Comm 176 | deewilsoncon

Reflections on Safety Assessment

(Originally published October 2024)

In last month’s commentary, I discussed the purpose or unintended effect of conflating different meanings of “child safety”. In policy documents or public relations releases, child safety usually refers to the mission of child protection, or as one of the goals of child welfare, along with permanency and well- being. However, in discussions of child protection policy and practice, child safety assessment is often compared to risk assessment, the difference between immediate safety concerns and the likelihood of future maltreatment during the next 6-12 months. The assessment of Safe vs. Unsafe is used to set a threshold for CPS intervention: an assessment of Unsafe child requires a CPS safety plan, while a determination of Safe may lead to a service plan or justify case closure.

Some well-informed persons inside or outside child welfare endorse untenable ideas such as “Children are either safe or unsafe,” and insist on equating ‘unsafe child’ with present danger and an observable safety threat. These statements often include an unspoken, implicit proviso: “For purposes of child protection policy and practice …,” which may describe a CPS practice framework or affirm how a safety framework ought to function. From this perspective, the patently false idea that “children on open CPS cases are either safe or unsafe,” or that children not at risk of imminent harm are therefore safe, is a statement about CPS policy and practice masked as an analysis of child safety.

In past decades, safety frameworks that focused CPS interventions on children in danger requiring an immediate CPS response appealed to managers and program developers in state and county run child welfare systems grappling with long-standing workload problems caused by understaffing, as well as the steady increase of prescriptive requirements in policy manuals. When child welfare offices face overwhelming and chronic workload pressures, it is inevitable that CPS units will narrow their involvement to the most serious cases involving serious injury, sexual abuse and/or risk of imminent harm. This is a process that occurs with or without the permission of managers who learn that when caseworkers have unmanageable workloads, policy requirements become aspirational (at best), and practice standards are sure to be compromised or collapse.

In these circumstances, it is understandable that child welfare managers have often adopted safety frameworks intended to bring consistency to assessment of immediate safety threats, while abandoning risk assessment and a commitment to services for families of children assessed as not in present danger.

In addition, for advocates and child welfare leaders who want to narrow the scope of CPS because they believe CPS involvement often does more harm than good, a practice framework which limits the meaning of child safety to present danger serves their policy agenda. However, there is a large cost to ignoring or minimizing the harms of chronic neglect and chronic multitype maltreatment that does not lead to serious physical injury or acute emotional harm requiring emergency room treatment or hospitalization. Chronically maltreated children are not safe in any reasonable, research-based child protection universe. These children have higher rates of serious illness, early mortality, detention and psychiatric hospitalization across the life span. (Jonson-Reid, et al, 2012).

Setting a threshold of “present danger” to initiate CPS interventions is bad policy when responding to allegations of neglect; early intervention to prevent situational neglect from become chronic neglect is of the utmost importance. The same is true of child protection practice on behalf of infants and toddlers; waiting until children, 0-3 are in danger to take action is a formula for increasing child maltreatment deaths and near deaths. The use of child safety thresholds that discourage early intervention in these cases endangers children’s lives.

The foundations of child safety

One of the most influential bodies of child welfare research ever completed in the U.S. was the psychiatrist, Rene Spitz’s, studies of foundling homes. Foundling “homes” were institutional settings in which infants and some toddlers without mothers for various reasons (e.g., death, illness, incarceration, abandonment) received care from nursing staff working shifts, with frequent examinations by physicians. These young children often received excellent physical care but lacked a mothering figure. Spitz found that children cared for in these settings had highly elevated rates of mortality; in one study 36 of 90 children in one facility died within two years vs. 0 deaths among 60 children in a detention facility that allowed incarcerated mothers to parent their child a few hours per day.

Spitz’s studies may be interpreted narrowly, i.e., as an indication of the harms of institutional care for young children, or more broadly (as I do) to suggest the effects of severe deprivation of nurturance on the immune systems of young children. (“The Science of Neglect”, 2012) Chronically maltreated children have high rates of chronic illness that begin in early life, conditions that may or may not be mitigated by foster care placement. Child safety in early childhood requires both dependable physical care and the consistent attention of a caregiver or caregivers strongly committed to a child’s well-being.

