The incidence of severe physical child abuse in Wales☆

Pergamon Child Abuse & Neglect 26 (2002) 267–276

The incidence of severe physical child abuse in Wales夞 J. R. Siberta,*, E.H. Paynea, A.M. Kempa, M. Barbera, K. Rolfea, R.J.H. Morganb, R.A. Lyonsc, I. Butlerd a

Cardiff Family Studies Research Centre, Department of Child Health, University of Wales College of Medicine, Academic Centre, Llandough Hospital, Penarth, Wales CF64 2XX, UK b Royal College of Paediatrics and Child Health Wales, Royal Glamorgan Hospital, Ynys Maerdy, Llantrisant, Wales, UK c University of Wales College of Medicine, Heath Park, Cardiff, Wales, UK d Keele University, Staffordshire, England, UK Received 28 November 2000; received in revised form 17 May 2001; accepted 23 May 2001

Abstract Objectives: The purposes of this study were: (1) to ascertain the incidence and nature of severe physical child abuse in Wales; (2) to ascertain the incidence of all physical abuse in babies under 1 year of age; and (3) to determine whether child protection registers (CPR) accurately reflect the numbers of children who are physically abused. Methods: This is a population-based incidence study based in Wales, UK, for 2 years from April 1996 through March 1998. Children studied were under the age of 14 with severe physical abuse consistent with the criminal law level of Grievous Bodily Harm. This included seven categories of injury (death; head injury including subdural hemorrhage; internal abdominal injury; physical injury in Munchausen Syndrome by Proxy including suffocation; fracture; burn or scald; adult bite). Cases were ascertained by a pediatrician surveillance reporting system (WPSU). A criterion for inclusion was multidisciplinary agreement that physical abuse had occurred (at case conference, strategy meeting, or Part 8 Review). The incidence of all babies under 1 year of age with physical abuse was also studied. Ascertainment of babies under the age of 1 year was undertaken from CPR as well as the WPSU. Results: Severe abuse is six times more common in babies [54/100,000/year (95% CI ⫾ 17.2)] than in children from 1 year to 4 years of age [9.2/100,000 (95% CI ⫾ 3.6)]. It is 120 times more common than in 5- to 13-year-olds [0.47/100,000 (95% CI ⫾ 0.47)]. This is mainly because two types of serious abuse (brain injury including subdural hemorrhage and fractures) are more common in babies under the age of 1 year than older children. Using data from two sources (the WPSU and CPRs), the

夞 The Wales Office of Research and Development funded this study. The preparation work for this paper was funded by the NSPCC. * Corresponding author. 0145-2134/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 5 - 2 1 3 4 ( 0 1 ) 0 0 3 2 4 - 6

268

J.R. Sibert et al. / Child Abuse & Neglect 26 (2002) 267–276

incidence of physical abuse in babies is 114/100,000 (CI 114 ⫾ 11.8) per year. This equates to 1 baby in 880 being abused in the first year of life. The largely rural Health Authority area in Wales had incidence figures for abuse in babies that were 50% of the three other predominantly urban Health Authority areas. Boys throughout the series were more at risk of being severely abused than girls (p ⬍ .025). Only 29% of the babies under 1 year of age on the CPR had actually been injured. Thirty percent of abused babies under the age of 1 year and 73% of severely abused children over the age of 1 year had caused previous concern to health professionals regarding abuse or neglect. Conclusions: Physical abuse is a significant problem in babies under the age of 1 year. Very young babies (under 6 months old) have the highest risk of suffering damage or death as a result of physical abuse. Severe abuse, in particular subdural hematoma and fracture, is much more common in babies than in older children. There is evidence of failure of secondary prevention of child abuse by health professionals, with a greater need to act on concerns regarding abuse and neglect. Interagency child protection work in partnership with parents should focus more on protecting babies under age 1 year from further abuse than on maintenance of the infant within an abusive home. The CPR is not intended as an accurate measure of children suffering abuse. It is a record of children requiring a child protection plan and must not be used as a measure of numbers of abused children. © 2002 Elsevier Science Ltd. All rights reserved. Keywords: Physical child abuse; Epidemiology; Incidence; Child protection register

