Infantile subdural hematoma in Japan: A multicenter, retrospective study by the J-HITs (Japanese head injury of infants and toddlers study) group.

OBJECTIVE: Subdural hematoma in infants or toddlers has often been linked to abuse, but it is not clear how many cases actually occur and how many are suspected of abuse. The purpose of this study was to investigate subdural hematoma in infants and toddlers in Japan.
METHODS: This multicenter, retrospective study reviewed the clinical records of children younger than 4 years with head trauma who were diagnosed with any finding on head computed tomography (CT) and/or magnetic resonance imaging (MRI), such as skull fracture and/or intracranial injury. A total of 452 children were included. The group suspected to have been abused was classified as nonaccidental, and the group considered to have been caused by an accident was classified as accidental. Subdural hematoma and other factors were examined on multivariate analysis to identify which factors increase the risk of nonaccidental injuries.
RESULTS: Of the 452 patients, 158 were diagnosed with subdural hematoma. Subdural hematoma was the most common finding intracranial finding in head trauma in infants and toddlers. A total of 51 patients were classified into the nonaccidental group, and 107 patients were classified into the accidental group. The age of patients with subdural hematoma showed a bimodal pattern. The mean age of the accidental group with subdural hematoma was significantly older than that in the nonaccidental group (10.2 months vs 5.9 months, respectively. p < 0.001). Multivariate analysis showed that patients under 5 months old, retinal hemorrhage, and seizure were significant risk factors for nonaccidental injury (odds ratio (OR) 3.86, p = 0.0011; OR 7.63, p < 0.001; OR 2.49, p = 0.03; respectively). On the other hand, the odds ratio for subdural hematoma was 1.96, and no significant difference was observed (p = 0.34).
CONCLUSIONS: At least in Japanese children, infantile subdural hematoma was frequently observed not only in nonaccidental but also in accidental injuries. In infants with head trauma, age, the presence of retinal hemorrhage, and the presence of seizures should be considered when determining whether they were abused. Subdural hematoma is also a powerful finding to detect abuse, but care should be taken because, in some ethnic groups, such as the Japanese, there are many accidental cases.

Introduction

Head trauma is one of the leading causes of pediatric emergency department visits, and the diagnosis of abusive head trauma (AHT) by neurological imaging is difficult in practice. Many studies have suggested that subdural hematoma is considered characteristic of AHT [1-5]. In Japan, the Ministry of Health, Labor and Welfare publishes the "Guide to Responding to Child Abuse" as a standard for child guidance centers to take temporary custody; the 2014 revision states that the case of the infant with subdural hematoma is highly likely to have been abused and should be treated with suspicion. However, the true incidence of subdural hematoma due to head trauma in infants and toddlers, regardless of abuse or accident, is unknown. In this study, the extent of subdural hematoma in cases where imaging studies were performed after head trauma was examined, and the findings and the extent to which abuse was suspected and not suspected are presented.

Methods

This multicenter, retrospective study reviewed clinical records of children younger than 4 years with head trauma who visited our institutions between January 2014 and August 2020. Patients with some imaging findings such as fracture or intracranial injury were included in the study. Patients without imaging or with only extracranial findings or no obvious findings on imaging were excluded. Two university hospitals, two children’s hospitals, and two general hospitals participated in this study, making it possible to obtain data closer to the real world in Japan. A total of 452 children were included in this study. From the medical charts, the sex and age of the child, mechanism of injury, physical and neurological findings, radiological findings, retinal hemorrhage, surgical intervention, notification to child guidance centers, temporary protection by child guidance centers, and criminal cases were extracted. The imaging studies examined in this study included not only computed tomography (CT), but also magnetic resonance imaging (MRI). Imaging findings were confirmed and recorded by a board-certified pediatric neurosurgeon at each institution. The information about the mechanism of injury was primarily determined from the medical history provided by the caregiver to a physician. If necessary, medical evaluation reports by child protection teams at each facility and adequate investigation reports by police and/or child guidance centers were requested. Because many head injuries in infancy are suspected to be caused by abuse, institutions with a child protection team to assess whether the head injury was caused by abuse were included. Patients were classified into the nonaccidental or accidental groups by our definition (Table 1).

Table 1

Definition of “nonaccidental” and “accidental”.

Nonaccidental			Accidental
Perpetrator confesses to abuse.	Perpetrator did not confess to abuse, but was prosecuted	CPT at each facility determined possible abuse, child guidance center notified, judged as possible abuse, taken into temporary custody	CPT at each facility determined possible abuse, child guidance center notified, judged as accident, not taken into custody	CPT at each facility determined no potential for abuse, and not reported to the child guidance center	Birth injury	Traffic accidents

Abbreviations: CPT, child protection team.

Our definition includes cases that have been taken into temporary custody by the Child Guidance Center, as well as cases that have been determined to be accidents and have not been notified to the Child Guidance Center. We cannot deny the possibility that some of these cases may be accidents or cases of abuse. For this reason, we also have studied cases of obvious abuse (those who confessed and those who were criminally prosecuted) with cases of traffic accidents, and the results were compared with the results from our classification.

