There is little understanding of the circumstances behind fatal pediatric opioid poisonings. Our objective was to characterize opioid fatalities according to child, family, and household factors.
We used data from the National Fatality Review-Case Reporting System to describe the circumstances behind the deaths of children 0 to 17 years of age who died of an opioid poisoning (ie, prescription opioid, heroin, illicit fentanyl) between 2004 and 2020. Decedents were stratified into age groups: 0–4, 5–9, 10–14, and 15–17 years.
The majority (65.3%) of the 1696 fatal opioid poisonings occurred in the child’s own home. Prescription opioids contributed to 91.8% of deaths, heroin contributed to 5.4%, and illicit fentanyl to 7.7%. Co-poisonings with nonopioid substances occurred in 43.2% of deaths. Among 0- to 4-year-olds, 33.9% died of homicide and 45.0% had a primary caregiver with a history of substance use/abuse. Among 10- to 14-year-olds and 15- to 17-year-olds, respectively, 42.0% and 72.8% of decedents had a history of substance use/abuse. In each age group, at least 25.0% of children were victims of previous child maltreatment.
A history of maltreatment and substance use—whether on the part of the caregiver or the child—are common factors surrounding fatal pediatric opioid poisonings, the majority of which occur in the child’s own home. Families with children of all ages would benefit from interventions focused on opioid prescribing, storage, disposal, and misuse. These findings also underscore the urgency of ensuring that access to naloxone becomes universal for families with a history of maltreatment and/or substance use.
Nearly 14 000 children died of opioid poisonings between 1999 and 2021.1 These deaths reflect the evolving nature of the opioid crisis: what was once an epidemic among young and middle-aged men is now taking a toll on Americans across the lifespan.2–9 In communities across the United States, children are increasingly exposed to prescription and illicit opioids, often with life-threatening consequences.10 Hospitalizations in children have more than doubled since the opioid epidemic began in the late 1990s,11 and the respective pediatric mortality rate has nearly tripled.1,12
Early on in the opioid crisis, which began 25 years ago, the vast majority of pediatric opioid poisonings were attributed to prescription opioids.2–4,11 Research now shows that fentanyl, a synthetic opioid that is increasingly manufactured illegally, is responsible for 94% of all pediatric deaths from opioids.1
Whether from illicit or prescription opioid, the majority of fatal pediatric opioid poisonings occur in the child’s home.1,12 Yet, there is little understanding of the circumstances behind these deaths. Previous national studies provide an empirical foundation for understanding the epidemiology of pediatric opioid poisonings in the United States and how age-specific mortality rates have changed since the epidemic began. These data, however, provide little insight into why children die of opioids and how circumstances vary across age groups. Although several recent studies on drug poisonings in general from prescription, over-the-counter, and/or illicit substances underscore the substantial contribution of opioids in pediatric poisoning deaths, these studies are limited to children younger than age 6 years.10,13,14 We are not aware of any study that extensively examines the circumstance surrounding opioid fatalities across the entire pediatric age spectrum.
Therefore, the objective of this study was to describe the circumstances surrounding fatal pediatric opioid poisonings with respect to child, family, and household factors using child death review (CDR) data. This objective aligns with the 2024 policy statement from the American Academy of Pediatrics stressing the importance of fatality review as a necessary step for informing strategies to mitigate preventable deaths in children.15
Methods
Data Source
Data were drawn from the National Fatality Review-Case Reporting System (NFR-CRS) from 2004 to 2020. NFR-CRS is a web-based tool used by multidisciplinary CDR teams to record information collected during fatality review meetings.16 The origins and structure of NFR-CRS and CDR, and their strengths and limitations, have been documented elsewhere.17 CDR teams rely on data gathered from birth and death certificates, medical records, death investigation records (eg, coroner, medical examiner, law enforcement), and legal and Child Protective Services (CPS) records.18 CDR teams are multidisciplinary and include representatives from coroner and medical examiner agencies, medicine, schools, public health, law enforcement, and CPS.18 NFR-CRS is managed by the National Center for Fatality Review and Prevention (National Center). More information about NFR-CRS, including the definitions of all variables collected, can be found on the National Center’s Web site.19
The study was considered exempt from approval by the Yale School of Medicine’s institutional review board.
