- Email: [email protected]
The epidemiology of child maltreatment
SYMPOSIUM: SAFEGUARDING CHILDREN
The epidemiology of child maltreatment
public health and surveillance from organisation such as the Centers for Disease Control. ‘Child maltreatment’ is defined as any act of commission or omission by a parent or other caregiver that results in harm, potential for harm, or threat of harm to a child.’ There is no requirement for this harm to be intended, and most cases of harm to children are perpetrated by parents or caregivers; for example, 82% of substantiated cases in one study from the United States. Other definitions and examples are shown in Table 1. The UK definition of psychological or emotional abuse is broad and includes interactions between child and parent or carer in which there are unintentional interactions with these adverse effects, as encompassed in the phrase: ‘The persistent emotional ill-treatment of a child such as to cause severe and persistent adverse effects on the child’s emotional development.’ Fabricated or induced illness (FII) is an established and currently preferred term in the UK for the condition that is also known as ‘Munchausen syndrome by proxy’, ‘paediatric symptom falsification’ or ‘factitious or fictitious disorder by proxy’. The American Academy of Pediatrics Committee on Child Abuse and Neglect has suggested three screening questions to identify FII: Are the history, symptoms and signs of disease credible? Is the child receiving unnecessary and harmful, or potentially harmful, medical care? If so, who is instigating the evaluations and treatment? These questions sound simple, yet identifying FII can be a ‘complex conundrum,’ as demonstrated by a case described in a Lancet editorial, a young boy whose physical examination was normal but, ‘whose medical history includes cerebral palsy, cystic fibrosis, diabetes, food allergies, and intolerance to light.’ It is particularly important to consider FII in chronic illnesses where symptoms are paroxysmal, such as asthma and epilepsy. NICE guidance suggests other factors that should alert us to the possibility of FII, including: Symptoms and signs appear only when the caregiver is present. Reported symptoms are observed solely by caregiver, and may sound biologically implausible. Responses to treatment are inexplicably poor. New symptoms appear as established symptoms resolve. Caregiver(s) seeks multiple medical opinions. The child’s daily and educational activities are disproportionately compromised.
Sahana Rao Andrew L Lux
Abstract Child maltreatment is a significant worldwide problem, with consequences that can include impaired physical and mental health throughout life, and adverse social and occupational effects that carry a heavy economic and social burden. One estimate of cost to the US economy in 2007, for example, was over US$100bn. In middle- and low-income countries, there have been fewer studies of incidence and prevalence, and the economic and social costs are harder to estimate. It is very difficult to obtain full case ascertainment even in the most severe forms of child maltreatment, and even for fatal cases. In general, self-reporting and surveys identify higher rates of child maltreatment than data from sources relating to official notification. This paper reviews the epidemiology of child maltreatment from a practical perspective, with a focus on factors in the social environment, and the clinical history and examination, that predict an increased risk of child maltreatment.
Keywords abusive head trauma; child abuse; childhood mortality; child maltreatment; child neglect; fabricated and induced illness
Introduction and background Child maltreatment is a major public health and social welfare problem in the UK and worldwide, and it remains a challenge over 150 years after Labbe’s first descriptions of typical injuries associated with such abuse. Part of the reason for it remaining a challenge is that the idea of systematic or sporadic child maltreatment, particularly in the child’s own home, has been met with incredulity. Breakthrough reports and studies were published in 1946, with Caffey’s description of long-bone fractures and subdural haematomas in infants, and in 1962, with Kempe and colleagues’ coining of the term ‘Battered-child syndrome.’ However, even 50 years on from Kempe’s description, the nature, causes, frequency and impact of child maltreatment remain controversial.
Incidence and prevalence of child maltreatment Case definitions There are case definitions for a variety of forms of child maltreatment, which is a collective term for the various forms of child abuse and neglect. Standard definitions are available for
The frequency of child maltreatment is challenging to estimate because of variation in definitions, the type of maltreatment being studied, and the comprehensiveness and quality of official statistics and of surveys. Estimates of the frequency of reported cases of maltreatment are more readily obtained from highincome than low- or middle-income countries. The burden of child maltreatment in four high-income countries has been summarised by Gilbert et al. (2009). In Australia in 2002e03, 3.34% of children were referred to official services and 0.68% were considered to have substantiated maltreatment. The Canadian Incidence Study of Reported Child Abuse and Neglect found that 2.15% of children were reported and investigated, with
Sahana Rao MB BS MRCPCH DCH is a Registrar in Paediatrics at Frenchay Hospital, Bristol, UK. Conflicts of interest: none declared. Andrew L Lux MBA MSc PhD FRCPCH is a Consultant in Paediatric Neurology at Bristol Royal Hospital for Children, Bristol, UK. Conflicts of interest: none declared.
PAEDIATRICS AND CHILD HEALTH 22:11
459
Ó 2012 Published by Elsevier Ltd.
SYMPOSIUM: SAFEGUARDING CHILDREN
estimates range between 5% and 35%. Such physical abuse included hitting with a fist or object, biting, kicking, or use or threat of use of a knife or other weapon. For any form of sexual abuse in the UK and US, cumulative prevalence is estimated to be 15e30% for girls and 5e15% for boys, with the respect estimates for penetrative sexual abuse being 5e10% and 1e5% respectively. A meta-analysis of worldwide studies has suggested that cumulative risk of any form of sexual abuse was 25.3% for girls and 8.7% for boys, with rates for penetrative sexual abuse being 5.3% and 1.9%, contact sexual abuse 13.2% and 3.7%, and noncontact sexual abuse 6.8% and 3.1% respectively. For severe emotional abuse cumulative incidence in the UK and US is in the range 4e9%, and for milder forms of psychological abuse, such as being told by parents or carers that a child is not wanted, it is estimated that there is a cumulative incidence of 10.3%.
Definitions of different forms of child maltreatment (based on Gilbert et al. (2009)) Definition Physical abuse
Sexual abuse
Psychological (or emotional) abuse Neglect
Intimate-partner violence
Intentional use of force or implements against a child that results in, or has the potential to result in, physical injury. Any completed or attempted sexual act, sexual contact, or non-contact sexual interaction with a child by a caregiver. Intentional behaviour that conveys to a child that he/she is worthless, flawed, unloved, unwanted, endangered, or valued only in meeting another’s needs. Failure to meet a child’s basic physical, emotional, medical/dental, or educational needs; failure to provide adequate nutrition, hygiene, or shelter; or failure to ensure a child’s safety. Any incident or threatening behaviour, violence, or abuse (psychological, physical, sexual, financial or emotional) between adults who are, or have been, intimate partners or family members, irrespective of sex or sexuality.
Validity and reliability Using the broader concept of ‘victimization‘ e which includes neglect, emotional abuse, theft or vandalism directed against children, assaults by siblings, dating violence and hate crimes e surveys in the United States have found an annual incidence of greater than 50% and self-reported annual incidence of child maltreatment to be as high as 13.6%. This demonstrates the substantially higher case ascertainment associated with selfreporting compared with notification or registration with official child-protection services. However, prospective methods of data collection may detect cases more sensitively than retrospective methods in studies of high-risk children. Using data from the Minnesota Longitudinal Study of Parents and Children, Shaffer et al. prospectively identified child maltreatment in 20.6% of firstborn children whose mothers were mostly of low socioeconomic status, with one-third not completing high school education. In that study, the rates ascertained from self-reporting questionnaires administered in adolescence were lower at 7.1%. Combining the two sources of reporting gave an overall rate of 22.9%. There are particular challenges with establishing the validity and reliability of data-collection instruments in the field of child and adolescent sexual abuse since direct questioning about experiences of sexual abuse is associated with complex methodological, legal and ethical difficulties. One approach to improving case ascertainment and reliability for child maltreatment from official data sources is to use a range of ICD-10 codes for injury rather than codes specific to maltreatment syndromes (such as ICD-10 T74, Y06 and Y07). This has been a trend in using Hospital Episode Statistics in the UK and accords with NICE CG89. The International Society for the Prevention of Child Abuse and Neglect (ISPCAN) has developed Child Abuse Screening Tool instruments that help to address deficiencies in previous screening tools due to cross-cultural, multicultural and multinational differences. Pilot studies have demonstrated these new instruments to have high sensitivity and internal consistency (Table 2).