IRB/ethics committee approval and a statement regarding patient consent

The protocol for this study was approved by the Ethics Committee of Kansai Medical University (No. 2019232). Need for written patient consent was waived by the ethics committee because data were deidentified. Institutional review board approval was obtained from all participants’ institutions prior to submitting cases for this study.

Statistical analysis

Statistical analysis was performed with JMP 14.2.0. Continuous variables were analyzed using the Wilcoxon rank-sum test. Univariate and multivariate analyses were performed to examine the relationship between nonaccidental injury and prognostic factors. Univariate logistic model was used to compare each prognostic factors. Variables were included in a multivariate logistic regression model if their p value in the univariate analysis was statistically significant. The odds ratios (ORs) and 95% confidence intervals (CIs) were calculated.

Results

Of the 452 patients, 58 were included in the nonaccidental group, and 394 patients were included in the accidental group. The demographics and age distribution of the patients are shown in Table 2 and Fig 1, respectively. The mean age of the nonaccidental group was 5.8 months, and the mean age of the accidental group was 12.1 months (p < 0.001). The number of patients in the nonaccidental group was highest at 2 months and 4 months, and in the accidental group at 4 months (Fig 1). Of the 452 patients, 158 were diagnosed with subdural hematoma. Subdural hematoma was the second most common finding on CT from head trauma in infants after skull fracture, and it was more common than epidural hematoma (Table 2). The causes of injury in the accidental group are summarized and described in Table 3.

Table 2

Comparison of demographic and clinical presentations.

		Total	Nonaccidental group	Accidental group
Total no		452	58 (12.8%)	394 (87.2%)
Mean age (months)		11.3	5.8	12.1
Male		305	40 (13.1%)	265 (83.9%)
Female		147	18 (12.2%)	129 (87.8%)
SDH	total	158	51 (32.3%)	107 (67.7%)
	bilateral SDH	50	25 (50.0%)	25 (50.0%)
	unilateral SDH	108	26 (24.1%)	82 (75.9%)
Skull fracture		306	18 (5.9%)	288 (94.1%)
EDH		100	4 (4.0%)	96 (96.0%)
SAH		55	11 (20.0%)	44 (80.0%)
Contusion		33	5 (15.2%)	28 (84.8%)
Brain edema		35	15 (42.9%)	20 (57.1%)
Retinal hemorrhage	yes	85	45 (52.9%)	40 (47.1%)
	no	154	12 (7.8%)	142 (92.2%)
	not examined	213	1 (0.5%)	212 (99.5%)
Seizure	yes	78	34 (43.6%)	44 (56.4%)
	no	374	24 (6.4%)	350 (93.6%)
Surgery		85	31 (36.5%)	54 (63.5%)

Abbreviations: SDH, subdural hematoma; EDH, epidural hematoma; SAH, subarachnoid hemorrhage.

Fig 1

Age distribution of all patients in this study.

A: Total, B: Nonaccidental, C: Accidental.

Table 3

List of cause of injuries in accidental groups.

Cause of injury		No.(%)
Birth injury		8(2.0%)
Traffic accident	Motor vehicle accident	27(6.9%)
	Bicycle accident	12(3.0%)
Falls from >2m		28(7.11%)
Falls from ≤2m	Caretakers tumbled while holding the child	42(10.7%)
	Accidentally dropped by caretakers	89(22.6%)
	Self-inflicted tumbles	38(9.6%)
	Falling from couches or bed	33(8.4%)
	Other falls from ≤2m	91(23.1%)
Other injury		14(3.6%)
Unknown		12(3.0%)
Total		394

The mean age was 8.8 months, and the number of patients showed a bimodal peak at 4 months and at 8 months (Fig 2A). Of the patients with subdural hematoma, 51 were classified as the nonaccidental group and 107 as the accidental group. The mean age of patients in the nonaccidental group was 5.9 months, and the mean age of patients in the accidental group was 10.2 months (p < 0.001). The number of patients was highest at 2 months and 4 months in the nonaccidental group (Fig 2B) and at 8 months and 10 months in the accidental group (Fig 2C). Retinal hemorrhage was present in 45 of 58 patients in the nonaccidental group and 40 of 394 patients in the accidental group.

Fig 2

Age distribution of patients with subdural hematoma.

A: Total, B: Nonaccidental, C: Accidental. There is a peak at 2 to 4 of months of age in the nonaccidental group and another peak at 8 to 10 of months of age in the accidental group.