Outcomes and Definitions
Inclusion Criteria for Identifying Opioid Fatalities
Deaths were limited to infants, children, and youth aged 0 to 17 years for whom (1) manner of death was categorized as unintentional, suicide, homicide, or undetermined, (2) the cause of death was a poisoning, and (3) the substance contributing to the death was a prescription opioid (including methadone), heroin, and/or illicit fentanyl/fentanyl analogs.
Infants for whom the descriptive text fields indicated likely neonatal opioid withdrawal syndrome (n = 53) were excluded.
Categorizing Prescription Versus Illicit Opioids
In the “Poisoning, Overdose, or Acute Intoxication” section of the NFR-CRS, users are asked to indicate the type of substance(s) involved in a poisoning death. The manner in which opioid data have been collected in this section has changed across versions of the NFR-CRS (version 1.0 in 2005 through version 6.0 in 2022). Although “prescription opioid pain medication” has always been an option, before version 6.0, methadone was collected separately from other substances and there was not a separate category for illicit fentanyl. In version 6.0, methadone was combined with other substances (eg, buprenorphine, naltrexone) into a new “medications for substance use disorder” category. Plus, a new “illicitly manufactured fentanyl/fentanyl analogs” category was added. National Center staff cross walked historical data into these new categories as appropriate.
In the present study, the “prescription opioid” category includes deaths in which either opioid pain medication or methadone (specified before version 6.0) were indicated as having contributed to the death. Methadone was included in this category because it is prescribed for both pain management and to treat opioid use disorder (OUD), but it may be diverted for nonmedical purposes or illicit use. Fentanyl is categorized as a prescription opioid if it was prescribed to the child or a household member.20 Otherwise, if the CDR team determined that the drug had been illegally manufactured and obtained illicitly it was categorized as “illicit fentanyl.” The new combined NFR-CRS category “medications for substance use disorder” was not used for analysis in this study because it combines potential opioid and non-opioid substances.
Statistical Analysis
Descriptive statistics were used to characterize opioid-related deaths among children aged 0 to 17 years by demographic characteristics, primary caregiver characteristics, incident circumstances, and other potential contributing factors. The social constructs of race and ethnicity are obtained from death certificates and reported because of known disparities in opioid deaths.21 All analyses were completed using SPSS.
Results
Characteristics of the Child
Sociodemographics
Of the 1696 children aged 0 to 17 years who died of an opioid poisoning between 2004 and 2020, 1097 (64.7%) were aged 15 to 17 years. Children aged 0 to 4 years accounted for the second highest number of deaths, 360 (21.2%), followed by 10 to 14 years, 181 (10.7%), and 5 to 9 years, 58 (3.4%). Further sociodemographic characteristics are shown in Table 1.
Characteristic . | n (%)a . |
---|---|
Age, y | |
0–4 | 360 (21.2) |
5–9 | 58 (3.4) |
10–14 | 181 (10.7) |
15–17 | 1097 (64.7) |
Sex | |
Male | 1086 (64.0) |
Female | 610 (36.0) |
Race | |
American Indian or Alaska Native | 54 (3.2) |
Asian | 22 (1.3) |
Black | 295 (17.4) |
Multiracial | 41 (2.4) |
White | 1218 (71.8) |
Missing/unknown | 66 (3.9) |
Ethnicity | |
Hispanic | 275 (16.2) |
Non-Hispanic | 1303 (76.8) |
