Abstract: Risk Factors for Child Maltreatment Recurrence and Re-Reporting: A Meta-Analysis of Studies (Society for Social Work and Research 25th Annual Conference - Social Work Science for Social Change)

All live presentations are in Eastern time zone.

257P Risk Factors for Child Maltreatment Recurrence and Re-Reporting: A Meta-Analysis of Studies

Tuesday, January 19, 2021
* noted as presenting author
Gerard Chung, MSW, PhD Student, University of North Carolina at Chapel Hill, Chapel Hill, NC
Paul Lanier, PhD, Associate Professor, University of North Carolina at Chapel Hill, Chapel Hill, NC
Daniel Gibbs, MSW, Doctoral student, University of North Carolina at Chapel Hill, Chapel Hill
Rie Masuda, PhD, Doctoral candidate, The University of Tokyo
Ding-Geng Chen, PhD, Professor, University of North Carolina at Chapel Hill, Chapel Hill, NC
Marie Yeo, MSW, Senior Manager, Fei Yue Community Services Singapore
Background and Purpose.
Children who have been maltreated are at increased risk of further maltreatment. Child protection services (CPS) are charged with the responsibility of ensuring that children who have been found to be victims of maltreatment are protected from further harm. One way that CPS strives to achieve this role is by providing effective risk assessments of the likelihood of recurrent maltreatment. Risk assessment tools such as the Structured Decision-Making tool are commonly used. In developing these tools, it is critical for predictive validity that the risk items are drawn from prior research, theory, or both, that are related to recurrent maltreatment. However, in the existing literature, there has been no quantitative synthesis of studies on the risk factors of maltreatment recurrence. With the aim of advancing social work science to improve risk assessment tools, this study conducts a series of meta-analyses of longitudinal studies examining the risk factors of maltreatment recurrence.

Comprehensive searches of published and non-published studies up to February 2020 were done in nine databases including Social Work Abstracts and PsycInfo. Search phrases related to “maltreatment”, “recurrence”, and “children” were used. We included studies of maltreatment of one or more types (abuse, neglect, sexual) and those with samples of children age 18 or younger. In total, 78 studies were identified for the meta-analyses. Since a new episode of maltreatment was defined as the recurrence or re-report of maltreatment at a follow-up time point, odds ratio was used as the effect size appropriate for a dichotomous outcome. Considering that the true value of the estimated effect size for the outcome might vary across studies, a random-effects model was used. Meta-regression was done to examine the impact of moderator variables (e.g., follow-up period, substantiation or non-substantiation outcome, control for services) on study effect size.


We analyzed a total of 122 risk factors and grouped them according to levels of child, caregiver, family, community, and case-related (Cicchetti & Lynch, 1993). At the child-level, children who are younger, have disabilities, or have behavioral/emotional problems have significantly higher odd ratios of experiencing maltreatment recurrence. At the caregiver-level, children with caregivers who abused alcohol/substances, have disabilities, childhood history of maltreatment, emotional problems, or criminal history have higher odds ratio of experiencing recurrent maltreatment. At the family-level, children from families with more children, active domestic violence, poverty, or single-parent family structure are more likely to experience maltreatment recurrence. Finally, cases that have prior reports of CPS involvement, substantiated status of initial report, and poor cooperation between family and CPS are more likely to have maltreatment recurrence.

Conclusions and Implications.
This study highlights the risk factors at different levels of the ecology of child maltreatment that are likely to result in maltreatment recurrence. This provides empirical evidence, currently lacking in social work research, to inform the development of risk assessment tools in CPS. The effect sizes found in this study can also inform the assignment of individual weights to risk factors to improve the predictive validity of risk assessment tools.