Introduction
Nonaccidental trauma (NAT) is a pervasive problem, with estimates that it represents 10% of all traumas evaluated in the emergency room for children less than 3 years old, with one out of every eight children in the USA falling victim before their 18th birthday [1,2]. Previous studies have identified many potential risk factors for NAT such as parental stress level, job loss, economic depression, and social isolation [1,3–6]. One such study showed that incidents of child maltreatment were positively correlated with unemployment rates and prior and current year foreclosure rates at the county level [3]. Similarly, a previous study showed that abusive head trauma in children under 5 increased significantly during the Great Recession of 2008 [4]. During the COVID-19 pandemic, the USA was faced with rising unemployment rates, nationwide school closures, and stay-at-home orders, all of which contributed significantly to exacerbating known NAT risk factors [1,7]. As these known risk factors rose, there was the potential for a corresponding rise in pediatric NAT [1].
Pediatric orthopedists have an important role in identifying NAT, as fracture is the second most common presentation of NAT after soft tissue damage and bruising, causing up to one-third of victims to require orthopedic treatment [8]. One previous study has shown that the pandemic and subsequent public health measures significantly impacted the types of fractures seen by orthopedists. Government restrictions on sports, travel, and elective surgeries were transformative for orthopedic practices, drastically reducing the frequency and types of injuries seen. The only injury type that did not decline in the time over the lockdown was femur fractures in elderly patients as the result of a fall [9]. This study did not evaluate NAT and therefore raises the question of whether the typical injuries seen in NAT may have also changed as a result of the pandemic lockdowns.
Studies evaluating the impact of the pandemic and associated public health measures on the incidence of NAT have been mixed [1,10–14] and there have been no studies characterizing the impact of the pandemic on orthopedic injuries resulting from NAT as the primary outcome. We sought to evaluate the impact of the COVID-19 pandemic on the frequency of NAT-associated fractures utilizing a large national multi-payer database. We hypothesized that, given the increase in risk factors affecting caregivers, we would see a corresponding rise in pediatric NAT-associated fractures.
Methods
Data source
We performed a retrospective cross-sectional study using de-identified data obtained from the Explore application on the Explorys Inc. platform (IBM, Armonk, New York, USA). The Explorys technology platform collects deidentified protected health information to meet Health Insurance Portability and Accountability Act (HIPAA) and Health Information Technology for Economic Clinical Health Act standards. For HIPAA-compliant statistical de-identification purposes, the Explorys Explore tool also does not allow reporting on sample sizes less than 10 and all population counts above 10 are rounded to the nearest 10. Information is standardized and normalized using a unified medical language system as it is aggregated into the Explorys data grid. At the time of this study, the Explorys platform contained inpatient and outpatient information on over 64 million patients representing greater than 15% of the US population from 360 distinct healthcare systems and roughly 920 000 healthcare providers.
Study population and statistical methods
We queried the Explorys Explore application using search terms from the Systematized Nomenclature for Medicine – Clinical Terms (SNOMED-CT) to identify patients under 18 years of age with a new diagnosis of NAT in 2019 and 2020. The following is a list of the inclusion terms used for this study: junior (age less than 18), victim of physical assault, victim of physical abuse, shaken baby syndrome, abuse by unrelated caregiver, abuse by relative of victim, physical abuse, child abuse, abuse, victim of child abuse, suspected victim of child abuse, and maltreatment syndromes. Exclusion terms included: adult victim of abuse, abuse of partner, and violent spouse. Results were then further analyzed to identify cases of NAT with a diagnosis of fracture within 21 days. Incidence of various fracture injuries in these cohorts was assessed using the following SNOMED-CT codes: fracture of bone, closed fracture of bone, fracture of lower limb, fracture of bone of head, fracture of upper limb, fracture of bones of trunk, fracture of shoulder, compression fracture, disorder of fracture healing, fracture of bone of nasal sinus, fracture of lumbar spine or pelvis, fracture of pelvis, fracture of vertebral column, fractures involving multiple body regions, multiple fractures, open fracture of bone, pathological fracture, fracture of sesamoid bone, stress fracture, and dislocation of joints.
Demographic data were compared between 2019 and 2020 where possible using chi-squared testing, and relative risk calculations for various fracture diagnoses included 95% confidence intervals (CIs).
