Client firearm assessment and safety counseling: the role of social workers.
CURRENT UNDERSTANDING OF SOCIAL WORKERS ASSESSMENT AND COUNSELING PRACTICE
When mental health practitioners work with clients and their families dealing with emotional and behavioral issues, it is important that they routinely assess and provide educational intervention on the risks of keeping a gun in the home or otherwise having access to a firearm. Although many clients may not initially present in crisis, the possibility of crisis remains significant with many clients; therefore, initial assessment and counseling on firearms is a proactive practice. Suicide is rare in the absence of psychiatric illness (Beautrais, 2006), and although mental health practitioners are key gatekeepers, the U.S. Department of Health and Human Services (2001) stated that gaps still exist in specialized assessment techniques and in risk factor recognition. This report points out that whereas at-risk individuals often seek professional help, only 18 percent of suicide decedents reported suicide ideation to a health professional (Robins, cited in U.S. Department of Health and Human Services, 2001). Furthermore, as many as 90 percent of suicide decedents carry a psychiatric diagnosis at the time of death (Conwell & Brent, 1995; Moscicki, 2001),indicating the need for proper assessment and recognition of risk factors by mental health practitioners. Although the present study is focused on mental health practitioners, this issue is still relevant to community organizers and policy practitioners in the mental health community.
An effort to study the attitudes, knowledge, and behaviors surrounding client gun ownership and access has been approached within the mental health practice fields of psychology, psychiatry, and medicine. Sullivan (2004) documented the practices and comfort level of psychologists in firearm assessment and discussion with clients. Gallagher (2002) collected data on psychiatrist's knowledge, attitude, and risk assessment practices of firearm-related suicide. In the field of medicine, the American Academy of Pediatrics recommended that physicians advise patients and families on firearm injury prevention (American Academy of Pediatrics Committee on Adolescents, 1992).
Research has demonstrated that clients are receptive to clinicians' counsel in this area. Sherman et al. (2001) illustrated the effectiveness of a multidisciplinary firearm risk management program in Ohio among high-risk mental health clients who expressed intent to commit suicide by means of a firearm upon inpatient admission to a mental health facility. The program was successful in eliminating access to firearms before discharge and thus neutralized this environmental risk factor for clients. In addition, studies in the medical field have demonstrated that parents and guardians are receptive to recommendations against having guns in the home (Haught, Grossman, & Connell, 1995) and to guidance about safe storage practices (Webster, Wilson, Duggan, & Pakula, 1992). Safe storage practices have been associated with a decrease in youth homicides and suicides, as well as a decline in unintentional shooting deaths among older youths (Cummings, Grossman, Rivara, & Koepsell, 1997).
With the many avenues and settings of services, the potential of the social work profession to reduce risk related to firearm access and to counsel about safe storage practices with all clients is immense. Although other professions have been urged to assess firearm availability and advocate for the removal of firearms in their patients' homes (Brent et al., 1988; Brent et al., 1991; Kaplan, Adamek, & Rhoades, 1998), little is known about the practices and techniques that social workers use with client populations. Although it has been suggested that clinical screening for firearm access is important in reducing the risk of suicide, homicide, and injury, firearm specific education and counseling techniques are not a required component of social work undergraduate, graduate, or continuing education programs. Therefore, it is unclear what these professionals know about this issue, to what degree this type of intervention occurs, and how effective such effort is.
The primary objective of this study was to document social workers' knowledge, attitudes, and behaviors regarding client firearm safety counseling. The two primary research questions are (1)What do social workers think, know, and do about client risk associated with firearms? and (2) what preliminary factors are associated with client firearm assessment and safety counseling?
Because there is a current lack of information on this issue among social workers, this study is exploratory in nature. Therefore, specific hypotheses will not be proffered and instead certain focal areas will be explored in relation to clinicians firearm assessment and safety counseling behaviors, such as attitudes, knowledge, demographics (of practitioner and clients), and clinician experiences with firearms and client injury or death by this means.