On univariate analysis, age under 5 months (OR 2.83, p < 0.001), subdural hematoma (OR 19.54, p < 0.001), retinal hemorrhage (OR 13.31, p < 0.001), brain edema (OR 6.52, p < 0.001), seizure (OR 12.10, p < 0.001), and surgical case (OR 7.22, p < 0.001) were significantly associated with increased odds of nonaccidental group (Table 4). On the other hand, the odds ratio of the nonaccidental group was significantly lower in cases with skull fracture (OR 0.17, p < 0.001) and epidural hematoma (OR 0.23, p = 0.006). A univariate analysis was conducted on 22 cases in the apparent abuse group, which included both confessed and criminally prosecuted cases, and 39 cases of traffic trauma, for a total of 61 cases. The results showed that there were significant differences in age under 5 months (OR 8.65, p = 0.0005), subdural hematoma (OR 45.71, p < 0.001), retinal hemorrhage (OR 15.30, p = 0.0034), brain edema (OR 4.71, p < 0.017), seizure (OR 18.75, p < 0.001), and surgical case (OR 40.8, p < 0.001) (Table 5). Similarly, the odds ratio of the apparent abuse group was significantly lower in cases with skull fracture (OR 0.10, p = 0.0002). Epidural hematoma (OR 0.12, p = 0.051) was just barely significant, but the OR was similarly low. These results were similar to the results of our analysis of cases classified as accidental and nonaccidental.

Table 4

Univariate analysis (nonaccidental vs accidental).

	n	Odds ratio	p value	Lower 95% CI	Upper 95% CI
Age under 5 months	452	2.84	0.0004 *	1.60	5.03
Male	452	1.17	0.60	0.65	2.12
Subdural hematoma	452	19.54	< .0001 *	8.60	44.40
Retinal bleeding	239	13.31	< .0001 *	6.43	27.54
Brain edema	452	6.52	< .0001 *	3.11	13.60
Skull fracture	452	0.17	< .0001 *	0.09	0.30
Contusion	452	1.236	0.68	0.46	3.33
Epidural hematoma	452	0.23	0.006 *	0.08	0.65
Subarachnoid hemorrhage	452	1.86	0.09	0.90	3.85
Seizure	452	12.10	< .0001 *	6.56	22.33
Surgery	452	7.23	< .0001 *	4.01	13.05

*Significant Abbreviations: CI, confidence interval.

Table 5

Univariate analysis of confessed or accused abused cases vs traffic accidental cases.

	n	odds ratio	p value	lower 95% CI	upper 95% CI
Age under 5 months old	61	8.65	0.0005 *	2.56	29.22
Male	61	2.04	0.2	0.68	6.09
Subdural hematoma	61	45.71	< .0001 *	8.62	242.31
Retinal bleeding	33	15.30	0.0034 *	2.46	95.19
Brain edema	61	4.71	0.017 *	1.33	16.70
Skull fracture	61	0.10	0.0002 *	0.029	0.33
Contusion	61	0.87	0.85	0.19	3.88
Epidural hematoma	61	0.12	0.051	0.014	1.01
Subarachnoid hemorrhage	61	0.74	0.65	0.20	2.76
Seizure	61	18.75	< .0001 *	4.77	73.74
Surgery	61	40.8	< .0001 *	8.72	190.92

*Significant Abbreviations: CI, confidence interval.

Of the 158 cases of subdural hematoma, 127 (80.4%) underwent fundus examination and were included in the multivariate analysis. On multivariate analysis, age under 5 months, retinal hemorrhage, and seizure were significantly associated with increased odds of nonaccidental injury (OR 3.86, p = 0.0011; OR 7.63, p < 0.001; OR 2.49, p = 0.03; respectively). On the other hand, subdural hematoma was not found to be significant (OR 1.96, p = 0.34) (Table 6). Multivariate analysis between the apparent abuse group and the accident group was not performed due to the small number of cases.

Table 6

Multivariate analysis of nonaccidental vs accidental.

	n	odds ratio	p value (Prob>ChiSq)	lower 95% CI	upper 95% CI
Under 5 months old	452	3.86	0.0011 *	1.71	8.71
Subdural hematoma	452	1.96	0.34	0.5	7.73
Retinal bleeding	239	7.63	< .0001 *	2.76	21.2
Brain edema	452	1.29	0.61	0.48	3.52
Skull fracture	452	1.03	0.96	0.36	2.98
Epidural hematoma	452	1.66	0.46	0.43	6.43
Seizure	452	2.49	0.03 *	1.1	5.69
Surgery	452	1.26	0.6	0.53	3.02

*Significant Abbreviations: CI, confidence interval.

Sensitivity, specificity, positive predictive value, and negative predictive value for subdural hematoma, retinal hemorrhage, and brain edema are summarized in Table 7. The sensitivity, specificity, and positive predictive value of subdural hematoma for the diagnosis of nonaccidental injury were 87.9%, 72.8%, and 32.0%, respectively.

Table 7

Sensitivity, specificity, positive predictive value, and negative predictive value.

