CFSA’s 2021 Internal Child Fatality Report: How not to learn from the past

CFSA’s newest Child Fatality Review Report focuses on the deaths of 29 children and young adults whose families were known to CFSA within five years of their deaths and whose cases were reviewed in 2021. This report’s toll includes one mother who lost two little children in one year–one who died of fentanyl poisoning and another left alone with a four-year-old sibling and a propped bottle. It included four overdose deaths from synthetic opioids and ten “non-abuse” homicides including the shooting of a six-year-old at 11pm outside a liquor store. It included a twelve-year-old who died of an untreated bacterial infection but had signs of abuse on her body. Most of the dead children’s families had been reported to CFSA at least four times in the past five years. Many of them had experienced investigations and received CFSA-supervised services. Nevertheless, these children died within five years of their contact with CFSA. But the agency’s child fatality reviewers made no recommendations to improve screening, investigation, or services. That is not a surprise, given the agency’s current tendency to minimize intervention in the lives of troubled families.

When a child known to a child welfare agency dies, a natural question is whether the agency could have prevented that death if it had done more or different things. For that reason, fatality review is an important way to assess the performance of a child welfare agency, both internally and externally. According to an appendix to the report, the mission of the CFSA Child Fatality Review (CFR) Unit is “to reduce the number of preventable child fatalities in the District of Columbia through identifying, evaluating, and improving programs and systems responsible for protecting and serving children and their families.” CFSA’s 17th annual Child Fatality Review Report, based on the work of the CFR Unit, was released on January 30, 2023.1

Before 2019, CFSA followed the common practice of including in a given year’s CFR report all of the fatalities it reviewed during that year, even if they occurred during previous years. This makes sense as these deaths were never discussed in earlier reports. But for the third year in a row, CFSA chose to eliminate some cases from its analysis based on when they occurred. The new CFR report includes only those deaths that occurred during 2021 and were reviewed in the same year. That means their analysis includes only 29 child fatalities instead of the total of 51 fatalities they actually reviewed in 2021. The rest of those fatalities occurred in 2018, 2019, and 2020. See the Note on Timing appended to this commentary for further discussion of this issue.

Manner of Death

The manners of death2 of the 29 young people whose cases are included in the body of the report are displayed in the pie chart below. About a third of these decedents were victims of “non-abuse homicide;” six (or 21 percent) died of accidents; four (or 14 percent) died of natural causes; and three (or 10 percent) died of “neglect homicide.” The other six children’s manners of death were “undetermined” or “unknown.”

Source: CFSA, Child Fatality Review Report: Statistics, Observations and Recommendations: 2021,; data plotted by Child Welfare Monitor DC=

Neglect Homicides

Three, or 10 percent of the deaths reviewed in this report, were labeled as “neglect homicides.”(There were no homicides attributed to abuse). All of the victims were age three or under, which is typical of child maltreatment fatalities nationally as well. One of the three was a 17-month-old with “thermal and scald” injuries. The two other fatalities both involved synthetic opioid toxicity, illustrating the spread of this crisis to the District of Columbia. The second fatality was a three-month-old who died of synthetic opioid (eutylone and fentanyl) toxicity and the third was a three-year-old boy dead of fentanyl toxicity. The families of the 17-month-old and the three-year-old both had more than ten hotline calls and had open in-home cases at the time of the fatality.

Non-abuse homicide

By far the most common manner of death reviewed in this report was “non-abuse homicide,” or homicide that was not the result of child abuse. Such “non-abuse homicides” were one-third of all deaths reviewed; eight of the decedents were male and two were female. One victim was only six years old; the remainder were aged 13 and older. Nine of these deaths were caused by gunshot wounds and one was caused by stab wounds. One of the victims was in foster care with a relative at the time of his death. He had been removed from his home in 2015 due to abuse and neglect.

Natural Causes

Three fatalities, or 10 percent of the deaths included in the report, were due to natural causes. All of these children had congenital anomalies. They included a four-day-old girl and a three-year old boy who were both born prematurely and were medically fragile. Both of their families had open cases with CFSA. The three-year-old had been placed with a foster parent experienced in caring for medically fragile children after his mother was determined to have neglected him and was determined to be ill-equipped to care for a medically fragile child.

