Community Participation and Belonging Among Formerly Homeless ...

Community Ment Health J (2014) 50:604–611 DOI 10.1007/s10597-013-9672-9

BRIEF REPORT

Community Participation and Belonging Among Formerly Homeless Adults with Mental Illness After 12 months of Housing First in Vancouver, British Columbia: A Randomized Controlled Trial Michelle L. Patterson • Akm Moniruzzaman Julian M. Somers

•

Received: 16 August 2012 / Accepted: 2 December 2013 / Published online: 10 December 2013 Ó Springer Science+Business Media New York 2013

Abstract This study examines community integration among homeless adults with mental illness 6 and 12 months after random assignment to Housing First (independent apartments or congregate residence) with support services or to treatment as usual (TAU). Residence in independent apartments was associated with increased ‘psychological integration’ for participants with less severe needs; however, no significant improvement in ‘physical integration’ was observed among any of the intervention groups. Analysis of individual items on the Psychological Integration subscale revealed that, compared to TAU, participants assigned to independent apartments were more likely to endorse statements related to the emotional components of community but not statements related to neighboring. Participants assigned to the congregate residence were more likely to endorse knowing their neighbors, but not interacting with neighbors or the emotional components of community. Findings are discussed in terms of housing program as well as broader contextual factors. Keywords Community integration Homelessness Severe mental illness

M. L. Patterson (&) A. Moniruzzaman J. M. Somers Faculty of Health Sciences, Simon Fraser University, 8888 University Drive, Blusson Hall Rm. 11010, Burnaby, BC V5A 1S6, Canada e-mail: [email protected] A. Moniruzzaman e-mail: [email protected] J. M. Somers e-mail: [email protected]

123

Introduction In the wake of deinstitutionalization, supported housing was developed to allow people with severe mental illness (SMI) to live in regular neighborhoods throughout the community. Off-site support was provided to facilitate housing stability and recovery goals, with the hope that people with SMI would participate and ‘‘integrate’’ into the mainstream community (Carling 1990). Housing First was developed to reach the chronically homeless, who often experience social exclusion as well as SMI and are seen as the ‘‘hardest to house’’ (Tsemberis and Eisenberg 2000). Housing First offers homeless people with SMI immediate access to independent apartments in various neighborhoods with a market lease, without any requirements around substance use or psychiatric treatment. Clients have access to treatment and a variety of services, but retain the right to choose their level of participation. Researchers from various disciplines have tried to define and measure community integration (CI) as well as related concepts such as social inclusion and citizenship (Dorvil et al. 2005; Ware et al. 2007; Wong and Solomon 2002). CI was originally conceptualized as physical presence in a neighborhood and was operationalized as the cumulative frequency of self-initiated participation and use of community resources (Segal et al. 1980). More recently, CI has been defined as a multi-dimensional construct including physical, social, and psychological integration (Nemiroff et al. 2011; Wong and Solomon 2002). Social integration refers to engagement in social interactions and developing social networks. Psychological integration, also called sense of community (MacMillan and Chavis 1996), refers to feeling that one belongs to a larger collective and includes perceived emotional safety, mutual benefit, and trust. This multi-dimensional approach

