Barriers to bystander interventions in suspected opioid-associated out-of-hospital cardiac arrests: A multiple methods study of 9-1-1 calls

Abstract

Introduction:

Opioid-associated out-of-hospital cardiac arrests (OA-OHCA) is a significant problem in the United States. Layperson interventions, including bystander CPR and naloxone may improve survival, but barriers may differ compared to other OHCA. This study aims to describe characteristics of 9-1-1 callers and patients in suspected OA-OHCAs and identify barriers to B-CPR and naloxone administration.

Methods:

This was a retrospective multiple methods study of transcribed 9-1-1 calls for suspected OHCA from two counties in North Carolina (5/2022–12/2023). Adult, non-traumatic OHCAs were included. Data were analyzed using descriptive statistics and Student’s t-test/Chi². We used thematic analysis and a combined deductive and inductive approach.

Results:

Patients with suspected OA-OHCA were younger than non-suspected OA-OHCA patients (39 vs 58 years [p<0.01]). Most patients were in a residence, however, this percentage was smaller in suspected OA-OHCA compared with non-suspected OA-OHCA (68% vs 88% [p<0.01]). Most callers in the suspected OA-OHCA group were a friend of the patient (35%), whereas most callers in the non-suspected OA-OHCA population were a family member (34%) [p<0.01]. Qualitative barriers unique to suspected OA-OHCA included: conflicting responsibilities, fear of drugs, and fear of the patient. Naloxone-specific barriers included lack of availability and lack of knowledge of use.

Conclusion:

We found significant differences in demographics between suspected OA-OHCA compared with non-suspected OA-OHCA. We also identified unique barriers in this population as well as previously described barriers which may be amplified in the setting of suspected drug use. A different approach towards cardiac resuscitation may be needed to maximize treatment and survival.

Background

Drug-related out-of-hospital cardiac arrests comprise a significant proportion of non-traumatic out-of-hospital cardiac arrests (OHCA) in the United States, up to 17% in some regions (1). These types of arrests are on the rise, with one setting reporting a 30% increase per year in drug-related OHCAs from 2015–2023 (1). A majority of these OHCAs are due to opioid overdoses, partly due to the widespread availability of potent synthetic opioids such as fentanyl (2). In 2023, over 80,000 deaths were attributed to opioid-associated OHCA (OA-OHCA) (3).

Intervention such as bystander cardiopulmonary resuscitation (B-CPR) have been shown to improve survival in OHCA and are recommended in OA-OHCA (4). Administration of naloxone is also recommended by the American Heart Association for layperson response in OA-OHCA and may improve outcomes in this population (5,6). Several barriers to B-CPR in OHCA have been identified, including physical, environmental, and social factors involving both the patient and bystander(s) when treating OHCAs. For example, delayed recognition of OHCA, the physical ability of the bystander to perform CPR, lack of knowledge on how to perform CPR, communication failure, and fear of litigation may all contribute to delays or lack of B-CPR (7–10). However, key barriers to bystander treatments in suspected OA-OHCA are largely unstudied. There are important differences between OA-OHCAs compared with non-OA-OHCAs. OA-OHCA patients are typically younger (11–14), more likely to experience unwitnessed arrests, and more likely have an initial non-shockable rhythm (12,13,15,16) compared with non-opioid related OHCAs. Similar to non-OA-OHCA, most OA-OHCAs occur in private locations, however, a larger proportion of drug-related arrests occur in public locations compared to non-drug-related and these patients are less likely to receive bystander CPR (12,16–18). The reasons for lower rates of B-CPR in this population are unclear. Examining barriers to B-CPR in this population may provide an understanding on how to address and improve public education and telecommunicator protocols.

Given the distinct differences in patient populations, environments and circumstances surrounding suspected OA-OHCAs and non-OA-OHCAs, we sought to understand whether OA-OHCAs have unique barriers to bystander interventions compared non-OA-OHCA. Understanding these differences may inform future public health interventions aimed at people who use drugs (PWUD). Through evaluating 9-1-1 calls for OHCA from 2 anonymized urban counties in North Carolina, this study aims to 1) compare patient and caller characteristics of suspected OA-OHCA and non-suspected OA-OHCA and 2) explore barriers to bystander assistance in suspected OA-OHCA, including B-CPR and Naloxone-administration.

