Accommodation‐based interventions for individuals experiencing, or at risk of experiencing, homelessness

Ciara Keenan; Sarah Miller; Jennifer Hanratty; Terri Pigott; Jayne Hamilton; Christopher Coughlan; Peter Mackie; Suzanne Fitzpatrick; John Cowman

doi:10.1002/cl2.1165

. 2021 May 18;17(2):e1165. doi: 10.1002/cl2.1165

Accommodation‐based interventions for individuals experiencing, or at risk of experiencing, homelessness

Ciara Keenan ^1,^✉, Sarah Miller ¹, Jennifer Hanratty ¹, Terri Pigott ², Jayne Hamilton ¹, Christopher Coughlan ¹, Peter Mackie ³, Suzanne Fitzpatrick ⁴, John Cowman ⁵

PMCID: PMC8356295 PMID: 37131929

Abstract

Background

Globally, almost 1.6 billion individuals lack adequate housing. Many accommodation‐based approaches have evolved across the globe to incorporate additional support and services beyond delivery of housing.

Objectives

This review examines the effectiveness of accommodation‐based approaches on outcomes including housing stability, health, employment, crime, wellbeing, and cost for individuals experiencing or at risk of experiencing homelessness.

Search Methods

The systematic review is based on evidence already identified in two existing EGMs commissioned by the Centre for Homelessness Impact (CHI) and built by White et al. The maps were constructed using a comprehensive three stage search and mapping process. Stage one mapped included studies in an existing systematic review on homelessness, stage two was an extensive search of 17 academic databases, three EGM databases, and eight systematic review databases. Finally stage three included web searches for grey literature, scanning reference lists of included studies and consultation with experts to identify additional literature. We identified 223 unique studies across 551 articles from the effectiveness map on 12th April 2019.

Selection Criteria

We include research on all individuals currently experiencing, or at risk of experiencing homelessness irrespective of age or gender, in high‐income countries. The Network Meta‐Analysis (NMA) contains all study designs where a comparison group was used. This includes randomised controlled trials (RCTs), quasi‐experimental designs, matched comparisons and other study designs that attempt to isolate the impact of the intervention on homelessness. The NMA primarily addresses how interventions can reduce homelessness and increase housing stability for those individuals experiencing, or at risk of experiencing, homelessness. Additional outcomes are examined and narratively described. These include: access to mainstream healthcare; crime and justice; employment and income; capabilities and wellbeing; and cost of intervention. These outcomes reflect the domains used in the EGM, with the addition of cost.

Data Collection and Analysis

Due to the diverse nature of the literature on accommodation‐based approaches, the way in which the approaches are implemented in practice, and the disordered descriptions of the categories, the review team created a novel typology to allow meaningful categorisations for functional and useful comparison between the various intervention types. Once these eligible categories were identified, we undertook dual data extraction, where two authors completed data extraction and risk of bias (ROB) assessments independently for each study. NMA was conducted across outcomes related to housing stability and health.Qualitative data from process evaluations is included using a “Best Fit” Framework synthesis. The purpose of this synthesis is to complement the quantitative evidence and provide a better understanding of what factors influenced programme effectiveness. All included Qualitative data followed the initial framework provided by the five main analytical categories of factors of influence (reflected in the EGM), namely: contextual factors, policy makers/funders, programme administrators/managers/implementing agencies, staff/case workers and recipients of the programme.

Main Results

There was a total of 13,128 people included in the review, across 51 reports of 28 studies. Most of the included studies were carried out in the United States of America (25/28), with other locations including Canada and the UK. Sixteen studies were RCTs (57%) and 12 were nonrandomised (quasi‐experimental) designs (43%). Assessment of methodological quality and potential for bias was conducted using the second version of the Cochrane Risk of Bias tool for Randomised controlled trials. Nonrandomised studies were coded using the ROBINS‐ I tool. Out of the 28 studies, three had sufficiently low ROB (11%), 11 (39%) had moderate ROB, and five (18%) presented serious problems with ROB, and nine (32%) demonstrated high, critical problems with their methodology. A NMA on housing stability outcomes demonstrates that interventions offering the highest levels of support alongside unconditional accommodation (High/Unconditional) were more effective in improving housing stability compared to basic support alongside unconditional housing (Basic/Unconditional) (ES=1.10, 95% confidence interval [CI] [0.39, 1.82]), and in comparison to a no‐intervention control group (ES=0.62, 95% CI [0.19, 1.06]). A second NMA on health outcomes demonstrates that interventions categorised as offering Moderate/Conditional (ES= 0.36, 95% CI [0.03, 0.69]) and High/Unconditional (ES = 0.22, 95% CI [0.01, 0.43]) support were effective in improving health outcomes compared to no intervention. These effects were smaller than those observed for housing stability. The quality of the evidence was relatively low but varied across the 28 included studies. Depending on the context, finding accommodation for those who need it can be hindered by supply and affordability in the market. The social welfare approach in each jurisdiction can impact heavily on support available and can influence some of the prejudice and stigma surrounding homelessness. The evaluations emphasised the need for collaboration and a shared commitment between policymakers, funders and practitioners which creates community and buy in across sectors and agencies. However, co‐ordinating this is difficult and requires sustainability to work. For those implementing programmes, it was important to invest time in developing a culture together to build trust and solid relationships. Additionally, identifying sufficient resources and appropriate referral routes allows for better implementation planning. Involving staff and case workers in creating processes helps drive enthusiasm and energy for the service. Time should be allocated for staff to develop key skills and communicate engage effectively with service users. Finally, staff need time to develop trust and relationships with service users; this goes hand in hand with providing information that is up to date and useful as well making themselves accessible in terms of location and time.

Authors' Conclusions

The network meta‐analysis suggests that all types of accommodation which provided support are more effective than no intervention or Basic/Unconditional accommodation in terms of housing stability and health. The qualitative evidence synthesis raised a primary issue in relation to context: which was the lack of stable, affordable accommodation and the variability in the rental market, such that actually sourcing accommodation to provide for individuals who are homeless is extremely challenging. Collaboration between stakeholders and practitioners can be fruitful but difficult to coordinate across different agencies and organisations.

1. PLAIN LANGUAGE SUMMARY

1.1. Accommodation‐based approaches help people remain healthy and stably housed

Accommodation‐based approaches are mostly effective for increasing housing stability and health outcomes, except for those which offer low support housing without behavioural conditions. These approaches led to worse outcomes related to housing stability and health than receiving nothing at all. Agencies working together and sharing resources such as time and staff creates a commitment to ending homelessness.

1.1.1. What is this review about?

Globally, almost 1.6 billion individuals lack adequate housing. Many accommodation‐based approaches have evolved to incorporate support and services beyond delivery of housing. This review looked at whether these approaches are effective on outcomes including housing stability, health, employment, crime, wellbeing, and cost for individuals experiencing or at risk of experiencing homelessness.

What is the aim of this review?

This Campbell systematic review of qualitative and quantitative evidence examines how useful accommodation‐based approaches are for people experiencing homelessness. The quantitative data summarises evidence from 28 studies, reported in 51 articles, mainly from North America. The qualitative data summarises evidence from 10 articles from high‐income countries.

1.1.2. What studies are included?

The quantitative research provides an overview of effectiveness findings from 28 intervention studies reported in 51 articles of accommodation‐based interventions. Twenty‐five out of the 28 studies are from the United States, two from Canada and one from the UK. The quality of the research is generally low and represents important weaknesses in the evidence base.

The qualitative data presents one evaluation based on an intervention conducted in the UK, two in Ireland, one in Australia, one across Europe and the remaining five carried out in North America; three in the United States and two in Canada. The quality of the evaluations was average and did not directly evaluate the effectiveness interventions discussed in this review.

1.1.3. Do accommodation‐based approaches help people experiencing homelessness?

Interventions which provide the highest levels of support and do not place rules on the person receiving the intervention are best at improving housing stability and health outcomes.

Interventions which offer the lowest levels of support and do not place rules on the person might harm those individuals. For those individuals, housing stability and health outcomes were worse than for all other interventions, including individuals who are not receiving any intervention at all.

1.1.4. What implementation factors affect how well accommodation‐based approaches work?

Staff, resources and time often impacted the delivery of accommodation programmes most. Programme managers knew that members of staff working on the ground took initiative and were capable in their roles. However, they need adequate training and time to build good relations with service users.

There is a tension in funding allocated between new and established services, which can cause issues when services collaborate. It can also impact upon the shared commitment to ending homelessness. Buy‐in at all levels of influence can impact how successful a programme is and how many people experiencing homelessness it can engage with.

1.1.5. What do the findings of this review mean?

Those interventions which are described as Basic/Unconditional (i.e., those that only satisfy very basic human needs such as a bed and food) harm people: meaning they had worse health and housing stability outcomes even when compared to receiving nothing at all. This invites questions on whether these types of accommodation‐based interventions should be discontinued so that other more suitable and effective offers of support can be made available.

Too few studies assess the cost, or important participant characteristics like age and gender. There are also gaps related to where the research is conducted. Most of the studies included are from the United States and Canada which have very different social welfare systems to those of the UK. The process evaluations were conducted in high‐income countries with different housing contexts and social welfare systems.

The studies were of average quality and not connected to the effectiveness studies, which presented issues when drawing connections between the available data. Researchers conducting studies into accommodation‐based interventions should consider evaluating and publishing the factors impacting upon the trial, reflecting on why the intervention did or did not work, and for whom.

1.1.6. How up‐to‐date is this review?

Quantitative studies were downloaded from the effectiveness Evidence and Gap Map on 12 April 2019. Qualitative reports were downloaded from the Process and Implementation Evidence and Gap Map on 10 May 2019.

2. BACKGROUND

2.1. The problem, condition or issue

Homelessness affects individuals who are experiencing life without safe, adequate, or stable housing. Conceived in this way, homelessness not only describes those individuals who are visibly homeless and living on the street, but also those precariously housed individuals who; stay in emergency accommodation, sleep in crowded or inadequate housing, and those who are not safe in their living environment. FEANTSA further classify individuals experiencing homelessness as those who are roofless, those who are houseless and those who experience insecure or inadequate housing (FEANTSA, 2005).

