We conducted a prospective pilot feasibility study of a convenience sample of parents and legal guardians (caregivers) of youth presenting to a pediatric emergency department with a BMH complaint between 14 January 2022 and 31 January 2023. Caregivers were enrolled in a 54-bed free-standing pediatric emergency department of a leading southeastern US tertiary pediatric healthcare system. This is a regional catchment hospital with an annual ED volume of 81,492 visits in 2022; of which, 1,821 (2.2%) were BMH encounters. This study was approved by the institutional review board at Emory University.

Participants were eligible for inclusion in the study if they were an English-speaking caregiver of a child of <18 years of age presenting with a BMH complaint or a BMH concern was raised during the ED visit, and the child was evaluated by an ED BMH specialist. Caregivers were eligible if the patient lived with them for any period during weeks/weekends.

Three ED BMH specialists (one licensed clinical social worker and two licensed professional counselors), each trained in LMC using the Counseling on Access to Lethal Means (CALM) online course detailed below, consented and enrolled caregivers for our study (20). These ED BMH specialists invited caregivers to participate in the study after providing their initial BMH assessment for the patient’s primary BMH concern(s). Enrollment occurred during their clinical shifts in an area separate from the child or teen patient whenever possible. Two of these three ED BMH specialists enrolled caregivers during a mix of weekday and weekend daytime hours (6:00 AM–6:30 PM) throughout the full study period; the third ED BMH specialist enrolled caregivers during nighttime hours (6:30 PM–6:00 AM), only during the last half (6 months) of the study period. All study consents were collected electronically via iPads and managed using Research Electronic Data Capture (REDCap) tools hosted at our institution. REDCap is a secure web-based software platform designed to support data capture for research studies (28, 29).

Study on-boarding and intervention were adapted for our study population based on a previous study by Miller and Salhi et al., including online training for ED BMH to facilitate LMC counseling, distribution of educational handouts and offsite storage handouts, and distribution of firearm safe storage devices (24, 30).

Study ED BMH specialists all completed mandatory training up to 2 weeks before enrollment, including a free, 2 h, online course, CALM through Zero Suicide (20). Additionally, ED BMH specialists participated in a virtual study training session where the research team reviewed the scope of local pediatric firearm injuries and suicide risk and associated mortality data, study aims, an overview of the study LMC protocol, and types of firearm safe storage devices. The training session reiterated key points from the CALM course, including primarily recommending offsite firearm storage, especially when the child or teen is in a period of crisis. If the family reported being unable to store offsite, the recommendation was secure firearm storage, including storing all guns locked and unloaded with ammunition stored separately and locked. Additional recommendations included locking up all prescription medications (especially narcotics, sleep aids, and analgesics) except those that may be life-saving (20). Offsite firearm storage considerations and those of triple safe storage are consistent with recommendations from the recent American Academy of Pediatrics (AAP) policy statement and technical report (31, 32). Finally, all study ED BMH specialists were required to complete institutional CITI-training certification so they could enroll and consent patients for the study. This was done to minimize any disruptions in care for BMH patients. All ED BMH specialists were given a study outline document with sample conversation scripts for LMC and summarized key messages from CALM training, as detailed above, to have on hand during enrollment.

ED BMH specialists were selected to conduct the study intervention of LMC since they are the primary resource in our ED, and they routinely introduce the topic of restriction of lethal means in their assessment and during discharge planning of BMH patients. Partner meetings between ED, nursing, BMH, and social work leadership were conducted over a period of months before study implementation to discuss the need for the intervention, receive feedback, and develop an informed approach regarding study logistics. The study was also introduced at an ED division meeting before enrollment so that all ED team members were aware of the study and that there would be minimal disruptions to patient care or flow. Study investigators checked in monthly, either via meetings or email, with BMH specialists to identify and address any challenges or concerns with more frequent engagement in the first few weeks of study enrollment.

In total, 50 caregivers enrolled in the study. The majority of caregivers were female (96%), Black (52%), non-Hispanic (76%), with a mean age of 40 (±8.12), and cared for 3 or more children regularly (52%) (Table 1). Among them, 22 endorsed having a firearm in their home. The mean age of the child presenting with the caregiver at the ED was 13 (± 2.66), and 62% were female. In total, 54% of presenting youth had a primary ED discharge diagnosis of suicidality (either suicidal ideation or suicide attempt).

There were no statistically significant differences between participant demographics and baseline characteristics when comparing non-firearm owners and firearm owner caregivers except for the age ranges of children cared for within the home. Firearm owner caregivers cared for a higher percentage of 15–17-year-olds within the home (p = 0.033).

For accrual, there was an 81% acceptance rate with 50 out of the 62 caregivers approached that met eligibility criteria enrolling (Figure 1). In total, 44% (n = 50) of caregivers reported having a firearm at home (Table 2). For attrition, 80% of participants who completed the initial ED intake survey completed the 1-week follow-up; while 28% of randomly sampled firearm owners (n = 11) completed the 1-month Zoom follow-up. For acceptability, more than 80% of participants agreed both immediately after the ED education intervention and 1 week later that the education given was useful and appropriate in the ED setting. For procedural fidelity, protocol deviations were low overall (n = 3 out of 50 patient encounters). For the ability to manage the study and implement the intervention, there were no reported disruptions to patient care or flow. Study check-ins with ED BMH specialists throughout the study period revealed barriers such as technical challenges with the iPad during the ED caregiver survey and the length of time for the study consent process. ED BMH specialists reported the following facilitators for conducting the intervention: pre-intervention training, taking the CALM course training, and the ability to provide immediate resources (educational material and devices) in hand during the LMC conversation.