Methods

Study design

We conducted a retrospective multiple methods study of recorded 9-1-1 calls for suspected OHCA from two counties in North Carolina. All sequential calls to the 9-1-1 call centers for County A and County B were collected from 8/2021–12/2023 and 5/2022–12/2023, respectively. Deidentified computer aided dispatch (CAD) notes and associated audio recordings of calls were obtained. A professional transcription service was used to transcribe all audio calls.

Setting

County and system demographics are described in Table 1. Each county utilizes a similar approach for call processing. Calls to 9-1-1 for each county are answered by public safety answering point (PSAP) by trained medical telecommunicators. All calls undergo a standardized process for Emergency Medical Dispatch (EMD). Each county uses a commercial EMD service with County A using Medical Priority Dispatch System (MPDS) and County B using Association of Public-Safety Communications Officials (APCO). Each system employs a structured set of protocols based on caller complaint which guides dispatch of medical resources and pre-arrival instructions (PAIs). Suspected cardiac arrest resources are dispatched and PAIs are provided via this standardized approach (Appendix).

Table 1:

County and state demographics, EMS system information and overdose data

County	County A	County B	North Carolina
Population per square mile (2020)	1,133.7	938.1	214.7
Urbanicity	Mixed urban/rural	Mixed urban/rural	--
Persons ≥ 18 years of age (%)	80.6	77.6	78.4
Sex (%)
Female	52.0	52.4	51.1
Racial demographics (%)
White alone	55.2	65.7	69.8
Black/African American alone	34.5	27.6	22.1
Asian alone	6.1	2.8	3.7
American Indian/Alaska Native alone	1.2	1.1	1.6
Native Hawaiian/Pacific Islander alone	0.2	0.2	0.2
Two or more races	2.9	2.6	2.7
Ethnicity (%)
Hispanic or Latino	15.4	15.3	11.4
Persons in poverty (%)	11.1	14.8	12.8
9-1-1 and EMS data (2023)
Total EMS call volume	57,249	51,112	--
Total suspected overdose 9-1-1 calls	903	932	--
Total non-traumatic cardiac arrests in adults ≥ 18 in which resuscitation was attempted	287	352	8,988
Drug overdose data (2023)
Annual rate of drug overdose deaths (per 100,000 residents)	40.7	40.0	41.0
Annual rate of overdose ED visits (per 100,000 residents)	138.3	139.0	114.9
Overdose deaths involving illicit opioids (%)	78.1	80.9	78.9

^*Data source: data.census.gov (52)

^**Data source: Opioid and Substance Use Action Plan Data Dashboard (NCDHHS) (53)

Study Sample and Population

Initial identification of suspected OHCA calls were identified by the 9-1-1 centers through two different methods given the differences in call processing systems. Calls from County A were identified through a final determinant code indicating a cardiac arrest response (09E01, 09E02, 31E01, 31D01) (Appendix). Calls from County B were included where telecommunicator CPR were provided. Calls from healthcare facilities or from third party callers were excluded. Calls in English of suspected adult (age ≥18 years), non-traumatic OHCA were reviewed. A cohort of suspected OA-OHCA calls was identified by reviewing transcripts and CAD notes for any mention of terminology suggestive of drug use: overdose, Narcan (or naloxone), “drugs”, specific opioids (e.g. fentanyl), or slang terms for drug use (i.e. “took something”). Any uncertainty in adjudication as a suspected OA-OHCA was reviewed by two emergency medicine (EM) and emergency medical services (EMS) trained physicians (MK and APJ) for a consensus agreement. A third EM/EMS physician (SP) provided final input if consensus was not achieved.

This study was approved by the Duke University Institutional Review Board (Pro00105060, Pro00114429).

Variables

Study team members reviewed call transcripts and CAD reports and created a template for call demographics and characteristics. Variables included suspected OA-OHCA, B-CPR, caller sex, patient sex, telecommunicator sex, call location, caller relationship, time of day, and number of bystanders at the OHCA. A call was coded as suspected OA-OHCA if there was mention of suspected drug overdose in the transcript or the CAD report (see description of identification above). Time of day was identified from the CAD report or transcript. Sex variables, call location, number of bystanders, caller relationship, and B-CPR were identified by trained study team members.

