“Lock to Live”: Development of a firearm storage decision aid to enhance lethal means counseling and prevent suicide

Abstract

Objectives:

We sought to: 1) clarify decision needs among suicidal adults with home firearm access; 2) identify accurate, unbiased, and acceptable approaches for content and messaging; and 3) develop a tablet-delivered decision aid for firearm storage options.

Methods:

Following international standards, we used stakeholder interviews to develop a decision aid for the decision, “what option(s) to choose to reduce home access to firearms for an adult at risk of suicide.” Participants were adults with: personal or family history of suicidal ideation or behaviors; firearm ownership or employment in a firearm range or store; involvement in suicide prevention field; or work as emergency department or other healthcare provider.

Results:

Through 64 interviews, we created the “Lock to Live” decision aid, which includes: 1) introduction specifying the decision; 2) clarification of preferences and logistics; 3) table of storage options; and 4) summary with specific next steps. The final tool had high user acceptability.

Conclusions:

Should the “Lock to Live” decision aid prove useful in a pilot feasibility trial and subsequent testing, it could enhance lethal means counseling and help prevent firearm suicide.

INTRODUCTION

Suicide is a leading – and growing – cause of death in the United States,¹ with firearms used in half of suicides.² Interventions to reduce access to firearms and other highly lethal methods of suicide are recommended;^3–5 a core approach is lethal means counselling (LMC) to limit home access.

Unfortunately, prior work shows that LMC addressing firearm access does not routinely occur during the care of patients with suicidal thoughts or attempts in emergency departments (EDs).^6–8 In a multi-site study, only half of suicidal patients had medical record documentation that a provider had asked about lethal means access,⁸ although at least one third of US adults live in a home with at least one firearm.⁹ Potential barriers to LMC include time demands in EDs,¹⁰ clinician unfamiliarity with firearms, and concerns about alienating patients who may be protective of their right to own firearms – though data suggest patients are generally receptive to LMC.^11–13 To date, published adult trials in EDs have examined training or protocols including LMC^14,15 but none has involved focused materials on firearms.

Given both the importance of firearm LMC and impediments in EDs, it is critical to explore ways of increasing counseling feasibility and patient engagement. A potent model is decision aids (DAs): tools that facilitate complicated health-related decisions by identifying the decision to be made, describing risks and benefits of options, assisting the patient in clarifying personal values, and activating the patient for decision-making.¹⁶ DAs have been shown to increase patient knowledge, decrease decisional conflict, and positively affect patient-provider communication.¹⁷

Guided by DA theory and development standards, we used a multi-stakeholder, iterative, user-centered approach to simultaneously 1) clarify decision support needs among ED patients with suicide risk and home firearm access; 2) identify content and messaging approaches which are accurate, unbiased, and acceptable to many stakeholders; and 3) develop a tablet-delivered, encounter-based DA describing firearm storage options.

METHODS

Study Design and Participants

As customary in DA development informed by international standards,¹⁸ our design began with theory. The Ottawa Decision Support Framework¹⁹ holds that meeting decisional needs (e.g., knowledge, conflict/uncertainty, and values) improves decision quality (i.e., a decision concurrently informed and reflective of the patient’s values). Simultaneously, we considered how to frame options (methods of temporary firearm storage) within a behavior change context. We followed principles of behavioral “nudging”,²⁰ seeking to present the most effortful options alongside easier options. This approach subtly “nudges” patients towards behavior change without obviating patient agency, as the comparison makes options seem more desirable than if they were presented by themselves.²¹

Our development team comprised individuals with varied clinical and research expertise, including an emergency physician, a geriatrician, social workers, and researchers in injury and suicide prevention, public health, decision science, and dissemination & implementation. One team member had a firearm at home, and our advisory panel included firearm retailers and instructors. We intended to design a DA for use with ED patients with acute suicide risk, with healthcare providers providing the DA as part of lethal means counseling. However, we hoped the DA could also be used outpatient clinical settings, at home, or elsewhere, and our recruitment and development processes reflected these goals.

