Are In-Home and Foster Care Services Associated With Risk of Maltreatment-Related Fatality and Near Fatality Following a Child Protection Case? Evidence From Pennsylvania

Abstract

Many children who experience child maltreatment fatalities (CMF) or near fatalities (CMNF) are previously known to the child welfare system (CWS). Although in-home services and foster care (FC) are tools for preventing CMF/CMNF, relatively little research has examined how each is associated with subsequent CMF/CMNF. Using a nested case-control design and linked administrative data from Pennsylvania, this study examined how in-home services and FC following an initial CWS case were associated with the odds of future CMF and CMNFs (n = 171 cases and 777 controls). Relative to receiving no services, in-home services were not associated with reduced odds of CMF/CMNF (OR = 1.27, p = 0.201), whereas FC was associated with lower odds (OR = 0.54, p = 0.061), though this association did not reach statistical significance at p < .05. Relative to in-home services, FC was significantly associated with 58% lower odds of CMF/CMNF (OR = 0.42, p = 0.010). Although replication with larger samples is needed, findings suggest that FC, but not in-home services, is associated with reductions in CMF/CMNFs in the years following CWS involvement.

Keywords

fatalities foster care in-home program

The federal government reports that approximately 2000 children die each year from abuse or neglect, at a rate that has steadily increased for the past five years of data (U.S. Department of Health & Human Services, 2024). Experts have long argued that the true count of these fatalities is far higher (Commission to Eliminate Child Abuse and Neglect Fatalities, 2016). An additional unknown number of children experience a “near fatality” – a term that lacks consistent definition (Campbell et al., 2021) but generally refers to a child whose abuse or neglect would have resulted in death if not for immediate medical attention. A sizable share of child maltreatment-related fatalities (CMFs) and near fatalities (CMNFs) involve children who were already known to the child welfare system (CWS) (U.S. Department of Health & Human Services, 2021). In these cases, it is often questioned why the CWS agency did not provide in-home services or place the child in foster care (FC) (Newman, 2024). It is arguably intuitive that placing a child who has been maltreated into FC should reduce their risk of experiencing subsequent CMF or CMNF. Yet, this has not been widely studied and many assert that FC does not prevent maltreatment of children and is among the least safe places for children (Raz & Sankaran, 2019; Roberts, 2023; Wexler, 2018).

Only one known study examined the association of both in-home services and foster care placement on CMF (Douglas, 2016). Using data from the National Child Abuse and Neglect Data System (NCANDS) and applying rare event logistic regression, Douglas (2016) found that receipt of in-home services (e.g., family support, case management, family preservation) was associated with reduced risk of CMF. When analyses were stratified by maltreatment type, both in-home services and foster care placement significantly lowered the odds of CMF for cases involving physical abuse, whereas only in-home services were protective in neglect cases. These findings suggest that service provision, whether delivered in-home or through placement, may reduce fatality risk. However, the underlying data were limited by inconsistent measurement standards across states for several key variables (Jones et al., 2024). Thus, evidence regarding these associations remains inconclusive.

Studies have linked FC placement with a reduction in the likelihood, by almost half, of all-cause mortality for substance-exposed infants (Lawler et al., 2025) and post-neonatal infant mortality due to medical causes (Schneiderman et al., 2021). However, CMFs are a small share of overall infant mortality, and factors that prevent infant mortality may not prevent CMF (or CMNF). For example, reductions in infant mortality could be explained by foster parents’ higher economic resources, education or training, or other supports, rather than by decreased exposure to abusive or negligent behavior. Moreover, because many children who enter FC return to their families of origin within 12-18 months, it is possible that FC has no long-term association with CMF/CMNF risk.

