Characteristics of Children Ages 1–17 Who Died of COVID-19 in 2020–2022 in the United States

Data for this study were obtained from the National Fatality Review-Case Reporting System (NFR-CRS). The NFR-CRS is a Web-based application used by >1350 CDR teams in the United States to record information collected and discussed during fatality review meetings.^10,11 Variables include demographic and social characteristics of the child, family, and supervisor at the time of incident; circumstances of the death, including death investigation information; and risk and protective factors. Information from death certificates is also recorded in the NFR-CRS. The NFR-CRS is managed by the National Center for Fatality Review and Prevention (National Center), a program of the Michigan Public Health Institute. The National Center is primarily a federally funded resource center that offers training and technical assistance to fatality review teams, and also provides data quality support. More information about the NFR-CRS can be found on the National Center’s Web site.¹¹ This research was ruled exempt by the Michigan Public Health Institute’s institutional review board.

The study population included children 1 to 17 years of age who died in 2020 to 2022 of COVID-19 and whose deaths were reviewed by a CDR team. A death was determined to be because of COVID-19 if at least 1 of the following criteria was met:

1. the cause of death code assigned by the National Center for Health Statistics using the World Health Organization's International Classification of Diseases, 10th Revision (ICD-10), was reported as “U07.1,”^12,13 because a few CDR teams also indicated text in the ICD-10 field, the terms “covid,” “sars,” and “coronavirus” were also used to identify deaths;

2. the death certificate indicated that the immediate cause of death was covid, sars, or coronavirus;

3. the death certificate indicated that the underlying disease that initiated events resulting in death was covid, sars, or coronavirus; or

4. the CDR team indicated that the cause of the death was COVID-19.

Because data entry often lags several months from the date of the CDR meeting, only deaths where data entry was indicated to be complete in the NFR-CRS were included.

This study summarizes demographic characteristics about the decedents, including race, ethnicity, type of insurance, and the population density of the incident site. Race and ethnicity, which are social constructs, are reported because of their relationship to health disparities, which have led to variations in health care and mortality rates across different racial and ethnic groups in the United States.^14,15 Race and ethnicity were reported by CDR teams in the NFR-CRS on the basis of information obtained from death certificates.

Eighty-six infants who died of COVID-19 during this same time period were also considered for study inclusion. However, because of the difficulty in distinguishing if COVID-19 was the causal factor for premature birth or other significant conditions (such as preeclampsia) which may have led to the demise, infant deaths were excluded from analysis.

COVID-19 vaccine eligibility was determined for children who, on the basis of their age, could have had a vaccine available to them before they contracted COVID-19. This only included 16-17-year-olds who died in 2021 to 2022 and 12- to 15-year-olds who died in 2022 (COVID-19 vaccines became available in December 2020 for individuals 16 years and older and in May 2021 for individuals 12 years and older).¹⁶ Because month of death is not available in the NFR-CRS, the authors limited the computation of COVID-19 vaccine eligibility for 12- to 15-year-olds to 2022 (and not 2021 and 2022) for a more conservative estimate.

Existing medical conditions among children who died of COVID-19 were identified using several data elements. The NFR-CRS includes a specific question to recognize if the child had a disability or chronic illness. If “yes” is selected to this question, more information is gathered on the type of disability or chronic illness. These include physical/orthopedic, mental health/substance abuse, cognitive/intellectual, sensory, and unknown. Further specification is permitted by allowing free text for these responses. In addition, the NFR-CRS includes death certificate data, including the immediate cause, the underlying disease that initiated the chain of events, and other significant conditions contributing to the death. For each decedent, the responses to these questions (including the free-text responses) were assessed, and unique medical conditions were identified. These medical conditions were systematically assigned to an associated organ system on the basis of ICD, 11th Revision, for Mortality and Morbidity Statistics (version: January 2023) independently by 2 clinicians/coauthors (D.P. and S.B.K.).¹⁷ Disagreements were resolved through case reviews and discussion. Identifying chronic conditions on the basis of data from the CDR teams’ assessment and death certificate enabled a more comprehensive reporting of chronic conditions. Most conditions were reported by CDR teams in response to the NFR-CRS question about the presence of a disability or chronic illness. In some cases, the addition of the death certificate data increased the identification of conditions that may have contributed to the child’s increased health risk related to COVID-19 infection.