Analysis

Two coauthors (MK and APJ) performed a thematic analysis of barriers to bystander interventions (B-CPR and naloxone administration) in the OA-OHCA population. We did not apply a specific framework to the qualitative analysis given the exploratory nature of this study and the specific population of interest. A combined inductive and deductive analysis was used to analyze transcripts. A preliminary codebook was developed using previously identified themes and codes pertinent to general OHCA from prior literature in addition to new codes identified through review of transcripts from County A (19). New codes not previously described in the scoping review and emerging from OA-OHCA patients were described as “unique” to OA-OHCA. This codebook was then used to code data from County B using an iterative process of feedback and review by the study team. Due to the qualitative nature of the themes, we presented themes without frequencies, as appropriate (20,21). NVivo (version 14) software was used to develop the initial codebook (Lumivero, Denver, CO). The final coding was completed using an Excel spreadsheet file (Microsoft Corporation, Redmond WA).

Quantitative data on caller, patient and telecommunicator demographics were analyzed using descriptive statistics using mean and standard deviation for age, and frequencies and percentages for all other categorical variables. A comparison of suspected OA-OHCA to non-OA-OHCA calls was performed using Pearson’s Chi-squared statistics for categorical data and a Student’s T test for continuous data. Analysis was performed in STATA (Stata 18, Austin, Texas).

Results

We identified a total of 377 patients meeting inclusion criteria for adult, non-traumatic OHCA (Figure 1). Of these, 28.4% (107/377) were identified as suspected OA-OHCA, with 16.2% (32/197) in County A and 41.2% (75/182) in County B.

Figure 1:

Inclusion and exclusion criteria for 9-1-1 transcripts from County A and County B

The mean age for suspected OA-OHCA patients was lower than non-OA-OHCA patients in the combined group (39 ±13 vs 58 ±20; p<0.01) (Table 2). County B had a lower percentage of females with suspected OA-OHCA compared to non-suspected OA-OHCAs (27% vs 43%; p=0.02). For both counties, most OHCAs occurred in a home or residence, however, this proportion was lower in suspected OA-OHCA, 68% compared to 88% in non-suspected OA-OHCA (p<0.01). Callers in suspected OA-OHCA were more frequently friends of the patient (34% vs 9% in non-suspected OA-OHCA [p<0.01]). No statistically significant differences in time of day, number of bystanders, or B-CPR were found in either county or combined data. In County A, only 38% of suspected OA-OHCA patients received B-CPR compared to non-suspected OA-OHCA (52%). One hundred percent of suspected OA-OHCA patients received B-CPR and 98% of non-suspected OA-OHCA patients received B-CPR (all suspected OHCA calls from County B were initially identified by provision of telecommunicator CPR).

Table 2:

Characteristics of suspected OA-OHCA vs non-suspected OA-OHCA

Attribute	County A			County B			Combined
	Suspected OA-OHCA n=32	Non-suspected OA-OHCA n=165	P-value	Suspected OA-OHCA n=75	Non-suspected OA-OHCA n=107	P-value	Suspected OA-OHCA n=107	Non-suspected OA-OHCA n=276	P-value
Age (mean±SD)	44± 15	58 ± 20	p<0.01	38 ± 11	58 ± 20	p <0.01	39±13	58 ± 20	p <0.01
Patient sex, n (%)							31 (29)	106 (39)	p=0.07
Women	11(35)	60 (37)	p=0.91	20 (27)	46 (43)	p=0.02
Location, n (%)									p<0.01
House/residence	18(56)	139(84)	p<0.01	55(73)	100(93)	p<0.01	73 (68)	239 (88)
Place of recreation	1(3)	0		0	0		1 (1)	0 (0)
Industrial	0	1(1)		0	0		0(0)	1 (1)
Public/commercial building	9(28)	19(12)		13(17)	2(2)		22 (21)	21 (8)
Street/highway	4(13)	6(4)		0	0		4 (4)	6 (2)
Car/Vehicle	0	0		2(3)	2(2)		2 (2)	2 (1)
Unknown/Other	0	0		5(7)	3(3)		5 (5)	3 (1)
Relationship to caller, n (%)									p<0.01
No Relationship	11(34)	24(15)	P=0.06	0 (0)	1(1)	p<0.01	11 (10)	25 (9)
Other/Unclear	8(25)	39(24)		21(28)	20(19)		29 (27)	59 (22)
Significant other	5(16)	39 (24)		7(9)	31(29)		12 (11)	70 (26)
Family member	6(19)	56 (34)		13(17)	37(35)		19 (18)	93 (34)
Friend	2(6)	7(4)		34(45)	18(17)		36 (34)	25 (9)
Time of day, n (%)									p=0.66
Day	20(63)	99(59)	p=0.79	32(43)	40(37)	p=0.47	52 (49)	139 (51)
Night	12(38)	66(40)		43(57)	67(63)		55 (51)	133 (49)
Bystanders with patient, n (%)
none	6(25)	64(44)	p=0.17	33(44)	44(41)	p=0.76	39 (39)	108 (43)	p=0.74
1 or more	14(58)	68(47)		40(53)	58(54)		54 (55)	126 (50)
Unknown	4(17)	13(9)		2(3)	5(5)		6 (6)	18 (7)
B-CPR, n (%)							--	--	--
No	20(63)	79(48)	p=0.13	0	2(2)	p=0.23
Yes	12(38)	86(52)		75(100)	104(98)

