top of page

How to Reduce All-Cause Child Mortality in Child Welfare  

(Originally published July 2021)

Last month’s Sounding Board summarized several studies of all-cause child mortality and of causes of child death other than child maltreatment for children with child welfare involvement in the U.S. and Australia during the past several decades. These studies found highly elevated rates of all- cause mortality, including deaths from medical causes, SIDS, accidents, and suicide, as well as from intentional and unintentional injuries, among children and older youth with CPS reports after controlling for poverty and other co-variates present at birth.


This commentary concurs with the argument of Richard Barth and Melissa Jonson-Reid in an article published in 2000 that


          “… there is a scientific – not just a journalistic basis – to suggest that the prevention of future death is a substantial need among children who have contact with child welfare services, yet there has been little systematic effort to assess the role of child welfare services in the prevention of later deaths among abused and neglected children.”


I also agree with these authors’ view of child safety:


          “We define child safety as including avoidance of an early preventable death … and from serious injury associated with ongoing maltreatment or other preventable harm.” 


This conceptualization of child safety is in stark contrast to the child safety framework utilized in Washington’s and Oregon’s child welfare systems, a safety framework focused on danger to children and risk of imminent harm.


This commentary seeks to answer the question: “What child welfare reforms are needed to reduce preventable child deaths from all causes, not just child maltreatment?” I also propose changes in the structure of child welfare and in the concept of child protection that will better protect the lives of the most vulnerable children in our society.  Barth and Jonson-Reid made a cogent research-based argument two decades ago for a child welfare mission that includes reduction of preventable child deaths and for performance indicators that measure agency progress toward that goal. Their argument was largely ignored for one main reason: child welfare policymakers and managers were resistant to being held accountable for a mission they could never achieve with available resources and through the administrative structure of child welfare. I was a child welfare manager in Washington State in 2000, and (like my peers) was strongly resistant to any expansion of child welfare mission that did not take account of resource limitations and inadequate staffing levels. Given chronic workload problems and limited funding for family support services, child welfare agencies have narrowed the scope of their responsibility for child protection, adopted “a current danger” approach to child safety, insisted on a clear distinction between child safety and child well-being and managed to performance indicators that leave them (and everyone else) uncertain as to what extent   child welfare involvement increases child safety.  


During the past year, an influential social justice movement has gained wide currency among child welfare advocates that proposes (in its most extreme form) the elimination of the current child welfare system. This movement is  more concerned with the harm to children and parents from child welfare interventions than with child maltreatment, especially child neglect which is sometimes referred to “just poverty.” In the view of many child advocates, a lower rate of child welfare involvement for children and families is better than a higher rate from intake through termination of parental rights, except for reunification rates. The idea that child welfare systems might be needed to reduce all-cause child mortality and the mortality of young adults, as well as near deaths and serious injuries among these groups, is not considered in the child welfare reform agendas I’ve read in recent months.      


Arguments against reorienting child welfare to reduce preventable child deaths 


I occasionally encounter the view that child death, especially child maltreatment death, is a low base rate phenomenon (about 2.5 per 100,000 in NCANDS statistics), and therefore too rare for prevention of child death to be adopted as key indicator of child safety. However, Barth and Jonson-Reid commented in 2000 that “For every child abuse death there are an estimated 9 seriously disabled and 70 seriously injured children (U.S. Advisory Board on Child Abuse and Neglect, 1995).” Furthermore, child maltreatment deaths are a small fraction of preventable child deaths from medical causes, SIDS, accidents, suicide, and substance abuse during adolescence. In a study pending publication in Pediatrics, the infant death rate from medical causes in California was found to be 13 per 10,000 with almost double the mortality odds for children with a single CPS report and 3 times more likely for children with more than one CPS report.  A recently published study of all-cause mortality for a birth cohort in South Australia found a mortality rate to age 16 of 2.7 per 1000 for the entire birth cohort, and much higher odds of mortality for children with CPS reports than for other children, after controlling for covariates.


An elevated rate of child mortality is only the initial stage of life-long sequelae for children with multiple adverse childhood experiences, including physical abuse, sexual abuse and emotional abuse, parental substance abuse or mental illness, recurrent domestic violence, or separation from a parent in childhood for whatever reason. Segal, et al,  comment that “the seminal ACE (ACEs) Study found adults self-reporting 6>ACEs … died an average of 20 years earlier than those reporting no ACEs.”  The same policies and practices required to reduce preventable child deaths have the potential to reduce early mortality of children with child welfare involvement across the life span.