Table 1
0.47% of the total remaining suspicious and 0.97% considered substantiated. In England, the rate for all social welfare referrals for children and adolescents in 2007 was 4.96%, and the estimated referral rate to social services for maltreatment, excluding neglect and intimate-partner violence, was 1.50%. In the United States, 4.78% of children were investigated for child maltreatment, and 1.21% were considered to have substantiated concerns. Of these, 12% were categorised as being due to multiple forms of maltreatment, and 11% as psychological abuse or unknown. In these four studies, the primary reason for referral was categorised as physical abuse in 10e28% of cases, sexual abuse in 7e10%, psychological abuse in 11e34%, and neglect in 34e60%. A subsequent review of variations in trends and policies relating to child maltreatment in six high-income countries (Australia, Canada, England, New Zealand, Sweden and US) showed that rates of referral to child-protection agencies varied by as much as a factor of 10 but that e after exclusion of the United States, which had a disproportionately high rate of violent child deaths and is a statistical outlier e the differences in maltreatment-related serious injury or violent death varied by less than a factor of two. Since there is greater variability in official notification than in the incidence of serious physical abuse, it is likely that official statistics are prone to a variety of systematic biases. In general, the cumulative incidence of child abuse ascertained by means of self-reporting is substantially higher but also subject to international variation. For physical abuse placing the child or young person at risk of harm e that is, excluding less serious violence such as hitting, slapping and grabbing e cumulative risk
PAEDIATRICS AND CHILD HEALTH 22:11
Deaths related to child maltreatment At the severe end of child maltreatment spectrum are cases in which the maltreatment leads to death. The WHO estimates that
460
Ó 2012 Published by Elsevier Ltd.
SYMPOSIUM: SAFEGUARDING CHILDREN
certificates and data from child welfare agencies, under ascertained cases substantially, and in the case of death certificates, by as much as 90%. However, in all three of these states, but combining two official sources of data, it was possible to ascertain over 90% of cases. Palusci et al. have studies formal capture-recapture methods for ascertaining maltreatment-related child deaths. They found that these methods are reliable for abuse-related deaths but continue to under-ascertain deaths related to neglect.
Examples of data-collection instruments in the field of child maltreatment Data-collection instruments ICAST-C, ICAST-P and ICAST-Ra Juvenile victimisation questionnaire ParenteChild Conflict Tactic Scales Youth Retrospective Questionnaire
Abusive head trauma (AHT)
a
International Child Abuse Screening Tool (child, parent and retrospective versions).
Infants are at particular risk of death from maltreatment due to abusive head trauma (AHT), also known as non-accidental head injury (NAHI) and inflicted traumatic brain injury (TBI). Many such cases fall within the spectrum of injuries described as the ‘shaken baby syndrome’, which is described as having three main elements: a clinical syndromic presentation, distinctive imaging features, and usually absence of a clinical history compatible with the nature and severity of the injuries. Clinical features: Acute and severe encephalopathy Extracranial features of injury, such as fractures, bruises or bites, in 70% of cases Retinal haemorrhages, in 65e90% of cases Imaging features: Intracranial extra-axial haemorrhage, usually subdural but sometimes subarachnoid Diffuse brain parenchymal damage Sometimes skull fractures and head soft-tissue swelling A 1-year survey by the British Paediatric Surveillance Unit (BPSU) found that incidence rates for subdural haemorrhage with or without effusion was 12.5 per 100,000 for children aged 0e2 years, with 7.1 per 100,000 being attributable to NAHI. There was a peak incidence in the middle of the first year, with overall incidence rates for age 0e1 year being 24.1 per 100,000, and 14.2 per 100,000 for cases attributable to NAHI. A Scottish study reported rates of shaken baby syndrome of 24.6 per 100,000 per annum in infants aged under 1 year, with a median incidence at age 2.2 months.
Table 2
there are 31,000 reported homicides per year in children aged less than 15 years, with many such deaths occurring in refugee camps or situations of armed conflict. However, the true number of childhood homicides is likely to be significantly higher, with many cases falsely reported as falls, burns or drowning. In 2003, UNICEF estimated that there were 3500 deaths each year due to child maltreatment in the Organisation for Economic Co-operation and Development (OECD) member states, which represented the world’s 30 wealthiest countries. UNICEF considers poverty and stress to be the factors with strongest association with child maltreatment, with drug and alcohol abuse being other very significant factors lying on a causal pathway that contributes to risk (Table 3). The countries with the lowest death rates from child maltreatment were Spain, Greece, Italy, Ireland and Norway, with rates up to 0.3 per 100,000 children over the 5-year period. The UK rate was 0.4 but rose to 0.9 when including deaths ‘of undetermined intent’. Mexico and the United States had rates that were substantially higher at 2.2. Portugal had an official rate of 0.4 per 100,000 children, but with inclusion of the large number of cases ‘of undetermined intent’ it fell to the worstranked place with a rate of 3.7. In the United States in 2008, there was estimated to be 1740 deaths related to child maltreatment, a rate of 2.3 per 100,000 children. Schnitzer and colleagues used multiple data sources in the states of California, Michigan and Rhode Island and found child maltreatment fatality rates of between 2.5 and 8.8 per 100,000 per annum. They found that official data sources, such as death
Causality, risk factors and predictors of child maltreatment Risk factors in the social environment A Netherlands study by Reijneveld and colleagues found that 5.6% of parents reported physically abusing their infants by the age of 6 months because of crying, with behaviours such as slapping, smothering or shaking. Cumulative rates specifically for shaking rose from 1.01% at age 1 month to 3.35% by age 6 months. Significant predictors for these behaviours (adjusted odds ratios more than 2.0) included a migrant labour family background, single-parent families, other factors relating to family composition (including adoption or non-biological parent in the home), both parents being unemployed, and parents judging that crying was excessive or worrying to them. Patterns of use of healthcare services can also indicate an increased risk, with adjusted odds ratio estimates of more than 2.5 associated with a single change in primary healthcare provider during the previous year, and more than 6.0 (95% CI 2.0e24.1) in cases where there has been two or more changes in primary healthcare provider during the previous year.
Proportion of child deaths in three age groups, and death rate per 100,000 children over a 5-year period, in OECD countries Age group
Proportion of child maltreatment deaths
Death rate (per 100,000 children)
Under 1 year 1e4 years 5e14 years
24% 31% 45%
6.1 1.9 1.1
Table 3
PAEDIATRICS AND CHILD HEALTH 22:11
461
Ó 2012 Published by Elsevier Ltd.
