Base Rates of Firearm Possession by Hospitalized Psychiatric Patients

Firearms are used in a majority of suicides and homicides in the United States ( 1 , 2 ). Surveys suggest that about 41% of American households own a firearm ( 3 ). The significance of these epidemiological findings for clinical assessment appears to have been neglected within the mental health professions, which have been inconsistent in their assessment of firearm possession. For example, a study of 267 outpatients receiving care for psychiatric and substance use disorders found that only 6% of the patients reported ever having been screened for gun ownership ( 4 ).

A key concept in risk management for suicide and violence is that reducing access to lethal methods can be life saving ( 5 ). Although clinical lore suggests that a thorough evaluation of patients at risk of suicide and violence includes consideration of the availability of means to commit the act, additional epidemiological data could inform evidence-based practice in this area. Some professional groups, such as the American Academy of Pediatrics ( 6 ), have argued that routine screening for firearms is desirable from the public health standpoint of injury prevention ( 7 ). Our review of the literature in mental health-related fields, however, suggests that a more common perspective is to not inquire about whether individual patients have firearms unless they manifest elevated risk (for example, recent ideation or history of suicidal or violent behavior), and even then, only when they acknowledge a plan involving use of a gun. The extent to which such a risk assessment strategy identifies high-risk individuals with firearms is unknown. This issue has particular relevance to inpatient care, given that the first few weeks after hospital discharge are associated with elevated risk of suicidal and violent behavior ( 8 , 9 ).

This study reviewed medical records to describe base rates of firearm possession among psychiatric inpatients. During the interval covered by our chart review, staff on an inpatient service had begun including routine questions about weapons in the safety assessment during the admission workup. We reviewed the charts of 100 consecutively admitted patients who had undergone routine screening and the charts of 100 consecutive admissions during a period several months previously in order to determine the rates of gun possession identified when clinicians conducted routine firearms screening versus the rates identified when case-specific decisions were made about whether to ask about firearms.

Methods

This brief report describes part of a larger study of hospitalization processes and civil commitment. Because it was a retrospective chart review study, the Committee on Human Research of the University of California, San Francisco, affirmed that informed consent was not necessary. Charts were retrospectively reviewed between February and August 2004.

In the fall of 2003 the facility had undertaken a performance improvement project concerning risk assessment, in which routine weapons screening had been incorporated into the workup completed by staff at the time patients were admitted to the locked psychiatric inpatient unit. Nurses ask routine screening questions about a variety of safety factors, including weapon possession, such as "Do you own a gun or is one available to you?" and "Do you own any other weapons?" Physicians are encouraged to ask patients about possession of firearms, and social workers are responsible for informing patients who have been civilly committed because of danger to self or others of a state law that prohibits purchasing a firearm for the next five years ( 10 ).

For our study, two graduate students in clinical psychology reviewed the charts of 100 consecutive unduplicated admissions from June 2003 to July 2003, before routine screening was implemented, and 100 consecutive admissions from November 2003 to December 2003, after routine firearms screening was in place. They reviewed charts for any notation that the patients reported firearms possession. In addition, they collected demographic and clinical information for risk factors for suicide and violence and rated standard measures of violence risk, including the Historical Clinical Risk Management-20 (HCR-20) ( 11 ) and the Psychopathy Checklist-Screening Version (PCL-SV) ( 12 ). Before making ratings for the study group, the chart reviewers reached a criterion level of interrater reliability (intraclass correlation coefficient) ( 13 ) of .75 or greater based on ten practice charts.

Data analysis involved description of base rates of firearm possession, and comparison of the proportion of patients reporting gun possession before and after routine screening was instituted, using chi square analyses that were corrected for continuity. To evaluate whether differences in base rates of firearm possession could be explained by differences in case mix, we compared the clinical and demographic characteristics of patients before and after routine screening was in place, using chi square analyses for categorical variables and t tests for continuous variables. We also evaluated whether clinical and demographic characteristics were associated with firearm possession, using chi square analyses for categorical variables and t tests for continuous variables.

