Contrasting perspectives and applications of the Good Samaritan Overdose Prevention Law: insights from people who use and distribute drugs in Rhode Island

Abstract

Background

The overdose crisis in the U.S is complex, with opioid overdoses compounded more recently with rising stimulant-involvement. Individuals who experience or witness overdoses often avoid calling 911 due to fear of criminal-legal system consequences. Many U.S. states have “Good Samaritan Laws” (GSLs) protecting those seeking help for overdoses, but research is lacking on the understanding and role of GSL in overdose response among people who use stimulants (PWUS) and criminal-legal populations. Our mixed-methods analysis focused on PWUS from a community-based sample in a state with GSL protections, Rhode Island, and participants incarcerated at the Rhode Island Department of Corrections (RIDOC) for charges related to drug distribution. We analyzed survey and interview data to discern differences in knowledge and attitudes regarding Rhode Island’s GSL in the two samples.

Results

Overall awareness of the GSL was widespread among the community sample (61.1%) and the RIDOC sample (63.3%); however, more RIDOC participants (70%) than PWUS in the community (42.9%) could accurately describe the law. Participants in both samples reported deeply rooted fears associated with the consequences of help-seeking and skepticism of protections provided by the GSL, particularly around criminal-legal consequences. While the Rhode Island GSL provides protections for probation or parole violations, many RIDOC participants were fearful that these protections would not apply to them, despite the law being amended to provide these protections. Additionally, participants from both samples reflected on differences in help-seeking behaviors among unexpected opioid and stimulant overdoses, discussing how their interpretation of overdose symptoms influenced decisions to call for help (e.g., call 911) or not.

Conclusions

Varying degrees of understanding of the Rhode Island GSL, along with fears of police involvement in overdose situations, underscore the need for targeted interventions, policy change, and re-education efforts about overdose prevention and the GSL. Findings can inform policy and professional practices associated with GSLs.

Introduction

The current opioid crisis, exacerbated by a toxic and unregulated drug supply, has underscored the need for effective overdose response measures. Most problematic within the volatile drug supply is fentanyl, which, when found unexpectedly in the stimulant supply (e.g., crack and powdered cocaine), can increase overdose fatalities (Hughto et al., 2022; Spencer et al., 2022). People who primarily use stimulants may lack opioid tolerance and may be less familiar with opioid overdose prevention (Hughto et al., 2022; Kelly et al., 2023). Relatedly, fatalities involving stimulants only, excluding fentanyl, are also on the rise, highlighting the need to better understand and respond to stimulant-involved overdose (CDC, 2025; Hoots et al., 2020; Hughto et al., 2025). Moreover, fear of legal repercussions can deter individuals from seeking medical attention during overdoses (Latimore & Bergstein, 2017; Wagner et al., 2021), contributing to approximately 107,941 overdose deaths in the U.S. in 2022 (Spencer et al., 2022).

In response to overdose surges, ‘Good Samaritan Laws’ (GSLs) for overdose intervention of fatal overdose prevention were introduced to provide selective legal immunity to individuals seeking medical help for themselves or others during drug-related emergencies (Green & Nace, 2024; Latimore & Bergstein, 2017; Moallef et al., 2021; Reader et al., 2022). First enacted in New Mexico in 2007, GSLs now exist in the majority of U.S. states (Lieberman & Davis, 2023). GSLs have evolved over time to be responsive to concerns among people who are at risk of overdose, incorporating provisions aimed at enhancing their effectiveness (Reader et al., 2022) and limiting the possible repercussions of criminal-legal system exposure (Popp, 2022). While some studies have found that GSLs increase emergency service calls (Jakubowski et al., 2018), others have identified persistent barriers to help-seeking and inconsistent application by criminal-legal system actors (Latimore & Bergstein, 2017; McLean, 2018; Pollini et al., 2006; Wagner et al., 2021). GSLs typically include caveats that allow police to arrest those calling for help or experiencing overdose if there are outstanding warrants, probation or parole violations, weapons possession, or if they are e suspected of drug distribution (Koester et al., 2017; Latimore & Bergstein, 2017; Wagner et al., 2019). Conversely, research has shown that GSLs providing specific protections against arrest are more effective than those only protecting from prosecution or simple possession charges (Hamilton et al., 2021; Reader et al., 2022; van der Meulen & Chu, 2022).

Research exploring perspectives of people who use drugs (PWUD) regarding GSLs offers valuable insights into the impact and efficacy of these laws, and have primarily focused on the impact of GSLs on reducing opioid overdose mortality (Hamilton et al., 2021; McLean, 2018). Several studies highlight varying degrees of understanding of GSLs and the protections they provide (Jakubowski et al., 2018; Schneider et al., 2020); however, less is known about the impact of GSLs on 911 help-seeking among people who use stimulants (PWUS). Most studies focused on GSLs awareness among PWUS suggest this population may not possess adequate knowledge of GSLs and may be less likely to call 911 following an overdose (Ambrose et al., 2016), relative to people who use opioids who may receive training on GSLs through their exposure to harm reduction programming (Evans et al., 2016). Such research is particularly relevant given the rise in stimulant overdose fatalities not involving opioids (Kariisa et al., 2021; Spencer et al., 2022). Similar to opioid overdoses, avoiding calling 911 may be motivated by fear of criminal-legal system consequences driven by disproportionate enforcement on racial and ethnic minorities (Kariisa et al., 2021; Rouhani et al., 2021). However, stimulant overdoses have a different clinical presentation and may advance to life-threatening conditions that necessitate medical attention. Therefore, there is a need to examine best practices related to help-seeking behaviors when these events occur (Hughto et al., 2025).