SYMPOSIUM: SAFEGUARDING CHILDREN
Risk factors identified from the medical history There are a number of factors that relate to the ‘role of the child’ in contributing to risks of maltreatment. The Avon Longitudinal Study of Pregnancy and Childhood (ALSPAC), for example, found odds ratios around 2.0 for factors such as unintended pregnancy, low birthweight, general poor health, and developmental problems. Overall, however, maltreatment was not associated with problems such as feeding difficulties, crying or temper tantrums, and it seems likely that parental perceptions and interpretations of childhood behaviour is more important. A study in New York found that child physical (but not sexual) abuse was associated with a history of incomplete immunization at ages 3 and 7 months, with adjusted odds ratios of 4.0 and 4.8 respectively. Other studies have shown that, with lower predictive values, maltreatment is more likely in children with a history of speech or other developmental delay, autism, cerebral palsy, or behavioural or psychological problems.
Factors associated with abusive (AHT) and non-abusive head trauma (nAHT) (Piteau et al.)
Predictors of child abuse identified by clinical examination Features of the clinical examination that predict the cause to be child abuse vary according to the presentation and need to be interpreted in the context of the clinical history. It should be borne in mind, however, that a plausible explanation for a physical injury might still be untrue, and it remains essential to gather other relevant information, such as a history of previous concerns about child maltreatment, before confidently attributing injuries to an accidental cause. A systematic review of burns and scald injuries has shown that intentional injuries tended to have the following characteristics: cause by immersion or a hot tap; location on the limbs, the perineum or buttocks, or both, and often symmetrically; and association with old fractures or injuries related to other events. A systematic review of skeletal fractures has shown that child abuse is a more likely cause where fractures are found in infants (aged less than a year) and toddlers (1e3 years old), and where there are multiple fractures. If major trauma has been excluded, rib fractures are strongly associated with physical abuse, with a pooled estimate of probability of 0.71 (0.42e0.91). Piteau et al. have reported their findings from a systematic review of abusive (AHT) and non-abusive head trauma (nAHT) (see Table 4). Kemp et al. performed a systematic review focussing specifically on neuroimaging features of AHT identified the following associated factors: Multiple subdural haemorrhages over the cerebral convexity Interhemispheric haemorrhages Subdural haemorrhages located in the posterior cranial fossa Neuroimaging features of hypoxic-ischaemic injury Cerebral oedema Factors associated with poorer outcome with AHT include seizures, and the need for endotracheal intubation and ventilation.
Associated variables
Adjusted odds ratiosa (95% CI)
Abusive
Inadequate history Retinal haemorrhage(s) Metaphyseal fracture(s) Subdural haemorrhage(s) Rib fracture(s) Skull fracture with intracranial injury Seizures at presentation or within 24 hours Apnoea at presentation Cerebral ischaemia Any bruising Long-bone fracture(s) Cerebral oedema Any bruisingb Diffuse axonal injury Subarachnoid haemorrhage(s) Vomiting Head and neck bruisingc Extradural (epidural) haemorrhage Scalp swelling Isolated skull fracture(s)
47.0 27.1 11.8 8.9 8.9 7.8
Not associated with trauma type
Non-abusive
(12.9e170.6) (15.7e46.8) (2.2e63.4) (6.8e11.7) (4.0e7.2) (1.1e57.1)
7.3 (3.0e17.3) 5.3 4.8 4.8 4.2 2.2 4.8 1.5 1.4
(2.3e12.1) (1.8e2.5) (1.6e14.1) (2.5e7.2) (1.1e4.5) (1.6e14.1) (0.3e8.3) (0.7e3.0)
0.9 (0.02e4.9) 0.42 (0.19e0.94) 0.15 (0.08e0.29) 0.12 (0.05e0.32) 0.01 (0.003e0.04)
a
ORs for AHT for all 24 studies except where original paper reported OR solely for high-quality studies; OR significantly below 1.0 implies that nAHT is a more likely cause. b The variable ‘Any bruising’ appeared to be mildly predictive of AHT after analysis of all the included studies, but it was not significant when including data solely from the studies regarded as being of higher methodological quality: OR 5.4 (0.9e31.4). c The variable ‘Head and neck bruising’ appeared to be associated with nAHT in all studies combined, but analysis of data solely from studies with higher methodological quality did not show a significant difference: OR 0.7 (0.3e1.6).
Table 4
emotional burden carried by the victims of child maltreatment, which is very difficult to quantify but can also have serious intergenerational effects. There are short- and long-term medical costs, future loss of earnings and productivity to the economy, child welfare costs, special education, and legal costs. A systematic literature review identified no studies using instruments to report quality-of-life (QOL) in children, but there were four studies that reported QOL outcomes in adult survivors of child maltreatment. Some indirect research on QOL losses have been based on jury awards, but these are likely to be inflated and biased upwards. Corso and colleagues challenge the estimated costs to the US economy in 2007, their estimate being revised downward from $103.8bn to $65.1bn. Mikton and Butchart conducted a ‘systematic review of reviews’ of prevention and intervention that examined factors such as home visits, parent education, strategies for child sexual
Impact and interventions The behavioural consequences in adulthood of child maltreatment include excessive risk-taking, criminal activity, a proneness to victimisation, and further perpetration of violence upon others. In addition to these costs, there is the enormous
PAEDIATRICS AND CHILD HEALTH 22:11
Type of trauma
462
Ó 2012 Published by Elsevier Ltd.
SYMPOSIUM: SAFEGUARDING CHILDREN
Gilbert R, Fluke J, O’Donnell M, et al. Child maltreatment: variation in trends and policies in six developed countries. Lancet 2012; 379: 758e72. Gonzalez-Izquierdo A, Woodman J, Copley L, et al. Variation in recording of child maltreatment in administrative records of hospital admissions for injury in England, 1997e2009. Arch Dis Child 2010; 95: 918e25. Greeley CS. The evolution of the child maltreatment literature. Pediatrics 2012; 130: 347e8. Greiner MV, Lawrence AP, Horn P, Newmeyer AJ, Makoroff KL. Early clinical indicators of developmental outcome in abusive head trauma. Childs Nerv Syst 2012; 28: 889e96. Hobbs C, Childs AM, Wynne J, Livingston J, Seal A. Subdural haematoma and effusion in infancy: an epidemiological study. Arch Dis Child 2005; 90: 952e5. Kemp AM, Dunstan F, Harrison S, et al. Patterns of skeletal fractures in child abuse: systematic review. BMJ 2008; 337: a1518. Kemp AM, Jaspan T, Griffiths J, et al. Neuroimaging: what neuroradiological features distinguish abusive from non-abusive head trauma? A systematic review. Arch Dis Child 2011; 96: 1103e12. Kempe CH, Silverman FN, Droegmueller W, Silver HK. The battered-child syndrome. JAMA 1962; 181: 17e24. Labbe J. Ambroise Tardieu: the man and his work on child maltreatment a century before Kempe. Child Abuse Negl 2005; 29: 311e24. Leeb RT, Paulozzi L, Melanson C, Simon T, Arias I. Child maltreatment surveillance. Uniform definitions for public health and recommended data elements. Atlanta, GA: Centers for Disease Control, 2008. Livingston JH, Childs A-M. The epidemiology of non-accidental head injury. In: Minns RA, Brown JK, eds. Shaking and other non-accidental injuries in children. London: MacKeith Press, 2005; 147e153. Macmillan HL, Wathen CN, Barlow J, Fergusson DM, Leventhal JM, Taussig HN. Interventions to prevent child maltreatment and associated impairment. Lancet 2009; 373: 250e66. Maguire S, Moynihan S, Mann M, Potokar T, Kemp AM. A systematic review of the features that indicate intentional scalds in children. Burns 2008; 34: 1072e81. Mikton C, Butchart A. Child maltreatment prevention: a systematic review of reviews. Bull World Health Organ 2009; 87: 353e61. Palusci VJ, Wirtz SJ, Covington TM. Using capture-recapture methods to better ascertain the incidence of fatal child maltreatment. Child Abuse Negl 2010; 34: 396e402. Piteau SJ, Ward MG, Barrowman NJ, Plint AC. Clinical and radiographic characteristics associated with abusive and nonabusive head trauma: a systematic review. Pediatrics 2012; 130: 315e23. Prosser LA, Corso PS. Measuring health-related quality of life for child maltreatment: a systematic literature review. Health Qual Life Outcomes 2007; 5: 42. Reijneveld SA, van der Wal MF, Brugman E, Sing RA, VerlooveVanhorick SP. Infant crying and abuse. Lancet 2004; 364: 1340e2. Runyan DK, Dunne MP, Zolotor AJ. Introduction to the development of the ISPCAN child abuse screening tools. Child Abuse Negl 2009; 33: 842e5. Schnitzer PG, Covington TM, Wirtz SJ, Verhoek-Oftedahl W, Palusci VJ. Public health surveillance of fatal child maltreatment: analysis of 3 state programs. Am J Public Health 2008; 98: 296e303. Scott D, Tonmyr L, Fraser J, Walker S, McKenzie K. The utility and challenges of using ICD codes in child maltreatment research: a review of existing literature. Child Abuse Negl 2009; 33: 791e808.
