The brutal circumstances of the death of Baby P are hard to bear, and made more harrowing still by the publication of photographs of him, showing the toddler as an icon of childish vulnerability. Gordon Brown says such deaths must never happen again, but the truth is that even if best practice were the norm, it would not be possible to prevent all children dying from abuse.
Our recent study, commissioned by the Department for Children, Schools and Families and carried out by a team from the University of East Anglia and the National Society for the Prevention of Cruelty to Children, examined 161 serious case reviews of child death and grave harm through abuse and neglect in England between April 2003 and March 2005. We concluded that the wealth of factors which raise or lower the risk of harm to the child was in most cases too complex for death or serious injury to have been predictable. In the current climate of moral outrage and panic, however, it seems this finding is unacceptable.
Rates of physical and sexual abuse appear to be falling in most high-income countries, including the UK, although not rates of neglect or psychological abuse. However, much abuse goes unrecognised (and hence not reported) and most abused children do not get help. Unlike Baby P, almost half of the children we studied were not known to the children's social care services, but their families were in contact with health visitors and midwives. Many families were also known only to specialist adult services for which the well-being and safety of children were not priorities (an omission, which must change). We found that many older, often suicidal, young people had histories of abuse and neglect. They were challenging to help but were failed by a range of agencies that squabbled over whether they had met their criteria for services. A recent Barnardo's study highlighted the demonisation of adolescents, whose vulnerability goes unnoticed.
Learning the lessons of the "worst cases" can lead to misinterpretation and bad decision-making. Although domestic violence, substance misuse and mental ill-health among parents were common, we found no cause-and-effect relationship between this and child death or serious injury. Repeat attendance at accident and emergency departments for babies with minor injuries, or hospital admission (as in Baby P's case), does, however, appear to be a marker of abuse, according to study evidence. Many families live in great adversity, which increases the risk of harm to children, but it is important to remember that most children do not suffer serious abuse in these circumstances.
With hindsight, the individual death of Baby P was preventable. He should have been protected because there were already heightened concerns for his safety. He was the subject of a multi-agency protection plan and had been seen on numerous occasions. The responsibility for protecting him was, appropriately, shared between many agencies, although blame is falling as usual on children's social care. This is wrong: the Children Act 2004 makes clear that all agencies are responsible for safeguarding minors. The reviews we studied reflected a preoccupation with boundaries and wrangles over which professional group was "responsible" for the child. It is a function of the new Local Safeguarding Children Boards to agree thresholds and ensure referrals. The extent to which the boards can achieve this will be a measure of their success.
Chaotic families, in which neglect often happens, tend to have complex histories and are confusing and overwhelming for practitioners. In Baby P's case there was not even a clear understanding of who was living in the household. The child is often "lost" in the midst of parental needs, hostility or even apparent co-operation. An assessment that this is a simple "neglect" case shuts off critical thinking and makes it hard to spot the combination of physical danger and other forms of abuse, as well as risks from others in the household, particularly new partners. Social workers should be (and evidence from audit studies says they usually are) approaching complex cases rigorously. Careful collection of information about the child, together with the parents' profile and history, should help them understand the stresses the parents are under - and what makes them, as well as new partners or visitors in the household, tick. It should also answer the question: "What's it like to be a child in this household?" What stops social workers thinking and acting systematically is the triple pressure from families, employers and bureaucratic demands. Work overload, a target- driven culture and poor support make workers stressed and ill or paralyse them into inactivity. Good support and supervision should not be solely about meeting performance indicators such as a reduction in numbers of children in care. This highly responsible role is crucial, but supervisors are not adequately trained or resourced. The vitriol attracted by high-profile child deaths will do nothing to bring good candidates into child protection work.
It may be that the threshold for removing children who are suffering chronic neglect is set too high. When social workers bring these cases to court they are often sent back to "start again" with the family; sometimes the consequences can be grave. Most child protection policy is effective, and when failings happen they are caused by human error. However, there is also a lack of guidance about acceptable standards of adequate parenting. This is a problem, not only for social workers, but for families themselves.
When the firestorm of the Baby P case is over, social workers and other professionals who protect children will need to be able to trust each other enough - and be trusted enough by the public - to start challenging their own practices and decisions.
Marian Brandon is a senior lecturer in social work at the University of East Anglia and lead author of a new study for the Department for Children, Schools and Families into child deaths from abuse and neglect in England