All-Cause Child Mortality Among Children
with Child Welfare Involvement
(Originally published June 2021)
In 2000, Richard Barth and Melissa Jonson-Reid published an article, “Outcomes After Child Welfare Services: Implications for the Design of Performance Measures,” in which they stated:
“The absence of knowledge about the safety of children in the child welfare
system after the close of children’s services creates a serious risk of
overemphasizing current indicators of child welfare service performance
(e.g., rate of reunification or length of stay in foster care). …if we allow …
outcome indicators to be selected so that they only reflect on the timing and
resolution of child welfare cases, rather than their real-world outcomes, we
leave child welfare vulnerable to continue to be funded at a level … (which)
dilutes the benefit that can be produced.” And “agreement on safety’s
superordinate status does not assure that we agree on what safety is or
how long it should be preserved. … we will argue that a focus on children’s
long-term physical well-being (safety) is a critical measure of the level of
need in the child welfare population …we define safety as including avoidance
of an early preventable death … and from serious injury associated with
ongoing maltreatment and other preventable harm.”
These comments, as pertinent today as they were twenty years ago, sound discordant during a period when child advocates are far more concerned with the harms done to families and children by child welfare interventions than by deaths or serious harm due to child maltreatment, much less by elevated mortality rates among children with child welfare involvement from a wide range of causes including birth abnormalities, disabilities, disease, SIDS, accidents, suicide or homicide during adolescence. In addition, many child welfare systems in the U.S., including Washington State and Oregon, utilize child safety frameworks that define child safety narrowly, i.e., as risk of imminent harm or danger. I have yet to come across a recent child welfare reform agenda that seeks to reduce preventable child deaths from multiple causes among children with CPS reports, or even acknowledges that children with open child welfare cases have much higher rates of all-cause mortality than children with no child welfare involvement.
Studies in the U.S. and Australia
There have been a surprising number of excellent studies of all-cause child mortality and of mortality due to causes other than child maltreatment during the past 30 years:
An early study of children in Washington State whose names were placed on a central registry ( i.e., found to be abused or neglected) found that these children were almost 3 times more likely to die from all causes prior to age 18 than children with no CPS reports. They were 18 times more likely to die of homicide before age 18, even though only one in five deaths were due to homicide. The extent to which other types of child maltreatment (such as neglect) caused or contributed to children’s deaths was uncertain. (Sabotta and Davis, 1992)
An unpublished study of almost 275,000 children born in in Washington State in the early 1990s found that Medicaid eligible children reported to CPS had at least double the rate of child mortality after the first month of life from 1 month to 4 years of age compared to Medicaid eligible children not reported to CPS. More than 90% of child deaths to age 4 occurred during the first year of life. Less than 10% (8.6%) of deaths of children with CPS reports were due to physical abuse. The main causes of early death were birth defects, complications of prematurity and pregnancy, infections, SIDS, and unintentional injuries. Children with CPS reports had much higher rates of diagnosed medical conditions and disabilities than children with no CPS involvement. The authors comment that “higher mortality rates among children experiencing CPS involvement may, in part, have been due to a higher frequency of physical ailments.” Low birth weight was strongly associated with post-neonatal (2-12 months of age) and early childhood deaths (1-4 years of age).
After the neonatal period (28 days following birth) children placed in foster care had a mortality rate per 1000 considerably lower (4.0) than the rate (7.0 ) for children reported to CPS but not placed in foster care. SIDS deaths were highly elevated among children with CPS reports; and were strongly associated with maternal substance abuse. “After controlling for maternal substance abuse, children with CPS involvement did not have an increased risk for SIDS,” the report states. (Cawthon & Hopps, 1997)
In a study pending publication in Pediatrics, birth records for more than 3,464,000 million children born in California between 2010 and 2016 were linked with deaths of infants to age 1. After controlling for covariates, including Medicaid eligibility, maternal age, birth weight and race/ethnicity, infants with a history of reported maltreatment died from medical causes 1.77- 3.27 times more frequently (depending on the number of CPS reports) than infants not reported to CPS. The authors state that “Among infants reported for maltreatment …When in foster care, the rate of death was roughly 50% lower.” And “as more reports were received, the associated risk (of death from medical causes) increased,” a finding that “held regardless of whether we included or excluded deaths during the neonatal period.” The authors assert that “Differences in death rates signal potential unmet service needs among infants who remain at home following reports.” (Schneiderman, et al, 2021)
An analysis of all-cause mortality for children, ages 1-18, in US foster care systems from 2003-16 found a 42% higher rate of child mortality for foster children, 1-14, than for children in the general population. The mortality rate for foster youth, 15-18, was close to the rate for this age group in the general population. One of the authors commented in an interview that “Clinicians should recognize that children in foster care are a particularly vulnerable population, though importantly, we are not concluding children in foster care have increased mortality because they are in foster care.” Foster children in all racial/ ethnic groups had an increased mortality rate, though the mortality rate for black children was more than 20% higher than the rate for white children. (Chaiyachati, B., et al, 2020)
A study of all children born in California between 1999-2006 “found that after adjusting for health and sociodemographic risk factors – including maternal age, infant birth weight, and prenatal care – infants previously reported as alleged victims of abuse and neglect had a rate of SUID (Sudden Unexpected Infant Death) that was more than 3 times higher than infants not reported.” Lack of prenatal care and low birth weight were strongly associated with SUID. (Putnam- Hornstein, et al, 2014)
A longitudinal study of all children born in California in 1999-2000 found that “most children who died of suicide in California had interacted with CPS at some point during childhood.” Children and youth reported to CPS but without a substantiated investigation or foster placement “have three times the risk of suicide compared with matched children who were never reported for maltreatment …” Children and youth with substantiated CPS reports and children with foster care placements “had 5 times the odds of suicide compared with children with no CPS involvement.” (Palmer, et al, 2021)
A study of 4.3 million children born in California between 1999 and 2006 found that “after adjusting for risk factors at birth, children with a prior allegation of maltreatment died from intentional injuries at a rate … 5.9 times greater than unreported children; and died from unintentional injuries at twice the rate of unreported children” during the first 5 years of life. This study found that a CPS report in and of itself was the strongest independent risk factor for death from injury in early childhood. (Putman-Hornstein, 2011).
