Archives of Suicide Research, 24:S260–S279, 2020

ISSN: 1381-1118 print/1543-6136 online

DOI: 10.1080/13811118.2019.1572555

Possession of Household
Firearms and Firearm-Related
Discussions with Clinicians
Among Veterans Receiving
VA Mental Health Care
Marcia Valenstein, Heather Walters, Paul Nelson Pfeiffer,
Dara Ganoczy, Mark Andrew Ilgen, Matthew Jason
Miller, Matthew Fiorillo, and Robert M. Bossarte

Objectives: To assess possession of household firearms among veterans

receiving mental health care and the frequency of their discussions with
clinicians about firearms. Methods: We surveyed random samples of veter-
ans receiving mental health care in each of five purposively chosen, geo-
graphically diverse VA facilities; 677 (50% of recipients) responded.
Results: 45.3% (95% CI 41.2, 49.3) of veteran respondents reported
household firearms; 46.9% of those with suicidal thoughts and 55.6%
with a suicide plan had household firearms. Only 27.5% of all veteran
respondents and 44% of those with recent suicidal ideation and household
firearms had had a firearm-related discussion with a clinician.
Discussion: Many veterans receiving mental health care can readily access
firearms, a highly lethal means for suicide. Increasing clinician-patient
discussions and health system efforts to reduce firearm access might reduce
suicide in this clinical population.
Keywords firearms, veterans, mental health providers, suicide risk

Suicide is the tenth leading cause of death 2017). Several, although not all, studies
and an important public health concern reported higher rates of suicide among com-
(CDC/National Center for Health Statistics, munity veterans than the general popula-
tion, with risk varying based on the specific
veteran population (Miller et al., 2009;
This work was authored as part of the
Contributor's official duties as an Employee of the
Miller et al., 2012; VHA Office of Suicide
United States Government and is therefore a work Prevention, 2016; Kang et al., 2015). The
of the United States Government. In accordance most recent estimate, in 2014, was that sui-
with 17 U.S.C. 105, no copyright protection is cide risk is 22% higher among veterans
available for such works under U.S. Law.

S260
 M. Valenstein et al.

compared with adults in the general U.S. calls have been shown to reduce suicidal
population after adjustment for age and sex behaviors among emergency department
(VHA Office of Suicide Prevention, 2016). users, including veteran users (B. Stanley
Younger veterans and veterans returning et al., 2018a). The Veterans Health
from recent conflicts appear to be at particu- Administration (VHA) has made systematic
lar risk (Kang et al., 2015; VHA Office of efforts to address firearm-related suicide
Suicide Prevention, 2016). Mental health risk. In many U.S. Department of Veterans
disorders are a potent risk factor for suicide Affairs (VA) facilities, suicide assessment
for veterans and for the general population protocols prompt clinicians to consider fire-
(Inskip, Harris, & Barraclough, 1998; Ilgen arm availability, and clinicians are encour-
et al., 2010), and suicide rates are higher aged to use suicide safety plans that address
among those receiving mental health care the availability of lethal means and ways
(Simon & VonKorff, 1998; Valenstein that the environment can be made safer
et al., 2009). (B. Stanley & Brown, 2008). The VA
Veterans predominantly used firearms Office of the Medical Inspector and the VA
as the means for suicide; 68% of male vet- Geriatrics and Extended Care Group have
eran suicide deaths and 41% of female vet- recommended that veterans with dementia
eran suicide deaths involved firearms (VHA be screened for access to firearms and that
Office of Suicide Prevention, 2016). This is the veterans and their family members
higher than in the general population where, receive firearm safety information (U.S.
in 2016, 57% of male and 31% of female Department of Veterans Affairs (VA)
suicide deaths involved firearms (Centers for Veterans Health Administration (VHA)
Disease Control Office of Statistics and Dementia Steering Committee, 2016). The
Programming, n.d.). Potentially contribu- Veteran Family Firearm Safety Program
ting to the disproportionate toll of firearm provides free cable gunlocks to veterans and
suicides among veterans, recent studies their families to address firearm-related
reported that approximately half (45%) of safety (Nalpathanchil, 2013).
all veterans are firearm owners (47% of Yet despite these efforts, VA patients
male veterans; 24% of female veterans) continue to have high rates of firearm-
(Cleveland, Azrael, Simonetti, & Miller, related suicide, suggesting that additional,
2017), compared with 22% of the general more effective, interventions or dissemin-
population (34% of males and 12% of ation efforts are needed to reduce firearm-
females) (Azreal, Hepburn, Hemenway, & related suicide. Critical to developing more
Miller, 2017). effective interventions that promote firearm
Voluntarily reducing firearm access safety is further information regarding the
among veterans in mental health care may prevalence and storage of firearms in this
be a practical and important avenue for clinical population, the reasons for keeping
reducing suicide risk in this clinical popula- firearms in the home, and the conversations
tion. One of the few suicide prevention that clinicians have with their at-risk patients
strategies with reasonable evidence for about firearm safety. However, to date, even
effectiveness is reducing access to lethal basic information about how often clinicians
means (Mann et al., 2005; Bagley, Munjas, discuss firearm safety with high-risk patients
& Shekelle, 2010), and suicide safety plans in VA mental health care is unknown.
that included lethal means assessment and There have been a few studies of fire-
counseling in conjunction with follow-up arms possession among a nationally

ARCHIVES OF SUICIDE RESEARCH S261

 Household Firearms and Related Discussions with Clinicians

representative sample of veterans in the One study reported that 32.3% of veterans
community, including veterans who self- seeking VA post-traumatic stress disorder
reported mental health conditions and (PTSD) residential care at two program
who self-reported some VA health care use sites owned a firearm (P. N. Smith et al.,
(23% reported VHA use) (Simonetti, 2015). Ascertaining firearm availability in
Azrael, & Miller, 2018; Cleveland et al., a broader sample of veterans in active VA
2017). Simonetti et al. found that one in mental health care and those who had a
three U.S. veteran firearm owners in their recent suicidal ideation or plan is likely to
community sample stored household fire- be helpful in developing interventions to
arms loaded and unlocked, and that the reduce risk.
prevalence of this storage practice was Understanding the frequency of
similar among those with and without self- patient-clinician discussions regarding fire-
reported suicide risk factors (i.e., self- arm access is also key to designing strategies
report of a mental health diagnosis) that mitigate risk, as clinician-patient dis-
(Simonetti, Azrael, & Miller, 2018). cussions are a likely component of any
Simonetti et al. also reported that a non- voluntary intervention to reduce firearm
significantly greater proportion of the vet- access. Yet little is known about the fre-
erans with firearms who also reported quency of such discussions in VA mental
VHA use stored a firearm locked and health care settings, despite the provision of
unloaded compared to those who did not suicide assessment protocols and safety
report VHA use (38.6% vs. 30.8%). In a planning templates that might encourage
second study using this sample, Simonetti these discussions. A few studies in commu-
et al. confirmed the above findings, and nity settings indicated that patient-
reported that having a firearm that was clinician discussions about firearm-related
loaded and unlocked was more common safety may be infrequent, even when
among community veterans who cited pro- patients are at high risk for suicide (Betz
tection as the reason they owned firearms, et al., 2016; Traylor, Price, Telljohann,
who owned more firearms, and who, King, & Thompson, 2010). Primary care
among other perceptions, thought that clinicians documented discussions about
having a firearm made people in the home firearm access for a small minority (15%) of
safer (Simonetti, Azrael, Rowhani-Rahbar, their VA primary care patients with positive
& Miller, 2018). suicide risk assessments and documented
However, to date, there have been few recommendations to reduce firearm access
studies of the prevalence of household fire- for only 2% (Dobscha et al., 2014).
arms or firearm storage practices among Additional data that may be helpful
the high-risk patients enrolled in and for clinicians and policymakers when
actively using VA mental health care, a addressing firearm-related suicide is the
population where clinicians and health sys- prevalence of multiple firearms and
tems can intervene to address firearm- handguns in the veterans’ households.
related safety. Prior studies of firearms and Securing multiple firearms may present
veterans actively receiving mental health logistical issues and require additional
services have been small or limited to a sin- planning by clinicians and patients, and
gle diagnosis or specialized setting (Heinz, handguns pose greater suicide risks than
Cohen, Holleran, Alvarez, & Bonn-Miller, long guns (Conwell et al., 2002). Prior
2016; Smith, Currier, & Drescher, 2015). studies on firearm owners in the general