We sought input from a range of stakeholder groups, guided by the theory that a DA built using language, visual messaging, and other design considerations consistent with the perspectives of diverse stakeholders would be most acceptable.²² Participants included adults (≥ 18 years) in the US without current suicidal ideation who were in at least one stakeholder group: 1) individuals with prior personal or family suicide ideation or attempts; 2) firearm owners or representatives from firearm groups, stores or ranges; 3) healthcare providers who care for suicidal patients (including in EDs); or 4) researchers, practitioners, other leaders in firearm suicide prevention. Participants were recruited through posted flyers (including in gun shops and EDs), emails, and advertisements on Twitter and Facebook (including Facebook groups for ED providers across the US). Participants received a $25 gift card. The study protocol was approved by the Colorado Multiple Institutional Review Board.

Development

Methods used to develop the LMC DA were adapted from those used previously by team members.^16,23 We drafted the initial DA based on international criteria¹⁸ and existing knowledge of LMC, with six iterative drafts before interviewing participants. These early drafts included sections typically found in DAs, customized for the decision of “which options to choose to reduce home access to firearms” in times of suicide risk: 1) introduction specifying the decision; 2) presentation of options, with pros and cons of each; 3) clarification of preferences, logistics and other issues to considered; 4) specific next steps to encourage and activate behavior change.

Early versions of the DA were on paper to allow easy adjustments of language, basic visual themes, and flow. The research team met with Cactus (a professional design firm) early in development to develop common goals for the DA based on decision science and user experience in a digital interface. Graphic designers at the firm created artwork and messaging consistent with written content, intending to create an appealing yet concise and intuitive user experience.

The one-on-one, semi-structured, confidential interviews were held online or in person. Participants were emailed study materials, including the latest DA draft and informed consent information, prior to the interview. Interviews lasted 30–45 minutes and were recorded for later transcription and de-identification. Interviews explored decision needs and feedback on iterative DA versions, including prior personal experiences and opinion of the DA’s ability to influence an adult with suicide risk. After the interview, participants completed an online survey about their demographics and the DA’s overall acceptability.²⁴

Following each conversation, interviewers wrote a de-identified reflective note covering the content of the interview and reactions to the DA. The development team discussed these notes during weekly meetings and maintained an audit document summarizing feedback and any changes made in response; when participant opinions conflicted, the development team considered feedback and made changes based on team consensus.

Once feedback about the basic content and flow of the DA was generally consistently positive across participants, Cactus transformed the paper version into a wireframe with basic web functionality, and continued interviews allowed feedback on the wireframe. The final version of the DA had a Flesch-Kincaid reading level of approximately 6^th to 7^th grade, and the final version was sent to all participants for acceptability testing using the Ottawa Acceptability Questionnaire.²⁴

RESULTS

Participants

Over the course of development, 29 DA versions were iteratively reviewed in interviews with 64 participants (Table 1). Nearly half of participants had lived experience with suicide ideation or attempts, and one third had experience with firearms (as owners or affiliation with a range or store). Additional review came through presentations to a patient advisory group,²⁵ to a state coalition of firearm suicide prevention experts, to the study team’s external advisory panel, and at an international suicide prevention conference.

Table 1.

Characteristics of participants (n=64 individuals)

Characteristic	n (%)
Age (median)	44 (IQR 35–52; range 23–70)
Stakeholder group affiliation (≥1 allowed)
Personal history of suicidal thoughts or attempt	14 (22)
Family member or friend had suicidal thoughts, attempt, or death	25 (39)
Firearm owner or enthusiast*	18 (28)
Work at/with firearm retailer, range, or organization	5 (8)
Work in suicide prevention (including volunteering)	37 (58)
Emergency department provider	9 (14)
Other healthcare provider	13 (20)
Work/affiliated with VA or other veteran service provider	8 (13)
Current or former law enforcement officer	4 (6)
Gender
Female	29 (45)
Male	27 (42)
Missing	8 (13)
Race (≥1 allowed)
White	49 (77)
African American	5 (8)
American Indian/Alaska Native	1 (2)
Asian	1 (2)
Hispanic	8 (13)
Veteran	9 (14)
Residence and employment location
Live in mostly rural area	9 (14)
Work in mostly rural area	9 (14)
Live in mostly urban area	40 (63)
Work in mostly urban area	38 (59)
Not currently working	2 (3)
Missing	15 (23)

^*Self-identified (e.g., participation in hunting, recreational target shooting, or advocacy for firearm rights, or employment at a firearm range/retail, regardless of personal firearm ownership)

Development

After reviewing early interviews from the first DA draft, we chose to focus the DA on adults with suicide risk rather than also on parents of adolescents with suicide risk, which would require different messaging according to our respondents. We also chose to make the DA usable by either individuals at risk or their families or friends rather than having separate versions. The DA title – “Lock to Live” – emerged from specific feedback from participants, including comparison with alternate titles.

