Skip to content

Topic Areas

Topic Areas

Target Population

Vulnerable adolescents aged 12-18 years, who can be identified as at risk of homelessness while still attending secondary schools, and their families

Target Population

Vulnerable adolescents aged 12-18 years, who can be identified as at risk of homelessness while still attending secondary schools, and their families

Program Overview

The COSS (Community of Schools and Services) Model of prevention/early intervention is a place-based and collective impact approach to the prevention of youth homelessness. The COSS Model is conceptually described in terms of four core foundations:

  • Community collaboration – a formalized collective consisting of schools and service providers
  • Annual cyclical population screening for all students using a series of risk indicators on the Australian Index of Adolescent Development [AIAD] survey instrument combined with local knowledge and a brief screening/engagement interview
  • A flexible and responsive practice framework with three levels of response:
    • Tier 1 – active monitoring
    • Tier 2 – short-term casework support
    • Tier 3 – wrap around case management for complex cases
  • Embedded outcomes monitoring and reporting to the community collective and practitioners, including a triennial project evaluation

Program Overview

The COSS (Community of Schools and Services) Model of prevention/early intervention is a place-based and collective impact approach to the prevention of youth homelessness. The COSS Model is conceptually described in terms of four core foundations:

  • Community collaboration – a formalized collective consisting of schools and service providers
  • Annual cyclical population screening for all students using a series of risk indicators on the Australian Index of Adolescent Development [AIAD] survey instrument combined with local knowledge and a brief screening/engagement interview
  • A flexible and responsive practice framework with three levels of response:
    • Tier 1 – active monitoring
    • Tier 2 – short-term casework support
    • Tier 3 – wrap around case management for complex cases
  • Embedded outcomes monitoring and reporting to the community collective and practitioners, including a triennial project evaluation

Contact Information

David Mackenzie

Dr. Tammy Hand

Contact Information

David Mackenzie

Dr. Tammy Hand

Program Goals

The goals of the COSS Model are:

  • Reduction in the risk of homelessness.
  • Continued engagement with education.
  • Increased functional health and well-being.
  • Active connection with community activities.
  • Active responsibility for self.
  • Improved family relations and natural supports.

Program Goals

The goals of the COSS Model are:

  • Reduction in the risk of homelessness.
  • Continued engagement with education.
  • Increased functional health and well-being.
  • Active connection with community activities.
  • Active responsibility for self.
  • Improved family relations and natural supports.

Logic Model

View the Logic Model (PDF) for Community of Schools and Services Model (COSS Model).

Logic Model

View the Logic Model (PDF) for Community of Schools and Services Model (COSS Model).

Essential Components

The essential components of The Community of Schools and Services Model (COSS Model) include:

  • The COSS Model is a place-based collective impact approach with a formal community-level architecture of collaborating schools and service providers.
  • Its methodology is based on identifying a range of risk(s) prior to crises and the model is an embedded with ongoing monitoring of support and measurement of outcomes.
  • The core features of the COSS Model are:
    • Community Collaboration and Shared Vision (Collective Impact):
      • Interagency agreements and partnerships:
        • Necessitates formal and informal collaboration between schools, youth services, family support agencies, and other community organizations
        • Goes beyond simple referrals to involve joint planning, decision-making, and coordinated service delivery
      • Common agenda and shared understanding:
        • All participating organizations share:
          • A clear, common vision of the problem (e.g., youth homelessness)
          • A joint strategy to address the problem
          • A shared vision of what a changed local service system would need to look like
        • Mutually reinforcing activities: While each partner organization may have distinct activities, they are coordinated and aligned to support the overall common agenda.
      • Early Identification and Proactive Screening:
        • Population-level screening:
          • A key innovation is the proactive identification of young people at risk—screening the entire state secondary school population in a community on an annual cyclical basis
          • Uses the proprietary Australian Index of Adolescent Development [AIAD] survey in a population screening process
        • Risk factor identification:
          • The AIAD survey, combined with other local knowledge, identifies various risk factors, including those related to:
            • Homelessness
            • Educational disengagement
            • Psychological distress
          • This allows for intervention before a crisis point is reached.
      • Flexible practice framework and early intervention support work with families:
          • Youth-focused and family-centered case management:
            • Support is provided directly to the young person as the primary client but also involves working with their family and natural support persons as appropriate
          • Differentiated response levels:
            • The model employs a multi-tiered approach, from active monitoring and secondary consultations (Tier One) to more intensive case management (Tier Two) and wrap-around case management with multiple support (Tier Three), ensuring responses are needs-based, comprehensive, and flexible, and tailored to individual needs
          • Integrated support:
            • Services are not delivered in discrete agency siloes but are delivered in a coordinated and often integrated manner, addressing multiple needs simultaneously (e.g., education, housing, and mental health).
    • Robust, embedded longitudinal monitoring and measurement of outcomes:
      • The model emphasizes continuous data collection to track progress and measure outcomes and to evaluate the effectiveness of interventions case by case and overall. This includes monitoring:
        • Reductions in risk factors
        • Improvements in educational engagement
        • Reduced entry into crisis services across individual, cohort, and community levels
      • All COSS community sites use consistent data collection and measurement tools to ensure alignment and accountability.
      • Continuous feedback loops ensure that information gathered from monitoring is used to refine practices and improve the overall support being delivered to the identified young people and their families.
    • Backbone support, data management, and fidelity guidance:
      • Dedicated project leads/coordinators at the local level, where there is often a dedicated lead worker or lead agency team to support the operation of the community collective.
      • Data management for COSS sites is a specialist and ongoing feature of the COSS Model and is provided by external backbone support agency.
      • Fidelity guidance is delivered via ongoing backbone support and through an Upstream Community of Practice network.
      • There is a need for broader systemic backbone support to coordinate and facilitate the implementation of the COSS Model across multiple communities.
      • Advocacy to governments and philanthropy for the resources to expand the number of COSS sites with a perspective for systemic change longer-term.
    • The COSS Model represents a shift from reactive, siloed service delivery to a proactive, collaborative, and data-driven place-based and collective impact approach focused on preventing issues from escalating and improving the long-term social and educational outcomes for vulnerable adolescents. This approach is premised on the propositions:
      • That schools are crucial sites for prevention and early identification; however, purely school-based student support interventions are inherently limited. By contrast, the COSS Model is a community-school linked model through a collective effort of schools and services, and
      • That community-level intervention requires a whole-of-community effort based on integrated activities and interventions rather than various separate programs operating in relative isolation from each other.

