“Generally, you get 86’ed because you’re a liability”: An application of Integrated Threat Theory to frequently witnessed overdoses and social distancing responses

Abstract

While rates of opioid overdose deaths in North American have increased exponentially in recent years, most overdoses are not fatal, especially when witnesses are present and can intervene. Previous research has found that some people who use drugs [PWUDs] trained in overdose response might cut social ties with frequent overdosers, leading to more solitary opioid use and risk of death if someone overdoses alone. To examine the phenomenon of social distancing of people who overdose frequently, we used data from fifty-two in-depth qualitative interviews collected in Southern California with PWUDs who had recently witnessed an opioid overdose. Transcripts were reviewed and coded thematically, using the Integrated Threat Theory (ITT) to conceptualize the observed phenomenon. ITT outlines how realistic and symbolic threats are experienced by a group. We found that while some participants acknowledged the role of adulterated street drugs in overdoses, individualized blame was nonetheless imposed. Accusations of careless drug use practices fostered negative stereotyping towards frequent overdosers. This was attributed to the need to summon 911 for rescue, which often resulted in police dispatch. The intergroup relationship between police and PWUDs is precarious as police pose realistic threats onto PWUDs - such as incarceration, eviction, and manslaughter - leading to intragroup anxiety among PWUDs about future overdose events, and labelled frequent overdosers as liabilities. These threats, and inter/intra-group conflict, explained one reason how and why non-fatal overdoses led to social distancing events. People who overdose frequently were also accused of breaking the norm of drug user surreptitiousness; a symbolic threat that endangered the group due to police exposure. Social distancing might dampen exposure to the protective effect of peer-led interventions such as take-home naloxone programs, increasing risk of overdose death. This phenomenon highlights how intergroup dynamics are driving intragroup processes. Suggestions for tailoring public health interventions are discussed.

INTRODUCTION

In North America, fatal opioid overdoses are escalating and have been a difficult-to-manage public health issue. In 2018 in the United States [U.S.], 46,802 persons died from opioid overdoses, representing 69.5% of all overdose deaths, with synthetic opioids like fentanyl and its analogues increasingly attributed (Wilson, Kariisa, Seth, Smith, & Davis, 2020). From a social-ecological perspective, overdose determinants are multifarious and exist on a continuum ranging from individual behaviors (e.g. solitary drug use, polydrug use) to community responses (e.g. stigma and social rejection) to policy directives (e.g. the effects of the criminalized drug environment) (Ciccarone, 2017; Davidson, 2003; Rhodes, 2002). An important factor from the policy level includes unregulated and contaminated street opioid products in the drug market, which result in unknown drug quality prior to consumption (Ciccarone, 2017). Notably, in spite of recognition by people who use drugs [PWUDs] of this policy-level issue impacting overdose risk (e.g., adulteration of street drugs with powerful opioids such as fentanyl), individual PWUDs are still sometimes blamed for overdoses because of their personal drug using practices (Bowles & Lankenau, 2018). For example, an individual might be accused of consuming too large of a drug dose that led to an overdose event.

Fellow PWUDs are often witnesses and first responders to opioid overdoses (National Institutes of Drug Abuse [NIDA], 2019; Wheeler et al., 2014). Responses to witnessed overdoses by PWUDs have included injecting the person overdosing with salt or milk, placing ice on their genitals, placing the person in an ice bath or cold shower, or slapping the person or causing pain to stimulate them (Bowles & Lankenau, 2018; Koester, S., Mueller, S. R., Raville, L., Langegger, S., & Binswanger, I. A., 2017; Pollini, R. A., McCall, L., Mehta, S. H., Celentano, D. D., Vlahov, D., & Strathdee, S. A., 2006; Tracy, M., Piper, T. M., Ompad, D., Bucciarelli, A., Coffin, P. O., Vlahov, D., & Galea, S., 2005, Tobin, Davey, & Latkin, 2005; Wagner et al. 2019). In each of these studies, participants mentioned reluctance, hesitation, or refusal to contact 911 due to prior bad experiences with police during overdose events, such as incarceration. In a recent study by Wagner et al. (2019), respondents described 911 as tantamount to contacting police thereby creating fear that calling 911 for any circumstance could result in harmful policing actions like arrest and damage to one’s reputation. In response to this issue, take-home naloxone programs, which began in the U.S. in 1996, have quickly expanded across the U.S. since 2010 (McDonald & Strang, 2016; Wheeler, Jones, Gilbert, & Davidson, 2015). Take-home naloxone programs are voluntary programs in which laypersons are trained to respond to witnessed overdoses by conducting rescue breathing, contacting 911, and administering the medication naloxone to reverse the effects of opioid overdose (Clark, Wilder, & Winstanley, 2013; McDonald & Strang, 2016; Wheeler, Jones, Gilbert, & Davidson, 2015). PWUDs are the primary intended audience for this intervention, as they are those most likely to witness overdoses and therefore are able to intervene expeditiously (World Health Organization, 2018). As such, the success of take-home naloxone programs is contingent on PWUDs consuming drugs in the presence of people who have naloxone and are willing to respond in the event of an overdose.