Community Ment Health J (2014) 50:604–611

emphasizes location, level of functioning, and normative social roles. Given that CI has been predominantly explored among housed individuals with SMI, it is not clear how people who have been homeless and marginalized from a variety of civic spaces, experiences, and social groups conceptualize and engage in communities (Townley and Kloos 2009; Yanos 2007). Moreover, the traditional definition of community as a monolithic based on location no longer applies to contemporary urban experience where a wide variety of niche communities are available to express unique aspects of one’s identity (Manzo 2003). The relationship between housing and CI is complex (Schutt 2011; Wolf et al. 2001). Living in independent apartments has been associated with greater satisfaction with housing compared to other residential settings, however, positive effects on other outcomes such as symptoms, physical health and substance use are mixed and may be moderated by variables such as neighborhood characteristics and tenant mix in a building (Schutt 2011). Among 95 adults with SMI receiving Assertive Community Treatment in small and mid-size Canadian towns, Prince and Prince (2002) found that perceptions of stigma were inversely related to self-reported psychological integration, particularly among participants with lower psychosocial functioning. However, it is not clear whether these findings generalize to adults with mental illness who have histories of homelessness in contemporary urban settings. Yanos et al. (2007) explored CI among a small sample of formerly homeless adults with mental illness after at least one year of stable housing in scattered-site or congregate residences in New York City. Perceived neighborhood social cohesion was strongly related to psychological integration but not to physical or social integration. Further, residence in independent apartments was significantly associated with greater independence and occupational functioning but no significant relationships were found between housing type and CI. Using a similar but larger sample, Gulcur et al. (2007) found that participants living in independent apartments reported greater perceived choice than those in congregate settings and that this was associated with higher levels of psychological integration. While interest in CI has grown, it is not clear how different housing environments facilitate CI, particularly psychological integration or sense of community, among homeless individuals with mental illness. Moreover, it is unknown how the relative severity or complexity of mental illness interacts with CI. Given the socially constructed nature of community and perceived belonging (Field 2011) and recent changes in urban communities, methods of assessing CI need to reflect current experiences. These experiences may be unique among formerly homeless people with SMI, who are often marginalized and may experience community differently

605

than the general population. When measuring community and social inclusion among this population, we must look more carefully at how different places and residential settings may influence participation in community as well as feelings of affiliation and belonging. This study focuses on self-report findings from a modified CI scale that was administered to 497 homeless adults with mental illness before random assignment to Housing First (scattered-site or congregate) or treatment as usual (TAU; no housing or supports through the study), and at 6 and 12 months follow-up. Informed by previous research, we predicted that physical integration would be higher among participants assigned to the congregate residence, but that psychological integration would be greater among participants assigned to independent apartments, and particularly among participants with less severe psychosocial needs.

Methods Participants and Sampling The Vancouver At Home Study is a randomized controlled trial involving homeless adults with mental illness in Vancouver, British Columbia. Study design and sample size were determined by the At Home/Chez Soi National Research Team, which coordinated activities at five different study sites. Details related to the study protocol, such as CONSORT, as well as measures and intervention arms have been reported elsewhere (Goering et al. 2011; Zabkiewicz et al. 2012). Eligibility criteria included legal adult status (19 years of age or over), presence of a current mental disorder on the MINI International Neuropsychiatric Interview (MINI; Sheehan et al. 1998), and being absolutely homeless or precariously housed (see Goering, et al. 2011 for details). Participants were recruited through referral from a wide variety of agencies providing services to homeless adults in Vancouver. (e.g., homeless shelters, drop-in centers, homeless outreach teams, hospitals, community mental health teams, criminal justice programs; see Somers et al. 2013 for details). Eligibility criteria were screened via telephone with the referral agent followed by, if appropriate, a face-to-face interview with the potential participant to more formally assess eligibility. All interviews were conducted by trained research interviewers who explained procedures and obtained informed consent. Institutional Research Ethics Board approval was obtained through Simon Fraser University and the University of British Columbia. If all eligibility criteria were met, an individual was enrolled as a participant and the baseline interview commenced consisting of a series of interviewer-administered questionnaires

123

606

addressing socio-demographic characteristics, mental illness, substance use, physical health, service use, community integration, and quality of life. All participants received a cash honorarium upon completion of the screener and the baseline interview. After completion of the baseline interview, 497 participants were randomly assigned to one of five study arms via computer algorithm based on their level of need. Level of need (‘‘high’’ or ‘‘moderate’’) was determined by a computer algorithm. The High Need (HN) group was based on a Multnomah Community Ability Scale (Barker et al. 1994) score of 62 or lower and current (hypo) manic episode or psychotic disorder on the MINI, as well as at least one of the following: legal involvement in the past year; substance dependence in the past month; two or more hospitalizations for mental illness in any one of the past 5 years. All other eligible participants were categorized as Moderate Needs (MN) (see Goering et al. 2011). Participants stratified as HN (n = 297) were assigned to one of three study arms: (1) Housing First with Assertive Community Treatment (ACT)1 (n = 90) (see footnote 1) in which participants could choose from up to three market lease apartments in a variety of neighborhoods and services were provided by a multi-disciplinary outreach team; (2) Congregate Housing2 with on-site support (see footnote 2) (CONG) (n = 107) in which participants had their own room and bathroom but shared amenity space with 100 other program participants and received three meals per day as well as activity programming and various health and social services on site; and (3) Treatment as Usual (HN-TAU) (n = 100) which provided no additional housing or support services beyond what was already available in the community. Participants stratified as MN (n = 200) were assigned to one of two study arms: (1) Housing First with Intensive Case Management (see footnote 1) (ICM) (n = 100) in which participants could choose from up to three market lease apartments in a variety of neighborhoods and services were provided by a team of outreach case managers who connected participants to existing services; and (2) MN-TAU (n = 100) as described above. This analysis is based on data from the baseline questionnaires from 297 HN participants and 200 MN participants recruited from the study’s inception in October 2009 to June 2011, and follow-up data at 6 and 12 months after the baseline interview. Based on baseline data, no differences were observed between the full sample (n = 497) and participants who were subsequently contacted for follow-up (n = 471; see Somers et al. 2013 for details).


