Poster #107 - Homelessness Among Families with Child Welfare Investigations: Prevalence, Co-Morbidities, and Case Opening Risks

• In Event: Friday Poster Reception

Abstract

Background and Purpose
Homelessness is a common factor for child protective services (referrals) and opening child welfare cases (Culhane et al., 2003; Farrell et al., 2017; Perlman & Fantuzzo). Homelessness exacerbates stress, and homeless families may have co-morbidities (e.g. mental health and substance use) that bring them to the attention of the child welfare system (Boxill & Beaty, 1990; Fowler & Schoeny, 2017; Hart-Shegos, 1999; Marcal, 2018). However, there is limited research on the prevalence of homelessness among families with child welfare investigations, and how the risk factors and outcomes of children in homeless families may vary from their non-homeless peers. This paper uses data from the Structured Decision Making (SDM) assessment used to investigate children referred to the San Francisco child welfare system between 2015-2023 to measure the proportion of children experiencing homelessness, their co-morbidities (e.g. mental health and substance use issues), and their risk for reaching an assessment level indicating that a child welfare case should open after investigation.

Methods
The co-occurring risk factors of homeless and non-homeless children are measured based on the proportion of homeless and non-homeless children with each risk factor identified on the SDM. The total risk factors on the SDM are summed and used to create a score indicating if a child welfare case should be opened. This total risk score is used to measure both the proportion of homeless and non-homeless children identified for case opening and ongoing child welfare services, and their likelihood of reaching this risk level using a multivariate regression model that includes child demographics and year of assessment. Descriptive statistics are used to assess the proportion of homeless children by state fiscal year, and co-morbidities of homeless and non-homeless children. Chi-square tests are used to determine if there is a statistically significant difference in the co-morbidities of homeless and non-homeless children. A logistic regression is used to determine the odds of a total risk assessment score reaching a level that indicates promotion to child welfare case opening for homeless and non-homeless children.

Findings
Overall, 9 percent children screened during this time period were identified as homeless. However, homelessness rates varied substantially by year, suggesting a need to clarify whether this trend reflects actual fluctuations or stems from an unclear definition of homelessness on screening tools. Homeless children have significantly more co-morbidities—more than twice as many children identified as homeless had caregiver mental health and substance abuse issues compared to non-homeless children. Homeless children also had more than three times the odds of having a total risk score reaching the level indicating that an on-going child welfare case should open.

Conclusion
These findings provide new information on the prevalence of homelessness among children investigated for child maltreatment. They underscore the importance of screening for homelessness and providing housing interventions to assist homeless families in addressing the various risk factors they face, thereby successfully closing their child welfare case.

This research does not necessarily reflect the views of the San Francisco Human Services Agency or the Wisconsin Department of Health Services.

Author

• Emily Rhodes, Wisconsin Department of Health Services
  Presenting Author