Example
	Nonaccidental	Accidental
	A	B	positive predictive value
	C	D	negative predictive value
	sensitivity	specificity
Subdural hematoma
	Nonaccidental	Accidental
yes	51	107	0.32
no	7	287	0.98
	0.88	0.73
Retinal hemorrhage
	Nonaccidental	Accidental
yes	45	40	0.53
no	12	142	0.92
	0.79	0.78
Brain edema
	Nonaccidental	Accidental
yes	15	20	0.43
no	43	374	0.90
	0.26	0.95
Age under 5 months
	Nonaccidental	Accidental
yes	37	151	0.20
no	21	243	0.92
	0.64	0.62
Seizure
	Nonaccidental	Accidental
yes	34	44	0.44
no	24	350	0.94
	0.59	0.89

Discussion

Based on the literature, AHT should be part of the differential diagnosis in infants and toddlers with subdural hematoma. In a systematic review, subdural hematoma and retinal hemorrhage were strongly associated with AHT (OR 8.92, 95% CI 6.77–11.74; and OR 27.12, 95% CI 15.70–46.84, respectively) [2]. Therefore, infants and toddlers with subdural hematoma and/or retinal hemorrhage due to short falls should be thoroughly investigated for possible AHT. In the present study, the results of univariate analysis showed an odds ratio of 19.5, with a 95% CI of 8.6–44.4, which is similar to previous reports. However, on multivariate analysis, the odds ratio dropped to 1.96, with a 95% CI of 0.5–7.73, which was not significant. The reason for this difference was thought to be the confounding of subdural hematoma with factors such as age, presence of retinal hemorrhage, and presence of seizures. Furthermore, in the present study, subdural hematoma was found in 87.9% of cases in the nonaccidental group, and child abuse represented 32.2% of all traumatic subdural hematomas. Unlike reports from other countries, we therefore consider that subdural hematoma is frequently associated with, but is not characteristic of AHT in Japan.

In Japan, subdural hematoma has long been known to occur due to minor head trauma in infants who are old enough to pull themselves up. This was known as Nakamura’s type 1, and many Japanese neurosurgeons were aware of its existence [6, 7]. However, all of these reports have been criticized for not being fully investigated for child abuse [8]. More recently, Amagasa et al. reported that subdural hematoma, retinal hemorrhage, and neurological sequelae due to short falls were not seen after witnessed falls in children younger than 2 years in Japan [5]. However, in their paper, only 14 cases of subdural hemorrhage were found, and it is difficult to assume that most subdural hematomas in Japan are caused by abuse. In recent years, there have been similar reports of infantile acute subdural hematoma due to accidental minor head trauma from other countries [9-12]. Hogberg et al. reported that 104 of 306 (34.0%) infants with subdural hematoma younger than 1 year in Sweden were caused by accidental falls [13].

The peak incidence of AHT was reported to be 2 to 4 months of age in studies from North America [14-16]. Nuno et al. explained that the cause of the peak was that frustrated parents often use shaking as a means of quieting an infant, and reports of excessive crying at approximately 2 months of age coincide with documented cases of AHT [15]. The present study also shows that cases of head trauma less than 5 months old are more likely to have been abused, which is consistent with previous reports. On the other hand, Fujiwara et al. reported that patients with AHT younger than 2 years in Japan had 2 peaks of age, at 2 to 4 months and at 7 to 9 months [3]. Similarly, there are some reports from Japan that the age distribution of the AHT group had two peaks of age [4, 17]. Ganesh et al. also questioned why the Japanese reports of subdural hematoma with retinal hemorrhage after minor trauma were so very different from the bulk of the world’s literature [18]. However, it was unclear why the data from Japan showed such a bimodal pattern. Fujiwara et al. suggested that the first peak was due to the association between shaken infant syndrome and the peak of crying, and the cause of the second peak, which was not found in Western countries, was the easy availability and taking of head CT in Japan [3]. Amagasa et al. suggested that intense crying, such as sleep-related night-time crying, which peaks at 6 to 8 months in Japanese infants, and sleep sharing could be possible factors for abuse in older infants in Japan [4]. However, there is no evidence to support their suggestions in these papers. In the present study, 158 cases of subdural hematoma were carefully reviewed by the child protection team of each hospital, and cases suspected of abuse were reported to the child guidance center for full consideration of possible abuse. As a result, 107 cases were determined to be accidental, whereas 51 cases were determined to be nonaccidental. Furthermore, the peak age of the nonaccidental cases was 4 months, which is similar to the peak age of abused children in other countries, whereas the peak age of the accidental cases was 8 to 10 months. The second peak in the incidence of subdural hematoma in this study indicates that a large number of accidental head injuries may be included. Furthermore, the previous reports from Japan may have misclassified patients as non-accidental in the older age peak.

Retinal hemorrhage that is bilateral, severe, and includes the posterior pole and peripheral hemorrhage has been reported as characteristic of AHT [19, 20]. In the present study, retinal hemorrhage was also significantly associated with increased odds of AHT. However, an eye examination was not always done within 48 hours of admission in the present study, and subtle retinal hemorrhage might disappear [21]. On the other hand, Scheller reported ten cases of retinal hemorrhage with no evidence of brain injury and suggested that clinicians should reassess the importance of retinal hemorrhage in the setting of suspected AHT [22]. Mechanisms that cause retinal hemorrhage other than vitreoretinal traction, such as abrupt increases in intracranial pressure, have also been suggested [23]. Needless to say, the presence of retinal hemorrhage does not prove AHT, but it is one of the important factors suggesting AHT.