The DA begins with information about reducing firearm access during times of suicide risk, moving to consideration of values and presentation of the decision options. Based on feedback on early drafts we switched the order of two sections, presenting the questions for values clarification first, followed by a listing of options (Table 2). Firearm storage options vary by location (in- versus out-of-home), type of locking device, cost, and legal constraints (e.g., background check requirements for firearm transfers). A goal of our DA is to facilitate a decision informed by both consideration of personal values impacting the choice and an understanding of the pros and cons of each option.¹⁹ After iterative modification, the four final issues presented for consideration are: 1) identification of “someone you trust to help you with safe storage”; 2) openness to storing firearms away from the home; 3) cost constraints; and 4) openness to options requiring a background check. Due to privacy considerations (below), individuals’ responses to these questions are not saved and do not alter the rest of the DA, such that all participants view the same full table of options. The questions are intended, instead, to prompt consideration of factors likely affecting storage choices.

Table 2.

“Lock to Live” decision aid sections, with key elements

	Imagery	Messaging	Key points	Rationale
Introduction	Hands with key	Validation of mental distress as common experience Hope and encouragement	Identification of decision (“temporary firearm storage when there is a risk of suicide”)	Engage and encourage tool use
Background	Man on bench	Temporary nature of “tough times”; small changes can make a big difference	Rationale for reducing firearm access (including statistics)	Many people do not know about safe storage as a method of suicide prevention and may feel the decision aid is “picking on” guns
Issues #1	Man with caring people	“You don’t have to do this alone” “Who is someone you trust to help you with safe storage?”	Encourages identification of other(s) who can help (requires selection to move on)	Safest if someone else locks or moves guns
Issues #2	Home and store	“How open are you to storing your firearms temporarily with someone else, away from your home?”	There are many options for storage	Some owners will not be open to any out-of-home options (e.g., from concern about not getting firearms back, or concern about home protection of family)
Issues #3	Price tag	“When looking for storage options, how concerned are you about cost?”	Varying costs for storage options	Storage costs range from free to expensive; encouraging unrealistic options may be counterproductive
Issues #4	Paper with check	“How open are you to storage options that involve a background check?”	Some options require background checks	Some owners will not be open to options that require background checks
Table of options	Expandable rows	Emphasizes choice, normalizes “storage options used by other firearm owners in your situation”	Displays out-of-home and at-home options with details, relative cost, and check-box	Bottom of page includes representative stories from other firearm owners to help with decision-making
Name contact	None	Encourages identification of someone who can help	Reiterates usefulness of having help; does not require input to move on	Safest if someone else locks or moves guns; also helpful if someone else helps with logistical details
Summary	Displays prior selections	Summarizes and praises effort	Lists crisis hotlines and option to print summary	Provides printable summary for individual to take home as reference
FAQ	At top of page	Responds to common questions	Addresses questions about other methods of suicide, laws, how to start conversations, and other issues	Allows further explanation while keeping text sparse on core pages

Storage options themselves are presented in a table displaying the option and its relative benefits and drawbacks. Separate tables for “Out of Home” and “In Home” storage methods (Figure 2) have rows allowing individuals to click and expand options. Each method includes a cost approximation (“free” to “$ $ $”). Guidelines recommend removing firearms from the home in times of suicide risk over locked storage at home,³ so we present “Out of Home” storage first. While participants generally felt individuals would not choose to store firearms with police, they agreed with including it as an option. Language in the table, and throughout the DA, intentionally is generalizable to any state and to any clinical or community setting. DA users can select multiple options, and those selected are displayed on the final screen, which can be printed for later reference. The final screen also includes empowering language and praise for working through a hard decision, along with suicide hotline numbers.

Figure 2.

Screen shot of table of options from decision aid, with expansion of sample rows

Acceptability

Twenty-three participants (36% of original sample) responded for a second $25 incentive. Self-identified affiliations included: lived experience with suicidal thoughts or an attempt personally (26%) or as a family member (30%); firearm ownership (26%); work in suicide prevention (61%); work in healthcare (44%); veteran or VA-affiliated (8.7%); and law enforcement (4.3%).