The medical conditions identified included nervous system disorders (eg, seizure disorder, spastic quadriplegia), congenital disorders (eg, chromosomal abnormalities, congenital malformations), obesity, respiratory disorders (eg, pulmonary fibrosis, asthma), neurodevelopmental disorders (eg, developmental delay, attention-deficit/hyperactivity disorder), endocrine disorders (eg, hypopituitarism, type 1 and type 2 diabetes), immune system disorders (eg, posttransplant, autoimmune disorder), cardiac disorders (eg, structural anomaly, arrythmia), cancers (eg, leukemia, rhabdomyosarcoma), gastrointestinal disorders (eg, liver failure, short bowel syndrome), and other. The other category represents various conditions that had counts of 5 or less and therefore were combined to protect anonymity. Examples of other conditions included prematurity, renal disorders, metabolic disorders, musculoskeletal disorders, sickle cell disease, and urological conditions.

Descriptive analyses were conducted by calculating frequencies and proportions. SPSS statistics version 29 for Windows (IBM Corp, Armonk, NY) was used for all analyses.

There were 15 243 deaths of children ages 1 to 17 years during 2020 to 2022 in the NFR-CRS on April 17, 2023, when the deidentified data file was prepared, and 183 deaths met the inclusion criteria for this study (Fig 1). The majority of deaths occurred in 2021 (57%), although it is possible that data entry for 2022 had not been completed by CDR teams at the time of the file download (Table 1).

One-third (33%) of deaths were 15-17-year-olds, 26% were 1- to 4-year-olds, 22% were 10-14-year-olds, and 18% were 5- to 9-year-olds. The mean age of decedents was 10.0 years (median 11.0 years). Children were most often male (56%). Over half (54%) were reported as white and 24% as Black. Eighteen percent were indicated as having Hispanic ethnicity. Though obtaining insurance status is often difficult for CDR teams (insurance status was missing in 46% for this sample), when insurance status was known, 74% were insured by Medicaid. Over the study period, nearly half of all incidents occurred in an urban area (47%), followed by suburban area (33%) and frontier/rural areas (13%). However, in 2022, a higher percentage of incidents occurred in suburban areas (46%) when compared with previous years (26% in 2021 and 37% in 2020; data not shown).

CDR teams reported that the person declaring the cause of death was a physician in at least 82% of deaths (Table 2). The majority of deaths did not have an autopsy (56%) or death investigation (54%). As far as exposure to COVID-19, CDR teams reported that 75% of children who died of COVID-19 were exposed to the virus within 14 days of death, and only 42% of children had contact with a health care provider within 1 month from date of death. More than one-third (n = 18, 36%) of youth were reported by CDR teams to have been eligible for COVID-19 vaccination, based on the decedent’s age and the Centers for Disease Control and Prevention’s (CDC’s) COVID-19 vaccine recommendations at the time of death. Of these, less than one-third of decedents (exact number is suppressed because of small numbers) were reported to have received the first dose of vaccine, and 6 did not. Vaccination data on the remaining decedents were missing.

