Detection of child abuse in emergency departments: a multi-centre study.

OBJECTIVE: This study examines the detection rates of suspected child abuse in the emergency departments of seven Dutch hospitals complying and not complying with screening guidelines for child abuse.
DESIGN: Data on demographics, diagnosis and suspected child abuse were collected for all children aged ≤18 years who visited the emergency departments over a 6-month period. The completion of a checklist of warning signs of child abuse in at least 10% of the emergency department visits was considered to be compliance with screening guidelines.
RESULTS: A total of 24 472 visits were analysed, 54% of which took place in an emergency department complying with screening guidelines. Child abuse was suspected in 52 children (0.2%). In 40 (77%) of these 52 cases, a checklist of warning signs had been completed compared with a completion rate of 19% in the total sample. In hospitals complying with screening guidelines for child abuse, the detection rate was higher (0.3%) than in those not complying (0.1%, p<0.001).
CONCLUSION: During a 6-month period, emergency department staff suspected child abuse in 0.2% of all children visiting the emergency department of seven Dutch hospitals. The numbers of suspected abuse cases detected were low, but an increase is likely if uniform screening guidelines are widely implemented.

Introduction

Child abuse is a serious public health problem with high morbidity and mortality; worldwide, 155 000 deaths occur annually in children as a result of abuse or neglect.1 Preventing recurrent abuse or recognising early abuse is difficult but essential if long-term effects are to be limited. Although victims of child abuse have higher emergency department use than the general paediatric population, child abuse unfortunately often remains unrecognised in the emergency department.2–6

Woodman et al4 found consistent evidence that physical abuse affects 1 in 11 children in the UK each year and estimated that about 1% of all injury-induced child visits to emergency departments are due to physical abuse. In the Netherlands, an estimated 107 200–160 000 children are victims of child abuse annually; however, hospital staff were responsible for only 6% of the 16 156 reports on child welfare in 2008.7–9

The overall impression is of suboptimal detection of child abuse in hospitals. Implementing a uniform screening protocol for child abuse in emergency departments could increase detection rates, leading to a decrease in the short- and long-term effects of child abuse.2 10 Therefore, the aim of the current study is to assess the detection rates of child abuse in emergency departments in hospitals complying and not complying with uniform screening guidelines and examine the characteristics of cases of suspected abuse.

Although child abuse victims have higher emergency department use than the general paediatric population, child abuse unfortunately often remains unrecognised in the emergency department.

Preventing recurrent abuse or recognising abuse at an early stage is essential to limit its long-term effects.

Staff suspected child abuse in only 0.2% of all children visiting seven Dutch emergency departments over 6 months.

The detection rate of suspected child abuse was higher in emergency departments complying than not complying with screening guidelines.

Hospitals should increase compliance with screening guidelines and implement policies to improve emergency department detection of child abuse.

Methods

The province of South-Holland in the Netherlands has a population of 3.5 million people which is served by 22 hospitals. So that the cohort would be representative, data were collected from emergency departments in seven hospitals (a university children's hospital, three teaching hospitals and three rural hospitals) which together have approximately 200 000 emergency department visitors annually.

Detection of child abuse

All children aged 0–18 years who visited the emergency departments with a new complaint over a 6-month period were included. We checked all data on cases of suspected abuse, and considered it to be a case of suspected child abuse if emergency department staff noted their suspicions in the medical record. Children who were specifically referred to the hospital with a suspicion of child abuse were not included in the number of cases detected in the emergency department. The definition used for child abuse was any form of threatening or violent physical, mental or sexual interaction with a minor which is perpetrated actively or passively by parents or other persons on whom the minor is dependent and causes or will probably cause physical or mental injury and serious harm to the minor.11

All hospitals had a checklist of warning signs of child abuse available in the emergency department. These checklists were either used for all children visiting the emergency department, or, in one hospital, only for children with trauma, and were mostly completed by emergency department nurses. If one of the warning signs on the checklist was ticked, the emergency department nurse informed the treating specialist that the suspicion of child abuse was increased. The frequency of checklist use varied greatly. Compliance with screening guidelines for child abuse in the emergency department was considered to be use of the checklist in at least 10% of all emergency department visits by children. The numbers of suspected abuse cases detected in the emergency department were compared between hospitals complying and not complying with the screening guidelines.

This study was approved by the Medical Ethical Committee of the Erasmus MC, University Medical Centre Rotterdam.