Results

The mean±SD age of the 200 patients was 44±17 years. A total of 77 patients (39%) were male, 134 (67%) were Caucasian, 26 (13%) were African American, 20 (10%) were Asian American, 14 (8%) were Hispanic, and five (2%) were from other ethnic backgrounds. Diagnoses included depressive disorders (93 patients, or 47%); substance-related disorders (50 patients, or 25%); schizophrenia (23 patients, or 12%); bipolar disorder, manic episode (ten patients, or 5%); and personality disorders (18 patients, or 9%). (Total exceeds 100% because of comorbidity.) A total of 163 patients (82%) were admitted involuntarily on the basis of emergency civil commitments.

The proportion of patients who acknowledged having firearms showed a large and statistically significant increase after staff began routinely inquiring about the subject ( χ² =5.16, df=1, p<.03). Of the 100 consecutive unduplicated admissions sampled before routine weapon screening, staff identified one patient as having a firearm. Of the 100 consecutive admissions sampled after the staff had adopted routine weapon screening, nine acknowledged owning or having access to firearms; seven owned firearms and two had access to firearms.

It is conceivable that the difference in the proportion of patients acknowledging firearm possession was due to differences in the diagnostic and demographic characteristics of the patients who had been hospitalized before and after implementation of routine screening. To evaluate this possibility, we compared the two groups on these variables. No significant differences between the two groups were identified on any of these variables.

We also evaluated the association between firearm possession and demographic and clinical characteristics that prior research has linked to risk of suicide and violence. Although our statistical power to detect such associations was modest, no significant associations were identified between gun possession and patients' historical, clinical, and risk management factors for suicide and violence or on standard measures of violence risk (for example, the HCR-20 and the PCL-SV). We did not identify any "profile" of patients who had access to guns.

Discussion

Past research suggests that psychiatric patients are rarely asked by clinicians about firearm possession ( 4 ). The results of this study suggest that, when patients are evaluated on an ad hoc basis, few patients are identified as having guns. If the clinician does not ask about firearms, patients may not volunteer this information. When specifically asked, a substantial minority of patients who were admitted for short-term psychiatric hospitalization acknowledged that they owned or had access to firearms. In this study, one patient out of 100 reported having a gun when screened on an as-needed basis. When all patients were asked about firearm possession, about one of every ten patients indicated that they did. Our data also suggest that firearm possession is not limited to a homogeneous subgroup of hospitalized patients but instead occurs across demographic and clinical groups.

These results represent a conservative basis for estimating base rates of firearm possession, because we relied on self-reports of patients upon admission to a locked unit, who may have had incentives to not report weapons that they actually had.

In view of the lethality potential associated with gun-related suicidal and violent behavior, our findings suggest the value of routine firearm screening in contexts in which patients are undergoing behavioral emergencies and therefore tend to be at elevated risk of impulsive suicidal or violent behavior—for example, inpatient units and emergency departments.

Once it has been determined that a patient possesses a firearm, the clinician may assess whether the individual patient is at risk of suicide or violence, and if the patient is at elevated risk, work with the patient to remove firearms from easy accessibility—specifically, ensure that any weapons in the patient's possession are placed in the hands of a third party. Options include involving family or significant others in weapon removal and having the patient turn in the gun to police. In our opinion it is not desirable to ask the patient to bring the weapon to the clinician, because this can increase the risk to the therapist. It can be helpful to address underlying concerns that may have been a motive for gun possession, such as fear or living in a dangerous neighborhood. Cultural issues need to be considered, such as whether the patient lives in a community where recreational use of firearms is popular. Even where it is culturally normative to possess firearms, it is useful to consider the benefits for risk management of removing firearms from easy accessibility among patients experiencing acute episodes of mental disorders that place them at elevated risk of suicide or violence.

Although these recommendations are based on clinical considerations, depending on the jurisdiction there may be legal restrictions on patients' eligibility to legally own firearms. For example, California has a mandatory prohibition on firearm ownership for five years for any patient placed on an involuntary civil commitment as a danger to self or others ( 10 ).

Conclusions

Given the current evidence of substantial base rates of firearms possession by hospitalized patients, strategies to manage their risk of suicide and violence may benefit from routinely assessing for firearms access.

Acknowledgments and disclosures

This study was supported in part by grant 21-MH-18261 from the National Institute of Mental Health. The authors thank Amber Green, B.A., and Corey Palatto, B.A., for their assistance.

The authors report no competing interests.