Essential Components

The essential components of The Community of Schools and Services Model (COSS Model) include:

  • The COSS Model is a place-based collective impact approach with a formal community-level architecture of collaborating schools and service providers.
  • Its methodology is based on identifying a range of risk(s) prior to crises and the model is an embedded with ongoing monitoring of support and measurement of outcomes.
  • The core features of the COSS Model are:
    • Community Collaboration and Shared Vision (Collective Impact):
      • Interagency agreements and partnerships:
        • Necessitates formal and informal collaboration between schools, youth services, family support agencies, and other community organizations
        • Goes beyond simple referrals to involve joint planning, decision-making, and coordinated service delivery
      • Common agenda and shared understanding:
        • All participating organizations share:
          • A clear, common vision of the problem (e.g., youth homelessness)
          • A joint strategy to address the problem
          • A shared vision of what a changed local service system would need to look like
        • Mutually reinforcing activities: While each partner organization may have distinct activities, they are coordinated and aligned to support the overall common agenda.
      • Early Identification and Proactive Screening:
        • Population-level screening:
          • A key innovation is the proactive identification of young people at risk—screening the entire state secondary school population in a community on an annual cyclical basis
          • Uses the proprietary Australian Index of Adolescent Development [AIAD] survey in a population screening process
        • Risk factor identification:
          • The AIAD survey, combined with other local knowledge, identifies various risk factors, including those related to:
            • Homelessness
            • Educational disengagement
            • Psychological distress
          • This allows for intervention before a crisis point is reached.
      • Flexible practice framework and early intervention support work with families:
          • Youth-focused and family-centered case management:
            • Support is provided directly to the young person as the primary client but also involves working with their family and natural support persons as appropriate
          • Differentiated response levels:
            • The model employs a multi-tiered approach, from active monitoring and secondary consultations (Tier One) to more intensive case management (Tier Two) and wrap-around case management with multiple support (Tier Three), ensuring responses are needs-based, comprehensive, and flexible, and tailored to individual needs
          • Integrated support:
            • Services are not delivered in discrete agency siloes but are delivered in a coordinated and often integrated manner, addressing multiple needs simultaneously (e.g., education, housing, and mental health).
    • Robust, embedded longitudinal monitoring and measurement of outcomes:
      • The model emphasizes continuous data collection to track progress and measure outcomes and to evaluate the effectiveness of interventions case by case and overall. This includes monitoring:
        • Reductions in risk factors
        • Improvements in educational engagement
        • Reduced entry into crisis services across individual, cohort, and community levels
      • All COSS community sites use consistent data collection and measurement tools to ensure alignment and accountability.
      • Continuous feedback loops ensure that information gathered from monitoring is used to refine practices and improve the overall support being delivered to the identified young people and their families.
    • Backbone support, data management, and fidelity guidance:
      • Dedicated project leads/coordinators at the local level, where there is often a dedicated lead worker or lead agency team to support the operation of the community collective.
      • Data management for COSS sites is a specialist and ongoing feature of the COSS Model and is provided by external backbone support agency.
      • Fidelity guidance is delivered via ongoing backbone support and through an Upstream Community of Practice network.
      • There is a need for broader systemic backbone support to coordinate and facilitate the implementation of the COSS Model across multiple communities.
      • Advocacy to governments and philanthropy for the resources to expand the number of COSS sites with a perspective for systemic change longer-term.
    • The COSS Model represents a shift from reactive, siloed service delivery to a proactive, collaborative, and data-driven place-based and collective impact approach focused on preventing issues from escalating and improving the long-term social and educational outcomes for vulnerable adolescents. This approach is premised on the propositions:
      • That schools are crucial sites for prevention and early identification; however, purely school-based student support interventions are inherently limited. By contrast, the COSS Model is a community-school linked model through a collective effort of schools and services, and
      • That community-level intervention requires a whole-of-community effort based on integrated activities and interventions rather than various separate programs operating in relative isolation from each other.