Accidental Deaths/Unsafe Sleep:

The manner of death was deemed to be accidental for six, or 21 percent of the deaths reviewed. Unsafe sleeping arrangements were involved in three of these six deaths of babies whose ages ranged from 19 days to two months old. In all of these cases, asphyxia was included as a cause of death. Of the remaining accidental deaths, two girls aged 16 and 17 died of opioid overdoses. Both deaths were part of a spate of fatalities in June 2021 that the police attributed to a tainted batch of fentanyl. The final accidental death was that of a four-year-old child who was hit by a car. An observer reported that he wandered away from his mother and two younger siblings before being hit.

Undetermined and unknown

Four of the fatalities were classified as undetermined because the autopsy findings were inconclusive. In one case, the mother left an 11-month-old and her four-year-old sibling sleeping alone in the home, and returned to find the baby unresponsive and foaming at the mouth. In a re-enactment using a doll, the mother demonstrated placing the child on her stomach with a bottle in her mouth in a way that could have impeded her breathing, but the medical examiner was unable to confirm asphyxia as a cause of the baby’s death. In the wake of the fatality, CFSA opened an in-home case to help the mother and her children. But that was not enough to save her three-year-old, who died within six months of opioid toxicity while the in-home case was still open, and was one of the three neglect homicide victims mentioned above. Unsafe sleep environments were involved in two of the other deaths for which the manner was undetermined.

The fourth death for which the manner was undetermined involved a twelve-year-old girl who reportedly collapsed after choking while eating soup. However, the hospital physician observed bruising on the child’s abdomen, back and legs, the mother was found to have abused her, and two siblings were removed from the home. The official cause of death was an untreated bacterial infection coupled with pneumonia but the manner could not be determined.

There were three deaths for which the manners were unknown. A seven-week old boy was reported missing by his father and is presumed dead. The mother was charged with suspected concealment or removal of the body and her other three children were removed from her. Unless there were two babies close to two months old who were reported missing in 2021 under the same circumstances, this is a case that received considerable media attention. The mother told the baby’s father that he had been removed by CFSA. Eventually she told police that she accidentally rolled over her son while under the influence of PCP, then panicked and threw his body into the trash. She was initially charged with murder but the charges were dropped as a body was never discovered. According to police, the mother was stabbed to death by the father in April 2022.

The remaining deaths for which the manner is unknown involved a newborn removed by Caesarian from a mother dying of Covid-19 and a seven-year-old who died in a house fire, for whom autopsy results are pending.

Decedents in Foster Care at TIme of Death

Two of the decedents were in foster care at the time they died. One was the eighteen-year-old who was living with a relative after being removed from his abusive mother, and who was the victim of a non-abuse gun violence homicide. The other victim was the medically fragile two-year-old, who was dependent on a gastrostomy tube and a tracheostomy vent, who died of natural causes and was living with a foster parent who specialized in caring for medically fragile children.

Family Risk Factors

The report provides some demographic information about the parents of the children who died, and that information is in line with research evidence that teen parenting, large families, and a parent’s history of maltreatment as a child are risk factors for child maltreatment. Parents of the children who died tended to be very young when they started having children. Sixty-nine percent of the mothers and 58 percent of the fathers were under age 21 at the birth of their first child. The youngest mother was 13 years old when she gave birth to her first child and the two youngest fathers were 16. Many of the dead children came from large families, which is more common among those who start having children at an early age. All but four of the 29 decedents had two or more siblings. Thirteen had four or more siblings; three had seven or more siblings, including two with 12 siblings and one with 10 siblings. A parent’s maltreatment history as a child is also known to be a risk factor; 13 of the 29 birth mothers had CFSA involvement as children, and both parents of two of the decedents were involved with CFSA involvement in their childhood.

Parents’ CFSA History as Caregivers

Nine of the 29 families reviewed in the report (or about a third) were involved with CFSA at the time of the fatality. Of these nine families, two had an open investigation and five had an open in-home case. Two had an open foster care case, but those are presumably the families of the two children who died while in foster care.