Introduction In the United Kingdom, professional and political debate on child abuse tends to be informed either by extraordinary events such as the death of children at the hands of their caregivers (Reder, Duncan, & Gray, 1993; Parton, 1999). This debate can also be informed by process measures such as the nature and rate of entries on child protection registers (CPR) (Creighton, 1987). The definitional and methodological difficulties (Belsey, 1993; Creighton, 1998) associated with incidence and prevalence studies in this area have been well documented (Creighton, 1998). The available data are generally regarded as insufficiently robust to provide baseline measures for the evaluation of service delivery strategies and individual interventions, especially in the context of a general shift toward evidence-based practice. Against this background, a population-based epidemiologic study of physical abuse in babies and severe physical abuse of all ages in Wales, UK, was undertaken. A successful experience was had using a monthly card return system from consultant pediatricians in the whole UK in our study of Munchausen Syndrome by Proxy (McClure, Davis, Meadow, & Sibert, 1996). It was therefore decided to use a similar system in Wales. Ascertainment in babies was confirmed using CPRs. Wales is one of the four nations that make up the United Kingdom; although the National Assembly in Cardiff governs Health and Social Services, the systems for child protection are similar to England. CPRs are maintained by the primary statutory agencies for child welfare (social services departments). Registers record details of those children for whom an interagency child protection plan has been agreed through a formal process of investigation and case review. Their purpose is to provide a reference point for those professionals working with a child and to alert others who may engage with such children and their

J.R. Sibert et al. / Child Abuse & Neglect 26 (2002) 267–276

269

families that such a plan exists. They do not therefore record all cases of child abuse (there is no general mandatory reporting system for child abuse in the UK), nor even all cases that have come to formal attention. Where no specific child protection plan is required (e.g., where a child has died or where a child is removed to substitute care), a child might not be entered on the register. The objectives of this study were to: Y To determine the incidence of severe physical child abuse in Wales; Y To determine the incidence of all physical abuse in babies; and Y To investigate the extent the CPR records abuse.

Methods The study included children between 0 and 14 years of age who had suffered severe physical abuse, and all babies under the age of 1 year who had been physically abused. In both categories, the child would ordinarily be resident in Wales and the injury identified by pediatrician. The cause of the injury was agreed in the multidisciplinary context as the result of abuse. Although young people between 14 years and their 18th birthday are subject to the Children Act 1989, not all would be seen by pediatricians. It was therefore decided not to include them. For the purposes of this study, abuse was understood as arising from a process where through actions that were directly injurious, the child suffered physical trauma at the hands of a caregiver or someone else in a position of trust or authority in relation to a child (Butler, 2000). Severe physical abuse was defined as compatible with the criminal law understanding in England and Wales of the offense of assault occasioning “grievous bodily harm” or more serious offenses such as murder or manslaughter. This included: Y Y Y Y Y Y Y

Death, Head injury including subdural hemorrhage, Internal abdominal injury, Physical Injury in Munchausen Syndrome by Proxy (including suffocation), Fracture, Burn or scald, Adult bite.

The study was conducted in the 2 year period from April 1, 1996 through March 31, 1998. This coincided with the major local government reorganization in Wales and the creation of a new set of social services and public health authorities. In Wales, a pediatrician would see all abused babies and severely abused children. Primary ascertainment of cases was from the Welsh Pediatric Surveillance Unit. This unit is managed jointly by pediatricians in Wales through the Welsh Pediatric Society (now the Welsh Committee of the Royal College of Paediatrics and Child Health) and the Department of Child Health, University of Wales College of Medicine. Pediatricians in Wales considered there was a need for a system that examined conditions that are too rare to be studied on a

270

J.R. Sibert et al. / Child Abuse & Neglect 26 (2002) 267–276

single district basis but not rare enough to warrant an all-UK study. The Welsh Pediatric Surveillance Unit was set up as a mechanism for this. It used the same methodology that had been successful with the British Pediatric Surveillance Unit (BPSU) (Hall & Nicoll, 1998). Cards are distributed each month to Consultant Pediatricians and Senior Clinical Medical Officers/Associate Specialists in Wales (approximately 90). Pediatricians return the cards even if there are no cases to report. The researcher is informed of any case report, and contacts the reporting pediatrician for further information. Reminders and requests to confirm reported cases were sent to all pediatricians in Wales at 6 month intervals as part of the ascertainment process for this study. Confirmation of our ascertainment was performed by studying all physical abuse in babies under 1 year of age. There was liaison with the 22 Social Services Departments in Wales for information on babies under 1 year of age entered on the CPR for physical abuse. The cases were matched individually, not using electronic means. This was possible for babies since it involved investigating 170 Social Services cases in a year cohort; it would have been impossible for all children on the CPR. Using these two sets of data, a capture-recapture analysis was performed (Lambert & Goldacre, 1998) to more accurately estimate the true incidence of severe physical abuse in babies under the age of 1 year. Capture-recapture analysis is an epidemiologic technique where incidence figures from two sources can be combined. Information is obtained on the cases ascertained by each source and those that are ascertained by both sources. From these figures, the likely true population can be estimated.