Missing/unknown | 118 (6.9) |
Primary caregiver: English-speaking | |
Yes | 1226 (72.3) |
No | 27 (1.6) |
Missing/unknown | 443 (26.1) |
Urbanicity | |
Urban | 661 (39.0) |
Suburban | 577 (34.0) |
Rural | 261 (15.4) |
Missing/unknown | 197 (11.7) |
Setting of incident | |
Child’s home | 1108 (65.3) |
Home of relative/friend/foster parent | 395 (23.3) |
Other setting (eg, hospital, roadway, hotel, motel) | 130 (7.7) |
Missing/unknown | 63 (3.7) |
Poison control called | |
Yes | 23 (1.4) |
No | 1176 (69.3) |
Missing/unknown | 497 (29.3) |
Resuscitation attempted | |
Yesb | 1253 (73.9) |
Emergency medical services | 1120 (89.4) |
Parent or relative | 181 (14.5) |
Health care professional | 125 (10.0) |
No/not applicable | 175 (10.3) |
Missing/unknown | 268 (15.8) |
Toxicology screen performed | |
Yes | 1596 (94.1) |
No | 10 (0.6) |
Missing/unknown | 90 (5.4) |
Autopsy performed | |
Yes | 1577 (92.8) |
No | 77 (4.5) |
Missing/unknown | 45 (2.7) |
Death scene investigation at place of incident | |
Yes | 1366 (80.6) |
No | 28 (1.7) |
Missing/unknown | 302 (17.8) |
CPS action or welfare services implemented in response to death | |
Yes | 404 (23.8) |
No/not applicable | 926 (54.6) |
Missing/unknown | 366 (21.6) |
CDR team determined that the death could have been prevented | |
Yes, probably | 1,334 (78.7) |
No, probably not | 92 (5.4) |
Team could not determine | 136 (8.0) |
Unknown | 134 (7.9) |
Characteristic . | n (%)a . |
---|---|
Age, y | |
0–4 | 360 (21.2) |
5–9 | 58 (3.4) |
10–14 | 181 (10.7) |
15–17 | 1097 (64.7) |
Sex | |
Male | 1086 (64.0) |
Female | 610 (36.0) |
Race | |
American Indian or Alaska Native | 54 (3.2) |
Asian | 22 (1.3) |
Black | 295 (17.4) |
Multiracial | 41 (2.4) |
White | 1218 (71.8) |
Missing/unknown | 66 (3.9) |
Ethnicity | |
Hispanic | 275 (16.2) |
Non-Hispanic | 1303 (76.8) |
Missing/unknown | 118 (6.9) |
Primary caregiver: English-speaking | |
Yes | 1226 (72.3) |
No | 27 (1.6) |
Missing/unknown | 443 (26.1) |
Urbanicity | |
Urban | 661 (39.0) |
Suburban | 577 (34.0) |
Rural | 261 (15.4) |
Missing/unknown | 197 (11.7) |
Setting of incident | |
Child’s home | 1108 (65.3) |
Home of relative/friend/foster parent | 395 (23.3) |
Other setting (eg, hospital, roadway, hotel, motel) | 130 (7.7) |
Missing/unknown | 63 (3.7) |
Poison control called | |
Yes | 23 (1.4) |
No | 1176 (69.3) |
Missing/unknown | 497 (29.3) |
Resuscitation attempted | |
Yesb | 1253 (73.9) |
Emergency medical services | 1120 (89.4) |
Parent or relative | 181 (14.5) |
Health care professional | 125 (10.0) |
No/not applicable | 175 (10.3) |
Missing/unknown | 268 (15.8) |
Toxicology screen performed | |
Yes | 1596 (94.1) |
No | 10 (0.6) |
Missing/unknown | 90 (5.4) |
Autopsy performed | |
Yes | 1577 (92.8) |
No | 77 (4.5) |
Missing/unknown | 45 (2.7) |
Death scene investigation at place of incident | |
Yes | 1366 (80.6) |
No | 28 (1.7) |
Missing/unknown | 302 (17.8) |
CPS action or welfare services implemented in response to death | |
Yes | 404 (23.8) |
No/not applicable | 926 (54.6) |
Missing/unknown | 366 (21.6) |
CDR team determined that the death could have been prevented | |
Yes, probably | 1,334 (78.7) |
No, probably not | 92 (5.4) |
Team could not determine | 136 (8.0) |
Unknown | 134 (7.9) |
CDR, child death review; CPS, Child Protective Services.
a Percentages might not sum to 100 because of rounding.
b Denominator for subtypes only includes those with a “yes” response.
Cause and Manner of Death
As shown in Table 2, 1557 (91.8%) deaths involved a prescription opioid. Heroin contributed to 92 (5.4%) deaths and illicit fentanyl to 130 (7.7%).