Social interactions within drug using social groups include drug use initiation, drug purchasing, and drug consumption; these interactions can shape drug using behaviors (Kumar, McNeely, & Latkin, 2016; West, 2019). Interactions among persons who use drugs also involve social support, resource sharing, assistance offsetting withdrawal, friendship, love, and camaraderie (Bourgois & Shonberg, 2009; Bowles & Lankenau, 2018; Lapinski, M. K., & Rimal, R. N., 2005, Syversten et al., 2014), many of which are features of typical social groups, drug-using or not. In addition to providing social support, drug using groups typically have established norms and expectations of appropriate behavior (Kumar, McNeely, & Latkin, 2016; Lapinski, M. K., & Rimal, R. N., 2005). Notably, a key norm within drug using groups is the commitment to maintain group surreptitiousness, which is critical because drug use is criminalized and drawing unwanted police attention exposes the group (Bonevski et al, 2014). Overdosing, although unintentional, could be one behavior that breaks this commitment, as 911 might be summoned for medical intervention thus exposing the group to police, who pose serious threats to the group’s freedom and social function.

In fact, in a study examining the impact of PWUDs participating in overdose prevention and naloxone trainings, Wagner et al. (2014) found that in spite of many positive effects associated with being trained, some PWUDs experienced some notable negative emotions such as fear, anger, burden, and regret following witnessed overdoses. In some instances, these negative effects resulted in “cutting social ties” (161) with persons who overdosed frequently (Wagner et al., 2014). Isolation from other drug-using peers as a consequence of frequent overdose is concerning because people who have experienced previous overdoses are at elevated risk for overdose death, a risk that could be further exacerbated if they are using alone and therefore no one is available to intervene (Caudarella, Dong, Milloy, Kerr, Wood, & Hayashi, 2016; Davidson et al., 2003; Stoove, Dietze, & Jolley, 2009).

In the present paper, we suggest that those who frequently overdose might experience social distancing imposed by their peers on account of an overdose necessitating contacting 911 for rescue, that in tandem summons police and accompanying threats (Latimore & Bergstein, 2017; Stephan & Stephan, 2000; Tobin, Davey, & Latkin, 2005; Wagner et al., 2019). Stephan & Stephan (2000) expanded on the origins and repercussions of such threats in the Integrated Threat Theory [ITT], a framework outlining how “realistic and symbolic” threats are experienced by a group. In this framework, “realistic threats” include those that pose potential harm to the group’s physical health, safety, and very existence. “Symbolic threats” include those that disrupt expected social behaviors and group norms, and could lead to realistic threats. Important components of the ITT include intragroup anxiety and negative stereotypes, which foster both realistic and symbolic threats. This framework has previously been applied to understanding prejudice between groups [intergroups] – such as discrimination by a dominant group against minorities (Stephans & Stephans, 2000). In the present study, we observed when overdose response techniques like shaking the victim, walking them around, inflicting pain, or injecting them with salt water (Pollini et al, 2006) failed, 911 was the only option for rescue. However, contacting 911 led to realistic threats emanating from police onto PWUDs (i.e. intergroup dynamics) and produced repercussions within PWUDs social groups (i.e. intragroup dynamics). Dovidio (2013) highlights the importance of linking the connection between intergroup threats and their impact on intragroup relations. In the present study, we aimed to expand on previous findings by Wagner et al. (2014) regarding cutting social ties following an overdose by uncovering how frequently overdosing might result in similar social distancing due to the threats they pose to other PWUDs in their social networks on account of possible law enforcement involvement summoned for lifesaving. We utilized the ITT to conceptualize this phenomenon.

METHODS

Setting

The data used for this study were collected for a larger mixed methods study designed to evaluate the effectiveness of a law enforcement naloxone program, in which Sheriff’s Deputies in San Diego County were trained in overdose prevention and equipped with naloxone to administer in the event they arrived at the scene of an overdose prior to Emergency Medical Services [EMS]. The larger study involved analysis of administrative data to determine the impact of the program on rates of calls to 911 for opioid overdose, comparing areas where law enforcement officers were equipped with naloxone to areas where they were not. In addition to collecting administrative data on 911 calls and opioid overdose mortality, we conducted qualitative interviews with PWUDs to examine their experiences witnessing opioid overdoses and their decision-making regarding calling 911, and with law enforcement officers about their experiences responding to opioid overdoses. The present study involves secondary analysis of the qualitative interviews with PWUDs.

Of note, at the time of interview, a take-home naloxone program had just started in San Diego through a local community health organization that operated two times per week from a mobile site, and also offered one-for-one syringe exchange. It has since expanded to three times per week (Family Health Centers San Diego, 2020).