Child safety includes a child’s sense of emotional security that comes from living with caregivers who a) dependably meet a child’s basic physical and emotional needs, b) provide protection from danger, c) claim the child as their own, d) support, recognize and reward pro-social development, e) engage in reasonable discipline and, f) seek out specialized services as needed. Parent advocates and other child welfare critics are right that foster care may do more harm than good when children are mistreated by foster parents or other children in the foster home, moved from home to home on short notice, given cocktails of psychotropic medications, or do not receive timely permanency. Foster care that lasts more than a few days is emotionally hazardous, though it is also true that many children, especially those without behavioral problems when placed, thrive in foster care.

In recent years, it has become increasingly common for advocates and child welfare leaders to ignore, minimize or misrepresent the harms of chronic or severe child maltreatment, while insulting scholars and others who disagree with them regarding the proposed abolition of child welfare. It cannot be said too strongly that the harms of chronic or severe maltreatment have not been overstated. In addition to negative effects on health and mortality across the life span already mentioned, developmental harms to children include:

deficits in executive functioning
difficulty with emotion regulation, i.e., liability to “meltdowns”
cognitive impairments, poor school performance
poor social skills, difficulty in making friends
elevated rates of detention and involvement in multiple systems of care during adolescence

Chronically maltreated children may not be in danger, but they are not safe from serious harm due to child maltreatment. To date, neither in-home service or foster care have been effective in responding to the developmental harms of chronic maltreatment among school age children. Both in-home services and foster placement with relatives or non-relatives often occur after children have been profoundly harmed by months or years of both neglect and abuse. To be more effective, CPS interventions must be both timely and focused on recognizing and mitigating development harm for children in chronically referring families.

A framework for assessing child safety

“Evolving a theoretical model of child safety in maltreating families,” (2006) by Tom Morton and Barry Salovitz, is one of the few articulate, insightful discussions of child safety assessment published in the U.S. during the past two decades. The strengths of this analysis include:

A framework that emphasizes the interactive relationships between and among safety threats, caregivers’ protective capacities and child vulnerabilities, such as young age, disabilities and difficulties with emotion regulation. They assert: “… the most unsafe scenario would be high threats, low protective capacities and high vulnerability.”
Includes caregivers’ histories of maltreatment, substance misuse, violence, mental illness, etc. as potential safety threats. While this may seem obvious, safety frameworks in some state child welfare systems (including Washington) require an “observable” safety threat to support an assessment of Unsafe, a rationale for ignoring or minimizing case history in safety assessments.
Bases reunification decisions on assessment of child safety for the “foreseeable future,” rather than a ‘present danger’ determination.
Most importantly, insists on explaining the underlying dynamics of safety threats, instead of listing, classifying or scoring them. They assert: “Safety models that only classify, without explaining, do not help child protection agencies identify the specific actions that will improve child safety in the short - or the long - term.”

Caseworkers who do not understand the dynamics of child torture are unlikely to recognize the meaning of caregivers’ systematic denial of food and water to a child who has been demeaned, dehumanized, and scapegoated in a family, while being beaten and forced to sleep in a closet, basement, or other tiny enclosure. There have been horrific child deaths in recent years, some in adoptive families, when caseworkers failed to recognize common indicators of torture, in part because these cases are infrequent, though no longer rare.

Morton and Salovitz comment: “Many instances of serious recurrence are preceded by substantially less serious initial maltreatment events.” CPS caseworkers must be able to recognize the meaning of minor “sentinel” injuries to an infant’s or toddler’s head, neck, torso, and recognize when a parent is becoming desensitized to injuries they have inflicted.

For some parents, the inconsolable crying of an infant is agonizingly painful due to their own troubled childhoods, while for others the experience is uncomfortable, but not intolerable. Some parents with histories of substance misuse can follow safe sleep guidelines most of the time, but not when they are extremely depressed or lonely.

As these examples suggest, ‘safety threat’ is a broad somewhat nebulous concept that encompasses much more than standard risk factors such as substance abuse, mental illness, domestic violence and severe poverty. Morton and Salovitz include the following in their list of possible safety threats:

Perceptions
Beliefs
Values
Needs
Motives
Personality
Behavioral tendencies
Physical and mental capabilities
Historical experiences
Neighborhood characteristics
Social or cultural norms

In this framework, just about anything that could reasonably be construed as a possible cause of serious maltreatment may be viewed as a safety threat, but there is no theory of child abuse or neglect that informs the identification of safety threats. These authors provide numerous cogent examples of various types of safety threats from their lengthy child welfare experience, but caseworkers with limited knowledge and experience assessing child maltreatment frequently have difficulty recognizing safety threats and understanding how to apply concepts such as “protective capacities” to actual cases.