Variables of Interest Demographic variables including age, ethnicity, education, duration of homelessness, as well as frequency of illicit drug use and chronic medical conditions were assessed by selfreport. Mental disorders, including substance dependence, were assessed using the MINI. ‘Less severe cluster’ included at least one of Major Depressive Episode, Panic Disorder, and Posttraumatic Stress Disorder; ‘severe cluster’ included at least one of Psychotic Disorder, Mood Disorder with Psychotic Features, and Manic/Hypomanic Episode. Daily illicit drug use over the past month was assessed using the Maudsley Addiction Profile (Marsden et al. 1998). Medical conditions were assessed using a self-report checklist of 30 chronic health conditions (lasting longer than 6 months). Infectious disease included positive self-report diagnosis of HIV, Hepatitis B or Hepatitis C. Based on baseline data, no significant differences were observed on the above variables between the full sample (n = 497) and participants who were subsequently contacted for follow-up (n = 471, p [ 0.05). The Community Integration Scales (CIS; Aubry and Myner 1996) were originally developed for individuals with SMI, and were modified for inclusion in the national study through cognitive interviewing to increase the relevance for homeless adults. Recommendations were made for item revisions and/or adjustments to the administration process based on relevance and comprehension (see Adair et al. 2011). The original scales consist of 37 items across three subscales: (1) Physical Integration originally assessed involvement in 12 activities over the past month. After pretesting, this subscale was reduced to seven items with four response options (yes, no, don’t know, declined) resulting in a maximum score of 7. (2) Psychological Integration originally assessed sense of belonging in one’s neighborhood using 12 items; four items were retained after pre-testing with response options consisting of a 5-point Likert scale from ‘‘strongly agree’’ to ‘‘strongly disagree’’ as well as ‘‘don’t know’’ and ‘‘declined,’’ resulting in a maximum score of 20. (3) Social Integration (13 items) was entirely eliminated after pre-testing due to lack of relevance for individuals who are homeless (e.g., receiving a ride from a neighbor, taking care of a neighbor’s home while they are away). Therefore, the modified CIS contained 11 items across two subscales. Analyses

1

Support services were available to participants assigned to ACT, CONG, and ICM but were not mandatory. The only requirement for housing was compliance with the terms of the rental lease and weekly visits with a case manager to ensure safety and well-being (Stefancic and Tsemberis 2007). 2 The CONG residence was located in downtown Vancouver in a neighborhood consisting of primarily businesses, including an innercity hospital and a number of affluent condominiums.

123

Bivariate comparisons of continuous variables (e.g., age, duration of homelessness) were conducted using t tests or one-way ANOVA as appropriate. Comparisons of categorical variables between groups were conducted using Pearson’s Chi square test. Due to the longitudinal nature of the data (repeated measures for each participant), generalized


