Scottish Government

Growing Support

10. Child protection

Preventative approaches

Child protection has come to mean official responses aimed at protecting children from neglect, physical, sexual or emotional abuse within their homes or families. Abuse affects children of any age, though neglect and physical abuse are most likely to implicate younger children.

This section looks at the effectiveness of preventing child abuse/maltreatment in vulnerable families and, where it has occurred, ways of preventing reoccurrence and providing help to those concerned. Therefore we shall be examining the effectiveness of primary, secondary and tertiary prevention ( see section 2). These three aspects of prevention operate at different ecological levels, although there is inevitably some overlap. Where primary and secondary prevention are concerned services are largely targeted at the community level. This may include a number of the interventions already reviewed in previous sections, such as helping to develop social support and community networks, education and the provision of social and day care facilities. In contrast tertiary prevention is more focused on the individual or family, addressing causal factors within the family, usually by seeking to improve competency and parenting skills.

One of the difficulties in assessing the effectiveness of child protection interventions is determining what the contributing factors to child abuse are. This appears to be largely dependent on the discipline or perspective that the study is based on e.g. sociology, genetics, medicine, psychology. However, it is rarely just one factor that provides the sole reason for the occurrence of child abuse. A number of studies have recognised this and have been designed so they can assess the impact of several specific factors and levels of causation, through adopting an integrative approach (Sameroff and Chandler 1975, Brofenbrenner 1979, Cicchetti and Rizley 1981). Macdonald and Winkley (1999:38) have summarised four core elements of the integrative approach. These provide a framework for developing an appropriate research process and using the findings to develop suitable social responses to social problems (in this case child abuse/maltreatment):

• the dynamic contribution of the different disciplines/perspectives and their cumulative impact over time;
• the fact that some factors increase the risk of abuse, whilst others have a protective effect, the cumulative factors approach ( see section 1);
• the fact that some factors have an enduring effect and some are circumstantial or transient; and
• that the understanding of abuse requires that we examine the interplay of factors at different levels of analysis - the ecological approach ( See section 1).

It is therefore increasingly apparent that child abuse/maltreatment is not the result of a single factor but is a multiply determined circumstance, event or in some cases a series of events - stress, poverty, psychological problems, parents' own poor childhood history, lack of social support, etc. (Olds and Henderson et al 1986).

Difficulties have been encountered when attempting to review the effectiveness of secondary and tertiary interventions in particular (Macdonald and Winkley 1999). The main problem has been the lack of 'purity' of the samples as a result of a blurred distinction between those at risk of abusing their children and those who have actually abused. Macdonald and Winkley suggest that this occurs as a result of the methods used to recruit participants to the study. Primarily they are recruited through referrals to and from child protection agencies, where the classification of families into one group or another is a difficult professional judgement. This is not to say that these studies are not valid or reliable, but caution should be adopted when interpreting the findings.

Primary prevention

Primary prevention is targeted at one or more stress factors that are known to have the capacity to provoke abuse. Tackling these stress factors, or enabling parents to cope with them in order to promote the general wellbeing of children are the characteristics of primary prevention. Aspects of primary prevention have been discussed in previous sections on social support, family centres and community networks, so we will not discuss them further at this stage. General strategies to tackle poverty or social exclusion also help create the conditions where family stress and hence child abuse become less likely.

Secondary prevention

In the context of child protection, secondary prevention is targeted at 'high risk' families who are sometimes identified through screening processes within health services. As a result of the increased knowledge around risk factors and the cumulative effect these have on families, it is believed that secondary prevention is an important facet of child protection, and can be more effective than the limited success that has been experienced with tertiary prevention. Not only does secondary prevention act as a preventive strategy to abuse and maltreatment, it also maximises the chances of good developmental outcomes for children. In addition it is sometimes possible to provide interventions that are not always part of child protection services, so stigma and labelling is reduced with positive consequences, as was also noted in relation to family centres in section 4 (Macdonald and Winkley 1999).

Effective intervention lies partly with the ability to identify groups or individuals at increased risk of abuse. Such groups include teenage mothers and parents who have been abused themselves. This identification of risk is important so that interventions can be provided to them. Likewise, in research terms, this may allow for random allocation to control and intervention groups which, over time, will provide evidence as to whether the intervention has been effective. It has been suggested that pre-natal screening provides an effective means of identifying at risk individuals.

