Introduction

The USA is in the midst of an opioid overdose epidemic, involving both pharmaceutical opioids and heroin.1,2 In addition to efforts to prevent overdoses from occurring, there is an urgent need to prevent overdoses that do occur from turning fatal. Opioid overdose is particularly amenable to intervention because death typically occurs more than an hour after the onset of an overdose, allowing time for resuscitation.3 Many overdoses are witnessed by others, making timely lifesaving possible.4 In many cases, however, help comes too late. Sometimes, response is delayed because witnesses do not recognize opioid overdose symptoms as life-threatening, or because they fear legal consequences of calling 911.5–7

A fairly robust research literature that addresses training those at risk for witnessing an overdose in how to recognize and intervene during an emergency exists.3,8–11 However, much less is known about law enforcement officers’ experience and perspectives on overdose. Engaging police in opioid overdose response is critical because they are often the first to arrive at the scene and can provide or enhance effective emergency response, especially in rural and poorly resourced areas. The police also routinely interact with many individuals at high risk for overdose and can facilitate primary and secondary prevention. Law enforcement institutions routinely engage in raising awareness and educational outreach on drug issues, including in schools with a young population that is exposed to prescription opioid use and abuse. Thus, law enforcement professionals and institutions are key stakeholders in building a comprehensive response to the overdose epidemic and should be partners in providing education substantiated by research findings.

Among heroin users, research indicates fear of police response as the most common barrier to not calling 911 during overdoses.12,13 In a Baltimore study, 37 % of injection drug users who did not call 911 during an overdose endorsed concerns about police as the most important reason they did not call.13 Several states have enacted laws, commonly called Good Samaritan laws, to encourage calling 911 during overdoses on controlled substances; these laws are in part modeled on college campus alcohol Good Samaritan policies.14 Overdose Good Samaritan laws had been adopted in ten states as of the end of 2012, but they have not yet been evaluated.15 Generally, the laws include provisions that provide immunity from criminal prosecution for drug possession to overdose victims and to those who seek medical aid. Eight states have passed laws that ease access to take-home-naloxone by allowing the prescription of naloxone (an opioid antagonist or antidote) to persons at risk for having or witnessing an overdose, enabling bystanders to quickly respond in the event of an overdose.3,15 Previous research suggests that police are sometimes under-informed, and often ambivalent to public health laws, especially those based in a risk reduction framework.16,17

In June 2010, Washington State became the second state to enact legislation to address these issues (Revised Code of Washington 69.50.315).18 The law included both a Good Samaritan immunity provision for overdose victims and bystanders who seek medical aid, as well as allowed for naloxone to be prescribed to those at risk for having or witnessing an opioid overdose. The law also explicitly allows naloxone to be carried and administered by lay persons. Clearly stated in the law is “The protection in this section from prosecution for possession crimes… shall not be grounds for suppression of evidence in other criminal charges.”

Given the persistent concern that interactions between police and people at the scene of an overdose may influence proclivity to call 911 during future overdoses, we examined the experiences, attitudes, and beliefs of Seattle police officers and paramedics with regard to overdose and the 2010 law. Paramedics were included in surveying to triangulate police officers’ responses and also to help address an often contentious point—“Should police be at the scene of overdoses?” We also describe how study findings were used by police to motivate and inform an overdose training video for patrol officers.

Methods

In order to evaluate the implementation and potential impacts of the law, Seattle Police Department (SPD) officers and Seattle Fire Department Medic One paramedics were surveyed. Police officers were surveyed during 2 weeks in the Fall of 2011. The surveys were administered in-person, at “roll calls” (staff meetings at the beginning of work shifts) at each of the five SPD precincts. Two study staff (Banta-Green and Schoeppe) administered the structured surveys. Roll calls take place at 3:00 am, 11:00 am, and 7:00 pm. Study staff attended each shift time once at each of the five precincts; totaling 15 shifts and yielding a representative sample of patrol officers in the SPD. Similar questions were asked of paramedics at a single staff meeting during the Fall of 2011. Note that paramedics are advanced life support trained and certified and respond to the most serious emergencies. Paramedics’ scope of practice includes administering naloxone.

