Article Text

Download PDFPDF

Are abused babies protected from further abuse?
  1. B A Ellaway1,
  2. E H Payne1,
  3. K Rolfe1,
  4. F D Dunstan1,
  5. A M Kemp1,
  6. I Butler2,
  7. J R Sibert1
  1. 1Departments of Child Health and Epidemiology, Statistics and Public Health, Cardiff University, Wales College of Medicine, Academic Centre, Llandough Hospital, Wales, UK
  2. 2Professor of Social Work, Keele University, Staffordshire, UK
  1. Correspondence to:
    Professor J Sibert
    Department of Child Health, University of Wales College of Medicine, Academic Centre, Llandough Hospital, Penarth CF64 2XX, Wales, UK;


A cohort of 69 physically abused babies under 1 year was followed for three years after the incident. Of the 49 babies who returned home following child protection investigations, 15 were further abused in the three year period, a re-abuse rate of 31%, which is grossly higher than the abuse rate in the general population.

  • babies
  • child abuse
  • follow up
  • outcome
  • physical abuse

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

There is little recent, systematic, or robust data available to provide a convincing case for the efficacy of any particular form of intervention to protect children from abuse.1–3 In the UK, Thoburn and colleagues4 acknowledge that the variables in their 1995 study were “soft and based on researcher ratings”; Farmer and Owen’s5 study is similarly based on a relatively small sub-sample. In particular, both internationally and domestically, there is a dearth of literature on the prevention of re-abuse in situations where abuse has been formally determined.

As part of a population based incidence study we ascertained a cohort of all the babies under the age of 1 year, physically abused in Wales between April 1996 and March 1998.3 We have now followed their progress over a period of three years after the abuse. We had the objective of ascertaining how effective child protection services are in actually protecting abused babies from further abuse, although we have not, for the purposes of the current study, sought to differentiate the various forms of child protection service operated in relation to each case. We have also studied information on the siblings before and after the original abuse.


We ascertained cases by a paediatrician surveillance reporting system (the Welsh Paediatric Surveillance Unit) and from Child Protection Registers in Wales. The full methodology is outlined in our first paper.6 The criterion for inclusion was multidisciplinary agreement that physical abuse had occurred (at case conference, strategy meeting, or Part 8 Review). We followed this cohort for three years, using information from social services and health sources. We used the same criterion to define abuse. We also gathered information on siblings both prior to the abuse and for the follow up period. We compared the re-abuse rate in babies with information that we have from the original study and from child protection registers. We compared a number of potential risk factors for abuse between the families of the group of babies who were allowed home and were re-abused with those who were allowed home and not re-abused. We chose these factors, because they were ascertainable from the information we had available and from our knowledge of factors that had been previously been shown to be connected with abuse. These factors were a criminal record in a parent, domestic violence, financial problems, the family being previously known to Social Services, mental health problems in a parent, a parent who was in care as a child, the baby being pre-term, previous conviction related to child abuse, previous injury due to physical child abuse, substance misuse, and young parent (under 18).

Sixty nine babies were identified over a two year period: an incidence of 113/100 000 (CI 80 to 152) per year. Of these babies, 39 had siblings born prior to the abuse. Eleven (28%) of these had been abused before the birth of the index case. In addition, there were serious child protection concerns by a professional, usually a heath visitor, in eight others. There was therefore abuse or serious concern over abuse in 19 (49%) of these families.

Of the whole cohort of 69 babies, five died from the abuse, one child went abroad and we have no further information, and 14 were permanently removed from the home, though one of these was re-abused during a contact visit. Of the remaining 49 babies who were returned home following child protection investigations, 15 were further abused in the three year period, a re-abuse rate of 31% (CI 20% to 45%), which is grossly different from the abuse rate in the general population.6

The re-abuse of the three year follow up consisted of physical abuse in eight children (including a fractured femur) and neglect in seven. Of the 15 re-abused children, 12 were allowed home after the subsequent abuse; three of these children were abused again, one three times. This child was therefore abused five times. Of the 49 babies returned home following child protection investigations, 35 had at least one sibling with 63 siblings in total. Of these, 11 siblings in seven families were re-abused in the three year period.

When we examined the comparison of risk factors between children who were re-abused and those who were not, none of the factors reached statistical significance (table 1).

Table 1

 Comparison of risk factors between children who were re-abused and those who were not


The risk of re-abuse of babies returned home after abuse is very high: more than 30%. The re-abuse was not just physical: there was neglect as well. Moreover not only were they at risk—their siblings were too. We have previously shown that 30% of abused babies had caused previous concern to health professionals regarding abuse or neglect.6 We now have shown that in abused babies with siblings, half of the siblings had shown previous child protection concerns.

All this represents a serious failure in secondary prevention in babies where the consequences of abuse can be death and disability. We must focus child protection services more on actually protecting babies and be more cautious where intervention involves their reintroduction to their families. These results should be seen in the context of Farmer and Owen’s5 finding that 25% of children placed on the child protection register were re-abused, and are consistent with re-abuse rates of between 26% and 31% reported in other studies.7,8 We can only speculate on the reasons why so many babies were re-abused. Our experience suggests that many social workers and health visitors have not been trained to recognise the increased risks of physical harm to babies that result from a previous episode of abuse.

Although there was no significant difference between risk factors in the families where re-abuse occurred, the profile of risk factors in the group as a whole is similar to that documented in abusing families by Browne and Herbert.9 They found both domestic violence and mental illness in 30–35% of abusing families compared to less than 6% of controls.


We thank all the paediatricians, nurses, and social workers in Wales who helped with this study.



  • Funding: We are grateful to the National Society for the Prevention of Cruelty to Children for funding this study. The Wales Office of Research and Development funded the original ascertainment of the cohort.

  • Possible conflict of interest: AMK, EHP, and JRS have previously been expert witnesses in child protection cases.

  • Ethical permission was given by the MREC for Wales.

Linked Articles

  • Atoms
    Howard Bauchner