Perceived Safety of Adults with Severe Mental Illness At a Community-Based Outpatient Program

Purpose:  Both actual and perceived safety may be influential components of environments that support healing and behavior change for adults with severe mental illness, particularly as this population typically has increased rates of victimization, homelessness, and lack of social support. Lower levels of perceived safety have been shown to predict higher levels of mental distress with homeless adults,  9/11 first responders, and urban youth living in high stress schools. Women generally report lower levels of perceived safety than men, though gender differences are debated in the literature. However, most previous studies have used limited measures of perceived safety that include one to three likert questions on perceived safety yet fail to capture other dimensions of safety, including feelings, judgments, and current levels of perceived safety, actual experiences of harm, and suggestions for safety improvements. The researchers searched for, but could not locate, a more dimensional scale of perceived safety for use in a mental health outpatient program. Thus, the following study developed and piloted a Perceived Safety Questionnaire (PSQ)  (approximately six pages of quantitative and qualitative questions on actual and perceived safety). This study addresses the following preliminary questions: What are the actual incidences of harm experienced by clients at a community-based outpatient program? What are clients' reported levels of perceived safety, including feeling, judgments, and values around safety?  Are there gender differences in these? Is actual harm correlated with levels of perceived safety, when controlling for gender, ethnicity, and age?

Methods:  This study used a self-report survey design to administer the Perceived Safety Questionnaire (PSQ). Sixty adults (31 men, 29 women, 33% white, 30% African American, 17% Asian/Pacific Islander, 12% Latino, 8% other) were purposively sampled from a community-based mental health outpatient program in San Francisco serving clients with severe mental illness who were frequent users of city crisis services. The researcher read each questionnaire item out loud and wrote down the client responses. Each participant was offered a $5 gift certificate to Walgreens upon completion of the questionnaire. 

Results:  17% of clients reported having experienced at least one incidence of physical harm at the program, 40%  reported experiencing sexual harassment at least once, and 52% reported experiencing emotional harassment at least once. Overall, two-thirds of clients reported feeling generally unsafe in at least one aspect of the program. 70% reported feeling generally unsafe some or all of the time in the neighborhood around the program, 50% reported feeling generally unsafe around other clients, and 35% reported feeling generally unsafe in the program milieu. Females were more likely to report lower perceived safety than males (F=.52, pvalue<.05). Clients who experienced emotional harassment were more likely to report lower levels of perceived safety (t=-2.573, pvalue<.001).  

Conclusions and Implications: Adults with severe mental illness are a vulnerable population who are often granted limited voice and agency in their recovery. A focus on perceived safety may help social workers and agencies to engage and empower this hidden stakeholder population. Implications for practice and policy will be discussed.