Overall acceptability was high: most reported the DA contained the right amount of information (87%) and was clear on most (52%) or all (48%) aspects. Most (78%) thought the DA presented options in a balanced way, while the other five (22%) thought it was “a little slanted towards out of home storage.” All participants rated the DA as either somewhat (52%) or very (48%) helpful; 87% said they would definitely recommend the DA to someone facing this decision, with the remaining 13% reporting they probably would. Participant feedback on the images and overall product were positive; as one said, “I think that it’s really professional, which I like because … the assumption is that it’s coming from somebody who knows what they’re talking about.”

Novel Development Issues

While the DA development process followed theory and international guidelines, our topic and approach generated novel considerations (Table 3).

Table 3.

Novel development issues, with representative quotes

Topic	Sample participant quote
Interface of decision support and behavior change	“So I think that this is the opportunity. If this is an emergency room and they’ve gotten that far, now is a great opportunity to get that person that they trust to say, ‘Would you be willing to have an open conversation with them and put a pact [for firearm storage] together?’ And then they need to be educated. ‘Could we do that?’ I’ll bet 80 percent of ‘em will say, ‘Yes, let’s talk to them.’ ”
Desire for language of hope and validation	“ ‘You support other people and take care of them. Now it’s time to take care of yourself and let folks help you,’ kind of thing or something that lets them know that the shoe can be on the other foot and it’s okay.”
Privacy concerns	“And it’s hard, now, ‘cause it’s such a prevalent part of gun control discussion, you know, people with serious psychiatric problems are the target of, like, whenever there’s a mass shooting or something, so people really have this ingrained belief on that.”

Interface of decision support and behavior change

Most DAs are grounded in the idea of equipoise among available options. In these frameworks, a “good” decision is one which is both informed and reflective of patients’ values and preferences for care.¹⁹ In contrast, behavior change theories are typically prescriptive, where a “good” decision is clearly linked to the best clinical outcome. Our DA was developed with consistent awareness of this tension: we wished to develop a DA that prompted patients and families to consider a variety of storage options within their own personal context and preferences, but that also subtly promoted options to make firearms least accessible to the person at risk. Participants also identified the ED setting as one where change might be more likely, especially when engaging family or friends (Table 3).

Desire for language of hope and validation

Patient education materials typically do not purposefully incorporate emotional expression. In our case, participants consistently reported a need to promote hope and compassion for those contemplating suicide. Many viewed hopeful, validating messaging as key to suicide intervention processes, as well as necessary to engaging patients in crisis. We included visual and text elements normalizing the emotional experience of being in crisis and emphasizing a hopeful narrative in which improved temporary firearm storage results in long-term safety and stability. Examples included imagery of caring friends or family (Figure 1) and language like, “Many people have gotten through times like this, and you can too.”

Figure 1.

Screen shot from decision aid demonstrating elements to convey confidentiality (footer language), hope, and validation of feelings

Privacy concerns

Confidentiality, especially as it related to recording gun ownership information in the electronic medical record, was of considerable importance to a variety of participants, with firearm owners citing credible protection of patient confidentiality as a prerequisite to any level of acceptability among firearm-owning patients. Prior limited work suggested some patients might be less willing to discuss firearm access if the information is recorded in the medical record, out of concern such information would lead to firearm confiscation;¹² of note, no federal or state law currently prohibits providers from counseling patients about firearms.²⁶ Concerns related to credibility in protecting confidentiality led to a number of design considerations. We limited the data-gathering capability of the web-based DA itself, including: 1) reducing the scope and type of data captured by the tool, such that individual-level analytics are not available; 2) not pursuing functionality to email the storage plan document to the patient or other requested recipients; and 3) presenting storage options identically across administrations, rather than individualizing presentation to the expressed values, preferences, and needs to each user (as individualized presentations would require stored data). At the bottom of each page, the DA included confidentiality language (“We’re not storing any of your personal information”) to reassure participants.