Circumstances, n (%)a . | Overall n = 1696 . | 0 − 4 y n = 360 . | 5 − 9 y n = 58 . | 10 − 14 y n = 181 . | 15 − 17 y n = 1097 . |
---|---|---|---|---|---|
Implicated opioidb | |||||
Any prescription opioidc | 1,557 (91.8) | 331 (91.9) | 58 (100.0) | 171 (94.5) | 997 (90.9) |
Heroin | 92 (5.4) | 16 (4.4) | 0 (0.0) | 7 (3.9) | 69 (6.3) |
Illicit fentanyl | 130 (7.7) | 27 (7.5) | 0 (0.0) | 7 (3.9) | 96 (8.8) |
Co-poisoning | |||||
Opioid(s) only | 963 (56.8) | 308 (85.6) | 45 (77.6) | 113 (62.4) | 497 (45.3) |
Opioid(s) and nonopioid(s)d | 733 (43.2) | 52 (14.4) | 13 (22.4) | 68 (37.6) | 600 (54.7) |
Manner of death | |||||
Unintentional (accident) | 1,171 (69.0) | 108 (30.0) | 32 (55.2) | 118 (65.2) | 913 (83.2) |
Suicide/self-inflicted | 162 (9.6) | 0 (0.0) | — | 44 (24.3) | — |
Homicide/assault | 136 (8.0) | 122 (33.9) | — | — | — |
Undetermined | 227 (13.4) | 130 (36.1) | 14 (24.1) | — | — |
Child history of prior maltreatment | |||||
Yes | 519 (30.6) | 96 (26.7) | 20 (34.5) | 64 (35.4) | 339 (30.9) |
No | 630 (37.1) | 161 (44.7) | 19 (32.8) | 67 (37.0) | 383 (34.9) |
Missing/unknown | 547 (32.3) | 103 (28.6) | 19 (32.8) | 50 (27.6) | 375 (34.2) |
Open CPS case for the child at time of child’s death | |||||
Yes | 156 (9.2) | 39 (10.8) | 11 (19.0) | 20 (11.0) | 86 (7.8) |
No | 1,222 (72.1) | 255 (70.8) | 36 (62.1) | 126 (69.6) | 805 (73.4) |
Missing/unknown | 318 (18.8) | 66 (18.3) | 11 (19.0) | 35 (19.3) | 206 (18.8) |
Child ever placed outside of home before death | |||||
Yes | 218 (12.9) | 25 (6.9) | 11 (19.0) | 26 (14.4) | 156 (14.2) |
No | 1,050 (61.9) | 255 (70.8) | 34 (58.6) | 115 (63.5) | 646 (58.9) |
Missing/unknown | 428 (25.2) | 80 (22.2) | 13 (22.4) | 40 (22.1) | 295 (26.9) |
Siblings placed outside of home before child’s death | |||||
Yes | 125 (7.4) | 42 (11.7) | — | 15 (8.3) | — |
No/not applicable | 990 (58.4) | 229 (63.6) | 33 (56.9) | 108 (59.7) | 620 (56.5) |
Missing/unknown | 581 (34.3) | 89 (24.7) | — | 58 (32.0) | — |
Child history of substance use/abusee | |||||
Yes | 877 (54.4) | — | — | 76 (42.0) | 799 (72.8) |
No/not applicable | 409 (25.4) | 237 (85.6) | 46 (79.3) | 49 (27.1) | 77 (7.0) |
Missing/unknown | 327 (20.3) | — | — | 56 (30.9) | 221 (20.1) |
Child received prior mental health servicese | |||||
Yes | 466 (28.9) | — | — | 56 (30.9) | 402 (36.6) |
No/not applicable | 475 (29.4) | 206 (74.4) | 29 (50.0) | 49 (27.1) | 191 (17.4) |
Missing/unknown | 672 (41.7) | — | — | 76 (42.0) | 504 (45.9) |
Caregiver history as perpetrator of child maltreatment for any child | |||||
Yes | 362 (21.3) | 115 (31.9) | 18 (31.0) | 42 (23.2) | 187 (17.0) |
No | 523 (30.8) | 122 (33.9) | 16 (27.6) | 52 (28.7) | 333 (30.4) |
Missing/unknown | 811 (47.8) | 123 (34.2) | 24 (41.4) | 87 (48.1) | 577 (52.6) |
Caregiver history of substance use/abuse | |||||
Yes | 439 (25.9) | 162 (45.0) | 17 (29.3) | 46 (25.4) | 214 (19.5) |