Recruitment and Eligibility

Data were collected between February 2017 and May 2018 by author 4, and three trained research assistants with qualitative data collection experience. Recruitment involved street-based engagement, chain referrals, and flyers with contact information posted in strategic venues. A screener was used to assess eligibility that included: 1) being over the age of 18; 2) having witnessed an overdose in the past year; 3) having used an illicit substance (other than cannabis) in the past three months; and 4) the witnessed overdose occurred within the jurisdiction of the Sheriff’s Department. In total, fifty-two participants were screened eligible and subsequently interviewed. The study protocol was approved by the University of California San Diego’s Institutional Review Board [IRB] (protocol #151797).

Data Collection & Preliminary Analysis

Interviews were conducted in a variety of semi-private and private locations including coffee shops, fast food restaurants, and participant’s homes. Each participant was given an informed consent document and the interviewer reviewed the study purpose, procedures, and the participants’ rights as research subjects, which included that they could stop the study at any time, did not have to answer questions they were uncomfortable answering, and information on how to contact the study’s principal investigators and IRB. UCSD IRB granted a waiver of documentation of informed consent because the consent document was the only piece of documentation that would link participants to the research study; therefore, participants were not asked to sign the consent form. The interviewers were trained and experienced qualitative investigators who used a semi-structured interview guide to facilitate enquiry. The study instrument contained open-ended and closed-ended questions designed to capture a small set of demographic indicators, recent experience witnessing an overdose, what occurred during the event, if others were present, and how persons responded, particularly as it relates to contacting 911 rescue services. The interview protocol allowed for follow-up probes on emergent and pertinent topics. Our iterative research approach allowed for emergent themes to be explored in further depth following preliminary analysis of initial transcripts. In the present study, the social consequences associated with frequently overdosing arose unprompted, and the study team proposed the interviewers further probe this phenomenon. The interviewers included probes such as, “So if a friend or people you know overdoses… this person, what happens to them after?”

Analysis

Interviews were digitally recorded and files were uploaded into a secure database for transcription by a professional transcription service. Verbatim transcripts were subsequently uploaded into Atlas.ti and coded by the first author. Open coding was conducted to organize data into broad topics including “perceived cause of OD,” “social consequence,” and “realistic threat.” Memos were written expanding on these codes and emergent themes and on the relationships between them. As part of the analytic process, we applied the Integrated Threat Theory to help conceptually arrange our data.

RESULTS

Participant Demographics

Of the 52 participants, 43 (83%) identified as male and 9 identified as female (17%). Of the men, the majority were non-Hispanic white (n=25; 48%), followed by Latino [8; 15.3%], Mixed-race [5; 9.6%], Black [3; 5.7%], American Indian [1; 1.9%], and Asian [1; 1.9%]. Women were also majority white [7; 77.8%], followed by Latina [2; 22.2%]. The mean age of participants was 42.1 years [range 22 to 66], however only one participant was under 30. Participants were all opioid users, and many reported using additional substances such as methamphetamines, cocaine, and benzodiazepines. Although not explicitly asked to each participant in the qualitative interview, many participants reported that they initiated drug use during adolescence and that they experienced fluctuating periods of abstinence over their drug using careers. For example, when asked about drug use one participant summed up, “On and off? My whole life” (Interview 2). Participants reported witnessing between 1 and 50 overdoses in their lifetime, and nearly all overdose events discussed in detail occurred within the last two months. Fewer than 10 participants mentioned ever possessing naloxone.

Perceived Causes of Overdose Among PWUDs

Participants were interviewed about their experience as witnesses to opioid overdoses, including their perceptions of why the overdose had occurred. Perceived causes fell into two groups: factors within the control of the person using drugs, (i.e. behavior patterns), and factors outside of their control, (i.e. macro or structural issues). As an example of the latter, participants described the impact of contaminants in heroin products, especially fentanyl (a high potency synthetic opioid), as a cause of overdose. As one participant stated,

I’ve lived all over the country, so it’s real risky, because you never know what you’re getting. Especially back east with all the fentanyl and stuff that’s out here too, but that’s the real scary stuff and I guess now the fentanyl from what I heard, some of it is Narcan [naloxone, an opioid antagonist] resistant, like the Narcan won’t work to bring you out of the fentanyl overdose (Interview 25).

Another participant suggested a combination of factors leading to overdose, while also blaming the individual who overdosed:

So, in the heroin scene … overdosing is viewed as a social faux pas, like it’s your fault [emphasis added]. Why would you do that, you know how to not do that, you know what I mean? And for most--and to be honest for the most part that’s the truth. You know you don’t drink alcohol with it, you know if you’ve been clean for any amount of time you’re not going to do the amount that you used to do. There’s certain things that every addict knows but then there’s also--you have, now you have stuff that’s got with fentanyl, so you just don’t know I guess really anymore, but people get angry at other people for overdosing especially if someone overdoses in their house [emphasis added] (Interview 19).