A list of licensed social workers was acquired from the state of Ohio on October 10, 2005. The social work licensure list procured from the state of Ohio contained information for 15,350 licensed social workers (LSWs) and 6,549 licensed independent social workers (LISWs). These two levels of social work licensure require a bachelor's degree in social work and a master's degree in social work and 3,000 hours of post-master's practice experience, respectively. From this initial list, addresses that were out of the country or were from a state not bordering Ohio were deleted. The final list contained 15,101 LSWs and 6,338 LISWs. The sample was stratified such that an approximately equal number of responses from practitioners in each level of licensure would be obtained.
Before a full mail survey was conducted, a pilot study was performed on a random sample of 50 social workers (25 LSWs and 25 LISWs) in the fall of 2005. This helped refine the survey questions, timing, and response categories before the full study was conducted.
A mail survey was used to gather information from licensed social workers in Ohio. Schutt (2001) stated that the typical response rate is 30 percent for mailed surveys unless extra steps are taken to increase this rate. Previous mail surveys of specialized social work practice groups have yielded response rates of 53.8 percent and 59.2 percent when using Dillman's (2000) suggestion of reminder postcards (Astor, Behre, Wallace, & Fravil, 1998; Joseph & Broussard, 2001).Although a reminder postcard was planned in the present study, a conservative response rate of 30 percent was still used because the list was not specialized. Also, it was important in this study to capture data from social workers who are active clinicians rather than from those in higher education, administration, and advocacy roles. Pilot participants were removed from the list. To achieve a 95 percent confidence level and a confidence interval of five within each licensure status, a 30 percent response rate must be achieved.
The survey instrument was designed from other surveys examining variables that influence firearm assessment and safety counseling in the medical profession (Becher, Cassel, & Nelson, 2000; Everett, Price, Bedell, & Telljohann, 1997), and particularly one conducted by Solomon, Duggan, Webster, and Serwint (2002). These surveys included variables on attitudes, behaviors, knowledge, training, and demographic variables pertinent to firearm assessment and safety counseling with clients. Expert consultants were also used to refine content area and instrumentation.
The survey was three pages in length. The first page asked social workers 23 questions about their attitudes, knowledge, and behaviors relating to firearms and counseling. Reponses were captured on a four-point ordinal scale in which 1 = strongly disagree, 2 = disagree, 3 = agree, and 4 = strongly agree. Respondents were instructed to answer on the basis of their past two years of practice, to capture current knowledge, attitudes, and behaviors and to create a boundary to standardize responses across participants.
The second page of the survey posed questions relating to respondent's own firearm ownership, formal training in firearm counseling, use of firearm safety counseling, previous client firearm injury and death, and effect of growing up with firearms. The final page of the survey collected demographic and practice information, including years at licensure level, practice population, work function, setting, sector, community type, and caseload.
The process of obtaining consent included a cover letter indicating the purpose of the study, its voluntary nature, and the anonymity of participation. Consent was implied if respondents returned the completed survey.
A total of 2,456 surveys were mailed out in January 2006. A response rate of 31 percent was reached. After removing surveys that were not usable because they lacked licensure indications, among other reasons, the surveys from 320 LSWs and 377 LISWs were analyzed, for a response rate of 28.4 percent.
The majority of the sample was female (85 percent), white (88 percent), had a master's degree (73 percent), and obtained a degree in social work (86 percent).Average age of participants was 47.5 years, and participants reported an average of about seven years at their licensure level. The largest portion of respondents indicated that their primary practice setting was mental health (40.6 percent) and that their primary population was adults (37.5).The majority of social workers indicated that direct practice was their primary work function (75.1 percent.
Firearm Assessment and Safety Counseling Analysis
Two variables examined how much social workers agreed or disagreed with the following two statements: (1)"I routinely assess if my clients own and have access to guns" and (2) "I routinely counsel my clients about firearm safety." Firearm assessment was defined for respondents to mean asking clients if they own or have access to guns. Firearm safety counseling was defined for respondents as counseling that covers risks of gun ownership, access to guns, and safe storage methods for guns.
Participants answered questions about the frequency of firearm assessment and safety counseling on the basis of their past two years of practice experience. Responses were captured on a four-point ordinal scale ranging from 1= strongly agree to 4 = strongly disagree (see Figure 1).
The majority of social workers did not report assessing for firearms or counseling on firearms safety on a routine basis within the past two years of their practice. When the four-point response categories were dichotomized into "yes/no" categories for analysis and interpretation, 34 percent of respondents agreed that they routinely assessed for firearm ownership and access, and 15.3 percent agreed they routinely counseled on firearm safety.
