News

A study of serious child abuse in Wales carries a cautionary message about the use of figures from the child protection register. Brian Morgan reports.

Research papers recently published in social work and medical journals have highlighted problems with the child protection register (CPR). The value of the CPR as a front-line instrument for helping to reduce child abuse may be viewed very differently by hospital health workers compared to social workers.

Doctors in accident and emergency or working in the community, who are the first to see children suspected of being abused, need reliable, up-to-date and easily accessible information about past concerns. The CPR is seemingly failing to provide this.

The most detailed of the recent evaluations of the CPR was a Welsh assembly and NSPCC-funded project looking at all discoverable cases of serious child abuse in Wales between 1996 and 1998.

The study, crucially, concludes that the best way to prevent the severe abuse of children is to take action when abuse is at the stage of only "minor episodes". It carries a cautionary message about the use of CPR figures as a measure of the extent of child abuse in a community. Wide variations in the true incidence of abuse have been uncovered in Wales.

The Welsh Paediatric Surveillance Unit (WPSU) gathered data from all parts of Wales for severe physical abuse of children up to the age of 14. Abuse was specified as being equivalent to grievous bodily harm at the criminal level. It also collated data for the incidence of physical abuse of babies under the age of 12 months, combining data from two different sources.

The WPSU uses a "yellow card" style reporting system serviced by paediatricians. The inclusion of cases in WPSU returns depends on multi-disciplinary agreement - at a strategy meeting, case conference or Working Together part 8 review - that abuse has taken place. For babies less than a year old, the researchers combined data from the WPSU and the CPR using "capture-recapture analysis". This technique enables a check to be made as to whether children are identified in just one or in both of the data sets. In this way the team produced more accurate estimates of the true incidence of abuse of very young children.

Some striking differences emerged in incidence by age, sex and locality. Severe abuse was six times more frequent in babies under a year old than in the next group up to five years of age, and 120 times more likely than in children over five. The three categories - subdural haemorrhage, fractures and death - account for these differences. The death rate was 10 times higher in babies under a year old than in the group one to five years of age. Boys were more likely to be abused than girls throughout the whole age range - the figures being 62 and 39 respectively, a highly significant difference, says the report.

There were too few cases identified for meaningful differences between local authorities to emerge from the study, but striking disparities were found between health authority areas. The three largely urban health authorities in south Wales had significantly higher abuse rates than the more rural health authority areas elsewhere. Bro-Taf Health Authority, which includes Cardiff, had an abuse rate of 9.4 per 1,000 babies less than a year old over the two year period, but Dyfed-Powys only 5.2.

The recording of names in the CPR is the result of a "process", say the authors, and this is not a substitute for more accurate systematic data collection, such as the WPSU. The report says the names of one-fifth of children who had been abused were not recorded on the CPR. A name not appearing on a CPR is no reassurance.

^l J R Sibert et al, "The incidence of severe physical child abuse in Wales" Child Abuse and Neglect, Volume 26, no 3, 2002

Brian Morgan is a social and health care writer.