Several themes around barriers to bystander interventions were identified in the OA-OHCA population (Table 3). These were classified into overarching categories of psychological, physical, communication and societal barriers specific to B-CPR. Psychological, physical and communication barriers have been previously described, however, a new category of “societal barriers” was identified, as well as a theme related to naloxone-specific barriers.

Table 3:

Coding themes and barriers to B-CPR

	OHCA barriers (previously described) (19)	OA-OHCA barriers (new themes)
Communication barriers	Communication failure Language barrier Noisy environment	Interference from bystanders Caller frustration/impatience
Physical barriers	Bystander’s physical limitations Patient physical attributes Physical access barriers Physical location barriers
Psychological barriers	Lack of confidence Emotional distress Fear of touching patient Medicolegal concerns	Fear of drugs Fear of patient
Societal barriers		Apathy to assistance Conflicting responsibilities Legal concerns
Naloxone barriers		Lack of availability Lack of knowledge on use

Three previously identified themes were identified within the larger category of communication barriers: communication failure, noisy environment, and language barriers. Two additional themes specific to this study were also identified: caller frustration/impatience and interference from bystanders. These themes describe specific scenarios in which the telecommunicator was unable to relay or had a delay in relaying B-CPR instructions.

Themes for physical barriers included patient physical attributes, bystander’s physical limitations, physical access barriers, and physical location barriers. These themes describe caller difficulties in physically performing B-CPR due to demographics or characteristics of the caller, patient, as well as the physical environment. No new physical barrier themes specific to OA-OHCA emerged in this cohort.

Bystanders experienced psychological barrier such as fear of drugs, fear of touching a patient, emotional distress, lack of confidence, and a fear of the patient. Fear of drugs and fear of patient were two themes that were deductively identified and unique to suspected OA-OHCA. In both of these themes, we found examples of callers who demonstrated hesitancy in providing bystander assistance in situations suspected to involve drugs. This could either be due to fear of contamination from the drug itself, or fear of the reactions from the patient who was suspected to have used drugs.

We identified a new broader category of societal barriers encompassing 3 themes: apathy to assistance, legal concerns, and conflicting responsibilities. These barriers and subsequent quotes describe cultural and societal norms around providing aid to unknown persons, as well as potential employment restrictions. Additionally, legal concerns, although previously described as “medicolegal concerns” was broadened to include uncertainties around the Good Samaritan law as well callers requesting withholding of a police response for fear of legal consequences.

Finally, the last category was naloxone-specific barriers. The two themes included lack of availability and lack of knowledge on use. In some scenarios, callers indicated they used up whatever supply they already had or that their supply had been taken. Other callers who had naloxone indicated that they did not know how to use it.