Additionally, research is needed to explore the views of people who distribute drugs (PWDD), including those incarcerated for charges related to drug distribution or manufacturing, as this population may be likely to encounter someone experiencing overdose and are positioned to intervene. Moreover, PWDD may also partake in drug use themselves, creating further challenges regarding help-seeking due to fear of criminal-legal involvement. PWDD and those incarcerated for drug-related crimes may also have a better understanding the limits of GSL protections due to proximity to the criminal-legal system. Arresting PWDD could result in drug market disruptions that further perpetuate overdose risk (del Pozo et al., 2024; Ray et al., 2023). Some states, like Rhode Island, have drug-induced homicide laws (Peterson et al., 2019) that can be enforced when a drug is supplied to a person who has overdosed, which may dissuade PWDD from calling 911 due to fear of arrest or incarceration (Beletsky, 2019; Carroll et al., 2023). Understanding these perspectives is crucial for tailoring interventions and policies that effectively address the complex dynamics of drug-related harm and overdose prevention (Bowles and Lankenau, 2019).

Rhode Island is a New England state heavily affected by the opioid crisis, an early adopter of the Good Samaritan Law (GSL), and has a long-standing program of overdose education and widespread naloxone provision, even in correctional settings (Martin et al., 2025; The Good Samaritan Overdose Prevention Act of 2016 R.I. Gen. Laws § 21-28.9-4, 2016, Green et al., 2015). Our study examined and contrasted the perspectives of PWUS and incarcerated PWDD in Rhode Island to explore how knowledge and awareness of the GSL differs between these groups. In comparing these groups, we focus on how the GSL is understood and interpreted to discern meaningful differences in how the GSL is conceptualized in the state. Further, because this study focused on both opioid and stimulant overdose events, our analysis explores whether there were meaningful differences in overdose responses from community members who use drugs versus those incarcerated for drug distribution at the Rhode Island Department of Corrections (RIDOC). Findings from this study can strengthen messaging about GSLs for both PWUS and PWDD, two critical stakeholder groups that may hold different concerns and perspectives towards the law.

Methods

This analysis draws from a larger mixed methods study of people who use or distribute drugs in two New England states, Massachusetts and Rhode Island, to identify risk and protective factors for stimulant and opioid-involved overdose (Hughto et al., 2024). All participants were surveyed, of which a purposive sub-sample completed an in-depth interview. We recruited (1) PWUS —cocaine (powdered or crack), methamphetamine, or street-obtained prescription stimulants—and (2) PWDD sentenced and incarcerated at RIDOC for charges related to drug distribution and/or manufacturing. Comparisons between these groups should be interpreted as reflecting differences in the current context (community vs. incarceration), rather than differences in past exposure to the criminal-legal system. Participants provided verbal consent to participate and were compensated $20 for completing a survey, and an additional $20 for completing an interview. The present analysis includes only participants from Greater Providence Rhode Island and RIDOC to compare community and incarcerated samples regarding knowledge and experience with Rhode Island’s GSL. Our mixed-methods approach used qualitative findings as the primary lens through which we interpreted and contextualized our quantitative results.

Community-based sample

Sampling procedures in the parent study involved both formative research with PWUD (Hughto et al., 2022) and review of surveillance data from Massachusetts and Rhode Island (Carroll et al., 2017; Shrestha et al., 2021). These processes informed both the locations where data were collected and overall study eligibility criteria. Participants of the parent study were then sampled through a modified respondent-driven sampling approach (see Hughto et al., 2024 for additional details). The Greater Providence Rhode Island community-based sample includes 56 participants recruited between March and April 2022 (Hughto et al., 2024). Eligible participants were 18 years or older; spoke English or Spanish; used an illicit stimulant in the past 30-days (e.g., powdered and crack cocaine, methamphetamine, or non-prescribed stimulants); and lived in or spent the majority of their time in the Greater Providence region.

Incarcerated sample

From April to July 2023, PWDD were recruited at RIDOC with the primary goal of learning how and why fentanyl enters into the stimulant supply and to further inquire on other adjacent topics (e.g., overdose response). Prior to data collection, a select group of 61 inmates were sent a letter that described the research and provided study contact information. Seven bushiness days after letters were sent, research staff screened and enrolled potential participants. All eligible participants in this sample were sentenced and incarcerated in RIDOC for manufacturing and/or distributing illicit drugs; incarcerated for ≤ three years; 18-years-old or older; and spoke and understood English or Spanish. Following screening processes, 30 participants were enrolled. All participants were first surveyed and then invited to be interviewed by research staff over the course of one hour. Interviews occurred in a private space where correctional staff were not present. Following site visits, staff documented post-visit ethnographic field notes further reflecting on emergent themes. For further details, see Kelly et al., 2025.