Results
Overall, 9500 records of pediatric NAT were identified for 2019 compared to 9350 for 2020. In 2019, 550 records were associated with a diagnosis of fracture versus 570 in 2020. Analysis of trends in NAT is limited to overall frequency. Incidence could not be calculated for this study because Explorys does not allow researchers to identify the total number of active pediatric records the system holds for a given year. In both 2019 and 2020, patients with a diagnosis of NAT were most likely to be female (54% in 2019, 55% in 2020), Caucasian (45% in 2019, 44% in 2020), covered by Medicaid (49% in 2019, 47% in 2020), and between the ages of 10 and 14 years old (35% in 2019, 33% in 2020). Similarly, in both 2019 and 2020, those with a diagnosis of NAT and fracture were most likely to be Caucasian (33% in 2019, 32% in 2020), and be covered by Medicaid (60% in 2019, 56% in 2020), however in contrast they were also most likely to be male (60% in 2019, 61% in 2020) and be between the ages of 0 and 4 years old (53% in 2019, 60% in 2020). Table 1 depicts descriptive characteristics for our patient cohorts.
Table 1 -
Demographics of pediatric nonaccidental trauma patients with and without fractures
|
Victims of pediatric NAT 2019 (%) |
Victims of pediatric NAT 2020 (%) |
P value |
NAT patients with concurrent diagnosis of fracture within 21 days 2019 (%) |
NAT patients with concurrent diagnosis of fracture within 21 days 2020 (%) |
P value |
| Total |
9500 |
9390 |
|
550 |
570 |
|
| Gender |
|
|
0.34 |
|
|
0.91 |
| Male |
4340 (46) |
4230 (45) |
|
330 (60) |
350 (61) |
|
| Female |
5150 (54) |
5160 (55) |
|
220 (40) |
230 (40) |
|
| Age |
|
|
|
|
|
|
| 0–4 |
1270 (13) |
1770 (19) |
|
290 (53) |
340 (60) |
|
| 5–9 |
2830 (30) |
2590 (28) |
|
150 (27) |
150 (26) |
|
| 10–14 |
3330 (35) |
3100 (33) |
|
70 (13) |
70 (12) |
|
| 15–19 |
2160 (23) |
2010 (21) |
|
30 (5) |
30 (5) |
|
| Race |
|
|
<0.001 |
|
|
0.23 |
| Caucasian |
4250 (45) |
4120 (44) |
|
180 (33) |
180 (32) |
|
| African American |
2150 (23) |
2380 (25) |
|
80 (15) |
100 (18) |
|
| Asian |
70 (1) |
70 (1) |
|
0 |
0 |
|
| Latin American |
130 (1) |
130 (1) |
|
0 |
0 |
|
| Hispanic/Latino |
100 (1) |
70 (1) |
|
0 |
0 |
|
| Native American or Alaskan Native |
50 (1) |
60 (1) |
|
0 |
0 |
|
| Multi-racial |
350 (4) |
290 (3) |
|
10 (2) |
10 (2) |
|
| Unknown |
790 (8) |
820 (9) |
|
30 (5) |
20 (4) |
|
| Other |
300 (3) |
230 (2) |
|
40 (7) |
30 (5) |
|
| Refused to classify |
140 (1) |
120 (1) |
|
0 |
0 |
|
| Insurance |
|
|
<0.001 |
|
|
<0.001 |
| Medicaid |
4640 (49) |
4440 (47) |
|
330 (60) |
320 (56) |
|
| Private |
3200 (34) |
2850 (30) |
|
170 (31) |
160 (28) |
|
| Unknown |
650 (7) |
830 (9) |
|
20 (4) |
50 (9) |
|
| Self pay |
1190 (13) |
1250 (13) |
|
40 (7) |
30 (5) |
|
| Other public |
710 (7) |
690 (7) |
|
10 (2) |
10 (2) |
|
| Other |
300 (3) |
230 (2) |
|
30 (5) |
20 (4) |
|
| Medicare |
130 (1) |
120 (1) |
|
10 (2) |
0 |
|
| Military |
90 (1) |
70 (1) |
|
0 |
0 |
|
| Dental |
20 (<1) |
10 (<1) |
|
0 |
0 |
|
NAT, nonaccidental trauma.