In total, 29 of all caregivers (58%, n = 50) and 11 of the caregivers with a firearm(s) at home (45%, n = 22) reported having not received prior education or counseling on firearm safe storage (Table 3). Only 12% of all caregivers reported having received prior firearm safety counseling from their child’s doctor’s office and 6% from their child’s prior ED visits. At baseline, 58% of all caregivers and 45% of caregivers with a firearm(s) at home reported not asking about firearms before their child/teen visited other homes. There were no significant differences between non-firearm owner caregivers and firearm owner caregivers for baseline firearm safety knowledge and the practice of asking about firearms in the home.

Among firearm owners (n = 22), 59% reported having one firearm at home; however, one-third of firearm owners reported having 3 or more guns at home (Table 4). Handguns (n = 18) were the most common type of firearm owned, followed by a shotgun (n = 9). Gun safes were the most commonly used type of safe storage device, followed by gun lock boxes. In total, 64% of firearm owners reported always using a safe storage device, while 14% never used one. A total of 77% of firearm owners reported storing all firearms currently locked, 72% storing firearms unloaded, and 77% with ammunition stored separately. Approximately one-third of firearm owners reported having stored firearms outside of their home when their child has been in a period of crisis.

In total, 13 firearm owners (59%, n = 22) requested a study safe storage device (Table 4). Among these firearm owners who requested a study safe storage device, 62% reported at baseline storing all firearms locked and 69% reported storing firearms unloaded. In comparison, for firearm owners who did not request a study device (n = 9), 100% reported at baseline storing all firearms locked and 78% storing firearms unloaded. In total, 15 safe storage devices were distributed over the course of the study, including 10 lock boxes and 5 trigger locks (Table 5).

At the 1-week follow-up, a greater proportion of all caregivers (+28%) and a greater proportion of firearm owners (+13%) self-reported that they had asked about firearms in the home before their child/teen visited other homes (Table 6). Overall, more than 85% of caregivers self-reported at baseline and 1-week post-study intervention that they will inquire in the future about firearms in other homes before their child/teen visits. Gun safety information from the ED intervention was reported to be shared with others by 45% of all caregivers (n = 40) in follow-up, including 74% of firearm (n = 19) owners. Changes in firearm storage practices at follow-up included a greater proportion of firearm owners reported storing all firearms currently locked (+23%). In total, 10 firearm owner caregivers self-reported in follow-up that they have removed firearms either temporarily or permanently from their homes. These firearm owners stated that keeping their children safe was the most common reason for removing their firearms. Among these firearm owners who reported removing a firearm from their home (n = 10), at baseline, 60% stored all firearms locked and 70% stored firearms unloaded. In comparison, for firearm owners who did not remove their firearms from their homes (n = 12), at the baseline, 89% stored all firearms locked and 70% stored firearms unloaded.

Three firearm owner participants completed the 1-month follow-up Zoom interview (Table 2; Figure 1). One of the three participants showed the study team where and how they stored their firearms at home via Zoom. Two of the three had taken a study lockbox and were currently using it to store their firearm locked. When asked what type of safe storage device we should provide to families, all three participants primarily recommended lockboxes.

There are several cautions when interpreting the findings of our study. First, the outcomes were self-reported and susceptible to recall and social desirability bias as participants may have misremembered or not been forthcoming regarding firearm ownership and firearm safety practices. Second, the small sample size and data being collected from a single center in the Southeastern US may not be generalizable to other populations. However, this is a region with both increasing suicide and firearm ownership rates. Third, while this was a convenience sample and potentially not reflective of firearm ownership rates, the reported gun ownership of 44% is only slightly lower than our state firearm ownership rates. Fourth, caregiver participants who reported access to firearms at home may not have been the primary firearm owners. Given our low video follow-up rates among firearm owners, we were unable to fully validate self-reporting. Fifth, being a convenience sample of participants identified by ED BMH specialists, the population enrolled may not be reflective of all caregivers of youth at risk for suicide. Sixth, our center had 24 h BMH specialist coverage and thus cannot determine if our training and implementation of LMC intervention would have the same results in emergency departments with different staffing models. However, the CALM program can be used and adopted by non-BMH providers. Seventh, we did not further categorize the BMH concern for which the child was diagnosed outside of suicidal thoughts or suicide attempts. Future work should evaluate the different types of BMH issues that youth are presenting with to better understand the population. Given that children with mental health disorders are at increased risk of suicide, we had our ED BMH specialists include all BMH complaints for the LMC intervention and not only those presenting with active suicidal thoughts or after a suicide attempt (45, 46). Finally, caregiver inclusion was limited to English speakers only, as our educational materials were not available in other languages.