Seizure has also been reported to be associated with AHT. (2) Ichord et al. confirmed that the higher rate of hypoxic-ischemic injury on diffusion-weighted MRI in AHT than in accidental head trauma is likely multifactorial, involving respiratory insufficiency, seizure, and intracranial space occupying lesions requiring neurosurgical intervention [24]. The higher incidence of seizure in AHT compared to accidental head trauma may also be related to the frequency and the pattern of hypoxic-ischemic injury lesions. Seizure in the acutely injured brain may exacerbate injury directly through excitotoxic mechanisms or indirectly by exacerbating respiratory insufficiency [25]. These studies suggest that management of infants and toddlers with head injury, and particularly from AHT, should include meticulous attention to optimizing ventilation, oxygenation, perfusion, and the diagnosis and treatment of seizure.

The current study has several limitations. First, it had a retrospective design, which can result in misclassification because missing values occur frequently. Although children with intracranial injury were included in the sample, the threshold to suspect intracranial injury and order head CT and/or MRI was not clearly defined a priori. In addition, fundoscopy and skeletal surveys were not completed in some cases. Second, the classification of a child as nonaccidental or accidental depends on the judgment of the child guidance center and the CPT of each hospital, which is often based on medical findings. However, there may be a problem with the accuracy of the past reports on which the judgment is based, and as a result, there is a possibility that the judgment may be affected. Third, the nonaccidental group is overrepresented because it includes all cases taken into temporary custody by the child guidance center. Further research should be conducted regarding the differences in the mechanisms of accidental and nonaccidental head injuries in infants and toddlers.

Conclusions

In the present study in Japan, unlike in other countries, more than half of the infantile subdural hematomas were determined to be accidental. This suggests that the likelihood of accidental subdural hematoma in infants may vary by ethnicity. On the other hand, cases younger than 5 months, cases with retinal hemorrhage, and cases with seizure were found to more likely have been abused, as in other countries. The diagnosis of abuse should not be simply judged by applying the standards of one particular region to other regions, but it is necessary to consider the possibility that there are cultural and racial differences in each region.

30 Dec 2021

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Alfio Spina, M.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified (1) whether consent was informed and (2) what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

If you are reporting a retrospective study of medical records or archived samples, please ensure that you have discussed whether all data were fully anonymized before you accessed them and/or whether the IRB or ethics committee waived the requirement for informed consent. If patients provided informed written consent to have data from their medical records used in research, please include this information

3. You indicated that you had ethical approval for your study. In your Methods section, please ensure you have also stated whether you obtained consent from parents or guardians of the minors included in the study or whether the research ethics committee or IRB specifically waived the need for their consent.

Additional Editor Comments (if provided):

The authors of this paper seek to expound the association of subdural hematoma and non-accidental trauma in infants in toddlers in Japan. The paper gives an interesting look into how Japan differers from the western world in accidental and non-accidental trauma. It also solidifies previously presented ideas that retinal hemorrhages, young age, seizures coupled with subdural hematoma have a high predictive value for non-accidental trauma. Overall, a very interesting paper but could use some added information.

1. It would be interesting to have a section that talks or chart that shows they types of accidental trauma they have, ie. falling out of bed or off a changing table, to see if it differs from other countries. This information would be especially nice to know on the 107 patients with SDH that were determined to be accidental traumas.

2. When you discuss the two peaks of age that has been found previously in Japanese studies but that was not seen in yours, are you attempting to say that maybe the previous studies had patients that were misclassified as non-accidental in the older age peak? This currently isn't clear in the paper

3. Lastly, what is the take away of this article for readers outside of Japan? How does it contribute or educate the global reader? These should be addressed in the conclusion.

Academic Editor: Reviewers found potential in your article. Please revise it according to the attached comments.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors present a large series of infants and toddlers with evaluation of head trauma to assess an association between subdural hemorrhage and abusive head trauma when accounting for other clinical signs and age. They find a low positive-predictive value, which reflects interactions between prevalence and test/associations.

51 / 452 is a strikingly high rate of non-accidental trauma, but this may be clarified by the selective denominator involving children with abnormal imaging rather than all patients presenting following trauma.

The categorization of non-accidental and accidental is at the authors’ discretion. There may be legal and cultural distinctions in the CPT decisions on taking a child into temporary custody, which is, as illustrated on Table 1, the critical dividing line. When reporting neurosurgical outcomes such as mRS after thrombectomy, the dichotomization has been used to improve statistical power, but the individual categories are also reported for further reader understanding. This may be an appropriate application of that concept as well. Alternatively, the authors may be able to conduct the univariate and multivariate analysis excluding the patients ‘taken into temporary custody’ to demonstrate that the findings are similar and robust, essentially independent of this borderline definition.

Is there a standardized criteria for CPT taking a child into temporary custody? Such as a local or national law or an institutional standard by the overseeing CPT department? If so, a short description of this may be appropriate.

Are there children taken into permanent custody but for which a perpetrator is not prosecuted?

Table 2 would typically have % listed as a total of the column (i.e. male 40 (69%))

In Table 3, if the significant variables had a p-value formatted with bold, it may improve reader focus.

