Antisocial behaviour

Aggression and fighting are part of normal child development and can help children to assert and defend themselves. Persistent, poorly controlled antisocial behaviour, however, is socially handicapping and often leads to poor adjustment in adults (Scott 1998). It occurs in 5% of children (Meltzer et al 2000), and its prevalence is rising (Rutter et al 1998). The children live with high levels of criticism and hostility from their parents and are often rejected by their peers.3 Truancy is common, most leave school with no qualifications, and over a third become recurrent juvenile offenders (Farrington 1995).

In adulthood, offending usually continues, relationships are limited and unsatisfactory, and the employment pattern is poor. Thelon term public cost from childhood for individuals with this behaviour is up to ten times higher than for controls and involves many agencies (Scott et al 2001b)

Antisocial behaviour accounts for 30-40% of referrals to child mental health services (Audit Commission 1999). Most referrals meet general clinical diagnostic guidelines for conduct disorder from ICD-10(international classification of diseases, 10th revision), which require at least one type of antisocial behaviour to be marked and persistent.

Rather fewer meet the diagnostic criteria for research, which for the oppositional defiant type of conduct disorder seen in younger children require at least four specific behaviours to be present (World Health Organisation 1993). The early onset pattern typically beginning at the age of 2 or 3 years is associated with comorbid psychopathology such as hyperactivity and emotional problems(Taylor et al 1996, language disorders, neuropsychological deficits such as poor attention and lower IQ, high heritability (Solberg et al1996), and lifelong antisocial behaviours (Moffitt 1993).

As a result of its prevalence and significant consequences, the management of these childhood behavioural problems has received an increasing level of attention, research and theory over recent years. Two of the more prominent interventions for the behavioural management of children are health visitors and Group Parenting Programmes. Each of these approaches will now be outlined and will be the focus of the systematic literature review to be discussed.

1.1 Health Visitors and behaviour management

The health visitor’s first task is to identify health care needs. Together with general practitioners, they provide the child health surveillance programme of immunisations, screening, and advice. They aim to identify those important conditions that parents might overlook and, for the rest, to help parents access professional expertise, voluntary agencies, and local facilities (NHS Executive 1996). Health visitors make key contributions regarding immunisation, breast-feeding, good nutrition and depression.

This role can extend to help make appropriate interventions regarding the management of child behavioural problems through home visits. Health visitors can help to identify problem situations and refer the parent/child to the right agency. Furthermore, they can advise the parent and help to equip them with the skills needed to effectively manage and reduce the behavioural problems. If the health visitor can meet the parent when the child is under 10 days old, or even at the ante-natal stage, then a trusting and effective relationship can be formed (Beecham 1997) which can have positive effects. It has been suggested that this is of particular relevance to subgroups such as single parents.

They have been shown tube less likely to attend health care environments for immunisations and their children appear to have more accidents around the home (Flemmingand Charlton 1998). These are clearly key issues within community service provision (Hall 1996). The health visitors can provide much needed support, particularly with the more vulnerable groups. This social support can have significant benefits during pregnancy/labour(Match and Sims 1992), after birth (Kumar et al 1993) and in reducing the probability that the mother will experience post-natal depression(Ray and Hornet 2000). The health visitor can therefore have a range of benefits for the parent and the child and the extent to which these benefits extend to the child’s behavioural problems merits consideration.

1.2 Group Parental Programmes

Harsh, inconsistent parenting is strongly associated with antisocial behaviour in children (Rutter et al 1998), but whether this is a cause or consequence or is due to a common genetic predisposition has been less clear (Farrington 1995). The pioneering work of Patterson and colleagues showed that parents had a causal role in maintaining antisocial behaviour by giving it attention and in extinguishing desirable behaviour by ignoring it (Patterson 1982).

Such findings have facilitated the development of group parenting programmes which aim to reduce children’s anti-social behaviour by working with parents. These programmes include the Webster-Stratton programme (Webster-Stratton and Hancock 1998) and the Solihull approach. They generally involve group sessions with parents of children who have behavioural problems.