Discussion

Our findings highlight some of the unique challenges associated with bystander assistance to patients with suspected OA-OHCAs. Many of these barriers stem from the characteristics of the population of patients experiencing OA-OHCAs as well as the unique circumstances surrounding these events. Similar to other studies, we found a higher proportion of patients in public locations experiencing drug-related OHCAs, although this was not statistically significant (1,16,17). This is likely due to a number of reasons, including the frequent use of public places for both drug acquisition and use, and housing instability (22). Individuals with substance use disorder are often challenged with unstable housing, which has also been shown to be a significant barrier in receiving treatment (23,24).

Approximately 46% of patients in the U.S. experiencing OHCA of suspected cardiac etiology receive B-CPR, compared to 38% in County A in our study (25). Others have described lower rates of B-CPR (23–48%) in patients with drug-related OHCA compared to non-drug-related OHCA (25–52%) (1,12,13,18). We also found that the majority of both suspected OA-OHCAs and non-OA-OHCAs occurred in private residences. In non-OA-OHCAs, B-CPR rates are lower in private residences compared to public (26). It is unclear if this is similar in OA-OHCAs, and our study was not powered to assess for this. However, a recent study of fatal opioid overdose deaths in Rhode Island found that bystanders were present in nearly 62% of deaths occurring in a private residence or home, yet only 29% of these patients received naloxone, much lower compared to public locations (27). This is reflected in our themes around naloxone, suggesting either a lack of available naloxone or lack of knowledge on how to use it. Availibility of naloxone in private locations will be limited by access to this medication in high-risk populations through overdose education and naloxone distribution (OEND) programs.

Although most suspected OA-OHCAs occurred in private residences, similar to other studies we found a larger proportion of patients experiencing OA-OHCA in public locations in both counties (25%) compared to 20% for non-OA-OHCA (12,16,17). Several reasons for bystander hesitancy in the general OHCA have been described, many of which correlate with our findings. However, a theme we identified, “apathy to assistance” has not been identified previously. Although in general, patients experiencing OHCAs in a public location generally have better survival (28), an increased frequency of public locations for suspected OA-OHCAs can present a unique challenge as these patients may not present as obviously in cardiac arrest. Current culture when encountering someone lying on the street is to leave them alone as it may be difficult to ascertain if the person is asleep or possibly unhoused. This cultural norm in our society is likely compounded by stigma against PWUD. We find this further highlighted in the psychological themes where bystanders demonstrated fear of touching the patient or fear of potential harm to themselves. This reluctance from unrelated bystanders has also been seen in delays in naloxone administration for suspected overdose (29).

Additionally, many of the patients in our study were encountered in places of business such as hotels. We identified a perceived association with increased drug use in these settings, with one caller specifically stating, “we’ve had a lot of drug issues going on at the hotel and he doesn’t want me to go in alone”, referring to a manager. We classified these as an additional societal barrier, “conflicting responsibilities”, reflecting policies that limit direct involvement or intervention from the caller (presumed employee) until a manager or supervisor has granted permission. The purpose of these policies or directives are unclear and may stem from a combination of safety concerns for the employee, stigma, and fear of litigation. One potential safety concern may be related to recent media reports around unintentional bystander or first responder overdose through contact or inhalation of drugs such as fentanyl (30). Unintentional overdoses from these types of exposures are not based in evidence (31,32), and these types of beliefs may perpetuate and reinforce stigma associated with PWUD, potentially leading to further delays in potentially life-saving interventions from bystanders.

These unique themes are closely interwoven with known societal stigma around drug use and PWUD. Many individuals continue to view addiction as a personal failure or choice, rather than a chronic disease. Stigma related to substance use been widely described among healthcare personnel, leading to inequities in care and negative interactions which result in avoidance of the healthcare system and limited access to treatment (33–36). Public stigma has also been described, resulting in discrimination and limiting community-based interventions (35). However, there is limited data on the effect of stigma on bystander treatment of individuals with suspected drug use experiencing a cardiac arrest. Addressing these barriers to bystander engagement are a critical aspect of continued public health and education efforts.