This narrative suggests that the participant believes that PWUDs are aware of drug consumption practices that elevate risk for overdose and should therefore behave in ways that reduce risk; failing to do so is therefore considered a “social faux pas.” Discussion of drug consumption behaviors and practices often included individualized blame, in which the person who overdosed was accused of consuming drugs irresponsibly or overindulgently leading to the overdose event. To illustrate, one participant stated, “you overdose because you’re being stupid or reckless (Interview 21),” and another mentioned, “Okay well basically I guess the guy, he just did too much, he’s always been known to (Interview 15).” Many participants discussed the risk of polydrug use,

Same thing with people who take those xany bars [Xanax, a benzodiazepine] or whatever; it’s not if, it’s when they’re going to go out. For some reason, it just doesn’t mix. My friend who I saved this last time, he’s an alcoholic, every day on the streets, he drinks. He’s overdosed like I said already, like three more times since then. So, he’s going to die soon, I told him. It’s sad, but he’s going to die, not if, it’s when (Interview 23).

Another participant agreed that polydrug use heightens overdose risk, highlighting the role of alcohol in exacerbating that risk,

A friend of mine is not only an opiate addict but he’s also an alcoholic and I don’t think a lot of people understand that if you mix alcohol with opiates--you’re probably going to overdose. It’s a very strong chance that you will and that’s his thing. This is probably the fifth or sixth time this guy has--and I’ve at least heard of that he’s overdosed. This is in public and in the middle of the day on the steps of the Buddhist temple that he’s just out and turned blue (Interview 19).

Persons known for consuming drugs in excess or in a manner known for causing overdoses, like polydrug use or drug injection, i.e., “The line is I wouldn’t be close with anybody using a needle anyways,” were branded as problematic within their networks due to perceived reckless drug consumption. Some participants stated that since drug use isn’t a compulsory activity, the onus of overdose is solely on the person who overdosed,

I guess you can go and blame the dealers, but who’s really responsible? The person that is putting it in their arm or whatever, they’re not forcing it on us. It is always sad, but then again, it’s like, when it goes that far, it’s just our fucked-up way of thinking and that’s like we want to get that dope, we know it’s that potent and we seek that out or whatever. It’s crazy (Interview 25).

The same participant assumed that they haven’t overdosed before because of their responsible practices, “I’ve never OD’d myself as far as I know. I’ve always been really careful,” suggesting that those who do overdose are careless. Another stated, “I’m not the greedy one who overdosed (Interview 23),” further suggesting that overindulgent drug consumption led to overdoses. The narratives of individualized blame suggested overdose was controllable and therefore careless drug consumption practices were perceived to be inconsiderately threatening the network through potential social consequences.

Intergroup Relations and Realistic Threats

The previous section described how respondents attributed blame and understood causal mechanisms of opioid overdoses they had witnessed. In this section, we describe how witnessing overdose events posed realistic threats for participants and contributed to onerous consequences. One participant summarized the overarching sentiment that was expressed by many: “that’s not good at all if you overdose, it’s all bad. Paramedics got to come, yeah, it’s a big scene (Interview 23),” and that participants “try not to call [911] because most of the time they’re going to have warrants. You really don’t want to go to the cops if you don’t have to (Interview 27).” This is acknowledged by participants due to the criminalization of drugs, e.g., “I was carrying on illegal activities there (overdose scene). No way… Under no conditions would I have called the cops” (Interview 2). Being witness to an overdose placed participants at risk of threats such as arrest and eviction on account of police presence, which often resulted in blame and anger towards the person who overdosed. To illustrate, a participant said,

You got to control your shit, you got to know how much you can do because you’re putting everyone in jeopardy, you know what I mean? You’re not let alone yourself but you’re putting our homes in jeopardy, our safety, our freedom. If you fall out, it’s usually your fault. You don’t overdose on accident (Interview 21).

This sentiment of protecting networks from police was furthered by the same participant, “You’re going to get people arrested [for overdosing], you might even get people killed. So, you definitely get people arrested. Get people evicted, and I’ve seen it happen (Interview 21).” One participant feared arrest in the aftermath of an overdose because “I feel fear for my life in a jail (Interview 32).”

Fear was an important underlying theme fueling anxiety and anger among participants. In one narrative, a participant elaborated on the emotional toll of overdose response,

It was a friend of mine and he had used with me a few times before but he wasn’t a normal user. And the day before, I’d given him some and actually administered, gave him the shot. And he said he didn’t barely felt it so it was the next day and I gave him about what I would use and he was behind me sitting on the bed and I shot him up and then I was making my shot and I asked him how he felt and he didn’t answer and I turned around and he was blue at that point and not breathing. And he fell off the bed and then I shook him around for a minute and then I immediately called 911 pretty much probably 30 seconds or so. And I called 911 and I was talking to the dispatch lady and she basically sent out the paramedics and then told me how to do CPR over the phone and I held the phone under--it was a regular house phone and I held it under my ear and did CPR and pumped his chest and hid the dope and all the same amount of time and it was extremely scary (emphasis added) especially being the one that administered (the opioid) … it was a good friend.