Sessions take place over a few months and involve the discussion of topics such as play, praise, limit setting, rewards and the handling of misbehaviour. The children do not attend the sessions. Video tapes aroused to provide examples of good and bad parenting behaviour and encourage the parents to talk about their experiences. This approach provides an alternative way of managing child behavioural problems rather than the need for health visitors to attend the parents’ homes.

1.3 Evaluating Health Interventions

Before selecting any health-related intervention it is vital that theyare assessed on a number of grounds through empirical research which investigates their effectiveness and efficacy. Within the NHS, cost restraints pose a significant issue and hence any intervention needs to provide value for money relative to other potential options (Royal College of Paediatric and Child Health 1997).

The Audit Commission(1997) reported that the annual maternity costs in England and Wales are £1.1 billion. Hence, any savings, or more cost-effective approaches, could have significant impacts on the financial performance of the NHS. Both group parenting programmes and health visiting have been evaluated within empirical research. Most of this research has taken place within America (Deal 1994).

The following review will consider this research in order to evaluate the use of group parenting programmes and home visits by health visitors with regards to their effectiveness and efficacy for managing child behavioural problems. The methodology employed within this research will now be outlined before ten relevant research studies are discussed and critically analysed. These findings will then be related to the research discussed in this introduction to the review before overall conclusions are drawn regarding the research question.

1.4 Method and search history

A systematic review aims to integrate existing information from comprehensive range of sources, utilising a scientific replicable approach, which gives a balanced view, hence minimising bias (Clarke& Oman 2001). In other words, a scientific approach will help to ensure that research evidence is either included or excluded based upon well-defined and standardised criteria.

This should ensure that the possible effects of researcher bias should be kept to a minimum. Berkley and Glenn (1999) also states that systematic reviews provide a means of integrating valid information from the research literature to provide a basis for rational decision making concerning the provision of healthcare. Literature reviews are important as they can help to consolidate the knowledge which is available on a given topic.

The main themes and findings can be highlighted and this information can inform the design, implementation and evaluation of future research. In this instance, the research evidence can be used to make recommendations and decisions regarding the use of health visitors and Group Parenting Programmes for behaviour management in children.

1.5 Reviewing process

Whenever one reviews or compares research reports, it is important that clear set of criteria are established upon which the evaluations can be made. Table 1 below outlines the global process which was used to conduct the literature review. This process was based upon that employed by Berkley et al (1999) It is important that such a framework is identified and used to structure a literature review so that all of the relevant stages are addressed and that limitations which could be associated with the methodology employed can be reduced where ever possible.

Table 1: Systematic Review (Summary of Framework)(Adapted from Berkley and Glenn 1999)
Identify the need Rationale, background information, existing work
Formulate problem and specify objectives
Background, problem specification, objectives
Develop review protocol
Design, resources, refinement
Literature search and study retrieval
Sources, search strategy, documenting a search strategy
Assessing studies for inclusion
Defined criteria, minimising reviewer bias, tables of studies included and excluded
Assessing and grading studies
Appraising checklists, hierarchies of evidence
Extracting Data Data collection forms, extraction methodology

Synthesizing data Qualitative overview, quantitative synthesis
Interpreting results Strength of evidence implications of results
Disseminating and implementing results
Methods of dissemination and implementation

In terms of the process used to review the selected research, the guidelines used by McInnis et al (2004) were adopted. These are displayed in Table 2 below:

Table 2: Core Principles Used in Reviewing Selected Research Articles (adapted from McInnis et al 2004)

Systematic reviews
Adequate search strategy
Inclusion criteria appropriate
Quality assessment of included studies undertaken
Characteristics and results of included studies appropriately summarized
Methods for pooling data
Sources of heterogeneity explored
Randomised controlled trials
Study blinded, if possible
Method used to generate randomisation schedule adequate
Allocation to treatment groups concealed
All randomised participants included in the analysis (intention to treat)
Withdrawal/dropout reasons given for each group
Cohort
All eligible subjects (free of disease/outcome of interested) selected or random sample
80% agreed to participate