No | 286 (16.9) | 54 (15.0) | 15 (25.9) | 36 (19.9) | 181 (16.5) |
Missing/unknown | 971 (57.3) | 144 (40.0) | 26 (44.8) | 99 (54.7) | 702 (64.0) |
Circumstances, n (%)a . | Overall n = 1696 . | 0 − 4 y n = 360 . | 5 − 9 y n = 58 . | 10 − 14 y n = 181 . | 15 − 17 y n = 1097 . |
---|---|---|---|---|---|
Implicated opioidb | |||||
Any prescription opioidc | 1,557 (91.8) | 331 (91.9) | 58 (100.0) | 171 (94.5) | 997 (90.9) |
Heroin | 92 (5.4) | 16 (4.4) | 0 (0.0) | 7 (3.9) | 69 (6.3) |
Illicit fentanyl | 130 (7.7) | 27 (7.5) | 0 (0.0) | 7 (3.9) | 96 (8.8) |
Co-poisoning | |||||
Opioid(s) only | 963 (56.8) | 308 (85.6) | 45 (77.6) | 113 (62.4) | 497 (45.3) |
Opioid(s) and nonopioid(s)d | 733 (43.2) | 52 (14.4) | 13 (22.4) | 68 (37.6) | 600 (54.7) |
Manner of death | |||||
Unintentional (accident) | 1,171 (69.0) | 108 (30.0) | 32 (55.2) | 118 (65.2) | 913 (83.2) |
Suicide/self-inflicted | 162 (9.6) | 0 (0.0) | — | 44 (24.3) | — |
Homicide/assault | 136 (8.0) | 122 (33.9) | — | — | — |
Undetermined | 227 (13.4) | 130 (36.1) | 14 (24.1) | — | — |
Child history of prior maltreatment | |||||
Yes | 519 (30.6) | 96 (26.7) | 20 (34.5) | 64 (35.4) | 339 (30.9) |
No | 630 (37.1) | 161 (44.7) | 19 (32.8) | 67 (37.0) | 383 (34.9) |
Missing/unknown | 547 (32.3) | 103 (28.6) | 19 (32.8) | 50 (27.6) | 375 (34.2) |
Open CPS case for the child at time of child’s death | |||||
Yes | 156 (9.2) | 39 (10.8) | 11 (19.0) | 20 (11.0) | 86 (7.8) |
No | 1,222 (72.1) | 255 (70.8) | 36 (62.1) | 126 (69.6) | 805 (73.4) |
Missing/unknown | 318 (18.8) | 66 (18.3) | 11 (19.0) | 35 (19.3) | 206 (18.8) |
Child ever placed outside of home before death | |||||
Yes | 218 (12.9) | 25 (6.9) | 11 (19.0) | 26 (14.4) | 156 (14.2) |
No | 1,050 (61.9) | 255 (70.8) | 34 (58.6) | 115 (63.5) | 646 (58.9) |
Missing/unknown | 428 (25.2) | 80 (22.2) | 13 (22.4) | 40 (22.1) | 295 (26.9) |
Siblings placed outside of home before child’s death | |||||
Yes | 125 (7.4) | 42 (11.7) | — | 15 (8.3) | — |
No/not applicable | 990 (58.4) | 229 (63.6) | 33 (56.9) | 108 (59.7) | 620 (56.5) |
Missing/unknown | 581 (34.3) | 89 (24.7) | — | 58 (32.0) | — |
Child history of substance use/abusee | |||||
Yes | 877 (54.4) | — | — | 76 (42.0) | 799 (72.8) |
No/not applicable | 409 (25.4) | 237 (85.6) | 46 (79.3) | 49 (27.1) | 77 (7.0) |
Missing/unknown | 327 (20.3) | — | — | 56 (30.9) | 221 (20.1) |
Child received prior mental health servicese | |||||
Yes | 466 (28.9) | — | — | 56 (30.9) | 402 (36.6) |
No/not applicable | 475 (29.4) | 206 (74.4) | 29 (50.0) | 49 (27.1) | 191 (17.4) |
Missing/unknown | 672 (41.7) | — | — | 76 (42.0) | 504 (45.9) |
Caregiver history as perpetrator of child maltreatment for any child | |||||
Yes | 362 (21.3) | 115 (31.9) | 18 (31.0) | 42 (23.2) | 187 (17.0) |
No | 523 (30.8) | 122 (33.9) | 16 (27.6) | 52 (28.7) | 333 (30.4) |