Patients diagnosed with bone fracture secondary to NAT were most likely to have closed fractures (96% in both 2019 and 2020). The most common locations of fractures were the lower limb, upper limb, head, trunk, and shoulder. The relative risk of fracture secondary to NAT in 2020 when compared to 2019 was greater but not to a statistically significant degree [relative risk, 1.05 (95% CI, 0.94–1.17)] (Table 2). The relative risk for each category of fracture diagnosis in 2020 was slightly higher, but not to a statistically significant degree.
Table 2 -
Type of fracture suffered by pediatric nonaccidental trauma patients in 2019 and 2020
| Diagnosis |
2019 pediatric NAT patients with a diagnosis of fracture within 21 days (%) |
2020 pediatric NAT patients with a diagnosis of fracture within 21 days (%) |
Relative risk (95% confidence interval) |
| Fracture of bone |
550 |
570 |
1.05 (0.94–1.17) |
| Closed fracture of Bone |
530 (96) |
550 (96) |
1.05 (0.94–1.18) |
| Fracture of lower limb |
220 (40) |
240 (42) |
1.1 (0.92–1.32) |
| Fracture of bone of head |
160 (29) |
180 (32) |
1.14 (0.92–1.41) |
| Fracture of upper limb |
180 (33) |
180 (32) |
1.01 (0.82–1.24) |
| Fracture of bones of trunk |
140 (25) |
140 (25) |
1.01 (0.8–1.28) |
| Fracture of shoulder |
60 (11) |
60 (11) |
1.01 (0.71–1.45) |
| Fracture of vertebral column |
10 (2) |
N/A
a
|
|
| Open fracture of bone |
10 (2) |
N/A
a
|
|
| Compression Fracture |
N/A
a
|
N/A
a
|
|
| Disorder of fracture healing |
N/A
a
|
N/A
a
|
|
| Fracture of long bone, as birth trauma |
N/A
a
|
N/A
a
|
|
| Fracture of lumbar spine or pelvis |
N/A
a
|
N/A
a
|
|
| Fracture of pelvis |
N/A
a
|
N/A
a
|
|
| Fractures involving multiple body regions |
N/A
a
|
N/A
a
|
|
| Multiple fractures |
N/A
a
|
N/A
a
|
|
| Pathological Fracture |
N/A
a
|
N/A
a
|
|
| Fracture of sesamoid bone |
N/A
a
|
0 |
|
| Fracture of bone of nasal sinus |
0 |
N/A
a
|
|
| Stress fracture |
0 |
0 |
|
| Dislocation of joint |
0 |
0 |
|
Data are divided by SNOMED-CT subheadings available in Explorys under the code ‘Fracture of Bone’. Percentages may add to more than 100% as a result of patients being given multiple diagnosis codes.
NAT, nonaccidental trauma.
aExplorys hides information for groups <10 for privacy purposes.
Discussion
Although the public health measures implemented during the COVID-19 pandemic were necessary to slow the spread of disease, the social and economic sequalae thereof may have had significant unintended consequences on other aspects of the population’s health. One area of particular concern has been NAT and its associated sequela such as fractures. Previous studies have shown that there is an increase in violence against children associated with school closures and in the summer when schools are not in session [15]. Over the course of the pandemic, there was an unprecedented degree of school and childcare facility closures. Additionally, rates of unemployment increased to the highest levels that have been recorded since data collection began on this topic [7]. Accordingly, several studies showed drastic increases in parental stress [16,17]. Together, these factors made it very likely that the rates of NAT would increase because of COVID-19 lockdown orders, however data to this point has been mixed [1,10–14].
Although the overall incidence of NAT could not be calculated for this study, the fact that the absolute number of pediatric NAT cases decreased from 2019 to 2020 makes an increase in the overall incidence of NAT less likely. In contrast to overall risk factors for NAT, the current literature has not elucidated the risk factors that may impact the incidence and injury pattern of fractures in pediatric NAT patients, nor has it been identified if the pandemic has influenced either of those outcomes. We found that there was no statistically significant increase in the relative risk of fracture injuries in pediatric victims of NAT in 2020 when compared to 2019, nor were there significant changes in the pattern distribution of those fractures.