S262 VOLUME 24
NUMBER S1
2020

 M. Valenstein et al.

population and in a nationally represen- on firearm-related safety practices among

tative sample of veteran firearm owners VA patients (Walters et al., 2012;
reported high rates of multiple gun own- Valenstein, Walters, Ganoczy, & Ilgen,
ership and handguns (Azreal et al., 2012). The study survey contained items
2017; Cleveland et al., 2017), suggesting addressing domains of firearm availability
this may be an issue for veterans in VA and storage; mental health symptoms;
mental health care. risky drinking; suicidal ideation, plans, and
Assessing reasons for maintaining fire- attempts; reasons for firearm access; and
arm availability is also likely to be import- interactions with clinicians regarding sui-
ant in efforts to voluntarily reduce firearm cide and firearms.
availability. The majority of firearm own- We distributed the survey by mail to a
ers in the United States and the majority total of 1,500 VHA patients, with 300
of veteran firearm owners in the commu- VHA patients being randomly selected
nity (both 63%) reported that protection from each of five VA facilities and their
against others is one of the primary reasons associated clinics. For patients to be eligible,
for ownership (Cleveland et al., 2017). VA administrative data had to indicate that
There were similar levels of concern about patients had used mental health, PTSD, or
a need for protection among veteran substance use services on at least two occa-
firearm owners with and without suicide sions in the prior 12 months. The VHAs
risk factors (Simonetti, Azrael, & Miller, were purposefully selected for geographic
2018). If high percentages of VA mental diversity, with six VA facilities approached
health patients maintain household fire- before five agreed to participate. The five
arms for protection, their concerns regard- study VA facilities were located in the fol-
ing security will need to be systematically lowing regions: Pacific, South Atlantic,
considered by clinicians in their discus- South Central, West Mountain South, and
sions about reducing firearm access. West Mountain North.
To collect data to address these issues Surveys were fielded between May 11
in an active clinical patient population, we and October 19, 2015, using a modified
surveyed random samples of VA patients Dillman method (Dillman, Smyth, &
who had used outpatient mental health Christian, 2009). Patients received a pre-
care services in the prior 12 months within notification letter, followed by a mailed
each of five geographically diverse VA facili- survey that included a $25 Amazon gift
ties and their associated community out- card as an incentive. As needed, patients
patient clinics. We queried veterans about received up to two additional mailings. Of
firearm possession, storage practices, risk the 1,500 mailed surveys, 146 mailings
factors for suicide, and discussions about were returned as undeliverable. Of the
firearms with their clinicians. The goal was 1,354 patients who may have received
to inform efforts to reduce firearm-related mailings, 677 surveys were returned, con-
suicide in this high-risk clinical population. stituting a 50% response rate.
In the multivariate analysis, nonres-
pondents to the survey mailings did not
METHODS differ significantly from respondents by sex
or by having a service connection of 50%
This study fielded a survey that was based or more. Service connection is based on
on prior qualitative and quantitative work disabilities resulting from diseases or

ARCHIVES OF SUICIDE RESEARCH S263

 Household Firearms and Related Discussions with Clinicians

injuries incurred or aggravated during check just one response. Responses

military service, with the level of service included: hunting or sport, protection,
connection reflecting limitation of work work, or some other reason.
time from disabilities. A veteran who is Suicide-related variables were ascer-
50% or more service connected has sub- tained using the following questions with
stantial disability and is eligible for many yes/no responses from the National
VA services. Comorbidity Survey Replication study
Individuals who received care in the (NCS-R): “Have you ever seriously
West Mountain South area were less likely thought about committing suicide?” “Have
to return surveys than those in the Pacific you seriously thought about committing
area (OR 0.62, 95% CI 0.44, 0.87). suicide in the past 12 months?” “Have you
Patients aged 45–64 years and those 65 ever made a plan for committing suicide?”
years or older were more likely to respond “Have you made a plan for committing
than patients younger than 45 years suicide in the past 12 months?” and “Have
(OR ¼ 1.73, 95% CI 1.31, 2.30; and you ever attempted suicide?”
OR ¼ 1.92, 95% CI 1.42, 2.59, Interactions with clinicians regarding
respectively). suicidal thoughts and firearm access were
ascertained using the following questions
with yes/no response options: “Have your
Measures clinicians ever asked whether you’ve had
suicidal thoughts?” and “Have you ever
For the primary outcome of house- talked with your clinicians about your
hold firearm access, survey recipients guns/access to guns as part of your health
were asked: “How many firearms or mental health treatment?”
(excluding BB guns) that are in working Since symptoms of patients in treat-
condition do you have at home, includ- ment for mental health conditions may
ing handguns, rifles, and shotguns now wax and wane, several mental health symp-
kept in or around your home?” tom measures were used to assess whether
Respondents were considered to have patients’ current symptoms were at clinic-
household firearms if they indicated they ally significant levels. Traditional cut
had one or more working firearms in or points of these scales were used to define
around their home. They were consid- clinically significant symptoms.
ered to have access to multiple firearms The severity of current depressive
if they reported having two or more fire- symptoms was ascertained using the
arms in or around their home. Patient Health Questionnaire-Eight
Survey recipients were asked: “Are any (PHQ-8), which has eight of the nine
of the guns at home handguns, such as pis- Diagnostic and Statistical Manual of Mental
tols or revolvers?” and “Is there a firearm Disorders (DSM–IV) criteria for major
in or around your home that is currently depressive disorder. The PHQ-8 has simi-
loaded and unlocked?” with yes responses lar psychometric properties to the widely
indicating the presence of a handgun or a used PHQ-9 and similar suggested cutoff
loaded and unlocked firearm, respectively. points. (Kroenke & Spitzer, 2002).
Survey recipients were asked: “What is Respondents who scored 10 were consid-
the main reason that there are guns in or ered to have clinically significant depressive
around your home?” and instructed to symptoms.

S264 VOLUME 24
NUMBER S1
2020

 M. Valenstein et al.

The severity of current PTSD symp- RESULTS

toms was ascertained using the PTSD
Checklist (PCL), a self-report measure Survey respondents answering firearm
assessing 17 DSM–IV symptoms of PTSD. questions reflected the demographics of
(Bliese et al., 2008). Respondents with the Veterans Affairs patient population,
scores 50 were considered to have signifi- with the majority being male (86.0%,
cant PTSD symptoms. 95% CI 83.1, 88.8) and white (63.6%,
An indicator for having either signifi- 95% CI 59.6, 67.6). Veterans in the sam-
cant PTSD or depression symptoms was ple were randomly chosen based on their
created and used in the multivari- use of mental health or substance use serv-
ate models. ices in the prior year, and they commonly
Risky drinking was assessed using the had significant mental health symptoms or
Alcohol Use Disorders Identification Test- risky drinking at the time of the survey
Consumption (AUDIT-C), which has (Online Tables 1 and 2).
been demonstrated to have high levels of Approximately,19.2% (95% CI 16.0,
internal consistency and test–retest reliabil- 22.5) of respondents reported suicidal
ity (Bush, Kivlahan, McDonell, Fihn, & ideation and 7.8% (95% CI 5.6, 10.0)
Bradley, 1998). We used cut point scores reported having a suicide plan within the
of three for women and four for men to past year. Approximately 19.9% (95% CI
indicate current risky drinking. 16.6, 23.1) reported at least one suicide
attempt in their lifetime.
Data Analysis Of the 590 respondents answering the
household firearms question, 267 (45.3%,
We calculated descriptive statistics for 95% CI 41.2, 49.3) reported having at
all survey items, using frequencies least one household firearm. Fully 46.9%
and means. (95% CI 37.6, 56.1) of those with suicidal
We conducted bivariate analyses for ideation in the past 12 months reported
patient report of household firearms and having a firearm in the home, as did
respondent demographic and clinical fac- 55.6% (95% CI 41.0, 70.1) of those with
tors, using Wilcoxon rank sum tests for a suicide plan in the past 12 months and
continuous variables and chi-square tests 38.6% (95% CI 29.7, 47.5) with a past
for dichotomous or categorical variables. A suicide attempt.
multivariate analysis assessed associations Among those with firearms, 73.4%
between the dependent variable of “has (95% CI 68.1,78.7) reported that they
household firearms” and respondents’ had more than one firearm in their home,
demographic and clinical variables. with a median of three firearms; 83.1%
Two additional multivariate analyses (95% CI 78.6, 87.6) indicated that they
assessed associations between the depend- had a handgun. Fully 38.5% (95% CI,
ent variables of “talking to a clinician 32.5, 44.4) of veterans with household
about suicide” and “talking with a clin- firearms responded that they had a loaded
ician about firearm access” and respond- and unlocked firearm in their home.
ents’ demographic and clinical variables. In bivariate analyses (Tables 1 and 2),
All statistical analyses were completed higher-income respondents, those who
using SAS software, Version 9.4 (SAS Inc., were married or who were living with
Cary, North Carolina). others, and Operation Enduring Freedom/