In this narrative, the commitment to lifesaving is confirmed. At the same time, it highlights the reality of feeling scared in such situations, especially when the participant was responsible for both saving a close friend’s life and hiding drugs that posed threats to their overall safety if discovered by law enforcement. Having to balance lifesaving and reducing the realistic threats by hiding drugs was noted as “extremely scary” and contributed to a stressful scenario.

The fear of arrest and/or other negative experiences with law enforcement was pervasive, yet often balanced by a desire to intervene to save a life. One participant stated that when police and paramedics showed up, police “basically told me I’m a scumbag and basically just yelled at me for a half an hour and then let me go,” and that it was “a pretty traumatic experience, really. But he lived, so that was a good thing.” This participant was belittled, ridiculed, and traumatized by police; but was grateful for the survival of his friend and having not been arrested. Another participant mentioned that he will only intervene in an overdose if he is confident it won’t result in arrest, “I would have most likely [intervened] depending unless I’m looking at being arrested (Interview 2).” Arrest was discussed by many participants, with one adding that overdose response is complicated by the overdose victim’s decision to consume drugs in the first place, which speaks to the intertwining of blame and realistic threats like arrest,

I definitely understand why people wouldn’t [call 911], because they’re scared, they think that they’re going to get in trouble, and I’ve been in jail for just possession and that’s all it would be. It’s not like you’re forcing him [the person who overdosed] to [use] drugs and then they overdose. It’s not like that, so I mean I can see why people are scared. No one wants to get arrested (Interview 11).

This participant’s narrative suggests that placing the burden of overdose response on one’s network of peers is troublesome; however, she followed up that, “Hopefully, I’m never in that situation ever again, but naloxone is a lot easier to come by now, so I might not even have to call EMTs sometimes. I’ve never had to do that, but I know that naloxone is a lot easier to get (Interview 11),” suggesting a possible reduced burden via naloxone availability. Another participant agreed, “But naloxone works pretty good and if it doesn’t there’s no way you could’ve saved them anyways (Interview 21).” One participant acknowledged the importance of lifesaving within the community,

Yeah. In general, you try to save (a person overdosing) --there’s a credo among heroin addicts. They try to save their own. They will try and save their own and go to extreme measures to do it. Get that guy in a shower, keep slapping him, cold water only. If worse comes to worse and they keep going out, you get ice… So, they will try to save the person because it could be them next. That’s the credo among IV users (Interview 19).

However, when rousing techniques like cold water and slapping were ineffective in reversing an overdose, overdose response was complicated by the credo of lifesaving and the need to contact 911 for lifesaving services due to accompanying realistic threats. Feeling forced into a position in which 911 needed to be called was frustrating for participants and led to intragroup anxiety. Overall, structural consequences such as the criminalization of drugs led to bystander vulnerability when witnessing an overdose, resulting in potential outcomes such as arrest, incarceration, and housing evictions. As a result, the person who overdosed experienced damaged social standing due to overdosing being the basis of the actualization of these realistic threats.

Intragroup Response to Overdose & Realistic Threats

Narratives of witnessed overdose events suggested that several mechanisms damage social standing within one’s drug use network, resulting in intragroup consequences such as social distancing or rejection. In this section, we discuss the various forms of distancing and damage to social relationships that occur within the group as a result of realistic threats and intragroup anxiety. When overdosing resulted in police presence, it represented the symbolic threat of breaking social expectations to keep a low profile and not draw unwanted attention. As such, overdosing was viewed as breaking an important unwritten commitment to clandestineness. To illustrate, a participant stated,

Yeah, because saving life’s important but what’s almost as important in this community is your reputation for solidarity or for keeping your mouth shut. It’s important because otherwise you don’t get served [i.e. cutoff from purchasing drugs or sharing drug doses among peers] and if you don’t get served then you can get sick and it’s sicker than any sickness anyone could ever experience is their life, I promise you that … But it’s enough to make any human to do whatever they have to (Interview 21).

This participant’s narrative indicates that while the community values the “credo” of lifesaving, overdose will simultaneously damage one’s reputation due to breaking the commitment of protecting the network from law enforcement, which might result in being barred from network participation. Critically, in this respondent’s narrative we see the cycle of how a symbolic threat (i.e. exposing the group to police) leads to intragroup anxiety of experiencing forced withdrawal if incarcerated, the onus of which is placed on the person who overdosed. He expressed that the violation of this important social norm could disrupt one’s ability to participate in drug sharing within the network, which in turn would result in the experience of further opioid withdrawal symptoms. This cycle could ultimately lead to elevation of risk for overdose death, since opioid use after even short periods of abstinence can elevate risk for fatal overdose.