Infancy and child death


Any potentially effective strategy for reducing preventable child deaths must focus on infants, 0-1, who make up the largest percentage of deaths (by far) from child maltreatment, medical causes, SIDS and “roll-over” suffocation deaths. Deaths of infants from medical causes far outnumber child maltreatment deaths. A large percentage of infant deaths occur during the first month of life, i.e., neonatal period. For example, in a study of infant deaths from medical causes in California pending publication in Pediatrics, neonatal deaths from 2010-16 outnumbered deaths in the post-neonatal period (29 days to first birthday) by more than 9-1.


In the 1990s, Washington State study of child deaths among children, 0-4, (discussed in last month’s Sounding Board) “More than one-third of deaths during the first 28 days of life – neonatal deaths – are caused by birth defects, complications of prematurity and pregnancy and infections.” For all age groups of young children, i.e., neonatal, post-neonatal and children, 1-4)  “biological risk factors were the strongest predictors of subsequent death … even after the neonatal period the risk of death continued to be related to the child’s physical condition at birth.”


It is the physical vulnerability of deceased young children and of children with CPS reports from birth that stands out in the Washington State study, including low birth weight, disabilities and chronic illnesses. A recently published study of all-cause mortality in South Australia among children with CPS reports replicated these findings. The two main risk factors for death to age 16 in South Australia for children after the neonatal period were “in hospital at 28 days” and birth abnormalities.           


In Washington State in the early 1990s, most CPS reports on young children were received in the post-neonatal period after half of children in the birth cohort had already died. Child welfare law and practice are not designed to save the lives of infants whose mothers do not receive prenatal care and/or engage in prenatal substance abuse, or quickly identify babies who are at greatly increased risk of death during infancy due to the combination of (a) serious physical health conditions and (b) parents with functional impairments due to substance abuse and/or mental health conditions. Many high-risk infants die unnecessarily in the first 28 days of life without any contact or assistance from a public agency and without notice or public outcry.


Lack of prenatal care, smoking and prenatal substance abuse are child protection issues, but this does not mean that the public policy response should be to define these behaviors as child maltreatment and initiate investigations of mothers reported to CPS. This is the kind of ham-fisted approach to child protection which has led to a social justice movement that seeks to dismantle child welfare systems. A better idea is to fund and empower public health agencies to reach out to pregnant women receiving publicly funded substance abuse or mental health services, and/or homeless persons with offers of assistance as early as possible prior to a child’s birth. The reduction of infant mortality was a triumph of public health during the 20th century, and one main reason the average life span in developed countries doubled from about 40 in 1920 to almost 80 in 2020, pre-pandemic (Johnson, 2021). It is time to utilize the expertise of public health systems once again to reduce preventable infant deaths.


If mothers with substance abuse issues and/or chronic mental health conditions refuse help in the prenatal period, low birth weight babies and babies with birth abnormalities should be flagged for services on the day of the child’s birth and their mother again offered services, including respite care on demand, by a public health nurse. If this offer of services is again refused, then a CPS report should be made and screened in as “risk only.” Public child welfare agencies should be funded to hire public health nurses and infant mental health specialists to work with these families; or be authorized to contract with local public health departments to take over case management responsibility.  


The public policy guideline should be that recognition of exceptional child vulnerability at any point during pregnancy; or following birth due to physical health problems should lead to early intervention services, rather than waiting for a CPS report. Furthermore, any child welfare system that waits for a high-needs infant to be clearly in danger, i.e., deemed unsafe, before intervening is risking preventable child deaths, however those deaths might be classified by fatality review teams. 


Reducing child maltreatment deaths  


Child maltreatment (CM) deaths have slowly and steadily increased in recent years, despite thousands of fatality reviews and hundreds of reports and studies of CM deaths. High profile CM deaths have repeatedly led to changes in leadership of state child welfare systems, and to the development of comprehensive child welfare reform agendas designed to prevent future CM deaths. In addition, a distinguished blue-ribbon commission released a report in 2016 with many recommendations for reducing CM deaths after more than a year of public meetings around the country. Why have these many persistent efforts failed to reduce CM deaths?


One possible reason that states and large counties have largely failed to reduce CM deaths is that reform agendas have continued to repeat the same mistakes while expecting a different result, e.g., by inundating public agencies with dozens, sometimes hundreds, of recommendations that could never be implemented in good faith. This is a management failure resulting from political pressure to demonstrate a high level of initiative and proactive efforts in pursuit of policy goals.


