, we facilitated four virtual focus group discussions with 20 HIV and substance use service providers to characterize the PrEP implementation environment for women in Baltimore City. Findings from this first phase guided our sampling plan and research questions for the study’s second phase, where we explored barriers and facilitators to PrEP willingness and uptake among WWID.

Between April and September 2021, we conducted telephone and face-to-face semi-structured, in-depth interviews with adult (≥ 18 years) cisgender women living in Baltimore City who reported injecting any drug not prescribed by health-care provider in the past 6 months. Due to COVID-19 restrictions and safety concerns we began recruiting women for telephone interviews (April–June 2021), then expanded to in-person recruitment after COVID-19 restrictions began to lift. For telephone interviews (n = 8, 30%), we recruited WWID through recruitment flyers distributed to partner organizations providing substance use services. Flyers directed callers to a study staff member, who screened individuals for study eligibility. For face-to-face interviews (n = 19, 70%), we approached women outside harm reduction service organizations (e.g. community-based organizations, mobile syringe services programs), shared study details and screened those expressing interest. We excluded women self-reporting HIV-positive status.

For WWID screening eligible and providing verbal informed consent, we administered a brief (~5 minutes) structured questionnaire measuring demographics, sexual partnership characteristics, substance use histories and PrEP knowledge. Interviews lasted 20–60 minutes in duration and explored various dimensions of substance use and HIV prevention, including lived experiences and daily priorities; settings and contexts in which WWID used drugs (e.g. solitary use); and health-care harm reduction and substance use service-seeking experiences.

Guided by the principles of emergent design, we incorporated additional questions interrogating the physical (i.e. indoor/private, outdoor/public) and social (i.e. alone, accompanied) environments of drug use, given the ubiquity of SDU in narratives early in data collection [21]. We specifically included questions exploring women’s preferences and experiences with SDU, including the frequency of SDU episodes, drivers and perceived risks of SDU, and strategies to minimize overdose risks when using drugs alone. Interviewed WWID received a $50 prepaid visa card for their participation. The Johns Hopkins Bloomberg School of Public Health Institutional Review Board approved the study protocol.

We audio-recorded interviews and submitted recordings for professional transcription. Interviewers checked each transcript for clarity and verified accuracy by comparing transcribed text to a random 5–10-minute recorded segment of the interview.

Analysis

We implemented a hybrid inductive–deductive, team-based thematic analysis to interrogate motivations for SDU and discern narrative patterns in SDU experiences [22]. First, we read transcripts line-by-line to gain familiarity and increase proximity to the data. Following multiple rounds of close reading, we generated and circulated memos summarizing initial impressions of the drivers and risks associated with SDU, from which we developed a preliminary set of textual codes. After refining these themes through additional rounds of transcript review and discussion, we collapsed these initial themes into more concise, discrete (focused) codes. We piloted the draft codebook by manually coding four transcripts in duplicate, identifying sources of (dis)agreement in code application.

After refining and finalizing the codebook, we imported transcripts and survey data into Dedoose version 9.0 (SocioCultural Research Consultants, Manhattan Beach, CA, USA) for data management and coding. Three interviewers independently applied codes to transcripts, and the first author (J.G.R.) reviewed all coded output for consistency and codebook fidelity. We exported coded text segments pertaining to SDU frequency and contexts (e.g. daily, public/private settings), harm reduction practices in the contexts of SDU (e.g. support from peers, naloxone possession) and drivers of SDU (e.g. convenience, stigma) and reviewed them horizontally (across interviews) to identify thematic salience [23]. Guided by the tenets of constant comparison, we reassembled coded text segments by populating analytical matrices, which plotted emerging themes surrounding SDU motivations and perceived risks against participant attributes captured quantitatively (from surveys) and qualitatively (from interviews) [24]. Lastly, continuous discussion among study team members helped to refine and crystalize themes and narrative patterns.

In our presentation of results, we define qualitatively anchored terms such as ‘many’ and ‘most’ as present in more than half of the transcripts; ‘some’ as present in more than a quarter but fewer than half of transcripts; and ‘few’ as present in fewer than a quarter of transcripts [25].

RESULTS

Table 1 presents demographic and behavioral characteristics of interviewed WWID (n = 27). The mean age was 39 years (range = 21–60 years). The racial composition was as follows: white (67%), black (26%) and American Indian/Alaska Native (15%). Only two participants identified as Hispanic/Latina (7%). Polysubstance use was ubiquitous (96%), with a majority of WWID reporting use of opioids (i.e. heroin, fentanyl) (96%), stimulants (i.e. crack, cocaine, methamphetamines) (93%) and alcohol (52%) in the past month. Nearly all participants reported using both opioids and stimulants in the past month (89%). Among WWID who reported current (i.e. past month) injection drug use (78%), most injected drugs daily (71%). A majority reported a history of sex work (89%), most of whom sold or traded vaginal and/or anal sex for money or things such as food, drugs, housing or favors in the past month (83%).