Subjects free of outcomes on interest at study inception
If groups used: comparable at baseline
Potential confounders controlled for
Measurement of outcomes unbiased (blinded to group)
Follow-up sufficient duration
Follow-up complete and exclusions accounted for ( 80% included in final analysis)
Case control
Eligible subjects diagnosed as cases over a defined period of time or defined catchment area or a random sample of such cases
Case and control definitions adequate and validated
Controls selected from same population as cases
Controls representative (individually matched)
80% agreed to participate

Exposure status ascertained objectively
Potential confounders controlled for
Measurement of exposure unbiased (blinded to group)
Groups comparable with respect to potential confounders
Outcome status ascertained objectively
80% selected subjects included in analysis

Cross-sectional/survey
Selected subjects are representative (all eligible or a random sample)
80% Subjects agreed to participate

Exposure/outcome status ascertained standardized way
Qualitative
Author’s position clearly stated
Criteria for selecting sample clearly described
Methods of data collection adequately described
Analysis method used rigorous (i.e., conceptualised in terms of themes/typologies rather than loose collection of descriptive material)
Respondent validation (feedback of data/researcher’s interpretation to participants)
Claims made for generalizability of data
Interpretations supported by data

The results of this analysis will be presented via the CAST tool. Thesis available in two formats. The first concerns the evaluation of qualitative research studies and the second provides a framework forth evaluation of studies which have used a randomised and controlled approach within their methodology. The use of such a framework can provide structure within the results section and ensure that the data is presented in a way which is easily read and understood by the reader.

1.6 Sources of data

The methodology employed within the research will involve obtaining data from three key sources: Computerised searches, Manual searches, and the Internet. Each of these data sources will now be considered in more detail.

1.6.1 Computer-based searches

The methods used in this research will include a detailed computerised literature search. Multiple databases, both online and CD–Rom will be accessed to retrieve literature because they cite the majority of relevant texts. (Ford and Miller 1999) The computerised bibliographic databases are:-

• MEDLINE
• EMBASE
• CINAHL
• PSYCHINFO
• British Nursing Info BNI
• Cochrane
• Science Direct (All Sciences Electronic Journals)
• Asia
• DETOC
• HMIC

However because articles may not be correctly indexed within the computerised databases, other strategies will be applied in order to achieve comprehensive search (Sindh & Dickson 1997).

1.6.2 Manual searches

A manual search will be performed to ensure that all relevant literature is accessed. The manual searches will include:-

• Books relevant to the topics from university libraries and web sites
• Inverse searching- by locating index terms of relevant journal articles and texts
• Systematically searching reference lists and bibliographies of relevant journal articles and texts

1.6.3 The Internet

The internet will provide a global perspective of the research topic and a searchable database of Internet files collected by a computer.
Sites accessed will include:-

• Department of Health
• National Institute of Clinical Excellence
• Google
• The British Medical Journal website (www.bmj.com)

1.7 Identification of key words

Databases use a controlled vocabulary of key words, in each citation. To assist direct retrieval of citations techniques Boolean logic will be applied using subject indexing, field searching and truncation to narrow the topic focus (Hicks 1996, Goodman 1993). As part of this approach, key words will be based on the components of the review question.

An imaginative and resourceful technique of searching electronic databases will be used including recognising the inherent faults in the indexing of articles. Misclassification and misspelling will be included in the searches with searches utilising keywords and the subheadings, (Hicks 1996). Based on these principles, the following search terms will be used in different combinations:

• Behaviour Management
• Children
• Anti-Social Behaviour
• Health Visitors
• Group Parenting Programmes
• Webster-Stratton
• Solihull
• Evaluation

Further search terms may be used within the methodology if they are identified within some of the initial search items. Whenever one is searching literature ‘sensitivity’ and ‘’specificity’ are important issues when conducting searches of research on a database. The searches need to be as ‘sensitive’ as is possible to ensure that as many of the relevant articles are located.