In the aftermath of an overdose event, frustration, resulting from fear of realistic threats, was directed towards the person who overdosed because of the symbolic threat of social consequences. For example, “I was like, ‘You motherfucker,’ [I was] ready to knock him back out. I was just like, ‘You’re up, now I’m going to knock you back out;’” and “I took a picture so I could show him later about how stupid he was and about this is why you don’t do this shit at my house, or just in general (Interview 32).” Another participant expressed anger, “This friend I’ve known since childhood - and last time he overdosed at my house he told me that he knew that the shot was too big … I was pissed (Interview 33).” Field notes from these interviews highlighted that these comments were made with energy and irritation. Following from this irritation, social distancing from persons who use overdosed frequently seemed the only option for group protection. As such, participants who overdosed frequently were labeled as liabilities:

“Once? Okay. But if you’re known for doing that [overdosing], you generally get shunned. Generally, get eighty-sixed as you call it because you’re a liability (Interview 21).”

The rate of what was considered frequent overdosing was not specifically quantified by participants; rather, it was noted that if someone often behaved in a manner that could lead to overdose, they were labelled as putting the group at risk. As such the threshold for tolerance of this behavior was based on a subjective assessment of how much risk it posed to the group. Another participant who said they had witnessed “countless overdoses” was asked about persons who overdose often and stated, “yeah you don’t want them around.” They went on to say,

Yeah, it’s just the fact they’re a liability… Even somebody that makes it, it’s just traumatic to be around that [an overdose]. And it brings a lot of heat onto the people that are around him, and if you give him the drugs, you could go in for manslaughter or something [emphasis added]. You shun that person, excommunicate him. And that’s the way it is. And it’s just for safety [emphasis added] … A lot of times you’re like, just go away, or you even leave … you just leave, and you don’t go back there, or don’t hang out with those people anymore, because it’s just a liability. Because I don’t do any illegal activities besides doing drugs, and it’s the fact is going to jail and shit, I can’t handle that, and I put my freedom over somebody (Interview 35).

Of note here is that the act of labeling someone as a liability is related to the intragroup anxiety of severe criminal justice sanctions – up to and including manslaughter charges – for individuals involved in the scene. These realistic threats, coupled with others discussed earlier such as loss of housing or eviction, led some participants to ban individuals known to overdose frequently from their social interactions, since this was the only means of group preservation available to them:

And she just yeah … she had the attitude of a younger addict where it’s like yeah, it’s not funny [overdosing], it’s my house, it’s really not funny. If you do that again, you can’t come here again. It’s common knowledge amongst addicts that if you fall out, we call it, you can’t come back (Interview 21).

Another participant offered a similar narrative, demonstrating how someone with a reputation as an overindulgent drug user at risk for overdose created a liability for the group,

And I was like honey I really don’t think it’s a good idea he comes over here no more because it’s not really working out. He gets so way over … And I know that since he lost a sibling to that, it would just kill me to have his mom have to hear of it, yeah. But he’s always been very extreme like that. He’s very extreme, really heavy user (Interview 15).

This narrative succinctly ties the reputation of being an “extreme” drug user, to the decision to socially distance from this person. Social distancing was confirmed by other participants, even if only temporarily, “Yeah, if it’s a big situation where I have to call 9-1-1 I try not to get high with them [the person who overdosed] again for a while,” but the same participant also acknowledged times where the break is permanent, “I decided to never get high with him again after that [the overdose] … too big of a risk (Interview 33).” Others confirmed the permanent distancing, e.g. “Just like, man, you put me through this shit. I don’t even want nothing to do with you (Interview 34).” Of note, this phenomenon was mostly discussed among participants untrained by take-home naloxone programs. One participant who wasn’t aware of take-home naloxone programs wished for one, “However, I think in first aid kits … they should have ones with Narcan and I think that’s something that should be available for people who are afraid to call 911 … because that’s scary [emphasis added]. (Interview 18)” For this participant, desire for take-home naloxone was rooted in entrenched fear of the 911 system.

In sum, frequently overdosing poses realistic threats for overdose witnesses that are fostered by symbolic threats and lead to intragroup anxiety. Because PWUDs have no control over intergroup threats posed by law enforcement, the only means of group protection and preservation is for the group to distance itself from persons who frequently overdose.

DISCUSSION

The intergroup dynamic between law enforcement and people who use drugs [PWUDs] is precarious for PWUDs because law enforcement personnel are positioned to impose serious consequences onto PWUDs in the aftermath of witnessed overdoses. In our study, we applied Integrated Threat Theory to the social repercussions surrounding witnessed overdose events, which revealed that when 911 was called during overdose emergencies, PWUDs feared the threat of serious consequences resulting from police presence and power. Intervening in a potentially fatal overdose by calling 911 was complicated by realistic threats to PWUDs (e.g., short and long term incarceration, housing eviction, and manslaughter) that stemmed from intergroup relations. These consequences then fueled an intragroup reaction among PWUDs in which the group felt anxiety about witnessing future overdoses and subsequently created distance from persons who frequently overdosed. Furthermore, because PWUDs faced such serious consequences from contacting 911, persons who overdosed frequently were labeled liabilities [i.e. threats] to the safety and freedom of the group. As a result, in some instances social ties between the group and the person who overdosed were severed. This could be especially perilous for the ostracized person as previously overdosing is a risk factor for future overdose and consuming drugs alone is a risk for fatally overdosing as there are no witnesses present who can intervene (Caudarella, Dong, Milloy, Kerr, Wood, & Hayashi, 2016; Davidson, 2003; Stoove, Dietze, & Jolley, 2009).