Ultimately, one main critique of this analysis is that the definition of non-accidental and accidental as the gold-standard truth for measuring outcomes. Here the outcome is measured by what clinicians and social workers deem to be suspected AHT. However, these decisions are made based on the constellation of multitrauma and HPI. Therefore, there is a concern about circular reasoning. The study aims to determine how strongly subdurals are associated with AHT. The definition of AHT is clinician and social work judgement of whether AHT exists. This judgement is based on expert opinion with limited data and no ability to confirm a historical event, but the existing opinion includes subdural hemorrhage and retinal hemorrhage as components of a suspicious trauma pattern. The clinicians and social workers therefore use SDH as one feature of their judgement, by which the study outcomes are measured. However, it remains unclear to the reviewer how to extricate the field from this circle – just that it must be acknowledged in the discussion.

Reviewer #2: I appreciate the author presenting this research article article. My comments as as follows

1. The major drawbacks of current stduy are the retrospective design and the nonaccidental group is overrepresented which made the applications of the results to the clinical unpredictable.

2. Table 5: the title is not clear. Sensitivity, specificity, positive predictive value, and negative predictive value ?

Table 5: How about the sensitivity, specificity, positive predictive value, and negative predictive value of seizure and

age under 5 months for diagnosis of AHT

Table 4: why epidural hematoma ans surgery were not evaluated in mutivariate analysis?

Table 2: the tilte is too simple : Overall results?

Fig.1: please clarify what do the x-axis and y-axis represnt?

Reviewer #3: The authors of this paper seek to expound the association of subdural hematoma and non-accidental trauma in infants in toddlers in Japan. The paper gives an interesting look into how Japan differers from the western world in accidental and non-accidental trauma. It also solidifies previously presented ideas that retinal hemorrhages, young age, seizures coupled with subdural hematoma have a high predictive value for non-accidental trauma. Overall, a very interesting paper but could use some added information.

1. It would be interesting to have a section that talks or chart that shows they types of accidental trauma they have, ie. falling out of bed or off a changing table, to see if it differs from other countries. This information would be especially nice to know on the 107 patients with SDH that were determined to be accidental traumas.

2. When you discuss the two peaks of age that has been found previously in Japanese studies but that was not seen in yours, are you attempting to say that maybe the previous studies had patients that were misclassified as non-accidental in the older age peak? This currently isn't clear in the paper

3. Lastly, what is the take away of this article for readers outside of Japan? How does it contribute or educate the global reader? These should be addressed in the conclusion.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Pokmeng See

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

26 Jan 2022

Reviewer #1:

Comment #1

The authors present a large series of infants and toddlers with evaluation of head trauma to assess an association between subdural hemorrhage and abusive head trauma when accounting for other clinical signs and age. They find a low positive-predictive value, which reflects interactions between prevalence and test/associations.

51 / 452 is a strikingly high rate of non-accidental trauma, but this may be clarified by the selective denominator involving children with abnormal imaging rather than all patients presenting following trauma.

The categorization of non-accidental and accidental is at the authors’ discretion. There may be legal and cultural distinctions in the CPT decisions on taking a child into temporary custody, which is, as illustrated on Table 1, the critical dividing line. When reporting neurosurgical outcomes such as mRS after thrombectomy, the dichotomization has been used to improve statistical power, but the individual categories are also reported for further reader understanding. This may be an appropriate application of that concept as well. Alternatively, the authors may be able to conduct the univariate and multivariate analysis excluding the patients ‘taken into temporary custody’ to demonstrate that the findings are similar and robust, essentially independent of this borderline definition.

Author’s Response: This suggestion was very helpful to improve the quality of this paper. Thank you very much. In accordance with the reviewer's suggestion, univariate analysis was performed, comparing traffic head injuries (car and bicycle accidents) and cases of apparent abuse (confessed or criminally prosecuted cases). The total number of cases was 61, which was clearly insufficient to perform multivariate analysis, so the multivariate analysis was not performed. The following sentences and table were added.

Change to Text: Lines 103-109 (Revised Manuscript with Track Changes)

Our definition includes cases that have been taken into temporary custody by the Child Guidance Center, as well as cases that have been determined to be accidents and have not been notified to the Child Guidance Center. We cannot deny the possibility that some of these cases may be accidents or cases of abuse. For this reason, we also have studied cases of obvious abuse (those who confessed and those who were criminally prosecuted) with cases of traffic accidents, and the results were compared with the results from our classification.

Change to Text: Lines 162-179 (Revised Manuscript with Track Changes)

On the other hand, the odds ratio of the nonaccidental group was significantly lower in cases with skull fracture (OR 0.17, p < 0.001) and epidural hematoma (OR 0.23, p = 0.006). A univariate analysis was conducted on 22 cases in the apparent abuse group, which included both confessed and criminally prosecuted cases, and 39 cases of traffic trauma, for a total of 61 cases. The results showed that there were significant differences in age under 5 months (OR 8.65, p =0.0005), subdural hematoma (OR 45.71, p < 0.001), retinal hemorrhage (OR 15.30, p =0.0034), brain edema (OR 4.71, p < 0.017), seizure (OR 18.75, p < 0.001), and surgical case (OR 40.8, p < 0.001) (Table 4). Similarly, the odds ratio of the apparent abuse group was significantly lower in cases with skull fracture (OR 0.10 , p =0.0002). Epidural hematoma (OR 0.12, p =0.051) was just barely significant, but the OR was similarly low. These results were similar to the results of our analysis of cases classified as accidental and nonaccidental.