Source: CFSA, Child Fatality Review Report: Statistics, Observations and Recommendations: 2021,

Table 7 shows that 10 families (or more than a third of the 29 families) had CFSA involvement within 12 months of the fatality – only one more family than was involved at the time of the fatality. It appears that all of them had an investigation within that period. However, with fewer families reported to have an in-home or foster care case than in Table 6, there must be some errors in the data; CFSA has not yet responded to a request for clarification.3

Source: CFSA, Child Fatality Review Report: Statistics, Observations and Recommendations: 2021,

Looking at the 17 families that did not have CFSA involvement at the time of the fatality (Figure K in the report)4 CFSA found that the time since they were involved varied from 1 to 56 months, and that 11 of these 17 families (65 percent) were last involved with CFSA more than 18 months before the fatality.

As shown in Figure J, the majority of the families had four or more reports to the hotline (known as referrals) during the five years preceding the fatality. Many of these reports were screened out, as shown in Figure Q. All but 4 of the 29 families had referrals that were screened out during the five-year-review period and ten of those families had five or more referrals screened out.

Source: CFSA, Child Fatality Review Report: Statistics, Observations and Recommendations: 2021,
Source: CFSA, Child Fatality Review Report: Statistics, Observations and Recommendations: 2021,

Table 8 summarizes the results of the reports received regarding these families. The large majority of these families (23 families or 79 percent of families) were investigated at least once , with an additional 6 families investigated twice. Twelve families (41 percent of families) had in-home cases within five years of the fatalities, with eight having one case, three having two cases, and one having three cases.5,6

Source: CFSA, Child Fatality Review Report: Statistics, Observations and Recommendations: 2021,

Figure L in the report (not reproduced here) shows that 16 families, or slightly more than half of the families, had at least one substantiated allegation in the five years before the fatality. Ten of these families had only one substantiation and the remaining families had between one and four substantiated allegations. The most common substantiated allegations were inadequate supervision and educational neglect (four families each), followed by physical abuse, failure to protect, medical neglect and indadequate food/clothing/hygiene (three families each).

4+ Staffings

One CFSA practice that is designed to prevent further harm to children known to the system is the “4+ Staffing.” CFSA conducts these meetings for families that have four or more referrals, with the last referral occurring within the past 12 months. These staffings are supposed to uncover gaps in past practice or service delivery that may have contributed to the repeated maltreatment and to find strategies to prevent future maltreatment. In the 2020 CFR report, CFSA stated that of the decedents’ families who had qualified for such a staffing during the five years before the fatality, all had received a 4+ Staffing. Clearly, this figure casts doubt on the effectiveness of these staffings. Unfortunately, the analogous section of the current report appears to focus on the percentage of eligible families that received a 4+ staffing after the fatality, a fact that is less relevant to the purpose of CFR.7

Siblings removed after child fatalities

A new report section states that 11 children were removed from four families in the wake of the child fatalities reviewed in the report.

  • After an 11-month old girl was left alone with a four-year-old sibling and a propped bottle, the agency opened an in-home case for the family. But when her three-year old brother died of fentanyl intoxication within six months of his sister’s death, the other children were removed and placed with kin. At the time of the writing of the report, their goal was reunification with the parent, who was said to be “receiving services, including grief counseling for the loss of more than one child within a short time frame, as well as substance use and housing.”
  • The siblings of the 12-year-old girl who died of an untreated bacterial infection but also had signs of physical abuse have already been returned home. CFSA states that the mother and children have received mental health services and the mother has completed court-ordered parenting education.
  • Of the three siblings of the infant boy who disappeared, one child was with her father at the time of the report’s being written and the other two were placed in foster care with a goal of reunification.7
  • The 17-month-old who was scalded to death had four siblings, ranging in age from five months to twelve years old, who were all removed from the mother in the aftermath of the toddler’s death. Two were placed with kin and two in a mysteriously named “non-foster care placement,” which probably connotes placement with kin outside the foster care system. There is no mention of reunification; perhaps this case qualified for the exemption from reunification that is allowed under certain aggravated circumstances.8

CFSA’s Findings and Recommendations

In its Summary of Critical Findings, the CFR report discusses specific areas that received “additional focus” in 2021, including screened-out referrals. Concerns about the accuracy of hotline decision-making have been expressed in the District and around the country. In a 2016 study, conducted by the Center for the Study of Social Policy, the court monitor in the LaShawn class action suit, reviewers disagreed with CFSA’s decision to screen out the referral in 27 percent of the referrals studied. However, the report does not suggest any review of policy or practice in screening out referrals. Instead, it cites the “prevention services” provided by the collaboratives and the family success centers. The report also devoted special focus to families involved with CFSA at the time of the fatality, unsafe sleep fatalities, and gun violence, but the report makes no suggestions about how to avoid such fatalities involving these factors.