Research on non-fatality CWS outcomes broadly suggests that FC prevents the recurrence of maltreatment in the near-term (Gross & Baron, 2022), but perhaps not long-term (Palmer et al., 2023; Jones & Jonson-Reid, 2023). In contrast, research generally finds that in-homes services do not reduce non-fatality CWS outcomes, such as rereport, with studies finding null (Connell et al., 2007) and positive (Chiang et al., 2025; Fuller & Nieto, 2014), associations. We observed a cohort of children with CWS cases (including alternative response) in Pennsylvania between 2015 and 2019, followed through September 2021. Applying a nested case-control design, we addressed the following questions: How are in-home services and foster care placement following an initial CWS case associated with the odds of 1) CMF or CMNF, 2) CMF specifically, and 3) CMNF specifically?

Methods

Data and Sample

We used linked CWS and Medicaid claims data from Pennsylvania on a cohort of children with CWS cases in 2015-2019. CWS cases were substantiated child protective services reports and validated general protective services (Pennsylvania’s alternative response program) reports. Children were followed from their first report in years 2015-2019 through September 2021. CWS records were probabilistically linked with Medicaid records for both the child and the biological mother, which allowed for the inclusion of additional risk factors not available in the CWS records. Sample inclusion criteria were: (1) child’s first CWS report was not a CMF/CMNF, (2) child was born after 2007 and enrolled in Medicaid at birth, and (3) child’s biological mother was identified and enrolled in Medicaid at the first CWS report. These criteria yielded 85,488 children. We identified 191 children with a subsequent CMF/CMNF. The case-control design (described below) did not yield any matches for 20 children, leaving a final sample of 171 CMF/CMNFs (59 CMFS and 112 CMNFs), which serve as the cases for the analysis.

Identification of Controls

A nested case-control design was implemented where each CMF/CMNF was matched with up to five controls from the population to maximize statistical power (Iwagami & Shinozaki, 2022). Matching incorporated ten characteristics that previous research has linked to both initial CWS contact and either CMF or CMNF: (1) child age at the 1-month level (U.S. Department of Health & Human Services, 2026), (2) child gender (male or female) (U.S. Department of Health & Human Services, 2026), (3) the child was identified as non-Hispanic Black (U.S. Department of Health & Human Services, 2026; Wilson et al., 2025), (4) child diagnosis of low-birthweight or premature birth (Putnam-Hornstein, 2011; Putnam-Hornstein et al., 2013; Spencer et al., 2006), (6) identification of physical abuse on the index report (Kennedy et al., 2020; Palusci & Covington, 2014; Putnam-Hornstein et al., 2013; Wilson et al., 2025), (7) mother diagnosis of substance use disorder as of one year prior to index report (Dean et al., 2024; Goldstein & Font, 2025; Kotch et al., 1999), (8) mother diagnosis of mental health disorder also as of one year prior to index report (Dean et al., 2024; Hammond et al., 2017; Kotch et al., 1999), (9) the biological mother was aged 19 or younger at birth of the child (Kotch et al., 1999; Segal et al., 2021), (10) an unrelated adult male partner of the biological mother was identified on the index report (Stiffman et al., 2002; U.S. Department of Health & Human Services, 2024), and (11) an indicator if the referral took place in the years 2018 or 2019.

These characteristics were selected as matching variables as prior literature has observed their associations with both CWS contact and CMF/CMNF (see citations following each characteristic). Because the sample is restricted to children born on Medicaid and mothers enrolled in Medicaid at the index report, cases and controls are implicitly matched on low-income household status (Farrell et al., 2017; Pelton, 2015). The grouped year matching variable (11) was included as there was a substantial increase in the frequency of CMF/CMNFs between 2017 and 2018. In selecting matching variables, we prioritized maintaining the largest possible sample, particularly to ensure that the number of cases who received foster care services was large enough to report without violating confidentiality (cell sizes of 11 or greater). Consequently, we do not match on county in our primary specification even though Pennsylvania’s CWS is county-administered; a county-matched alternative specification is provided in Appendix Table A3.

The final sample consisted of 777 controls. The design was “nested” in that each case was matched with controls who were in the risk set at the time the CMF/CMNF occurred. To qualify as a control, an individual had to have the same combination of the 10 matching variables described above and have been observed over the same time interval as the matched CMF/CMNF. The Stata command sttocc was used to create the nested case-control dataset. Of the 171 cases, 141 (82%) were matched with five controls, five (3%) with four controls, nine (5%) with three controls, nine (5%) with two controls, and 7 (4%) with one control.

