Chaiyachati, B.H., Gaither, J.R., Hughes, M., Foley-Schain, K., & Leventhal, J.M. (2018). Preventing child maltreatment: Examination of an established statewide home visiting program. Child Abuse & Neglect, 79, 476-484.
Over two million families across the U.S. receive federally funded or state and locally funded home-visiting services, however, the effectiveness of these services on preventing child maltreatment is still unclear. This study examined the effects of home visiting on preventing child maltreatment in the state of Connecticut by linking data from the Nurturing Families Network (NFN) home-vising program to that from the state’s Child Protection Services (CPS) agency. Using propensity score matching to reduce selection bias, the authors compared investigated reports of maltreatment, substantiated reports of maltreatment, and out-of-home placements in socially high-risk families who received home-visiting services to a matched sample of families who were eligible for home-visiting but chose not to participate.
The final sample included 7,386 families: 2,662 were in the home-visiting group and 4,724 in the comparison group. After matching, there were 2,280 families in each group. Bivariate associations both before and after matching were assessed using χ2 tests. To account for the timing at which substantiations and out-of-home placements occurred, the authors used Cox proportional hazards regression to generate hazard ratios. They also assessed time to first substantiation and out-of-home placement using Kaplan-Meier curves.
After matching, there were no statistically significant differences between groups in rates of CPS investigation (21.1% vs. 20.9%) and out-of-home placement (2.7% vs. 3.6%). There was a significantly lower occurrence of any substantiated child maltreatment in the home-visiting group than in the comparison group (7.8% vs. 9.9%). Home visiting was also associated with lower rates of substantiated neglect (7.5% vs. 9.7%). Results showed that substantiated reports of maltreatment among families receiving home-visiting services occurred later in the child’s life than families in the comparison group. The authors concluded that participation in a statewide home-visiting program did demonstrate a significant decrease in substantiations of child maltreatment for families, and this article underscores the potential benefit of these services.
The researchers used administrative data to conduct a deterministic data linkage between families involved or eligible for the NFN program to families involved with CPS. The Revised Early Identification (REID) screening instrument assessed the risk of maltreatment and identified socially high-risk families. The REID is comprised of 17 factors (e.g., maternal age, single parent status, unstable housing) and has been used to identify participants for home visiting programs in previous studies. Elements from the REID instrument and other covariates (including CPS follow-up time) were used for the propensity score models.
The study had the following limitations: (1) it is a retrospective observational study of a well-established statewide home-visiting program and not a randomized controlled trial, however, the use of propensity score matching minimized bias; (2) 26% (2,633) of families were excluded from the full sample of eligible families (largely due to missing dates of birth), which may have contributed to a bias in the results; (3) administrative data has a number of inherent limitations, including the availability of additional covariates, errors, and missing data; and (4) it is unknown whether disclosing information about participation in home-visiting services may bias worker decisions regarding substantiation and out-of-home placement.