Program Delivery

Recommended Intensity

The number of contacts and the intensity of support work varies in terms of a determination of the appropriateness of tiered support (i.e., 1, 2, and 3) in each case. A typical profile would be: Tier 1 response, ~25% of clients – up to 20 hours/year; Tier 2 response ~50% of clients – 20-100 hours/year; Tier 3 response ~ 25% of clients – 100-160 hours/year


Recommended Duration

The episodes of intervention with adolescents and their families are as necessary and determined by a tiered determination with flexibility for a youth on Tier 1, which a watching brief with occasional direct contacts, to move into a case-work Tier 2 mode or even an intensive Tier 3 wrap-around family support mode, if, and when, their situation changes. The duration of engagement with the COSS Model in a community is for student’s entire time in secondary school. Once identified as at risk or whenever a student is identified as at risk they are considered as part of the cohort for which the COSS Model remains active on a ‘whenever support is needed’ basis.


Delivery Settings

This program is typically conducted in a(n):

  • Adoptive Home
  • Birth Family Home
  • Community Daily Living Setting
  • Community-based Agency / Organization / Provider
  • Foster / Kinship Care
  • School Setting (Including: Day Care, Day Treatment Programs, etc.)

Homework

This program does not include a homework component.


Resources Needed to Run Program

The typical resources for implementing the program are:

In terms of resources for the COSS Model, human resources are a key issue.

  • The lead agency of the local community collective is an existing funded Specialist Homelessness (crisis) service provider capable of leading a community collaboration as required by the COSS Model. The number of staff funded under the COSS Model ideally is determined by an algorithm based on the number of identified at-risk adolescents.
  • In Australia, the participating secondary schools typically have at least one dedicated in-school welfare support position and at least a part-time psychology trained counsellor. However, the complement of in-school student support staff varies across the Australian states and territories.

The Upstream Australia backbone support staff provide the core backbone support activities and are funded as part of service delivery costs to of funding COSS local lead agencies.

Program Delivery

Recommended Intensity

The number of contacts and the intensity of support work varies in terms of a determination of the appropriateness of tiered support (i.e., 1, 2, and 3) in each case. A typical profile would be: Tier 1 response, ~25% of clients – up to 20 hours/year; Tier 2 response ~50% of clients – 20-100 hours/year; Tier 3 response ~ 25% of clients – 100-160 hours/year


Recommended Duration

The episodes of intervention with adolescents and their families are as necessary and determined by a tiered determination with flexibility for a youth on Tier 1, which a watching brief with occasional direct contacts, to move into a case-work Tier 2 mode or even an intensive Tier 3 wrap-around family support mode, if, and when, their situation changes. The duration of engagement with the COSS Model in a community is for student’s entire time in secondary school. Once identified as at risk or whenever a student is identified as at risk they are considered as part of the cohort for which the COSS Model remains active on a ‘whenever support is needed’ basis.


Delivery Settings

This program is typically conducted in a(n):

  • Adoptive Home
  • Birth Family Home
  • Community Daily Living Setting
  • Community-based Agency / Organization / Provider
  • Foster / Kinship Care
  • School Setting (Including: Day Care, Day Treatment Programs, etc.)

Homework

This program does not include a homework component.


Resources Needed to Run Program

The typical resources for implementing the program are:

In terms of resources for the COSS Model, human resources are a key issue.

  • The lead agency of the local community collective is an existing funded Specialist Homelessness (crisis) service provider capable of leading a community collaboration as required by the COSS Model. The number of staff funded under the COSS Model ideally is determined by an algorithm based on the number of identified at-risk adolescents.
  • In Australia, the participating secondary schools typically have at least one dedicated in-school welfare support position and at least a part-time psychology trained counsellor. However, the complement of in-school student support staff varies across the Australian states and territories.

The Upstream Australia backbone support staff provide the core backbone support activities and are funded as part of service delivery costs to of funding COSS local lead agencies.

Manuals and Training

Prerequisite/Minimum Provider Qualifications

COSS Model early intervention workers would normally be expected to have a degree in social work, social science, youth work, or some other equivalent tertiary welfare qualification and be able to undertake support work with vulnerable adolescents and work with their family members as well – a mode of practice described in Australia as youth-focused family-centered supportive case work.


Manual Information

There is a manual that describes how to deliver this program.


Program Manual(s)

The COSS Model manual is not publicly available except to agencies implementing the COSS Model in a formal collaboration project that includes the Upstream Australia team who are the innovative developers and providers of backbone support. For access to further information about the COSS Model, contact the training contact below.


Training Information

There is training available for this program.

Training Contact

Training Type/Location:

The first six months to one year of implementing the COSS Model in a community is used to undertake a process of community-building of the formal collaboration between the secondary schools in a community and the community services provider(s) whose role is doing the youth and family work. This stage involves education and training as well as planning for the implementation of the COSS Model. Some training is provided through online webinars and training sessions and there are regular meetings conducted onsite to build close relations amongst the project stakeholders and a shared vision for a changed local service system.