estimating equations (GEE) with an exchangeable correlation structure were used to model the independent association between study arm and CI subscale scores. Effects for HN and MN study arms (compared to the relevant TAU group) were estimated separately in both univariate and multivariable models. Subscale total scores were analyzed separately using GEE linear regression analysis (Gaussian distribution with an identity link). Many people who live in contemporary urban settings do not know or regularly interact with their neighbors, particularly single people who live in apartment buildings (Vancouver Foundation 2012). Therefore, we separately examined the four items from the Psychological Integration subscale (‘‘I know most of the people who live near me,’’ ‘‘I interact with the people who live near me,’’ ‘‘I feel at home where I live,’’ and ‘‘I feel like I belong where I live’’). The responses to these items were categorized into binary outcomes (agreement or disagreement). Agreement consisted of three responses (‘‘strongly agree,’’ ‘‘agree,’’ and ‘‘neither’’) while disagreement consisted of ‘‘strongly disagree’’ and ‘‘disagree’’. The proportion of agreement was modeled using GEE logistic regression analysis (binomial distribution with a logit link). For total subscale scores as well as individual items, multivariable models were developed for HN and MN groups separately controlling for the following variables (continuous unless otherwise indicated): follow-up time, age at enrolment, lifetime duration of homelessness, age first homeless, gender (male, female), ethnicity (Aboriginal, Caucasian, other), marital status (single, other), education (incomplete high school, other), Mental Disorder—less severe cluster (yes, no), Mental Disorder—severe cluster (yes, no), Mental Disorder— 2 or more (yes, no), Substance Dependence (yes, no), multiple medical conditions (yes, no), infectious disease (yes, no), daily illicit drug use (yes, no), and detained by police (yes, no). These variables were chosen based on past research and a priori hypothesized relationships between Housing First and CI. Mean substitution for individual missing items was used to obtain the total score on both subscales if no more than half the items were missing (recommended by Aubry, personal communication, 3 November 2011). Missing data for each of the two CIS subscales across three time points (baseline, 6, 12 months) ranged from 0.6 % to 4.2 %. Missing values for other covariates, ranging from 0 to 4 %, were not included in the analysis.

Results Sample Description Of the 497 participants interviewed at baseline, 457 (92 %) were re-interviewed at 6 months and 444 (89 %) at 12 months follow-up. Overall, 471 (95 %) participants

607

completed either the 6 or 12 months follow-up interview and 430 (87 %) completed both interviews. At baseline, the mean age was 40.8 (SD = 11.0) years; the mean age when first homeless was 30.3 (SD = 13.3) years; and the majority of the sample was male (73 %) and Caucasian (56 %). In the month prior to enrollment, 58 % of participants met criteria for Substance Dependence and 25 % reported daily illicit drug use. With regards to mental disorders on the MINI, 73 % met criteria for at least one ‘‘severe’’ disorder and 53 % met criteria for at least one ‘‘less severe’’ disorder. Community Integration Subscale Scores With regard to the Physical Integration subscale, there was no significant change in the mean score over time for any of the HN or MN groups (p [ 0.05). All Housing First and TAU groups reported engaging in approximately two of the seven activities in the past month at each time point. However, in the multivariable model, a trend toward greater physical integration was observed for the CONG group at 12 months (p = 0.076). On the Psychological Integration subscale, significant improvement was observed at both 6 and 12 months for the ICM group compared to MNTAU after controlling for potential confounders (13.1 vs. 11.5 at 6 months; 12.6 vs. 11.9 at 12 months; p B 0.05). However, there was no significant increase in mean subscale score over time for either of the HN Housing First groups compared to HNTAU (p [ 0.05). Psychological Integration Item Analysis Psychological integration was further examined by comparing the proportions of each study arm that endorsed each of the four items comprising this subscale using GEE logistic regression (see Table 1). Compared to HNTAU, the CONG group was significantly more likely to agree with the statement ‘‘I know most of the people who live near me’’ (AOR = 1.54) whereas the ACT group was significantly less likely to agree with the same statement (AOR = 0.65). Compared to HNTAU, neither CONG nor ACT groups were more likely to report interacting with the people who live near them (AOR = 1.00 and 0.88 respectively). For the MN groups, ICM participants were no more likely than MNTAU to report knowing most neighbors (AOR = 0.70) or interacting with neighbors (AOR = 0.97). The ACT group was significantly more likely than HNTAU to agree with feeling at home where they live (AOR = 1.77) and with feeling like they belong where they live (AOR = 1.56). Compared to HNTAU, CONG participants were not more likely to agree with feeling at home (AOR = 1.10) or feeling like they belong where they

123

123

live (AOR = 1.24). The ICM group was also significantly more likely than MNTAU to agree with feeling at home where they live (AOR = 2.05) and with feeling like they belong where they live (AOR = 1.99; see Table 1).