Firearm assessment and safety counseling were retained as "yes/no" dichotomized variables and 21 variables regarding firearm attitude, behavior, and knowledge questions were also dichotomized to serve as predictor (independent) variables. Binary logit regression was used to estimate the models, with results presented as a percentage change in the probability of the respondent answering "yes" to the dependent measure. At the bivariate level of analysis, many personal and environmental barriers were significantly associated with routine firearm assessment and counseling as shown in Table 1.
[FIGURE 1 OMITTED]
Those measures with the statistically significant and positive change in odds were variables tapping into whether the client was depressed or suicidal and motivation by media coverage. If a client was depressed, the odds of routine firearm assessment and counseling by a social worker increased by 348.5 percent and 450 percent, respectively. In addition, social workers were asked if when assessing for suicide they routinely ask about gun access. If this was endorsed, it increased the odds of routine firearm assessment and safety counseling across the board by 381.6 percent and 232.9 percent. Also, social workers were asked if when assessing for suicide risk they routinely provide safety counseling for clients. If this statement was endorsed, it increased the odds of routine firearm assessment and safety counseling across the board by 401.3 percent and 278.5 percent, respectively.
The media's coverage of gun-related issues also affected responses to routinely assessing for firearms and counseling on firearm safety. If a social worker stated that media coverage motivated her or him to address gun-related issues with clients, the odds of routine firearm assessment and safety counseling increased by 320.6 percent and 465.9 percent, respectively.
Other variables significantly decreased the odds that these behaviors would have occurred in their practice in the past two years. For example, if the social workers endorsed the statement that they were not adequately trained on the topic of firearm safety, the change in odds that routine firearm assessment and safety counseling occurred decreased by 67 percent and 86.4 percent, respectively. If they thought it was not their responsibility to discuss firearm safety with clients, the change in odds that routine firearm assessment and safety counseling occurred decreased by 59 percent and 76.2 percent, respectively. Among respondents who believed that their clients were not at risk of firearm injury, the change in odds that routine firearm assessment and safety counseling occurred decreased by 79.4 percent and 68.4 percent, respectively. If they thought their clients were not at risk to harm others with a firearm, the change in odds that routine firearm assessment and safety counseling occurred decreased by 73 percent and 56 percent, respectively. If they endorsed the statement that there are more important topics to discuss, the change in odds that routine firearm assessment and safety counseling decreased by 52.7 percent and 69.3 percent, respectively. If social workers reported that they were uncomfortable with the topic of firearms, the change in odds that routine firearm assessment and safety counseling occurred decreased by 56.6 percent and 52.6 percent, respectively. And finally, if social workers reported not being aware of the risks associated with firearms in the homes of clients, the change in odds that routine firearm assessment and safety counseling occurred decreased by 60.3 percent and 52.9 percent, respectively.
To determine whether other variables exhibited a relationship to firearm assessment and safety counseling, multivariate analyses were conducted. This multivariate analysis also allows the estimation of a coefficient while controlling for other, possibly competing, effects on the dependent variable. Ordered logistic regression models were estimated using both firearm assessment and safety counseling as dependent variables. Each analysis included the following independent variables, which were dichotomously coded: race (white and all other races), gender, client ever injured or killed, gun access or ownership (by respondent), licensure status (LSW and LISW), firearm safety training, and community type (urban, suburban, and rural). The analyses also included interval level independent variables of age of respondent and percentage of male clients served. Demographics were used to determine whether basic clinician characteristics could be identified as influencing their behaviors.
Other independent variables were chosen because of their use in other studies as independent variables (Becher et al., 2000; Everett et al., 1997; Solomon et al., 2002; Sullivan, 2004; Webster et al., 1992). In this study, most clinicians said they did not have a client injured (66 percent) or die (85.4 percent), most said they did not live in a home with firearms (74 percent) or personally own a gun (91 percent), and the majority had not received training on firearm safety counseling (76.2 percent). In addition, most (52.3 percent) reported that their clients lived in an urban area, 37.5 percent reported that their clients lived in a suburban area, and 9.6 percent reported that their clients lived in a rural area. The average percentage of male clients served was 45 percent.