Prior to administration of the in-person paper survey, the following information was presented to the participants, verbally and in writing: research project background and overall aims; survey aims; survey instructions; and a description of the voluntary and anonymous nature of survey responses. The survey was distributed by study staff and returned to study staff in an opaque envelope. The self-administered anonymous survey took less than 5 min to complete. A single pilot test of the survey and study procedures was conducted during a police department training with only minor wording changes made as a result. Survey procedures were reviewed by the University of Washington Human Subjects Division and determined to be exempt from human subjects’ regulations. Permission for conducting the survey was obtained from an Assistant Chief of the SPD and the medical director of Medic One.

The survey questions used for the analyses presented here address experience with “serious heroin or opiate medication (e.g., OxyContin, methadone, Vicodin) overdoses” with a description of the symptoms of an overdose provided. We also asked about their knowledge of the law and level of support for and opinions of each provision of the law.

A description of the law was included in the survey towards the end, immediately preceding questions inquiring about officers’ opinion and the reasons for their opinions of each of the provisions of the law. Opinions were asked with a five-point Likert scale response: very important, somewhat important, neutral, not very important, and not important at all. For the sake of brevity, we report aggregated responses with the first two categories “important” and the last two categories “unimportant.” Following each of the Likert scale opinion questions was a space for a narrative response about the reason for their opinion. The final question on the police survey asked about intended behavior at future overdoses with multiple response options provided.

Open-ended responses were exported into Atlas.ti qualitative data analysis software, where they were analyzed by a research team member for common themes. The general process of developing themes and aggregating data into groups of themes was an iterative cycle whereby all comments were summarized by the content of comments made, as themes emerged subcategories were created and the comments re-read and aggregated. This process was repeated on the entire set of data until a complete (all comments included and all placed into a subcategory), minimal, and meaningful set of subcodes was created. This process is one of data reduction via sequential analysis.19 Other data are presented descriptively; given the exploratory nature of the research, tests of statistical significance were not applied to the data.

Results

Surveys were completed by 97 % (251 of 258) of police officers present at roll calls. The respondents represent 50 % of the patrol officers in the city of Seattle. The median of years of service as a police officer was 10.0, with a mean of 11.3 years. Only 5 % of police respondents knew that lay persons could legally possess and administer naloxone whether they are at risk for having or witnessing an overdose (Table 1). Sixteen percent knew of the existence of the Good Samaritan law, of whom 46 % knew that the law applied to both overdose victims and bystanders. Among the 36 officers who knew of the law, 58 % correctly indicated that the charges the law provided immunity for included possession, but 28 % incorrectly reported it covered other drug charges, and 6 % thought it covered warrants/probation/parole. Among all officers surveyed, 1 % felt they had received clear guidance on the law from the police department.

Table 1 Police officers’ survey responses
Full size table

Ninety-three percent of police respondents had attended a serious opioid overdose (defined in the survey) in their career, with 64 % having attended one in the past year. While 77 % of officers felt it was important they were at the scene of an overdose to protect medical personnel, a minority, 34 %, indicated it was important they were present for the purpose of enforcing laws. Arrest during the last overdose officers encountered was rare, with only 1 % of overdose victims and 1 % of bystanders being arrested. In cases in which no arrest was made, 25 % reported confiscating drugs or paraphernalia.

The majority, 62 %, indicated the law would not change their behavior at a future overdose because they would not have arrested anyone at the scene of an overdose anyway. Smaller proportions indicated they would be less likely to arrest (14 %), did not know what they would do (20 %), or would continue to arrest people at the scene of an overdose (4 %).

Opinions of the drug possession immunity component of the law were somewhat more negative than positive with 20 % somewhat or strongly in support, 31 % neutral, and 45 % somewhat or strongly against the law among the 240 who answered the question with 4 % indicating they did not know what they thought. A total of 123 officers provided a response to the opinion question as well as the narrative follow-up question.

The most common type of response among those with a positive view of the law was that lifesaving is more important than a drug charge (12 of 28 positive responses). An example of a typical response was:

“Drug possession charges are not that serious; no use having someone not get medical attention over.”