New table has been added (Table 5)

Comment #2

Is there a standardized criteria for CPT taking a child into temporary custody? Such as a local or national law or an institutional standard by the overseeing CPT department? If so, a short description of this may be appropriate.

Author’s Response: In Japan, there is no law stipulating the standards for temporary protection by child guidance centers, but our Ministry of Health, Labor and Welfare's Child Abuse Response Guide, revised in 2014, states that cases of infants with subdural hematomas are highly likely to have been abused and should be treated with suspicion. The following sentences were added.

Change to Text: Lines 65-68 (Revised Manuscript with Track Changes)

In Japan, the Ministry of Health, Labor and Welfare publishes the "Guide to Responding to Child Abuse" as a standard for child guidance centers to take temporary custody; the 2014 revision states that the case of the infant with subdural hematoma is highly likely to have been abused and should be treated with suspicion.

Comment #3

Are there children taken into permanent custody but for which a perpetrator is not prosecuted?

Author’s Response: There are some children who are taken into permanent custody, even if the perpetrator is not prosecuted.

Comment #4

Table 2 would typically have % listed as a total of the column (i.e. male 40 (69%))

Author’s Response: The Table has been revised according to the reviewers' suggestions.

Change to Table: % has been listed as a total of the column of Table 2

Comment #5

In Table 3, if the significant variables had a p-value formatted with bold, it may improve reader focus.

Author’s Response: The Table has been revised according to the reviewers' suggestions.

Change to Table: In Table 4 (Formerly Table 3), significant p-value have been formatted with bold.

Ultimately, one main critique of this analysis is that the definition of non-accidental and accidental as the gold-standard truth for measuring outcomes. Here the outcome is measured by what clinicians and social workers deem to be suspected AHT. However, these decisions are made based on the constellation of multitrauma and HPI. Therefore, there is a concern about circular reasoning. The study aims to determine how strongly subdurals are associated with AHT. The definition of AHT is clinician and social work judgement of whether AHT exists. This judgement is based on expert opinion with limited data and no ability to confirm a historical event, but the existing opinion includes subdural hemorrhage and retinal hemorrhage as components of a suspicious trauma pattern. The clinicians and social workers therefore use SDH as one feature of their judgement, by which the study outcomes are measured. However, it remains unclear to the reviewer how to extricate the field from this circle – just that it must be acknowledged in the discussion.

Author’s Response: The possibility that judgments of abuse may fall into the circular argument is described in the discussion

Change to Text: Lines 304-308 (Revised Manuscript with Track Changes)

Second, the classification of a child as nonaccidental or accidental depends on the judgment of the child guidance center and the CPT of each hospital, which is often based on medical findings. However, there may be a problem with the accuracy of the past reports on which the judgment is based, and as a result, there is a possibility that the judgment may be affected.

Reviewer #2:

Comment #1

1. The major drawbacks of current study are the retrospective design and the nonaccidental group is overrepresented which made the applications of the results to the clinical unpredictable.

Author’s Response: We believe that the limitation of this paper is the possibility of overrepresented nonaccidental group due to the boundaries we have set. So, in order to check the accuracy of this boundary between nonaccidental and accidental, we compared the results of univariate analysis of apparent abuse and traffic accidents, as described in the response to Reviewer #1, and checked whether the boundary we set between nonaccidental and accidental is reasonable. In both cases, the items that showed a significant difference were almost identical, and we consider that the boundary we divided was reasonable.

Change to Text: Please see response to Reviewer #1’s comment #1

Comment #2

2. Table 5: the title is not clear. Sensitivity, specificity, positive predictive value, and negative predictive value?

Comment #3

Table 5: How about the sensitivity, specificity, positive predictive value, and negative predictive value of seizure and

age under 5 months for diagnosis of AHT

Author’s Response: The title of the Table has been revised according to the reviewers' suggestions. And we have added a table of sensitivity, specificity, positive and negative predictive values for seizure and age under 5 months.

Change to Text: Table 7 (formerly Table 5) has been modified.

Table 4: why epidural hematoma ans surgery were not evaluated in mutivariate analysis?

Author’s Response: Initially, we reduced the number of items to be analyzed in order to improve the accuracy of the multivariate analysis, but as pointed out by the reviewers, we included all items that were significantly different in the univariate analysis. The numbers changed slightly, but the items that showed significant differences remained the same.

Change to Table and Text: Table 6 (formerly Table 4) has been modified.

Lines 44-52 (Revised Manuscript with Track Changes)

(odds ratio (OR) 3.86, p = 0.0011; OR 7.63, p < 0.001; OR 2.49, p = 0.03; respectively). On the other hand, the odds ratio for subdural hematoma was 1.96, and no significant difference was observed (p = 0.34).

Table 2: the tilte is too simple : Overall results?

Author’s Response: The title of the Table has been revised according to the reviewers' suggestions.