There are several reasons we may not have seen the anticipated increase in NAT or any changes in the risk and injury pattern of NAT-associated fractures. Regarding risk of fracture, our results may indicate that increasing social isolation, job loss, and parental stress may not necessarily increase a child’s chance of suffering a fracture due to NAT. In that case, we would expect the total number of fractures due to NAT to remain proportional to the overall incidence of NAT in a given year. A second possibility is that government interventions implemented at the same time as lockdowns strengthened the social safety net mitigating the impact of national stay-at-home orders. From a theoretical standpoint, the primary reason researchers believed there would be an increase in NAT is due to worsening job loss, increasing levels of parental stress, and a higher degree of social isolation. However, over the course of the pandemic, in addition to instituting measures aimed at slowing the spread of COVID-19, local and national governments implemented measures such as stimulus payments, increased unemployment benefits, student loan deferments, and eviction moratoriums with the goal of alleviating the social and economic repercussions of their actions. It is possible that these interventions were protective against NAT and associated fractures, by reducing the stress and economic uncertainty that has been proven to increase the risk of NAT.
Limitations and opportunities for further study
Our study does have several limitations. Explorys relies on EMR data to identify patients diagnosed with NAT using a variety of possible diagnosis codes. Underrepresentation of the diagnosis of NAT has always been a challenge, and therefore as more diverse populations have been equally affected by the pandemic, there is the possibility of increased missed cases [1]. Additionally, it has been hypothesized that public health measures such as the closure of both schools and childcare facilities may have also functioned to decrease detection of NAT [18]. Over the course of 2020, a significant proportion of children had their in-person interactions limited to their immediate household. As a result, they had fewer interactions with adults from outside their living situation who could potentially have recognized and reported signs of NAT. In the USA, 67% of substantiated child abuse or neglect reports come from victim-serving professionals with 19% of those reports coming from education personnel [18]. Even when cases of NAT are identified, anecdotally it can be a challenge to ensure cases are properly coded as providers are sometimes resistant to using the ICD-10 diagnosis of child abuse [19]. Both decreased detection as well as under-coding of NAT would lead to underestimations of the incidence of fractures due to NAT. One way to potentially identify the degree of missed cases would be to examine the incidence of nonunion and malunion fractures identified as a result of NAT in the months following lockdown protocols.
Other limitations of our study are similar to those that impact all large database studies, primarily the potential for variability between the data systems of participating institutions. Chart review of our data to evaluate the internal validity of the study was not possible due to the privacy measures in place. Additionally, to comply with HIPAA standards for protection of personal health information this data system could only reports data to the nearest 10, impacting the precision of our data. However, because these limitations impact data from both 2019 and 2020, they should have less of an impact on overall data trends. Additionally, although they do not specifically examine orthopedic manifestations of NAT, two similar studies offer some external validity to our findings. The first study utilizes data specifically from 52 US children’s hospitals also found that fractures as a result of NAT did not change significantly during the COVID-19 pandemic [14]. The second examines trends in orthopedic trauma from multiple UK centers over the course of the weeks of lockdowns. This study showed a non-statistically significant increase in child abuse reporting, with no increase in [20]. Therefore, we view these limitations as acceptable trade-offs allowing reporting on population-based trends that otherwise would be impossible to study. Future studies should focus on verifying the results of this study. Perhaps the most thorough way of doing so would be continued publication of institutional-level data followed by subsequent meta-analysis.
Conclusion
Pediatric orthopedic surgeons are very likely to care for children and adolescents who are victims of NAT. Despite the significant social stressors of the pandemic, we did not find evidence of increased pediatric fractures as a result of NAT compared to the prior year, nor did we note a change in the distribution of fracture location. One potential hypothesis to explain these results is that government interventions implemented at the same time as the pandemic mitigated the family stress of the pandemic and were protective against NAT. As healthcare professionals, their societies, and government agencies consider the potential unintended consequences of public health policy, is important to understand how the incidence and fracture patterns may have changed to better facilitate the timely identification of future NAT cases and inform future policy.
Acknowledgements
Conflicts of interest
There are no conflicts of interest.
References
1. Rodriguez CM, Lee SJ, Ward KP, Pu DF. The perfect storm: hidden risk of child maltreatment during the Covid-19 pandemic. Child Maltreat 2021; 26:139–151.