Survey and variables

The survey collected socio-demographics, drugs use, overdose history, knowledge of naloxone and overdose prevention policies, attitudes towards treatment, contamination of the drug supply, packaging and drug distribution, and use of protective tools like naloxone, fentanyl test strips, and drug checking. Participant survey responses were coded (yes/no) for analysis: awareness of the GSL; able to describe the GSL correctly in their own words, prior experience with an opioid overdose, and whether this occurred after consumption of opioids or after use of stimulants. Experiences with last witnessed opioid overdose and last witnessed stimulant overdose included: whether the participant personally called 911; personally administered naloxone; had any police interaction, and whether this interaction was positive/neutral vs. negative; had any interaction with Emergency Medical Services (EMS), and whether this interaction was positive/neutral vs. negative; familiarity with naloxone, including awareness, knowing where to get naloxone, and whether they carry or carried naloxone. The survey was programmed in Qualtrics for ease of administration.

Interviews

Community survey respondents were invited to participate in a qualitative interview if their survey responses warranted qualitative exploration (e.g., participant had unique or extensive use patterns of stimulants and opioids, overdose experiences, etc.) (n = 22) (Hughto et al., 2024). All 30 RIDOC participants completed an in-depth interview. Interviews assessed similar domains as the survey in greater depth, although there were some notable differences with respect to the topics covered in each interview guide (available upon request). Both community and RIDOC participants were asked to reflect on opioid and stimulant overdose responses, naloxone administration and interactions with first responders. Community participants were asked specific questions about the GSL, whereas RIDOC participants organically and unprompted reflected on GSL in discussions about overdose response.

Analysis

Analysis of quantitative data employed chi-square tests of independence using SAS OnDemand for Academics to explore associations between sample type (community or RIDOC) and the following survey variables: awareness of the GSL and ability to correctly describe it; characteristics of personally experienced opioid overdoses; characteristics of last witnessed opioid overdose and characteristics of last witnessed stimulant overdose. Two-tailed statistical significance was established at .

Transcribed qualitative interviews were assigned a randomly generated pseudonym, coded and analyzed in Dedoose version 9.0.17 and thematically analyzed by the research team (Braun and Clarke, 2006). While coding procedures are detailed elsewhere (see Hughto et al., 2024), briefly, interviews with PWUS were coded using a codebook developed by three trained coders and the third author. We used deductive codes consistent with the interview guide and inductive codes that emerged during an open coding process. This codebook was adapted to code interviews with PWDD following a similar deductive and inductive process. Following training and a reliability assessment, transcripts were independently coded, reviewed, and discussed at weekly coding meetings to resolve uncertainties and ensure consistent code application.

The current qualitative analysis was limited to interview content specific to GSL. Community sample excerpts tagged with a GSL code were exported into a separate document for analysis. Because RIDOC interviews included no specific GSL questions, our analysis is limited to questions and codes about overdose experience, many of which contain emergent perspectives on the GSL. To improve the rigor of our approach, we conducted a keyword search across all RIDOC transcripts for “GSL”, “Good Sam” and “Good Samaritan Law” to pull data not reflected in existing codes; the excerpts were also exported for analysis. All qualitative data were reviewed and thematically analyzed using Braun & Clark’s (2006) six-phase framework, employing an inductive approach to allow for themes to emerge from the data. The first author reviewed and applied secondary codes to the data, drafted reflexive analytic memos about emergent GSL-related themes. These themes were then refined and discussed collaboratively by the first and senior author.

The current study’s research protocol was reviewed and approved by the Institutional Review Board of Brown University as well as the RIDOC Medical Research Advisory Group.

Results

Characteristics of community and RIDOC samples

Table 1 details the demographic characteristics of both samples. Community participants (n = 56) primarily identified as male (64.3%); ages 35–44 (28.6); were securely housed (58.9%); and reported using powdered cocaine (57.1%), crack (87.5%), and fentanyl (53.6%) in the last 30-days. RIDOC participants (n = 30) were also predominantly male (86.7%), ages 18–34 (63.3%), and reported using powdered cocaine (40%), crack (23.3%), and fentanyl (26.7%) in the 30 days prior to incarceration.

Table 1.

Participant characteristics by sample

	Providence Community Sample (N = 56)		Rhode Island Department of Corrections (RIDOC) Sample (N = 30)
Characteristic	n	%	n	%
Gender identity
Male	36	64.3%	26	86.7%
Female	19	33.9%	4	13.3%
Nonbinary/gender nonconforming	1	1.8%	0	0.0%
Age in years, median (IQR)	46.0 (35.5–56.5)	-	33.0 (28.0–40.0)	-
Age in years, categorical
18–34	9	16.1%	19	63.3%
35–44	16	28.6%	7	23.3%
45–54	15	26.8%	4	13.3%
55–64	13	23.2%	0	0.0%
65+	3	5.4%	0	0.0%
Race
Non-Hispanic White	22	39.3%	8	26.7%
Non-Hispanic Black	13	23.2%	5	16.7%
Hispanic or Latinx	15	26.8%	13	43.3%
Another race	6	10.7%	4	13.3%
Educational attainment
Some high school or less	21	37.5%	8	26.7%
High school graduate or GED	22	39.3%	14	46.7%
Some college or more	13	23.2%	8	26.7%
Securely housed	33	58.9%	-	-
Weekly income, median (IQR)	250.0 (199.0–400.0)	-	-	-
Ever arrested	51	91.1%	-	-
Ever incarcerated	44	78.6%	-	-
Currently on probation	20	35.7%	-	-
Released in the past year	12	21.4%	-	-
Total years incarcerated, median (IQR)	-	-	8.5 (5.0–13.0)	-
Substances used in the past 30 days orpast 30 days prior to incarceration
Fentanyl or heroin	30	53.6%	8	26.7%
Powdered cocaine	32	57.1%	12	40.0%
Crack cocaine	49	87.5%	7	23.3%
Methamphetamine	13	23.2%	4	13.3%
Ecstasy/MDMA	4	7.1%	10	33.3%
Experienced opioid overdose, ever	24	42.9%	11	36.7%
Number of experienced overdoses, median (range)	0 (0–20)	-	0 (0–20)	-
Services used
Methadone, current	14	25.0%	-	-
Buprenorphine, current	6	10.7%	-	-
Fentanyl test strip use, ever	19	33.9%	9	30.0%