Individualized blame imposed onto a person who overdosed frequently led to a reputation as an overindulgent drug user spurring symbolic threats, such as exposure, within the intragroup. This contrasted with the importance of having a reputation for upholding the norm of surreptitiousness within networks of PWUDs. Subjective norms, such as the commitment to clandestineness, play an important role in group identity, which is threatened by external realistic threats (Dovidio, 2013). Bonding manifests from respecting and supporting group identity [as PWUDs hidden from law enforcement] and respecting groups norms [like maintaining surreptitiousness]. Participants faced an ethical dilemma in the presence of an overdose event; they felt an obligation to contact 911 to save their peer, but this was complicated by the pact of secrecy among PWUDs, violation of which could result in the actualization of realistic threats onto the group. In our study these threats - often the result of policy directives - shaped intragroup dynamics. Other recent research has similarly highlighted how policy directives and community overdose response impact and shape intragroup dynamics. For example, Farrugia et al. (2019) discussed conflict that can arise in the aftermath of overdose events, particularly as it relates to withdrawal caused by naloxone. Kolla & Strike (2019) uncovered the complexity faced by PWUDs working in satellite harm reduction sites providing overdose prevention guidance and drug use observation to peers that was hampered by structural factors such as threat of eviction, lack of supervised consumption services, and criminalization. In these studies, as in the present study, drug use is conceptualized as operating within a policy and social risk environment (Rhodes, 2002) that drives drug use harms and shapes intragroup dynamics.

Bourgois and Schonberg (2011) highlighted the intragroup dynamics of PWUDs by describing the “moral economy” in which PWUDs share resources and drugs. These give-and-take interactions are a means of survival and off-putting withdrawal symptoms. This phenomenon was noted in the current study by a participant who also highlighted that symbolic threats would effectively cut one off from the “moral economy,” which is one form of social distancing. Beyond the basic interactions of resource and drug sharing, the bonds of love, support, and friendship are also severed when social ties are cut, which is likely painful and possibly traumatic for all involved (Syvertsen et al. 2015; Simons & Singer, 2006). It could also interrupt relationships, or result in the ostracizing of drug using couples as their bond and reliance on one another are strong (Syversten et al., 2014).

Reluctance to contact 911 among PWUDs is well-documented (Latimore & Bergstein, 2017; Tobin, Davey, & Latkin, 2005). Fear of incarceration, distrust of police, prior police exposure and mistreatment, and forced withdrawal in jail are known barriers to calling 911 during an overdose emergency. In spite of ample evidence, these issues persist. Our participants commonly mentioned experiences of incarceration and losing their housing because of police involvement during overdose events. Latimore and Bergstein (2017) and Wagner et al., (2019) add fear of losing child custody, encountering stigma, and facing violence from drug sellers to the list of possible repercussions onto PWUDs for contacting 911 during an overdose event. Such palpable, precarious threats to safety and wellbeing have created an environment in which contacting 911 during an overdose is exceptionally hazardous for PWUDs. Our findings add to the current knowledge base by revealing the social processes that underlie barriers to calling 911 and subsequent social implications. Having a reputation for regularly engaging in overdose-causing behaviors led to being labelled a liability as some overdoses require 911 for rescue. As noted throughout the present study, 911 directives resulting in police presence during an overdose created fear-inducing, serious threats; however, they could possibly be mitigated through program and policy changes that might in turn reduce social distancing responses and solitary drug use.

Public Health Implications

Severing social ties following an overdose event was first introduced by Wagner et al. (2014) who reported that participants distanced themselves from persons who frequently overdosed in their presence as a means of coping with the emotional stress accompanying overdose events. Participants in their study reported attempts to limit their association with “particularly risky network members” [161]. The need to support community members in coping with the stress and trauma associated with frequently responding to overdoses has also been described by researchers in Vancouver, Canada, where the burden of opioid overdose is particularly severe (Shearer, Fleming, Fowler, Boyd, & McNeil, 2018). The current work adds to this growing body of literature by demonstrating that the emotional burden might also be fear-based, due of the risk of social and legal consequences that might occur from the need to contact 911. Wagner et al.’s (2014) previous findings are important for this study as participants in their study were all trained by take-home naloxone programs, and they uncovered many associated positive effects including enhanced confidence responding to an overdose, greater sense of control, reductions in risky drug use, and feelings of heroism and pride. They suggested that these positive effects might mitigate some of the negative effects of witnessing an overdose. Most participants in the present study were not familiar with take-home naloxone programs, were untrained, or weren’t aware of where or how to access naloxone; but did clearly state a desire to respond to overdoses and uphold the “credo” to save lives within their network. To that end, expanding take-home naloxone programs, supporting peer responders in dealing with the trauma of witnessing overdoses, and discussing the causes of opioid overdoses with PWUDs might be an important approach in lessening ostracism and cutting social ties with persons who frequently overdose.