Results Incidence of severe physical abuse in children of all ages The cases and incidence figures of severe physical abuse, ascertained by pediatricians in Wales, are shown in Table 1 in three age bands. The incidence figures obtained are: babies, 54/100,000/year (95% CI ⫾ 17.2), children from 1 year to 4 years of age, 9.2/100,000 (95% CI ⫾ 3.6), and 5- to 13-year-olds, 0.47/100,000 (95% CI ⫾ 0.47). For babies, further information has also been obtained from CPRs ascertaining eight further cases. Using capture-recapture analysis, the estimated real number of cases of severe physical abuse in babies was 52 cases (95% CI ⫾ 8). The population of babies under the age of 1 year in Wales was 35,200. This gives an incidence of severe physical abuse in babies of 73.4/100,000 (CI ⫾ 8.2) per year. There is a marked difference with age in the incidence of severe abuse to a level of grievous bodily harm or more. Using our estimates, the incidence of severe abuse in babies less than 1 year of age was six times greater than in children from 1 year to 5 years of age. It was 120 times more frequent than in older children over 5 years of age. This is because of three categories of severe abuse: subdural hemorrhage, fractures, and death being much more common in babies than older children. Injury rates were calculated in 6 month age periods for these two most common causes of severe abuse: brain injury and fracture (Table 2). It is shown that brain injuries (including subdural hemorrhage) and fractures are more

J.R. Sibert et al. / Child Abuse & Neglect 26 (2002) 267–276

271

Table 1 Severe physical abuse in three age bands Age group

Under 1 year

Age group total population

35,200

1–4 years

141,200

5–13 years

424,500

WPSU cases 2 years

Incidence of severe abuse/ 100,000/ year from WPSU data

Extra cases ascertained from CPR

Revised no. of cases from capture/ recapture calculation

Incidence of severe abuse/ 100,000/ year from WPSU and CPR data

38 (5 deaths) 26 (2 deaths) 4 (0 deaths)

54 (95% CI ⫾ 17) 9.2 (95% CI ⫾ 3.6) 0.47 (95% CI ⫾ 0.47)

8 N/A

51.7 (95% CI ⫾ 5.8) N/A

73.9 (95% CI ⫾ 19.5) N/A

N/A

N/A

N/A

common in babies under the age of 1 year than older children. They are at their most frequent in the first 6 months of age. Deaths. Seven deaths from child abuse were ascertained during the 2 year period of the study. All of these deaths occurred in children under 5 years of age. The death rate is 10 times higher in babies than children from 1 year to 5 years of age. Detailed analysis of the seven deaths shows that: Y Y Y Y

Four died from subdural hemorrhage and shaken baby syndrome; One died from abdominal injury; One died from nonaccidental suffocation; and One died from nonaccidental drowning.

Injuries received. In addition to the 17 children identified with a brain injury or subdural hemorrhage secondary to physical child abuse in Wales and the 49 children who sustained fractures, less frequent types of abuse included: Y Burn and scald: 11 cases were ascertained, one of which was associated with a fractured clavicle; there were four cigarette burns; Table 2 Age of child and incidence of two types of serious abuse confidence intervals in brackets Age range (months)

Number of cases brain injury & subdural

Incidence/100,000/year

Number of cases fracture

Incidence/ 100,000/year

0–5 6–11 12–17 18–23 24–29 30–59 60–119

12 3 1 0 1 0 0

34.0 (14.8–53.2) 8.5 (0.23–22.6) 2.8 (0.0–8.29) 0 2.8 (0.0–8.29) 0 0

20 14 6 5 3 1 0

56.8 (31.8–81.5) 39.8 (27.3–51.9) 17.0 (3.36–30.2) 14.2 (6.4–22.0) 8.5 (0.0–17.9) 0.6 (0.0–1.9) 0