DISCUSSION

We successfully developed a web-based decision aid for “what option(s) to choose to reduce home access to firearms” for an adult at risk of suicide. Through interviews with participant from stakeholder groups - firearm owners, clinicians who would use the DA, family or friends of individuals with suicide risk, and the suicide prevention field - and through collaboration with a professional design firm, we identified key issues pertinent to the decision and created a tool ready for pilot testing. Our process was grounded in theory and followed international development standards, yet we also identified unique issues related to our subject matter. Major themes arising illustrated the need for a DA that incorporated both messages of hope and educational information and that was nonjudgmental, trustworthy, accurate, and concise yet thorough.

Our work contributes to the knowledge base concerning DA development, as we faced several unique challenges stemming from this decision context. First, the interface of behavior change and shared decision making is novel. Others have wrestled with related issues – for example, how to encourage colon cancer screening while informing patients of the available screening options.²⁷ But in our case, some individuals may not have yet recognized there is even a decision to be made, increasing the challenge of both educating and motivating them through the DA. Additionally, DAs do not routinely address the costs involved in the options, largely due to both the logistical (changes in cost) and ethical challenges that this can create for clinicians.^28,29 We handled costs using dollar signs to acknowledge the importance of cost without overwhelming a conversation with cost specifics likely to change with time and context.

Specific feedback also called to attention benefits of coupling firearm storage with medication storage. Participants across our stakeholder groups identified the need to comprehensively address access to lethal means. In this line of logic, any conversation about reducing access to lethal means creates an opportunity to improve home safety while an individual is in crisis. Firearm owners in our sample also consistently indicated that any effort to highlight the specific lethality of firearms would make the DA less acceptable to the subset of firearm owners who see any discussion about firearms within the context of healthcare as part of a hidden gun control agenda.¹² We addressed the issue of other lethal means through language (e.g., “Keeping dangerous items, like firearms, out of reach during this time can save a life”) and specific information under the FAQ section; a future parallel module focused on medications may be useful.

Relatedly, considerations regarding tailoring to the specific audience of interest featured prominently in our design process. Suggestions regarding the use of language or imagery associated with the culture of gun ownership were made both by participants and members of the design team (including “patriotic” imagery, specific use of language highlighting “freedom” and “rights”, and co-opting messaging from firearm groups). Previous efforts to develop public health messaging for safe storage of firearms advocated cultural appropriation as a method to improve acceptability among firearm owners,³⁰ echoing a more general debate about tailoring to specific demographic groups which occurs in the development of other decision aids.^31,32 We elected to avoid this imagery, out of concern we might alienate any individuals who would might feel those messages were insincere – or worse, represented a caricature. This may be especially true when the DA is used within an ED (and with individuals who have already been identified as having guns at home), but it may be that messaging intended for varied geographic settings or community versus healthcare settings would warrant a different approach.³³

Limitations include participation bias; at least one potential participant declined an interview because he worried the recording would be used for gun control advocacy. We did assure participants of confidentiality, recruit from a range of settings, and use snowball sampling. A second limitation is that interviewees were predominantly white. While data show that both firearm ownership³⁴ and firearm suicide² are more common among whites than other demographic groups, the specific views and needs of diverse communities should be addressed to determine acceptability across racial and ethnic groups. Finally, the current version of the DA is intended specifically for use with adults, though feedback was clear on the need for a similar tool for guardians or parents of children or adolescents with suicide risk. Our group is collaborating to develop such a tool.

CONCLUSION

This is the first study to create a decision support tool specifically for suicidal patients for firearm storage to augment lethal means counseling. The “Lock to Live” decision aid could help support both the patients’ autonomy and the clinicians in having these conversations – and, ultimately, help prevent firearm suicide. A pilot randomized controlled trial (NCT03478501) will test the DA’s acceptability and feasibility of use in three EDs, with plans for a subsequent trial in real-world ED settings. Future work should clarify possible settings for DA use and how it might be tailored for populations based on age, suicide risk (e.g., chronic depression versus acute suicide ideation), or other characteristics.

What is already known on this subject:

• For individuals at risk of suicide, reducing access to firearm and other lethal means can save lives.

• Though recommended, counseling by healthcare providers about reducing access to lethal means often does not occur.

• There are many ways to reduce firearm access in times of suicide risk – deciding which ones to use can be difficult.

What this study adds:

• We developed a web-based decision aid for “what option(s) to choose to reduce home access to firearms for an adult at risk of suicide.” The final tool had high acceptability.

• A patient-centered decision aid related to firearm storage options may address barriers and enhance counseling and subsequent behavior change.