Global data suggests that at least 1.6 billion people lack adequate housing (Habitat for Humanity, 2017). In the European context this figure continues to rise across all European Union member states except for Finland where homelessness has been on the decline since 1987 (FEANTSA, 2017; Y‐Foundation, 2017). Crisis, a charity based in the UK, estimated that in 2019 England acknowledged 57,890 households as homeless. In Wales, homelessness threatened 9,210 households and in Scotland, 34,100 individual applications were assessed for homelessness status (Fitzpatrick et al., 2016). Finally, Northern Ireland have an estimated 18,200 households experiencing homelessness according to a recent report (Fitzpatrick et al., 2020).

Without access to adequate housing, individuals experience multiple adverse effects including; exposure to disease, poverty, isolation, mental health issues, prejudice and discrimination, and are under constant and significant threat to their personal safety. Therefore, having access to safe, stable and adequate housing is internationally recognised as a basic human right (OHCHR, 2009) and is central to create the conditions whereby the population can live healthy, safe and happy lives.

2.2. The intervention

Homelessness is recognised as a multifaceted issue and many accommodation‐based approaches have evolved across the globe to incorporate additional support and services beyond delivery of housing, while other interventions deliver only temporary housing which is insufficient to meet people's basic needs. Through amalgamation of global ideas, the progression of evidence‐based policy and practice, and further establishment of welfare states, classification of accommodation‐based approaches is varied and represents the diversity in how the interventions were formed. The number of interventions which now exist, coupled with inconsistent descriptions of interventions and their elements (e.g., different models of housing, support services, expectations of engagement, etc.), has rendered current categorisations meaningless. Therefore, it was deemed necessary to group interventions based on their components, rather than their name. Later in this review, we describe how the review team created a novel and meaningful typology to categorise included interventions, however, initially we will briefly describe some of the familiar interventions that establish this evidence base.

2.2.1. Housing First

Housing First interventions offer housing to people experiencing homelessness with minimal obligation or preconditions being placed upon the participant. The Housing First programme, as conceived by Tsemberis (Tsemberis & Eisenberg, 2000), had clear principles which other researchers have since deviated from. However, most Housing First programmes share some common themes: (i) the participant is provided access to permanent housing immediately, without conditions, (ii) decisions around the location of the home and the services received are made by the user, (iii) support and services to aid the individual recovery are offered alongside housing placement, (iv) social integration with local community and meaningful engagement with positive activities is encouraged. Housing First is based on the principle that housing should be made available in the first instance and preconditions such as sobriety and involvement in treatment programmes are unnecessary barriers placed upon people who are homeless. Through the removal of these common obstacles, it is believed that the individual has a better chance of achieving stabilisation in appropriate housing and feeling more willing or able to accept treatment. In the original Pathways model of Housing First, housing provision is offered through scattered sites, which is where user choice is emphasised and housing is distributed (scattered) among existing rental properties. A key variation in the model has been the use of congregate housing where a property is reserved solely for the use of individuals experiencing homelessness. There is significant debate about the potential differences in effectiveness of these two models (Mackie et al., 2017).

2.2.2. Rapid rehousing

The rapid rehousing approach seeks to provide accommodation to individuals experiencing homelessness as quickly as possible. Generally, the rapid rehousing approach will identify available accommodation, aid with application, rent and moving in and the provision of case management to support access to other services. Rapid rehousing might provide the service user with a short‐term subsidy to assist with rent, rent in advance, help with rent arrears or help with moving. Generally, rapid rehousing targets those persons experiencing homelessness who have lower support needs and are less likely to require substantial access to services. The amount of support provided through a rapid rehousing approach is usually time limited.

2.2.3. Hostels

Hostels provide accommodation to meet short‐term housing needs. Homeless hostels often impose strict rules on the people who stay there relating to abstinence, behaviour and curfews. The individuals who use hostels vary but may include individuals, including those with pets, families and couples who are homeless. There is no clear definition on what constitutes a hostel and the provision offered will vary across councils, counties, and countries. Sleeping arrangements are variable too, with some offering dormitory style sleeping alongside communal kitchen, living, and shower areas while others have bedsit flats. The type of support offered by a homeless hostel is often determined by the resources available and individuals they can house. There are examples of in‐house support services such as: residents having a support plan to move to more stable accommodation; practical help with form filling and obtaining necessary governmental documents to continue education or gain employment; or treatment for substance abuse or mental health issues. This support is sometimes provided by other outside organisations separate to the hostel.

2.2.4. Shelters

Homeless shelters are typically viewed as a basic form of temporary accommodation where a bed is provided in a shared space overnight which a requirement for the individual to vacate the space during the day. One of the key features of a homeless shelter is that it is transitory and not usually seen as a stable form of accommodation as the individual are often in overcrowded buildings, and often subjected to physical altercations, theft, substance abuse, and unhygienic sleeping conditions. Like hostels, homeless shelters often place additional requirements on potential users including night‐time curfews. Additional services that may or may not be provided by the homeless shelter are warm meals for dinner and breakfast or support from volunteers and staff who help individuals make connections to other services. However, similarly to hostels, some support may be offered by external organisations and not by the shelter itself. Shelters and hostels are often defined in different ways in the UK and the United States, where these models are often used. Even within these categories there is substantial variability on the services that are provided and the conditions in which the facilities operate. Due to some of the common elements between shelters and hostels, which have now been outlined, the interventions are often described interchangeably in the global context, even if that masks some of the heterogeneity in provision.

2.2.5. Supported housing

Supported housing is an umbrella term for various accommodation‐based approaches and therefore an extremely complex intervention type. When providers describe their approach as supported housing, the intervention will typically combine housing with additional supportive services as an integrated package. The housing offered can be permanent or temporary; nonabstinent contingent or abstinent‐contingent; staffed group homes, community based or in a private unit; and the subsidies towards rent also vary. Supportive services will be offered directly to the individual or through referrals to the relevant body. Supportive services might include those to help with mental health issues, substance misuse, those interventions which increase access to health services, support to continue education or find employment, help with accessing benefits, or those services which focus on social aspects of the individual's life such as positive interactions with society, or community engagement. Due to the inconsistencies in the approach which “supported housing” takes, and the wide range of housing and support offered through supported housing interventions, it is incredibly difficult to group supported housing as a homogenous set of interventions for which to compare effectiveness to other groups of accommodation‐based approaches.

2.2.6. Conclusion

In homelessness literature, there is difficulty both in defining homelessness and the interventions which seek to benefit individuals (FEANTSA, 2017). Suttor argues that while it may be advantageous to create interventions tailored to an individual's unique needs, there is a need to classify approaches (Suttor, 2016). Indeed, most commentators acknowledge the challenges of lack of clear definition of the many terminologies used to describe accommodation‐based interventions. One example of this is highlighted in a study which identified 307 unique terms across 400 articles on supported accommodation (Gustafsson et al., 2009). Additionally, the Housing First model initially seems like an approach where categorisation is straightforward, however, there exists significant inconsistencies regarding implementation. Various researchers observe that this may be due to the way the Housing First model has deviated from the original “Pathways to Housing” intervention (Tsemberis & Eisenberg, 2000) due in part to additional services and support (Johnson et al., 2012; Phillips et al., 2011).

2.3. A new typology

One such typology already exists and is based on an international evidence review of 533 interventions for rough sleepers (Mackie et al., 2017). This review was led by one of the current review authors and identified characteristics of various types of temporary accommodation, namely shelters and hostels. The review team adapted this typology to inform the development of categories for the accommodation‐based interventions. This process was undertaken alongside Lipton and colleagues' (Lipton et al., 2000) descriptive categorisation of low, moderate, or high intensity housing which is based on the degree of structure and level of independence offered to their 2937 study participants. A further category (housing only) was added to allow for interventions which focused on providing accommodation for an extended period without further support or services offered. It was deemed to be more than just meeting the basic needs of the individual, but not intense enough to meet the criteria of the moderate category, as individuals were not receiving any additional services or help.

To develop the typology further, we used an iterative decision model. First, the review team selected a random sample of five accommodation‐based interventions included in the Evidence and Gap Map (EGM) of homelessness interventions (White et al., 2020), upon which this review is based. Second, two review team members independently coded the characteristics, hypotheses and concepts related to each intervention and compared notes. This independent analysis of the sampled papers ensured both objectivity and consistency in this step of the process and allowed the reviewers to investigate substantial amounts of data without bias or a predetermined hypothesis. Third, emerging themes were collated, and reviewers communicated to better understand the patterns which appeared through the sampled studies. Finally, through this iterative process we conclude that the most suitable way to create meaningful categorisations would be based around the intensity (defined as the level of the support offered).

Furthermore, interventions varied on the conditions the user was required to abide by. These conditions include needing to be sober from alcohol and/or drugs, abstain from criminal activity or to gain employment after a certain amount of time. To accurately incorporate these conditions into the categories, it must be stated whether the intervention required such a behavioural condition (conditional) or whether there were no behavioural conditions imposed (unconditional). The typology is described below and presented in Table 1.

Table 1.