Number of days/hours:

Varies according to the readiness of the lead agency and its staff as well as school welfare staff for implementing the COSS Model.

Implementation resources such as pre-implementation materials, implementation support, fidelity measures, implementation manual, and previous implementation research are available via the program’s Contact Information link.

Manuals and Training

Prerequisite/Minimum Provider Qualifications

COSS Model early intervention workers would normally be expected to have a degree in social work, social science, youth work, or some other equivalent tertiary welfare qualification and be able to undertake support work with vulnerable adolescents and work with their family members as well – a mode of practice described in Australia as youth-focused family-centered supportive case work.


Manual Information

There is a manual that describes how to deliver this program.


Program Manual(s)

The COSS Model manual is not publicly available except to agencies implementing the COSS Model in a formal collaboration project that includes the Upstream Australia team who are the innovative developers and providers of backbone support. For access to further information about the COSS Model, contact the training contact below.


Training Information

There is training available for this program.

Training Contact

Training Type/Location:

The first six months to one year of implementing the COSS Model in a community is used to undertake a process of community-building of the formal collaboration between the secondary schools in a community and the community services provider(s) whose role is doing the youth and family work. This stage involves education and training as well as planning for the implementation of the COSS Model. Some training is provided through online webinars and training sessions and there are regular meetings conducted onsite to build close relations amongst the project stakeholders and a shared vision for a changed local service system.

Number of days/hours:

Varies according to the readiness of the lead agency and its staff as well as school welfare staff for implementing the COSS Model.

Implementation resources such as pre-implementation materials, implementation support, fidelity measures, implementation manual, and previous implementation research are available via the program’s Contact Information link.

Relevant Published, Peer-Reviewed Research

What is included in the Published, Peer-Reviewed Research section?

  • MacKenzie, D., Hand, T., & Gill, P. (2024). The ‘Community of Schools and Services’(COSS) Model of Early Intervention: A system-changing innovation for the prevention of youth homelessness. Youth, 4(3), 1305–1321. https://doi.org/10.3390/youth4030082

    Type of Study: Other quasi-experimental

    Participants: Not specified

    Sample / Population:

    • Age — 12–18 years
    • Race/Ethnicity — Not specified
    • Gender — Not specified
    • Status

      Participants were secondary students at-risk for homelessness.

    Location/Institution: Regional City of Albury, 550 km from Sydney, in the State of New South Wales

    Summary:

    The purpose of the study was to present the findings of the Embedded Development and Outcomes Measurement (EDOM) report, which is a feature of the Community of Schools and Services (COSS) Model. Participants were identified youth who became homeless with 2 years post identification. Measures utilized include the Australian Index of Adolescent Development (AIAD) Survey. Results indicate that when COSS Model support is delivered to identified at-risk students, (1) 40–50% of individuals are no longer at such high risk of homelessness 12-months later; (2) only 3–5% of students identified as at risk of homelessness and supported through the COSS Model sought assistance from local homelessness services in the following two years; and (3) the flow of adolescents (12–18 years) into the local homelessness services was reduced by 40% from 2019 to 2023. Limitations include that the outcomes data are based on a single COSS Model community and the use of an interrupted times series (ITS) design.

    Length of controlled postintervention follow-up: None.

Relevant Published, Peer-Reviewed Research

What is included in the Published, Peer-Reviewed Research section?

  • MacKenzie, D., Hand, T., & Gill, P. (2024). The ‘Community of Schools and Services’(COSS) Model of Early Intervention: A system-changing innovation for the prevention of youth homelessness. Youth, 4(3), 1305–1321. https://doi.org/10.3390/youth4030082

    Type of Study: Other quasi-experimental

    Participants: Not specified

    Sample / Population:

    • Age — 12–18 years
    • Race/Ethnicity — Not specified
    • Gender — Not specified
    • Status

      Participants were secondary students at-risk for homelessness.

    Location/Institution: Regional City of Albury, 550 km from Sydney, in the State of New South Wales

    Summary:

    The purpose of the study was to present the findings of the Embedded Development and Outcomes Measurement (EDOM) report, which is a feature of the Community of Schools and Services (COSS) Model. Participants were identified youth who became homeless with 2 years post identification. Measures utilized include the Australian Index of Adolescent Development (AIAD) Survey. Results indicate that when COSS Model support is delivered to identified at-risk students, (1) 40–50% of individuals are no longer at such high risk of homelessness 12-months later; (2) only 3–5% of students identified as at risk of homelessness and supported through the COSS Model sought assistance from local homelessness services in the following two years; and (3) the flow of adolescents (12–18 years) into the local homelessness services was reduced by 40% from 2019 to 2023. Limitations include that the outcomes data are based on a single COSS Model community and the use of an interrupted times series (ITS) design.

    Length of controlled postintervention follow-up: None.