However, the main reason for ineffective reforms is the inability of public agencies to assimilate the “hidden in plain sight” lesson from studies of child maltreatment deaths: it is the combination of children with much higher than average vulnerabilities resulting from physical health conditions (including neuro-disabilities and difficult temperaments) being cared for by parents with below average capacity for the care of children due to substance abuse and/or mental health conditions, and untreated trauma from childhood or interpersonal violence in intimate relationships that endangers children’s lives.  The main dynamic that leads to CM deaths is the same dynamic that leads to deaths of young children from causes other than child maltreatment. Until public agencies zero in on the mismatch between young children with a high level of special needs due to physical health conditions and disabilities and parents with limited capacity to care for a child due to functional impairments, there will not be a reduction in CM deaths or in all-cause mortality among young children reported to CPS.


The inadequacy of CPS practice with mothers of infants is described in Putnam-Hornstein’s, et al, 2016 study of infants rereported to CPS after an initial CPS investigation in California. Almost one in five (18%) of infants reported to CPS were removed from the parent’s home at the time of the first report; 82% of infants remained in the parent’s home. Only one in ten of the parents whose infants remained in the home received services. 60% of these infants were rereported to CPS within 5 years, and the highest rate of rereport was for infants with physical health conditions and/or whose mothers had received no prenatal care. CPS agencies in California showed little or no interest in providing services to mothers whose infants were not placed in foster care, regardless of their physical health conditions or disabilities. The findings of this study reflect clueless CPS practice.


One of the most glaring deficiencies of CPS programs is the lack of investment over decades in in-home safety planning. There is scarce research on which to base guidelines for in-home safety plans, a lack of evidence which leaves practitioners dependent on flawed practice wisdom, or their own instincts. In 2015, I met with CPS supervisors in a Southern state, one of whom expressed the view that in-home safety plans “are not worth the price of the paper they’re written on.” A middle manager in the state stated: “we develop these plans, put them in case records and then we pray.”  Even in states where safety plans are utilized in a conscientious fashion by CPS caseworkers, there is no reason for confidence in the effectiveness of these plans in protecting infants, given inadequate research and the lack of safety-oriented services (such as childcare and respite care) to support families. In these circumstances, depending on in-home safety plans to protect infants assessed by CPS caseworkers to be “unsafe,” (not most infants) is not responsible practice.


Until public agencies - in cooperation with scholars and other professionals-  develop a knowledge base to guide development and implementation of in-home safety plans, infants assessed to be in danger must be placed with relatives; or in foster care; or in residential substance abuse treatment facilities that take both mothers and babies; or in day treatment programs such as Oregon’s Relief Nurseries; or in comprehensive programs such as the Sche’lang’en Village developed by the Lummi Nation in Washington State.  


In addition, child advocates, practitioners and judicial officers should reflect on the finding of the study (mentioned above) of infant deaths from medical causes in California, i.e., that foster care placement reduced infant deaths from medical causes by half. The Washington State study of all-cause mortality in early childhood during the early 1990s found that foster care had a large effect on reduction of child deaths among children with CPS reports after the neonatal period. This does not mean that greatly increasing the number of young children in foster care is a good idea. However, it does indicate the importance of foster care in reducing all-cause mortality among young children with CPS reports, while (at the same time) states make large new investments in family support services for troubled parents whose children have physical health conditions and/or disabilities.


Next month’s Sounding Board will discuss strategies for reducing all-cause mortality among children 1-5 and 6-17.




Barth, R. & Jonson-Reid, M., “ Outcomes After Child Welfare Services: Implications for the Design of Performance Measures,” (2000), Children and Youth Services Review, Vol. 22, Nos. 9-10.  


Cawthon, L. & Hopps, D., “Mortality of CPS Clients in Washington State from Birth to Age Four,” (1997)  Research and Data Analysis, Department of Social and Health Services, Olympia, Washington.  


Johnson, S., Extra Life: A Short History of Living Longer (2021), Riverhead Books, New York, NY.


Putnam-Hornstein, E., Simon, J. ,Eastman, A. & Macgruder, J., ”Risk of Re-reporting Among Infants Who Remain at Home Following Alleged Maltreatment,” Child Maltreatment, published  online Nov. 21, 2014.    


Schneiderman, J., Prindle, J. &, Putnam-Hornstein, E., “Infant Deaths from Medical Causes Following a Maltreatment Report,” (pending publication), Pediatrics.


Segal, L., Doidge, J., Armfield, J., Gnanamanickam, E., Preen, D., Brown, D. & Nguyen, H., “Association of Child Maltreatment with Risk of Death During Childhood in South Australia,” (2021), JAMA Network Open, 4 (6).   


Segal, L. Armfield, J., Gnanamanickam, E., et al, “Child maltreatment and mortality in young adults,” (2021), Pediatrics, 147(1).  ©   


​©Dee Wilson    


Index of all Sounding boards


bottom of page