This may be a particularly salient issue with regards to the evaluation of behavioural management techniques for children as the number of appropriate entries may be limited. Thus an attempt to locate as many of these articles as possible becomes a more relevant and important objective. Furthermore, the search needs to be ‘specific’. In other words, it needs to be efficient where appropriates that a higher number of the articles identified through a database search can be included and hence the time allocated to reviewing articles which are ultimately of no relevance, can be kept at inacceptable level.

1.8 Inclusion/Exclusion criteria

In order that a manageable quantity of pertinent literature is included in this study, it is essential that inclusion and exclusion criteria are applied. In order that a diverse perspective of the topic is examined broad criteria will be used. (Benignant 1997). However, it is important to note that a balance needs to be achieved through which the scope of the inclusion criteria is sufficiently wide to include relevant articles whilst also being sufficiently specific such that the retrieval of an unmanageable set of articles is avoided.

1.8.1 Inclusion criteria:

The articles which are highlighted within the proposed searches will be assessed in terms of whether or not they meet the following criteria. Each article will need to be viewed as appropriate with regards to all of these constraints if they are to be included in the final analysis.

• A literature review encompassing all methodologies will be applied ( Pettigrew 2003)
• International studies will be included
• Available in English
• Relate to the evaluation of Health Visitors and/or Group Parenting Programmes
• Focus on the behaviour of young children
1.8.2 Exclusion criteria

The articles highlighted by the searches will also be assessed in terms of whether or not they fulfil the following exclusion criteria. If a potential relevant article meets one or more of these criteria then they will be immediately excluded from the data set and will not be included within the analysis stage of the methodology.

• It is not the purpose of this review to discuss the development of behavioural management interventions so studies focusing on this will be excluded
• Literature in a foreign language will be excluded because of the cost and difficulties in obtaining translation.
• Research reported prior to 1990 will not be included within this review.

1.9 Consideration of ethical issues

Any research involving NHS patients/service users, carers, NHS data, organs or tissues, NHS staff, or premises requires the approval of ankhs research ethics committee (REC).(DH 2001) A literature review involves commenting on the work of others, work that is primarily published or in the public domain. This research methodology does not require access to confidential case records, staff, patients or clients so permission from an ethics committee is not required to carry out there view.

However, it is essential to ensure that all direct quotes are correctly referenced. Permission must be sought from the correspondent before any personal communication may be used. All copyrights need tube acknowledged and referenced. The researcher will also act professionally when completing this report and ensure that research is identified, reviewed and reported accurately and on a scientific basis. The analyses of the ten selected articles will now be summarised.

2.0 Results and CAST tool

Based on the inclusion and exclusion criteria for this literature review, a set of ten research studies were selected. They will now be analysed using the CAST Tool.

Article 1: Morrell and Walters (2000)