A number of interventions have been identified as effective in providing support to parents thought to be at risk of abuse/maltreatment. In particular home visiting by professionals, paraprofessionals and lay volunteers pre- and post-natally to women in high-risk groups has been found to be effective with respect to child protection, as in relation to broader family tensions considered in section 6. Although there are a number of studies in this area, some are more methodologically rigorous than others and the strength of their findings varies accordingly (Olds and Kitzman 1993, Macmillan 1994a, Clemant and Tourginy 1997). Macdonald and Winkley identify the study by Macmillan et al as the most methodologically secure review to date. In summary Macmillan et al found that long-term home visiting programmes for parents with one or more risk factors (single parenthood, poverty, teenage parent status) was effective in the prevention of child physical abuse. However, the available evidence regarding the provision of short term home visiting schemes was inconclusive.

The impact of a RCT reported by Olds et al (1997) provides a good example of an effective long-term home visiting scheme. It was specifically designed to improve particular aspects of maternal and child functioning and therefore provided interventions that were directly related to known risk factors. The study consisted of 400 predominantly white women all first-time mothers. Those allocated to the control group received ante-natal care and screening only, and the intervention group received home visiting.

The following findings were reported by the study:

• The rate of abuse or neglect amongst poor, unmarried teenage mothers was considerably higher in the control group compared with the intervention group (19 per cent compared with 4 per cent).
• Children aged 1-2 years in the intervention group were 40 per cent less likely to have visited a physician for a physical injury or ingestion of a poisonous substance.
• No change occurred in the number of referrals for child maltreatment in the two years following the study, although the children in the intervention group had 87 per cent fewer visits to the physician for physical injury or ingestion of poisonous substances compared with the control group. The intervention group children also lived in better environments, which were more conducive to safety and intellectual and emotional development.

Although the study did prove effective in a number of ways it was evident that no change occurred in the rate of referrals for child maltreatment. As Olds et al indicated, this may be partly because child abuse could be over reported in the intervention group as they were under closer scrutiny than the comparison group. The authors suggest that a longer period of intervention is perhaps required to make any impact on this issue as different problems emerge as children get older, which parents may not have the capacity to deal with. In addition the environmental and socio-economic problems already faced by such families are long standing and may need constant attention over a longer period of time.

Tertiary prevention

This level of prevention refers to action taken with respect to families where abuse is known to have occurred. It would appear that there is limited reliable evidence regarding the effectiveness of tertiary prevention (Oates and Bross (1995)). After setting four methodological criteria they sought to identify studies that evaluated child and/or parent-centred interventions and only identified a total of 16 studies in this area. Five of these studies were child-centred interventions and the remaining 11 were parent interventions. An additional review by Oates and Bross, covering the period from 1983 to 1992 and using different criteria, showed a total of 25 studies. None of these addressed routine service provision, but focused on special interventions for helping abused children (13 studies) or interventions aimed at helping parents (12 studies) (Macdonald and Winkley 1999).

There appears to be little evidence to suggest what is effective in tertiary prevention by way of legal or administrative interventions, including investigations, case conferences, registration, court and children's hearings decisions. This is largely because studies lack methodological rigour as a result of a lack of experimental designs, small sample sizes, high rates of attrition and inadequate outcome indicators. On the other hand, considerable research has documented the effects of these processes on parents and (mainly older) children (DH 1995; Hill and Tisdall 1997). Among the immediate consequences are distress and anger by many parents, and emotional trauma by many children. A longitudinal study by Gibbons et al (1995) showed that (a) rates of re-abuse were high (25-30 per cent) and (b) children registered for neglect had poorer long-term outcomes than those registered for other reasons. It is in the nature of long-term follow-up studies, though, that they relate to past intervention, in this case during the first half of the 1980s. However, other studies of more recent investigations and registrations indicate similar rates of re-abuse (DH 1995).

More evidence is available regarding therapeutic interventions. These mostly show that behavioural and cognitive approaches are effective in helping to prevent abuse in high risk cases and prevent abuse from re-occurring in families where it has already been a problem. Through such approaches parents are taught skills, learn appropriate response and reinforcement patterns for dealing with a child's 'difficult' behaviour, understanding of children, develop anger management, and modify their attributions towards children (Gough 1999).