Values are n (%) unless otherwise indicated

IQR Interquartile range. Dashes (–) indicate data not collected for that sample

Knowledge and awareness of GSL – community vs. RIDOC

As shown in Table 2, awareness of GSL in Rhode Island was high among PWUS in the community (61.1%, n = 33/56) and among the RIDOC sample (63.3%, n = 19/30). However, a significantly greater proportion of RIDOC participants were able to correctly describe the law compared to PWUS living in the community (70.0%, n = 21/30 vs. 42.9%, n = 24/56; ). These findings highlight that, while both groups were mostly aware of the existence of the GSL, their understanding of its provisions differed.

Table 2.

Good Samaritan law awareness, overdose experience, and 911 help seeking by sample

Variable	Response	Providence Community Sample (N = 56)		RIDOC Sample (N = 30)		Test Statistic (df)	p-value
		n	%	n	%
Good Samaritan Law (GSL)
Aware of GSL	Yes	33	61.1%	19	63.3%	1.27 (1)	0.259
	No	21	38.9%	11	36.7%
	Missing	2		0
Correctly described GSL	Yes	24	42.9%	21	70.0%	3.93 (1)	0.048ᵃ
	No	32	57.1%	9	30.0%
	Missing	0		0
Personally experienced opioid overdose
Ever experienced opioid overdose	Yes	24	42.9%	11	36.7%	0.39 (1)	0.533
	No	32	57.1%	19	63.3%
	Missing	0		0
Overdose after any intentional consumption of opioids	Yes	19	79.2%	11	100.0%	0.20 (1)	0.656
	No	5	20.8%	0	0.0%
	Missing	32		19
Overdose after only consuming stimulants	Yes	6	25.0%	2	18.2%	2.67 (1)	0.102
	No	18	75.0%	9	81.8%
	Missing	32		19
911 help seeking at last witnessed opioid overdose
Any witnessed opioid overdose	Yes	42	75.0%	18	60.0%	2.99 (1)	0.084
	No	14	25.0%	12	40.0%
	Missing	0		0
Personally called 911 for opioid overdose	Yes	12	28.6%	3	16.7%	0.95 (1)	0.329
	No	30	71.4%	15	83.3%
	Missing	14		12
Personally administered naloxone for opioid overdose	Yes	20	47.6%	0	0.0%	14.12 (1)	<0.001ᵃ
	No	22	52.4%	18	100.0%
	Missing	14		12
Any police interaction	Yes	19	45.2%	6	33.3%	0.86 (1)	0.354
	No	23	54.8%	12	66.7%
	Missing	14		12
Positive/neutral police interaction	Yes	14	73.7%	4	66.7%	0.11 (1)	0.739
	No	5	26.3%	2	33.3%
	Missing	37		24
Any EMS interaction	Yes	26	61.9%	7	38.9%	2.17 (1)	0.141
	No	16	38.1%	11	61.1%
	Missing	14		12
Positive/neutral EMS interaction	Yes	24	92.3%	7	100.0%	0.57 (1)	0.449
	No	2	7.7%	0	0.0%
	Missing	30		23
911 help seeking at last witnessed stimulant overdose
Any witnessed stimulant overdose	Yes	18	32.1%	22	73.3%	12.85 (1)	<0.001ᵃ
	No	38	67.9%	8	26.7%
	Missing	0		0
Personally called 911 for stimulant overdose	Yes	4	22.2%	2	9.1%	1.34 (1)	0.247
	No	14	77.8%	20	90.9%
	Missing	38		8
Any police interaction	Yes	8	44.4%	4	18.2%	4.92 (1)	0.027ᵃ
	No	10	55.6%	18	81.8%
	Missing	38		8
Positive/neutral police interaction	Yes	3	37.5%	1	25.0%	0.19 (1)	0.665
	No	5	62.5%	3	75.0%
	Missing	48		26
Any EMS interaction	Yes	10	55.6%	3	13.6%	8.57 (1)	0.003ᵃ
	No	8	44.4%	19	86.4%
	Missing	38		8
Positive/neutral EMS interaction	Yes	9	90.0%	3	100.0%	0.33 (1)	0.569
	No	1	10.0%	0	0.0%
	Missing	46		27