Based on its fatality reviews, CFSA’s ICFR Committee approved three recommendations: revision of the critical event and child fatality review policies, integration of child fatality review data into the new computerized case management database currently being developed, and finalizing a Memorandum of Understanding with DC Health “to provide monthly data on applicable fatalities to CFSA to facilitate the timely review of child fatalities.” It is notable that all these recommendations address the child fatality review process itself. There are no recommendations for changes in policy or practice related to screening, investigations, or services. It is also significant that the following language about the purpose of ICFR recommendations which was included in the 2020 report does not appear in this one: “The CFSA ICFR committee makes recommendations concerning appropriate actions that may possibly avert future fatalities.” Perhaps this language was deleted because none of this year’s recommendations are aimed at averting future fatalities, just about amending the fatality review process.


It is important to remember why we study child fatalities. These tragic deaths are the tip of an iceberg – the visible consequence of recurring abuse and neglect after at least one incident of alleged maltreatment was reported to the agency. When a child remains in the home after services end, we may not know that maltreatment has continued unless the child dies. Some children known to CFSA die for causes that are not related to abuse and neglect, like the newborn delivered early from the mother dying of COVID-19, or the three children with congenital abnormalities who died of natural causes. But research shows that simply having a report of child maltreatment increases the risk of deaths from all causes, not just abuse or neglect. Thus, many of the fatalities included in the report may be a consequence of ongoing maltreatment, even if the manner of death was not found to be maltreatment.

Some of the deaths with a manner labeled as accidental or unknown may have been due to neglect, like some of the eight babies who died while sleeping in unsafe arrangements. In total, CFSA reported that unsafe sleep factors were present in eight of the 10 deaths to children aged two or under. In my years of service on the citywide Child Fatality Review Committee,1 I have seen numerous cases of children dying in unsafe sleep environments in families with histories of child welfare involvement. Almost invariably, the parents have used marijuana, alcohol or other intoxicants before lying down with the baby, and they failed to wake up as their children struggled to breathe. With unimpaired parents, these sleeping arrangements might not result in death. This is why a study found that adjusting for risk factors at birth (including low birth weight and late or absent prenatal care), the rate of Sudden Unexplained Infant Death (SUID) was more than three times greater among infants who had been previously reported for past maltreatment than among infants who had not been reported.

And then there are the ten deaths from “non-abuse homicide.” The connection between child maltreatment and violent death became obvious to me soon after I started sitting on the citywide Child Fatality Review Committee. I learned that many young victims of homicide grew up in families with long histories of reports to CFSA regarding lack of supervision, school absences, physical abuse and other concerns. In some cases CFSA screened out the reports or found no maltreatment; in other cases services were provided. But the maltreatment continued. Many of these families exhibited chronic child neglect, which occurs “when a caregiver repeatedly fails to meet a child’s basic physical, developmental, and/or emotional needs over time.” Many of these children, with histories of trauma and little support at home or connection to school, eventually found belonging in the streets and took up violent and illegal activities.

Of course, we do not know how many of the gun violence victims included in this report came from abusive or neglectful homes or were involved in violence themselves. But according to police reports, motives for five of the nine homicides included retaliation for robbery and gang-related activities, implying the victims were involved with such activities. Relatives of the 18-year-old female victim of another homicide indicated that she was involved in a “volatile” relationship with her killer. The six-year-old whose death is reviewed here was obviously an innocent bystander. But she was shot while walking to a liquor store with her parents late on a Friday night. And federal prosecutors stated that the father of the six-year-old shot on July 16, 2021 was involved in the violent drug trade on the street where his daughter was shot, and the judge indicated that he did not understand what the child was doing in that area at 11:00 PM.