A total of 124 decedents (68%) had 1 or more medical condition (excluding COVID-19), with the majority of these (86%) identified by the yes response to the disability or chronic illness question in the NFR-CRS (Table 3). A total of 276 medical conditions (among 124 decedents) were documented by CDR teams. The most frequent medical conditions reported were nervous system disorders (19%), congenital disorders (14%), obesity (12%), respiratory disorders (including asthma) (12%), and neurodevelopmental disorders (10%). Of the 124 decedents with medical conditions, 43% had 1 medical condition, 23% had 2 medical conditions, and 35% had 3 or more medical conditions. Of the children with an underlying condition, the mean number of medical conditions was 2.2. In addition, 30% of decedents were also noted to have experienced significant inflammation (eg, fever, laboratory evidence of inflammation, and involvement of 2 or more organs) that required hospitalization in the week before death, with 9 children diagnosed with multisystem inflammatory syndrome in Children.

To our knowledge, this is the largest descriptive study of the demographic characteristics, circumstances of death, and preexisting medical conditions of children who died of COVID-19 in the United States during the first 3 years of the pandemic. Through its multidisciplinary process, CDR teams documented valuable information that can be used to better characterize and understand these deaths. The findings from our study were consistent with previous research that attributed increased risk for mortality from COVID-19 among children with their age, sex, race, and underlying medical conditions.^5,18–21 

Unlike other mortality studies, these CDR data provide information about the time from exposure to death, contact with health care providers, and COVID-19 vaccine eligibility. Three-fourths of the children in this study were reported to have been exposed to COVID-19 within 2 weeks of death. This time frame is a shorter duration when compared with other COVID-19 mortality studies that reported median times from exposure to death of >18 days in adults.^22,23 This finding emphasizes for caregivers and providers that access to effective medical care is of paramount importance, especially for children with underlying medical needs, and may result in saving children’s lives. Less than half of the children were reported to have contact with a health care provider in the 30 days before death, further underscoring the importance of timely medical interventions in children with COVID-19.

Mortality data from the CDC for 2020 to 2021 and 2022 (provisional) indicate that COVID-19 death rates among children were highest among ages 15 to 17 years and ages 1 to 4 years compared with other age groups.⁴ Our findings also suggest a U-shaped curve in mortality trend by age for adolescents aged 15 to 17 years and children aged 1 to 4 years making up a higher proportion of COVID-19 deaths than children aged 5 to 9 years and 10 to 14 years. This is consistent with larger-scale epidemiologic studies of child mortality of COVID-19.^2,18,19,24 

CDC data for the same time period also show that COVID-19 death rates were highest among Black and American Indian/Alaska Native children compared with other racial groups.⁴ COVID-19 death rates for Hispanic and non-Hispanic children were the same rate for this time period.⁴ Although rates cannot be calculated using NFR-CRS data, Black children in this study were overrepresented among the deaths (24%) relative to the US population (15%).⁴ This highlights the need to accelerate efforts to significantly improve health and well-being and eliminate health disparities among the Black community.^1,4,5 Ensuring access to high-quality, timely care that addresses the legacy of racism is critically important.^6,25,26 

Though children with preexisting chronic medical conditions are at increased risk for severe illness, hospitalization, and death from COVID-19, studies are mixed on which chronic conditions are most highly associated with increased mortality risk.^5,21,27 The most reported chronic conditions in studies looking at children who died of COVID-19 were neurologic conditions, developmental conditions, chronic lung disease (including asthma), obesity, and cardiovascular conditions.^5,20,28–31 The results from our study found that nervous system disorders (eg, epilepsy, cerebral palsy) and congenital disorders were the most common conditions among decedents, followed by more commonly reported medical conditions associated with severe COVID-19 illness such as obesity and respiratory disorders. For this study, congenital disorders were defined broadly as structural or functional anomalies that occur during intrauterine life, including congenital anomalies, chromosomal abnormalities, or inborn errors of metabolism. It is estimated that 6% of births worldwide are considered to have a congenital disorder, compared with 14% of children in our study.³² The preponderance of children with congenital disorders in these data could indicate an additional vulnerability of children with complex or multisystem congenital disorders; alternatively, it may be related to a sampling bias in the deaths that were reviewed by CDR teams. Given the literature suggesting that specifically asthma and obesity are associated with more severe COVID-19, we reported these conditions separately from their broader system category, but found that they were not as common in our study population as reported in other studies of COVID-19 morbidity.