2. Kocher MS, Kasser JR. Orthopaedic aspects of child abuse. J Am Acad Orthop Surg 2000; 8:10–20.
3. Frioux S, Wood JN, Fakeye O, Luan X, Localio R, Rubin DM. Longitudinal association of county-level economic indicators and child maltreatment incidents. Matern Child Health J 2014; 18:2202–2208.
4. Berger RP, Fromkin JB, Stutz H, Makoroff K, Scribano PV, Feldman K, et al. Abusive head trauma during a time of increased unemployment: a multicenter analysis.
Pediatrics 2011; 128:637–643.
5. Conrad-Hiebner A, Byram E. The temporal impact of economic insecurity on child maltreatment: a systematic review. Trauma Violence Abuse 2020; 21:157–178.
6. Centers for Disease Control and Prevention. Violence Prevention: Risk and Protective Factors. 2021.
https://www.cdc.gov/violenceprevention/childabuseandneglect/riskprotectivefactors.html. [Accessed 24 February 2022]
7. Falk G, Romero P, Nicchitta I, Nyhof E. Unemployment rates during the COVID-19 pandemic. Congressional Research Service. 2021.
https://sgp.fas.org/crs/misc/R46554.pdf. [Accessed 24 February 2022]
8. Sink EL, Hyman JE, Matheny T, Georgopoulos G, Kleinman P. Child abuse: the role of the orthopaedic surgeon in nonaccidental trauma. Clin Orthop Relat Res 2011; 469:790–797.
9. Testa G, Sapienza M, Rabuazzo F, Culmone A, Valenti F, Vescio A, Pavone V. Comparative study between admission, orthopaedic surgery, and economic trends during Covid-19 and non-Covid-19 pandemic in an Italian tertiary hospital: a retrospective review. J Orthop Surg Res 2021; 16:601.
10. Lawson M, Piel MH, Simon M. Child maltreatment during the COVID-19 pandemic: consequences of parental job loss on psychological and physical abuse towards children. Child Abuse Negl 2020; 110:104709.
11. Paul AR, Adamo MA. Non-accidental trauma in pediatric patients: a review of epidemiology, pathophysiology, diagnosis and treatment. Transl Pediatr 2014; 3:195–207.
12. Swedo E, Idaikkadar N, Leemis R, Dias T, Radhakrishnan L, Stein Z, et al. Trends in U.S. Emergency Department Visits Related to Suspected or Confirmed Child Abuse and Neglect Among Children and Adolescents Aged <18 Years Before and During the COVID-19 Pandemic - United States, January 2019-September 2020. MMWR Morb Mortal Wkly Rep 2020; 69:1841–1847.
13. Sidpra J, Abomeli D, Hameed B, Baker J, Mankad K. Rise in the incidence of abusive head trauma during the COVID-19 pandemic. Arch Dis Child 2021; 106:e14.
14. Kaiser SV, Kornblith AE, Richardson T, Pantell MS, Fleegler EW, Fritz CQ, et al. Emergency visits and hospitalizations for child abuse during the COVID-19 pandemic.
Pediatrics 2021; 147:e2020038489.
15. Leaman L, Hennrikus W, Nasreddine AY. An evaluation of seasonal variation of nonaccidental fractures in children less than 1 year of age. Clin Pediatr (Phila) 2017; 56:1345–1349.
16. Brown SM, Doom JR, Lechuga-Peña S, Watamura SE, Koppels T. Stress and parenting during the global COVID-19 pandemic. Child Abuse Negl 2020; 110:104699.
17. Calvano C, Engelke L, Bella J, Kindermann J, Renneberg B, Winter SM. Families in the COVID-19 pandemic: parental stress, parent mental health and the occurrence of adverse childhood experiences—results of a representative survey in Germany. Eur Child Adolesc Psychiatry 2021; 1:1–13.
18. Campbell A. An increasing risk of family violence during the Covid-19 pandemic: strengthening community collaborations to save lives. Forensic Sci Int. 2020; 2:100089.
19. Hayman L. Challenges of Coding “Child Abuse”. ICD10 Monitor. 2018.
https://www.icd10monitor.com/challenges-of-coding-child-abuse
20. Sugand K, Park C, Morgan C, Dyke R, Aframian A, Hulme A, et al. Impact of the COVID-19 pandemic on paediatric orthopaedic trauma workload in central London: a multi-centre longitudinal observational study over the “golden weeks”. Acta Orthop 2020; 91:633–638.