Missing/unknown | 811 (47.8) | 123 (34.2) | 24 (41.4) | 87 (48.1) | 577 (52.6) |
Caregiver history of substance use/abuse | |||||
Yes | 439 (25.9) | 162 (45.0) | 17 (29.3) | 46 (25.4) | 214 (19.5) |
No | 286 (16.9) | 54 (15.0) | 15 (25.9) | 36 (19.9) | 181 (16.5) |
Missing/unknown | 971 (57.3) | 144 (40.0) | 26 (44.8) | 99 (54.7) | 702 (64.0) |
CPS, Child Protective Services; –, data either not applicable or suppressed to protect confidentiality.
a Percentages might not sum to 100 because of rounding.
b Percentage sum exceeds 100 as substances from more than one category could have contributed to the death.
c Includes methadone and prescribed fentanyl.
d Nonopioids include non-opioid prescription substances (eg, antidepressants, antipsychotics), nonopioid illicit drugs (eg, cocaine, methamphetamine), over-the-counter drugs, and other substances (eg, alcohol). Nonopioid medications for substance use disorder (eg, buprenorphine, naltrexone) are not included.
e Infants younger than age 1 y excluded (n = 83).
Nonopioid medications/substances contributed to 733 (43.2%) deaths. Antidepressants, 193 (26.3%); benzodiazepines, 148 (20.2%); and alcohol, 132 (18.0%) were the most frequently found co-occurring substances (data not shown).
As shown in Table 2, manner of death was most commonly attributed to accidental ingestion for the sample overall (69.0%), but among children aged 0 to 4 years, 122 (33.9%) were attributed to homicide. For those aged 10 to 14 years, 44 (24.3%) deaths were attributed to suicide.
History of Previous Maltreatment and CPS Involvement
Among the overall sample, 519 (30.6%) had a history of previous maltreatment (eg, documentation through CPS, autopsy, law enforcement, medical records) (Table 2). In each age group, between one quarter and one third of children had a history of previous maltreatment. Notably, for this variable, a response was missing/unknown for 547 (32.3%) children.
Before death, 218 (12.9%) children had been placed outside of the home (eg, foster care). At the time of death, 156 (9.2%) children had an open CPS case (Table 2).
Adolescent Substance Use
Of the 1278 children aged 10 to 17 years, 875 (68.4%) had a history of substance use/abuse. When stratified by younger versus older adolescents, of the 181 children aged 10 to 14 years, 42.0% had a history of substance use/abuse versus 72.8% of the 1097 15- to 17-year-olds. Receipt of previous mental health services was found for 56 (30.9%) of those aged 10 to 14 years and 402 (36.6%) of those aged 15 to 17 years (Table 2).
Characteristics of Family and Household
Setting of Incident
As shown in Table 1, among all children and adolescents, 1108 (65.3%) poisoning incidents occurred in the child’s own home, whereas 395 (23.3%) occurred at the home of a relative, friend, or foster parent; for 130 (7.7%), the incident occurred elsewhere (eg, roadway, hotel, motel).