Measures

The dependent variable was indication of a CMF or CMNF on a subsequent report. A CMF was identified if the child died due to the maltreatment identified on the report. A CMNF was identified if the child would have died if not for immediate medical care. The primary independent variable was CWS response, which was time-varying within the first 90 days following the index report. The measure had three categories: none, in-home, and foster care. In-home services included both prevention and family preservation services. During the first 90 days, a child’s CWS response status could change from none to in-home, none to foster care, or in-home to foster care. For example, if a child was not placed in foster care, and began receiving in-home services 10 days after the index report, they would have two records. The first would be censored at day nine. The second would start on day 10 and continue for the duration of the child’s observation window. If the child was placed in foster care, then their status did not change, regardless of whether they returned home within the 90-day period. For cases, the last record was included in the regression. Given the nested case control design, the record chosen for controls was the record where the start date was before the case’s last record’s start date and the censor date was after the date of the CMF/CMNF. In other words, the control’s record had to span the observation period of the case’s final record.

We also included a time-varying indicator denoting the onset of the COVID-19 time period. New records were created starting on March 30^th, 2020 for children still being observed. Prior literature has linked the COVID-19 time period to increased odds of child maltreatment-related medical encounters (Rebbe et al., 2023) but has yet to examine how it associates with CMF/CMNFs.

Analysis

Odds ratios (ORs) for all independent variables were estimated using conditional logistic regression, which stratifies the sets of cases/controls. Using unconditional logistic regression yields biased estimates of ORs in a case-control sampling design. Models were fit for each outcome: CMF or CMNF; CMF only; and CMNF only.

Results

Sample Description

Table 1 presents the descriptive statistics for the nested case-control sample broken out by case (CMF/CMNF) and control (neither outcome) status for the full sample. Descriptives for the population from which the controls are sampled are also given. Due to confidentiality concerns for cell sizes 10 or less, we cannot provide descriptives broken out by CMF and CMNF. Compared to the population, cases/controls were significantly younger. Of the cases, 90% were aged two or younger at the index report. Compared to the population, cases were more likely to be male, and had similar rates of identifying as non-Hispanic Black. Rates of diagnoses of low-birthweight or premature birth were significantly higher in cases/controls than the population. Cases/controls had significantly lower rates of having a teen mother at birth and slightly lower rates of having an unrelated adult male partner of the biological mother present on the index report. We note that examining the bivariate distributions alone between the cases and the population does not suffice to identify a variable as a risk factor for CMF/CMNF, particularly without conditioning on age. Compared to controls, cases had significantly higher rates of receiving in-home services in the first 90 days post-report (44.4 vs 35.6%) and significantly lower rates of foster care (7.0 vs 13.8%).

Table 1.

Descriptives for the “Population” and Nested Case-control Sample. Unless Otherwise Noted, Percents are Given in Parentheses

	Population	Nested case-control sample, N = 948
	Population	Cases	Controls
N	85488	171	777
Variable used for matching
Child-level
Age in years (M, SD)	3.5 (3.0)	0.8 (1.5)^a	0.8 (1.5)
Gender
Female	41066 (48.0)	72 (42.1)	325 (41.8)
Male	44422 (52.0)	99 (57.9)	452 (58.2)
Non-hispanic black	20617 (24.1)	42 (24.6)	187 (24.1)
Low birthweight/Premature birth	7949 (9.3)	29 (17.0)^a	123 (15.8)
Physical abuse	7569 (8.9)	^b	^b
Biological mother-level
Substance use disorder	45099 (52.8)	99 (57.9)	475 (61.1)
Mental health disorder	46756 (54.7)	99 (57.9)	453 (58.3)
Age <20 at Child’s birth	13893 (16.3)	16 (9.4)^a	45 (5.8)
Male paramour/Stepfather in household	15143 (17.7)	24 (14.0)	103 (13.3)
Covariates for regression
CWS Response
None	45885 (53.7)	83 (48.5)	393 (50.6)
In-home	32252 (37.7)	76 (44.4)^b	277 (35.6)
Foster care	7351 (8.6)	12 (7.0)^b	107 (13.8)

Note. CMF = child maltreatment fatality, CMNF = child maltreatment near fatality.