Typology: summary of categories

	Type of accommodation	Support	Conditionality
Basic/conditional	Interventions that meet the user's basic human needs only, for example, providing bed and other basic subsistence such as food.	There are no named additional services or support offered to the user. This type of intervention focuses more on the short‐term benefit to the user	Conditions such as sobriety or punctuality apply
Basic/unconditional	Interventions that meet the user's basic human needs only, for example, providing bed and other basic subsistence such as food.	There are no named additional services or support offered to the user. This type of intervention focuses more on the short‐term benefit to the user	Accommodation is not conditional on adherence to rules such as sobriety or punctuality
Housing only/conditional	Accommodation provided for an extended period	Without additional support or services	Behavioural expectations are imposed on users, for example, they must enter paid employment within six months
Housing only/unconditional	Accommodation provided for an extended period	Without additional support or services	The participant is not required or obligated to meet any behavioural expectation to retain their housing
Moderate support/conditional	Accommodation provided for an extended period	The level of support and type of service offered will remain general and aimed towards a group of homeless individuals, and not specific to individual personal needs	Expectations on behaviour in place for example signing a contract agreeing to abstain from drugs and or alcohol
Moderate support/unconditional	Accommodation provided for an extended period	The level of support and type of service offered will remain general and aimed towards a group of homeless individuals, and not specific to individual personal needs	Accommodation not conditional on engagement (though engagement may be encouraged)
High support/conditional	Accommodation provided for an extended period	Assertive, individualised services and interventions for users. They often focus specifically on the personal needs of the user	Expectations such as abstinence from alcohol and drugs in place
High support/unconditional	Accommodation provided for an extended period	Assertive, individualised services and interventions for users. They often focus specifically on the personal needs of the user	No behavioural expectation such as sobriety placed on the user
No intervention (control groups)	None provided	None provided except basic information about other services	Not applicable

Study title	Name of intervention(s)	Complexity of needs	Sample size	Age	Sex	Design	Typology
Appel (2012)	Keeping “Home project”: Housing First approach	Poor Mental Health and Substance abuse issues	61	Mean: Intervention, 45.9 (range 26–63); Control, 39.7 (no range mentioned)	Male: Intervention, 26 (80.8%); Control, 19 (63.3%). Female: Intervention, 5 (19.2%); Control, 11 (36.7%)	Non‐RCT	High unconditional
Brown (2016)	Housing First	Poor Mental Health (70.9%) and Substance abuse issues (75.8%)	182	Mean, 42.79; SD, 11.14	Male: 73.6%; Female: 26.4%	Non‐RCT	High unconditional
Buchanan (2006)	Respite Care	Poor Physical Health	225	Mean (SD), Intervention, 43 (9); Control, 44 (10)	Male: Intervention, 78% (125); Control, 81% (52). Female: Intervention, 22% (36); Control, 19% (12)	Non‐RCT	Moderate conditional
Buchanan (2009)	The Chicago Housing for Health Partnership (CHHP)	Poor Physical Health and substance abuse issues	94	Intervention, mean 45 (SD 6.9); Control, mean 43 (SD 7.7)	Male: Intervention, 39 (72%); Control; 43 (84%). Female: Intervention, 15 (28%); Control, 8 (16%)	RCT	High—conditional and unconditional, depending on participant
Cheng (2007)	HUD‐VASH	Poor Physical Health and substance abuse issues	460	Not specified	Not specified	RCT	Moderate unconditional
Gilmer (2010)	Full‐Service Partnerships (FSP)	Poor Mental Health and Incarceration	363	Mean (SD), Intervention, 44 (9); Control, 43 (11)	Male: Intervention, 131/209 (63%); Control, 97/154 (63%). Female: Intervention, 78/209 (37%); Control, 57/154 (37%)	Non‐RCT	High conditional
Goering (2011)	Chez Soi—Housing First	Poor Mental Health	2131	40.89 (SD, 11.23)	Male: 1508 (67.9%); Female: 603 (31.2); Other: 20 (.9%)	RCT	High unconditional
Goldfinger (1999)	Group homes, or Independent Apartments.	Poor Mental Health and Substance abuse issues	110	Mean, 38	(Intervention only) Male: 85 (72%); Female 33 (28%)	RCT	High unconditional
Gulcur (2003)	HOUSING FIRST Continuum of Care	Poor Mental Health and Incarceration and Substance abuse issues	199	Not specified	Male: 173; Female: 52	RCT	High unconditional
Howard (2011)	Crisis House or Patient preference of Crisis House	Poor Mental Health	44	Mean (SD), 37.5(11.1)	Female: 102 participants (100%)	RCT	Moderate conditional
Hwang (2011)	Supportive housing	None specified	112	22 people aged 17–0, 90 people aged 31 and over	81 male 72% 30 (27%) women one (1%) transgendered individual	Non‐RCT	High conditional
Kertesz (2007)	ACH Vs NACH Vs Control	Poor psychical Health, poor mental health, incarceration and substance abuse issues.	99	Mean (SD): ACH: 38.25 (2.61); N‐ACH: 41.25 (3.19); Control: 38.5 (2.61)	Male: ACH, 74.6% (47); N‐ACH, 76.8% (50); Control/No Housing, 76.8% (50). Female: ACH, 25.4% (16); N‐ACH, 24.2% (16); Control/No Housing, 24.2% (16)	RCT	Intervention 1, high conditional. Intervention 2, high unconditional
Larimer (2009)	Housing First	Poor Physical health and substance abuse issues	134	Mean (SD), Overall, 48 (10); Intervention: 48 (9); Control: 48 (11)	Male, 94% (126); Female, 6% (8)	Non‐RCT	High unconditional
Levitt (2013)	Home to stay	None Specified		Mean (SD), Intervention: 33.5 (8.7); Control: 33.9 (7.2)	Not specified	RCT	Moderate conditional
Lim (2017)	NYNY IIII	Substance Abuse Issues, care leaver and high risk of harm and/or exploitation	895	Mean overall: 18.6	Male, 510; Female, 385	Non‐RCT	High unconditional
Lim (2018)	New York Supportive Housing Program	Poor Mental Health and Substance abuse issues	330 families/2827 individuals	Number of people in each group: 18–34 years, 16%; 35–44 years, 26%; 45–54 years, 38%; ≥55 years, 20%	Male, 70% of 2827; Female, 30% of 2827	Non‐RCT	High unconditional
Lipton (2000)	Intervention 1: High Intensity Housing Intervention 2: Moderate Intensity Housing Intervention 3: Low Intensity Housing	Poor Mental Health and Substance abuse issues	2937	Mean (SD), 40.3(10.3)	Male 67% n = 1980; Female 33%, n = 957	Non‐RCT	Intervention 1, high conditional. Intervention 2, moderate conditional
McHugo (2004)	Integrated housing services programme	Poor Mental Health	113	Ranged, 21–60 years	Male: Intervention, 29 (47.5%); Control, 29 (48.3%). Female: Intervention, 32; Control, 31	RCT	High unconditional
Milby (1996)	Enhanced Care	Substance abuse issues	176	Age, mean (SD) Usual Care, 35.7 (6.2); Enhanced Care, 36.0 (6.6)	Male Intervention 54 (87.1%) Usual Care 50 (72.5%) Female Usual Care 8 (12.9) Enhanced Care 19 (27.5)	RCT	High conditional
Milby (2000)	Day treatment+	Poor Mental Health and Substance Abuse Issues	110	Mean (SD), DT, 39.1 (7.5); DT+, 37.3 (7.2); Al, 38.1 (7.4)	Male: DT, 45 (83%); DT+, 39 (70%); All, 84 (76%). Female: DT, 9 (17); DT+, 17 (30); All, 26 (24%)	RCT	High conditional
Milby (2008)	CM and CM+ (Contingency Management plus behavioural day treatment)	Substance abuse issues	206	CM, 39.5 (7.2); CM+, 40.6 (7.1)	Male: CM, 77 (74.8%); CM+, 73 (70.9%). Female: CM, 26 (25.2); CM+, 30 (29.1	RCT	High conditional
O'Connell (2012)	Housing and Urban Development–Veterans Affairs Supported Housing (HUD‐VASH) Intensive Care Management (ICM)	Poor physical Health, Poor mental health, substance abuse issues	207	Age (median + SD), TAU, 42.3 + −7.5; ICM, 44.0 + −6.3; HUD‐VASH, 41.8 + −7.1	100% Male	RCT	Moderate unconditional
Sadowski (2009)	Housing and Case management (HOUSING FIRST)	Poor mental health and substance abuse issues	357	25 and over	Male, 310/405 75%; Female, 95/405 25%	RCT	High unconditional
Shern (1997)	New York Street Study—Specialised Housing	Poor Mental Health	168	37.5 (9.01) years	Male, 72%; Male, (644). Female, 28%; Female, (250)	RCT	Moderate conditional
Siegel (2006)	Supported Housing (SH) vs community residences (CR)	Poor Physical Health, poor mental health and Substance abuse issues	47	Mean (SD): Stratum 1 (N‐10): SH, 34.7 (7.9); CR, 36(7.9) (n‐37). Stratum 2: SH, 41.6 (11.2) (N‐18); CR, 41.3 (10) (N‐28). Stratum 3: SH, 47.4 (10.1) (N‐39); CR, 41.6 (8.5) (N‐7)	Male: 91/139 across three strata of interventions. Female: 48/139 Across three strata of interventions	Non‐RCT	Intervention 1, high unconditional. Intervention 2, high conditional
Sosin (1996)	Housing and case management Case management alone	Substance abuse issues	419	Mean, 35; Housing, 35.5; case management, 35.2; control, 34.6	Male: 74.5% 312. Female: 25.5% 107	Non‐RCT	Moderate conditional
Srebnik (2013)	Housing First	Poor Physical Health, poor mental health and substance abuse issues	60	Intervention mean, 51.3; SD, 9.2. Control mean, 50.0; SD, 6.9	Male: 52 (87%); Intervention, 21 (72%); Comparison, 31 (100%); Female: 8 (13%); Intervention, 8 (28%); Comparison, 0 (0%)	Non‐RCT	High unconditional
Stefancic and Tsemberis (2007)	Pathways Consortium (Housing First)	Poor mental health and substance abuse issues	392	over 18 to be eligible, otherwise not specified	Male:74.23% 193; Pathways, 71, 67.6%; Consortium, 83, 79.8%; Control, 39, 76.5%. Female: 25.77%. 67; Pathways, 34, 32.4%; Consortium, 21, 20.2%; Control, 12, 23.5%	RCT	High unconditional

Intervention type	Housing stability (59)^a	Health (65)	Crime and justice (12)	Employment and income (13)	Capabilities and wellbeing (23)
Basic/conditional (0 studies)^b	6	2
Basic/unconditional (0 studies)	7	2	5
Housing only/conditional (0 studies)
Housing only/unconditional (0 studies)	4	5		2	9
Moderate support/conditional (six studies)	5	6			3
Moderate support/unconditional (two studies)	16	22			4
High support/conditional (nine studies)	5	11	3	4	3
High support/unconditional (15 studies)	45	33	9	8	9
No intervention (control groups) (25 studies)	30	47	7	12	20