TITLE Costs and effectiveness of community post-natal support workers: Randomised controlled trial
AUTHORS Morrell CJ and Walters PS
SOURCE British Medical Journal, 2000: 321, 593-598
QUESTION 1: FOCUS This research was sufficiently focussed on assessing the cost effectiveness of a series of home visits by a health visitor. It aimed to determine the cost of this intervention compared to that which would be normally incurred through the maternity process. It also aimed to investigate the health benefits of these individual home visits for the mothers and children involved.
QUESTION 2: APPROPRIATENESS A randomised controlled trial was employed within the methodology of this research as it provided a group with which the results of the women in the intervention group could be compared. Therefore the progress of women who had recently given birth could be monitored and analysed to see if there were any significant differences as a result of the attendance of a Community post-natal support worker.
QUESTION 3: ALLOCATION A total of 623 women who had recently given birth were recruited for the study at a university teaching hospital. They were randomly allocated to either the intervention group (N = 311)or the control group (N = 312). The only requirement for inclusion in the study was that the participants were giving birth. Participants were not matched for factors such as their age, marital status or whether or not it was their first child. It was presumed that such individual differences would be controlled for by the random allocation of the participants within the relatively large sample. Subsequent analysis of the characteristics of those in the sample revealed that there was no significant differences in terms of age between the intervention and the control group. Neither did they differ on a set of88 socio-economic details.
QUESTION 4: BLINDED The intervention participants were not blind to the fact that they were receiving help from a support worker. No detailed information is given of the control group and of what their perception and knowledge of the research was. Inevitably the support workers themselves knew that they were in the intervention group. The potential, however, for observational bias was relatively small as the dependent variables were provided by the participant. As they had nuclear interest in demonstrating that the intervention had made appositive effect when it actually had not, this should have helped to ensure that the data given were accurate accounts of what had actually happened. .
QUESTION 5: ACCOUNTED FOR Of the 623 participants who were recruited for the original study, a total of 551 participants completed the whole study through to the follow up stage. The cases of drop out were due tithe participants not wanting to complete the course of home visits or because they did not return the questionnaires at the follow up stage.
QUESTION 6: FOLLOW-UP A range of questionnaires were completed by the participants at the six week and six month follow up stages. It would have been interesting to combine this approach with a more qualitative method, such as a focus group, such that a more in-depth data set could be gained to supplement the quantitative data.
QUESTION 7: CHANCE The study employed a relatively large sample of 551 participants.
QUESTION 8: FINDINGS Therefore were no significant health benefits associated with the intervention at the six week or six month follow up periods. The cost of the intervention to the NHS was £815 for the intervention group and £639 for the control group. There were no differences between the groups in terms of their use of the social services and in personal costs.
QUESTION 9: PRECISE The study provides p values which indicates that there are no significant benefits associated with this intervention despite it being significantly more expensive.
QUESTION 10: OUTCOMES As a result of the relatively large sample it would appear that these results could be generalised to other simple hospital situations in the UK. Based on the statistics provided, one would not recommend this intervention in terms of the health benefits. Having said this, it was a popular intervention with the women who received it and this may have value in itself.

Article 2: Scott et al (2001a)

TITLE Multi-centre controlled trial of parenting groups for childhood anti-social behaviour in clinical practice.
AUTHORS Scott S, Spender Q, Dolan M, Jacobs B and Ashland H
SOURCE British Medical Journal, 2001, 323, 194
QUESTION 1: FOCUS This research was sufficiently focused on the evaluation of a specific programme for a specific age group and set of behaviours.
QUESTION 2: APPROPRIATENESS A sample of 141 3-8 year olds were allocated to either receive the intervention or to go on a waiting list(control group). Allocation was based on the date of referral This was an appropriate approach for this research study as it enabled the effects of the intervention programme to be evaluated.
QUESTION 3: ALLOCATION The controlled trial approach was used as the allocation procedure should help to ensure that the children in the intervention and control groups exhibited equivalent anti-social behaviour and hence individual differences could be controlled for.
QUESTION 4: BLINDED The participants were blind to the allocation stage of the methodology. The participants were aware that they were taking part in an evaluation study. The people who rated video tapes on the parent participants and their children was blind to whether the participant had been in the intervention group or in the control group. Therefore the ratters were blind to treatment and condition.
QUESTION 5: ACCOUNTED FOR A total of 31 participants dropped out of the study as they did not attend a sufficient number of the intervention sessions.
QUESTION 6: FOLLOW-UP Participants were followed up five to seven months after the base line stage. Six measures of child behaviour were taken as well as one measure of parenting behaviour. This is inacceptable follow up period for this form of study. A long term follow-up, however, would have helped to establish the permanence of any significant changes which result from the intervention.
QUESTION 7: CHANCE A power calculation was reported in this study and the sample size exceeds that which is recommended. Thus it could be argued that sufficient steps have been taken to minimise the possible influence of chance.
QUESTION 8: FINDINGS The referred children who took part in the study were highly anti-social. A significant reduction was observed in taint-social behaviour of those within the intervention group. The behaviour of those within the control group was found to remain constant. The praise given by parents was found to increase three fold by those in the intervention group and to decrease by a third for those in the control group.
QUESTION 9: PRECISE Confidence levels are provided within the statistical section of the study. Based on these it could be concluded that the parental group behavioural programme does have a significant impact on serious anti-social behaviour among children.
QUESTION 10: OUTCOMES The large sample and sound methodology employed within this research would lead one to conclude that these results could be generalised to children of similar ages and with similar levels of anti-social behaviour.