While PWUDs have repeatedly demonstrated interest in lifesaving, and expressed desire for an emergency overdose rescue system devoid of police dispatch altogether (Wagner et al, 2019; Wagner et al, 2014; Wheeler, Jones, Gilbert, & Davidson, 2014), improving the protections offered by Good Samaritan Laws could work to mitigate the realistic threats imposed by police when 911 is contacted for overdose rescue. 911 Good Samaritan Laws are designed to provide prosecutorial immunity for persons acting in good faith towards a person in a precarious situation, like an opioid overdose (California Legislative Information, 2012). California Legislature enacted a Good Samaritan Law in January, 2013, under Health and Safety Code Section 11376.5. The Good Samaritan Law exempts prosecution for possession of minor drug and paraphernalia possession when 911 is called to an overdose scene; however, there are important limitations to the protections. First, police might still ask witnesses for identification and search for existing warrants, which is an emerging deterrent to the success of these laws (Latimore & Bergstein, 2017; Wagner et al., 2019). Second, witnesses might still face arrest and spend some time incarcerated before a judge will dismiss the case because of immunity protections. Third, Code Section 11376.5 is clear that this law does not provide immunity to those who are currently on probation or parole, and presence at the scene might result in probation or parole violations, even if the person was engaged in lifesaving activities. Furthermore, those who might have sold, provided, given, or exchanged the drug complicit in the overdose are not protected and might be charged with manslaughter-related charges (California Legislative Information, 2012). As such, witnesses risk facing prosecution if the drugs were being shared, as they often are within drug using networks (Bourgois & Schonberg, 2010; Koester, Glanz & Barón, 2005). Recently, overdose witnesses across the U.S. are increasingly facing homicide-associated charges for complicity in overdose fatalities, a phenomenon that could directly negate the impact of Good Samaritan laws and serve as a deterrent to contacting 911 during an overdose emergency (Drug Policy Alliance, 2017; Goldensohn, 2018). This overdose-related manslaughter policy directive and Good Samaritan Law clash is an area worthy of further investigation.

Limitations

This study is not without limitations. Our probing on the topic of cutting social ties in the aftermath of overdoses did not start until we identified it as an emergent topic after approximately 17 interviews. As such, some narratives might not have been probed to the extent needed to uncover this phenomenon among all participants in the larger study, thereby leading to underreporting. Similar to other research about sensitive topics, considerations of social desirability bias should be acknowledged; respondents might have minimized reports of what they perceived to be undesirable behaviors or attitudes. However, given the high amount of saturation on the topic that we achieved in the current analysis, we judge these threats to be minimal. Because this was a small sample in a single setting, generalizability is limited, especially as policies and the implementation of Good Samaritan Laws differ locale to locale. While fear of incarceration, eviction, or loss of child custody is known (Tobin, Davey, & Latkin, 2005) it is unknown if social distancing is a response outside of Southern California, creating a call for future research to have a better grasp of the extent of this phenomenon. Nonetheless, small and localized samples are generally considered a hallmark of qualitative research, which confers advantages such as the ability to iteratively respond to emergent concepts as the research evolves, and the ability to provide thick narrative description that yields insight into processes, meanings, and social dynamics (Sandelowski 2003). In addition, the application of Integrated Threat Theory bolsters the strength of our conclusions and sets the stage for future research, which suggests transferability to other contexts.

CONCLUSION

Responding to potentially fatal overdoses is complicated for PWUDs. Witnessing potentially fatal overdoses can be emotionally harrowing, which is compounded when overdose response requires calling 911, threatening burdensome and serious social consequences for witnesses. In the current study, we observed that these social consequences, coupled with individualized blame for overdose occurrence, resulted in persons who frequently overdose being labeled as threats to the safety of the whole social network. As such, members of the network might create social distance from or fully reject persons who overdose frequently as a means of preserving group safety. Ostracized persons could then consume drugs alone, which places them at higher risk of fatal overdose (Davidson, 2003). The available data demonstrate that this phenomenon is occurring due to forces beyond the control of PWUDs, such as lack of access to stable potency heroin products and increasingly hazardous social consequences associated with law enforcement responses to overdose events. The intragroup response of social distancing was a means of asserting the only control possible to mitigate the potential threats to the network and preserve group safety. Structural reforms to address these threats could include: providing additional support and coping resources for PWUDs who respond to overdoses, expanding take-home naloxone programs, and improving protections offered by extant Good Samaritan Laws.

Highlights of “Generally, you get 86’ed because you’re a liability”: An application of Integrated Threat Theory to frequently witnessed overdoses and social distancing responses.

• Opioid overdoses attended by 911 for rescue often result in police dispatch

• The tense intergroup dynamic between police and drug users poses realistic threats

• These burdensome threats led to intragroup anxiety among networks of drug users

• Reducing such threats was curtailed by cutting social ties with frequent overdosers

• Perilous intergroup dynamics could be addressed through improved policy directives