^aIndicates measures where the cases are significantly different (p < 0.05) than the population.

^bIndicates measures where the cases are significantly different (p < 0.05) than the controls.

^cCell sizes n < 11 are omitted due to confidentiality concerns.

Regression Results

Table 2 presents the ORs for the three conditional logistic regression models for the full sample, and the CMF and CMNF specific subsamples. We caution that, given the relative rarity of CMF/CMNF generally and the small share of cases/controls that experienced an intervention (FC or in-home services), even large coefficients are unlikely to be statistically significant at p < .05. Exact p-values are provided in Table 2. Receiving in-home services within 90 days of report was associated with elevated odds of CMF/CMNF relative to no services (OR = 1.27, p = 0.201), but was not significant. Compared to no services, being placed in foster care within 90 days of the index report was associated with 46% lower odds of CMF/CMNF (OR = 0.54, p = 0.061) and was also not significant. Compared with in-home services, foster care placement was significantly associated with 58% lower odds of CMF/CMNF (OR = 0.42, p = 0.010). The COVID-19 time period was associated with 46% increased odds of CMF/CMNF (OR = 1.46, p = 0.156), but not significantly. When county is added as a matching variable, the significance of the ORs are similar, though their magnitudes increase (see Table A3 in Appendix).

Table 2.

Odds Ratios for Conditional Logistic Regression Models. 95% Confidence Intervals Given in Brackets

N	CMF/CMNF 171 cases + 777 controls	CMF 59 cases + 263 controls	CMNF 112 cases + 514 controls
CWS response
In-home (Ref = None)	1.27 [0.88-1.84], p = 0.201	1.22 [0.63-2.35], p = 0.558	1.31 [0.84-2.05], p = 0.233
Foster care (Ref = None)	0.54 [0.28-1.03], p = 0.061	0.36 [0.12-1.10], p = 0.073	0.65 [0.29-1.46], p = 0.297
Foster care (Ref = In-Home)	0.42* [0.22-0.82], p = 0.010	0.30* [0.10-0.92], p = 0.036	0.5 [0.22-1.12], p = 0.092
COVID-19	1.46 [0.87-2.47], p = 0.156	2.67* [1.15-6.23], p = 0.023	1.01 [0.51-1.99], p = 0.978

Note. CMF = child maltreatment fatality, CMNF = child maltreatment near fatality.

*p < 0.05.

Subsample analyses separating CMF and CMNF cases revealed consistent trends with regard to both in-home services and foster care. Although, we caution against overinterpretation given event rarity, we note that the point estimates for both FC and in-home services are very similar in the CMF and CMNF subsamples. Of note, however, the elevated odds of CMF/CMNF during the COVID-19 time period for the full sample appears to be specific to CMFs The COVID-19 time period was associated with 2.67 times odds of CMFs (OR = 2.67, p = 0.023), but was not associated with CMNFs (OR = 1.01, p = 0.978)

Discussion

This study examined how the use of in-home services and foster care placement were associated with the likelihood of later CMFs and CMNFs among children with confirmed child welfare cases in Pennsylvania. We found that initiating in-home services within 90 days of the index report did not reduce the odds of CMF/CMNF; rather, the coefficients indicated elevated risk relative to no formal response. Foster care placement was associated with approximately half the odds of CMF/CMNF compared to both no services and in-home services—indicating a potential benefit for the most vulnerable children. Although our estimates are somewhat imprecise due to the rarity of CMF/CMNF -- especially within foster care—they are similar to estimates reported in other studies of child mortality with larger samples (e.g., Lawler et al., 2025; Schneiderman et al., 2021). However, we offer several cautions and considerations related to these findings.