A fear of liability has been previously identified for non-OA-OHCAs, however, these carry a different weight in PWUD (37). Many dispatch protocols automatically send law enforcement officers to all overdose calls. This may disincentivize people who use drugs and their friends and family from calling 9-1-1 (15,16,38), as the following quote from a caller demonstrates: “No police arrival, please. Just EMS.” Naloxone administration during an overdose and performing CPR in OA-OHCA are time-critical actions, yet a constant and prevailing fear of legal ramifications leads to unwillingness to call 9-1-1 (10,39,40). This fear persists despite the widespread implementation of Good Samaritan laws (GSLs) in the United States (38,41) partly due to the limited protections offered which vary greatly from state to state (42). In general, most GSLs provide some degree of legal protection to individuals who attempt to provide good faith assistance to individuals experiencing medical emergencies. However, there is significant variation between different states with respect to the breadth and degree of protections offered (42). In North Carolina (NC), the limited immunity protects those who experience an overdose or request aid for an individual who has overdosed and act in good faith while doing so but individuals must provide their name to EMS or law enforcement officers and immunity does not apply if the incident occurs during execution of a warrant (43). Some states have harsh penalties for fentanyl overdoses, drug-induced death and drug delivery resulting in death, further discouraging bystanders from calling EMS or rendering aid (44). Hamilton et al found that states that enacted GSLs protecting from arrest and naloxone access laws had 10% lower rates of death from opioid overdose by year two of passing GSLs within the state (45). Further investigations are needed to confirm these findings.

Increasing public initiatives, especially in targeted populations, will provide greater awareness and understanding of state GSLs. Another potential solution is integrating a brief explanation of the states GSL to the bystander by the telecommunicator, especially if LEOs will be dispatched. However, there is concern that this approach could cause a delay in the time to hands-on-chest. The American Society of Addiction Medicine (ASAM) recently created a public policy statement that outlines their support for strengthening GSLs (44). In the statement, ASAM seeks amendments in federal and state GSLs to include a better description of the people who could benefit from these protections and create protections against both arrest and prosecution for a wide array for drug-related offenses. They also recommend increasing public awareness of GSL for individuals involved in an overdose event but also for LEOs and emergency medical personal who may have limited understanding of their states GSLs (44,46). The creation and implementation of a homogenous GSL with consistent immunities in all states and improved public outreach could improve awareness and potentially lower overdose mortality.

Modifying response codes to replace LEOs with behavioral health or addiction professionals such as peer support specialists for suspected overdoses is another consideration. However, police may play a crucial role in scene safety in suspected overdoses. Additionally, often law enforcement can respond faster than behavioral health professionals and EMS clinicians, resulting in an earlier administration of naloxone (47,48).

Given our findings as well as previous studies comparing differences in suspected drug-related OHCA and primarily cardiac OHCAs, a different approach is needed to proactively educate and prepare potential bystanders during these events (Figure 2). Although current guidelines suggest the use of layperson naloxone in suspected OA-OHCA, the primary focus is still on high-quality CPR (6). Overdose education and naloxone distribution (OEND) programs have been widely implemented over the last several years as a part of aggressive public health efforts. These interventions have been demonstrated to increased long-term bystander knowledge in recognizing and treating overdoses and reduce mortality from opioid overdoses (49). These types of programs have been extremely effective in reaching target populations through engagement via EMS, emergency departments, harm reduction agencies, and public locations (50,51). However, these programs rarely focus on treatment of cardiac arrest through bystander CPR.

Figure 2:

Framework for understanding bystander willingness to engage in resuscitative efforts and possible solutions

Community-focused OEND programs are a prime opportunity to implement bystander CPR education as patients suffering from opioid ODs are at high risk of progressing to cardiac arrest. However, this type of education needs to be targeted and tailored to this specific audience, with input from people with lived experience. Additionally, our findings demonstrate some unique barriers to bystander interventions in this population, highlighting a need to improve communication between callers (or bystanders) and telecommunicators in these specific circumstances. Pre-arrival instructions may require a different approach compared to non-OA-OHCAs to overcome specific barriers to assistance in this population.

Limitations

This study had several limitations. The two participating 9-1-1 centers used different protocols to process emergency calls, resulting in different approaches to identifying suspected OHCAs. This is reflected in findings of County B, with 100% of patients receiving B-CPR. Due to the process used to identify calls, only those in which B-CPR was provided were included for analysis. This likely results in missed calls in which the bystander did not perform B-CPR despite telecommunicator instructions, thus resulting in a selection bias and possibly important missed barriers/themes. Thus, County A likely reflects a more accurate picture of B-CPR rates in this population. Despite these limitations, we believe that these data provide an accurate representation of barriers to a layperson initiating B-CPR in suspected OA-OHCAs, which has not previously been described. Another limitation is that our method of identifying suspected OA-OHCA was based on statements made by the caller or CAD notes. Some calls with suspected drug use may have been missed as we were unable to link with the EMS chart. However, our study was intended to focus primarily on upstream efforts and barriers specific to interactions between the caller and telecommunicator, which hinged upon information available to the telecommunicator during the time of the call.