Values are n (%) unless otherwise indicated

Percentages calculated using non-missing responses only

DF Degrees of freedom; EMS Emergency medical services, GSL Good Samaritan Law, RIDOC Rhode Island Department of Corrections

^a p < 0.05

Interview data converged with the survey data such that RIDOC participants tended to articulate a greater awareness of the GSL and a better overall understanding of its application during interviews relative to PWUS in the community. Participants from RIDOC spoke about their knowledge of GSL, reflecting on their lack of trust in the law, and also demonstrating a nuanced understanding of criminal consequences. ‘Luke’, an incarcerated participant who has a history of prior substance use, noted that, while GSL may afford some protections in an overdose situation, police may leverage information from the overdose scene for future action. He shared:

It [the Good Samaritan Law] -- applies in my situation because I was out cold and I didn’t know somebody was calling. So, all they could do when they came to me in a car, was take my drugs. But if they [Police] come to your house and they see everything [drugs] -- they going back two days later, building a report and raiding your house two days later. They didn’t arrest you that day, but -- now they know what you are and what you’re doing…I mean, for me in a car, all they could do was take my drugs. So, that’s fine, but coming to your house, that’s a different story.

Conversely, our quantitative results helped explain that PWUS in the community were less likely to correctly describe the GSL. Even when community participants were familiar with GSL, they described a general lack of understanding about GSL among their peers and discussed the burden of needing to educate others. In an interview, ‘Curtis’, a community participant, noted “It’s very frustrating to me how few people seem to know about the Good Samaritan Law. I feel like I’m constantly telling people about that”. Similarly, another community participant, ‘Justin’, commented on the lack of consistent knowledge of GSL among community members, stating:

It’s very important, because some people… they don’t even know [about the GSL], and you have one person that knows, like myself, and I tell you call the ambulance. But even if they don’t, I’m going to…because I know [the GSL]. So, they’re not going to be scared to get charged now. You know what I’m saying?

While there was recognition of insufficient and inconsistent knowledge of GSL, some community participants acknowledged the importance of the law in providing what they believed were necessary protections. For instance, ‘Thomas’, a community participant, shared his perceptions of GSL:

It’s good [the GSL] because I really don’t care what’s going on, if I got warrants at the time, I’m too worried about the person getting help at the time, so it doesn’t really matter.

Here, Thomas describes how GSL aligns with his altruistic stance to responding to an overdose yet his remark conveys a level of misunderstanding about the law; there is no language in the Rhode Island statute that provides protections for outstanding warrants. Conversely, RIDOC participants, like ‘Cedric’, more consistently communicated how they would respond:

“I’m not playing with somebody else’s life. Can’t have that on my conscience. I’m calling [the police], but I might leave…I might leave, make sure somebody’s there with them, but I’m not sticking around for the cavalry to show”.

Our survey results showed that RIDOC participants had an overall better understanding of GSL than community participants. However, there were meaningful qualitative differences as community members often conveyed that the law was not well understood broadly and shared a sense of responsibility to educate others regarding its protections. That said, both samples shared a similar distrust toward and fear of engaging with the police due to how the police might leverage these situations.

Criminal-legal system involvement as a barrier to calling 911

Most PWUS in the community had some level of involvement with the criminal-legal system within their lifetime: 91.1% had been arrested and 78.6% had been incarcerated. Among those with a history of incarceration, 45.5% were on probation and 27.3% had been released within the past year. Within the RIDOC sample, for both lifetime incarcerations and total years incarcerated, the median number was 7.5 (IQR = 4.0–16.0) and 8.5 years (IQR = 5.0–13.0), respectively. With respect to laws like GSL that provide protections against arrest (in certain circumstances), in some instances, people from both samples qualitatively described being fearful of the consequences of calling 911 due to their history of criminal-legal system involvement. For instance, ‘Victor’, a community participant, shared:

I called 911 on my way out, on my phone. Just to make sure that they got there because, when I saw that dude finish the shot, I said, I can’t depend on him to make the [call] you know? He could go out, too. So, but I didn’t want any part of it because there’s consequences… My name is bad enough with the law. I didn’t need to be mixed up in that shit. But I did what I had to do, you know. I did what I could do.

While community members like Victor may be willing to call 911, fears about being present when the police arrive on scene persist. Some community participants indicated that they knew the protections afforded by GSL, but were still hesitant to be associated with an overdose event due to past interactions with the criminal-legal system or the appearance of being aligned with the police. ‘Barbara’, another community participant, reflected on these fears stating:

They [PWUD] don’t want to be affiliated with, you know [police], or be known as [a snitch] … I’m like, I don’t care. I was here. You know? Just because I’m here, doesn’t mean I’m affiliated with [police] or in with that crowd.

As articulated by Barbara, the perception of colluding with the police among PWUD may present an additional barrier to help-seeking during an overdose event, particularly when the caller has a history of engagement with the criminal-legal system and community members are unaware of GSL’s existence and protections.

Despite the Rhode Island GSL being specifically amended to provide additional protections against probation and parole violations for individuals who call 911 after a suspected overdose, participants from RIDOC often expressed fear of these violations as a barrier to calling. These participants reasoned that if they were to call 911 for personally experiencing or witnessing an overdose, it would result in a violation of the conditions of their probation or parole. Because a violation could result in incarceration, participants anticipated legal consequences for help-seeking or receiving medical care for an overdose. ‘Tristan’, from RIDOC, stated:

For me, right? I was on parole. I was afraid to get help because it would be a violation of my parole, which it was. When – I overdosed, I violated my parole that way… Because I would be punished.