TABLE 1.

Demographic and behavioral characteristics of women who inject drugs in Baltimore, Maryland (n = 27).

Characteristics	Number (n)	Percentage (%)
Interview modality
Face-to-face (in-person)	19	70.4
Telephone	8	29.6
Age, in years (mean, range)	38.7	21-60
Race and ethnicity
White	18	66.7
Black	7	25.9
American Indian/Alaska Native	4	14.8
Hispanic/Latina (any race)	2	7.4
Substance classes used, past month
Opioids (heroin, fentanyl)	26	96.3
Stimulants (crack, cocaine, methamphetamines, speed)	25	92.6
Alcohol	14	51.9
Marijuana	9	33.3
Club drugs (ecstasy/MDMA, Rohypnol, GHB)	7	25.9
Non-prescribed pain medications (OxyContin, codeine)	7	25.9
Polysubstance use, past month
No	1	3.7
Yes	26	96.3
Time since last injection drug use
< 1 month	21	77.8
1-3 months	6	22.2
Injection drug use frequencya
Daily	15	71.4
Non-daily	6	28.6
Shared injection equipment, past 6 months
No	19	70.4
Yes	8	29.6
Time since last sex workb
< 1 month	20	74.1
1–3 months	1	3.7
> 12 months	3	11.1
Never	3	11.1

^aMeasured among those who reported injection drug use in the past month;

^bdefined as selling or trading vaginal or anal sex for money or things such as food, drugs, housing or favors.

MDMA = methylenedioxymethamphetamine; GHB = gamma-hydroxybutyric acid.

Most participants qualitatively reported regularly using drugs in the presence of others and preferring accompanied drug use to SDU. This was primarily driven by safety concerns surrounding overdose response, satisfying social expectations to use drugs with others (i.e. using drugs with romantic partners who purchased drugs) and requiring or offering mechanical assistance for injecting drugs (i.e. ‘needing help getting on’).

Nevertheless, many WWID reported experiences with SDU, despite stated preferences for accompanied drug use. For women who preferred using drugs in the presence of others, these SDU experiences were sporadic or occasional. The few women who preferred using alone reported frequent (i.e. nearly daily) SDU episodes. Motivations for using drugs alone clustered around three primary themes: (1) managing withdrawal symptoms, (2) preferences for privacy during drug use episodes and (3) safety concerns surrounding accompanied drug use (see Fig. 1).

FIGURE 1.

Competing risks and drivers of solitary drug use emerging from narratives of women who inject drugs in Baltimore, Maryland

‘Getting well’ and numbing the pain of dopesickness

WWID situated SDU episodes in the daily management of debilitating opioid withdrawal symptoms (‘dopesickness’), which participants prioritized among a constellation of competing needs. One participant experiencing homelessness viscerally detailed the severity of her opioid withdrawal symptoms, illustrating how ‘getting well’ took precedence over other priorities.

I start off my day ‘getting well’, which is getting my heroin because if I don’t, it affects my bones. It affects my body… It affects everything. I’m throwing up, going to the bathroom, don’t want to move, sweating… You cannot function.

In the context of ‘chasing’ opioids to cope with painful withdrawal symptoms, WWID reported using drugs alone out of convenience, especially participants who lived alone. They prioritized withdrawal management over other drug use preferences, including where or with whom they used drugs. One participant described how the severity of her withdrawal symptoms determined the settings and places where she used heroin, explaining how her ‘dopesickness’ could drive her to use alone, despite her preference for using drugs at home with her partner present.

I don’t like being by myself because I’m afraid I might die. At least somebody’s around me [at home]… I just feel safer. I like to be home, but if I really have to get it [heroin] in me, I’ll do it out here [on the street].

Preferences for privacy in the context of drug use stigma and social pressures to share substances

Another key driver of SDU was the desire for privacy when using substances. Women’s narratives emphasized two discrete motivations for privacy: one attributed to internalized shame surrounding drug use and another to increased agency over the drug use experience. For some WWID, stigma related to injection drug use motivated their preference for using alone. One WWID experiencing homelessness reported routinely injecting drugs alone, often in alleyways, in an attempt to stay hidden from others, stating: ‘I don’t like people to look at me as a junkie’. Others, similarly, described their SDU to be driven by keeping their drug use secretive and hidden from others, including family and roommates. One woman who lived with her mother explains her preference for keeping her drug use private motivated her decision to always inject alone:

[My mother] doesn’t like it… I prefer to be alone… I like being by myself… I’d prefer it if she didn’t even know.