abuse prevention, abusive head trauma prevention, multicomponent interventions, media-based interventions, and support and mutual aid groups. This found that only 0.6% of published evidence related to low- and middle-income countries. In the UK, two home-visiting programmes e the Nurse-Family Partnership and Early Start e have been shown to reduce child maltreatment and associated physical injuries.
Summary and conclusions Child maltreatment remains a significant global problem in all of its many forms. This review has briefly described the problems associated with reliably estimating its incidence and prevalence, and some of the features from the social environment, and the clinical history and examination, that should alert the paediatrician and other healthcare workers to the possibility of maltreatment contributing to the child’s presentation. This is particularly challenging in the case of fabricated or induced illness because in that circumstance the perpetrator will take significant steps to conceal or distract. There remains substantial work to be done on identifying the most effective means of preventing, detecting and effectively intervening with child maltreatment at individual, family and population levels. A
FURTHER READING Andrews G, Corry J, Slade T, Issakadis C, Swanston H. Child sexual abuse. In: Comparative quantification of health risks: global and regional burden of disease attributable to selected major risk factors, vol. 2. Geneva: World Health Organization, 2004. 1851e1940. Barlow KM, Minns RA. Annual incidence of shaken impact syndrome in young children. Lancet 2000; 356: 1571e2. Barber MA, Davis PM. Fits, faints, or fatal fantasy? Fabricated seizures and child abuse. Arch Dis Child 2002; 86: 230e3. Caffey J. Multiple fractures in the long bones of infants suffering from chronic subdural hematoma. Am J Roentgenol 1946; 56: 163e73. Centers for Disease Control. Child maltreatment: facts at a glance. Available from: http://www.cdc.gov/ViolencePrevention/pdf/cm-datasheeta.pdf; 2010 (accessed 12 August 2012). Corso PS, Fertig AR. The economic impact of child maltreatment in the United States: are the estimates credible? Child Abuse Negl 2010; 34: 296e304. DCSF. Referrals, assessments and children and young people who are the subject of a child protection plan or are on child protection registers: year ending 31 March 2007. London: n.p., 2008. Print. Editorial. Fabricated or induced illness by carers: a complex conundrum. Lancet 2010; 375: 433. Finkelhor D, Ormrod R, Turner H, Hamby SL. The victimization of children and youth: a comprehensive, national survey. Child Maltreat 2005; 10: 5e25. Friedlaender EY, Rubin DM, Alpern ER, Mandell DS, Christian CW, Alessandrini EA. Patterns of health care use that may identify young children who are at risk for maltreatment. Pediatrics 2005; 116: 1303e8. Gilbert R, Widom CS, Browne K, Fergusson D, Webb E, Janson S. Burden and consequences of child maltreatment in high-income countries. Lancet 2009; 373: 68e81.
PAEDIATRICS AND CHILD HEALTH 22:11
463
Ó 2012 Published by Elsevier Ltd.
SYMPOSIUM: SAFEGUARDING CHILDREN
Shaffer A, Huston L, Egeland B. Identification of child maltreatment using prospective and self-report methodologies: a comparison of maltreatment incidence and relation to later psychopathology. Child Abuse Negl 2008; 32: 682e92. Sidebotham P, Heron J, ALSPAC Study Team. Child maltreatment in the “Children of the Nineties:” the role of the child. Child Abuse Negl 2003; 27: 337e52. Stirling J, American Academy of Pediatrics Committee on Child Abuse and Neglect. Beyond Munchausen syndrome by proxy: identification and treatment of child abuse in a medical setting. Pediatrics 2007; 119: 1026e30. Stockwell MS, Brown J, Chen S, Vaughan RD, Irigoyen M. Is underimmunization associated with child maltreatment? Ambul Pediatr 2008; 8: 210e3. Trocme N, MacMillan H, Fallon B, De Marco R. Nature and severity of physical harm caused by child abuse and neglect: results from the Canadian incidence study. CMAJ 2003; 169: 911e5. UNICEF. A league table of child maltreatment deaths in rich nations. Florence: UNICEF Innocenti Research Centre, 2003. US DHHS. Child maltreatment 2006. Washington, DC: US Government Printing Office, 2008. Print. Walsh C, Jamieson E, MacMillan H, Trocme N. Measuring child sexual abuse in children and youth. J Child Sex Abus 2004; 13: 39e68. World Health Organization. Child maltreatment: fact sheet number 150; Available from: http://www.who.int/mediacentre/factsheets/fs150/en/ index.html (accessed 12 August 2012). Wilson RG. Fabricated or induced illness in children. Munchausen by proxy comes of age. BMJ 2001; 323: 296e7. Woodman J, Brandon M, Bailey S, Belderson P, Sidebotham P, Gilbert R. Healthcare use by children fatally or seriously harmed by child maltreatment: analysis of a national case series 2005e2007. Arch Dis Child 2011; 96: 270e5.
PAEDIATRICS AND CHILD HEALTH 22:11
Practice points C
C
C
C
C
C
464
Child maltreatment is a significant worldwide problem, with consequences that can include impaired physical and mental health throughout life, and adverse social and occupational effects that carry a heavy economic and social burden. A conservative estimate of the cost to the US economy in 2007 was approximately US$65bn, which was a challenge to an earlier estimate of over US$100bn. In middle- and low-income countries, the economic and social costs are harder to estimate but are extremely important. Epidemiological studies have tended to come from highincome countries, but it is estimated in broad terms that 20% of women and 5e10% of men report being sexually abused in childhood, and 25e50% of children report being the victims of physical abuse. Severe forms of physical abuse are associated with significant childhood mortality, most commonly in infants and toddlers and often due to the ‘shaken baby syndrome’, which is strongly associated with an inadequate or implausible history, retinal haemorrhage(s), metaphyseal fracture(s) and subdural haematoma(s). One of the most challenging forms of child abuse to diagnose and manage is fabricated or induced illness (FII) since the perpetrator(s) will make all efforts to conceal and distract healthcare professionals from the information required to make the diagnosis. The World Health Organization (WHO) recommends a ‘multisectoral’ approach to preventing child maltreatment, with the most effective intervention likely to be parental support and teaching positive parenting skills.