Intervention type	Housing/unconditional	Moderate/conditional	Moderate/unconditional	High/Conditional	High/Unconditional	No intervention
Basic/unconditional	−0.97 (−2.15, 0.21)	−0.83 (−1.84, 0.18)	−0.87 (−1.87, 0.13)	−0.69 (−1.71, 0.32)	−1.10 (−1.82, −0.39)	−0.48 (−1.32, 0.36)
Housing only/unconditional	.	0.14 (−1.04, 1.31)	0.10 (−0.85, 1.06)	0.28 (−0.85, 1.41)	−0.13 (−1.07, 0.81)	0.49 (−0.46, 1.45)
Moderate/conditional		.	−0.03 (−1.03, 0.96)	0.14 (−0.87, 1.15)	−0.27 (−0.98, 0.44)	0.35 (−0.48, 1.18)
Moderate/unconditional			.	0.17 (−0.69, 1.04)	−0.24 (−0.93, 0.46)	0.39 (−0.19, 0.97)
High/conditional				.	−0.41 (−1.13, 0.31)	0.21 (−0.44, 0.86)
High/unconditional					.	0.62 (0.19, 1.06)

Intervention type	Housing/unconditional	Moderate/conditional	Moderate/unconditional	High/conditional	High/Unconditional	No intervention
Basic/unconditional	−0.19 (−0.94, 0.55)	−0.45 (−1.00, 0.09)	−0.36 (−0.92, 0.20)	−0.31 (−0.85, 0.23)	−0.32 (−0.75, 0.12)	−0.10 (−0.57, 0.38)
Housing/unconditional	.	−0.26 (−0.95, 0.43)	−0.17 (−0.87, 0.54)	−0.12 (−0.80, 0.57)	−0.12 (−0.73, 0.48)	0.10 (−0.54, 0.74)
Moderate/conditional		.	0.09 (−0.30,0.49)	0.15 (−0.28, 0.57)	0.14 (−0.19, 0.47)	0.36 (0.03, 0.69)
Moderate/unconditional			.	0.05 (−0.37, 0.47)	0.04 (−0.32, 0.41)	0.27 (−0.05, 0.58)
High/conditional				.	−0.01 (−0.33, 0.31)	0.21 (−0.06, 0.49)
High/unconditional					.	0.22 (0.01, 0.43)

Study name	Name of intervention	Intervention type	Location	Setting	Population
Austin et al. (2014)	Housing and Urban Development Veteran's Affairs Supportive Housing	High/unconditional support	USA	Veteran's Affairs facilities	Veterans
Burt (2009)	Various permanent supportive housing projects	Not specified	USA	Various	Long term homeless and homeless populations with mental illnesses
Busch‐Geertsema (2013)	Housing First Europe	High/unconditional support	Europe	Amsterdam: Discus Housing First Copenhagen (ACT in Copenhagen)Glasgow (Turning Point Scotland Housing First) Lisbon (Casas Primeiro) Budapest (Pilisi Forest Project)	Homeless populations with complex needs
Greenwood (2005)	Dublin Housing First Demonstration Project	High/unconditional support	Ireland	Dublin	People with significant histories of homelessness
HRDC (2003)	National Homeless Strategy	Not specified	Canada	Various community‐ based case studies	People experiencing homelessness, particularly youth and Aboriginal groups
Keast (2008)	Responding to Homelessness Strategy	High/conditional support	Australia	Various service settings across Brisbane, the Gold Coast and Townsville	Various: newly homeless users, users with complex and multiple needs, users in persistent homelessness
Lawlor and Bowen (2017)	Limerick Youth Housing	High/unconditional support	Ireland	Limerick city and Midwest region of Ireland	Young people who are homeless or at risk of becoming homelessness
Pleace and Bretherton (2013)	Camden Housing First	High/unconditional support	UK	Camden hostels	People with co‐morbidities experiencing recurrent homelessness
Sewel (2016)	Supported lodgings services	High/conditional support	UK	Supported lodgings services for care experienced young people	Young people in care who are yet ready for independent living
Turner Research (2015)	Haven's Way Programme	High/conditional support	Canada	Haven's Way congregate housing	Vulnerable young women experiencing considerable housing instability

Study name	Theory of change	TAU	Tx Length	Dosage	Personnel delivering intervention	Personnel working with users	Intervention funding	Organisations involved	Participants
Austin et al. (2014)	To end veteran homelessness by 2015 using and Housing First model by providing permeant rental vouchers	NR	NR	NR	Housing First coordinator, HUD‐VASH programme manager	Housing specialist, Case workers, substance use or mental health specialist	Federal funding	DoH and Urban Development	95 facility managers and front‐ line staff interviewed
Burt (2009)	To reduce people experiencing long term homelessness with Predevelopment work on permanent supportive housing projects, collaborating with public officials and key stakeholders to stimulate increased commitment.	NR	Varied across the projects	Usually weekly meetings with case worker	Corporation for supportive housing	Case workers, health care specialists	Conrad N. Hilton Foundation	Los Angeles Council	35 staff and stakeholders interviewed
Busch‐Geertsema (2013)	To use the Housing First approach in various European cities to house individuals who are homeless	No control group	Varied between cities	Weekly home visits from staff	Case managers	Case managers	European Commission	Housing First Europe, National Board of Social Services, GISS Bremen	22 participants interviewed
Greenwood (2005)	To use Housing First (scattered site) in to house those who are currently homeless	Usual housing services	Three months at times to house	Weekly face to face meetings with programme staff	Team leader	Programme staff: key workers, substance specialist, nurse, employment support worker	European Commission	Dublin Region Homeless Executive	30 participants surveyed, 26 interviewed
HRDC (2003)	To provide housing and other services to vulnerable groups	No control groups due to ethical reasons	Varies between projects	Not described, probably varied across the projects	Various	Various	Federal funding	Local organisations to the projects	Interviews collected from 91 youth and Aboriginals
Keast (2008)	Integrate services to provide a smoother pathway to housing for people who are homeless	NR	NR	NR	Service staff	Homelessness workers	Queensland Government	DoH, Department of Communities, Queensland Health, Department of Justice and Attorney‐General, Queensland Police, Queensland Corrective Services, Department of Tourism, Fair Trade and Wine Industry Development	241 different projects but number of participants not specified
Lawlor and Bowen (2017)	The provision of secure, appropriate and supported accommodation allows young people the space and stability to plan for the future	NR	Varies from 3 to 18 months and more	Varied between twice a week and twice a month	Tusla and Focus staff	Key workers and social workers	Human Dignity Foundation, Focus Ireland and Tusla	Focus Ireland, Tusla, Limerick City and County Council	14 semi structured interviews with young people
Pleace and Bretherton (2013)	The provision of secure, appropriate and supported accommodation allows people the space and stability to plan for the future	NR	Up to 18 months	Weekly home visits	Camden Council staff	Case managers	Camden Council	Camden Council, University of York, SITRA	13 service users interviewed
Sewel (2016)	Providing care experienced young people with supported lodgings will assist in a young person developing the confidence and capability to live an independent adult life	NR	Up to 6 months before discharge	2‐3 face to face meetings per week with support worker, tapering to once a month as placement continues	Supported lodgings providers	Support worker	Local authority funding	Barnardos	7 service staff surveyed, 11 service staff, 14 young people and 20 supported lodgings providers interviewed
Turner Research (2015)	Providing young girls with supported lodgings will assist in a young person developing the confidence and capability to live an independent adult life	NR	Average of 8 months	Weekly case manager meetings	Boys and Girls Clubs of Calgary	Case manager	Safe Haven Foundation of Canada (Housing Firstc), private donors and Alberta Human Services.	Human Services, Government of Alberta, Boys & Girls Clubs of Calgary, Safe Haven Foundation	18 participants and 5 stakeholders interviewed

Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk
Allocation concealment (selection bias)	Unclear risk
Blinding of participants and personnel (performance bias)	Unclear risk
Blinding of outcome assessment (detection bias)	Unclear risk
Incomplete outcome data (attrition bias)	Unclear risk
Selective reporting (reporting bias)	Unclear risk
Other bias	Unclear risk

1. Bibliographic information
Article ID	FREETEXT
Linked articles	FREETEXT
Extracted by	FREETEXT
Checked by	FREETEXT
Year of publication	FREETEXT
Type of publication	1. Journal Article 2. Book/book chapter 3. Government report 4. Conference proceedings 5. Presentation 6. Thesis or Dissertation 7. Unpublished report 8. Other (please specify)
Location of study The location in which the study is set not* where the study authors are based*.	1. UK 2. ROI 3. Rest of Europe 4. USA 5. Canada 6. South America 7. Central America 8. Oceania 9. Middle‐East 10. Asia 11. Africa 12. Other (Please Specify)
	Not Specified
Study funding sources	1. Research council funding 2. University scholarships and bursaries 3. Salaried research assistantships from university departments 4. Grants or loans from trusts and charities 5. Local enterprise initiatives 6. Company sponsorship 7. Government loans 8. EU Scholarships 9. Industry sponsorship 10. Other (please specify)
Possible conflicts of interest	1. Yes, possible/definite conflict of interest 2. No, study appears to be free of CoI 3. Can't tell