By comparison, a few other participants preferred SDU to avoid undesirable interactions with others while using drugs. One woman, who reported selling sex as her primary source of income, explained her preference for using heroin alone in order to avoid uncomfortable encounters with other PWUD who could coax or pressure her to share her drugs.

There are times when I don’t want to be around anybody else because I don’t want anyone to ask me for anything… Not that I’m greedy—it’s just sometimes you want to do something by yourself… I don’t want to share with someone. I just want the whole feeling for me… Especially considering how most of my money comes. Sometimes I’m really like, ‘No, I just had to do this and that, and no, I don’t want to share’.

‘If I passed out, they could do anything to me’: searching for safety in SDU

A few WWID also expressed safety concerns with accompanied drug use or using drugs around others. While most participants explained that they felt safer using drugs around others, especially romantic partners or trusted friends/peers, the few WWID who preferred using drugs alone perceived accompanied drug use as more dangerous than SDU. One participant who reported always using drugs alone explains that using drugs around others could render her more vulnerable to physical harm, especially if she were to ‘nod off’ or become incapacitated while using:

I think it’s more dangerous to inject around somebody because if I passed out, they could do anything to me.

SDU motivated by safety concerns reflected WWID’s mistrust of others in their environment, including peers, sex work clients and others involved in the drug economy (e.g. dealers). One participant, who described a recent experience with attempted kidnapping while ‘dating’ (selling sex), explains that her decision to use drugs alone stemmed from concerns surrounding her safety when using drugs around others in her neighborhood.

I just don’t want to be around this neighborhood anymore… because of the people around here… They like to rob you, mace you, set you up… for drugs… They’re schiesty. I can’t trust them.

‘Don’t let me die’: strategies utilized to minimize risks of SDU

WWID disclosing SDU experiences overwhelmingly acknowledged the dangers of using drugs alone, specifically those related to overdose risk. Among women reporting frequent SDU episodes or expressing preferences for unaccompanied drug use, they described three strategies they implemented to maximize safety when using drugs alone: (1) naloxone possession, (2) communicating to peers when using alone and (3) using drugs in public spaces.

Most WWID, regardless of drug consumption behaviors or preferences, reported personal access to naloxone and self-efficacy administering naloxone in the event of an opioid overdose. For participants using drugs alone, access to naloxone offered some security against a solitary overdose:

I always keep the Narcan on me, so if I feel a little funny, I know to hit myself with that before it [the heroin] hits me really hard.

Among women reporting SDU, most also alluded to taking active steps to prevent an overdose by ‘spotting’ (letting others know when they intended to use drugs alone). One woman, who preferred using alone in a vacant house where she lived, explained that she would usually inform others in her vicinity of her intentions to use alone, should she overdose:

I’ll let someone know… I’m going up here [to the vacant house] to ‘get on’ [inject drugs], just for my safety… ‘If you don’t hear from me in so and so [time], can you check in on me?’

Lastly, a few participants, particularly those experiencing homelessness, reported explicitly using drugs in public spaces (e.g. alleyways, parks, cemeteries) to increase their visibility in the event of an overdose. One participant explained that she always injected drugs outside the abandoned house where she stayed, so others could respond if she overdosed:

If I’m inside the house, and I go out [overdose], nobody would even know that I went out. If I’m outside, then they at least see that I’m out, and they could hit me with Narcan.

Another woman experiencing homelessness explained that when using drugs alone, she would always find public spaces frequented by other PWUD, who could intervene should she overdose:

In an alley, this little patch of woods everybody gets high in… There’s usually other people around… I keep Narcan on me constantly, and if there’s somebody else there, I always say the same thing: ‘Not that I think there’s going to be a problem, but just in case, there is Narcan in my bag if you need it. Please don’t let me die.’

DISCUSSION

Our interviews revealed that while most WWID expressed preferences for accompanied drug use, experiences with SDU were pervasive and shaped by the physical and social environments in which drug use occurred (see Fig. 1), expanding insights into this understudied topic. Importantly, WWID acknowledged the heightened risk of overdose when using drugs alone and reported specific practices, from naloxone possession to public drug use, to minimize these harms. In other words, WWID took active steps to facilitate timely overdose detection and response during episodes of SDU, which were shaped by the realities of their housing statuses and the absence of public health interventions to safely support these practices. Studies describing SDU have characterized some of the motivations of using drugs alone, but few have described strategies women implement to minimize risks associated with SDU [16]. Conceptualizing WWID’s agency to mitigate harms related to SDU is critical to identifying harm reduction strategies that prioritize safety while dignifying WWID’s autonomy and respecting their preferences related to optimal settings and contexts for drug use.