First, in-home services are disproportionately targeted to families with substantial risk factors as an alternative to foster care. Although the case-control design allowed for efficient analysis of rare outcomes, it is possible that unmeasured factors, such as those influencing decisions about provision of in-home services or foster placements, could contribute to potential selection bias. In addition, in-home services vary in quality, intensity, and duration and thus may be too inconsistent in their implementation to produce a meaningful reduction in CMF/CMNF. Our administrative records did not capture the quality, intensity, or fidelity of in-home or placement services delivered, nor access to informal supports or other environmental factors that may influence outcomes. Further research is needed to better understand the circumstances in which in-home services are successful in reducing risk for CMF/CMNF, which may reflect specific features of the child and family or attributes of the services and case management process. It is possible that the presence of services gives a false sense of safety or that services are offered without preventing ongoing exposure to abuse or neglect. It is also important to note that CMF and CMNF events are rare, even among children known to CWS. For example, of the 32,252 children who received in-home services in the population, only 76 (0.25%) experienced a known CMF/CMNF (see Table 1). Yet, CMFs and especially CMNFs are likely undercounted overall. Both CMFs and CMNFs may be undercounted due to inadequacies in the reporting or investigative process; CMNFs are also subject to imprecisions in definitions. Potential CMF/CMNFs with active CWS involvement may also receive greater scrutiny that cases without active involvement, such that CMF/CMNFs are more likely to be identified during periods of in-home services or foster care. Future research is needed to determine whether findings hold when using alternative measures of CMF/CMNF.

Moreover, although foster care placement reduced the risk of CMF/CMNF for children receiving an intervention, there were nevertheless some CMF/CMNFs among children who experienced foster care placement following the index report. Given that states have an explicit responsibility for the safety of children in their custody, and that foster parents, which include both unrelated and kinship caregivers, are screened, trained, approved, and monitored by the child welfare agency, any CMF/CMNF in foster care is highly concerning. The rarity of these events creates greater barriers to identifying patterns and opportunities to improve the process through which non-relative and kinship foster parents are approved and monitored. Yet, such actions are essential not only for the protection of children but also to ameliorate the lack of trust many have in the system (Volk et al., 2025).

Lastly, the finding that the COVID-19 time period is associated with increased odds of CMFs but not CMNFs was unexpected. Prior research found that maltreated children were less likely to be reported during the COVID-19 time period (Marmor et al., 2023), but that child maltreatment-related medical encounters increased (Rebbe et al., 2023). However, it remains unclear why these processes may differentially impact CMFs and CMNFs. One possibility would be related to decreased utilization or access to emergency medical services (EMS). For example, EMS incidents decreased during the COVID-19 time period, while the percent of EMS incidents that responded to a death doubled (Lerner et al., 2020). Nevertheless, more research needs to be conducted to describe why CMFs and CMNFs would be differentially impacted.

Conclusion

This study of Pennsylvania child maltreatment fatalities and near fatalities raises concerns about the capacity of in-home services – as currently delivered –to protect children from serious harm At the same time, we find that foster care may reduce serious harm. These findings elevate the importance of continuous efforts to improve the quality and efficacy of in-home services. Such services must be carefully matched to families’ actual safety needs, with clearer thresholds for when placement is the more appropriate course. Although foster care is not a universal solution, its potential to reduce immediate risk of severe harm should not be overlooked. Future research should examine the quality and timing of services, the role of different types of in-home and out-of-home services and supports across family contexts or risk profiles, and how to more effectively align system responses with both the severity and type of risk present in child maltreatment cases.

Supplemental Material

Supplemental Material - Are In-Home and Foster Care Services Associated With Risk of Maltreatment-Related Fatality and Near Fatality Following a Child Protection Case? Evidence From Pennsylvania

Supplemental Material for Are In-Home and Foster Care Services Associated With Risk of Maltreatment-Related Fatality and Near Fatality Following a Child Protection Case? Evidence From Pennsylvania by Dylan Jones, Christian M. Connell and Sarah A. Font in Child Maltreatment

Footnotes

ORCID iDs

Dylan M. Jones

Christian M. Connell

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Supplemental Material

Supplemental material for this article is available online.