2. Participant information
Recruitment setting Where were participants recruited from?	1. Clinical setting 2. Accommodation for individuals experiencing homelessness 3. Family home 4. The street 5. Community setting 6. Referred by friends or family 7. Referred by medical health professional 8. Housing Agency 9. Other (Please specify)
Homelessness Status at intake Describe the housing status of the sample at intake and/or any information given about housing status prior to intake. Tick all that apply and try to extract numbers were available. Homelessness is defined as those individuals who are sleeping 'rough' (sometimes defined as street homeless), those in temporary accommodation (such as shelters and hostels), those in insecure accommodation (such as those facing eviction or in abusive or unsafe environments), and those in inadequate accommodation (environments which are unhygienic and/or overcrowded).	1. Sleeping 'Rough' (or rooflessness) 2. Temporary Accommodation 3. Insecure Accommodation 4. Inadequate Accommodation 5. Involuntary sharing, for example, domestic violence 6. Hidden/concealed homelessness 7. Other (please specify)
	Not Specified
Geographical context Where participants receive treatment?	1. Urban 2. Rural 3. Suburban 4. Mixed 5. Other (please specify)
	Not Specified
Gender % (actual number)	FREETEXT
Age Extract mean age, SD and range. Choose multiple options if the analysis is reported separately for different age groups.	1. Under 25 2. 25 and Over
Complexity of needs What other challenges does the individual face, if any, aside from the risk or experience of homelessness? High Risk of Harm and/or Exploitation ‐ For example, women in shelters, newcomer families, refugee/asylum seeker, care leavers	1. Poor Physical Health 2. Poor Mental Health 3. Incarceration 4. Substance Abuse Issues 5. Care leaver 6. Limited access to integrated support services 7. High Risk of Harm and/or Exploitation 8. Other (please specify) Not RelevantNot Specified
Mental health status	1. Receiving treatment 2. Not receiving treatment 3. Other (please specify)
	Not relevant
	Not Specified
Substance use status	1. Receiving treatment 2. Not receiving treatment 3. Other (please specify)
	Not relevant
	Not Specified
Homelessness status Homelessness is defined as those individuals who are sleeping “rough” (sometimes defined as street homeless), those in temporary accommodation (such as shelters and hostels), those in insecure accommodation (such as those facing eviction or in abusive or unsafe environments), and those in inadequate accommodation (environments which are unhygienic and/or overcrowded).	1. Sleeping “rough” 2. Temporary accommodation 3. Insecure accommodation 4. Inadequate accommodation 5. Other (please specify)
	Not Specified
Family vs. No Family Family = any child involved Non‐family = single person or couple without children If mixed sample select both and describe	1. Family 2. Non‐Family
	Not Specified
Sample size of treatment groupNumber of people assigned to treatment. If more than one treatment group extract all and be clear which group is which.	FREETEXT
Sample size of control group Number of people assigned to control. If more than one control group extract all and be clear which group is which.	FREETEXT

3. Intervention information
How many intervention arms in this trial? List how many study arms there are and given each a name. e.g. Intervention = Critical Time Intervention; Control = Treatment as usual If there is more than one intervention arm go to the "Study Arm" tab and add the RELEVANT study arms. You must then extract data for each relevant study arm.	FREETEXT
Name of intervention Write in the name of the program, intervention, or treatment under study. This may be specific like “critical time intervention” or it may be something more generic like “supported housing”	FREETEXT
Briefly Describe the intervention Briefly describe the intervention, what participants are offered and any important factors such as conditionality, nature of housing, case management, substance abuse treatment included etc.	FREETEXT
Theory of change How does the intervention aim to bring about change? What is the underlying theoretical rationale for why the intervention might work to improve outcomes? If not specified write "not specified"	FREETEXT
What is the size of accommodation/How many beds?	FREETEXT
Duration of treatment period from start to finish In the dosage items, we are interested in the amount of treatment received by the participants. If the treatment was delivered directly to participants, the authors will probably provide at least some information about dosage and you can code these items accordingly. If minimal information is provided, you should try to give estimates for these items if you can come up with a reasonable estimate.	FREETEXT
Timing Frequency of contact between participants and provider/program activity	1. Once a month2. Less than weekly3. Once a week4. 1‐2 times a week5. 2 times a week6. 2‐3 times a week7. 3 times a week8. 3‐4 times a week9. 4 times a week10. Daily contactCan't Estimate
Length of each individual session How long does each contact/session last?	FREETEXT
Study Personnel The primary individual/s who have direct contact with the participants served by the program. If the report is the author's dissertation (or based on the author's dissertation), then code as "Graduate Researcher". If the delivery is performed by graduate or undergraduate students assisting the author then select "Grad/Undergrad Students". Code “Self‐directed” for studies where electronic/computer programs are used. If the intervention is solely environmental i.e. community housing, then code “environmental change”	1. Graduate Researcher 2. Grad/Undergrad Students 3. Author 4. Homelessness professional
	Includes case manager, social worker, outreach worker
	5. Peers 6. Interventionist (Not Hired by Researcher) 7. Interventionist (Hired by Researcher) 8. Self‐Directed 9. Medical Professionals 10. Other (please specify)
	Not Specified
Did provider receive specialised training? This refers to whether or not the “interventionist” received specialised training to equip them to deliver the intervention proficiently.	1. Yes 2. The interventionist IS program developer 3. No
	Not specified
Resource requirements Time, staff, housing provision etc	FREETEXT
Cost	FREETEXT

4a. Study Design
Design The studies included in all reviews must include an intervention group and at least one untrained control group. Control groups can include placebo, no treatment, waitlist, or treatments vs “treatment as usual.” Any study which includes one group pre‐test/post‐test or in which a treatment group is only compared to another treatment group will not be eligible for inclusion.	1. Randomised control trial
	Individual or cluster randomised
	2. Non‐randomised control trial
What do control subjects receive?	1. Placebo 2. Treatment as usual 3. No treatment 4. Not specified
1. Placebo (or attention) treatment. Group gets some attention or a sham treatment 2. Treatment as usual. Group gets “usual” handling instead of some special treatment. 3. No treatment. Group gets no treatment at all.
Unit of allocation Individual (i.e., some were assigned to treatment group, some to comparison group) Group (i.e., whole subsets assigned to treatment and comparison groups) Regions (i.e., region assigned as an intact unit)	1. Individual 2. Group 3. Regions 4. Other (Please Specify)
Not Specified
Method of assignment	1. Randomly after matching 2. Randomly without matching 3. Regression discontinuity design 4. Cluster assigned 5. Wait list control 6. Non‐random, but matched 7. Other (Please Specify)
Method of group assignment. How participants/units were assigned to groups. This item focuses on the initial method of assignment to groups, regardless of subsequent degradations due to attrition, refusal, etc. prior to treatment onset.
1. Randomly after matching, yoking, stratification, blocking, etc. The entire sample is matched or blocked first, then assigned to treatment and comparison groups within pairs or blocks. This does not refer to blocking after treatment for the data analysis. 2. Randomly without matching, etc. This also includes cases when every other person goes to the control group. 3. Regression discontinuity design: quantitative cutting point defines groups on some continuum (this is rare). 4. Cluster assigned, this is to be used in cluster assignment studies only, specify the number of clusters in the treatment group and the number of clusters in control. 5. Wait list control or other quasi‐random procedure presumed to produce comparable groups (no obvious differences). This applies to groups which have individuals apparently randomly assigned by some naturally occurring process, e.g. first person to walk in the door. The key here is that the procedure used to select groups doesn't involve individual characteristics of persons so that the groups generated should be essentially equivalent. 6. Non‐random, but matched: Matching refers to the process by which comparison groups are generated by identifying individuals or groups that are comparable to the treatment group using various characteristics of the treatment group. Matching can be done individually, e.g., by selecting a control subject for each intervention subject who is the same age, gender, and so forth, or on a group basis.
	Not Specified
Was there >20% attrition in either/both groups? Attrition occurs when participants are lost from an intervention over time or over a series of sequential processes. Studies may describe this as “lost to follow‐up,” or “drop outs.”	FREETEXT

4b. Non‐random studies
How were groups matched? If matching was used prior to assignment of condition, how were groups matched?	1. Matched on Pre‐test measure 2. Matched on personal characteristics 3. Matched on demographics 4. Groups weren't matched 5. Other (please specify)
	Not specified
Was the equivalence of groups tested at pre‐test?	FREETEXT
Results of statistical comparisons of pre‐test differences	1. No statistically significant differences 2. Significant differences judged unimportant by coder 3. Significant differences judged of uncertain importance by coder 4. Significant differences judged important by coder 5. Other (please specify)
Were there pre‐test adjustments?	FREETEXT

5. Qualitative information
Qualitative methods used	FREETEXT
Data analysis technique and procedure	FREETEXT
Was the intervention implemented as intended?	1. Yes 2. No
	Not specified
How was this measured?	FREETEXT
What implementation and process factors impact intervention delivery?	1. Contextual factors 2. Policy makers/funders 3. Programme managers/Implementing agency, 4. Staff/case workers 5. Recipients