Data collection and processing

Data were collected over 6 months in each hospital. In five of the seven hospitals this period covered August 2007 to January 2008. For logistical reasons, the collection period was November 2007 to April 2008 in one hospital and March 2008 to August 2008 in another. For all children data were collected on demographics, reason for the emergency department visit, the referrer, the treating specialist and the diagnosis at discharge. We used emergency department triage systems and (electronic) patient files, and if available, data from the screening checklist for child abuse. Based on all this information, we assessed whether or not emergency department staff suspected child abuse.

Statistical analysis

The χ2 test was used to compare baseline categorical variables between hospitals complying and not complying with screening policy, between cases screened and not screened for child abuse, and between children with and without suspected abuse. The Wilcoxon rank test was used for the continuous variable age. Analyses were performed using the statistical package SPSS 15.0. Statistical significance was defined as p<0.05.

Results

Emergency department visitors

During the 6-month study period, 24 472 children aged ≤18 years visited one of the seven emergency departments with a new complaint. Of these, 56% (13 754) were male, and more than 40% (10 546) were 4 years of age or younger. The majority (52%) visited the emergency department without being referred, and most (60%) had a surgical problem (eg, a fracture).

Of all children, 54% visited an emergency department which complied with screening guidelines. The patients in these hospitals were older, less often self-referred, and more often treated by a paediatrician than those who visited hospitals not complying with screening guidelines (table 1).

Table 1

Characteristics of all emergency department visitors aged ≤18 years over 6 months in seven Dutch hospitals classified as complying or not complying with screening policy

	Hospitals complying with screening policy	Hospitals not complying with screening policy	Total	p Value*
Emergency department visitors	13 109	11 363	24 472
Age in years (range)	7.6 (0–18)	6.0 (0–18)	6.9 (0–18)	<0.001
0–4	4799 (37%)	5747 (51%)	10 546 (43%)
5–8	2233 (17%)	1938 (17%)	4171 (17%)
9–12	2210 (17%)	1366 (12%)	3576 (15%)
13–18	3829 (29%)	2157 (19%)	5986 (25%)
Sex (male)	7353 (56%)	6401 (56%)	13 754 (56%)	0.58
Referrer
Self-referral	5505 (43%)	6937 (63%)	12 442 (52%)	<0.001
General practitioner	2921 (23%)	2622 (24%)	5543 (23%)
Other	4469 (35%)	1383 (13%)	5852 (25%)
Treating specialist
Surgeon	7616 (58%)	6924 (61%)	14 540 (60%)	<0.001
Paediatrician	4641 (36%)	3106 (28%)	7747 (32%)
Other	833 (6%)	1257 (11%)	2090 (9%)
Completed checklists	4726 (36%)	43 (0.4%)	4769 (20%)	<0.001
Suspected child abuse cases in the emergency department	43 (0.3%)	9 (0.1%)	52 (0.2%)	<0.001

Continuous variable (age) calculated with Wilcoxon rank test; categorical variables calculated with the χ2 test.

Screening

Overall, the checklist was completed for 36% of emergency department visitors (ranging from 16% to 56%) in hospitals complying versus 0.4% in hospitals not complying with the guidelines. The rate of suspected child abuse was higher in hospitals complying versus hospitals not complying with screening guidelines (0.3% vs 0.1%, p<0.001) (table 1).

The 4769 children who were screened for child abuse were significantly younger than those not screened (p<0.001). Children screened for abuse were less often self-referrals and were more often treated by a paediatrician. Significantly more cases of suspected abuse arose in children screened than not screened for child abuse (0.8% vs 0.1%, p<0.001) (table 2).

Table 2

Characteristics of all emergency department visitors aged ≤18 years over 6 months in seven Dutch hospitals classified by cases screened or not screened

	Cases screened for child abuse	Cases not screened for child abuse	p Value*
Emergency department visitors	4769	19 167
Age in years (range)	6.2 (0–18)	6.9 (0–18)	<0.001
0–4	2052 (43%)	8352 (44%)
5–8	976 (21%)	3112 (16%)
9–12	901 (19%)	2593 (14%)
13–18	838 (18%)	4937 (26%)
Sex (male)	2690 (56%)	10 772 (56%)	0.85
Referrer
Self-referral	2105 (45%)	10 258 (54%)	<0.001
General practitioner	1390 (30%)	4044 (21%)
Other	1274 (27%)	4865 (25%)
Treating specialist
Surgeon	2800 (59%)	11 432 (60%)	<0.001
Paediatrician	1739 (37%)	5894 (31%)
Other	227 (5%)	1753 (9%)
Suspected child abuse cases in he emergency department	40 (0.8%)	12 (0.1%)	<0.001

Continuous variable (age) calculated with Wilcoxon rank test; categorical variables calculated with the χ2 test.