Discussion

* p B 0.05

Multivariable models were conducted separately for HN and MN study arms. Each model was controlled for the following variables (continuous unless otherwise indicated): follow-up time, age at enrolment, lifetime duration of homelessness, age first homeless, gender (male, female), ethnicity (Aboriginal, Caucasian, other), marital status (single, other), education (incomplete high school, other), Mental Disorder—less severe cluster (yes, no), Mental Disorder—severe cluster (yes, no), Mental Disorder—2 or more (yes, no), Substance Dependence (yes, no), multiple medical conditions (yes, no), infectious disease (yes, no), daily illicit drug use (yes, no), and detained by police (yes, no)

1.99 (1.31, 3.03)*

Reference Reference

1.71 (1.19, 2.46)* 2.05 (1.32, 3.17)*

Reference Reference

1.65 (1.15, 2.38)* 0.97 (0.67, 1.40)

Reference Reference

0.92 (0.65, 1.30) 0.70 (0.47, 1.06)

Reference Reference MNTAU

Reference

MN study arms (n = 185) ICM 0.75 (0.52, 1.10)

1.56 (1.05, 2.33)*

Reference Reference Reference Reference Reference Reference

1.61 (1.12, 2.33)*

HNTAU

Reference

1.24 (0.82, 1.86) 1.16 (0.81, 1.68) 1.10 (0.72, 1.69)

1.77 (1.14, 2.77)* 1.70 (1.15, 2.53)*

1.05 (0.72, 1.55) 1.00 (0.66, 1.51)

0.88 (0.60, 1.29) 1.00 (0.69, 1.43)

1.02 (0.71, 1.49)

0.69 (0.47, 1.01)* ACT

1.54 (1.00, 2.36)* 1.59 (1.08, 2.36)* CONG

HN study arms (n = 286)

AOR (95 % CI)

0.65 (0.43, 0.97)*

UOR (95 % CI) AOR (95 % CI) UOR (95 % CI) AOR (95 % CI) UOR (95 % CI) UOR (95 % CI)

I feel at home where I live I interact with the people who live near me I know most of the people who live near me Item

Table 1 GEE logistic regression analysis for HN and MN study arms by individual items on the Psychological Integration subscale

AOR (95 % CI)


I feel like I belong where I live

608

At follow-up, the majority of Housing First participants reported moderate levels of physical and psychological integration within their new neighborhoods. No differences were observed on the Physical Integration subscale between HN or MN scattered-site Housing First participants and their respective control groups (no housing or supports through the study) over time; however, improvements among the CONG group were marginally significant at 12 months. With regard to Psychological Integration, no differences were observed in total subscale score between the three HN groups; however, among MN participants, the ICM group showed significantly higher subscale total scores at both 6 and 12 months compared to the control group. These findings confirm previous research that shows that psychological integration is particularly challenging for individuals with high levels of psychosocial impairment (Prince and Prince 2002). Psychological Integration was explored further by looking at items comprising the subscale. Participants in the CONG group, who were living with 100 fellow participants, were more likely to agree with knowing most of their neighbors but not interacting with their neighbors compared to HNTAU. ACT participants, who were living in independent apartments throughout the city, significantly disagreed with knowing most of their neighbors. Not surprisingly, living with peers in a building with shared amenity space and meals allowed for opportunities to get to know one’s neighbors. However, ACT participants were more likely than HNTAU to agree with statements that reflect the emotional component of community (feeling at home and like they belong where they live), whereas there was no significant difference between CONG and HNTAU on these items. Participants in the ICM group were no more likely to agree with the two neighboring statements than their counterparts in MNTAU; however, they were more likely to agree with the two emotional statements than the control group. These results expand previous findings indicating that many people with mental illness who have been homeless face challenges when transitioning to independent apartments and living alone (Yanos et al. 2004). Many formerly homeless people who are stably housed continue to be marginalized by poverty, are excluded from the labor market, and live in social isolation (Nemiroff et al. 2011). Longstanding histories of residential instability are