Conclusions

This study of 9-1-1 calls from two counties in North Carolina identified important differences between suspected OA-OHCA and non-suspected OA-OHCAs as well as unique barriers to bystander interventions. We found that suspected OA-OHCA patients were on average younger, with a smaller proportion receiving B-CPR (County A) and a larger proportion occurring in public locations. Although many of the identified barriers align with previously identified OHCA barriers, we identified unique barriers in this population (fear of patient, fear of drugs, conflicting responsibilities) which may further impact bystander interventions. These findings indicate a need for a more tailored public health and education approach to address unique barriers to bystander interventions in this population.

Table 4:

Coding themes/barriers in suspected OA-OHCA patients with example quotes

Barrier	Example quote
Communication barriers
Caller frustration/impatience Definition: Situations in which the caller indicates or responds with frustration or anger towards questions asked by the telecommunicator.	Caller 1: “I need to go clear my g*dd**n [inaudible] [00:01:28].” Dispatcher: “How old is the patient?” Caller 2: “I don’t know how old he is.” Caller 1: “G*dd**n, what you want to write a book?” Dispatcher: “I don’t. Can you step away from him?” Caller 1: “[Inaudible – crosstalk] I know why so many people die with them motherf*****s, man. Why you talk to us for? Hang up on that m [inaudible – crosstalk] –” (County A - A156) Dispatcher: “[Redacted], take a deep breath and talk to me. I’ve got to ask you some questions, okay?” Caller: “Just send the f*****g paramedics over. Like, d**n.” Dispatcher: “Ma’am, is he breathing?” Caller: “No, he’s not [inaudible] [00:01:11] send the paramedics over you f*****g idiot.” (County B - 14174)
Interference from bystanders Definition: Scenarios in which multiple bystanders or callers are contributing to the situation or interfering with effective communication with the telecommunicator.	Caller: “The lady on the phone is telling me to do that. What are you talking about?” Bystander: “Yeah, but I have the Narcan –” Caller: “Okay, well then, do it. Just do it.” Bystander 1: “[Inaudible] Please don’t make me.” Dispatcher: “If you’re giving him the Narcan we’re good.” Bystander: “[Inaudible] Just do it, f**k him.” Caller: “It’s on him. [Inaudible] the matter, just focus on him. Focus on him, not me. Focus on him.” Bystander: “Everybody shut the f**k up. Please.” Caller: “Focus on him.” Bystander: “Hold up. He ain’t breathing.” Dispatcher: “Ma’am, did he give him the Narcan?” Caller: “Did you give him the Narcan? She’s asking.” (County A - A243) Caller: “He’s dying. Help me. [Redacted]” Bystander: “[Redacted], you gotta keep it together. They’re coming.” Caller: “Come on. Come on, hurry up. Hurry up.” Bystander: “Get out of the way. Get out of the way.” Caller: “Let me get up. Help me. Help me. He did drugs. He took drugs.” Bystander: “They know that. They know that.” (County B - 28392)
Psychological barriers
Fear of drugs Definition: Caller demonstrates a fear of drug involvement in the situation, thus causing delayed or lack of evaluation/treatment	Caller: Okay, I’m sorry, but these people are getting freaking drugs in, and I’m scared that’s what it is. And I don’t want one getting, I don’t want that to happen. I don’t know sir, down. I can’t get him up. I’m sorry ma’am, this is just, this is the first time on my watch this has happened. Yeah. Are you still with me? (County A - A121)
Fear of patient Definition: Fear that the patient may cause the caller or bystander physical harm.	Caller: “And I don’t know if he’s alive. I’m really honestly afraid to touch his neck or whatever.” Dispatcher: “Okay. Yeah.” Caller: “‘Cause if he is alive, I don’t know how he’s gonna respond.” (County A - A176)
Societal barriers