References

Campbell

K. A.

Wood

J. N.

Lindberg

D. M.

Berger

R. P.

(2021). A standardized definition of near-fatal child maltreatment: Results of a multidisciplinary Delphi process. Child Abuse & Neglect, 112(3), 104893. https://doi.org/10.1016/j.chiabu.2020.104893

Chiang

C.-J.

Kim

Yang

M. Y.

Jonson-Reid

LaBrenz

Chen

J.-H.

(2025). Navigating the services maze: Assessing service needs, referrals, receipts, and child protective services re-report and foster care entry outcomes. Child Abuse & Neglect, 159, 107164. https://doi.org/10.1016/j.chiabu.2024.107164

Commission to Eliminate Child Abuse and Neglect Fatalities . (2016). Within our reach: A national strategy to eliminate child abuse and neglect fatalities. Government Printing Office.

Dean

J. R.

Kaczor

Lorenz

Mason

Simonton

(2024). Characteristics of child abuse fatalities: Insights from a statewide violent death reporting system. Child Abuse & Neglect, 149, 106649. https://doi.org/10.1016/j.chiabu.2024.106649

Douglas

E. M.

(2016). Testing if social services prevent fatal child maltreatment among a sample of children previously known to child protective services. Child Maltreatment, 21(3), 239–249. https://doi.org/10.1177/1077559516657890

Goldstein

E. G.

Font

S. A.

(2025). Prevalence and treatment of maternal substance use disorder in child welfare. JAMA Health Forum, 6(3), Article e250054. https://doi.org/10.1001/jamahealthforum.2025.0054

Hammond

Eastman

A. L.

Leventhal

J. M.

Putnam-Hornstein

(2017). Maternal mental health disorders and reports to child protective services: A birth cohort study. International Journal of Environmental Research and Public Health, 14(11), 1320. https://doi.org/10.3390/ijerph14111320

Iwagami

Shinozaki

(2022). Introduction to matching in case-control and cohort studies. Annals of clinical epidemiology, 4(2), 33–40. https://doi.org/10.37737/ace.22005

Kennedy

J. M.

Lazoritz

Palusci

V. J.

(2020). Risk factors for child maltreatment fatalities in a national pediatric inpatient database. Hospital Pediatrics, 10(3), 230–237. https://doi.org/10.1542/hpeds.2019-0229

10.

Kotch

J. B.

Browne

D. C.

Dufort

Winsor

(1999). Predicting child maltreatment in the first 4 years of life from characteristics assessed in the neonatal period. Child Abuse & Neglect, 23(4), 305–319. https://doi.org/10.1016/s0145-2134(99)00003-4

11.

Lawler

Dronavalli

Uebel

Burns

Bajuk

Page

Lee

Dickson

Green

Dicair

Eastwood

Oei

J.-L.

(2025). Association between out-of-home care and mortality in children with prenatal drug exposure: A retrospective population-based cohort study. https://doi.org/10.1136/archdischild-2025-328474

12.

Lerner

E. B.

Newgard

C. D.

Mann

N. C.

(2020). Effect of the coronavirus disease 2019 (COVID-19) pandemic on the U.S. emergency medical services system: A preliminary report. Academic Emergency Medicine: Official Journal of the Society for Academic Emergency Medicine, 27(8), 693–699. https://doi.org/10.1111/acem.14051

13.

Marmor

Cohen

Katz

(2023). Child maltreatment during COVID-19: Key conclusions and future directions based on a systematic literature review. Trauma, Violence, & Abuse, 24(2), 760–775. https://doi.org/10.1177/15248380211043818

14.

Newman

(2024). A 4-Year-Old starved to death at home. How were the signs missed? New York. https://www.nytimes.com/2024/10/23/nyregion/jahmeik-modlin-dead-nyc-acs.html.Newman, 2024.