ARCHIVES OF SUICIDE RESEARCH S265

 Household Firearms and Related Discussions with Clinicians

TABLE 1. Veterans in VA Mental Health Care: Respondent Characteristics and Household

Firearms (N 5 590b)

Has Household Firearms

Characteristic (%, 95% CI)a P value
All veteran respondents 45.3 (95% CI 41.2, 49.3)
Sex (n ¼ 563a) 0.18
Male (n ¼ 484) 47.3 (95% CI 42.9, 51.8)
Female (n ¼ 79) 39.2 (95% CI 28.5, 50.1)
Age group (n ¼ 567a) 0.08
<45 years old (n ¼ 108) 51.9 (95% CI 42.4, 61.3)
45–64 years old (n ¼ 245) 40.4 (95% CI 34.3, 46.6)
65 years old (n ¼ 214) 48.6 (95% CI 41.9, 55.3)
Marital status (n ¼ 562a) 0.01
Married/cohabitating (n ¼ 323) 51.7 (95% CI 46.3, 57.2)
Divorced/separated (n ¼ 134) 41.8 (95% CI 33.4, 50.1)
Widowed (n ¼ 28) 25.0 (95% CI 0.09, 41.0)
Single (n ¼ 77) 39.0 (95% CI 28.1, 49.9)
Race/ethnicity (n ¼ 558a) 0.06
White (n ¼ 355) 49.9 (95% CI 44.7, 55.1)
Black (n ¼ 79) 40.5 (95% CI 29.7, 51.3)
Latino/Hispanic (n ¼ 66) 33.3 (95% CI 21.9, 44.7)
Other (n ¼ 58) 44.8 (95% CI 32.0, 57.6)
Education (n ¼ 554a) 0.63
Some high school (n ¼ 21) 33.3 (95% CI 13.2, 53.5)
High School diploma (n ¼ 125) 44.0 (95% CI 35.3, 52.7)
Some college/associate’s degree (n ¼ 272) 47.1(95% CI 41.1, 53.0)
Bachelor’s degree (n ¼ 86) 48.8 (95% CI 38.3, 59.4)
Graduate degree (n ¼ 50) 52.0 (95% CI 38.2, 65.8)
Household income (n ¼ 534a) <0.0001
<$20,000 (n ¼ 137) 29.2 (95% CI 21.6, 36.8)
$20,001–$40,000 (n ¼ 146) 48.0 (95% CI 39.8, 56.0)
$40,001–$50,000 (n ¼ 79) 48.1 (95% CI 37.1, 59.1)
$50,001–$75,000 (n ¼ 86) 54.7 (95% CI 44.1, 65.2)
$75,001–$100,000 (n ¼ 49) 55.1 (95% CI 41.2, 69.0)
>$100,000 (n ¼ 37) 73.0 (95% CI 58.7, 87.3))
Living with others (n ¼ 559a) 0.003
Yes (n ¼ 441) 49.4 (95% CI 44.8, 54.1)
No (n ¼ 118) 33.9 (95% CI 25.4, 42.4)
Era of service (n ¼ 555a)
OEF/OIF 0.03
Yes (n ¼ 131) 55.0 (95% CI 46.4, 63.5)
No (n ¼ 424) 43.9 (95% CI 39.1, 48.6)
(Continued)

S266 VOLUME 24
NUMBER S1
2020

 M. Valenstein et al.

TABLE 1. (Continued).
Has Household Firearms
Characteristic (%, 95% CI)a P value
Gulf War 0.03
Yes (n ¼ 122) 54.9 (95% CI 46.1, 63.7)
No (n ¼ 433) 44.1 (95% CI 39.4, 48.8)
Vietnam War 0.84
Yes (n ¼ 275) 46.9 (95% CI 41.0, 52.8)
No (n ¼ 280) 46.1 (95% CI 40.2, 51.9)
Korean War 0.59
Yes (n ¼ 22) 40.9 (95% CI 20.4, 61.5)
No (n ¼ 533) 46.7 (95% CI 42.5, 51.0)
World War II 0.03
Yes (n ¼ 9) 11.1 (95% CI 9.4, 31.6)
No (n ¼ 546) 47.1 (95% CI 42.9, 51.3)
Other 0.12
Yes (n ¼ 69) 37.7 (95% CI 26.2, 49.1)
No (n ¼ 4 86) 47.7 (95% CI 43.3, 52.2)
VA facility geography (n ¼ 590a) 0.02
South Atlantic (n ¼ 119) 40.3 (95% CI 31.5, 49.2)
South Central (n ¼ 120) 50.0 (95% CI 41.1, 58.9)
West Mountain South (n ¼ 105) 49.5 (95% CI 40.0, 59.1)
West Pacific (n ¼ 127) 34.7 (95% CI 26.4, 42.9)
West Mountain North (n ¼ 119) 52.9 (95% CI 44.0, 61.9)

Notes. anumber answering both demographic question and household firearm access question; bnumber answer-
ing household firearm access question; VA ¼ U.S. Department of Veterans Affairs; OEF/OIF = Operation
Enduring Freedom/Operation Iraqi Freedom.

Operation Iraqi Freedom (OEF/OIF) and likely to have household firearms, but
Gulf War veterans were more likely to PTSD or depression symptoms and sui-
report household firearms, as were those cide-related variables were not significantly
receiving treatment in the South Central, associated with firearms in the home.
West Mountain South, and West A total of 64.6% (95% CI 58.9,70.4)
Mountain North. Patients who reported of veterans with access to firearms indi-
hazardous alcohol use were also more cated that protection was their main rea-
likely to have household firearms. son for having firearms in or around the
In multivariate analyses (Online Table home, while 23.2% (95% CI 18.1, 28.3)
3), patients with higher incomes and those indicated that their main reason for having
living with others were more likely to have firearms was for hunting or sport.
firearms in the home. Respondents receiv- As outlined in Table 3, 87.9% (95%
ing services in the Pacific region were less CI 85.2, 90.6) of patients reported their
likely to have household firearms than clinician had asked them about suicidal
those in the South Central facility. thoughts. Among those who discussed sui-
Patients with risky drinking were more cidal thoughts with clinicians, slightly over

ARCHIVES OF SUICIDE RESEARCH S267

 Household Firearms and Related Discussions with Clinicians

TABLE 2. Clinical Characteristics and Household Firearms (n 5 590b)