Additional References

Additional References

Topic Areas

Topic Areas

Target Population

Vulnerable adolescents aged 12-18 years, who can be identified as at risk of homelessness while still attending secondary schools, and their families

Target Population

Vulnerable adolescents aged 12-18 years, who can be identified as at risk of homelessness while still attending secondary schools, and their families

Program Overview

The COSS (Community of Schools and Services) Model of prevention/early intervention is a place-based and collective impact approach to the prevention of youth homelessness. The COSS Model is conceptually described in terms of four core foundations:

  • Community collaboration – a formalized collective consisting of schools and service providers
  • Annual cyclical population screening for all students using a series of risk indicators on the Australian Index of Adolescent Development [AIAD] survey instrument combined with local knowledge and a brief screening/engagement interview
  • A flexible and responsive practice framework with three levels of response:
    • Tier 1 – active monitoring
    • Tier 2 – short-term casework support
    • Tier 3 – wrap around case management for complex cases
  • Embedded outcomes monitoring and reporting to the community collective and practitioners, including a triennial project evaluation

Program Overview

The COSS (Community of Schools and Services) Model of prevention/early intervention is a place-based and collective impact approach to the prevention of youth homelessness. The COSS Model is conceptually described in terms of four core foundations:

  • Community collaboration – a formalized collective consisting of schools and service providers
  • Annual cyclical population screening for all students using a series of risk indicators on the Australian Index of Adolescent Development [AIAD] survey instrument combined with local knowledge and a brief screening/engagement interview
  • A flexible and responsive practice framework with three levels of response:
    • Tier 1 – active monitoring
    • Tier 2 – short-term casework support
    • Tier 3 – wrap around case management for complex cases
  • Embedded outcomes monitoring and reporting to the community collective and practitioners, including a triennial project evaluation

Contact Information

David Mackenzie

Dr. Tammy Hand

Contact Information

David Mackenzie

Dr. Tammy Hand

Program Goals

The goals of the COSS Model are:

  • Reduction in the risk of homelessness.
  • Continued engagement with education.
  • Increased functional health and well-being.
  • Active connection with community activities.
  • Active responsibility for self.
  • Improved family relations and natural supports.

Program Goals

The goals of the COSS Model are:

  • Reduction in the risk of homelessness.
  • Continued engagement with education.
  • Increased functional health and well-being.
  • Active connection with community activities.
  • Active responsibility for self.
  • Improved family relations and natural supports.

Logic Model

View the Logic Model (PDF) for Community of Schools and Services Model (COSS Model).

Logic Model

View the Logic Model (PDF) for Community of Schools and Services Model (COSS Model).

Essential Components

The essential components of The Community of Schools and Services Model (COSS Model) include:

  • The COSS Model is a place-based collective impact approach with a formal community-level architecture of collaborating schools and service providers.
  • Its methodology is based on identifying a range of risk(s) prior to crises and the model is an embedded with ongoing monitoring of support and measurement of outcomes.
  • The core features of the COSS Model are:
    • Community Collaboration and Shared Vision (Collective Impact):
      • Interagency agreements and partnerships:
        • Necessitates formal and informal collaboration between schools, youth services, family support agencies, and other community organizations
        • Goes beyond simple referrals to involve joint planning, decision-making, and coordinated service delivery
      • Common agenda and shared understanding:
        • All participating organizations share:
          • A clear, common vision of the problem (e.g., youth homelessness)
          • A joint strategy to address the problem
          • A shared vision of what a changed local service system would need to look like
        • Mutually reinforcing activities: While each partner organization may have distinct activities, they are coordinated and aligned to support the overall common agenda.
      • Early Identification and Proactive Screening:
        • Population-level screening:
          • A key innovation is the proactive identification of young people at risk—screening the entire state secondary school population in a community on an annual cyclical basis
          • Uses the proprietary Australian Index of Adolescent Development [AIAD] survey in a population screening process
        • Risk factor identification:
          • The AIAD survey, combined with other local knowledge, identifies various risk factors, including those related to:
            • Homelessness
            • Educational disengagement
            • Psychological distress
          • This allows for intervention before a crisis point is reached.
      • Flexible practice framework and early intervention support work with families:
          • Youth-focused and family-centered case management:
            • Support is provided directly to the young person as the primary client but also involves working with their family and natural support persons as appropriate
          • Differentiated response levels:
            • The model employs a multi-tiered approach, from active monitoring and secondary consultations (Tier One) to more intensive case management (Tier Two) and wrap-around case management with multiple support (Tier Three), ensuring responses are needs-based, comprehensive, and flexible, and tailored to individual needs
          • Integrated support:
            • Services are not delivered in discrete agency siloes but are delivered in a coordinated and often integrated manner, addressing multiple needs simultaneously (e.g., education, housing, and mental health).
    • Robust, embedded longitudinal monitoring and measurement of outcomes:
      • The model emphasizes continuous data collection to track progress and measure outcomes and to evaluate the effectiveness of interventions case by case and overall. This includes monitoring:
        • Reductions in risk factors
        • Improvements in educational engagement
        • Reduced entry into crisis services across individual, cohort, and community levels
      • All COSS community sites use consistent data collection and measurement tools to ensure alignment and accountability.
      • Continuous feedback loops ensure that information gathered from monitoring is used to refine practices and improve the overall support being delivered to the identified young people and their families.
    • Backbone support, data management, and fidelity guidance:
      • Dedicated project leads/coordinators at the local level, where there is often a dedicated lead worker or lead agency team to support the operation of the community collective.
      • Data management for COSS sites is a specialist and ongoing feature of the COSS Model and is provided by external backbone support agency.
      • Fidelity guidance is delivered via ongoing backbone support and through an Upstream Community of Practice network.
      • There is a need for broader systemic backbone support to coordinate and facilitate the implementation of the COSS Model across multiple communities.
      • Advocacy to governments and philanthropy for the resources to expand the number of COSS sites with a perspective for systemic change longer-term.
    • The COSS Model represents a shift from reactive, siloed service delivery to a proactive, collaborative, and data-driven place-based and collective impact approach focused on preventing issues from escalating and improving the long-term social and educational outcomes for vulnerable adolescents. This approach is premised on the propositions:
      • That schools are crucial sites for prevention and early identification; however, purely school-based student support interventions are inherently limited. By contrast, the COSS Model is a community-school linked model through a collective effort of schools and services, and
      • That community-level intervention requires a whole-of-community effort based on integrated activities and interventions rather than various separate programs operating in relative isolation from each other.