272

J.R. Sibert et al. / Child Abuse & Neglect 26 (2002) 267–276

Table 3 Previous professional concern regarding abuse Previous abuse (including CPR) No No Yes Yes

Previous neglect (including CPR) No Yes No Yes

Babies under 1 year

Children 1–4 years

Children 5–14 years

40 7 3 7

6 6 4 10

2 1 0 1

Y Serious abdominal injury: found in three cases from the study, causing one death; Y Adult bite: four children had nonaccidental bites inflicted by adults; Y Physical injury in Munchausen by Proxy (MSbP): four cases were ascertained that fell into the category of Munchausen by Proxy or induced illness with physical injury over the period of the study; one child (3-month-old boy) died of nonaccidental suffocation; and one child (15-month-old boy) died of nonaccidental drowning. Analysis of the cases by sex Boys throughout the series were more at risk of being severely abused than girls. In the series as a whole, there were 39 girls and 62 boys. This difference is statistically significant (p ⬍ 0.025). For the series of babies under 1 year of age, there were 28 girls and 43 boys; for the series over 1 year of age, there were 11 girls and 19 boys. Previous professional concerns regarding abuse The referring pediatrician was asked whether there had been any previous professional concerns about abuse or neglect (Table 3). In severely abused children over the age of 1 year, there had been a previous history of professional concerns regarding child abuse and neglect in 22 of 30 cases (73%). In babies under 1 year of age, there had been less opportunity for concerns to emerge simply because they were younger. Nevertheless, 17 of 57 cases (30%) caused professional concerns regarding abuse and neglect before the episode of severe physical abuse. Analysis of the children who died shows that in the cases of five of the seven children, there were previous major concerns about abuse, although only two had been on the CPR. The two that did not have concerns were both babies, aged 2 and 3 months of age. Incidence of all physical abuse in babies under one year of age It was ascertained by pediatricians from the Welsh Pediatric Surveillance Unit (WPSU) that 57 babies had actually been physically abused, and 49 babies from CPR. Thirty-five babies were ascertained by both methods, with 22 from the WPSU alone and 14 from the CPR alone. Using capture-recapture analysis, the estimated real number of cases of abuse in

J.R. Sibert et al. / Child Abuse & Neglect 26 (2002) 267–276

273

Table 4 Cases of all physical abuse 1/4/96-31/3/98 by health authority area Health authority

Cases over 2 yr period

Pop ⫻ 1000 Under 1 yr old

Age appropriate incidence/100,000/ year

Bro-Taf Morgannwg Gwent North Wales Dyfed-Powys

23 14 16 12 6

9.4 5.9 7.2 7.7 5.2

122 (CI 74–176) 116 (CI 56–180) 111 (CI 52–166) 78 (CI 34–122) 57 (CI 11–104)

babies was 80 cases (95% CI 80 ⫾ 8.3). This gives an incidence of all physical abuse in babies of 114/100,000 (CI 114 ⫾ 11.8) per year, or 1 in 880 babies abused in the first year of life. The CPR Our enquiry of local authorities in Wales about babies under the age of 1 year that had been placed on to the CPR under the category of physical abuse in the same 2 year period (April 1996-March 1998), ascertained 170 cases initially. Only 49 of these (29%) had actually suffered any physical injury. It was found that children entering the CPR under the category of physical abuse included children who were thought to be at risk of abuse because: Y Siblings had been abused, Y There was domestic violence, or Y The lifestyle of the parents presented risks of physical injury. Some local authorities, a minority, were able to differentiate in their routinely collected statistics between the children who had actually been abused and for whom physical abuse was considered likely in the future. Analysis of cases by locality Cases of abuse in babies were analyzed by local authorities. However, the numbers were too small in each authority for meaningful analysis. Nevertheless, upon analysis of cases by Health Authority area, clear differences emerged (Table 4), with statistically significant differences between the higher rates in the three largely urban South Wales Authorities and the largely rural Dyfed Powys (p ⬍ 0.001). North Wales had intermediate rates.