The four-year-old who was hit by a car when he wandered away from his mother might also be a victim of neglect, but we do not know because no case details are provided in this report. And as for the teenage girls dead from tainted fentanyl, we will never know what kind of home environment they had and if that contributed to their drug use. But the connection between childhood maltreatment and later substance abuse is well-known.

Of course CFSA understands the linkage between abuse and neglect and all causes of death. That’s why it studies all deaths of children known in the past five years, not just those due to maltreatment. CFSA had many opportunities to intervene in the lives of the 29 children discussed in the report and their families. Nine of the families were involved with CFSA at the time of the fatality. Ten of the families were involved with CFSA within a year of the fatality. Within five years of the fatality, 25 of the families had at least one screened-out report, 23 of the families had at least one investigation, and 12 of the families had at least one in-home case.

Obviously it is concerning that two fatalities occurred in families that were being investigated by CFSA and five occurred in families that had an open in-home case, In which a CFSA worker is expected to visit a family from weekly or twice a month, depending on its perceived need. One has to wonder whether any red flags were missed by the workers who were investigating or monitoring these families. Among the families that had an open in-home case at the time of the fatality were the families of both the 17-month-old and the three-year-old who died of synthetic opioid toxicity. One cannot help wondering how frequently and thoroughly the in-home workers interacted with the families, without noticing that the parents were still using opioids. Nobody wants to remove more children, but perhaps they needed to be removed, and some of these children may have had a protective relative ready and eager to receive them.

I do not mean to say that CFSA could have saved all of the children that they touched who later died. But perhaps it could have prevented some of these deaths. The agency could have chosen to devote special attention to parents who were very young when they started childbearing, families with many children, and parents who were involved with CFSA as children. It could have screened in more referrals, substantiated more allegations, opened more cases, monitored families more effectively, provided more intensive and effective services, or involved the court in more in-home cases. And if necessary it could have removed more children, preferably to a protective relative or other known adult.

The total of 29 children who were known by the end of 2021 to have died after being touched in some way by CFSA will certainly rise as more fatalities are identified and reviewed. (The total for 2020 is 40 so far.) But if present practices continue, the remaining fatalities will not be included in an annual report, except for one or two tables in the appendix. Therefore, the annual reports understate the number of children who died after being touched by CFSA. Strangely, the report authors do not seem to understand the significance of the smaller universe, making comments like “There was a decrease in total infant fatalities in 2021.” That decrease, from 16 to 10, may not mean much when the total number of fatalities reviewed was 29 instead of 40.

It is concerning that all of the report’s recommendations concern the practice of fatality review itself. There are no recommendations to improve CFSA’s practice in conducting the hotline, investigations and in-home and foster care cases. It is hard to avoid thinking that CFSA’s ideology of non-intervention and family preservation has contributed to both these deaths and the lack of recommendations aimed at preventing such deaths in the future. No child welfare agency has a crystal ball. They all have to strike a balance between the harms caused by intervening in families where there is no maltreatment and not intervening in maltreating families. But CFSA has declared its preference clearly. As stated in the report, “For over a decade, CFSA has invested in safely keeping families together and developing robust prevention strategies to help support them.” But these “robust prevention strategies” were not enough to protect the children who died in 2021. For CFSA, these deaths may simply be collateral damage.

The information about the siblings who were removed from their families after the fatalities – a new addition to the report – is quite disturbing. The idea that the agency is still working for reunification with the mother who lost two children in one year – one left alone with a four-year-old and a propped bottle and the other poisoned by fentanyl – and had ten hotline reports in the five preceding years is concerning. And that the siblings of the abused child who died of a bacterial infection are back with their mother after she received mental health services and parenting education is concerning as well. But it is not clear what CFSA could have done differently without action from the DC Council. The DC Code requires CFSA to make “reasonable efforts” to reunify children with their families except in certain aggravating circumstances which probably would not have been found in these cases.9

As always, CFSA’s internal child fatality report is distressing. It includes two deaths in one year to the same family, ten non-abuse homicides, several deaths involving unsafe sleep, and two high-profile child deaths: an infant whose body was disposed of by the mother and a six-year-old who was killed as part of a drug war involving her father. We learned of four deaths – two of small children and two of teenagers – involving the new scourge of fentanyl and other synthetic opioids. CFSA has made the decision to avoid intervention and prioritize family preservation above child safety, and the report contains no recommendations for improving the agency’s efforts to spot and address abuse and neglect. If CFSA is not going to make any recommendations to protect children in the CFR report, then one might question the report’s purpose and utility.