6. Assessing quality in RCTs (Cochranes ROB2 tool)
Domain 1: Risk of bias arising from the randomization process
1.1 Was the allocation sequence random?	1. Yes 2. Probably yes 3. Probably No 4. No
1.2 Was the allocation sequence concealed until participants were enrolled and assigned to interventions?	1. Yes 2. Probably yes 3. Probably No 4. No
1.3 Did baseline differences between intervention groups suggest a problem with the randomization process?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk‐of‐bias judgement	1. Low 2. High 3. Some concerns
Optional: What is the predicted direction of bias arising from the randomization process?	1. Favours experimental 2. Favours comparator 3. Towards null 4. Away from null 5. Unpredictable
Domain 2: Risk of bias due to deviations from the intended interventions (effect of assignment to intervention)
2.1. Were participants aware of their assigned intervention during the trial?	1. Yes 2. Probably yes 3. Probably No 4. No
2.2. Were carers and people delivering the interventions aware of participants' assigned intervention during the trial?	1. Yes 2. Probably yes 3. Probably No 4. No
2.3. If Y/PY/NI to 2.1 or 2.2: Were there deviations from the intended intervention that arose because of the experimental context?	1. Yes 2. Probably yes 3. Probably No 4. No
2.4. If Y/PY to 2.3: Were these deviations from intended intervention balanced between groups?	1. Yes 2. Probably yes 3. Probably No 4. No
2.5 If N/PN/NI to 2.4: Were these deviations likely to have affected the outcome?	1. Yes 2. Probably yes 3. Probably No 4. No
2.6 Was an appropriate analysis used to estimate the effect of assignment to intervention?	1. Yes 2. Probably yes 3. Probably No 4. No
2.7 If N/PN/NI to 2.6: Was there potential for a substantial impact (on the result) of the failure to analyse participants in the group to which they were randomized?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk‐of‐bias judgement	1. Low 2. High 3. Some concerns
Optional: What is the predicted direction of bias due to deviations from intended interventions?	1. Favours experimental 2. Favours comparator 3. Towards null 4. Away from null 5. Unpredictable
Domain 3: Missing outcome data
3.1 Were data for this outcome available for all, or nearly all, participants randomized?	1. Yes 2. Probably yes 3. Probably No 4. No
3.2 If N/PN/NI to 3.1: Is there evidence that result was not biased by missing outcome data?	1. Yes 2. Probably yes 3. Probably No 4. No
3.3 If N/PN to 3.2: Could missingness in the outcome depend on its true value?	1. Yes 2. Probably yes 3. Probably No 4. No
3.4 If Y/PY/NI to 3.3: Do the proportions of missing outcome data differ between intervention groups?	1. Yes 2. Probably yes 3. Probably No 4. No
3.5 If Y/PY/NI to 3.3: Is it likely that missingness in the outcome depended on its true value?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk‐of‐bias judgement	1. Low 2. High 3. Some concerns
Optional: What is the predicted direction of bias due to missing outcome data?	1. Favours experimental 2. Favours comparator 3. Towards null 4. Away from null 5. Unpredictable
Domain 4: Risk of bias in measurement of the outcome
4.1 Was the method of measuring the outcome inappropriate?	1. Yes 2. Probably yes 3. Probably No 4. No
4.2 Could measurement or ascertainment of the outcome have differed between intervention groups?	1. Yes 2. Probably yes 3. Probably No 4. No
4.3 If N/PN/NI to 4.1 and 4.2: Were outcome assessors aware of the intervention received by study participants?	1. Yes 2. Probably yes 3. Probably No 4. No
4.4 If Y/PY/NI to 4.3: Could assessment of the outcome have been influenced by knowledge of intervention received?	1. Yes 2. Probably yes 3. Probably No 4. No
4.5 If Y/PY/NI to 4.4: Is it likely that assessment of the outcome was influenced by knowledge of intervention received?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk‐of‐bias judgement	1. Low 2. High 3. Some concerns
Optional: What is the predicted direction of bias in measurement of the outcome?	1. Favours experimental 2. Favours comparator 3. Towards null 4. Away from null 5. Unpredictable
Domain 5: Risk of bias in selection of the reported result
5.1 Was the trial analysed in accordance with a pre‐specified plan that was finalized before unblinded outcome data were available for analysis?	1. Yes 2. Probably yes 3. Probably No 4. No
Is the numerical result being assessed likely to have been selected, on the basis of the results, from…
5.2. … multiple outcome measurements (e.g. scales, definitions, time points) within the outcome domain?	1. Yes 2. Probably yes 3. Probably No 4. 4. No
5.3 … multiple analyses of the data?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk‐of‐bias judgement	1. Low 2. High 3. Some concerns
Optional: What is the predicted direction of bias due to selection of the reported result?	1. Favours experimental 2. Favours comparator 3. Towards null 4. Away from null 5. Unpredictable
Overall risk of bias
Risk‐of‐bias judgement	1. Low 2. High 3. Some concerns

7. Assessing quality in Non‐random control trials (ROBINS‐I tool)
Bias due to confounding
1.1 Is there potential for confounding of the effect of intervention in this study? If N/PN to 1.1: the study can be considered to be at low risk of bias due to confounding and no further signalling questions need be considered	1. Yes 2. Probably yes 3. Probably No 4. No
If Y/PY to 1.1: determine whether there is a need to assess time‐varying confounding:	1. Yes 2. Probably yes 3. Probably No 4. No
1.2. Was the analysis based on splitting participants' follow up time according to intervention received?If N/PN, answer questions relating to baseline confounding (1.4 to 1.6)If Y/PY, go to question 1.3.	1. Yes 2. Probably yes 3. Probably No 4. No
1.3. Were intervention discontinuations or switches likely to be related to factors that are prognostic for the outcome? If N/PN, answer questions relating to baseline confounding (1.4 to 1.6) If Y/PY, answer questions relating to both baseline and time‐varying confounding (1.7 and 1.8)	1. Yes 2. Probably yes 3. Probably No 4. No
Questions relating to baseline confounding only
1.4. Did the authors use an appropriate analysis method that controlled for all the important confounding domains?	1. Yes 2. Probably yes 3. Probably No 4. No
1.5. If Y/PY to 1.4: Were confounding domains that were controlled for measured validly and reliably by the variables available in this study?	1. Yes 2. Probably yes 3. Probably No 4. No
1.6. Did the authors control for any post‐intervention variables that could have been affected by the intervention?	1. Yes 2. Probably yes 3. Probably No 4. No
Questions relating to baseline and time‐varying confounding
1.7. Did the authors use an appropriate analysis method that controlled for all the important confounding domains and for time‐varying confounding?	1. Yes 2. Probably yes 3. Probably No 4. No
1.8. If Y/PY to 1.7: Were confounding domains that were controlled for measured validly and reliably by the variables available in this study?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk‐of‐bias judgement	1. Low 2. Moderate 3. Serious 4. Critical
Optional: What is the predicted direction of bias due to confounding?	1. Favours experimental 2. Favours comparator 3. Unpredictable
Bias in selection of participants into the study
2.1. Was selection of participants into the study (or into the analysis) based on participant characteristics observed after the start of intervention? If N/PN to 2.1: go to 2.4	1. Yes 2. Probably yes 3. Probably No 4. No
2.2. If Y/PY to 2.1: Were the post‐intervention variables that influenced selection likely to be associated with intervention?	1. Yes 2. Probably yes 3. Probably No 4. No
2.3 If Y/PY to 2.2: Were the post‐intervention variables that influenced selection likely to be influenced by the outcome or a cause of the outcome?	1. Yes 2. Probably yes 3. Probably No 4. No
2.4. Do start of follow‐up and start of intervention coincide for most participants?	1. Yes 2. Probably yes 3. Probably No 4. No
2.5. If Y/PY to 2.2 and 2.3, or N/PN to 2.4: Were adjustment techniques used that are likely to correct for the presence of selection biases?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk‐of‐bias judgement	1. Low 2. Moderate 3. Serious 4. Critical
Optional: What is the predicted direction of bias due to selection of participants into the study?	1. Favours experimental 2. Favours comparator 3. Towards null 4. Away from null 5. Unpredictable
Bias in classification of interventions
3.1 Were intervention groups clearly defined?	1. Yes 2. Probably yes 3. Probably No 4. No
3.2 Was the information used to define intervention groups recorded at the start of the intervention?	1. Yes 2. Probably yes 3. Probably No 4. No
3.3 Could classification of intervention status have been affected by knowledge of the outcome or risk of the outcome?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk‐of‐bias judgement	1. Low 2. Moderate 3. Serious 4. Critical
Optional: What is the predicted direction of bias due to classification of interventions?	1. Favours experimental 2. Favours comparator 3. Towards null 4. Away from null 5. Unpredictable
Bias due to deviations from intended interventions
If your aim for this study is to assess the effect of assignment to intervention, answer questions 4.1 and 4.2
4.1. Were there deviations from the intended intervention beyond what would be expected in usual practice?	1. Yes 2. Probably yes 3. Probably No 4. No
4.2. If Y/PY to 4.1: Were these deviations from intended intervention unbalanced between groups and likely to have affected the outcome?	1. Yes 2. Probably yes 3. Probably No 4. No
If your aim for this study is to assess the effect of starting and adhering to intervention, answer questions 4.3 to 4.6
4.3. Were important co‐interventions balanced across intervention groups?	1. Yes 2. Probably yes 3. Probably No 4. No
4.4. Was the intervention implemented successfully for most participants?	1. Yes 2. Probably yes 3. Probably No 4. No
4.5. Did study participants adhere to the assigned intervention regimen?	1. Yes 2. Probably yes 3. Probably No 4. No
4.6. If N/PN to 4.3, 4.4 or 4.5: Was an appropriate analysis used to estimate the effect of starting and adhering to the intervention?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk‐of‐bias judgement	1. Low 2. Moderate 3. Serious 4. Critical
Optional: What is the predicted direction of bias due to deviations from the intended interventions?	1. Favours experimental 2. Favours comparator 3. Towards null 4. Away from null 5. Unpredictable
Bias due to missing data
5.1 Were outcome data available for all, or nearly all, participants?	1. Yes 2. Probably yes 3. Probably No 4. No
5.2 Were participants excluded due to missing data on intervention status?	1. Yes 2. Probably yes 3. Probably No 4. No
5.3 Were participants excluded due to missing data on other variables needed for the analysis?	1. Yes 2. Probably yes 3. Probably No 4. No
5.4 If PN/N to 5.1, or Y/PY to 5.2 or 5.3: Are the proportion of participants and reasons for missing data similar across interventions?	1. Yes 2. Probably yes 3. Probably No 4. No
5.5 If PN/N to 5.1, or Y/PY to 5.2 or 5.3: Is there evidence that results were robust to the presence of missing data?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk of bias judgement	1. Low 2. Moderate 3. Serious 4. Critical
Optional: What is the predicted direction of bias due to missing data?	1. Favours experimental 2. Favours comparator 3. Towards null 4. Away from null 5. Unpredictable
Bias in measurement of outcomes
6.1 Could the outcome measure have been influenced by knowledge of the intervention received?	1. Yes 2. Probably yes 3. Probably No 4. No
6.2 Were outcome assessors aware of the intervention received by study participants?	1. Yes 2. Probably yes 3. Probably No 4. No
6.3 Were the methods of outcome assessment comparable across intervention groups?	1. Yes 2. Probably yes 3. Probably No 4. No
6.4 Were any systematic errors in measurement of the outcome related to intervention received?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk of bias judgement	1. Low 2. Moderate 3. Serious 4. Critical
Optional: What is the predicted direction of bias due to measurement of outcomes?	1. Favours experimental 2. Favours comparator 3. Towards null 4. Away from null 5. Unpredictable
Bias in selection of the reported result
Is the reported effect estimate likely to be selected, on the basis of the results, from…
7.1. … multiple outcome measurements within the outcome domain?	1. Yes 2. Probably yes 3. Probably No 4. No
7.2 … multiple analyses of the intervention‐outcome relationship?	1. Yes 2. Probably yes 3. Probably No 4. No
7.3 … different subgroups?	1. Yes 2. Probably yes 3. Probably No 4. No
Risk of bias judgement	1. Low 2. Moderate 3. Serious 4. Critical
Optional: What is the predicted direction of bias due to selection of the reported result?	1. Favours experimental 2. Favours comparator 3. Towards null 4. Away from null 5. Unpredictable
Overall risk of bias
Risk‐of‐bias judgement	1. Low 2. Moderate 3. Serious 4. Critical