Essential Components

The essential components of The Community of Schools and Services Model (COSS Model) include:

  • The COSS Model is a place-based collective impact approach with a formal community-level architecture of collaborating schools and service providers.
  • Its methodology is based on identifying a range of risk(s) prior to crises and the model is an embedded with ongoing monitoring of support and measurement of outcomes.
  • The core features of the COSS Model are:
    • Community Collaboration and Shared Vision (Collective Impact):
      • Interagency agreements and partnerships:
        • Necessitates formal and informal collaboration between schools, youth services, family support agencies, and other community organizations
        • Goes beyond simple referrals to involve joint planning, decision-making, and coordinated service delivery
      • Common agenda and shared understanding:
        • All participating organizations share:
          • A clear, common vision of the problem (e.g., youth homelessness)
          • A joint strategy to address the problem
          • A shared vision of what a changed local service system would need to look like
        • Mutually reinforcing activities: While each partner organization may have distinct activities, they are coordinated and aligned to support the overall common agenda.
      • Early Identification and Proactive Screening:
        • Population-level screening:
          • A key innovation is the proactive identification of young people at risk—screening the entire state secondary school population in a community on an annual cyclical basis
          • Uses the proprietary Australian Index of Adolescent Development [AIAD] survey in a population screening process
        • Risk factor identification:
          • The AIAD survey, combined with other local knowledge, identifies various risk factors, including those related to:
            • Homelessness
            • Educational disengagement
            • Psychological distress
          • This allows for intervention before a crisis point is reached.
      • Flexible practice framework and early intervention support work with families:
          • Youth-focused and family-centered case management:
            • Support is provided directly to the young person as the primary client but also involves working with their family and natural support persons as appropriate
          • Differentiated response levels:
            • The model employs a multi-tiered approach, from active monitoring and secondary consultations (Tier One) to more intensive case management (Tier Two) and wrap-around case management with multiple support (Tier Three), ensuring responses are needs-based, comprehensive, and flexible, and tailored to individual needs
          • Integrated support:
            • Services are not delivered in discrete agency siloes but are delivered in a coordinated and often integrated manner, addressing multiple needs simultaneously (e.g., education, housing, and mental health).
    • Robust, embedded longitudinal monitoring and measurement of outcomes:
      • The model emphasizes continuous data collection to track progress and measure outcomes and to evaluate the effectiveness of interventions case by case and overall. This includes monitoring:
        • Reductions in risk factors
        • Improvements in educational engagement
        • Reduced entry into crisis services across individual, cohort, and community levels
      • All COSS community sites use consistent data collection and measurement tools to ensure alignment and accountability.
      • Continuous feedback loops ensure that information gathered from monitoring is used to refine practices and improve the overall support being delivered to the identified young people and their families.
    • Backbone support, data management, and fidelity guidance:
      • Dedicated project leads/coordinators at the local level, where there is often a dedicated lead worker or lead agency team to support the operation of the community collective.
      • Data management for COSS sites is a specialist and ongoing feature of the COSS Model and is provided by external backbone support agency.
      • Fidelity guidance is delivered via ongoing backbone support and through an Upstream Community of Practice network.
      • There is a need for broader systemic backbone support to coordinate and facilitate the implementation of the COSS Model across multiple communities.
      • Advocacy to governments and philanthropy for the resources to expand the number of COSS sites with a perspective for systemic change longer-term.
    • The COSS Model represents a shift from reactive, siloed service delivery to a proactive, collaborative, and data-driven place-based and collective impact approach focused on preventing issues from escalating and improving the long-term social and educational outcomes for vulnerable adolescents. This approach is premised on the propositions:
      • That schools are crucial sites for prevention and early identification; however, purely school-based student support interventions are inherently limited. By contrast, the COSS Model is a community-school linked model through a collective effort of schools and services, and
      • That community-level intervention requires a whole-of-community effort based on integrated activities and interventions rather than various separate programs operating in relative isolation from each other.