Discussion Epidemiology This study has demonstrated that physical abuse is a major problem for babies under the age of 1 year (Creighton, 1992). It is a real threat to their health and welfare. Our figures are

274

J.R. Sibert et al. / Child Abuse & Neglect 26 (2002) 267–276

broadly in line with both Kempe’s (1971) survey of newborns (85 per 100,000) and Baldwin and Oliver’s (1975) case study (100 per 100,000). However, Ards and Harrell (1993) demonstrated a much lower figure of 31 per 100,000 in the US. Death, fracture, and brain injuries from abuse are more common in babies than in older children, with the greatest risk in the first 6 months of life. Indeed although our numbers are small, it is likely that other forms of physical abuse such as burns, nonaccidental suffocation, and abdominal injury are also more common in babies than older children. It is known from previous work (Jayawant et al., 1998) that subdural hemorrhage has a mortality of 25% and resultant disability for half the surviving children. Babies are therefore at risk not only of dying from abuse but also of significant disability as well. It is believed that these figures are applicable to the whole United Kingdom. The findings indicate that the prevention of abuse in babies must be recognized as a significant problem for health professionals. Concerns of this nature have been previously voiced by the National Commission of Inquiry into the Prevention of Child Abuse (Childhood Matters, 1996) and by Speight and Wynne (2000). Clear figures are held to back-up previous professional perceptions. The practical implications of these findings are that there is a need for increased emphasis on the protection of children by secondary prevention. This might take the form of clear clinical guidelines for health professionals and more effective targeted training. This study shows that abuse in babies is most common in the three Health Authorities areas in South Wales. They have broadly similar social mixes and have similar incidences. The more rural North Wales and Dyfed Powys Health Authority Areas have lower rates, probably in keeping with their social class mix. Abuse is likely to be linked to the higher rate of socio-economic deprivation in these largely urban areas. Much more emphasis is needed on the community as a whole supporting parents with young children. Munchausen Syndrome by Proxy is an important subject but needs a population bigger than Wales to study it. The Department of Child Health, University of Wales College of Medicine has now completed such a study based on the British Isles. This used the British Pediatric Surveillance Unit and was in collaboration with Professor Sir Roy Meadow in Leeds (McClure et al., 1996, Davis et al., 1998). There was severe abdominal injury in three cases. There are important issues of diagnosis and incidence to determine, and there is a need for an all-UK study in this important area. The child protection process There is no evidence that the CPR is not fulfilling its primary role of recording those children who are in need of a child protection plan. It has been demonstrated however that the present system of recording cases by process is an unsatisfactory measure of recording the number of abused children. Episodes of severe abuse in 21% to 24% of cases are not recorded on the CPR. These results echo previous findings which suggest that of every 10 children referred, only two will be registered (ADSS, 1987, Giller, Gormley, & Williams, 1992; Gibbons, 1993). This has implications for the interpretation of the common practice of checking the CPR for reassurance when a child presents with evidence of injury. The fact that a child’s name is on the CPR is certainly cause for extra caution, but the fact that a child’s name is not on the CPR is no reassurance.

J.R. Sibert et al. / Child Abuse & Neglect 26 (2002) 267–276

275

During the course of the study, several cases were ascertained that did not meet the criteria for diagnosis where there have been concerns about abuse that have not been satisfactorily managed in the multidisciplinary context, including the case conference. What is perhaps equally worrying is that the majority of local authorities in Wales cannot readily identify children that have actually been severely physically abused. It is vital to have accurate and comparable incidence figures for this problem to monitor progress and determine the effectiveness of management and prevention. To develop an evidence base for clinical governance, information systems are needed to monitor the real incidence and nature of child abuse. This will require cooperative interagency work about definitions and information-sharing and will need to involve pediatricians as well as social workers. Preventing abuse This study confirms that the best way of preventing severe abuse is by dealing with more minor episodes of abuse. Pediatricians have indicated that there was a high level of previous concern regarding neglect and physical abuse in this study. The analysis of previous abuse in cases of subdural hemorrhage also confirms this (Jayawant et al., 1998). Previous physical child abuse of an infant is a significant risk factor for a subdural hemorrhage and must be taken seriously by child protection agencies. This is also borne out very much by an analysis of the seven deaths in the study. The study of abuse has demonstrated that boys are abused more commonly than girls are. This must be taken into account when risks are considered. The capabilities of a pediatrician card return system (the WPSU) were found to be very valuable in investigating the important subject of Severe Child Abuse. Capture-recapture methods were used to estimate the true incidence of abuse in babies in Wales. It is realized that the use of these methods for children can be questioned (Jarvis, Lowe, Avery, Levene, & Cormack, 2000). However, babies were sampled from two sources (the WPSU and the CPR) from the same population (Wales) and over the same time period. The presence on one list was not contingent on the presence of the other (one-fifth of children diagnosed as abused at Case Conference were not put on the CPR). It was also possible to exactly identify and match individual children. It is believed the use of these methods gives a valuable estimate of how common abuse is in the UK, and any lack of independence would tend to underestimate abuse.