It should be noted that there appears to be a paucity of UK studies in this area, and the majority of studies found for this review have largely come from the US. Few are recent. For example a study carried out by Wolfe et al (1981) compared a cognitive behavioural assessment intervention with a standard service for a total of 16 families where parents had abused their children or were found to be at high risk of abusing. Parents were ordered by the court to attend or had been given the opportunity to attend rather than be taken to court by the child protection agency. The intervention group (eight families) received group-based and individual support. The group support consisted of all parents meeting for two hours per week over a period of eight weeks, and these sessions included: developing an understanding of child development; problem solving and modelling of appropriate child management; and self-control training. The individual support consisted of weekly home visits by the project co-ordinator who acted as a facilitator in building on the skills developed as a result of the group sessions. Other more specific work was also carried out. The control group (eight families) received bi-weekly visits from statutory services to monitor the safety of the child in the home and also received community resources.

The results showed that none of the families in the intervention group had been reported or suspected of abuse during treatment, and a year after the programme had finished all cases had been closed as parents were thought to be functioning appropriately (Macdonald and Winkley 1999). In the control group only two families had not had their cases closed. One had a further report of abuse and the other remained under supervision. In addition to a higher rate of case closure, the families in the intervention group were also reported to have improved child management skills, fewer child behavioural problems and caseworkers observed there were fewer problems generally.

Macdonald and Winkley cite other studies which reported similar findings, although they indicate that few are RCTs (Crimmins et al 1984, Egan 1983, Wolfe and Sandler 1981, Wolfe et al 1981). The few British studies in this area that have been reviewed do not show any conclusive evidence regarding effectiveness (Smith and Rachman 1984, Nicol et al 1988).

Other cognitive approaches have been identified, which focus on the wider issues that are often problematic for families where abuse has occurred or there is a high risk of it occurring. Macdonald and Winkley (1999:112) provide a comprehensive overview of two approaches - behavioural family therapy and ecobehavioural approaches. The latter approach was developed as 'Project 12 Ways', which provided assessment and 12 core services to families where abuse had occurred. The project ran from 1979-1985 and catered for 50-100 families per year. A five-year follow-up was undertaken of 700 families, 352 of which attended the project and the rest were allocated into a comparison group with no project intervention. The study showed that families who attended the project had lower rates of abuse over all years apart from one year when the rate was similar to that of the comparison group. This was particularly interesting considering the intervention group were thought to have more severe problems than the comparison group (Lutzker and Rice 1987). Although these results appeared impressive the researchers found that over time the gap in incidence between the intervention and comparison group became less, even though it was still statistically significant. This is attributed to several factors - families dropped out of the programme and/or they did not receive adequate help in resolving existing problems. As we have seen a number of times in this report, research on different types of service suggests that prolonged or additional support is needed so that improvements made early on can be maintained.

Social network interventions

Moving away from the area of therapeutic intervention, the final part of this chapter examines the effectiveness of social network interventions, already briefly discussed in section 7. Although intense professional support is required for families who abuse or are suspected of abuse, there is evidence that a combination of formal and informal support can provide an effective intervention, especially in terms of preventing the re-occurrence of abuse.

Evidence comes from work carried out in the US by Gaudin et al (1990) who undertook a RCT of 88 families who had been identified as neglecting their children. Participants were recruited on a voluntary basis from the caseloads of social workers at the Department of Children's Services. Those who were allocated to the control group continued to receive standard services. Those allocated to the intervention group received 'holistic care management' from a project social worker who took on the role of case manager, which included advocacy and brokering of formal services, as well as facilitating the development of supportive social contacts.

The study showed significant improvements in several aspects for families in the intervention group at both six and 12 months. These results were obtained using three well validated tools (Edington et al 1980, Halper and Jones 1981, Polanski et al 1981). Families in the intervention group showed an increase in the size of support networks compared with the initial baseline measures, but no change was reported in the control group. At 12 months both groups' respondents reported a significant increase in the perceived supportiveness of their networks, but this change was bigger in the intervention group. There were significant improvements in age-appropriate expectations of parents in the intervention group and use of corporal punishment had decreased compared with baseline measures. However, the authors do point out that these improvements still remained below that of average non-neglectful parents. Using standardised parenting measures, it was reported that after nine months of participating, parents in the intervention group improved their parenting from neglectful or severely neglectful to marginally adequate parenting and almost 60 per cent of cases were closed as a result of these improvements. These results were undoubtedly positive, but as pointed out by the authors, those who participated in the study did so voluntarily, thereby indicating a will to address their parenting deficits. The findings, although significant would need to be carefully applied to parents who were not so compliant and easy to engage. As noted in section 6, a Canadian study also showed that work in a small neighbourhood aimed at improving the linkages of vulnerable mainly young mothers to informal support was linked with a reduction in known abuse in the locality (Fuchs 1995).