Program Delivery

Recommended Intensity

The number of contacts and the intensity of support work varies in terms of a determination of the appropriateness of tiered support (i.e., 1, 2, and 3) in each case. A typical profile would be: Tier 1 response, ~25% of clients – up to 20 hours/year; Tier 2 response ~50% of clients – 20-100 hours/year; Tier 3 response ~ 25% of clients – 100-160 hours/year


Recommended Duration

The episodes of intervention with adolescents and their families are as necessary and determined by a tiered determination with flexibility for a youth on Tier 1, which a watching brief with occasional direct contacts, to move into a case-work Tier 2 mode or even an intensive Tier 3 wrap-around family support mode, if, and when, their situation changes. The duration of engagement with the COSS Model in a community is for student’s entire time in secondary school. Once identified as at risk or whenever a student is identified as at risk they are considered as part of the cohort for which the COSS Model remains active on a ‘whenever support is needed’ basis.


Delivery Settings

This program is typically conducted in a(n):

  • Adoptive Home
  • Birth Family Home
  • Community Daily Living Setting
  • Community-based Agency / Organization / Provider
  • Foster / Kinship Care
  • School Setting (Including: Day Care, Day Treatment Programs, etc.)

Homework

This program does not include a homework component.


Resources Needed to Run Program

The typical resources for implementing the program are:

In terms of resources for the COSS Model, human resources are a key issue.

  • The lead agency of the local community collective is an existing funded Specialist Homelessness (crisis) service provider capable of leading a community collaboration as required by the COSS Model. The number of staff funded under the COSS Model ideally is determined by an algorithm based on the number of identified at-risk adolescents.
  • In Australia, the participating secondary schools typically have at least one dedicated in-school welfare support position and at least a part-time psychology trained counsellor. However, the complement of in-school student support staff varies across the Australian states and territories.

The Upstream Australia backbone support staff provide the core backbone support activities and are funded as part of service delivery costs to of funding COSS local lead agencies.

Program Delivery

Recommended Intensity

The number of contacts and the intensity of support work varies in terms of a determination of the appropriateness of tiered support (i.e., 1, 2, and 3) in each case. A typical profile would be: Tier 1 response, ~25% of clients – up to 20 hours/year; Tier 2 response ~50% of clients – 20-100 hours/year; Tier 3 response ~ 25% of clients – 100-160 hours/year


Recommended Duration

The episodes of intervention with adolescents and their families are as necessary and determined by a tiered determination with flexibility for a youth on Tier 1, which a watching brief with occasional direct contacts, to move into a case-work Tier 2 mode or even an intensive Tier 3 wrap-around family support mode, if, and when, their situation changes. The duration of engagement with the COSS Model in a community is for student’s entire time in secondary school. Once identified as at risk or whenever a student is identified as at risk they are considered as part of the cohort for which the COSS Model remains active on a ‘whenever support is needed’ basis.


Delivery Settings

This program is typically conducted in a(n):

  • Adoptive Home
  • Birth Family Home
  • Community Daily Living Setting
  • Community-based Agency / Organization / Provider
  • Foster / Kinship Care
  • School Setting (Including: Day Care, Day Treatment Programs, etc.)

Homework

This program does not include a homework component.


Resources Needed to Run Program

The typical resources for implementing the program are:

In terms of resources for the COSS Model, human resources are a key issue.

  • The lead agency of the local community collective is an existing funded Specialist Homelessness (crisis) service provider capable of leading a community collaboration as required by the COSS Model. The number of staff funded under the COSS Model ideally is determined by an algorithm based on the number of identified at-risk adolescents.
  • In Australia, the participating secondary schools typically have at least one dedicated in-school welfare support position and at least a part-time psychology trained counsellor. However, the complement of in-school student support staff varies across the Australian states and territories.

The Upstream Australia backbone support staff provide the core backbone support activities and are funded as part of service delivery costs to of funding COSS local lead agencies.

Manuals and Training

Prerequisite/Minimum Provider Qualifications

COSS Model early intervention workers would normally be expected to have a degree in social work, social science, youth work, or some other equivalent tertiary welfare qualification and be able to undertake support work with vulnerable adolescents and work with their family members as well – a mode of practice described in Australia as youth-focused family-centered supportive case work.


Manual Information

There is a manual that describes how to deliver this program.


Program Manual(s)

The COSS Model manual is not publicly available except to agencies implementing the COSS Model in a formal collaboration project that includes the Upstream Australia team who are the innovative developers and providers of backbone support. For access to further information about the COSS Model, contact the training contact below.


Training Information

There is training available for this program.

Training Contact

Training Type/Location:

The first six months to one year of implementing the COSS Model in a community is used to undertake a process of community-building of the formal collaboration between the secondary schools in a community and the community services provider(s) whose role is doing the youth and family work. This stage involves education and training as well as planning for the implementation of the COSS Model. Some training is provided through online webinars and training sessions and there are regular meetings conducted onsite to build close relations amongst the project stakeholders and a shared vision for a changed local service system.