Clinical Characteristics Has Firearm Access (%, 95% CI) P value

All veterans in sample 45.3 (95% CI 41.2, 49.3)
Current PTSD symptoms (n ¼ 577a) 0.87
Yes (n ¼ 229) 45.0 (95% CI 38.5, 51.4)
No (n ¼ 348) 45.7(95% CI 40.5, 50.9)
Current depressive symptoms (n ¼ 574a) 0.85
Yes (n ¼ 279) 45.5 (95% CI 39.7, 51.4)
No (n ¼ 295) 44.8 (95% CI 39.1, 50.4)
Current depressive or PTSD symptoms (n ¼ 586a) 0.76
Yes (n ¼ 323) 46.1 (95% CI 40.7, 51.6)
No (n ¼ 263) 44.9 (95% CI 38.9, 50.9)
Risky alcohol use (n ¼ 573a) 0.02
Yes (n ¼ 118) 55.1 (95% CI 46.1, 64.1)
No (n ¼ 455) 42.9 (95% CI 38.3, 47.4)
Suicidal ideation ever (n ¼ 580a) 0.17
Yes (n ¼ 287) 42.5 (95% CI 36.8, 48.2)
No (n ¼ 293) 48.1 (95% CI 42.4, 53.8)
Suicidal ideation past 12 mos. (n ¼ 577a) 0.73
Yes (n ¼ 111) 46.9 (95% CI 37.6, 56.1)
No (n ¼ 466) 45.1 (95% CI 40.5, 49.6)
Suicide plan ever (n ¼ 576a) 0.56
Yes (n ¼ 149) 43.6 (95% CI 35.7, 51.6)
No (n ¼ 427) 46.4 (95% CI 41.6, 51.1)
Suicide plan past 12 mos. (n ¼ 574a) 0.17
Yes (n ¼ 45) 55.6 (95% CI 41.0, 70.1)
No (n ¼ 529) 45.0 (95% CI 40.8, 49.2)
Suicide attempt ever (n ¼ 574a) 0.09
Yes (n ¼ 114) 38.6 (95% CI 29.7, 47.5)
No (n ¼ 460) 47.4 (95% CI 42.8, 52.0)
Notes. aanswered both clinical symptoms question and household firearms access question;
bnumber answering household firearm access question; PTSD ¼ post-traumatic stress disorder.

half (53.2%; 95% CI 48.7, 57.7) were firearms had talked to a clinician about
asked about suicidal thoughts at every men- firearms as part of their treatment. Of
tal health visit. Among patients who patients who had suicidal ideation in the
reported that they had suicidal thoughts in past 12 months, 35.5% (95% CI 28.2,
the past 12 months, 96.3% (95% CI 92.7, 39.8) had talked to their clinicians about
99.9) had talked with a clinician firearms and, of those who had both
about suicide. household firearms and recent suicidal
Only 27.5% (95% CI 23.8, 31.2) of ideation, 44.0% (95% CI 30.2, 57.8) had
all patients and 34.0% (95% CI 28.2, talked to their clinicians about fire-
39.8) of patients who had household arm access.

S268 VOLUME 24
NUMBER S1
2020

 M. Valenstein et al.

TABLE 3. Clinical Characteristics and Clinician Inquiries About Suicide (N 5 563) and Firearm
Access (N 5 560)

Clinician Inquired Clinician Inquired

About Suicidal About Firearm
Thoughts (%) P value Access (%) P value
All veterans in sample 87.9 (95% CI 85.2, 90.6) 27.5 (95% CI 23.8, 31.2)
Significant PTSD symptoms <0.001 <0.001
(553a; 549b)
Yes (n ¼ 223a; n ¼ 222b) 96.4 (95% CI 94.0, 99.0) 36.0 (95% CI 29.7, 42.2)
No (n ¼ 330a; n ¼ 327b) 82.7 (95% CI 78.6, 86.8) 21.7 (95% CI 17.2, 26.2)
Significant depressive <0.001 <0.01
symptoms (549a; 545b)
Yes (n ¼ 268a; n ¼ 267b) 93.7 (95% CI 90.7, 96.6) 34.1 (95% CI 28.4, 39.8)
No (n ¼ 281a; n ¼ 278b) 83.3 (95% CI 78.9, 87.6) 21.9 (95% CI 17.1, 26.8)
Risky alcohol use 0.14 0.90
(548a; 546b)
Yes (n ¼ 114a; n ¼ 114b) 92.1 (95% CI 87.2, 97.1) 27.2(95% CI 19.0, 35.4)
No (n ¼ 434a; n ¼ 432b) 87.1 (95% CI 83.9, 90.3) 27.8(95% CI 23.6, 32.0)
Suicidal ideation ever <0.001 <0.01
(560a; 555b)
Yes (n ¼ 276a; n ¼ 274b) 96.0 (95% CI 93.7, 98.3) 32.9 (95% CI 27.3, 38.4)
No (n ¼ 284a; n ¼ 281b) 79.9 (95% CI 75.3, 84.6) 21.4 (95% CI 16.6, 26.1)
Suicidal ideation past 12 <0.01 0.03
mos. (557a; 552b)
Yes (n ¼ 107 a; n ¼ 107b) 96.3 (95% CI 92.7, 99.9) 35.5 (95% CI 26.4, 44.6)
No (n ¼ 450a; n ¼ 445b) 85.8 (95% CI 82.3, 89.0) 25.2 (95% CI 21.1, 29.2)
Suicide plan ever <0.001 <0.001
(556a; 551b)
Yes (n ¼ 143a; n ¼ 144b) 96.5 (95% CI 93.5, 99.5) 41.7 (95% CI 33.6, 49.7)
No (n ¼ 413a; n ¼ 407b) 85.2 (95% CI 81.8, 88.7) 22.1 (95% CI 18.1, 26.1)
Suicide plan past12 mos. 0.04 <0.001
(554a; 550b)
Yes (n ¼ 43a; n ¼ 44b) 97.7 (95% CI 93.2, 100) 50.0 (95% CI 35.2, 64.8)
No (n ¼ 511a; n ¼ 506b) 87.3 (95% CI 84.4, 90.2) 25.3 (95% CI 21.5, 29.1)
Suicide attempt ever 0.02 0.15
(554a; 550b)
Yes (n ¼ 108a; n ¼ 110b) 94.4 (95% CI 90.1, 98.8) 32.7 (95% CI 24.0, 41.5)
No (n ¼ 446a; n ¼ 440b) 86.3 (95% CI 83.1, 89.5) 25.9 (95% CI 21.8, 30.0)
Had firearm access 0.01 <0.01
(563a; 560b)
Yes (n ¼ 256a; n ¼ 256b) 91.8 (95% CI 88.4, 95.2) 34.0 (95% CI 28.2, 39.8)
No (n ¼ 307a; n ¼ 304b) 84.7 (95% CI 80.7, 88.7) 22.0 (95% CI 17.4, 26.7)
(Continued)

ARCHIVES OF SUICIDE RESEARCH S269

 Household Firearms and Related Discussions with Clinicians

TABLE 3. (Continued).
Clinician Inquired Clinician Inquired
About Suicidal About Firearm
Thoughts (%) P value Access (%) P value
Had firearm access and idea- 0.07 <0.01
tion past 12 mos.
(557a; 552b)
Yes (n ¼ 49a; n ¼ 50b) 95.9 (95% CI 90.4, 100) 44.0 (95% CI 30.2, 57.8)
No (n ¼ 508a; n ¼ 502b) 87.0 (95% CI 84.1, 89.9) 25.5 (95% CI 21.7, 29.3)
Notes. aanswered clinical symptom question and the question regarding clinician inquiry about suicide;
b
answered clinical symptom question and the question regarding clinician inquiry about firearms; PTSD ¼ post-
traumatic stress disorder.

In multivariate analyses that assessed 95% CI 63.6, 78.3) or by both the patient
the likelihood of clinician inquiries regard- and the clinician (18.9%, 95% CI 12.6,
ing suicidal thoughts, patients who were 25.2). Patients seldom reported that they
living with others, who had suicidal idea- were the primary initiator of a discussion
tion in the past 12 months, or who had about firearm access.
higher levels of depression or PTSD symp-
toms were more likely to report talking
with their clinicians about suicidal DISCUSSION
thoughts (Table 4). Middle-aged patients
(45–64 years) were less likely to talk with Approximately 45% of VA patients in
their clinicians about suicide than younger active mental health care in five VA facili-
patients (<45 years), while those at the ties reported having firearms in their
South Atlantic and West Mountain North households, a finding that is similar to the
facilities were more likely to have those percentage of U.S. veterans in the commu-
discussions about suicide compared to nity who reported owning a firearm in
patients at the West Pacific facility. 2015 (44.9%) (Cleveland et al., 2017).
In multivariate analyses assessing the This is substantially higher than the per-
likelihood of clinician inquiries about fire- centage of persons in the general U.S.
arms, patients with significant depression population (33%) who reported in 2002
or PTSD symptoms were twice as likely to that they kept a firearm in the home
discuss firearms with their clinicians (Smith, Davern, Freese, & Hout, n.d.). It
(Table 4). Those with a recent suicide plan is also higher than the percentage of U.S.
were also more likely to talk to their households that in 2004 reported owning
clinicians about firearms. However, a firearm (32.6%) (Hamilton, Lemeshow,
lifetime suicide attempts and suicidal idea- Saleska, Brewer, & Strobino, 2018), and
tion in the past 12 months were not the percentage of individuals in the general
significant predictors of whether firearms population in 2015 (22%) who reported
were discussed. owning a firearm (Azreal et al., 2017).
When discussions with clinicians As might be anticipated in a sample of
regarding firearm access occurred, most veterans in active mental health care, many
were initiated by the clinician (71.0%, had high levels of current mental health