Lessons from child mortality studies
Child deaths due to physical abuse are a small fraction of preventable child deaths. More children die from neglect related accidents than from physical abuse, though the extent to which unintended injuries are due to neglect in large mortality studies is uncertain. Mortality rates of children are much higher in early childhood, especially infancy, than during the school age years. Most deaths of young children with CPS reports are due to medical conditions, including disease, disability and SUID. A 2017 needs assessment of children in Oregon’s foster care system found the following:
Drug affected infant ........................9.2%
Low birth wight/premature birth...5.7%
Heart disease/heart problems........2.9%
Broken bones/inflicted injuries.......1.9%
Failure to thrive.................................1.8%
Fetal alcohol syndrome....................1.6%
Brain abnormalities......................... 0.2%
Traumatic brain injury.......................0.4%
Anal or genital warts/herpes............0.4%
Shaken baby syndrome.....................0.2%
Approximately one-third of foster children had externalizing behavior problems such as adjustment disorder, conduct disorder and/or oppositional defiant disorder; and more than 40% of foster children had internalizing disorders such as depression, anxiety, or PTSD. The needs assessment classified 42% of foster children and youth as “high need.” It’s likely that many of these children had a combination of serious physical ailments and behavior problems which present a formidable challenge to parents, foster parents, and unlicensed kinship care providers. (Bellatty, et al, 2019)
In their study of mortality in early childhood in Washington State, Cawthon and Hopps found that “low birth weight was a strong predictor of post-neonatal and early childhood death. Extremely low birth weight children were nearly five times more likely to have died as normal birth weight children.” Low birth weight was associated with lack of (or late) prenatal care. “Across the three Medicaid study groups with CPS involvement, at least half of all children who died were diagnosed with a medical condition or birth defect, compared to one third of children in the No CPS Medicaid group.” Cawthon and Hopps assert that “Prevention of neonatal death is directed primarily at improving medical treatments during pregnancy and the neonatal period,” i.e., lack of prenatal care may endanger the lives of children with low birth weight or birth abnormalities. The mothers of 22.9% - 28.5% of children with CPS reports had “intermediate or inadequate prenatal care.”
Cawthon and Hopps found that maternal substance abuse was a major risk factor for child mortality by age 4. Children of mothers with diagnosed substance abuse were 3.59 times more likely to die in early childhood than children born to mothers without diagnosed substance abuse. Maternal substance abuse accounted for highly elevated rates of SIDS deaths among children with CPS reports.
After controlling for risk factors at birth, African American children had lower than average mortality in early childhood, and Native American children had an average mortality rate in the Washington State study.
Studies of all-cause mortality for children with child welfare involvement and the California study of infant deaths from medical causes summarized above suggest one main “hidden in plain sight” lesson: most preventable child deaths in early childhood are due to the combination of children with serious medical conditions and/or disabilities being cared for by caregivers with extreme functional impairments due to substance abuse, mental illness and/or untreated trauma resulting from domestic violence and/or childhood histories of abuse and neglect. It is the combination of medically fragile/ disabled children with caregivers who have limited capacity to care for their special needs absent long term intensive family support services – not one or the other – that leads to highly elevated rates of preventable deaths among children with prior CPS reports.