common among this population and have been shown to restrict one’s ability to make social ties and participate in community (Schutt 2011). However, low levels of interaction with neighbors may be part of a broader phenomenon affecting the general population, particularly in large urban settings (Colombo et al. 2001). A recent survey of 3,841 residents in Metro Vancouver found that one-third reported difficulty making new friends, 40 % of renters did not know the names of at least two neighbors, and most respondents had not participated in a community event in the past year (Vancouver Foundation 2012). A low level of community connection and engagement in the general population would make it very difficult for people with mental illness and histories of homelessness who are living in scattered-site apartments to develop new social ties. Despite these challenges, it is encouraging that participants with both high and moderate needs assigned to scattered-site apartments reported feeling at home and a sense of belonging in their new neighborhoods. Previous research suggests that this sense of affiliation may be related to perceived choice (Gulcur et al. 2007). Participants in the ACT and ICM groups could choose from up to three apartments from various neighborhoods. People who are homeless typically have few choices, and those they do have are typically constrained by structural factors such as poverty and inequity (Dovey et al. 2001). The availability of safe, good quality housing may also have contributed to the sense of affiliation observed in the scattered-site groups (Nemiroff et al. 2011). Without the rental supplement provided through the project, most participants could not afford good quality housing in Vancouver, which for many years has had a vacancy rate less than one percent. Prior to enrollment, the majority of participants lived in the innercity, which has a high concentration of rooming houses and hotels. While many of these buildings have been renovated in recent years, the quality of housing is generally very poor. Furthermore, living in these buildings is accompanied by considerable stigma for many people who are homeless and have mental illness (Fast et al. 2010; Patterson et al. 2012). Another factor that may contribute to the observed sense of belonging is the intensive off-site supports available to participants assigned to scattered-site Housing First. Participants assigned to scattered-site housing received intensive outreach supports from an ACT team (high needs) or an ICM team (moderate needs). Support workers met with participants in their new neighborhoods and worked toward recovery goals which included connecting people to resources and meaningful activities. However, despite these supports, participants did not report improvement in physical integration (e.g., use of community resources), and many reported activities were linked to program involvement (e.g., group outings). The exception was

609

participants in the CONG group, who had access to organized sporting events, art groups, and volunteer activities on-site. Further, the community of peers in the building may have increased the likelihood of participant engagement both in and outside the building. Nevertheless, this subscale may need revision to increase its relevance for contemporary urban populations and for people living in poverty. For example, many people no longer watch movies in theatres or attend church or concerts on a monthly basis. Inclusion of social networking sites and engagement in communities that are not based around one’s place of residence may also increase the relevance of this subscale. Acculturation into a homeless lifestyle might be a barrier to CI, particularly for participants with long histories of homelessness and whose social contacts revolve around substance use (Vangeest and Johnson 2002). The social exclusion and marginalization that often accompanies homelessness may generate emotional distress and coping strategies that have profound effects on multiple areas of functioning (Fast et al. 2010). Given the long duration of homelessness in our sample (49 % had been homeless for a total of 3 years or more), one year of stable housing is a short period to adjust to living indoors and in a new neighborhood. Initially, people may feel more connected to homeless peers or professionals but after a period of adjustment and as their recovery progresses, they may begin to participate and feel supported in more diverse settings within the broader community (Wong et al. 2006). Thus, it is important to view CI as a dynamic, non-linear process that includes building independent living skills and overcoming various psychosocial barriers. Despite the strengths of our sample (e.g., low attrition rates, structured diagnostic interviews, randomized controlled trial), several limitations must be considered. Our measure of CI was based on an instrument originally developed for adults with SMI in 1996 and was not very relevant for homeless or formerly homeless people in a contemporary urban setting. As a result, a number of items were removed, including the entire Social Integration subscale (Adair et al. 2011); no psychometric data are available for the modified scale. Also, it is not clear whether our findings generalize to congregate residences in geographic or urban locations different to the one in our study. Participants were assigned to one congregate residence so we were not able to compare the effects of different locations or the impact of choice of location or tenant mix on CI. Future research could explore objective and subjective assessments of neighborhood characteristics, housing quality, tenant mix, and how these variables relate to CI in different housing settings. Even once someone receives good-quality housing, poverty and lack of meaningful

