An Evaluation of Family-to-Family Model Intervention for Families Involved in the Child Welfare System

Background:  Successful family reunification is achieved only about fifty percent of the time when children are in foster care. Parents’ ability to access and complete court ordered services are paramount in determining whether the family can achieve reunification. The complex needs of these families require comprehensive interventions. The strategies implemented in the L.A. County Pomona Family First project were: 1) Find and maintain foster and kinship families; 2) Build community partnerships; 3) Provide Team Decision Making (TDM) meetings; and 4) Create self-evaluation tools.  A reduced caseload requirement (15 cases) for case workers was also implemented.

 Methods: A non-equivalent groups design was used.  A total  sample of 157 families was drawn from the child welfare administrative database (50 intervention and 50 comparison)were matched using five criteria: 1) a substantiated child maltreatment allegation, 2) primary caregiver had substance use problems, 3) at least one child was removed from the family home 4) families had not received prior child welfare services, and 4) the same caseworker provided services to both groups. Propensity score matching without replacement was applied to the sample based on the following criteria: 1) the number of children in the family,2) the reason for child removal, 3) the age of the children,4) household composition, 5) housing needs, 6) identified mental health, 7) substance use, and 8) domestic violence problems of the primary caregiver. This matching process resulted in 48 PFFP group families and 48 for comparison using a match tolerance of 0.5. The independent variables were: 1) number of workers assigned to the family, 2) number of worker visits with the family; and 3) community partnerships.  Primary dependent variables were 1) the number of days the case was open, 2) the child placement at case closure and at one-year follow up, 3) the number of days in placement, 4) subsequent substantiated child maltreatment reports at one year.

 Results: Bivariate analysis revealed a significant difference between the intervention families and the comparison families for the days the children were in placement.  Chi square analysis found that at case closure 37% of the intervention families’ children remained out of the home compared to 69% of the comparison families’ children.  At 52 weeks (1 year ) almost 40% of the intervention  families were reunified compared to 16% of the comparison families.   A cox regression model revealed after controlling for the time the case was open families having multiple workers would [(0.283-1.0)x 100%=] likely take  76%  longer to  reunify than families with  one worker. Examining the impact of caseload size we found that families whose workers had a reduced caseload size allowing them to see them more frequently decreased the time to reunification by 22%.  For every day that access to drug testing and outpatient treatment was delayed, there was a 2% increase in the time to reunification  

Implications: The data suggest that a community partnership model that incorporate family engagement, enhanced service provider accessibility, reduced caseloads, and one caseworker for each family can produce stable reunification outcomes for families.