Randomized Trial to Improve Safe Firearm Storage

Study Design and Population We conducted a single-center, prospective, randomized controlled trial with the emergency department of a free-standing tertiary care level-1 pediatric trauma center. Due to the nature of the study, investigators were not blinded to the interventions received by participants. This study was approved by the Cincinnati Children's Hospital Medical Center (CCHMC) Institutional Review Board. Study protocol, counseling handout, and surveys are available upon request to study team.

The study population consisted of firearm-owning caregivers of patients presenting to the PED for emergent MH evaluation. As only patients under the age of 18 can receive psychiatric care at our facility, all children of enrolled caregivers were less than 18 years of age.

Survey Development Survey content was derived from the expert opinion of PED physicians, injury prevention researchers, and psychiatry staff involved in the care of MH patients. Pre-counseling surveys contained questions regarding caregiver demographics as well as firearm and medication storage practices. Follow-up surveys repeated questions about storage behaviors with additional questions regarding how participants viewed PED-based counseling, use of the provided locks, removal of lethal means from the home, and the purchase of additional safety devices after PED encounter.

Counseling handouts and surveys were piloted for readability and content among 12 firearm-owning caregivers.17 Inclusion criteria revised to include all gun-owning caregivers rather than only those with unsafely stored weapons. Potential participants felt more comfortable relaying storage practices via anonymous survey as opposed to face-to-face with counselor during screening process. Content regarding another lethal mean, medication, was added to reduce perceived judgement surrounding firearm ownership.

Study Procedure All enrollment occurred within the primary PED of CCHMC between June 28, 2021 and February 10, 2022. Potential participants were identified by documented chief complaint of "Psychiatric Evaluation", which includes, but is not limited to, those presenting for: suicidal ideation/attempt, homicidal ideation/attempt, aggression, behavior change, and hallucinations. The Psychiatric Intake and Response Center (PIRC) team is made up of social workers and an attending psychiatrist that consult on all patients presenting to the PED for psychiatric evaluation. Caregivers who met inclusion criteria and received consultation by PIRC team were considered for enrollment. A convenience sample of caregivers were screened for enrollment by the study principal investigator (PI) or Clinical Research Coordinators (CRCs) specifically trained in study recruitment and lethal means counseling. CRCs were present in the PED 8am to 12pm on weekdays and 11am to 9pm on weekends. The PI aided in recruitment, as needed during high patient volumes.

Caregivers were approached for enrollment after patient had been evaluated by both emergency physicians and PIRC team. If more than one caregiver was present, one caregiver was chosen by the family for participation. Participants provided electronic consent prior to completing survey questions. Neither consent nor assent was required from patients as no protected health information was collected. Eligibility screening and safety counseling with caregiver occurred outside of patient room, physically distant from patient, in order to avoid raising awareness of firearms in the home. Baseline survey, detailing caregiver demographics as well as storage habits of medications and firearms, was completed prior to counseling.

Enrollment occurred in parallel with a 1:1 allocation ratio. Enrolled caregivers were randomized into one of two study arms based on date of enrollment. Participants enrolled on odd-numbered dates were allocated to the control group (LMC alone), while participants enrolled on even-numbered dates were allocated to the intervention group (LMC + 2 cable-style gun locks). Participants completed the electronic-based survey on a tablet with study team member present to clarify questions. The REDCap® application was used for data collection.

The control arm of the study received standardized LMC from the study PI or CRC as well as a 1-page handout summarizing the counseling recommendations. The intervention arm received the same counseling and handout with the additional provision of 2 cable-style gun locks. Counseling provided by study team was derived from the Suicide Prevention Resource Center's "Counseling on Access to Lethal Means" training module18 and the "Store It Safe" campaign19 from the Ohio Chapter of the American Academy of Pediatrics (AAP). Guidance focused on securement of dangerous items in the household, such as medications and firearms, with locking devices-or more preferably removing these items from the home, even if temporarily. Particular attention was given to the AAP's recommendations on safe storage of firearms, which states that all firearms should be kept locked, unloaded, with ammunition secured separately19. The cable locks were SnapSafe Cable Padlock (Item No. 75281). Cable locks were chosen as they have several advantages over other types of gun locks, including: (1) they are near-universally applicable to both handguns and long guns (2) cable locks are the least expensive type of gun locks, often costing 5-10x less than even the most basic gun safes/lockboxes (3) the possibility of unintentional discharge during securement process is much lower than that of devices such as a trigger locks and (4) even if a gun is kept loaded, a cable lock will prevent the firing pin from striking a bullet/shell. Caregivers were provided with instructions included within the package of the provided gun locks and advised to store keys away from the firearm in a location that children would not have access. All patients and caregivers received usual care from PIRC team which includes a standardized safety checklist and instructions on increased supervision of child. Counseling provided by study team members was supplementary to usual care.

A link to a follow-up survey was sent to participants 4 weeks after completion of baseline survey. This survey was distributed via text message using functionality within the REDCap® application and sent to the mobile phone of the participant completing the initial survey. If a participant did not complete the follow-up survey, reminders were sent at 3-day intervals up to 3 additional times. Participants received $10 gift cards upon completion of each survey.

Statistical Analysis A sample size of 200 patients was determined based on a 95% confidence interval with an 80% power to detect a 30% difference in primary outcome-based on described outcome differences in previous clinic-based trials.14 Sample size calculations accounted for an estimated 30% loss to follow-up based on prior survey-based projects within CCHMC PED.11 Wilcoxon rank sum and Fisher's exact tests were used to analyze differences between primary comparison groups at both baseline and follow-up, for continuous and categorical variables, respectively. Odds ratios and McNemar's test were used to compare baseline vs follow-up securement rates within and between study groups. Logistic regression-with adjustments for patient and caregiver age, gender and race-was used to compare change in proportion reporting firearm securement between study arms.