Girvin, H., DePanfilis, D., & Daining, C. (2007). Predicting program completion among families enrolled in a child neglect preventive intervention. Research on Social Work Practice, 17(6), 674-685.
Understanding which clients are more likely to complete services and under what circumstances is an important issue in child welfare. Using a sample of 154 caregivers who were deemed at risk of child neglect, this study developed a model to predict which clients completed the Family Connections (FC), a social work child neglect preventive program. The effectiveness of this program (both the 3 month and 9 month version) in preventing child neglect was previously evaluated (DePanfilis & Dubowitz, 2005). While there was no control group in the original evaluation, making findings of effectiveness highly tentative, both the 3 and 9 month groups showed decreases in parenting stress, caregiver depressive symptoms, and everyday stress. As well, both groups had increased positive parenting attitudes, parental sense of competence, and social support. In the current study, families were randomly assigned to receive either a 3 month or 9 month intervention. Results indicate that those receiving the 3 month version of FC were more likely to complete service than those receiving the 9 month version. Within groups, completers reported a stronger working alliance with their workers than non-completers. The final statistical model found three significant predictors of service completion: depressive symptoms (caregivers with more depressive symptoms were more likely to complete); interpersonal relationship with the worker (a stronger relationship predicted completion); and treatment group status (families assigned to the 3 month program were more likely to complete services than those assigned to 9 months). Findings from this study contradict the findings of previous studies that “difficult” clients (e.g., those who are depressed and/or using drugs) are less likely to complete services. Additionally, the authors stress that given the lack of significant differences in outcomes between the 3 and 9 month groups, consideration should be given to the notion that shorter treatment periods may be more appropriate and realistic for this population. Finally, this study serves as a reminder that the strength of the helping relationship is a significant predictor of many positive outcomes including treatment completion.
Of the original 154 caregivers, 136 completed a research interview, and were included in the study. Caregivers accepted to the program were those where concern regarding neglect existed, but at a lower level than would warrant a CPS investigation, and where there were at least two additional risk factors related to the child(ren) (e.g. child behaviour problems, child physical/developmental/learning disability). Caregivers were not currently involved with CPS at the time of service onset, and agreed to participate in the program. CPS status prior to the program was defined as the existence of an indicated or substantiated report of child abuse/neglect with the local child welfare agency. History of drug use was operationalized as the caregiver staying high for a whole day or more in the past 12 months, and was collected through self-report. Depressive symptoms were measured using the Center for Epidemiological Studies-Depressed Mood Scale (Radloff, 1977). Everyday stressors were measured by the Everyday Stress Index (Hall, Williams & Greenberg, 1985). The worker-client relationship was measured using the Helping Relationship Inventory-Client (Poulin & Young, 1997). Caregivers’ satisfaction with their social workers was measured using the Parent Outcome Interview (Magura & Moses, 1986). The dependent variable, service completion was based on a coding at the end of treatment: a) completed treatment; b) directly refused further service; c) whereabouts known but unavailable; and d) whereabouts unknown.