  1. It is important to distinguish CFSA’s internal child fatality reports from the annual reports of the citywide Child Fatality Review Team, which covers all deaths of young people up to age 18 and some deaths of those aged 19-21.
  2. ‘Manner of death” refers to the circumstances that caused the death, as opposed to “cause of death,” which refers to the specific disease or injury that led to the death.
  3. It is also unclear why only one child is counted as having a foster care case since two of the children were in care at time of death.
  4. This count of 17 families that did not have CFSA involvement at the time of the fatality is inconsistent with Table 6 and the associated text, which says that 20 families did not have active CFSA involvement at the time of the fatality. 
  5. The heading “Reports” in this table is confusing but I have followed the lead of CFSA in describing the Table’s data and am awaiting clarification from the agency.
  6. Family assessments were an alternative to an investigation for low-risk cases, and are no longer being used by CFSA.
  7. I have asked CFSA to clarify the meaning of their data on families that received a 4+ staffing but have not received an answer as of the date of this publication.
  8. If this is indeed the case referred to earlier, this reunification will not take place as the mother has been killed by the baby’s father. It is possible that the oldest daughter’s father is a different person and that she can remain with him.
  9. See DC Code § 4–1301.09a. Reasonable efforts,,Reasonable%20efforts.,the%20family%20by%20the%20Agency.

A Note about Timing

Until 2019, CFSA’s internal child fatality reports covered the agency’s reviews of all fatalities of children whose families were known to the agency within five years of their death. This is normal for child fatality review panels, which often have a long time lag before reviewing a case. But in 2019, ICFR decided to include in its review only the deaths that actually occurred in the year they were reviewed. As I have described, one problem with that approach is that it is clearly impossible to review all deaths that occurred in a given calendar year during that same year. Deaths that occur or become known near the end of the year clearly cannot be reviewed during that same year. Perhaps as a result of that realization, the CFR Unit included in the 2020 child fatality report those child deaths that occurred in 2020 and were reviewed during the same year or in the first three months of CY 2021. That decision gave them a total of 40 cases that were included in the report. However, this year, CFSA, as in 2019, including only the cases reviewed during the calendar year, missing any cases that were reviewed in the first quarter of 2022 or later. This is particularly strange because this report was released so much later than the last report with respect to the calendar year reported on – January 2023 as opposed to October 2021. So they had more time, not less, to include an extra quarter of reviews.

There is another problem with limiting reviews to the current calendar year, which is that deaths occurring in earlier years are missed. In Appendix A and B, the ICFR reports that it actually reviewed 51 deaths in 2021. Thirteen of these deaths occurred in CY 2020 and were apparently reviewed in the first quarter of 2021 and included in the 2020 report. However another 14 fatalities that occurred during 2018 and 2019 were presented to the ICFR committee during 2021 but not included in the report. These fatalities were never included in previous analyses, nor will they be included in the future, so most of the information on these fatalities will never be released to the public, aside from some demographic and cause and manner of death data provided in Appendix C.

There is also an internal inconsistency between the 2021 report and the appendices. The report contains reviews of 29 cases. Appendix A states that “the ICFR Committee reviewed 51 fatalities during CY 2021; all 51 fatalities helped to inform practice and policy recommendations that potentially reduce future child fatalities.” It states that 13 of these fatalities occurred in 2020 and were included in the previous report, and another 14 of these fatalities occurred in 2018 and 2019. That means that out of the 51 fatalities reviewed in 2021, only 24 (51 minus 27) occurred in 2021. Yet, the 2021 analysis includes 29 fatalities. I have asked CFSA about the discrepancy but have not received an answer as of the date of publishing this commentary.