15.

Palusci

V. J.

Covington

T. M.

(2014). Child maltreatment deaths in the U.S. national child death review case reporting system. Child Abuse & Neglect, 38(1), 25–36. https://doi.org/10.1016/j.chiabu.2013.08.014

16.

Putnam-Hornstein

(2011). Report of maltreatment as a risk factor for injury death: A prospective birth cohort study. Child Maltreatment, 16(3), 163–174. https://doi.org/10.1177/1077559511411179

17.

Putnam-Hornstein

Cleves

M. A.

Licht

Needell

(2013). Risk of fatal injury in young children following abuse allegations: Evidence from a prospective, population-based study. American Journal of Public Health, 103(10), e39–e44. https://doi.org/10.2105/AJPH.2013.301516

18.

Raz

Sankaran

(2019). Opposing family separation policies for the welfare of children. American Journal of Public Health, 109(11), 1529–1530. https://doi.org/10.2105/AJPH.2019.305327

19.

Rebbe

Reddy

Huang

J. S.

Kuelbs

C. L.

Putnam-Hornstein

(2023). Counts and child protection reports of diagnosed child maltreatment before and after the COVID-19 pandemic onset. Child Abuse & Neglect, 146(1), 106450. https://doi.org/10.1016/j.chiabu.2023.106450

20.

Roberts

(2023). Building a World Without Family Policing [The Law and Political Economy Project]. https://lpeproject.org/blog/building-a-world-without-family-policing/.Roberts, 2023.

21.

Schneiderman

J. U.

Prindle

Putnam-Hornstein

(2021). Infant deaths from medical causes after a maltreatment report. Pediatrics, 148(3), Article e2020048389. https://doi.org/10.1542/peds.2020-048389

22.

Segal

Doidge

Armfield

J. M.

Gnanamanickam

E. S.

Preen

D. B.

Brown

D. S.

Nguyen

(2021). Association of child maltreatment with risk of death during childhood in South Australia. JAMA Network Open, 4(6) Article e2113221. https://doi.org/10.1001/jamanetworkopen.2021.13221

23.

Spencer

Wallace

Sundrum

Bacchus

Logan

(2006). Child abuse registration, fetal growth, and preterm birth: A population based study. Journal of Epidemiology & Community Health, 60(4), 337–340. https://doi.org/10.1136/jech.2005.042085

24.

Stiffman

M. N.

Schnitzer

P. G.

Adam

Kruse

R. L.

Ewigman

B. G.

(2002). Household composition and risk of fatal child maltreatment. Pediatrics, 109(4), 615–621. https://doi.org/10.1542/peds.109.4.615

25.

U.S. Department of Health & Human Services . (2024). Administration for children and families, administration on children, youth and families, children’s bureau. Child Maltreatment 2022. https://www.acf.hhs.gov/cb/data-research/child-maltreatment

26.

U.S. Department of Health & Human Services . (2021). Administration for children and families, administration on children, youth and families, children’s bureau. Child Maltreatment 2022. https://acf.gov/cb/report/child-maltreatment-2019

27.

U.S. Department of Health & Human Services . (2026). Administration for children and families, administration on children, youth and families, children’s bureau. Child Maltreatment 2022. https://acf.gov/sites/default/files/documents/cb/cm2024.pdf

28.

Volk

Christie

Phi . (2025). Philly removes neglected children from homes to keep them safe. But they often face worse abuse in foster care placements. The Philadelphia Inquirer.

29.

Wexler

(2018). Do spikes in foster care contribute to more child abuse deaths? Youth Today. https://youthtoday.org/2018/02/panic-foster-care-actually-contributes-spike-child-abuse-deaths/

30.

Wilson

R. F.

Yue

Thomas

K. E.

Kota

K. K.

Betz

C. J.

(2025). Racial and ethnic differences in fatal child abuse and neglect and the intersection of community poverty: U.S., 2003 to 2022. Child Protection and Practice, 4(1), 100108. https://doi.org/10.1016/j.chipro.2025.100108