A summary of the results of factors evaluated for association with the outcome variables is presented in Table 2. The results demonstrate the percentage change in the odds of being coded in a higher category (for example, responding "strongly agree" to a statement in the survey versus responding "agree") for firearm assessment and for safety counseling with an increase in the independent variable. This is discussed in terms of the odds of registering more agreement with routine firearm assessment and safety counseling with a change in the independent variable, holding all others constant.
Starting with the most significant results, both previous firearm safety training and gender of the social worker significantly increased the odds of registering more agreement in both routine firearm assessment and safety counseling. If a social worker reported having prior firearm safety training, the odds of registering more agreement with routine assessment increased by 257.7 percent, and safety counseling by 479.3 percent. This illustrates the magnitude of firearm training for enhancing social work practice protocol.
Being female increased the odds of registering more agreement with routine firearm assessment and safety counseling by 65 percent and 75.3 percent, respectively. However, this should be interpreted with caution because many more men are in administrative positions and therefore are not likely to engage with clients in this respect.
Higher licensure status and ever having a client injured or killed by a firearm significantly influenced routine firearm assessment only. Clinicians who had a client injured or killed by a firearm were 71.1 percent more likely to register more agreement with routine firearm assessment. In addition, the odds of registering more agreement for routine firearm assessment were 35.8 percent greater for those with the LISW.
Other variables in the model did not demonstrate a significant association with routine firearm assessment and safety counseling. Race, age, and gun ownership status of respondents did not affect the odds of routine firearm assessment and safety counseling. The gender of clients was also examined to determine whether the odds would significantly increase for routine firearm assessment and safety counseling with an increasing proportion of male clients. Having more or fewer male clients did not affect the likelihood of routine firearm assessment and safety counseling. Last, the model also controlled for the effect of population density of the practitioners' primary region of practice. Social workers who indicated that they practice primarily in an urban or a suburban area were no more or less likely to register more agreement with routine firearm assessment and safety counseling with clients than were practitioners practicing in rural areas.
Although this study provides useful insight, limitations must be recognized. First, because this was an exploratory study, questions were not normed, making it difficult for comparative purposes. Another limitation was nonrespondent bias. The lower response rate of the study limits the generalizability of the results. Other counseling professions were not represented in this study. In addition, the sample of social workers was limited to those in Ohio. Finally, survey questions tapped into only the past two years of practice for these respondents.
The most common method of death by suicide is by firearms, which accounted for 51.6 percent of all deaths in 2004 (Centers for Disease Control and Prevention, 2004).Thus, the National Strategy for Suicide Prevention (U.S. Department of Health and Human Services, 2001) focus on eliminating or restricting access to the lethal means of firearms is warranted. Their report points out that social workers and other mental health practitioners are on the frontline in assessing and treating individuals at increased risk of suicidal behavior and are thus in a unique position to assess and counsel on firearm risks and safe storage as measures to reduce self-harm. Increasing the proportion of mental health practitioners who routinely assess for firearms and counsel on safety would help address this important environmental risk factor for suicide.
Developing trainings for clinicians on specific assessment and intervention techniques is aligned with the goal of "reduc[ing] factors that increase the risk of suicide" set forth by the Ohio Suicide Prevention Plan (Ohio Department of Mental Health and the Ohio Coalition for Suicide Prevention, 2002, p. 9). The present research gathered information from social workers regarding firearm assessment and intervention techniques being used with clients, highlighting social worker attitudes, knowledge, practices, and potential training needs. When best practices are documented and incorporated into mental health practice protocol, there is a strong potential to decrease clients' risk of homicide, suicide, and injury related to firearms.
From the findings, it is clear that the majority of social workers are not assessing for the presence of firearms or counseling on firearms safety on a routine basis. Only 34 percent of social workers strongly agreed that they routinely assessed for firearm ownership and access within the past two years of practice. Only 15.3 percent reported providing firearm safety counseling on this variable. To put this in perspective with the medical field, about 50 percent of pediatric residents in one recent study reported routine firearm counseling on the issue of firearms with adolescents and their parents (Solomon et al., 2002).The medical field has the professional body of the American Academy of Pediatrics strongly encouraging their physicians to counsel families in this area and has published recommendations to follow. Unfortunately, social work does not have a parallel organization encouraging this agenda to reduce this environmental risk of self-harm for clients. However, we can learn from the practices, educational interventions, and common barriers that the medical field has faced with this issue to further this important practice issue.