123

610

activities can still dictate their daily choices and sense of belonging. Further qualitative research could explore the dynamic and conflicting experiences that influence how people with histories of homelessness experience living in various different neighborhoods and communities. Most mental health and housing research has focused on developing services to change individuals rather than potential changes to communities or society that would support marginalized individuals. Access to safe, affordable housing and treatment for mental disorders and substance use is critical to social inclusion as well as access to employment and recreation opportunities. Service providers and policy makers have the opportunity to facilitate social inclusion by ensuring services are not class-based. This includes expanding services for homeless individuals to communities outside of poor neighborhoods as well as mobile outreach and service delivery programs. Social action is not just required to increase the participation of marginalized citizens but also to modify the power relationships between various social groups. Acknowledgments Special thanks to the service teams who are working with participants to increase their sense of community. The authors also thank the At Home/Chez Soi Project collaborative at national and local levels; National project team: J. Barker, PhD (2008–2011) and C. Keller, National Project Leads; P. Goering, RN, PhD, Research Lead; approximately 40 investigators from across Canada and the US; 5 site coordinators; numerous service and housing providers; and persons with lived experience. Conflict of interest This research was funded by Health Canada and the Mental Health Commission of Canada. The views expressed herein solely represent the authors.

References Adair, C., Holland, A., Patterson, M., Mason, K., Hwang, S., & Goering, P. (2011). Cognitive interviewing methods for questionnaire pre-testing in homeless persons with serious mental disorders. Journal of Urban Health, 89, 36–52. Aubry, T., & Myner, J. (1996). Community integration and quality of life: A comparison of persons with psychiatric disabilities in housing programs and community residents who are neighbours. Canadian Journal of Community Mental Health, 15, 5–20. Barker, S., Barron, N., McFarland, B. H., et al. (1994). A community ability scale for chronically mentally ill consumer. Community Mental Health Journal, 30, 459–472. Carling, P. (1990). Major mental illness, housing, and supports: The promise of community integration. American Psychologist, 45, 969–975. Colombo, M., Mosso, C., & DePiccoli, N. (2001). Sense of community and participation in urban contexts. Journal of Community and Applied Social Psychology, 11, 457–464. Dorvil, H., Morin, P., Beaulieu, A., & Robert, D. (2005). Housing as a social integration factor for people classified as mentally ill. Housing Studies, 20, 497–519. Dovey, K., Fitzgerald, J., & Choi, Y. (2001). Safety becomes danger: Dilemmas of drug-use in public space. Health and Place, 7, 319–331.