Ó 2012 Published by Elsevier Ltd.
The epidemiology of child maltreatment
public health and surveillance from organisation such as the Centers for Disease Control. ‘Child maltreatment’ is defined as any act of commission or omission by a parent or other caregiver that results in harm, potential for harm, or threat of harm to a child.’ There is no requirement for this harm to be intended, and most cases of harm to children are perpetrated by parents or caregivers; for example, 82% of substantiated cases in one study from the United States. Other definitions and examples are shown in Table 1. The UK definition of psychological or emotional abuse is broad and includes interactions between child and parent or carer in which there are unintentional interactions with these adverse effects, as encompassed in the phrase: ‘The persistent emotional ill-treatment of a child such as to cause severe and persistent adverse effects on the child’s emotional development.’ Fabricated or induced illness (FII) is an established and currently preferred term in the UK for the condition that is also known as ‘Munchausen syndrome by proxy’, ‘paediatric symptom falsification’ or ‘factitious or fictitious disorder by proxy’. The American Academy of Pediatrics Committee on Child Abuse and Neglect has suggested three screening questions to identify FII: Are the history, symptoms and signs of disease credible? Is the child receiving unnecessary and harmful, or potentially harmful, medical care? If so, who is instigating the evaluations and treatment? These questions sound simple, yet identifying FII can be a ‘complex conundrum,’ as demonstrated by a case described in a Lancet editorial, a young boy whose physical examination was normal but, ‘whose medical history includes cerebral palsy, cystic fibrosis, diabetes, food allergies, and intolerance to light.’ It is particularly important to consider FII in chronic illnesses where symptoms are paroxysmal, such as asthma and epilepsy. NICE guidance suggests other factors that should alert us to the possibility of FII, including: Symptoms and signs appear only when the caregiver is present. Reported symptoms are observed solely by caregiver, and may sound biologically implausible. Responses to treatment are inexplicably poor. New symptoms appear as established symptoms resolve. Caregiver(s) seeks multiple medical opinions. The child’s daily and educational activities are disproportionately compromised.
Sahana Rao Andrew L Lux
Abstract Child maltreatment is a significant worldwide problem, with consequences that can include impaired physical and mental health throughout life, and adverse social and occupational effects that carry a heavy economic and social burden. One estimate of cost to the US economy in 2007, for example, was over US$100bn. In middle- and low-income countries, there have been fewer studies of incidence and prevalence, and the economic and social costs are harder to estimate. It is very difficult to obtain full case ascertainment even in the most severe forms of child maltreatment, and even for fatal cases. In general, self-reporting and surveys identify higher rates of child maltreatment than data from sources relating to official notification. This paper reviews the epidemiology of child maltreatment from a practical perspective, with a focus on factors in the social environment, and the clinical history and examination, that predict an increased risk of child maltreatment.
Keywords abusive head trauma; child abuse; childhood mortality; child maltreatment; child neglect; fabricated and induced illness
Introduction and background Child maltreatment is a major public health and social welfare problem in the UK and worldwide, and it remains a challenge over 150 years after Labbe’s first descriptions of typical injuries associated with such abuse. Part of the reason for it remaining a challenge is that the idea of systematic or sporadic child maltreatment, particularly in the child’s own home, has been met with incredulity. Breakthrough reports and studies were published in 1946, with Caffey’s description of long-bone fractures and subdural haematomas in infants, and in 1962, with Kempe and colleagues’ coining of the term ‘Battered-child syndrome.’ However, even 50 years on from Kempe’s description, the nature, causes, frequency and impact of child maltreatment remain controversial.
Incidence and prevalence of child maltreatment Case definitions There are case definitions for a variety of forms of child maltreatment, which is a collective term for the various forms of child abuse and neglect. Standard definitions are available for
The frequency of child maltreatment is challenging to estimate because of variation in definitions, the type of maltreatment being studied, and the comprehensiveness and quality of official statistics and of surveys. Estimates of the frequency of reported cases of maltreatment are more readily obtained from highincome than low- or middle-income countries. The burden of child maltreatment in four high-income countries has been summarised by Gilbert et al. (2009). In Australia in 2002e03, 3.34% of children were referred to official services and 0.68% were considered to have substantiated maltreatment. The Canadian Incidence Study of Reported Child Abuse and Neglect found that 2.15% of children were reported and investigated, with
Sahana Rao MB BS MRCPCH DCH is a Registrar in Paediatrics at Frenchay Hospital, Bristol, UK. Conflicts of interest: none declared. Andrew L Lux MBA MSc PhD FRCPCH is a Consultant in Paediatric Neurology at Bristol Royal Hospital for Children, Bristol, UK. Conflicts of interest: none declared.
PAEDIATRICS AND CHILD HEALTH 22:11
459
Ó 2012 Published by Elsevier Ltd.
SYMPOSIUM: SAFEGUARDING CHILDREN
estimates range between 5% and 35%. Such physical abuse included hitting with a fist or object, biting, kicking, or use or threat of use of a knife or other weapon. For any form of sexual abuse in the UK and US, cumulative prevalence is estimated to be 15e30% for girls and 5e15% for boys, with the respect estimates for penetrative sexual abuse being 5e10% and 1e5% respectively. A meta-analysis of worldwide studies has suggested that cumulative risk of any form of sexual abuse was 25.3% for girls and 8.7% for boys, with rates for penetrative sexual abuse being 5.3% and 1.9%, contact sexual abuse 13.2% and 3.7%, and noncontact sexual abuse 6.8% and 3.1% respectively. For severe emotional abuse cumulative incidence in the UK and US is in the range 4e9%, and for milder forms of psychological abuse, such as being told by parents or carers that a child is not wanted, it is estimated that there is a cumulative incidence of 10.3%.
Definitions of different forms of child maltreatment (based on Gilbert et al. (2009)) Definition Physical abuse
Sexual abuse
Psychological (or emotional) abuse Neglect
Intimate-partner violence
Intentional use of force or implements against a child that results in, or has the potential to result in, physical injury. Any completed or attempted sexual act, sexual contact, or non-contact sexual interaction with a child by a caregiver. Intentional behaviour that conveys to a child that he/she is worthless, flawed, unloved, unwanted, endangered, or valued only in meeting another’s needs. Failure to meet a child’s basic physical, emotional, medical/dental, or educational needs; failure to provide adequate nutrition, hygiene, or shelter; or failure to ensure a child’s safety. Any incident or threatening behaviour, violence, or abuse (psychological, physical, sexual, financial or emotional) between adults who are, or have been, intimate partners or family members, irrespective of sex or sexuality.
Validity and reliability Using the broader concept of ‘victimization‘ e which includes neglect, emotional abuse, theft or vandalism directed against children, assaults by siblings, dating violence and hate crimes e surveys in the United States have found an annual incidence of greater than 50% and self-reported annual incidence of child maltreatment to be as high as 13.6%. This demonstrates the substantially higher case ascertainment associated with selfreporting compared with notification or registration with official child-protection services. However, prospective methods of data collection may detect cases more sensitively than retrospective methods in studies of high-risk children. Using data from the Minnesota Longitudinal Study of Parents and Children, Shaffer et al. prospectively identified child maltreatment in 20.6% of firstborn children whose mothers were mostly of low socioeconomic status, with one-third not completing high school education. In that study, the rates ascertained from self-reporting questionnaires administered in adolescence were lower at 7.1%. Combining the two sources of reporting gave an overall rate of 22.9%. There are particular challenges with establishing the validity and reliability of data-collection instruments in the field of child and adolescent sexual abuse since direct questioning about experiences of sexual abuse is associated with complex methodological, legal and ethical difficulties. One approach to improving case ascertainment and reliability for child maltreatment from official data sources is to use a range of ICD-10 codes for injury rather than codes specific to maltreatment syndromes (such as ICD-10 T74, Y06 and Y07). This has been a trend in using Hospital Episode Statistics in the UK and accords with NICE CG89. The International Society for the Prevention of Child Abuse and Neglect (ISPCAN) has developed Child Abuse Screening Tool instruments that help to address deficiencies in previous screening tools due to cross-cultural, multicultural and multinational differences. Pilot studies have demonstrated these new instruments to have high sensitivity and internal consistency (Table 2).