S270 VOLUME 24
NUMBER S1
2020

 M. Valenstein et al.

TABLE 4. Logistic Regression Analyses Assessing Correlates of Clinician Inquiries About Suicide
Ideation and Firearms

Assessment of Firearm
Assessment of Ideation Availability
Variable OR 95% CI P value OR 95% CI P value
Male gender 1.39 0.50, 3.86 0.53 1.10 0.59, 2.04 0.78
Age <45 years old ref ref
Age 45–64 years old 0.17 0.05, 0.64 <0.01 0.59 0.34, 1.03 0.06
Age >65 years old 0.33 0.08, 1.34 0.12 0.60 0.32, 1.11 0.10
White ref ref
Black 1.05 0.37, 2.97 0.92 0.81 0.41, 1.59 0.54
Hispanic 0.61 0.23, 1.59 0.31 0.65 0.31, 1.35 0.25
Other 0.69 0.23, 2.11 0.51 0.90 0.44, 1.86 0.78
<$20,000 income ref ref
$20,001–$40,000 1.20 0.48, 2.97 0.69 0.84 0.44, 1.57 0.58
$40,001–$50,000 0.90 0.32, 2.52 0.85 0.92 0.45, 1.91 0.83
$50,001–$75,000 1.08 0.40, 2.93 0.88 1.32 0.67, 2.61 0.42
$75,001–$100,000 1.67 0.32, 8.76 0.55 1.55 0.67, 3.61 0.31
>$100,000 5.96 0.64, 55.80 0.12 1.23 0.48, 3.19 0.67
Living with others 4.67 2.24, 9.73 <0.001 0.90 0.51, 1.57 0.71
South Central site ref ref
South Atlantic site 1.49 0.50, 4.48 0.48 1.40 0.69, 2.84 0.35
West Mountain South site 1.27 0.44, 3.72 0.66 1.60 0.78, 3.27 0.20
West Mountain North site 2.84 0.85, 9.52 0.09 1.51 0.75, 3.03 0.25
Pacific site 0.48 0.18, 1.24 0.13 1.36 0.68, 2.73 0.38
Hazardous alcohol use 1.57 0.65, 3.80 0.32 0.79 0.46, 1.35 0.39
PTSD/depression symptoms 2.85 1.40, 5.81 0.004 2.18 1.37, 3.47 0.0009
Suicide ideation prior 12 mos. 5.21 1.22, 22.31 0.03 0.92 0.48, 1.75 0.79
Suicide plan prior 12 mos. 0.89 0.08, 9.48 0.93 2.60 1.04, 6.52 0.04
Suicide attempt ever 2.43 0.84, 7.03 0.10 1.15 0.66, 2.02 0.62
Has firearm access 1.45 0.70, 3.00 0.31 1.61 1.03, 2.53 0.04
Note. PTSD ¼ post-traumatic stress disorder.

symptoms, risky drinking patterns, and past year and 19.9% reported a suicide
other suicide risk factors. Thus, this popu- attempt in their lifetime, one of the stron-
lation had increased suicide risks associated gest predictors of eventual death from sui-
with a high prevalence of household fire- cide (Bostwick, Pabbati, Geske, &
arms (Brent, 2001; Anglemyer, Horvath, McKean, 2016). Concerns about the over-
& Rutherford, 2014) and increased risks all prevalence of household firearms in this
from mental health and substance use con- treatment population is heightened by the
ditions (Ilgen et al., 2010). A total of high prevalence of household firearms,
19.2% reported suicidal ideation in the particularly among these subsets of patients

ARCHIVES OF SUICIDE RESEARCH S271

 Household Firearms and Related Discussions with Clinicians

with suicidal ideation, plans, or attempts. plans to reduce suicide risk. Our data indi-
Fully 46.9% of veteran patients with sui- cate that the VA has largely been successful
cidal ideation and 55.6% of those with a in facilitating discussions of suicidal
suicide plan in the past 12 months thoughts, with 87.9% of patients in VA
reported having a firearm in the home; mental health care and 96.3% of those
38.6% of respondents who had a past sui- with suicidal ideation in the past
cide attempt had a firearm in the home. 12 months reporting they had discussions
In addition, a substantial percentage with their clinicians about sui-
of veterans in mental health care stored at cidal thoughts.
least one firearm in an unsafe manner. A Unfortunately, despite high levels of
total of 38.5% of respondents with house- clinician inquiries regarding suicidal
hold firearms indicated they had an thoughts, only a minority (27.5%) of VA
unlocked and loaded firearm in their mental health patients reported they had
home, a storage practice that leaves a fire- talked to a clinician about firearms as part
arm ready for immediate use and has been of their treatment. Indeed, only a minority
associated with increased risk of suicide (34.0%) of patients who had household
(Shenassa et al., 2004; Conwell et al., firearms and only 44.0% of patients who
2002). These findings are in keeping with had both household firearms and recent
Simonetti et al. who reported that 32.7% suicidal ideation reported having discussed
of a representative sample of U.S. veterans firearms with clinicians. Even though clini-
in the community stored a firearm cians are more likely to discuss firearms
unlocked and loaded, and that 38.6% of with patients who have high levels of men-
the subset of veterans with firearms who tal health symptoms and several suicide
reported using VHA services had an risk factors, in almost every subgroup of
unlocked and loaded firearm (Simonetti, patients examined, the majority reported
Azrael, & Miller, 2018; Simonetti, Azrael, never speaking with a clinician about fire-
Rowhani-Rahbar, et al., 2018). Estimates arms. Only among patients who had a sui-
of having a loaded and unlocked firearm cide plan in the past 12 months, did at least
were lower in the general population. For 50% report having had such a discussion.
example, nationally representative data Because we asked whether patients
from the 2015 National Firearms Survey had discussed firearms with their clinicians
found that 15.7% of new firearm owners as part of their treatment, we do not know
and 33.1% of long-standing firearm own- what percentage had only a brief exchange
ers stored at least one household firearm about firearms and what percentage had
loaded and unlocked (Wertz, Azrael, more extensive discussions that focused on
Hemenway, Sorenson, & Miller, 2018). temporary removal or other ways of mak-
Morgan et al. reported that approximately ing household firearms less accessible.
18.8% of Washington State residents with Thus, the percentage of patients who
firearms had a firearm that was unlocked received counseling to reduce their firearm
and loaded (Morgan, Gomez, & Rowhani- access may be even lower than the percen-
Rahbar, 2018). tages noted above. We also do not know
The VA has made vigorous efforts to whether some veterans without household
promote structured discussions between firearms had been motivated to remove
clinicians and patients who are at risk for their firearms as a result of prior discus-
suicide and to develop individualized safety sions with clinicians.