The most common type of narrative response among those opposed to the law was that drug use is a criminal activity (28 of 63 negative responses), for example:

“Drug use is illegal and user shouldn’t be immune to prosecution because they OD [overdose].”

Neutral responses were relatively common and the most common response type (12 of 32 neutral responses) indicated ambivalence between saving a life and the victim’s illegal activity followed in frequency by a belief that the law will not change anything (8 of 32 neutral responses) exemplified by:

“I feel people should respect the law, but peoples’ lives are more important”

“I wasn’t enforcing drug laws at OD’s before the law.”

Opinion regarding the element of the law that allows potential overdose victims to possess and administer the opioid antidote naloxone was more evenly split with 28 % in support, 30 % neutral, and 34 % opposed. Of those who answered the initial question, 44 % (99 of 223) provided comments. Most of the positive comments (21 of 29 responses) indicated their reasoning simply as naloxone saves lives, a typical quote was:

“Don’t like drug use but it’s more an illness than a crime and it would be good to have something on hand that can assist with an OD.”

Half of the comments (49 of 99) were negative towards the naloxone component of the law, with the most common responses (18 of 49) stating that naloxone should be administered by medical professionals only. Other common responses were that naloxone enables drug use (11 of 49) and concerns that naloxone could be misused, abused, or have negative effects (8 of 49). Representative quotes include:

“Narcan [brand name of naloxone] might be administered incorrectly and should only be administered by MDs/Medics.”

“Would not want Narcan to be a safety net allowing greater use of drug.”

No single theme dominated the comments of those who responded “neutral.”

Note that for both questions, those who initially indicated a neutral response were less likely to provide written comments.

Paramedics’ Responses

Half of Seattle’s paramedics were surveyed (n = 28). All had previously been at the scene of an opioid overdose, 89 % in the prior year (Table 2). They reported that in their experience, police rarely or never arrested people who overdosed or others at the scene. Three quarters reported that police were always or usually at overdoses. Eighty-five percent of paramedics in King County indicated that from their perspective, it was important for police to be at overdose scenes to help protect their safety. Few, only 2 of 27 who answered the question, had heard of the “Good Samaritan overdose prevention” law.

Table 2 Paramedics’ survey responses
Full size table

Discussion

We found that police officers have limited awareness of the Good Samaritan law in Washington State, that they report attending overdoses primarily to ensure the safety of medical personnel, and that they rarely arrest overdose victims or witnesses, although they do confiscate drugs and related paraphernalia with some frequency. In an informal check of the validity of these findings, we compared the results to parallel surveys conducted with paramedics and found good concordance. Similarly, based on a 2011 survey at Seattle area syringe exchange programs with 355 heroin users, 62 % reported police presence at the most recent witnessed overdose to which paramedics responded; just one person reported to have been arrested.20 The majority of heroin users surveyed (88 %) indicated they were more likely to call 911 during a future overdose after being informed about the law.

While we focused on heroin users for this study, the majority of fatal opioid overdoses actually now involve pharmaceutical opioids (usually with other drugs) with the sources of the drugs, motivations for use, and legality of use often unclear.1,21 Interventions that include opioid overdose education, including local laws, may become even more important as indicators suggest a transition from pharmaceutical opioids to heroin in the research literature and mass media.22–24 Indicator data also point to recent increases in heroin use and among young adults in particular.2,25 These younger users may not have hardened perceptions of law enforcement, so messaging directed at them prior to their first encounter with law enforcement may be valuable.