Apathy to assistance Definition: Caller indicates a distinct aversion to, or disinterest in assisting patient.	Caller: “Ma’am. I’m not a doctor. I’m getting ready to go in and eat some Mexican food, and I saw this guy on the sidewalk.” Dispatcher: “Okay. Are you able to check on the person to see if he’s breathing or not?” Caller: “No, ma’am. No, ma’am. I’m not walking over there. I don’t know what the situation is. I’m just trying to call in. If he’s having a heart attack or if he’s dying, then y’all can send somebody.” (County A - A297a) Bystander: “Can I just move?” Caller: “No.” Bystander: “I’m going outside.” Caller: “Sit your a** down.” Bystander: “Ma’am, I’ll be right out here. I need a cigarette.” Dispatcher: “Is there anybody there –” Caller: “I can’t deal with this. It’s your fault.” Bystander: “It ain’t my fault. I didn’t get into Ice.” (County B - 23265)
Legal concerns Definition: Caller expresses concern over or questions the legal ramifications of assisting patient.	Caller: “You guys have the Good Samaritan law out here, right?” Dispatcher: “I’m not sure, why? Tell me what happened, sir?” Caller: “I just – I mean, she took a tiny bit of fentanyl and like she’s – just like her eyes are open, but she’s not there, and I’m just scared. I’m just scared.” (County A - A204) Caller: “No. And can you make sure no police arrival, please.” Dispatcher: “You what?” Caller: “No police arrival, please. Just EMS.” (County B - 06432)
Conflicting responsibilities Definition: 1) Caller states they cannot assist patient due to other needs (i.e. providing childcare, work responsibilities, etc.). 2) Caller is unable to assist without obtaining permission from someone else, such as a work supervisor.	Caller: “I’m sorry. My manager – we’ve had a lot of drug issues going on at the hotel and he doesn’t want me to go in alone.” (County A - A152)
Naloxone-specific barriers
Lack of availability Definition: Caller or bystander indicates either naloxone/Narcan is unavailable or any existing supply has already been administered.	Caller: “I gave him one Narcan, that’s all I had. We, we...” (County A - A116) Dispatcher: “Okay. So, he’s got a history of using heroin? Do you have any Narcan?” Caller: “No. I don’t. Somebody took it.” (County B - 25543) Caller: “I’ve had Narcan, but I don’t know where it’s at and the only one I have is the one for injection and I don’t have the needle for it. It’s the one already pre-mixed in the needle, but no needle in.” (County B - 43831)
Lack of knowledge on use Definition: Caller or bystander indicates or demonstrates a lack of knowledge on how to administer naloxone.	Dispatcher: “Okay. Give me one moment here. If there’s a defibrillator available send someone to get it now and tell me when you have it, okay?” Caller: “I don’t believe there’s one available. I do have someone available with Narcan, though. How do I administer that?” [Inaudible – crosstalk]. (County A - A144) Caller: “I’ve got Narcan but I don’t know how to use it, right there.” (County B - 57707)

Abbreviations:

OHCA

out-of-hospital cardiac arrest

OA-OHCA

opioid-associated out-of-hospital cardiac arrest

T-CPR

telephone CPR

B-CPR

bystander CPR

PWUD

people who use drugs

PSAP

public safety answering point

CAD

computer aided dispatch

EMD

Emergency Medical Dispatch

MPDS

Medical Priority Dispatch System

APCO

Association of Public-Safety Communications Officials

PAI

pre-arrival instructions

EMS

emergency medical services

EM

emergency medicine

GSL

Good Samaritan laws

LEO

law enforcement officer

NAL

naloxone access laws

OEND

overdose education and naloxone distribution

Appendix: Medical Priority Dispatch System

Determinant Code	Determinant Descriptor
9-CARDIAC OR RESPIRATORY ARREST / DEATH
9-E-1	Suspected Workable arrest (not breathing/ineffective breathing): Not breathing at all
9-E-2	Suspected Workable arrest (not breathing/ineffective breathing): Uncertain breathing
31-UNCONSCIOUS / FAINTING (NEAR)
31-D-1	Unconscious – AGONAL/INEFFECTIVE BREATHING
31-E-1	INEFFECTIVE BREATHING