S272 VOLUME 24
NUMBER S1
2020

 M. Valenstein et al.

Risky drinking behaviors, a well-estab- and to the percentage of gun owners in the
lished risk factor for suicide (Borges et al., general U.S. population (63%) who cited
2017), were more common among gun- protection as one of the primary reasons
owning mental health patients in our for having a firearm (Azreal et al., 2017).
study. Living with others and having Simonetti et al. found that community
higher incomes, as has been observed in veterans who endorsed concerns about
nationally representative surveys of U.S. protection were more likely to store their
adults in general (Azreal et al., 2017) and firearms loaded and unlocked (Simonetti,
in a nationally representative sample of Azrael, Rowhani-Rahbar, et al., 2018).
veterans (Cleveland et al., 2017), were also Exploring veteran and nonveteran patient
associated with a greater likelihood of hav- concerns and finding alternative ways to
ing household firearms. Thus, clinicians ensure personal and family safety may be
may need to pay particular attention to the essential if patients are to voluntarily
need for discussing firearm safety with remove firearms from their home or store
these groups of patients. The fact that their firearms more safely. Clinicians and
patients with household firearms were health systems might consider exploring or
more likely to live with others than alone providing logistical or financial support for
also raises the opportunity of bringing other means of protection when firearms
family members or friends into firearm- are voluntarily removed from the home
related suicide prevention efforts. such as more secure home entrances, locks,
This study provides additional data fences, floodlights, home security services,
that are germane to clinicians’ discussions or nonlethal self-defense devices.
with patients about firearm-related suicide Given that discussions about firearms
risk. Like firearm owners in the general are largely initiated by clinicians, add-
population and veterans in the community itional encouragement for clinicians to
(Cleveland et al., 2017; Azreal et al., bring up this topic with more of their
2017), VA mental health patients who had patients and to encourage at-risk patients
household firearms were likely to have to bring up this topic with their clinicians
more than one firearm and to have hand- appears warranted. To this end, barriers
guns. Thus, clinicians will need to discuss and facilitators to discussing firearm-
firearm-related safety practices with these related safety and effective ways to engage
factors in mind, determining how each patients in reducing firearm access during
firearm in the home might be more suicide safety planning need to be studied
securely stored or removed in urgent situa- further. Additional training of clinicians to
tions, and paying particular attention have effective conversations about means
to handguns. reduction may also be helpful, with a
Clinicians will also need to address recent study showing that safety planning
many patients’ concerns regarding protec- that included means reduction counseling
tion of themselves and their family mem- reduced suicidal behaviors (Stanley
bers. Approximately 65% of veterans with et al., 2018b).
firearms in this study indicated that pro- A study of college undergraduates,
tection was the main reason for having a including those with access to firearms,
firearm, similar to the percentage of gun- suggested that using terms such as means
owning veterans (63%) in a recent safety rather than means restriction may be
national sample (Cleveland et al., 2017) more acceptable when discussing safer

ARCHIVES OF SUICIDE RESEARCH S273

 Household Firearms and Related Discussions with Clinicians

storage of firearms (Stanley, Hom, Rogers, (Pietrzak, Johnson, Goldstein, Malley, &
Anestis, & Joiner, 2017). Studies indicat- Southwick, 2009; Sayer et al., 2010).
ing that adults in the United States often We collected data on current depres-
do not understand that firearm access sive symptoms, PTSD symptoms, and
increases suicide risk (Simonetti, Azrael, & risky drinking, which are important risk
Miller, 2018; Anestis, Butterworth, & factors for suicide, but did not collect data
Houtsma, 2018) have suggested that clini- on symptoms specific to other mental
cians may need to address patient percep- health disorders associated with suicide
tions that household firearms make people such as bipolar disorder, schizophrenia,
in the home safer, rather than less safe. In and other substance use disorders. Patients
addition, there is a need to convey more with these disorders were included in the
effectively that keeping an easily accessible sample, and symptoms specific to these
firearm at home puts everyone in the disorders may have had different associa-
home at higher risk for suicide (Conner, tions with household firearm availability
Azrael, & Miller, 2018). Additional and clinician inquiries regarding suicidal
research is needed on optimal approaches ideation or firearm access.
for clinicians attempting to discuss firearm We included a combined question
safety with their patients and on the effect- regarding the current storage of firearms
iveness of existing training programs for that asked veterans if they had at least one
firearm that was both loaded and
clinicians on means counseling (Sale
unlocked. Although some type of com-
et al., 2018).
bined question is often included in surveys
regarding firearm accessibility, the manner
in which firearm storage questions are
LIMITATIONS framed may impact respondent responses,
resulting in differing estimates of this less
When interpreting study findings, readers safe storage practice. We note there are
should consider several caveats. many different types of locks for firearms
First, our survey response rate was (i.e., trigger locks, cable locks, gun safes,
50%, which raises a concern regarding and lock boxes), that individuals may have
selection among survey respondents. lock preferences, and that the term loaded
Comparison of respondents to nonres- and unlocked may mean different things to
pondents indicated that younger veterans different respondents. Simonetti et al.
were significantly less likely to complete reported that veterans who owned cable
the survey, as were patients in the West locks were more likely to report having a
Mountain South area compared to the loaded and unlocked firearm, while those
Pacific area. However, survey respondents who owned lock boxes were less likely to
did not differ significantly from nonres- report a loaded and unlocked firearm
pondents in sex or service connection, a (Simonetti, Azrael, Rowhani-Rahbar,
variable assessing service-related disability. et al., 2018).
The response rate was also similar to or Finally, the survey frame was not a
higher than other published surveys of vet- random sample of all veterans receiving
erans in the community, with recent VA mental health treatment, but rather a
mailed surveys to veterans obtaining random sample of veterans receiving treat-
response rates between 27% and 62% ment in each of five VA facilities that were

S274 VOLUME 24
NUMBER S1
2020

 M. Valenstein et al.

purposively chosen to achieve geographic substantial counseling regarding reduction

diversity. Patients receiving treatment in of access to their firearms.
these five facilities might not be entirely Health system and educational inter-
representative of VA facilities throughout ventions to increase clinicians’ discussions
the country. Our findings for veterans of firearm-related safety may be an import-
receiving mental health care in the VA sys- ant starting point for efforts to voluntarily
tem also may not be generalizable to veter- reduce access during high-risk periods
ans receiving mental health care from (Allchin, Chaplin, & Consortium for
other providers. Risk-Based Firearm Policy, 2017). Such
efforts may include specific training of
clinicians in counseling to reduce access to
CONCLUSION lethal means. Clinician-patient discussions
will need to take into account that VA
Reducing access to lethal means that are patients with household firearms are likely
commonly used in suicides is one of the to have more than one firearm, to have
more promising strategies for preventing handguns, and to be concerned about per-
suicide deaths. To prevent veteran suicide, sonal protection. Further research is
seven in ten of which involve firearms, needed on the barriers to such conversa-
effectively addressing firearm access is key. tions, on how clinicians can best engage
In this study, veterans in mental health patients in efforts to reduce firearm-related
care reported higher levels of household risk, and on the acceptability of a variety
firearms than the general population, with of systematic health system interventions
almost half (45.3%) reporting firearms in to voluntarily reduce firearm access during
their home. Within this high-risk treat- high-risk periods (Barber & Miller, 2014).
ment population, those who had particu-
larly high suicide risks because of recent
suicidal ideation or plans also reported AUTHOR NOTE
high rates of household firearms.
The Veterans Health Administration Marcia Valenstein, VA Ann Arbor Center
has made vigorous efforts to promote clin- for Clinical Management Research, Ann
ician inquiries about suicidal thoughts and Arbor, Michigan, USA; Department of
safety discussions between clinicians and Psychiatry, University of Michigan, Ann
their at-risk patients that include consider- Arbor, Michigan, USA.
ation of lethal means (Stanley & Brown, Heather Walters, VA Ann Arbor
2008). However, even though a large Center for Clinical Management Research,
majority of VA mental health patients had Ann Arbor, Michigan, USA.
been asked about suicide, only a minority Paul Nelson Pfeiffer, VA Ann Arbor
(27.5%) had ever discussed firearms with Center for Clinical Management Research,
their clinicians. Indeed, only 44% of Ann Arbor, Michigan, USA; Department
patients who had both household firearms of Psychiatry, University of Michigan,
and recent suicidal ideation reported dis- Ann Arbor, Michigan, USA.
cussing firearms with a clinician as part of Dara Ganoczy, VA Ann Arbor Center
their treatment. Likely, even smaller per- for Clinical Management Research, Ann
centages of these patients received Arbor, Michigan, USA.