Other participants from RIDOC echoed similar sentiments but spoke more candidly about their fears of being issued a warrant and arrested following an overdose. When reflecting on GSL, Luke discussed how overdose situations can feel unpredictable and uncontrollable because the outcome of police involvement hinges on officers’ discretion. He stated:

So, normally with warrants, they [police] will leave you alone, sometimes depending on the cops in the parks they [PWUD] run into, they’ll take them anyway. And then if you are on probation or parole, also depending on who the cops are, as well. There could be a different story. You could get hit with a violation of one or the other [probation or parole].

While the law also enshrined possible incentives for calling 911 during an overdose event in Rhode Island, like negotiating lesser sentences, these incentives were not discussed by any participants. Together, these findings demonstrate that despite high awareness of GSL, its effectiveness may be undermined by persistent legal fears shaped by past and current criminal-legal system involvement.

Considerations for calling 911 by overdose event type

Across both samples, there were no statistically significant differences in having a police or EMS interaction during their last witnessed or experienced opioid overdose events or the quality of these interactions (positive/neutral vs. negative) (Table 2). In interviews, participants in both samples described similar experiences when responding to unexpected opioid overdoses that occurred when fentanyl was present in a stimulant like crack or powder cocaine. In these situations, conventional opioid overdose responses like administering naloxone and calling 911 were described by participants as being needed. ‘Adriana’ from RIDOC, who exclusively reported using stimulants, explained why she sought help following an overdose event when a friend used crack cocaine and began to exhibit opioid overdose symptoms:

They were nodding off after they [used crack cocaine] -- they were up for a little bit, and then they started crashing. So then, so all of a sudden, I’m like, “Are you okay?” They were like, “Nah. You need to call 911”. So that’s when I called 911.

Additional considerations for calling 911 during unexpected opioid overdose events were shaped by participants knowledge of the drug-induced homicide law in Rhode Island. Both community and RIDOC participants expressed a fear of incarceration for homicide due to selling an adulterated product containing fentanyl. ‘Joanne’, a community participant, reflected on a conversation with her dealer following an overdose she experienced after smoking crack:

“Oh shoot, I can’t sell this to nobody,” and I said, “No, you can’t”. You better throw that shit away because if I’m the first overdose you had out of that batch. So, if you’re going to sell it, it’s either you’re taking the risks…you better sell it to people who do fentanyl. If not, you know, you’re going to end up killing somebody.

Several PWDD described needing to carefully navigate GSL and the drug-induced homicide law in these situations. ‘Marcelo’, from RIDOC, discussed an instance where he called 911 after selling a bag to a client because he was made aware that she had overdosed:

I had already called 911 before I even knocked on the door [of client]. I was nervous, you know? I was nervous that my shit [cocaine] that I served the lady -- I mean, I could’ve been charged with her death. I don’t know if it could’ve been led back to me, you know what I’m saying? So, instead of doing just 7 years, I could’ve been doing 25 years, you know what I’m saying?

As described by Marcelo, calling 911 following an overdose event may be the preferred option, not only to save another person’s life, but also because being charged with both homicide and distribution is worse than distribution charges alone. In this way, proactively calling 911 in an effort to prevent death serves a self-protective legal purpose while also mitigating overdose risk.

In addition to unexpected opioid overdoses, in the survey, participants in both samples reflected on stimulant overdose events. In our sample, a greater proportion of RIDOC participants reported witnessing a stimulant overdose within their lifetime compared to PWUS in the community (73.3%, n = 22/30 vs. 32.1%, n = 18/56; ). Similar to their opioid overdose response behaviors, there was no difference by sample for personally calling 911 during their last witnessed stimulant overdose (22.2% PWUS vs. 9.1%, PWDD, ). Compared to PWUS in the community, PWDD were significantly less likely to report having any police interaction (18.2%, n=4/22 vs. 44.4%, n=8/18; ) and any EMS interaction (13.6%, n=3/22 vs. 55.6%, n=10/18; ) at their last witnessed stimulant overdose.

Qualitatively, there were also differing perceptions regarding when to seek help during stimulant overdoses. For instance, ‘Ty’ from RIDOC, who had history of stimulant and opioid use, was asked how he would respond to a stimulant overdose. He explained:

Well, if you needed to call 911, I guess people call 911. But usually, it’s just like drink some water…Like a cloth on your head. You’d be all right… Like if I ever felt somebody was on [opioid] overdose, I’d call 911.

Ty described alternatives for addressing stimulant overdose symptoms while demonstrating how participants linked calling 911 with opioid overdose symptoms rather than stimulant overdose symptoms. Similarly, ‘Anita’, a participant from RIDOC who only used stimulants, described how differences in the type of overdose event can influence help seeking behaviors:

I had an opioid overdose outside my house one day. The person was on the sidewalk [overdosed]. So, I called 911. I gave them Narcan, but, I wasn’t even with the person at the time. They were walking’ down my street and fell out. Versus if you’re completely high and obliviated on crack-cocaine, and someone’s shooting’ coke and the heart gets ready to explode and stops, it’s a little bit different. Cause now they say it as, “Well, maybe you coulda helped that person,” or, “Maybe you could have done something different with that person,” you know? So, it’s a little bit different.