Most police and paramedics surveyed believed it was important for police to be at the scene of an overdose to help ensure the safety of medical personnel. This finding is important in light of concerns expressed locally and in the research literature about the presence of police at the scene of an overdose.5,12 Importantly, just a third of police felt it was important to be at the scene of an overdose to enforce laws. Some may view this as still too high a proportion and raises the point that the immunity provision in Washington State is quite narrow, whereas several states passed legislation in 2013 with broader immunity such as New Jersey which provided protection from revocation of probation or parole. A quarter of police did confiscate drugs or paraphernalia when arrests were not made. The perceptions and motivations of all of the players at overdose scenes would ideally be made transparent in order to promote public health. There is evidence from another major US city that medic response to emergencies is delayed in overdose cases while medics wait for police to arrive, this may not be the case in Seattle which has a dense urban environment and very rapid response times by both police and paramedics.26 However, this issue is an important one for communities to grapple with in a proactive manner. While fear of police is often cited as a concern during overdoses, anecdote and some research also point to positive experiences between drug users and first responders at overdoses.6

Poor policy penetration and implementation gaps observed here have been identified in a previous research. It is not atypical for police officers to exhibit substantial gaps in knowledge of the law related to public health interventions.16 Similarly, research among drug users and other at-risk groups has identified pervasive lack of clarity about the law.27–30 Our findings underscore the critical role of policy evaluation to evidence-based tailoring of public health laws in order to guarantee their positive impact.

The results of this survey were utilized in public health efforts. Survey results were shared with Seattle Police Department commanders, county prosecutors, and health department staff. The police department training unit produced a training video that includes the narcotics commander introducing the law; the prosecutor describing the Good Samaritan component; and the medical director for public health describing the nature of opioid overdoses, how naloxone works, and the research evidence demonstrating the impact of distributing take-home-naloxone on public health. The video has been shown at Seattle Police Department roll calls, and the narcotics commander has also shared it at national law enforcement conferences. It may be viewed at www.stopoverdose.org. In addition, potential bystanders receiving overdose trainings in the Seattle area are informed that police have been trained on the Good Samaritan law and the legality of take-home-naloxone. At this stage, we do not know whether these efforts have led to a change in perceptions, 911 calls, or the rate of overdoses.

Engaging police is a critical component in a comprehensive strategy for curbing the US opioid overdose epidemic. Law enforcement activities, such as searches, questioning, and arrest at a scene of an overdose, represent an important, but by no means the only aspect of the possible role of police in overdose prevention. Given their routine contact with high-risk groups and central role in school-based and other drug education, police are also uniquely situated to promote opioid overdose awareness and response in the community. Law enforcement policies and outreach can encourage help-seeking during overdose events by mollifying concerns about legal repercussions to victims or witnesses.5 Emergency dispatchers—often police department employees—are uniquely positioned to instruct bystanders on lifesaving measures that can substantially reduce the risk of death of injury from opioid overdose before professional help arrives. Equally critical, police officers provide key security coverage for paramedics, as reflected in our findings.

This study is limited by several factors including the reliance upon self-report by respondents; however, we did triangulate information about arrest with paramedics and heroin users. We also surveyed half of each first responder type strengthening the representativeness of the results. This was an exploratory study, and therefore we conducted descriptive analyses. Generalizability is limited by the single city evaluated. Social, political, drug-using, and demographic factors might yield different results in other settings. Nonetheless, the approach to evaluating public health law may be readily transferable to other localities and healthcare topics.

Many opportunities for future research exist. Follow-up survey work with police to determine the degree and nature of the impact of the overdose training would be valuable. More substantial law enforcement training could be designed and evaluated. The impact of Good Samaritan laws on actual number and rates of 911 calls and related health outcomes is worth evaluating. However, we have found that in a time of dramatic drug policy changes as well as dramatic changes in heroin and pharmaceutical opioid use, separating out the impacts of a law may be difficult. As Good Samaritan and naloxone laws are increasingly passed by state legislators, some manner of mandating community education and program evaluation within existing agencies and funding streams would be valuable.

Conclusion

This study demonstrates that merely passing Good Samaritan laws encouraging help-seeking and lifesaving intervention during an overdose is perhaps necessary, but is decidedly insufficient. Funding for implementation has not been included in any states’ similar laws. National media have begun reporting on Good Samaritan laws and noted problems with implementation, including improper arrests of victims and help-seekers otherwise immune under the new provisions.31,32 We found that vocal support from state and local officials, open conversation, local data, and a common interest in healthy communities may help facilitate implementation. Public safety and public health need not be antagonistic, but rather can potentiate each other to manage this and other disease epidemics.