Primary Caregiver Substance Use
Among the overall sample, 25.9% had a primary caregiver with a history of substance use/abuse (Table 2). Among children aged 0 to 4 years and 5 to 9 years, 45.0% and 29.3%, respectively, had a caregiver with a history of substance use/abuse. These percentages were lower for the caregivers of the older children. Specifically, 25.4% for children aged 10 to 14 years and 19.5% for children aged 15 to 17 years (Table 2).
Perpetrators of Maltreatment and CPS Involvement
As shown in Table 2, overall, 21.3% of children had a caregiver with a history as the perpetrator of maltreatment for any child (ie, not necessarily the deceased child). This percentage was higher for the younger children: 31.9% of children 0 to 4 years of age and 31.0% of children 5 to 9 years of age. For children 10 to 14 years and 15 to 19 years, 23.2% and 17.0%, respectively, had a caregiver with a history as the perpetrator of maltreatment.
Discussion
To our knowledge, this is the first study to examine the circumstances behind the deaths of infants, children, and adolescents younger than 18 years of age who died of prescription opioids, heroin, and/or illicit fentanyl between 2004 and 2020. There are 4 key findings from this analysis of nearly 1700 pediatric opioid fatalities: (1) most children were poisoned in their own home, (2) the majority of poisonings were from prescription opioids, but heroin and illicit fentanyl contributed to nearly 1 in 10 fatalities and co-ingestions of nonopioid substances to nearly one half, (3) a history of substance use was reported for a substantial number of both young and older adolescents, and (4) at least 1 in 4 children within each age category had a history of maltreatment. The first 2 of these key findings are consistent with what has been noted in national epidemiologic studies,1,12 but to our knowledge the latter 2 findings have not been previously reported and thus contribute to the field a better understanding of the extent to which substance use and maltreatment are common circumstances surrounding fatal pediatric opioid poisonings.
Regarding previous research, national studies of US deaths certificate data show that the majority of children and adolescents die at home as opposed to in a medical setting (ie, inpatient, outpatient, or emergency departments).1,12 This is consistent with the findings from this current study, which examined the setting where the poisoning occurred (as opposed to the setting of death); we found that 64% of poisoning events occurred in the child’s own home.
Previous research has also shown that although prescription opioids have historically accounted for the majority of deaths, mirroring trends seen among adults, fentanyl is now the primary driver of the pediatric crisis.1 A recent study showed that pediatric deaths from fentanyl have risen more than 30-fold since 2013, and in 2021, fentanyl was responsible for all but 6% of pediatric opioid fatalities.1 In this current study, we found that heroin and illicit fentanyl contributed to nearly 1 in 10 fatalities, with illicit fentanyl poisonings exceeding those from heroin. Moreover, we found that co-ingestions of nonopioid substances, such as benzodiazepines and alcohol, contributed to nearly half of all deaths, which is also consistent with what is known of the contributing role that sedatives play in opioid-related deaths in both children and adults.1,12,22
It is in the area of adolescent deaths that the present study provides important new findings, particularly in relation to substance use and child maltreatment. Related to substance use, previous national studies of opioid hospitalizations and deaths have indicated that opioid misuse and abuse is a public health concern mainly for older adolescents (ie, 15–19-year-olds).1,11,12 Thus, our finding in this current study that 73% of adolescents in this age category had a history of substance use/abuse is in line with this finding. Yet, we were surprised to also find that 42% of 10- to 14-year-olds had a history of substance use/abuse. Moreover, we found that more than one third of both young and older adolescents had received mental health services in the past.
Regarding maltreatment, previous national studies rely on International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10), codes to classify manner of death. In general, national data attribute 25% of opioid-related deaths in children younger than age 5 to homicide,12 which would indicate abuse or criminal neglect. For older children and adolescents, homicide is rarely listed as a cause of opioid-related death.12 In contrast, the present study relies on data from multiple sources, including medical and autopsy records, law enforcement, and CPS; the results indicate that maltreatment is a concern for children and adolescents across age groups. We found for each age group that at least 1 in 4 had a history of maltreatment in the past, and the occurrence was greater than 1 in 3 for both 5- to 9-year-olds and 10- to 14-year-olds. Many of the children who died were known to come from families with previous child welfare involvement. Nearly 13% percent of children had been previously placed outside of the home and 9% had an open CPS case at the time of their death. One fifth of children had a caregiver or supervisor with a history as the perpetrator of maltreatment related to the child who died or another child.