Article 3: Harrington et al (2000)

TITLE Randomised comparison of the effectiveness and costs of community and hospital based mental health services for children with behavioural disorders.
AUTHORS Harrington R, Peters S, Green J, Byford S, Woods J and McGowan R.
SOURCE British Medical Journal, 2000, 321, 1047-1050
QUESTION 1: FOCUS The research focused on the evaluation of a community based versus a hospital based delivery of mental health services for children with behavioural disorders. The question set was relatively broad including both the costs and effectiveness of the approaches but it was sufficiently focused on specific programmes.
QUESTION 2: APPROPRIATENESS The parent/child participant pairing were randomly allocated to receive the behavioural programme either at community location or at the hospital. This allocation was performed bay researcher who was independent of the study. The allocation was performed using stratified sampling between the two different health authorities involved in the research.
QUESTION 3: ALLOCATION This randomisation was performed such that no bias within the allocation procedure could have an influence on the results. The potential of parental expectations as a confounding variable was also acknowledged and assessed. No significant difference was found between the two groups on this variable.
QUESTION 4: BLINDED At the observational stage of the research theatre was blind to the treatment group of the participants. This was demonstrated when they tried to identify the location which different participants had received the intervention. Their performance on this task was no better than chance.
QUESTION 5: ACCOUNTED FOR A full set of data was available for 115 out of the 141 participants who took part in the research. The drop outs occurred through non-attendance to the programme sessions or no data being provided at the follow up stage.
QUESTION 6: FOLLOW-UP The participants were followed up one year after the base line stage.
QUESTION 7: CHANCE The sample size was selected based on the size of the effect which was required by the purchaser and the provider’s agreements regarding whether the programme would be accepted for wider implementation.
QUESTION 8: FINDINGS It was reported that there were no significant differences between the intervention groups in terms of the parents’/teachers reports of the child’s behaviours, the parents ‘criticisms of the child and the impact of the child’s behaviour on the family. Parental depression was identified as a significant problem and variable which predicted the outcome of the child’s behaviour assessments.
QUESTION 9: PRECISE The ultimate finding of this research was fairly specific in suggesting that the location in which a parental behavioural management programme was delivered did not have significant impact on the child’s behaviour. It appears more important that a range of services are made available, including those which address parental depression.
QUESTION 10: OUTCOMES The large sample and the use of two different health care authorities would lead one to conclude that these findings could be generalised to other areas of the UK.

Article 4: Buts et al (2001)

TITLE Effectiveness of home intervention for perceived child behavioural problems and parental stress in children with utero drug exposure
AUTHORS Buts AM, Pulpier M, Marino N, Belcher M, Leers M and Royall R.
SOURCE Archives of Paediatric and Adolescent Medicine, 2001, 155, 1029-1037
QUESTION 1: FOCUS This research project was specifically focused on evaluating a home intervention programme which aimed to educate and provide support for parents of children with perceived behavioural problems.
QUESTION 2: APPROPRIATENESS Participants were mothers who had recently given birth at one of two urban based hospitals in Baltimore, USA. They were randomly allocated to either receive the home visits or to be given the standard care package which would usually be given.
QUESTION 3: ALLOCATION Random allocation was used to overcome any potential bias which could have been present if the researchers had allocated the participants. This enabled an assessment of the relative benefits of the home intervention to be determined over and above that which would be associated with standard care.
QUESTION 4: BLINDED The data obtained within the study was via questionnaires completed by the parental participants. They were blind at the allocation stage of the study but clearly they knew that they had been either exposed or not exposed to the home visit intervention. The child behaviour ratings were given by an independent observer.
QUESTION 5: ACCOUNTED FOR A total of 100 participants took part in the study. A sample of 51 participants comprised the standard care control group with 49 being in the intervention group. The details of the dropout rates were not clear.
QUESTION 6: FOLL