8. Assessing quality in Qualitative studies (White and Keenan tool)
Are the evaluation questions clearly stated?	1. Yes 2. No
Is the qualitative methodology described?	1. Yes 2. No
Is the qualitative methodology appropriate to address the evaluation questions?	1. Yes 2. No 3. Insufficient detail
Is the recruitment or sampling strategy described?	1. Yes 2. No
Is the recruitment or sampling strategy appropriate to address the evaluation questions?	1. Yes 2. No 3. Insufficient detail
Are the researcher's own position, assumptions and possible biases outlined?	1. Yes 2. No
Have ethical considerations been sufficiently considered?	1. Yes 2. No 3. Insufficient detail
Is the data analysis approach adequately described?	1. Yes 2. No
Is the data analysis sufficiently rigorous?	1. Yes 2. No
Is there a clear statement of findings?	1. Yes 2. No
Are the research findings useful?	1. Yes 2. No

Study title	Intervention name	Intervention description	Control	Scattered versus congregate	Intervention setting and time	Typology
Appel (2012)	Keeping Home project" "Housing First approach	Keeping Home uses the Housing First approach to address the needs of homeless, mentally ill substance abusers. Specifically, Keeping Home secures market‐rate, scattered site apartments for seriously mentally ill methadone maintenance treatment patients and then uses assertive community treatment to address patients service needs	Usual treatment: available housing and support services	Scattered site	New York City, USA. Between March 2005 and March 2007	High unconditional
Brown (2016)	Housing First	Held at a 75‐unit single‐site HOUSING FIRST program. Residents were not required to abstain from substance use, nor are they required to participate in mental health or substance abuse treatment. However, staff provides assertive engagement to encourage participation. Participants in the HOUSING FIRST program received a breadth of intensive, consumer‐driven support services based on personal need and interest along with housing	Usual treatment: comparison group participants received usual care, comprising access to a variety of supports, including outpatient mental health and/or substance abuse treatment, sobering services, shelter, and/or other supportive housing programs	Congregate (75‐unit single site)	Washington, USA	High unconditional
Buchanan (2006)	Respite care	Provides short‐term shelter along with access to additional services (food, case management and referrals to permanent housing is provided). Conditional upon being capable of living in a drug and alcohol free environment	No intervention	Congregate	Chicago, Illinois, USA. Between October 1, 1998, and December 31, 2000	Moderate conditional
Buchanan (2009)	The Chicago Housing for Health Partnership (CHHP)	Two strands within the intervention, one is respite care and case management the other, for those unable to abstain from alcohol or drugs, is referral to shelter with case management	Unclear	Usual treatment: Referrals to overnight shelters or to interim housing providers. All usual care participants were eligible to receive case management through an existing Ryan White program in the hospital‐affiliated HIV/AIDS clinic	Chicago, Illinois, USA. October 2003–May 2006	High, conditional and unconditional, depending on participant
Cheng (2007)	HUD‐VASH	HUD‐VASH: VA intensive case management were linked to Section 8 housing vouchers, which provide rent subsidies to low‐income individuals with disabilities	Usual treatment: standard VA homeless services, which consisted of short termbroker case management linking users to VA and community services	Not specified	San Francisco, California (N = 107); San Diego, California (N = 91); New Orleans, Louisiana (N = 165); and Cleveland, Ohio (N = 97).Around 1992	Moderate unconditional
Gilmer (2010)	Full‐Service Partnerships (FSP)	The FSP programs implemented in San Diego County provide a combination of subsidized permanent housing and team based services with a focus on rehabilitation and recovery	No treatment	Some scattered, some congregate	San Diego, Caliifornia USA. Between October 2005‐June 2008	High conditional
Goering 2011	Chez Soi—Housing First	The Housing First model used here involves the provision of permanent, private housing units to qualifying individuals, with consumer choice on services and housing location being fundamental	Usual treatment: treatment as usual in each of the current systems of care; does not mean “no care”	Mix of both depending on city and intervention given. Congregate specifically in vancouver	Five Canadian cities of Vancouver, Winnipeg, Toronto, Montreal and Moncton	High unconditional
Goldfinger (1999)	Assigned to group homes, or assigned to independent apartments.	The seven group homes each accommodated between 6 and 10 tenants with shared living, dining, recreational, and kitchen facilities, but separate bedrooms. They provided 24‐h daily coverage, with project staff constantly at hand. All study participants were provided with an intensive case manager. Each participant met with the case manager at least once a week for counselling, hands‐on help with daily activities, and help with access to needed services	High unconditional: participants were given an individual apartment and case manager, all were encouraged,but not required, to participatein community mental healthcentre programs55 scattered housing	63 to congregate housing	Boston, Massachusetts, USA. 18 Months (January 1991 and March 1992)	High unconditional
Gulcur (2003)	Housing First	The experimental Housing First programme offered immediate access to independent housing without requiring psychiatric treatment or sobriety	Continuum of Care: users begin with outreach programmes and drop‐in centres, and then progress through a series of congregate living arrangements with varying levels of on‐site support	Independent housing (scattered)	New York, New York, USA	High unconditional
Howard (2011)	Crisis House or Patient preference of Crisis House	Womens Crisis Housing was available at two London centres where there were also psychiatric wards. The women's crisis houses were located in ordinary residential streets, had a domestic atmosphere and were funded by the UK National Health Service (NHS). Staff were available 24 h per day and were either nurses or healthcare workers with a background in mental health.	Usual Treatment: Usual NHS psychiatric hospital care	Congregate – Crisis houses	London, UK, Around 2007	Moderate conditional
Hwang (2011)	Supportive housing	Tenants had access to a shared kitchen facility, a drop‐in center offering meals and outreach services, as well as a medical and dental clinic providing free services. Individuals accepted into the program received rental subsidies and paid rent‐geared‐to‐income (not exceeding 30% of income). Specifically, tenant support workers assisted residents with mental illness to transition into the housing program.	Usual Treatment, but waitlist (no housing)	Congregate, 84 units in one building	Toronto, Canada. Between 2005 and 2008	High conditional
Kertesz (2007)	ACH Vs NACH Vs Control	All participants were offered behavioral day treatment services and randomly assigned to receive either no program‐provided housing during treatment (no housing), program provided housing contingent upon abstinence (abstinence‐contingent housing), or program‐ provided housing not contingent upon abstinence (nonabstinence‐contingent housing).	Usual outpatient treatment program	Not specified	Birmingham, Alabama, United states	Intervention 1, high conditional; Intervention 2, high unconditional
Larimer (2009)	Housing First	Participants were given a room at an apartment building. No treatment requirements, but on‐site case managers work to engage residents about substance use and life goals. Meals and on‐site health care services are also offered. No behavioural conditions are placed on participants.	Waitlist	Not specified	King county, Washington, United states	High unconditional
Levitt (2013)	Home to stay	The Home to Stay pilot was focused on 3 strategies: moving families out of shelter rapidly using a locally funded, temporary housing subsidy; securing sufficient household income to enable families to pay market rent on expiration of the subsidy; and connecting families to community‐based services that would help them to maintain housing stability after the termination of Home to Stay services.	Usual Treatment, servicestaffing and delivery vary widely across the 150New York City DHS family shelters, but sheltercaseworkers are typically assigned mixed caseloadsof approximately 25 families.	Not specified	New York City, New York, USA. Around 2010–2012	Moderate conditional
Lim (2017)	NYNY IIII	The NYNY III program was for young adults aging out of foster care provides affordable housing and access to various supportive services to help achieve independent lives, including case management, job training, and education support, and provides connections to health services.	Usual Treatment: government‐subsidized housing programs	Not specified	New York City, New York, USA. 2007–2010	High unconditional
Lim (2018)	New York Supportive Housing Program	The program followed the Housing First model with housing placement not being contingent on adhering to treatment or services	No treatment	Not specified	New York City, New York, USA. 2007–2010	High unconditional
Lipton (2000)	Intervention 1: High Intensity Housing Intervention 2: Moderate Intensity Housing Intervention 3: Low Intensity Housing	High‐intensity settings are exclusively dedicated to persons with serious mental illness. It had the highest intensity of housing requirements and services. Residents in moderate intensity settings have their own rooms or studio apartments, and cooking facilities were shared. Tenants at low‐intensity sites hold legally binding leases for individual furnished private rooms or small studio type apartments, with access to either shared or private bathrooms and cooking facilities.	No Intervention	Intervention 1: Congregate Intervention 2: Congregate Intervention 3: Mix of congregate and Scattered	New York City, New York, USA. 1990–1995	Intervention 1, high conditional; Intervention 2, moderate conditional
McHugo (2004)	Integrated housing services program	The integrated housing services approach provided comprehensive mental health services through intensive case management and housing services through dedicated teams that controlled a variety of housing settings	The ACT teams assisted users in finding and affording housing	Mostly congregate housing	Washington, USA	High unconditional
Milby (1996)	Enhanced Care	Some received abstinent contingent work therapy (44.9% of EC subjects) and housing (37.7% of EC subjects). Work therapy, day treatment focused on relapse prevention, social skills training, goal review, weekend planning, community resources for homeless persons.	Usual Treatment: twice weekly group and individual counselling, referral for medical treatment, housing and other agencies. Less frequent aftercare visits.	Not specified	Birmingham, Alabama, United States. Recruited between April 1990‐August 1991, not done beyond this.	High conditional
Milby (2000)	Day treatment +	Day treatment met week days from 07:30‐14:00 h and included lunch and transportation to and from shelters. The following groups were conducted: participant governed morning meeting, process group, AIDS education, relapse prevention training, goal development, goal review, assertiveness training, and many more services.	Usual treatment: once weeklyindividual and group counselling to substance abuseday treatment which included unique interventions forhomelessness in addition to substance abuse.	Group housing: congregate	Birmingham, Alabama, United States	High conditional
Milby (2008)	CM and CM+ (Contingency Management plus behavioral day treatment)	The study compared the abbreviated treatment model (offering contingency‐managed housing, vocational training and work therapy, i.e. CM) to the full treatment model (CM+), which included the same interventions plus extensive behavioral day treatment	Offering contingency‐managed housing, vocational training and work therapy.	Not specified	Birmingham, Alabama, USA. Participants were recruited, randomized, and treated from 2001 to 2004	High conditional
O'Connell (2012)	Housing and Urban Development–Veterans Affairs Supported Housing (HUD‐VASH) Intensive Care Management (ICM)	HUD VASH: intensive case management [ICM] plus rent subsidy vouchers ICM: Intensive case management only	Usual Treatment	Not specified	New Haven, Connecticut, USA. Around 1992–1995	Moderate unconditional
Sadowski (2009)	Housing and Case management (HOUSING FIRST)	Subjects are offered interim housing upon discharge from enrolling hospitalizations, followed by stable housing within 90 days. They have a case manager at each stage (hospital, respite/interim housing, and stable housing)	Usual Treatment: these patients receive usual social services for hospital discharge planning	Mix of scattered and congregate	Chicago, Illinois, USA. Around 2008	High unconditional
Shern (1997)	New York Street Study—specialised housing	The Boston project compared congregate consumer‐run housing with independent living. In New York specialized housing for homeless persons with severe mental illness was the primary housing resource. The San Diego project tested the importance of Section 8 housing certificates in obtaining and maintaining housing.	Usual treatment, usually some sort of case management	Varied. Some congregate. Some scattered	Baltimore, Boston, San Diego, New York City, USA	Moderate conditional
Siegel (2006)	Supported Housing vs community residences	Two types of supported housing. (1) tenants, mostly living alone, reside in studio or one‐bedroom apartments. No preconditions (e.g. sobriety) for housing, with ACT also provided. (2) Sobriety a requirement. Could be asked to leave if they did not maintain good neighbour status—for example, without too many complaints from neighbors. On‐site crisis services are continuously available to tenants in coordination with a New York City psychiatric emergency service	Community Residences: The three agencies that operate thenine community residences provideintensive supportive services in theresidences, which are only for personswith mental illness.	Supported Housing (congregate) vs. community residences, control (scattered)	New York City, USA	Intervention 1, high unconditional; Intervention 2, high conditional
Sosin (1996)	Housing and case management Case management alone	Housing and case management: This is conditional on abstinence and complying with treatment plan. The housing intervention provided the case management model along with supported housing. Brokerage case management without direct provision of services	No Treatment—"The users placed in the control condition were referred by the relevant short‐term program staff to an outpatient or inpatient substance abuse agency, to welfare offices''	Not specified	Chicago, United States. Recruitment around 1991–1992, intervention took place around 1995	Moderate conditional
Srebnik (2013)	Housing First	Intervention is characterized by rapid placement from homelessness directly into permanent housing, supported by assertive on‐site engagement and services but no requirement to participate or to achieve or maintain sobriety	No treatment	Congregate: BAH set aside 20 units at a newly renovated downtown building	King County, the City of Seattle, Washington, USA	High unconditional
Stefancic and Tsemberis (2007)	Pathways (HOUSING FIRST) Consortium (HOUSING FIRST)	Pathways to Housing (HOUSING FIRST) and Consortium (HOUSING FIRST) both provide Housing First services, in the form of independent scatter‐site apartments and ACT. Pathways had a more straightforward HOUSING FIRST approach, whereas Consortium was a newly formed mix of treatment and housing agencies from within the county but with no prior experience operating Housing First	Usual Treatment: usual array of services that included shelter basedprograms and transitional housing	Scatter Site	New York City, USA. 2000–2002	High unconditional

