Children's Voice Article
Risk Assessment and Decisionmaking in Child Protection
By Sue Steib
A distraught father telephones child protective services (CPS) to report he has learned from a neighbor that his estranged wife has left their three young children at home alone at least twice during the last week.
- Protection of children from abuse and neglect
- Safe maintenance of children in their own homes whenever possible
The agency intake worker says CPS cannot accept his report for investigation because the information is secondhand and he is not reporting the children are home alone at the time of the call. The worker notes in the file that this report may have been prompted by a child custody dispute.
Is this response typical? Might the father have received a different response if he had spoken with another individual or called another agency?
Accepting all reports, however questionable, may seem the better choice, since the alternative might deny children needed protection. But is this realistic when agencies are trying to allocate limited resources to the most serious situations? Is accepting more reports warranted when CPS agencies substantiate less than a third of the reports investigated?
Questions about decisionmaking also arise when reports of child abuse and neglect are investigated. Can a caseworker trust a parent who admits he lost control and was wrong to badly bruise his child during a spanking? Is this situation different from that of a parent who also bruises her child but believes harsh discipline is necessary for the child's own good? What's the likelihood either child will be injured again, perhaps more severely, if left in the care of his or her parents?
Public concern, even outrage, is understandable when CPS agencies fail to intervene and abuse or neglect results in a child's injury or death. But how much do we know about what's involved in determining whether a child is abused or neglected? How straightforward are such decisions?
Increasingly, research is shedding light on ways in which child welfare professionals can improve the accuracy and consistency of their decisions using written tools. Such tools serve two broad functions--to structure risk assessment for child maltreatment, and to guide decisions about intervention.
The development of risk assessment protocols in child protection began in the early 1980s. By 1991, all but a few states were using some type of written instrument to structure CPS decisions. This trend is not unique to child protection--research in other fields has shown that structured protocols can improve clinical decisions.
As reasonable as the use of risk assessment tools might be, developing and implementing them has not been easy. Questions and controversy have arisen about definitions of abuse and neglect and the kinds of measures and statistical methods used. Risk itself is variously defined at different points in the decisionmaking process. As in the case illustration that opens this article, the concept may be applied in decisions about accepting a report or, as in the second example, about the disposition of a confirmed report. Child welfare professionals also may rely on risk assessments to distinguish immediate danger from threats of long-term harm, or to guide decisions about specific service needs.
Tools for Assessing Risk
Risk assessment systems commonly fall into two broad categories: consensus-based and actuarial, with some representing a composite of the two. Strictly defined, consensus-based instruments are derived from opinions of experts in the field or reviews of professional literature; actuarial instruments comprise factors demonstrated through research to predict the recurrence of abuse or neglect, usually in the jurisdiction where the model is used.
The distinctions between these two models are often blurred. Many consensus instruments, for instance, use only items for which empirical support exists. Some researchers believe the two models are becoming more similar as they evolve.
In the field of child welfare, the best-known consensus-based instruments include the Washington Assessment of Risk Matrix, developed in Washington State; the Child Abuse and Neglect Tracking System, developed in Illinois; and the Child at Risk Field System (CARF) from ACTION for Child Protection in North Carolina. CARF, in turn, has been the basis for developing tools that are more specialized to determine risk of immediate harm, such as the Child Endangerment Risk Assessment Protocol used in Illinois.
Many states and counties have adapted these models for their jurisdictions. Other jurisdictions have developed their own tools or use models developed for family assessment that is more comprehensive, such as the Family Assessment Form or the Child Well Being Scales.
The best-known actuarial system is Structured Decision Making (SDM), developed by the Children's Research Center of the National Council on Crime and Delinquency in California. SDM includes a range of instruments designed for intake screening, assessing immediate and longer-term risks of abuse and neglect, and managing cases in which the level of risk may require ongoing services for the child and family.
Developed in collaboration with each of the dozen different jurisdictions using SDM in the United States and Australia, the system looks slightly different from one site to the next. Michigan's SDM system was a finalist for the 2002 Innovations in American Government Award from Harvard University's John F. Kennedy School of Government and the Council for Excellence in Government.
Research has demonstrated all of these systems can improve the consistency and accuracy of child protection decisions. Studies have not supported initial fears that risk assessment tools would be used to justify using less-qualified staff. Rather, the best systems require well-qualified staff capable of using both structured protocols and professional judgment.
How an agency implements a system may be as important as the model itself. Caseworkers often view formalized risk assessment as more useful for agency accountability than for making better decisions for children and families. Some studies have shown that assessment tools are not used in making decisions at all, but only applied after the fact to meet documentation requirements and
justify actions caseworkers have already taken. Other research has found evidence that caseworkers intentionally inflate initial assessments of risk to ensure a family's access to limited services.
The child welfare field has made much progress in structuring decisions about risk, yet work remains. Some researchers suggest that studies consider not just prediction of future maltreatment, but how well measures of risk concur with other established measures of individual and family functioning. Further research should also examine how the interaction of individual factors affects the overall risk picture for a child and family.
Child welfare agencies are moving away from a strict investigative model, which ties services to substantiating a report of abuse or neglect, to one that offers families at lower-risk a "softer" approach, foregoing a formal finding of maltreatment in favor of a comprehensive assessment of family needs and then providing services based on the family's needs. Substantiation, therefore, can no longer be a proxy for maltreatment. Research must identify indicators of actual risk, both to determine which families are appropriate for an assessment approach and to guide decisions about services and disposition.
Implementation is another critical issue that hasn't been fully explored. What studies have been conducted suggest that caseworker skepticism and resistance, inadequate training, or manipulation can set up the best system for failure.
Formal risk assessment is now mandated by federal law. But to fulfill the promise of such systems, we must identify clear paths to ensure that agencies successfully incorporate them into their daily work.
Sue Steib is a Senior Staff Consultant for CWLA's Research to Practice Initiative.
- The U.S. Department of Health and Human Services requires federally funded state child welfare services to undergo Child and Family Services Reviews (CFSRs) to assess state performance in three outcome areas--child safety, child permanence, and child and family well-being. By identifying related practices and programs that research has shown to be successful in these outcome areas, CWLA's Research to Practice Initiative (R2P) seeks to help states explore the evidence base for improving their programs. For more information, visit the R2P website at www.cwla.org/programs/r2p, or e-mail R2P@cwla.org.
Back to Top Printer-friendly Page Contact Us