Firearm Safety Discussions Between Clinicians and U.S. Adults Living in Households With Firearms: Results From a 2019 National Survey

Background: Several U.S. medical organizations urge clinicians to counsel patients about firearm risk (1). How often these conversations occur and what topics are covered are unknown. Objective: To estimate the proportion of adults in gunowning households who have discussed firearm safety with a clinician and to characterize these encounters. Methods: Data were obtained from the second National Firearms Survey, conducted online 30 July 2019 to 11 August 2019 by the survey firm Ipsos. The study sample comprised adults living in homes with firearms, regardless of whether they were gun owners themselves. Participants were sampled from Ipsos KnowledgePanel, a probability-based web panel of approximately 55000 noninstitutionalized, English-speaking adults that was designed to represent the U.S. population, excluding active military service personnel. Weighting commands in Stata, version 16.1 (StataCorp), were used to generate descriptive statistics using weights provided by Ipsos. Poststratification weights adjusted for nonresponse and underor overcoverage from the study-specific sample design relative to expected demographic distributions from the U.S. Census Current Population Survey and the American Community Survey. These weights also adjusted for characteristics not available from national surveys, such as gun ownership, from weighted KnowledgePanel data. Harvard University's Institutional Review Board approved the study. Respondents were asked, “Has a physician or other health care practitioner ever spoken to you about firearm safety?” Those answering affirmatively were asked where discussions occurred (outpatient medical visit, outpatient mental health visit, emergency department, or other setting); who the patient was (the respondent, another adult, or a child); and, for each patient–setting combination affirmed, what was discussed in the most recent such encounter. Participant characteristics included sociodemographics, presence of children in the home, and personal firearm ownership status. Findings: The National Firearms Survey included 4030 adult respondents, all of whom lived in homes with firearms (completion rate, 65%); 4011 answered all firearm safety questions. Of all respondents, 7.5% (95% CI, 6.6% to 8.6%) had ever discussed firearm safety with a provider (12.0% [CI, 9.9% to 14.6%] of those living with children vs. 5.3% [CI, 4.4% to 6.3%] in homes without children) (Table 1). Most encounters involved an outpatient medical visit (not shown). Of respondents spoken to about firearms, 48.0% (CI, 41.1% to 54.9%) said that locking all firearms was discussed at their most recent visit, 31.8% (CI, 23.6% to 38.7%) that storing ammunition separately from firearms was discussed, and 15.9% (CI, 11.3% to 21.9%) that removing firearms from the home was covered (Table 2). Removing firearms was rarely discussed when the patient was a child (4.1% [CI, 1.1% to 9.4%]); when the patient was the respondent or another adult, however, conversations about removal were reported by one quarter or half of respondents, respectively. Discussion: Fewer than 10% of adults who currently live in homes with firearms report having ever discussed firearm safety with a clinician. When firearm discussions occurred, the most common advice across all clinical settings was to lock all household firearms. Whether additional safety advice was offered varied by context. In pediatric settings, removing firearms from the homewas rarely discussed. By contrast, and consistent with empirical evidence favoring removal over in-home storage for adults (2), when an adult other than the respondent was the patient, removal was discussed in more than half of encounters in which firearm safety was discussed (and in one quarter of such encounters when the respondent was the patient). To our knowledge, this is the first study to address the frequency and content of firearm safety conversations in a national sample. Our study has limitations, perhaps the most consequential being that we focused exclusively on adults currently living in homes with firearms. On the basis of a prior study that found that few former gun owners gave up their firearms because of safety concerns (3), we do not expect that expanding the denominator to adults who had ever lived in a home with firearms would appreciably alter our main findings. Recall bias may have led to underestimates. If so, our findings nevertheless reflect a clinically relevant outcome: encounters salient enough to be recalled. Our study suggests that efforts to date have not adequately promoted firearm safety discussions in clinical settings. Addressing clinician concerns that these discussions will negatively affect the physician–patient relationship (4) in light of recent evidence that patients are usually open to such conversations (5) may be a good place to start.