Anita’s quote illustrates how both opioid and stimulant overdoses can be life-threatening without response, while also acknowledging the unique nature of, and differences in bystander responses to, stimulant versus opioid overdoses.

Participants also described how the police response to people experiencing a stimulant overdose shaped overdose response approaches. For example, in our interview with Anita she noted that, “The [police] don’t really like to be around people at all, high on cocaine that are yelling and going’ like, a little overboard with it, you know?” Similarly, Luke from RIDOC discussed his uncertainty regarding how police would respond during these events. He explained:

It’s bad. Like, they [police] because they don’t know what’s gonna happen and I mean, I get it. They don’t know what’s gonna happen. You could just pull a knife and charge them or whatever. I mean, I get it but at the same time, it’s like, come on? But, yeah, they [PWUS] get a little outta hand. And then it’s always the EMS that are trying to calm you down and it’s always the cops that are trying to get you riled up”.

Several participants like Luke believed that police might exacerbate stimulant overdoses by improperly responding or failing to de-escalate stimulant overdose situations. These narratives are consistent with our quantitative findings that RIDOC participants reported fewer interactions with police or EMS during stimulant overdose situations. The uncertainty of how police may handle stimulant overdose events may increase reticence to seek help.

Discussion

This mixed methods study explored and compared the perspectives of PWUS in the community and incarcerated PWDD in Rhode Island, uncovering differences in their understanding and application of the GSL. We found that PWUS in Rhode Island were less likely to accurately describe the GSL than the incarcerated RIDOC sample, suggesting a need to further investigate the dissemination and application of the law in Rhode Island. Prior research indicates that bystanders with accurate GSL knowledge are more likely to call 911 compared to those with incorrect or incomplete understandings (Jakubowski et al., 2018). Consistent with the literature, our research found that PWUS in the community-based sample who were familiar with the GSL felt more compelled to intervene and call for help during overdose situations and felt a stronger sense of responsibility to educate others about the law. However, while greater knowledge of the law is beneficial, deep fears of, mistrust in, and contact with the criminal-legal system remain prominent and evidently drive reticence to seek help; therefore, increasing knowledge of the GSL alone may not always translate into traditional help-seeking behaviors like calling 911 (Jakubowski et al., 2018; Latimore & Bergstein, 2017; van der Meulen & Chu, 2022).

The PWDD we interviewed were savvy to GSL limitations likely applicable to their circumstances (i.e., drugs packaged for distribution on scene) and therefore largely refrained from help-seeking. This is likely because PWDD from the RIDOC sample, were more acutely aware of the consequences of criminal-legal involvement, whereas in some instances participants from the community may have not fully understood the protections of the GSL. Given that our community sample was comprised of people who use stimulants, it may be that people who use drugs other than opioids may benefit from more education on GSL application and protections. Alternatively, PWUS may utilize informal practices to mitigate stimulant overdose symptoms without formal intervention (i.e., calling 911) and may not be as well connected with harm reduction organizations that provide overdose response education and training on GSL typically to people who use primarily opioids (Evans et al., 2016; Kelly et al., 2023).

There were also discrepancies between the law on the books and the perception of GSL protections. Some participants suggested that police misinterpret circumstances where the GSL should protect PWUD or leverage information during overdose response calls for subsequent criminal pursuits. Despite the intended goals of GSLs, practical application by police officers allows for discretion through their institutionalized lens (Beletsky et al., 2005). Studies have shown that despite having GSLs, police do not always consistently apply these laws and cannot always accurately define to whom or when they apply (Banta-Green et al., 2013; Carroll et al., 2020; Koester et al., 2017; Xavier et al., 2021), though police trainings have been shown to improve GSL knowledge (Saucier et al., 2016). Increasing transparency on GSL protections may encourage help-seeking behavior in PWUD and PWDD (Jakubowski et al. 2024), yet persistent fears of criminal-legal system contact may hinder their doing so. Future research should explore the need for targeted police interventions, along with further investigating police discretion in GSL application across U.S jurisdictions.

Similar to other New England states (e.g., Massachusetts, Vermont), the Rhode Island GSL contains protections extending to probation and parole aimed to mitigate concerns of calling 911 during an overdose (The Good Samaritan Overdose Prevention Act of 2016 R.I. Gen. Laws § 21-28.9-4, 2016), which were included because of high community supervision rates and the need for extended immunities (Bender, 2016). Despite this language, RIDOC participants communicated fears that calling for help during an overdose would violate conditions of their probation or parole. In some instances, participants described calling for help and subsequently being punished or incarcerated, suggesting greater oversight of GSL application is needed. Targeted interventions should focus on educating currently and formerly incarcerated people on all aspects of the GSL to mitigate fears of violating probation or parole when seeking help during an overdose. While violation fears were particularly salient within our RIDOC sample likely due to the fact that they were currently incarcerated at the time of the interview; these views were less commonly expressed by our community-based sample. Additionally, we found that Rhode Island’s drug-induced homicide law impacted decision-making related to calling 911. Prior research found that these laws dissuaded people for calling for help (Peterson et al., 2019), our research extended these findings highlighting that PWDD were aware of these laws and fearful of being given harsher sentences if they were to be charged with a homicide following an overdose. In turn, decision-making in these instances reflected dealers pulling products off the market or intervening by calling 911. Further research on the possible synergy between GSLs and drug induced homicide law effects is warranted.