Table 1
0.47% of the total remaining suspicious and 0.97% considered substantiated. In England, the rate for all social welfare referrals for children and adolescents in 2007 was 4.96%, and the estimated referral rate to social services for maltreatment, excluding neglect and intimate-partner violence, was 1.50%. In the United States, 4.78% of children were investigated for child maltreatment, and 1.21% were considered to have substantiated concerns. Of these, 12% were categorised as being due to multiple forms of maltreatment, and 11% as psychological abuse or unknown. In these four studies, the primary reason for referral was categorised as physical abuse in 10e28% of cases, sexual abuse in 7e10%, psychological abuse in 11e34%, and neglect in 34e60%. A subsequent review of variations in trends and policies relating to child maltreatment in six high-income countries (Australia, Canada, England, New Zealand, Sweden and US) showed that rates of referral to child-protection agencies varied by as much as a factor of 10 but that e after exclusion of the United States, which had a disproportionately high rate of violent child deaths and is a statistical outlier e the differences in maltreatment-related serious injury or violent death varied by less than a factor of two. Since there is greater variability in official notification than in the incidence of serious physical abuse, it is likely that official statistics are prone to a variety of systematic biases. In general, the cumulative incidence of child abuse ascertained by means of self-reporting is substantially higher but also subject to international variation. For physical abuse placing the child or young person at risk of harm e that is, excluding less serious violence such as hitting, slapping and grabbing e cumulative risk
PAEDIATRICS AND CHILD HEALTH 22:11
Deaths related to child maltreatment At the severe end of child maltreatment spectrum are cases in which the maltreatment leads to death. The WHO estimates that
460
Ó 2012 Published by Elsevier Ltd.
SYMPOSIUM: SAFEGUARDING CHILDREN
certificates and data from child welfare agencies, under ascertained cases substantially, and in the case of death certificates, by as much as 90%. However, in all three of these states, but combining two official sources of data, it was possible to ascertain over 90% of cases. Palusci et al. have studies formal capture-recapture methods for ascertaining maltreatment-related child deaths. They found that these methods are reliable for abuse-related deaths but continue to under-ascertain deaths related to neglect.
Examples of data-collection instruments in the field of child maltreatment Data-collection instruments ICAST-C, ICAST-P and ICAST-Ra Juvenile victimisation questionnaire ParenteChild Conflict Tactic Scales Youth Retrospective Questionnaire
Abusive head trauma (AHT)
a
International Child Abuse Screening Tool (child, parent and retrospective versions).
Infants are at particular risk of death from maltreatment due to abusive head trauma (AHT), also known as non-accidental head injury (NAHI) and inflicted traumatic brain injury (TBI). Many such cases fall within the spectrum of injuries described as the ‘shaken baby syndrome’, which is described as having three main elements: a clinical syndromic presentation, distinctive imaging features, and usually absence of a clinical history compatible with the nature and severity of the injuries. Clinical features: Acute and severe encephalopathy Extracranial features of injury, such as fractures, bruises or bites, in 70% of cases Retinal haemorrhages, in 65e90% of cases Imaging features: Intracranial extra-axial haemorrhage, usually subdural but sometimes subarachnoid Diffuse brain parenchymal damage Sometimes skull fractures and head soft-tissue swelling A 1-year survey by the British Paediatric Surveillance Unit (BPSU) found that incidence rates for subdural haemorrhage with or without effusion was 12.5 per 100,000 for children aged 0e2 years, with 7.1 per 100,000 being attributable to NAHI. There was a peak incidence in the middle of the first year, with overall incidence rates for age 0e1 year being 24.1 per 100,000, and 14.2 per 100,000 for cases attributable to NAHI. A Scottish study reported rates of shaken baby syndrome of 24.6 per 100,000 per annum in infants aged under 1 year, with a median incidence at age 2.2 months.
Table 2
there are 31,000 reported homicides per year in children aged less than 15 years, with many such deaths occurring in refugee camps or situations of armed conflict. However, the true number of childhood homicides is likely to be significantly higher, with many cases falsely reported as falls, burns or drowning. In 2003, UNICEF estimated that there were 3500 deaths each year due to child maltreatment in the Organisation for Economic Co-operation and Development (OECD) member states, which represented the world’s 30 wealthiest countries. UNICEF considers poverty and stress to be the factors with strongest association with child maltreatment, with drug and alcohol abuse being other very significant factors lying on a causal pathway that contributes to risk (Table 3). The countries with the lowest death rates from child maltreatment were Spain, Greece, Italy, Ireland and Norway, with rates up to 0.3 per 100,000 children over the 5-year period. The UK rate was 0.4 but rose to 0.9 when including deaths ‘of undetermined intent’. Mexico and the United States had rates that were substantially higher at 2.2. Portugal had an official rate of 0.4 per 100,000 children, but with inclusion of the large number of cases ‘of undetermined intent’ it fell to the worstranked place with a rate of 3.7. In the United States in 2008, there was estimated to be 1740 deaths related to child maltreatment, a rate of 2.3 per 100,000 children. Schnitzer and colleagues used multiple data sources in the states of California, Michigan and Rhode Island and found child maltreatment fatality rates of between 2.5 and 8.8 per 100,000 per annum. They found that official data sources, such as death
Causality, risk factors and predictors of child maltreatment Risk factors in the social environment A Netherlands study by Reijneveld and colleagues found that 5.6% of parents reported physically abusing their infants by the age of 6 months because of crying, with behaviours such as slapping, smothering or shaking. Cumulative rates specifically for shaking rose from 1.01% at age 1 month to 3.35% by age 6 months. Significant predictors for these behaviours (adjusted odds ratios more than 2.0) included a migrant labour family background, single-parent families, other factors relating to family composition (including adoption or non-biological parent in the home), both parents being unemployed, and parents judging that crying was excessive or worrying to them. Patterns of use of healthcare services can also indicate an increased risk, with adjusted odds ratio estimates of more than 2.5 associated with a single change in primary healthcare provider during the previous year, and more than 6.0 (95% CI 2.0e24.1) in cases where there has been two or more changes in primary healthcare provider during the previous year.
Proportion of child deaths in three age groups, and death rate per 100,000 children over a 5-year period, in OECD countries Age group
Proportion of child maltreatment deaths
Death rate (per 100,000 children)
Under 1 year 1e4 years 5e14 years
24% 31% 45%
6.1 1.9 1.1
Table 3
PAEDIATRICS AND CHILD HEALTH 22:11
461
Ó 2012 Published by Elsevier Ltd.