Number of days/hours:

Varies according to the readiness of the lead agency and its staff as well as school welfare staff for implementing the COSS Model.

Implementation resources such as pre-implementation materials, implementation support, fidelity measures, implementation manual, and previous implementation research are available via the program’s Contact Information link.

Manuals and Training

Prerequisite/Minimum Provider Qualifications

COSS Model early intervention workers would normally be expected to have a degree in social work, social science, youth work, or some other equivalent tertiary welfare qualification and be able to undertake support work with vulnerable adolescents and work with their family members as well – a mode of practice described in Australia as youth-focused family-centered supportive case work.


Manual Information

There is a manual that describes how to deliver this program.


Program Manual(s)

The COSS Model manual is not publicly available except to agencies implementing the COSS Model in a formal collaboration project that includes the Upstream Australia team who are the innovative developers and providers of backbone support. For access to further information about the COSS Model, contact the training contact below.


Training Information

There is training available for this program.

Training Contact

Training Type/Location:

The first six months to one year of implementing the COSS Model in a community is used to undertake a process of community-building of the formal collaboration between the secondary schools in a community and the community services provider(s) whose role is doing the youth and family work. This stage involves education and training as well as planning for the implementation of the COSS Model. Some training is provided through online webinars and training sessions and there are regular meetings conducted onsite to build close relations amongst the project stakeholders and a shared vision for a changed local service system.

Number of days/hours:

Varies according to the readiness of the lead agency and its staff as well as school welfare staff for implementing the COSS Model.

Implementation resources such as pre-implementation materials, implementation support, fidelity measures, implementation manual, and previous implementation research are available via the program’s Contact Information link.

Relevant Published, Peer-Reviewed Research

What is included in the Published, Peer-Reviewed Research section?

  • MacKenzie, D., Hand, T., & Gill, P. (2024). The ‘Community of Schools and Services’(COSS) Model of Early Intervention: A system-changing innovation for the prevention of youth homelessness. Youth, 4(3), 1305–1321. https://doi.org/10.3390/youth4030082

    Type of Study: Other quasi-experimental

    Participants: Not specified

    Sample / Population:

    • Age — 12–18 years
    • Race/Ethnicity — Not specified
    • Gender — Not specified
    • Status

      Participants were secondary students at-risk for homelessness.

    Location/Institution: Regional City of Albury, 550 km from Sydney, in the State of New South Wales

    Summary:

    The purpose of the study was to present the findings of the Embedded Development and Outcomes Measurement (EDOM) report, which is a feature of the Community of Schools and Services (COSS) Model. Participants were identified youth who became homeless with 2 years post identification. Measures utilized include the Australian Index of Adolescent Development (AIAD) Survey. Results indicate that when COSS Model support is delivered to identified at-risk students, (1) 40–50% of individuals are no longer at such high risk of homelessness 12-months later; (2) only 3–5% of students identified as at risk of homelessness and supported through the COSS Model sought assistance from local homelessness services in the following two years; and (3) the flow of adolescents (12–18 years) into the local homelessness services was reduced by 40% from 2019 to 2023. Limitations include that the outcomes data are based on a single COSS Model community and the use of an interrupted times series (ITS) design.

    Length of controlled postintervention follow-up: None.

Relevant Published, Peer-Reviewed Research

What is included in the Published, Peer-Reviewed Research section?

  • MacKenzie, D., Hand, T., & Gill, P. (2024). The ‘Community of Schools and Services’(COSS) Model of Early Intervention: A system-changing innovation for the prevention of youth homelessness. Youth, 4(3), 1305–1321. https://doi.org/10.3390/youth4030082

    Type of Study: Other quasi-experimental

    Participants: Not specified

    Sample / Population:

    • Age — 12–18 years
    • Race/Ethnicity — Not specified
    • Gender — Not specified
    • Status

      Participants were secondary students at-risk for homelessness.

    Location/Institution: Regional City of Albury, 550 km from Sydney, in the State of New South Wales

    Summary:

    The purpose of the study was to present the findings of the Embedded Development and Outcomes Measurement (EDOM) report, which is a feature of the Community of Schools and Services (COSS) Model. Participants were identified youth who became homeless with 2 years post identification. Measures utilized include the Australian Index of Adolescent Development (AIAD) Survey. Results indicate that when COSS Model support is delivered to identified at-risk students, (1) 40–50% of individuals are no longer at such high risk of homelessness 12-months later; (2) only 3–5% of students identified as at risk of homelessness and supported through the COSS Model sought assistance from local homelessness services in the following two years; and (3) the flow of adolescents (12–18 years) into the local homelessness services was reduced by 40% from 2019 to 2023. Limitations include that the outcomes data are based on a single COSS Model community and the use of an interrupted times series (ITS) design.

    Length of controlled postintervention follow-up: None.

Additional References

Additional References

Date CEBC Staff Last Reviewed Research: January 2025

Date Program's Staff Last Reviewed Content: January 2026

Date Originally Loaded onto CEBC: January 2026