Conversely, participants differed in their perceptions on when bystanders should call 911 based on the type of overdose events. We distinguished between unexpected opioid overdoses (i.e., overdoses caused by fentanyl unexpectedly present in a stimulant) and stimulant overdoses (i.e., overdoses caused solely by the effects of using stimulants). Our findings found no significant difference between the community and incarcerated samples in calling for help during opioid overdoses. However, we found that RIDOC participants were significantly less likely to call 911 during stimulant overdose events due to fears of stimulant overdose effects and potentially harmful police interactions with someone overdosing on stimulants (i.e., overamping) (Mansoor et al., 2022). Further, research on PWUS by Ambrose et al. (2016) found that reported stimulant use by the people who experienced an overdose was associated with lower odds of calling 911, suggesting a need for behavioral interventions focused on educating PWUS on best practices around overdose response. In Rhode Island, the GSL uses language attributed to “drug or alcohol overdose” and does not use expansive language to describe what “drug” means (The Good Samaritan Overdose Prevention Act of 2016 R.I. Gen. Laws § 21-28.9-4, 2016.). Our findings suggest the need for comprehensive overdose education curriculum to specifically address the GSL in application to opioid and stimulant overdoses, including signs, symptoms, and best practices for responding to both types of overdoses (Hughto et al., 2024).

Finally, the majority of participants in both samples communicated mistrust and fear of engaging with police following an overdose. This commonality suggests that fear surrounding help-seeking may be largely based on the extensive criminal-legal system involvement histories that many participants shared (McLean, 2018). Our findings point to opportunities for educating police on proper application of the GSL, and providing alternative pathways for help-seeking where individuals can call for help during an overdose without police involvement (e.g., emergency medical professional or peer trained first responders) may mitigate this fear (Formica et al., 2022; Wagner et al., 2019). De-implementing policies of universal police co-response in suspected overdose is another strategy, since such calls are rarely “combative” and minimize criminal-legal involvement on scene (Becker, 2021; Macmadu et al., 2022).

Limitations

There are several important limitations to our study. We only examined GSL in one state, yet states vary in the legal protections afforded to help-seekers and those experiencing a suspected overdose. Some GSLs provide protection from arrest, probation or parole violations, prosecution for drug charges, or other protections like sentence mitigation (Hamilton et al., 2021). Thus, our findings may not be transferable to states with different GSLs. Additionally, our research focused primarily on PWUS and did not include perspectives from people who only use opioids or other drugs. Given the illegality of most drugs in the U.S., we believe the help-seeking behaviors and perspectives of PWUS we describe are relevant if not directly generalizable to people using other drugs. Interview questions about GSL were limited to community participants due to time constraints, despite that RIDOC participants commonly discussed GSL in relation to prior overdose experiences and responses; this might have limited our ability to directly compare and contrast perspectives. Finally, the incarceration setting and experience may have influenced reporting of traumatic events like an overdose, potentially contributing to under-reporting and limited accounts among the RIDOC sample. While we took measures to support study participants throughout survey and interview administration, we recognize the potential trauma associated with an overdose experienced, witnessed, or responded to and the need for support and resources in the community and institutions like RIDOC (Shearer et al., 2019; Song et al., 2023).

Conclusion

Our research highlighted varying degrees of understanding of the Rhode Island GSL between PWUS in the community and those incarcerated in RIDOC for drug distribution. Examining both these samples is meaningful as it provided context into how PWDD and PWUS, understand the GSL. Findings indicate the need to clarify definitions and expand education related to the GSL and responding to stimulant overdoses, with a focus on PWUS and criminal-legal system actors to increase help-seeking and reduce overdose deaths. Future studies should consider policy interventions and educational and behavioral interventions to address these gaps.

Authors’ contributions

JS, Original Draft - Writing/Editing, Conceptualization, Methods - Qualitative , Data Collection; DP Writing/Editing, Conceptualization, Methods - Quantitative; JWH, Writing/Editing, Conceptualization, Data Collection, Funding Acquisition; SSR, Writing/Editing, Data Collection; SZ, Writing/Editing, Formatting; PJAK, Writing/Editing, Data Collection; SV, Writing/Editing, Data Collection; JD, Review, Editing, Funding Acquisition; TCG, Writing/Editing, Conceptualization, Methods - Qualitative, Funding Acquisition. All authors reviewed manuscript.

Funding

This work was supported by the Centers for Disease Control and Prevention (CDC), grant number [5R01CE003353] (MPI: Rich, Hughto, Green ). Rich, Hughto, and Green are supported by COBRE on Opioids and Overdose funded by the National Institute of General Medical Sciences of the National Institutes of Health under grant P20GM125507. The content is solely the responsibility of the authors and does not necessarily represent the official views of the CDC. The funder was not involved in the collection, analysis, or interpretation of study data.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethical approval and consent to participate

This study involved human subject and was reviewed and approved by the Brown University Institutional review board. Verbal consent was obtained from participants in this study.

Competing interests

The authors declare no competing interests.