SYMPOSIUM: SAFEGUARDING CHILDREN
Risk factors identified from the medical history There are a number of factors that relate to the ‘role of the child’ in contributing to risks of maltreatment. The Avon Longitudinal Study of Pregnancy and Childhood (ALSPAC), for example, found odds ratios around 2.0 for factors such as unintended pregnancy, low birthweight, general poor health, and developmental problems. Overall, however, maltreatment was not associated with problems such as feeding difficulties, crying or temper tantrums, and it seems likely that parental perceptions and interpretations of childhood behaviour is more important. A study in New York found that child physical (but not sexual) abuse was associated with a history of incomplete immunization at ages 3 and 7 months, with adjusted odds ratios of 4.0 and 4.8 respectively. Other studies have shown that, with lower predictive values, maltreatment is more likely in children with a history of speech or other developmental delay, autism, cerebral palsy, or behavioural or psychological problems.
Factors associated with abusive (AHT) and non-abusive head trauma (nAHT) (Piteau et al.)
Predictors of child abuse identified by clinical examination Features of the clinical examination that predict the cause to be child abuse vary according to the presentation and need to be interpreted in the context of the clinical history. It should be borne in mind, however, that a plausible explanation for a physical injury might still be untrue, and it remains essential to gather other relevant information, such as a history of previous concerns about child maltreatment, before confidently attributing injuries to an accidental cause. A systematic review of burns and scald injuries has shown that intentional injuries tended to have the following characteristics: cause by immersion or a hot tap; location on the limbs, the perineum or buttocks, or both, and often symmetrically; and association with old fractures or injuries related to other events. A systematic review of skeletal fractures has shown that child abuse is a more likely cause where fractures are found in infants (aged less than a year) and toddlers (1e3 years old), and where there are multiple fractures. If major trauma has been excluded, rib fractures are strongly associated with physical abuse, with a pooled estimate of probability of 0.71 (0.42e0.91). Piteau et al. have reported their findings from a systematic review of abusive (AHT) and non-abusive head trauma (nAHT) (see Table 4). Kemp et al. performed a systematic review focussing specifically on neuroimaging features of AHT identified the following associated factors: Multiple subdural haemorrhages over the cerebral convexity Interhemispheric haemorrhages Subdural haemorrhages located in the posterior cranial fossa Neuroimaging features of hypoxic-ischaemic injury Cerebral oedema Factors associated with poorer outcome with AHT include seizures, and the need for endotracheal intubation and ventilation.
Associated variables
Adjusted odds ratiosa (95% CI)
Abusive
Inadequate history Retinal haemorrhage(s) Metaphyseal fracture(s) Subdural haemorrhage(s) Rib fracture(s) Skull fracture with intracranial injury Seizures at presentation or within 24 hours Apnoea at presentation Cerebral ischaemia Any bruising Long-bone fracture(s) Cerebral oedema Any bruisingb Diffuse axonal injury Subarachnoid haemorrhage(s) Vomiting Head and neck bruisingc Extradural (epidural) haemorrhage Scalp swelling Isolated skull fracture(s)
47.0 27.1 11.8 8.9 8.9 7.8
Not associated with trauma type
Non-abusive
(12.9e170.6) (15.7e46.8) (2.2e63.4) (6.8e11.7) (4.0e7.2) (1.1e57.1)
7.3 (3.0e17.3) 5.3 4.8 4.8 4.2 2.2 4.8 1.5 1.4
(2.3e12.1) (1.8e2.5) (1.6e14.1) (2.5e7.2) (1.1e4.5) (1.6e14.1) (0.3e8.3) (0.7e3.0)
0.9 (0.02e4.9) 0.42 (0.19e0.94) 0.15 (0.08e0.29) 0.12 (0.05e0.32) 0.01 (0.003e0.04)
a
ORs for AHT for all 24 studies except where original paper reported OR solely for high-quality studies; OR significantly below 1.0 implies that nAHT is a more likely cause. b The variable ‘Any bruising’ appeared to be mildly predictive of AHT after analysis of all the included studies, but it was not significant when including data solely from the studies regarded as being of higher methodological quality: OR 5.4 (0.9e31.4). c The variable ‘Head and neck bruising’ appeared to be associated with nAHT in all studies combined, but analysis of data solely from studies with higher methodological quality did not show a significant difference: OR 0.7 (0.3e1.6).
Table 4
emotional burden carried by the victims of child maltreatment, which is very difficult to quantify but can also have serious intergenerational effects. There are short- and long-term medical costs, future loss of earnings and productivity to the economy, child welfare costs, special education, and legal costs. A systematic literature review identified no studies using instruments to report quality-of-life (QOL) in children, but there were four studies that reported QOL outcomes in adult survivors of child maltreatment. Some indirect research on QOL losses have been based on jury awards, but these are likely to be inflated and biased upwards. Corso and colleagues challenge the estimated costs to the US economy in 2007, their estimate being revised downward from $103.8bn to $65.1bn. Mikton and Butchart conducted a ‘systematic review of reviews’ of prevention and intervention that examined factors such as home visits, parent education, strategies for child sexual
Impact and interventions The behavioural consequences in adulthood of child maltreatment include excessive risk-taking, criminal activity, a proneness to victimisation, and further perpetration of violence upon others. In addition to these costs, there is the enormous
PAEDIATRICS AND CHILD HEALTH 22:11
Type of trauma
462
Ó 2012 Published by Elsevier Ltd.
SYMPOSIUM: SAFEGUARDING CHILDREN
Gilbert R, Fluke J, O’Donnell M, et al. Child maltreatment: variation in trends and policies in six developed countries. Lancet 2012; 379: 758e72. Gonzalez-Izquierdo A, Woodman J, Copley L, et al. Variation in recording of child maltreatment in administrative records of hospital admissions for injury in England, 1997e2009. Arch Dis Child 2010; 95: 918e25. Greeley CS. The evolution of the child maltreatment literature. Pediatrics 2012; 130: 347e8. Greiner MV, Lawrence AP, Horn P, Newmeyer AJ, Makoroff KL. Early clinical indicators of developmental outcome in abusive head trauma. Childs Nerv Syst 2012; 28: 889e96. Hobbs C, Childs AM, Wynne J, Livingston J, Seal A. Subdural haematoma and effusion in infancy: an epidemiological study. Arch Dis Child 2005; 90: 952e5. Kemp AM, Dunstan F, Harrison S, et al. Patterns of skeletal fractures in child abuse: systematic review. BMJ 2008; 337: a1518. Kemp AM, Jaspan T, Griffiths J, et al. Neuroimaging: what neuroradiological features distinguish abusive from non-abusive head trauma? A systematic review. Arch Dis Child 2011; 96: 1103e12. Kempe CH, Silverman FN, Droegmueller W, Silver HK. The battered-child syndrome. JAMA 1962; 181: 17e24. Labbe J. Ambroise Tardieu: the man and his work on child maltreatment a century before Kempe. Child Abuse Negl 2005; 29: 311e24. Leeb RT, Paulozzi L, Melanson C, Simon T, Arias I. Child maltreatment surveillance. Uniform definitions for public health and recommended data elements. Atlanta, GA: Centers for Disease Control, 2008. Livingston JH, Childs A-M. The epidemiology of non-accidental head injury. In: Minns RA, Brown JK, eds. Shaking and other non-accidental injuries in children. London: MacKeith Press, 2005; 147e153. Macmillan HL, Wathen CN, Barlow J, Fergusson DM, Leventhal JM, Taussig HN. Interventions to prevent child maltreatment and associated impairment. Lancet 2009; 373: 250e66. Maguire S, Moynihan S, Mann M, Potokar T, Kemp AM. A systematic review of the features that indicate intentional scalds in children. Burns 2008; 34: 1072e81. Mikton C, Butchart A. Child maltreatment prevention: a systematic review of reviews. Bull World Health Organ 2009; 87: 353e61. Palusci VJ, Wirtz SJ, Covington TM. Using capture-recapture methods to better ascertain the incidence of fatal child maltreatment. Child Abuse Negl 2010; 34: 396e402. Piteau SJ, Ward MG, Barrowman NJ, Plint AC. Clinical and radiographic characteristics associated with abusive and nonabusive head trauma: a systematic review. Pediatrics 2012; 130: 315e23. Prosser LA, Corso PS. Measuring health-related quality of life for child maltreatment: a systematic literature review. Health Qual Life Outcomes 2007; 5: 42. Reijneveld SA, van der Wal MF, Brugman E, Sing RA, VerlooveVanhorick SP. Infant crying and abuse. Lancet 2004; 364: 1340e2. Runyan DK, Dunne MP, Zolotor AJ. Introduction to the development of the ISPCAN child abuse screening tools. Child Abuse Negl 2009; 33: 842e5. Schnitzer PG, Covington TM, Wirtz SJ, Verhoek-Oftedahl W, Palusci VJ. Public health surveillance of fatal child maltreatment: analysis of 3 state programs. Am J Public Health 2008; 98: 296e303. Scott D, Tonmyr L, Fraser J, Walker S, McKenzie K. The utility and challenges of using ICD codes in child maltreatment research: a review of existing literature. Child Abuse Negl 2009; 33: 791e808.