As a routine practice protocol, it is important to understand what factors were related to an increase in the occurrence of firearm assessment and safety counseling practice behaviors with social workers. From the bivariate analysis, factors that contribute to the decreased likelihood that routine firearm assessment and safety counseling would occur include lack of training in this area, lack of awareness of risks associated with firearms, discomfort with topic, not a social work responsibility, lack of time, not feeling their clients are at risk, and more important topics to discuss. These factors are similar to barriers perceived in the medical field: insufficient time, inadequate training, discomfort with topic, and a belief that patients do not need firearm safety counseling (Everett et al., 1997; Solomon et al., 2002). Clearly these barriers are areas easily addressed in targeted educational trainings.
Although these factors appear to decrease the probability that firearm assessment and counseling will occur, other variables indicate opportunities to be fostered. The effect of media in influencing routine firearm assessment and safety counseling techniques is apparent from these results. Pediatric residents in one study also believed that gun-related media motivated them to counsel more on this topic--those who had this belief were almost twice as likely to counsel routinely on firearm safety (Solomon et al., 2002). Continuing and improving existing forms of professional education and service campaigns is an obvious path for action. This can also include public campaigns highlighting the correlation between mental illness (depression) and firearm use.
Social workers indicated a significantly increased probability of firearm assessment and counseling if the client was depressed or suicidal. This is a positive finding for prevention of suicide by firearm. However, as a proactive and preventive measure, routine firearm assessment and safety counseling should occur as a standard part of all treatment protocol across all populations and settings, even for those clients who do not immediately present as depressed or suicidal.
An important finding from this research, indicated by the multivariate regression model, is that some type of firearm safety training increases the probability that firearm assessment (275 percent) and counseling (479 percent) would occur by social workers. Thus, training in this area would provide great potential to increase an important assessment and safety counseling protocol for mental health service providers. The audience for this type of training is considerable, with more than 76 percent of social workers in this study reporting that they had never received training or education about counseling clients on firearm safety. Comprehensive training on this topic can address many of the barriers mentioned earlier, such as lack of awareness of firearm risks to clients, importance of topic, and comfort with topic. In addition, trainings can provide agencies with educational resources, handouts, and avenues for the provision of safe storage hardware.
This study represents an initial attempt to document social workers' attitudes, knowledge, and behaviors about client firearm safety assessment and counseling. Although the focus of this article is on practitioners, these findings can extend to community organizers and policy professionals in the mental health community. More research of this nature is needed to better understand this issue among social workers and other mental health practitioners. Additional research, both quantitative and qualitative, would help create comprehensive training material on this topic. Given that social workers are key gatekeepers, their reported lack of routine firearm assessment and safety counseling is disheartening for suicide and injury prevention efforts. The irrefutable evidence that suicide by firearms accounts for the majority of these types of deaths begs for suicide, homicide, and injury prevention protocols aimed at restricting this lethal means among high-risk populations.
Original manuscript received January 23, 2007
Final revision received November 28, 2007
Accepted November 30, 2007
American Academy of Pediatrics Committee on Adolescents. (1992). Firearms and adolescents. Pediatrics, 89, 784-787.
Astor, R.A., Behre, W.J., Wallace, J. M., & Fravil, K.A. (1998). School social workers and school violence: Personal safety, training, and violence programs. Social Work, 43, 223-232.
Beautrais, A. (2006, June 12). Epidemiology of suicide. Paper presented at the Center for the Study and Prevention of Suicide, Summer Research Institute, Rochester, NY.
Becher, E. C., Cassel, C. K., & Nelson, E.A. (2000). Physician firearm ownership as a predictor of firearm injury prevention practice. American Journal of Public Health, 90, 1626-1628.
Brent, D., Perper, J., Allman, C., Moritz, G., Wartella, M., & Zelenak, J. (1991). The presence and accessibility of firearms in the homes of adolescent suicides: A case control study. JAMA, 266, 2989-2995.