Nevertheless, focusing safety assessments on the relationship between parental strengths, functional impairments and child vulnerabilities such as chronic illness, disability and/or cranky temperament is a big step in the right direction. Child welfare training programs should offer caseworkers multiple opportunities based on actual cases to recognize parental strengths and deficits in caring for children with complex health and mental health needs; and how to immediately assist parents and children in these circumstances. Assessment tools can also flag interactions of parent characteristics and child vulnerabilities that often lead to risk of imminent harm.

Risk assessment, safety assessment and child maltreatment deaths

Cases of children whose deaths are due to child maltreatment usually have high risk factors such as substance abuse, mental illness and domestic violence, i.e., they look much the same as other high-risk families. Nevertheless, actuarial risk assessment tools do a poor job of predicting low base rate phenomena such as child fatality for reasons explained by Daniel Kahneman in Thinking Fast and Slow, i.e., actuarial tools are not designed to predict highly unlikely events. Furthermore, most risk assessment tools were developed to predict recurrence of child maltreatment regardless of severity, not the occurrence of serious maltreatment. It took program developers decades to understand that, if their agency’s goal is to reduce child maltreatment deaths on open/recently open cases, risk assessment is an inherently flawed strategy.

Safety assessment frameworks, on the other hand, were developed to anticipate and prevent serious injuries resulting from child maltreatment by recognizing risk of imminent harm. In theory, safety assessment frameworks should improve performance in preventing child maltreatment deaths and near deaths. This has not occurred in Washington State where risk assessment, according to DCYF’s annual self-assessment, has fallen into disuse and where DCYF instead has “bet the bank” on its safety framework. Child maltreatment deaths and near deaths have greatly increased in Washington since 2019, due in large part to a large increase in critical incidents resulting from opioid overdoses and unsafe sleep practices. Nationally, there has been a steady, gradual increase in child maltreatment deaths, as reported in NCANDS data over the past decade.

Safety assessment frameworks have not been more effective – or as effective - in protecting children on open/recently open cases from serious maltreatment than risk assessment models in Washington, despite their narrower focus on present danger. Why not?

1) Assessment tools of whatever type achieve nothing unless their use

leads to effective services or safety plans. In-home safety planning is

arguably the weakest element of child protection programs, both in

Washington and nationally. It is difficult to understand why child welfare systems committed to a safe reduction in foster care have not made major investments in improved safety planning, given the large reduction in foster care expenditures that have occurred in recent years.
Furthermore, child protection programs have struggled to find effective approaches to preventing fatalities from unsafe sleep practices or opioid overdoses when parents have a history of substance misuse, other than foster care, residential programs for mothers and their infant, or live in support from a friend or relative.

2) Safety assessments and safety plans in Washington have been influenced by the social milieu reflected in the Keeping Families Together Act and by the DCYF policy goal of reducing foster care by 50%. In past years, some risk assessment studies found that caseworkers’ case planning goals influenced their risk assessment documentation. Rather than assessments guiding case plans, a caseworker’s intent to avoid foster care placement or close the case may influence her/his risk assessment or safety assessment.

3) Safety frameworks that require caseworkers to explain the dynamics, i.e., the causal connections, of safety threats are difficult for inexperienced caseworkers to learn and apply. In lieu of understanding, use of a safety framework may deteriorate into frequently used jargon that has little or no meaning.

The main deficiency of DCYF’s safety framework is its requirement that an “observable” safety threat must be present to support an assessment of “Unsafe,” a guideline which appears to ignore or minimize case history in determinations of ‘present danger’. It is difficult to understand how a child welfare system could make a mistake of this magnitude, or promptly fail to correct it following reviews of child maltreatment deaths and near deaths often preceded by chronic multitype maltreatment.

References

Jonson-Reid, M., Kohl, P. & Drake, B., “Child and Adult Outcomes of Chronic Child Maltreatment” (2012), Pediatrics, vol. 129, #5.

Kahneman, D., Thinking Fast and Slow (2011), Farrar, Straus & Giroux, New York City.

Morton, T. & Salovitz, B., “Evolving a theoretical model of child safety in maltreating families” (2006), Child Abuse and Neglect, 30.

Spitz, R., First Year of Life: A Psychoanalytic Study of Normal and Deviant Object Relations (1965), International Universities Press, New York City.

“The Science of Neglect: The Persistent Absence of Responsive Care Disrupts the Developing Brain: Working Paper No. 12,” (2012), Center on The Developing Child, Harvard University, Cambridge, Mass.

See past Sounding Board commentaries

©Dee Wilson

deewilson13@aol.com