ARCHIVES OF SUICIDE RESEARCH S275

 Household Firearms and Related Discussions with Clinicians

Mark Andrew Ilgen, VA Ann Arbor REFERENCES

Center for Clinical Management Research,
Ann Arbor, Michigan, USA; Department Allchin, A., & Chaplin, V. Consortium for Risk-
of Psychiatry, University of Michigan, Based Firearm Policy. (2017). Breaking through
Ann Arbor, Michigan, USA. barriers: The emerging role of healthcare provider
training programs in firearm suicide prevention.
Matthew Jason Miller, Department of
Retrieved from http://efsgv.org/wp-content/
Health Sciences, Northeastern University uploads/2017/09/Breaking-through-Barriers-
Bouve College of Health Sciences, Boston, September-2017-Consortium-for-Risk-Based-
Massachusetts, USA. Firearm-Policy-FINAL.pdf
Matthew Fiorillo, School of Public Anestis, M. D., Butterworth, S. E., & Houtsma,
Health, University of Michigan, Ann C. (2018). Perceptions of firearms and suicide:
Arbor, Michigan, USA. The role of misinformation in storage practices
and openness to means safety measures. Journal
Robert M. Bossarte, West Virginia
of Affective Disorders, 227, 530–535. doi:
University Injury Control Research 10.1016/j.jad.2017.11.057
Center, West Virginia University, Anglemyer, A., Horvath, T., & Rutherford, G.
Morgantown, West Virginia, USA; (2014). The accessibility of firearms and risk for
Department of Behavioral Medicine and suicide and homicide victimization among
Psychiatry, West Virginia University household members: A systematic review and
School of Medicine, Morgantown, West meta-analysis. Annals of Internal Medicine,
160(2), 101–110. doi:10.7326/M13-1301
Virginia, USA.
Azreal, D., Hepburn, L., Hemenway, D., & Miller,
Correspondence concerning this article M. (2017). The stock and flow of US firearms:
should be addressed to Marcia Valenstein, Results from the 2015 National Firearms
VA Ann Arbor Center for Clinical Survey. Russell Sage Foundation Journal of Social
Management Research, 2800 Plymouth Science, 3(5), 38–57.
Road, Ann Arbor, MI 48109-2800, USA; Bagley, S. C., Munjas, B., & Shekelle, P. (2010). A
Department of Psychiatry, University of systematic review of suicide prevention programs
for military or veterans. Suicide Life Threat
Michigan, Ann Arbor, MI 48109-0624,
Behavior, 40(3), 257–265. doi:10.1521/
USA. Email: [email protected] suli.2010.40.3.257
Barber, C. W., & Miller, M. J. (2014). Reducing a
suicidal person’s access to lethal means of sui-
SUPPLEMENTAL DATA cide: a research agenda. American Journal of
Preventive Medicine, 47(3), S264–S272. doi:
Supplemental data for this article can
10.1016/j.amepre.2014.05.028
be accessed at publisher’s weblink. Betz, M. E., Miller, M., Barber, C., Beaty, B.,
Miller, I., Camargo, C. A., & Boudreaux, E. D.
(2016). Lethal means access and assessment
FUNDING among suicidal emergency department patients.
Depression and Anxiety, 33(6), 502–511. doi:
This work was supported by the 10.1002/da.22486
Department of Veterans Affairs Health Bliese, P. D., Wright, K. M., Adler, A. B., Cabrera,
Services Research and Development O., Castro, C. A., & Hoge, C. W. (2008).
Validating the primary care posttraumatic stress
(RRP 07-284) and the Department of
disorder screen and the posttraumatic stress dis-
Veterans Affairs Center of Excellence for order checklist with soldiers returning from
Suicide Prevention, Canandaigua, New combat. Journal of Consulting and Clinical
York. M. Miller’s time was supported in Psychology, 76(2), 272–281. doi:10.1037/0022-
part by the Joyce Foundation. 006X.76.2.272

S276 VOLUME 24
NUMBER S1
2020

 M. Valenstein et al.

Borges, G., Bagge, C. L., Cherpitel, C. J., Conner, Dobscha, S. K., Denneson, L. M., Kovas, A. E.,
K. R., Orozco, R., & Rossow, I. (2017). A Corson, K., Helmer, D. A., & Bair, M. J.
meta-analysis of acute use of alcohol and the risk (2014). Primary care clinician responses to posi-
of suicide attempt. Psychological Medicine, 47(5), tive suicidal ideation risk assessments in veterans
949–957. doi:10.1017/S0033291716002841 of Iraq and Afghanistan. General Hospital
Bostwick, J. M., Pabbati, C., Geske, J. R., & Psychiatry, 36(3), 310–317. doi:10.1016/
McKean, A. J. (2016). Suicide attempt as a risk j.genhosppsych.2013.11.007
factor for completed suicide: Even more lethal Hamilton, D., Lemeshow, S., Saleska, J. L.,
than we knew. American Journal of Psychiatry, Brewer, B., & Strobino, K. (2018). Who owns
173(11), 1094–1100. doi:10.1176/appi.ajp. guns and how do they keep them? The influence
2016.15070854 of household characteristics on firearms owner-
Brent, D. A. (2001). Firearms and suicide. Annals ship and storage practices in the United States.
of the New York Academy of Sciences, 932, Preventive Medicine, 116, 134–142. doi:
225–239. discussion; 239–240. 10.1016/j.ypmed.2018.07.013
Bush, K., Kivlahan, D. R., McDonell, M. B., Fihn, Heinz, A. J., Cohen, N. L., Holleran, L., Alvarez,
S. D., & Bradley, K. A. (1998). The AUDIT J. A., & Bonn-Miller, M. O. (2016). Firearm
alcohol consumption questions (AUDIT-C): An ownership among military veterans with PTSD:
effective brief screening test for problem drink- A profile of demographic and psychosocial corre-
ing. Ambulatory Care Quality Improvement lates. Military Medicine, 181(10), 1207–1211.
Project (ACQUIP). Alcohol use disorders identi- doi:10.7205/MILMED-D-15-00552
fication test. Archives of Internal Medicine, Ilgen, M. A., Bohnert, A. S. B., Ignacio, R. V.,
158(16), 1789–1795. doi:10.1001/archinte. McCarthy, J. F., Valenstein, M. M., Kim,
158.16.1789 H. M., & Blow, F. C. (2010). Psychiatric diag-
CDC/National Center for Health Statistics (2017). noses and risk of suicide in veterans. Archives of
Suicide and self-inflicted injury. Centers for General Psychiatry, 67(11), 1152–1158. doi:
Disease Control and Prevention. Retrieved from 10.1001/archgenpsychiatry.2010.129
https://www.cdc.gov/nchs/fastats/suicide.htm Inskip, H. M., Harris, E. C., & Barraclough, B.
Centers for Disease Control Office of Statistics and (1998). Lifetime risk of suicide for affective dis-
Programming (n.d). Web-Based Injury Statistics order, alcoholism and schizophrenia. British
Query and Reporting System (WISQARS). Journal of Psychiatry, 172(1), 35–37. doi:
Retrieved from https://www.cdc.gov/injury/wis- 10.1192/bjp.172.1.35
qars/fatal.html Kang, H. K., Bullman, T. A., Smolenski, D. J.,
Cleveland, E. C., Azrael, D., Simonetti, J. A., Skopp, N. A., Gahm, G. A., & Reger, M. A.
& Miller, M. (2017). Firearm ownership (2015). Suicide risk among 1.3 million veterans
among American veterans: Findings from the who were on active duty during the Iraq and
2015 National Firearm Survey. Injury Afghanistan wars. Annals of Epidemiology, 25(2),
Epidemiology, 4(1). doi:10.1186/s40621-017- 96–100. doi:10.1016/j.annepidem.2014.11.020
0130-y Kroenke, K., & Spitzer, R. (2002). The PHQ-9: a
Conner, A., Azrael, D., & Miller, M. (2018). new depression and diagnostic severity measure.
Public opinion about the relationship between Psychiatric Annals, 32(9), 509–521. doi:10.3928/
firearm availability and suicide: Results from a 0048-5713-20020901-06
national survey. Annals of Internal Medicine, Mann, J. J., Apter, A., Bertolote, J., Beautrais, A.,
168(2), 153–155. doi:10.7326/M17-2348 Currier, D., Haas, A., … Hendin, H. (2005).
Conwell, Y., Duberstein, P. R., Connor, K., Suicide prevention strategies: A systematic
Eberly, S., Cox, C., & Caine, E. D. (2002). review. Journal of the American Medical
Access to firearms and risk for suicide in middle- Association, 294(16), 2064–2074. doi:10.1001/
aged and older adults. American Journal of jama.294.16.2064
Geriatric Psychiatry, 10(4), 407–416. Miller, M., Barber, C., Azrael, D., Calle, E. E.,
Dillman, D. A., Smyth, J. D., & Christian, L. M. Lawler, E., & Mukamal, K. J. (2009). Suicide
(2009). Internet, mail, and mixed-mode surveys: among US veterans: A prospective study of
The tailored design method. Hoboken, NJ: Wiley. 500,000 middle-aged and elderly men. American