Although the NFR-CRS data on COVID-19 vaccine eligibility are highly missing, of the 18 children documented to be eligible to receive the vaccine, less than one-third received the first dose before death. Although vaccine uptake among children in the United States tended to be poor because of the belief that children experienced less severe disease, this finding emphasizes the need for parents and providers to address vaccine hesitancy and improve vaccine uptake to reduce mortality among the most medically vulnerable children.^33,34 

There are several limitations in the data. Although CDR exists in all 50 states, CDR is not standardized across the United States, and not all CDR programs use the NFR-CRS. CDR programs differ across states in multiple ways including team composition, level of funding, and case inclusion criteria. Furthermore, COVID-19 significantly impacted CDR teams’ ability to review deaths in a timely manner, because many team members were deployed to the pandemic response and did not meet to maintain isolation protocols. Teams may not have finished reviews for deaths occurring in 2022 at the time these data were downloaded (April 2023), because typically there is a 7- to 10-month lag between date of death and completing data entry in the NFR-CRS. For these reasons, rates cannot be calculated, and temporal trends cannot be assessed.

An additional limitation includes the potential for underascertainment of COVID-19 deaths by using ICD-10 codes to identify the study population. Because the assignment of ICD-10 codes from death certificates by the National Center for Health Statistics generally happens after the CDR review, ICD-10 codes in the NFR-CRS are highly missing. Using CDR team determination of COVID-19 as the cause of death, in addition to information on contributing and underlying causes from death certificates, resulted in identification of the majority of deaths because of COVID-19 in our study; however, more complete ICD-10 information in the NFR-CRS may have resulted in the identification of additional deaths.

Finally, the multidisciplinary nature of CDR ensures information is obtained from multiple sources for a robust and thorough discussion. However, the child’s medical records may not have been available for all team reviews. This may have contributed to an underreporting of vaccination eligibility and status and preexisting medical conditions.

Despite these limitations, the NFR-CRS has several strengths. The ability to quickly add and deploy questions related to the emerging COVID-19 pandemic and provide support to teams to help them understand and keep up with rapidly emerging medical information related to children is a unique strength of the NFR-CRS and the National Center in their training and technical assistance functions. Details on death investigation, autopsy, preexisting medical conditions, COVID-19 exposure time frame, COVID-19 vaccine eligibility, and time since previous contact with a health care provider offer important contextual details not available from other mortality data sources. Identifying prevention strategies and service enhancements is an essential component of CDR programs and the results of this study offer several opportunities, such as strengthening CDR’s ability to partner with health departments and health systems to access immunization registries and expanding community partnerships to address barriers to accessing vaccines and vaccine hesitancy. To prepare for future public health emergencies, building and strengthening linkages between CDR teams and vaccine registries will strengthen our emergency preparedness system.

The results of our study indicated that the majority of children died within 14 days of exposure to COVID-19. This brings heightened awareness to the need for protecting children during emerging infectious illnesses and pandemics that then prioritizes prevention efforts, such as vaccines and timely access to effective medical care. Continued strengthening of CDR teams’ access to data including information about a child’s medical history, including vaccination status, as well as collaborations with community partnerships, will strengthen fatality reviews, inform future pandemic preparedness, and ensure that we develop improved public health systems to protect our nation’s children.

We thank Deepa S. Joshi, MD, MPH, medical officer for the Division of Child, Adolescent, and Family Health, Health Resources and Services Administration, for her assistance in reviewing the medical conditions in our study.

CDC

Centers for Disease Control and Prevention

CDR

child death review

COVID-19

coronavirus disease of 2019

ICD-10

International Classification of Diseases, 10th Revision

National Center

National Center for Fatality Review and Prevention

NFR-CRS

National Fatality Review-Case Reporting System