123

Community Ment Health J (2014) 50:604–611 Fast, D., Shoveller, J., Shannon, K., & Kerr, T. (2010). Safety and danger in downtown Vancouver: Understandings of place among young people entrenched in an urban drug scene. Health and Place, 6(1), 51–60. Field, D. (2011). Emotional refuge? Dynamics of place and belonging among formerly homeless individuals with mental illness. Emotion, Space and Society, 4, 258–267. Goering, P., Streiner, D., Adair, C., Aubry, T., Barker, J., Distasio, J., et al. (2011). The At Home/Chez Soi trial protocol: a pragmatic, multi-site, randomised controlled trial of a Housing First intervention for homeless individuals with mental illness in five Canadian cities. BMJ Open, 1(2), e000323. Gulcur, L., Tsemberis, S., Stefancic, A., & Greenwood, R. (2007). Community integration of adults with psychiatric disabilities and histories of homelessness. Community Mental Health Journal, 43, 211–228. MacMillan, D. W., & Chavis, D. M. (1996). Sense of community: A definition and theory. Journal of Community Psychology, 14, 6–23. Manzo, L. (2003). Beyond house and haven: Toward a revisioning of emotional relationships with places. Journal of Environmental Psychology, 23, 47–61. Marsden, J., Gossop, M., Stewart, D., Best, D., Farrell, M., Strang, J. (1998). The Maudsley addiction profile (MAP): A brief instrument for treatment outcome research. Development and user manual. London UK, National Addiction Centre/Institute of Psychiatry. Accessed on April 25, 2012 at http://www.iop.kcl.ac. uk/iopweb/blob/downloads/locator/l_346_MAP.pdf. Nemiroff, R., Aubry, T., & Klodawsky, F. (2011). From homelessness to community: Psychological integration of women who have experienced homelessness. Journal of Community Psychology, 39, 1003–1018. Patterson, M., Markey, M., & Somers, J. (2012). Multiple paths to just ends: Using narrative interviews and timelines to explore health equity and homelessness. International Journal of Qualitative Methods, 11(2), 132–151. Prince, P., & Prince, C. (2002). Perceived stigma and community integration among clients of assertive community treatment. Psychiatric Rehabilitation Journal, 25, 323–331. Schutt, R. (2011). Homelessness, housing, and mental illness. Cambridge, Massachusetts: Harvard University Press. Segal, S., Baumohl, J., & Moyles, E. (1980). Neighborhood types and community reaction to the mentally ill: A paradox of intensity. Journal of Health and Social Behavior, 21, 345–359. Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., et al. (1998). The mini-international neuropsychiatric interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. Journal of Clinical Psychiatry, 59(Suppl 20), 22–33. Somers, J., Patterson, M., Moniruzzaman, A., et al. (2013). Vancouver at home: Pragmatic randomized trials investigating Housing First for people who are homeless and mentally ill. Trials, 14, 365. Stefancic, A., & Tsemberis, S. (2007). Housing First for long-term shelter dwellers with psychiatric disabilities in a suburban county: a 4-year study of housing access and retention. Journal of Primary Prevention, 28, 265–279. Townley, G., & Kloos, B. (2009). Development of a measure of sense of community for individuals with serious mental illness residing in community settings. Journal of Community Psychology, 37, 362–380. Tsemberis, S., & Eisenberg, R. (2000). Pathways to housing: Supported housing for street-dwelling homeless individuals with psychiatric disabilities. Psychiatric Services, 51, 487–493. Vancouver Foundation (2012). Connections and engagement: A survey of metro Vancouver. Sentis Market Research. Retrieved

Community Ment Health J (2014) 50:604–611 July 5, 2012 from http://www.vancouverfoundation.ca/connect% 2Dengage/2012survey.htm. Vangeest, J., & Johnson, T. (2002). Substance abuse and homelessness: Direct or indirect effects? Annals of Epidemiology, 12(7), 455–461. Ware, N., Hopper, K., Tugenberg, T., Dickey, B., & Fisher, D. (2007). Connectedness and citizenship: redefining social integration. Psychiatric Services, 58(4), 469–474. Wolf, J., Burnam, A., Koegel, P., Sullivan, G., & Morton, S. (2001). Changes in subjective quality of life among homeless adults who obtain housing: A prospective examination. Social Psychiatry and Psychiatric Epidemiology, 36, 391–398. Wong, Y., Metzendorf, D., & Min, S. (2006). Neighborhood experiences and community integration: Perspectives from mental health consumers and providers. Social Work in Mental Health, 4, 45–59. Wong, Y., & Solomon, P. (2002). Community integration of persons with psychiatric disabilities in supportive independent housing: A conceptual model and methodological considerations. Mental Health Services Research, 4, 13–28.

611 Yanos, P. (2007). Beyong ‘‘Landscapes of Despair’’: The need for new research on the urban environment, sprawl, and the community integration of persons with severe mental illness. Health and Place, 13, 672–676. Yanos, P., Barrow, S., & Tsemberis, S. (2004). Community integration in the early phase of housing among homeless persons diagnosed with severe mental illness: Successes and challenges. Community Mental Health Journal, 40, 133–150. Yanos, P., Felton, B., Tsemberis, S., & Frye, V. (2007). Exploring the role of housing type, neighborhood characteristics, and lifestyle factors in the community integration of formerly homeless persons diagnosed with mental illness. Journal of Mental Health, 16, 703–717. Zabkiewicz, D., Patterson, M., Frankish, C. J., & Somers, J. (2012). The Vancouver at home study: overview and methods of a Housing First trial among individuals who are homeless and living with mental illness. Journal of Clinical Trials,. doi:10. 4172/2167-0870.1000123.

123