ARCHIVES OF SUICIDE RESEARCH S277

 Household Firearms and Related Discussions with Clinicians

Journal of Epidemiology, 170(4), 494–500. doi: in an insured population. American Journal of

10.1093/aje/kwp164 Epidemiology, 147(2), 155–160. doi:10.1093/
Miller, M., Barber, C., Young, M., Azrael, D., oxfordjournals.aje.a009428
Mukamal, K., & Lawler, E. (2012). Veterans Simonetti, J. A., Azrael, D., & Miller, M. (2018).
and suicide: A reexamination of the national Firearm storage practices and risk perceptions
death index–linked national health interview among a nationally representative sample of U.S.
survey. American Journal of Public Health, veterans with and without self-harm risk factors.
102(S1), S154–S159. doi:10.2105/AJPH.2011. Suicide and Life-Threatening Behavior. doi:
300409 10.1111/sltb.12463
Morgan, E. R., Gomez, A., & Rowhani-Rahbar, A. Simonetti, J. A., Azrael, D., Rowhani-Rahbar, A.,
(2018). Firearm ownership, storage practices, & Miller, M. (2018). Firearm storage practices
and suicide risk factors in Washington State, among American veterans. American Journal of
2013–2016. American Journal of Public Health, Preventive Medicine, 55(4), 445–454. doi:
108(7), 882–888. doi:10.2105/AJPH.2018. 10.1016/j.amepre.2018.04.014
304403 Smith, P. N., Currier, J., & Drescher, K. (2015).
Nalpathanchil, L. (2013). March 6). VA offers free Firearm ownership in veterans entering residen-
gun locks to help prevent vet suicides. National tial PTSD treatment: Associations with suicide
Public Radio. Retrieved from https://www.npr. ideation, attempts, and combat exposure.
org/2013/03/07/173609220/va-offers-free-gun- Psychiatry Research, 229(1–2), 220–224. doi:
locks-to-help-prevent-vet-suicides 10.1016/j.psychres.2015.07.031
Pietrzak, R. H., Johnson, D. C., Goldstein, M. B., Smith, T., Davern, M., Freese, J., & Hout, M. (n.d).
Malley, J. C., & Southwick, S. M. (2009). General Social Surveys, 1972–2016 [machine-read-
Perceived stigma and barriers to mental health
able data file]. National Science Foundation.
care utilization among OEF-OIF veterans.
Retrieved from https://gssdataexplorer.norc.org/
Psychiatric Services, 60(8), 1118–1122. doi:
trends/Civil%20Liberties?measure¼owngun
10.1176/appi.ps.60.8.1118
Stanley, B., & Brown, G. (2008). Safety plan
Sale, E., Hendricks, M., Weil, V., Miller, C.,
treatment manual to reduce suicide risk: Veteran
Perkins, S., & McCudden, S. (2018).
version. Retrieved from http://r.search.yahoo.
Counseling on access to lethal means
com/_ylt¼A0LEVvt8qNpYPE8AgwQnnIlQ;_
(CALM): An evaluation of a suicide preven-
ylu¼X3oDMTByOHZyb21tBGNvbG8DYm-
tion means restriction training program for
YxBHBvcwMxBHZ0aWQDBHNlYwNzcg–/
mental health providers. Community Mental
Health Journal, 54(3), 293–301. doi:10.1007/ RV¼2/RE¼1490753788/RO¼10/RU¼http
s10597-017-0190-z %3a%2f%2fwww.mentalhealth.va.gov%2fdocs
Sayer, N. A., Noorbaloochi, S., Frazier, P., %2fva_safety_planning_manual.pdf/RK¼0/
Carlson, K., Gravely, A., & Murdoch, M. RS¼kbH1ilhac_hYGPQMwnKlTdLfpJk-
(2010). Reintegration problems and treatment Stanley, B., Brown, G. K., Brenner, L. A.,
interests among Iraq and Afghanistan combat Galfalvy, H. C., Currier, G. W., Knox, K. L.,
veterans receiving VA medical care. Psychiatric … Green, K. L. (2018). Comparison of the
Services, 61(6), 589–597. doi:10.1176/ safety planning intervention with follow-up vs
ps.2010.61.6.589 usual care of suicidal patients treated in the
Shenassa, E. D., Rogers, M. L., Spalding, K. L., emergency department. JAMA Psychiatry, 75(9),
Roberts, M. B., Shenassa, E. D., Rogers, M. L., 894–900. doi:10.1001/jamapsychiatry.2018.
… Roberts, M. B. (2004). Safer storage of fire- 1776
arms at home and risk of suicide: A study of Stanley, I. H., Hom, M. A., Rogers, M. L.,
protective factors in a nationally representative Anestis, M. D., & Joiner, T. E. (2017).
sample. Journal of Epidemiology & Community Discussing firearm ownership and access as part
Health, 58(10), 841–848. doi:10.1136/ of suicide risk assessment and prevention:
jech.2003.017343 “Means Safety” versus “Means Restriction.”
Simon, G. E., & VonKorff, M. (1998). Suicide Archives of Suicide Research, 21(2), 237–253.
mortality among patients treated for depression doi:10.1080/13811118.2016.1175395

S278 VOLUME 24
NUMBER S1
2020

 M. Valenstein et al.

Traylor, A., Price, J. H., Telljohann, S. K., King, health treatment. Presented at the Department of
K., & Thompson, A. (2010). Clinical psycholo- Veterans Affairs Health Services Research and
gists’ firearm risk management perceptions and Development/QUERI National Conference,
practices. Journal of Community Health, 35(1), National Harbor, Maryland. Retrieved from
60–67. doi:10.1007/s10900-009-9200-6 https://www.hsrd.research.va.gov/meetings2012/
U.S. Department of Veterans Affairs (VA) Veterans abstract-display.cfm?RecordID¼473
Health Administration (VHA) Dementia Steering VHA Office of Suicide Prevention. (2016). Suicide
Committee. (2016). VHA Dementia Steering among veterans and other Americans, 2001–2014.
Committee recommendations for dementia care in the Retrieved from https://www.mentalhealth.va.
VHA health care system 2016. Department of
gov/docs/2016suicidedatareport.pdf
Veterans Affairs. Retrieved from https://www.va.
Walters, H., Kulkarni, M., Forman, J., Roeder, K.,
gov/GERIATRICS/docs/VHA_DSC_RECOMME
Travis, J., & Valenstein, M. (2012). Feasibility
NDATIONS_SEPT_2016_9-12-16.pdf
and acceptability of interventions to delay gun
Valenstein, M., Kim, H. M., Ganoczy, D.,
McCarthy, J. F., Zivin, K., Austin, K. L., … access in VA mental health settings. General
Olfson, M. (2009). Higher-risk periods for sui- Hospital Psychiatry, 34(6), 692–698. doi:
cide among VA patients receiving depression 10.1016/j.genhosppsych.2012.07.012
treatment: Prioritizing suicide prevention efforts. Wertz, J., Azrael, D., Hemenway, D., Sorenson, S.,
Journal of Affective Disorders, 112(1–3), 50–58. & Miller, M. (2018). Differences between new
doi:10.1016/j.jad.2008.08.020 and long-standing US gun owners: Results
Valenstein, M., Walters, H., Ganoczy, D., & Ilgen, from a national survey. American Journal of
M. (2012). Firearm access and potential interven- Public Health, 108(7), 871–877. doi:10.2105/
tions to increase safety for VA patients in mental AJPH.2018.304412

ARCHIVES OF SUICIDE RESEARCH S279

Copyright of Archives of Suicide Research is the property of Taylor & Francis Ltd and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.