Foster care in early childhood in some U.S. states appears to have reduced child mortality to age 4 by 50% or more. In a recently published study in JAMA, children born in South Australia between 1986-2003 who had one or more CPS reports had highly elevated all-cause mortality from 1 month to 16 years of age, much like their counterparts in the U.S. Children placed in foster care had a higher mortality rate ratio (MMR – 3.79) than the MMR for children with CPS reports only (2.69); and with investigations not substantiated (3.16), and with substantiated maltreatment (2.93). (Segal, et al, 2021) In a separate study of all-cause mortality of young persons, 16-33, in South Australia by this same group of researchers published recently in Pediatrics, young persons placed in foster care at age 3 or older had a mortality rate more than double the rate of children placed in foster care before age 3 (4.69 vs. 1.75), a finding which suggests the possibility that foster care has different effects on risk of mortality among both children and young adults depending on the age at entry-into-care.
Implications for child welfare policy and practice
The lack of concern, or even interest, among policymakers and child advocates in reducing preventable child deaths from causes other than child maltreatment is partly the result of persistent attempts in the U.S. over several decades to organize child welfare completely around reports of alleged child abuse and neglect, both to control costs and in response to overwhelming workloads. Some child welfare leaders and child advocates do not understand how unusual this is, both in comparison with other developed countries and historically in the U.S. The conflation of child protection with CPS investigation/ assessment of alleged abuse and neglect is not defensible for anyone who (like Barth and Jonson-Reid) believes child welfare systems should be accountable for reducing preventable deaths of vulnerable children, given that most of these deaths are not due to child maltreatment, at least not directly. Mandatory reporting of suspected child abuse and neglect and investigations of allegations of maltreatment contained in CPS reports has had widespread support across the political spectrum in this country until recently. It’s uncertain whether a child welfare system which views child protection more broadly (i.e., concern with reducing preventable deaths from all causes in vulnerable child populations) would have the same degree of support. Science cannot have the final say in determining child welfare mission, a public policy decision that depends on social values embodied in law.
I’m sure there are policymakers, child advocates and child welfare leaders who view elevated child mortality rates among children with CPS reports as inevitable, given the level of social inequities in this country. However, there is nothing inevitable about denial and passivity in response to information regarding preventable child deaths. This is a choice, one that must be rejected.
For child advocates and child welfare staff who agree that reduction of preventable deaths from all causes should be part of the child welfare mission, there is an urgent need to reflect on the large changes in law and child welfare policy and practice necessary to achieve this goal. These changes include an enhanced child protection role for public health and public mental health agencies, changes in intake screening criteria; and in safety frameworks utilized by CPS investigators; and in the types of resources available to children and families with open child welfare cases. Foster parents with the skills required to care for medically fragile children and for children with behavior problems must be recruited and given adequate training, supports and compensation, as well as professional status. These foster parents should be given a new role, i.e., the support and coaching of birth parents with ‘high needs’ children.
In next month’s Sounding Board, I will discuss key components of a more effective response in several public systems to elevated rates of child mortality from multiple causes among children with CPS reports.
Barth, R. & Jonson-Reid, M., “Outcomes After Child Welfare Services: Implications for the Design of Performance Measures,” Children and Youth Services Review, Vol.22, Nos. 9-10 , 2000.
Belatti, P.& Gibson, W. “Identifying Capacity Needs for Children within the Oregon Child Welfare System. Summary Document, Office of Reporting, Research, Analytics and Implementation, Oregon Department of Human Services, 2019.
Chaiyachati ,B., Wood, J., Mitra, N., “All-Cause Mortality Among Children in the US Foster Care System,” JAMA Pediatrics, 174 (9), January 2020.
Cawthon, L. & Hopps, D., “Mortality of CPS Clients in Washington State from Birth to Age Four,” unpublished; Research and Data Analysis, Department of Social and Health Services, Olympia, Washington, 1997; available on request by sending me an email.
Palmer, L., Prindle, J.& Putnam-Hornstein, E., “A Population-Based Examination of Suicide and Child Protection System Involvement,” Journal of Adolescent Health, pending publication, 2021.
Putnam-Hornstein, E., Schneiderman, J., Cleves, M., Magruder, J. & Krous, H., “A Prospective Study of Sudden Unexpected Infant Death after Reported Maltreatment,“
Journal of Pediatrics, 164, 2014.
Putnam-Hornstein, E., “Report of maltreatment as a risk factor for injury death: a prospective birth cohort study,” Child Maltreatment, 16 (3), 2011.
Segal, L, Armfield, J, Gnanamanickam, E., Preen, D., Brown, D. & Nguyen, H., “Child Maltreatment and Mortality in Young Adults,” Pediatrics, 147 (1), January 2021.
Segal, L., Doidge, J., Armfield, J., Gnanamanickam, G., Preen ,D., Brown, D. & Nguyen, H., “Association of Child Maltreatment with Risk of Death During Childhood in South Australia,” JAMA Network / Open, 4(6), 2021.
Schneiderman, J., Prindle, J. & Putnam- Hornstein, E., “Infant Deaths from Medical Causes Following a Maltreatment Report,” Pediatrics, pending publication, 2021.