Brent, D., Perper, J., Goldstein, C., Kolko, D., Allan, M., Allman, C., & Zelenak, J. (1988). Risk factors for adolescent suicide. A comparison of adolescent suicide victims with suicidal inpatients. Archives of General Psychiatry, 45, 581-588.
Centers for Disease Control and Prevention, National Center for Injury Prevention and Control (producer). Web-based Injury Statistics Query and Reporting System (WISQARS). (2004). Retrieved February 16, 2007, from http://www.cdc.gov/ncipc/wisqars/ default.htm
Colby, I., & Dziegielewski, S. (2001). Introduction to social work: The people's profession. Chicago: Lyceum Books.
Conwell, Y., & Brent, D. (1995). Suicide and aging. I: Patterns of psychiatric diagnosis. International Psychogeriatrics, 7, 149-164.
Cummings, P., Grossman, D., Rivara, E, & Koepsell, T. (1997). State gun safe storage laws and child mortality due to firearms. JAMA, 278, 1084-1086.
Dillman, D. (2000). Mail and Internet surveys (2nd ed.). New York: John Wiley & Sons.
Everett, S., Price, J., Bedell, A., & Telljohann, S. (1997). Family practice physicians' firearm safety counseling beliefs and behaviors. Journal of Community Health, 22, 313-324.
Gallagher, S. (2002, November 11). Psychiatrist's knowledge, attitude, and practices concerning firearm-related suicide risk assessment (Abstract #52433). Retrieved February 16, 2007, from http://apha.confex.com/apha/130am/ techprogram/paper_52433-htm
Haught, K., Grossman, D., & Connell, F. (1995). Parents' attitudes toward firearm injury prevention counseling in urban pediatric clinics. Pediatrics, 96, 649-653.
Joseph, A, & Broussard, A. (2001). School social workers and structured inequality: A survey of attitudes and knowledge of tracking. School Social Work Journal, 25, 59-75.
Kaplan, M., Adamek, M., & Rhoades, J. (1998). Prevention of elderly suicide: Physicians' assessment of firearm availability. American Journal of Preventive Medicine, 15, 60-64.
Long, J. S., & Freese, J. (2001). Regression models for categorical dependent variables using Stata. College Station, TX: Stata Press.
Long, J. S., & Freese, J. (2006). Regression models for categorical dependent variables using Stata (2nd ed.). College Station, TX: Stata Press.
Moscicki, E. K. (2001). Epidemiology of completed and attempted suicide: Toward a framework for prevention. Clinical Neuroscience Research, 1, 310-323.
Ohio Department of Mental Health and the Ohio Coalition for Suicide Prevention. (2002). Ohio's suicide prevention plan. Columbus: Author.
Schutt, R. (2001). Investigating the social world: The process and practice of research. Thousand Oaks, CA: Sage Publications, Pine Forge Press.
Sherman, M., Burns, K., Ignelzi, J., Raia, J., Lofton, V., Toland, D., Stinson, B., Tilley, J., & Coon, T. (2001). Firearm risk management in psychiatric care. Psychiatric Services, 52, 1057-1061.
Solomon, B., Duggan, A., Webster, D., & Serwint, J. (2002). Pediatric residents' attitudes and behaviors related to counseling adolescents and their parents about firearm safety. Archives of Pediatric and Adolescent Medicine, 156, 769-775.
StataCorp. (2005). Stata statistical software: Release 9 [Computer software]. College Station, TX: Author.
Sullivan, G. (2004). Assessment of firearm access. Behavioral Emergencies Update: Section on clinical emergencies and crises, 1 (Spring/Summer). Retrieved February 16, 2007, from http://www.apa.org/divisions/div12/ sections/section7/news/sp04/access.html
U.S. Department of Health and Human Services. (2001). National strategy for suicide prevention: Goals and objectives for action. Washington, DC U.S. Public Health Service.
Webster, D., Wilson, M., Duggan, A., & Pakula, L. (1992). Parents' beliefs about preventing gun injuries to children. Pediatrics, 89, 908-914.
Wolfe, R., & Gould, W. (1998).An approximate likelihood-ratio test for ordinal response models. Stata Technical Bulletin, 22, 193-213.