Cases of suspected child abuse

After excluding 23 children specifically referred with suspected abuse, a suspicion of child abuse arose in 0.2% (52) of all remaining emergency department visitors. Cases of suspected child abuse were on average 3.9 years old, and 33/52 were aged ≤4 years. In 40 of the 52 cases, a checklist of warning signs of abuse was completed. Most suspicions concerned physical abuse, with fractures being the most often reported diagnosis (table 3).

Table 3

Diagnoses of suspected cases of child abuse in seven Dutch hospitals over 6 months

Type of abuse	Diagnosis	Children (n)
Physical abuse		37 (71%)
	Fractures/luxation	18
	Bruises/wounds/burns/contusion	15
	Inflicted traumatic brain/head injury	4
Neglect		13 (25%)
	Various diagnoses (intoxication, wounds, commotiocordis, infection)
Sexual abuse		2 (4%)
	Genital wound

Discussion

During a 6-month study period, a suspicion of child abuse arose in 0.2% of children aged ≤18 years who visited the emergency departments of seven hospitals, and in 0.3% of children aged ≤4 years. The significantly higher detection rate in hospitals complying with screening guidelines for child abuse (0.3%) compared to those not complying (0.1%) shows the importance of increased situational awareness for improving detection of child abuse. Checklists were completed in 40 (77%) of the 52 cases of suspected abuse but in only 19% of the total population, although the use of checklists in suspected cases might have been intentional selection.

Implementation of a structured screening protocol, therefore increasing the situational awareness of child abuse, might result in a higher detection rate of suspected child abuse. A checklist of warning signs of child abuse could be part of such a protocol, but a validated checklist is currently not available.12 If a checklist were used, it might result in more cases of suspected child abuse being identified and would be a first step in improving the detection rate of actual cases of child abuse. The use of a checklist in every child visiting the emergency department would result in universal screening to identify a high risk group.13 Unfortunately, an increase in the sensitivity of child abuse detection would lead to a decrease in specificity. Therefore, a protocol with clear guidelines on how to manage suspicions of child abuse is required. Where a case is suspected, it is very important that the physician informs the parents about his or her concerns without accusing anyone. Unfortunately, fear among physicians and other emergency department staff of making a false accusation can lead to failure to report cases of suspected abuse.

Screening for child abuse in the emergency department is not standard policy in most countries (eg, USA, Canada and the UK), but did become mandatory in the Netherlands in 2009.14 Earlier studies on screening (each including 2000–4422 patients) reported higher detection rates of suspected abuse (range 1.1–1.4%) than the 0.3% rate identified in the present study.15–17

We found that the detection rate of suspected child abuse was much higher when a checklist of warning signs of abuse was completed. Comparison of screened cases with those not screened showed that emergency department staff completed the checklist more often in children who were younger, were referred by a general practitioner or were treated for a paediatric complaint.

In the present study, children suspected of being abused were younger than the average child in the emergency department. The younger the child, the more vulnerable he or she is, the higher the risk that an injury requires medical attention, and the higher the chance that emergency department staff suspect abuse.18 However, because child abuse can affect children of all ages, emergency department staff must be aware of the risk in all children visiting the emergency department to avoid missing cases of child abuse.19

Physical abuse is the most common type of child abuse detected in the emergency department,6 as shown in the present study. Neglect and emotional and sexual abuse are more difficult to identify in an emergency department setting but also require attention. Overall, child abuse remains an under-reported problem. This can be attributed to, for example, inadequate knowledge and training of professionals regarding recognition of abuse injuries, unwillingness to report suspicions of abuse, and variations in what is considered to be abuse.18

Some limitations of the present study need to be addressed. First, we present cases of suspected abuse. Since abuse was not yet confirmed, this could have led to an overestimation of the detection rate of child abuse. Second, cases of suspected abuse might have been missed because only one of the hospitals systematically registered such cases. Finally, for optimal data comparison the same time period should have been used in all hospitals. However, due to logistical problems this was not possible in two of the participating centres.

The strengths of this study are the relatively long observational period, the large number of children, the inclusion of all patients (≤18 years old) who visited the emergency departments with a new complaint, and the fact that of the results are representative of various emergency department settings.

In summary, the detection rates of suspected child abuse in children who visited an emergency department were very low (0.2%). However, the detection rate of suspected abuse was higher in hospitals where emergency department staff complied with screening guidelines than in hospitals with non-compliant emergency department staff. We recommend that hospitals encourage compliance with screening guidelines, implement strict policies to improve the detection rate of suspected child abuse in emergency departments, and use the results of these interventions to develop an optimal screening protocol for emergency departments. Further research is recommended on how to identify genuine cases of child abuse among the high risk group of suspected cases identified by screening.