The findings from this study also underscore the extent to which substance misuse/abuse in the family is a concern for pediatric opioid poisonings. More than one quarter of children and adolescents overall had a primary caregiver with a history of substance use/abuse, and for children younger than age 5 years, it was nearly one third. Among these young children, for nearly 1 in 5, their caregiver was impaired by drugs or alcohol at the time of the child’s death.
All these findings have important implications for reducing pediatric opioid poisonings. In particular, the finding that the majority of fatal pediatric opioid poisonings occur in the home has important policy and practice implications. In general, family-centered interventions are needed related to opioid storage, misuse, and disposal. Health care providers, including providers who prescribe opioids to adults, whether for pain management or for OUD, have a responsibility to ensure that their patients know of the dangers that opioids pose to all in the family and understand how to store and dispose of opioids to decrease a child or adolescent’s exposure to the drug.
Parents and caregivers should also understand how to use naloxone in the event that a child is poisoned. Moreover, even though naloxone is now available over the counter, interventions are needed to ensure that the medication gets into the homes of high-risk families, for whom research suggests that many will not follow through with obtaining the drug.23 Thus, it is critically important to remove as many barriers as possible so that caregivers have naloxone on hand. However, given that nearly half of all the deaths in this study involved a contributing nonopioid substance, naloxone must not be seen as the only solution to this complex problem.
Beyond these commonsense measures, harm-reduction strategies that increase the availability of OUD treatment of adults and adolescents are urgently needed. Research suggests that adolescents rarely receive timely OUD treatment or mental health services.24 However, with treatment expansion, it is important for providers to consider not only the patient with OUD in isolation but also the entire family and community at large.
Limitations
Our study has several key limitations. First, not all states report to the NFR-CRS database, and reporting criteria vary by state; therefore, these data cannot be used to calculate mortality rates; during the 17 years of our study (2004–2020), nationally, there were 3835 deaths from opioids in children younger than 18 years of age,25 so our study population of 1696 children represents ∼44% of the opioid-related deaths in this age group. Second, state participation in the NFR-CRS changes over time, and not all participating states permit their data to be included in the deidentified datasets that are released for research purposes. Thus, the assessment of temporal trends is not possible. Third, our analysis revealed that there were missing data for several variables, including demographic data. Notably, insurance and income data were missing for more than 75% of children and, thus, are not reported.26 Fourth, version 6.0 of the NFR-CRS data resulted in more specific reporting of types of opioid medications, which makes the results concerning fentanyl poisonings difficult to interpret. Fifth, for poisonings classified as unintentional, we are unable to differentiate between children who accidentally ingested opioids and those who intentionally did so. Sixth, the definition of child maltreatment includes referrals that might not have been substantiated, so there is the potential for misclassification. Furthermore, there are substantial missing data for the maltreatment variables (eg, 32% for the variable “child history of prior maltreatment”); thus, the results we report likely underestimate the severity of the problem.
Conclusions
In this examination of the circumstances surrounding the deaths of nearly 1700 children who died of prescription opioids, heroin, and/or illicit fentanyl between 2004 and 2020, we found that there were common circumstances across age groups. Most children die in their own homes from prescription opioids, but the illicit fentanyl crisis has exacerbated the opioid epidemic to the point where commonsense solutions that focus solely on opioid prescribing, storage, and disposal are unlikely to substantially reduce pediatric deaths. Moreover, the extent of substance use/abuse and maltreatment as circumstances surrounding opioid fatalities speaks to the urgent need for harm-reduction strategies and treatment initiatives that prioritize the entire family. Finally, naloxone should be made universally available to all families with a history of maltreatment and/or substance use.
Dr Gaither conceptualized and designed the study and drafted the initial manuscript; Drs Bechtel and Leventhal conceptualized and designed the study and critically reviewed and revised the manuscript; Ms Mintz and Ms McCollum assisted with the analyses and critically reviewed and revised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
References
Competing Interests
CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no potential conflicts of interest to disclose.
Comments