Karen Slovak, PhD, is assistant professor of social work, Ohio University, Zanesville, 1425 Newark Road, Zanesville, OH 43701; e-mail: Slovak@ohio.edu. Thomas W. Brewer, PhD, is assistant professor, Department of Justice Studies, and research fellow, Institute for the Study and Prevention of Violence, Kent State University, Kent, OH. Karen Carlson, PhD, is associate professor, Department of Social Work, Ohio University, Athens, OH. This research was funded by the Ohio Department of Mental Health, grant number 06.123.
Table 1: Bivariate Predictors of Routine Firearm Assessment and Safety Counseling Assess Change in Probability Variable b (%) Routinely counsel on firearm safety with suicidal clients 1.61 401.3 *** Routinely assess for guns with suicidal clients 1.57 381.6 *** Routinely discuss firearm safety if client is depressed 1.50 348.5 *** Media coverage has motivated Clients not at risk to harm self 1.43 320.6 *** Clients not at risk to harm others -1.58 -79.4 *** Clients safer with guns -1.31 -73.0 *** Not adequately trained -1.23 -70.7 * Not aware of suicide/homicide -1.11 -67.0 *** risks associated with guns -.92 -60.3 *** Not responsibility of social worker -.89 -59.0 *** Uncomfortable with topic More important topics -.84 -56.6 *** Would counsel more if materials -.75 -52.7 *** available .33 39.8 My advice doesn't matter Concerned I will offend client -.35 -29.5 Counseling clients on firearm -.28 -24.0 safety would be effective .20 22.0 Not enough time Proper training would give me -.16 -14.9 credibility .03 3.0 Likely to support gun control Firearm counseling is major .03 2.8 public health issue .02 1.6 Clients won't be truthful about guns .004 .4 Counsel Change in Probability Variable b (%) Routinely counsel on firearm safety with suicidal clients 1.33 Routinely assess for guns with suicidal clients 1.20 232.9 ** Routinely discuss firearm safety if client is depressed 1.71 450.3 *** Media coverage has motivated Clients not at risk to harm self 1.73 465.9 *** Clients not at risk to harm others -1.15 -68.4 *** Clients safer with guns -0.84 -56.0 ** Not adequately trained -1.67 -81.1 *** Not aware of suicide/homicide -2.00 -86.4 *** risks associated with guns -.75 -52.9 * Not responsibility of social worker -1.43 -76.2 *** Uncomfortable with topic More important topics -.75 -52.6 * Would counsel more if materials -1.18 -69.3 *** available .23 25.6 My advice doesn't matter Concerned I will offend client -.52 -40.5 * Counseling clients on firearm -.02 -1.5 safety would be effective .66 94.0 * Not enough time Proper training would give me -1.08 -66.0 *** credibility .30 34.3 Likely to support gun control Firearm counseling is major -.25 -21.8 public health issue 1.70 18.2 Clients won't be truthful about guns -.56 -42.8 * * p [less than or equal to] 05. ** p [less than or equal to] .01. *** p = [less than or equal to] .001. Table 2: Multivariate Predictors of Firearm Assessment and Counseling Assessment Counseling Change in Change in Probability Probability Variable b (%) b (%) Firearm safety 1.32 275.7 *** 1.76 479.3 *** training Client injured .54 71.1 * .25 27.85 or killed by gun Gender .50 65.0 * .55 73.7 * Licensure status .31 35.8 * .21 22.8 Urban .29 33.0 -.08 -7.3 Suburban .12 12.7 -.14 -13.4 Gun access/ownership -.09 -8.8 -.13 -12.4 Race .04 4.1 .06 6.2 Age -.01 -1.35 -.006 -.6 Male clients .003 .3 .001 .1 Note: An underlying assumption of ordered logit model, known as the parallel regression assumption, states that the slope coefficient estimates generated by the series of binary regressions are equal and only the intercept changes (Long, 1997). This assumption can be tested with an approximate likelihood-ratio test developed by Wolfe and Gould (1998; as described by Long and Freese, 2001) for the Stata statistical software package (StataCorp, 2005). Violation of the parallel regression assumption is not indicated in this model. * ps [less than or equal to] .05. ** p [less than or equal to] .01. *** ps [less than or equal to] 001.