abuse prevention, abusive head trauma prevention, multicomponent interventions, media-based interventions, and support and mutual aid groups. This found that only 0.6% of published evidence related to low- and middle-income countries. In the UK, two home-visiting programmes e the Nurse-Family Partnership and Early Start e have been shown to reduce child maltreatment and associated physical injuries.
Summary and conclusions Child maltreatment remains a significant global problem in all of its many forms. This review has briefly described the problems associated with reliably estimating its incidence and prevalence, and some of the features from the social environment, and the clinical history and examination, that should alert the paediatrician and other healthcare workers to the possibility of maltreatment contributing to the child’s presentation. This is particularly challenging in the case of fabricated or induced illness because in that circumstance the perpetrator will take significant steps to conceal or distract. There remains substantial work to be done on identifying the most effective means of preventing, detecting and effectively intervening with child maltreatment at individual, family and population levels. A
FURTHER READING Andrews G, Corry J, Slade T, Issakadis C, Swanston H. Child sexual abuse. In: Comparative quantification of health risks: global and regional burden of disease attributable to selected major risk factors, vol. 2. Geneva: World Health Organization, 2004. 1851e1940. Barlow KM, Minns RA. Annual incidence of shaken impact syndrome in young children. Lancet 2000; 356: 1571e2. Barber MA, Davis PM. Fits, faints, or fatal fantasy? Fabricated seizures and child abuse. Arch Dis Child 2002; 86: 230e3. Caffey J. Multiple fractures in the long bones of infants suffering from chronic subdural hematoma. Am J Roentgenol 1946; 56: 163e73. Centers for Disease Control. Child maltreatment: facts at a glance. Available from: http://www.cdc.gov/ViolencePrevention/pdf/cm-datasheeta.pdf; 2010 (accessed 12 August 2012). Corso PS, Fertig AR. The economic impact of child maltreatment in the United States: are the estimates credible? Child Abuse Negl 2010; 34: 296e304. DCSF. Referrals, assessments and children and young people who are the subject of a child protection plan or are on child protection registers: year ending 31 March 2007. London: n.p., 2008. Print. Editorial. Fabricated or induced illness by carers: a complex conundrum. Lancet 2010; 375: 433. Finkelhor D, Ormrod R, Turner H, Hamby SL. The victimization of children and youth: a comprehensive, national survey. Child Maltreat 2005; 10: 5e25. Friedlaender EY, Rubin DM, Alpern ER, Mandell DS, Christian CW, Alessandrini EA. Patterns of health care use that may identify young children who are at risk for maltreatment. Pediatrics 2005; 116: 1303e8. Gilbert R, Widom CS, Browne K, Fergusson D, Webb E, Janson S. Burden and consequences of child maltreatment in high-income countries. Lancet 2009; 373: 68e81.
PAEDIATRICS AND CHILD HEALTH 22:11
463
Ó 2012 Published by Elsevier Ltd.
SYMPOSIUM: SAFEGUARDING CHILDREN
Shaffer A, Huston L, Egeland B. Identification of child maltreatment using prospective and self-report methodologies: a comparison of maltreatment incidence and relation to later psychopathology. Child Abuse Negl 2008; 32: 682e92. Sidebotham P, Heron J, ALSPAC Study Team. Child maltreatment in the “Children of the Nineties:” the role of the child. Child Abuse Negl 2003; 27: 337e52. Stirling J, American Academy of Pediatrics Committee on Child Abuse and Neglect. Beyond Munchausen syndrome by proxy: identification and treatment of child abuse in a medical setting. Pediatrics 2007; 119: 1026e30. Stockwell MS, Brown J, Chen S, Vaughan RD, Irigoyen M. Is underimmunization associated with child maltreatment? Ambul Pediatr 2008; 8: 210e3. Trocme N, MacMillan H, Fallon B, De Marco R. Nature and severity of physical harm caused by child abuse and neglect: results from the Canadian incidence study. CMAJ 2003; 169: 911e5. UNICEF. A league table of child maltreatment deaths in rich nations. Florence: UNICEF Innocenti Research Centre, 2003. US DHHS. Child maltreatment 2006. Washington, DC: US Government Printing Office, 2008. Print. Walsh C, Jamieson E, MacMillan H, Trocme N. Measuring child sexual abuse in children and youth. J Child Sex Abus 2004; 13: 39e68. World Health Organization. Child maltreatment: fact sheet number 150; Available from: http://www.who.int/mediacentre/factsheets/fs150/en/ index.html (accessed 12 August 2012). Wilson RG. Fabricated or induced illness in children. Munchausen by proxy comes of age. BMJ 2001; 323: 296e7. Woodman J, Brandon M, Bailey S, Belderson P, Sidebotham P, Gilbert R. Healthcare use by children fatally or seriously harmed by child maltreatment: analysis of a national case series 2005e2007. Arch Dis Child 2011; 96: 270e5.
PAEDIATRICS AND CHILD HEALTH 22:11
Practice points C
C
C
C
C
C
464
Child maltreatment is a significant worldwide problem, with consequences that can include impaired physical and mental health throughout life, and adverse social and occupational effects that carry a heavy economic and social burden. A conservative estimate of the cost to the US economy in 2007 was approximately US$65bn, which was a challenge to an earlier estimate of over US$100bn. In middle- and low-income countries, the economic and social costs are harder to estimate but are extremely important. Epidemiological studies have tended to come from highincome countries, but it is estimated in broad terms that 20% of women and 5e10% of men report being sexually abused in childhood, and 25e50% of children report being the victims of physical abuse. Severe forms of physical abuse are associated with significant childhood mortality, most commonly in infants and toddlers and often due to the ‘shaken baby syndrome’, which is strongly associated with an inadequate or implausible history, retinal haemorrhage(s), metaphyseal fracture(s) and subdural haematoma(s). One of the most challenging forms of child abuse to diagnose and manage is fabricated or induced illness (FII) since the perpetrator(s) will make all efforts to conceal and distract healthcare professionals from the information required to make the diagnosis. The World Health Organization (WHO) recommends a ‘multisectoral’ approach to preventing child maltreatment, with the most effective intervention likely to be parental support and teaching positive parenting skills.
Ó 2012 Published by Elsevier Ltd.