The extant drug overdose literature characterizes SDU as unsafe [3, 14, 26] but, in our qualitative study, some women who preferred using drugs alone described SDU as a safer alternative to accompanied drug use due to threats of violence or robbery when using around others. Given differences in men and women’s conceptualization of injection-related practices (i.e. syringe sharing) reported in other studies, the gendered manifestations of violence featuring prominently in women’s narratives help to contextualize the ubiquity of SDU among WWID in the present study [27, 28]. For some WWID, the physical dangers of using drugs around untrustworthy others, especially in the context of selling sex to purchase drugs, outweighed the risks of using drugs alone, which many felt could be minimized through feasible harm reduction practices. Because safety concerns shape women’s decision-making regarding the settings and contexts in which they use drugs [11, 16, 29, 30], revisiting traditional naloxone distribution models that exclude the needs of women who prefer SDU are urgently needed. Supervised consumption sites with private booths, or other spaces where women can consume drugs purchased elsewhere under trained/trusted supervision, are one solution for detecting overdoses that would otherwise occur in isolation while reducing safety concerns surrounding violence [5, 31, 32]. There is also burgeoning interest in leveraging overdose detection technologies (e.g. mobile applications, naloxone autoinjectors) to address gaps in overdose response for women who use drugs alone [33, 34].

WWID also prioritized withdrawal symptom management in their decision-making surrounding the settings and contexts in which they used drugs. Other studies have attributed SDU to the perceived convenience of using drugs alone, especially when alleviating symptoms of ‘dopesickness’ [30, 35]. Supervised consumption sites would, therefore, insufficiently appeal to women whose pressing withdrawal symptoms might demotivate travel to these sites [36]. Mobile technologies that reduce the window between an overdose emergency and response require neither the presence of bystanders nor accompanied drug use [37–39]. These may be acceptable alternatives to existing harm reduction interventions for WWID, who—relative to men—may have fewer resources to address ‘dopesickness’ and autonomy over the drug use process, particularly in romantic partnerships where women may be expected to share drugs with intimate male partners [40, 41]. Additionally, expansion of drug-checking programs could mitigate some risks associated with SDU [42, 43].

Preferences for SDU also emerged in the context of internalized drug use stigma and unwanted attention from others during drug use episodes. Injection-related stigma is a particularly critical dimension of SDU that underpins gendered differences in drug use practices and behaviors [10, 13, 30]. For instance, reconciling social meanings of gender identity and injection drug use, which can compete with norms governing women’s caregiving identities, has been linked to WWID’s avoidance of substance use services, including syringe service programs [40]. Similarly, intentional avoidance of others when using drugs has been shown in prior research to remove external distractions and prevent harassment by others, facilitating an undisturbed and more pleasant drug consumption experience [10, 16, 29]. Anticipated stigma and the desire for privacy when using drugs has important implications for harm reduction interventions, particularly supervised consumption sites. Using booths or private rooms and privacy at points of entrance and exit may appeal to women whose SDU practices are influenced most saliently by stigma or privacy concerns [16, 44, 45].

Our findings are subject to several limitations. First, we recruited women through flyers distributed to community-based organizations and street intercept outside of harm reduction services. Insights gleaned from interviews may, therefore, not reflect the motivations and experiences of WWID not accessing harm reduction or other social services, whose drug use behaviors and environments may be incomparable to those of WWID represented in our study [46, 47]. Secondly, because SDU was not the primary focus of the parent study, our purposive sampling approach was not guided by traditional tenets of thematic saturation for this singular topic [48]. Nevertheless, following the principles of emergent design [21, 22], we integrated questions related to SDU motivations and experiences following its emergence in WWID’s narratives. Lastly, we relied exclusively upon interviews to explore drivers and risks associated with SDU and did not leverage other qualitative methods (e.g. prolonged observations, focus group discussions with service providers), through which complementary or competing insights might have emerged.

Nevertheless, our study of the competing risks and drivers of SDU among WWID uniquely identified the agency women exert over their overdose risk when using drugs alone, contributing to the nascent US literature on this topic. Our findings help reconceptualize ‘safety’ in the context of SDU, as WWID emphasized the physical dangers of using drugs around others, which the extant overdose literature tends to characterize as a safer alternative to SDU. Given that the perceived rewards of using drugs alone for some WWID outweighed the risks, which many felt could be minimized through harm reduction practices (i.e. naloxone possession, ‘spotting’, public drug use), gender-responsive alternatives to naloxone distribution programs, including supervised consumption sites and overdose detection technologies, are imperative to timely overdose detection and response for people using drugs alone. Future studies should examine the acceptability and feasibility of these novel overdose detection and response interventions among women, ensuring they are adequately tailored to the needs and priorities of women engaging in SDU.