Source	Q	df	p
Total	471.76	17	< 0.001
Within designs	455.15	15	< 0.001
Between designs	16.61	2	< 0.001

Comparison	k	prop	nma	direct	indir.	Diff	z	p‐value
2:04	0	0	−0.9714	.	−0.9714	.	.	.
2:05	0	0	−0.8316	.	−0.8316	.	.	.
2:06	0	0	−0.8665	.	−0.8665	.	.	.
2:07	0	0	−0.6928	.	−0.6928	.	.	.
2:08	3	1	−1.1043	−1.1043	.	.	.	.
2:09	0	0	−0.4795	.	−0.4795	.	.	.
4:05	0	0	0.1398	.	0.1398	.	.	.
4:06	1	0.55	0.1048	0.2045	−0.0163	0.2208	0.23	0.8218
4:07	0	0	0.2786	.	0.2786	.	.	.
4:08	1	0.61	−0.133	−0.2197	0.001	−0.2208	−0.23	0.8218
4:09	0	0	0.4919	.	0.4919	.	.	.
5:06	0	0	−0.035	.	−0.035	.	.	.
5:07	0	0	0.1388	.	0.1388	.	.	.
5:08	3	1	−0.2728	−0.2728	.	.	.	.
5:09	0	0	0.3521	.	0.3521	.	.	.
6:07	0	0	0.1737	.	0.1737	.	.	.
6:08	0	0	−0.2378	.	−0.2378	.	.	.
6:09	4	0.9	0.3871	0.4095	0.1887	0.2208	0.23	0.8218
7:08	1	0.31	−0.4115	0.2092	−0.6911	0.9003	1.14	0.2561
7:09	3	0.77	0.2133	0.0106	0.9109	−0.9003	−1.14	0.2561
8:09	7	0.86	0.6249	0.6934	0.2014	0.492	0.77	0.4417

Intervention	P‐score (fixed)	P‐score (random)
8	0.9996	0.833
4	0.8067	0.6739
6	0.6075	0.6178
5	0.104	0.576
7	0.3201	0.4727
9	0.0768	0.2643
2	0.5854	0.0623

PERMALINK

Accommodation‐based interventions for individuals experiencing, or at risk of experiencing, homelessness

Ciara Keenan

Sarah Miller

Jennifer Hanratty

Terri Pigott

Jayne Hamilton

Christopher Coughlan

Peter Mackie

Suzanne Fitzpatrick

John Cowman

Abstract

Background

Objectives

Search Methods

Selection Criteria

Data Collection and Analysis

Main Results

Authors' Conclusions

1. PLAIN LANGUAGE SUMMARY

1.1. Accommodation‐based approaches help people remain healthy and stably housed

1.1.1. What is this review about?

1.1.2. What studies are included?

1.1.3. Do accommodation‐based approaches help people experiencing homelessness?

1.1.4. What implementation factors affect how well accommodation‐based approaches work?

1.1.5. What do the findings of this review mean?

1.1.6. How up‐to‐date is this review?

2. BACKGROUND

2.1. The problem, condition or issue

2.2. The intervention

2.2.1. Housing First

2.2.2. Rapid rehousing

2.2.3. Hostels

2.2.4. Shelters

2.2.5. Supported housing

2.2.6. Conclusion

2.3. A new typology

Table 1.

2.4. How the intervention might work

2.5. Why it is important to do this review

2.5.1. Previous reviews

Differences in population

Differences in outcomes of interest

Differences in analytic methods

Inclusion of qualitative studies

3. OBJECTIVES

4. METHODS

4.1. Criteria for considering studies for this review

4.1.1. Types of studies

4.1.2. Types of participants

4.1.3. Types of interventions

4.1.4. Types of outcome measures

Primary outcomes

Secondary outcomes

Types of settings

4.2. Search methods for identification of studies

4.2.1. Electronic searches

4.2.2. Searching other resources

4.3. Data collection and analysis

Figure 1.

4.3.1. Description of methods used in primary research

4.3.2. Criteria for determination of independent findings

4.3.3. Selection of studies

4.3.4. Data extraction and management

Details of study coding categories

4.3.5. Assessment of ROB in included studies

4.3.6. Measures of treatment effect

Statistical procedures and conventions

4.3.7. Unit of analysis issues

4.3.8. Dealing with missing data

4.3.9. Assessment of heterogeneity

4.3.10. Assessment of reporting biases

4.3.11. Data synthesis

4.3.12. Subgroup analysis and investigation of heterogeneity

Treatment of qualitative research

4.3.13. Sensitivity analysis

4.3.14. Summary of findings and assessment of the certainty of the evidence

5. RESULTS

5.1. Description of studies

5.1.1. Results of the search

Source	Q	df	p
Total	1852.22	16	<.001
Within designs	1795.81	13	<.001
Between designs	56.41	3	<.001

comparison	k	prop	nma	direct	indir.	Diff	z	p‐value
2:04	0	0	−0.1934	.	−0.1934	.	.	.
2:05	0	0	−0.4547	.	−0.4547	.	.	.
2:06	0	0	−0.3603	.	−0.3603	.	.	.
2:07	0	0	−0.309	.	−0.309	.	.	.
2:08	2	1	−0.317	−0.317	.	.	.	.
2:09	0	0	−0.0952	.	−0.0952	.	.	.
4:05	0	0	−0.2613	.	−0.2613	.	.	.
4:06	0	0	−0.1669	.	−0.1669	.	.	.
4:07	0	0	−0.1157	.	−0.1157	.	.	.
4:08	1	1	−0.1236	−0.1236	.	.	.	.
4:09	0	0	0.0982	.	0.0982	.	.	.
5:06	1	0.42	0.0944	0.0884	0.0988	−0.0104	−0.03	0.9796
5:07	0	0	0.1457	.	0.1457	.	.	.
5:08	2	0.58	0.1377	0.1992	0.052	0.1472	0.43	0.6649
5:09	1	0.3	0.3595	0.2436	0.4095	−0.1659	−0.45	0.654
6:07	0	0	0.0513	.	0.0513	.	.	.
6:08	0	0	0.0433	.	0.0433	.	.	.
6:09	3	0.8	0.2651	0.2631	0.2734	−0.0104	−0.03	0.9796
7:08	1	0.27	−0.0079	0.261	−0.1066	0.3676	0.99	0.3213
7:09	4	0.82	0.2138	0.1479	0.5155	−0.3676	−0.99	0.3213
8:09	7	0.8	0.2218	0.2766	−0.0015	0.2781	1.05	0.2947

Intervention	P‐score (fixed)	P‐score (random)
5	1	0.8215
6	0.8333	0.6716
8	0.6667	0.6157
7	0	0.5947
4	0.2224	0.4297
9	0.4995	0.1974
2	0.2781	0.1693