Change to Table: The title of the Table 3 (formerly Table 2) has been modified.

Fig.1: please clarify what do the x-axis and y-axis represnt?

Author’s Response: Thank you for pointing this out. We have filled in what the X and Y axes indicate.

Change to Figure:　In Figure 1 and 2, we have filled in what the X and Y axes indicate.

Reviewer #3: The authors of this paper seek to expound the association of subdural hematoma and non-accidental trauma in infants in toddlers in Japan. The paper gives an interesting look into how Japan differers from the western world in accidental and non-accidental trauma. It also solidifies previously presented ideas that retinal hemorrhages, young age, seizures coupled with subdural hematoma have a high predictive value for non-accidental trauma. Overall, a very interesting paper but could use some added information.

1. It would be interesting to have a section that talks or chart that shows they types of accidental trauma they have, ie. falling out of bed or off a changing table, to see if it differs from other countries. This information would be especially nice to know on the 107 patients with SDH that were determined to be accidental traumas.

Author’s Response: Thank you for your interest in the cause of the injury.We have included in Table 3 the causes of injury in the accident group and the number of patients by cause of injury.

The reviewer would also like to know what is the cause of the accidental subdural hematoma injury. However, we believe that a detailed study between the cause of injury and the subdural hematoma is needed. In addition, if the cause of injury of subdural hematoma is discussed in this paper, the paper would be too long. Therefore, we are preparing another paper that examines the relationship between the cause of injury and subdural hematoma. Please understand that we do not describe the causes and number of subdural hematoma injuries in this manuscript to prevent the current paper from being dual publication with the other paper that we are working on. We will show a table with data on subdural hematoma in addition to the entire accident group for the reviewer's reference only.

Cause of injury All Subdural Hematoma

Birth injury 8(2.0%) 6(5.6%)

Traffic accident Motor vehicle accident 27(6.9%) 6(5.6%)

Bicycle accident 12(3.0%) 1(0.9%)

Falls from >2m 28(7.11%) 4(3.7%)

Falls from ≤2m Caretakers tumbled while holding the child 42(10.7%) 9(8.4%)

Accidentally dropped by caretakers 89(22.6%) 10(9.3%)

Self-inflicted tumbles 38(9.6%) 24(22.4%)

Falling from couches or bed 33(8.4%) 18(16.8%)

Other falls from ≤2m 91(23.1%) 16(14.9%)

Other injury 14(3.6%) 6(5.6%)

Unknown 12(3.0%) 7(6.5%)

Total 394 107

Change to Text: Lines 101-102 (Unmarked); Lines 112-113 (Revised Manuscript with Track Changes)

The causes of injury in the accidental group are summarized and described in Table 3.

In the Table 3 (new Table), list of cause of injuries in accidental group have been presented.

2. When you discuss the two peaks of age that has been found previously in Japanese studies but that was not seen in yours, are you attempting to say that maybe the previous studies had patients that were misclassified as non-accidental in the older age peak? This currently isn't clear in the paper

Author’s Response: We tried not to be direct, but it may have been confusing and inappropriate. We have clearly stated that the previous studies may have misclassified patients as non-accidental in the older age peak as indicated.

Change to Text: Lines 275-276 (Revised Manuscript with Track Changes)

Furthermore, the previous reports from Japan may have misclassified patients as non-accidental in the older age peak

3. Lastly, what is the take away of this article for readers outside of Japan? How does it contribute or educate the global reader? These should be addressed in the conclusion.

Author’s Response: Our results show that there are regional differences in subdural hematoma in infants. The reason for this is unknown, but it is necessary to consider the possibility that there are cultural and racial differences. Therefore, it is necessary to discuss the diagnosis of abuse based on the actual situation in each region.

Change to Text: Lines 318-321 (Revised Manuscript with Track Changes)

The diagnosis of abuse should not be judged by applying the standards of one particular region to other regions, but it is necessary to consider the possibility that there are cultural and racial differences in each region.

From editorial office:

3. You indicated that you had ethical approval for your study. In your Methods section, please ensure you have also stated whether you obtained consent from parents or guardians of the minors included in the study or whether the research ethics committee or IRB specifically waived the need for their consent.

Change to Text: Lines 110-115 (Revised Manuscript with Track Changes)

IRB/ethics committee approval and a statement regarding patient consent

The protocol for this study was approved by the Ethics Committee of Kansai Medical University (No. 2019232). Need for written patient consent was waived by the ethics committee because data were deidentified. Institutional review board approval was obtained from all participants’ institutions prior to submitting cases for this study.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

10 Feb 2022

Infantile subdural hematoma in Japan: A multicenter, retrospective study by the J-HITs (Japanese Head injury of Infants and Toddlers study) group

PONE-D-21-35151R1

Dear Dr. Nonaka,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Alfio Spina, M.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

16 Feb 2022

PONE-D-21-35151R1

Infantile subdural hematoma in Japan: A multicenter, retrospective study by the J-HITs (Japanese Head injury of Infants and Toddlers study) group

Dear Dr. Nonaka:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Alfio Spina

Academic Editor

PLOS ONE