Acknowledgments The authors are grateful to all the pediatricians, nurses, and social workers throughout Wales who helped with this study.

References ADSS (Association of Directors of Social Services). (1987). Child abuse: incidence of registrations for child abuse between 1985 and 1986. Reading: Berkshire Social Services Department.

276

J.R. Sibert et al. / Child Abuse & Neglect 26 (2002) 267–276

Ards, S., & Harrell, A. (1993). Reporting of child maltreatment: a secondary analysis of the National Incidence Surveys. Child Abuse & Neglect, 17 (3), 337–344. Baldwin, J. A., & Oliver, J. E. (1975). Epidemiology and family characteristics of severely-abused children. British Journal of Preventive & Social Medicine, 29 (4), 205–221. Belsey, M. A., (1993). Child abuse: measuring a global problem. World Health Statistics Quarterly, 46 (1), 69 –77. Butler, I. (2000). Child abuse. In D. Howe (Ed.), Encyclopedia of social work (pp. 48 – 49). Oxford: Blackwell. Childhood Matters. (1996). Report of the National Commission of Inquiry into the Prevention of Child abuse. London: Stationary Office. Creighton, S. J. (1987). Trends in child abuse. London: NSPCC. Creighton, S. J. (1992). Child abuse trends in England and Wales 1988 –1990 and an overview from 1973–1990. London: NSPCC. Creighton, S. J. (1998). Patterns and outcomes. In K. Wilson & A. James (Eds.), The child protection handbook (pp. 5–26). London: Bailliere and Tindall. Davis, P., McClure, R. J., Rolfe, K., Chessman, N., Pearson, S., Sibert, J. R., & Meadow, S. R. (1998). Procedures placement and risks of further abuse after Munchausen by Proxy, non-accidental poisoning and non accidental suffocation. Archives of Disease in Childhood, 78, 217–221. Gibbons, J. (1993). Operation of child protection registers. University of East Anglia: Social Work Development Unit. Giller, H., Gormley, C., & Williams, P. (1992). The effectiveness of child protection procedures: an evaluation of child protection procedures in four ACPC areas. London: Social Information Systems Ltd. Hall, S. M., & Nicoll, A. (1998). The British Paediatric Surveillance Unit—a pioneering method for investigating the less common disorders of childhood. Child: Care, Health & Development, 24, 129 –143. Jarvis, S. N., Lowe, P., Avery, A., Levene, S., & Cormack, M. (2000). Children are not goldfish-recapture techniques and their application to injury data. Injury Prevention, 6, 46 –50. Jayawant, S., Rawlinson, A., Gibbon, F., Price, J., Schulte, J., Sharples, P., Sibert, J. R., & Kemp, A. M. (1998). Subdural haemorrhages in infants: population based study. British Medical Journal, 317, 1558 –1561. Kempe, C. H. (1971). Paediatric implications of the battered baby syndrome. Archives of Disease in Childhood, 46 (245), 28 –37. Lambert, T. W., & Goldacre, M. J. (1998). Career destinations seven years on among doctors who qualified in the United Kingdom in 1988: postal questionnaire survey. British Medical Journal, 317, 1429 –1431. McClure, R. J., Davis, P. M., Meadow, S. R., & Sibert, J. R. (1996). Epidemiology of Munchausen Syndrome by Proxy, non-accidental poisoning, and non-accidental suffocation. Archives of Disease in Childhood, 75, 57– 61. Parton, N. (1999). Ideology, politics and policy. In O. Stevensson (Ed.), Child welfare in the UK (pp. 3–21). Oxford: Blackwell Science. Reder, P., Duncan, S., & Gray, M. (1993). Beyond blame: child abuse tragedies revisited. London: Routledge. Speight, N., & Wynne, J. (2000). Is the Children Act failing severely abused and neglected children? Archives of Disease in Childhood, 82, 192–196.