The Cost-Effectiveness of UK Parenting Programmes for Preventing Children’s Behaviour Problems – A Review of the Evidence

• Report title: The Cost-Effectiveness of UK Parenting Programmes for Preventing Children’s Behaviour Problems – A Review of the Evidence
• Report author: Madeleine Stevens, research fellow, Personal Social Services Research Unit, London School of Economics and Political Science, London, Child and Family Social Work (2014)

In March 2012, the UK government announced a pilot of free parenting classes to parents of all children under five, on the same day as an independent panel reported that poor parenting was widely blamed for the UK riots of August 2011. The presence of serious behaviour problems in childhood, affecting nearly five per cent of UK children is strongly associated with later delinquency and criminality, and several recent UK policy documents have emphasised the economic advantages of preventative early intervention (Field 2010; Allen 2011; Munro 2011).

Parenting programmes designed to help parents respond to difficult behaviour and strengthen family relationships are promoted in Graham Allen’s review of early intervention and have been widely implemented in the UK. Substantial long-term savings have been predicted; however, children’s social care services have pointed out the difficulties of demonstrating cost effectiveness over the short term and have experienced budget cuts, particularly to preventative services.

This paper reviews the economic evidence available to support or question claims that parenting programmes can be a cost-effective component in early intervention packages.

Parenting programmes vary in format and theoretical approach, with aims including improvement in parents’ and children’s emotional wellbeing and reduction in abuse, and are delivered by a variety of practitioners including social care, health and education professionals, often in multiagency settings.

The research focuses on programmes addressing children’s behaviour problems, typically delivering eight to 12 weekly two-hour sessions, often group-based, following a structured behavioural programme.

Evidence on whether parenting programmes are cost-effective

Research (2002) by Richardson estimates that half of the children with behaviour problems get better without intervention, so a control group allows researchers to see how change in the intervention group compares with improvement that might have occurred anyway.

Researchers found three completed UK primary economic evaluations of parenting programmes using RCTs (Harrington 2000; Muntz 2004; Edwards et al.2007). Information on three further studies still in progress was also found. Two of the economic evaluations are cost-effectiveness analyses, i.e. they have estimated the cost of a unit of improvement in child behaviour resulting from a parenting programme. However, the results cannot be directly compared as different target groups, outcome measures and interventions were used.

Muntz studied children aged two to 10 years already referred to child and adolescent mental health services, comparing a group receiving standard services with one receiving intensive psychological intervention. There was a four-year follow-up showing improvements in both groups at six months but only sustained at four years in the intensive group. The study concluded that a unit of improvement in the outcome measure (the externalising T-scale of the child behaviour checklist) could be achieved by the intervention with a saving (of £224) compared with controls. However, the difference in incremental cost effectiveness between groups was non-significant and the study is considered a pilot.

Edwards looked at the Incredible Years (IY) parenting programme delivered through Sure Start, aimed at parents of children aged 3–5 years with potentially problematic behaviour, but not yet referred. About 116 families were included with 73 in the intervention group. Based on the post-intervention assessment the intervention was considered cost-effective compared with waiting list controls. It cost £5,486 to bring the child with the highest intensity score to below the clinical cut-off point and £1,344 for the average child. The programme was found to be more cost-effective for those at highest risk of developing conduct disorder. The cost per unit of improvement on the intensity score of the Eyberg child behaviour inventory (ECBI) was £73.

Although this study found significant improvements in the primary outcome measure, a meta-analysis using a pooled effect size (across all outcomes) categorised it as nonsignificant (Piquero et al. 2009).

The third evaluation based on an RCT (Harrington 2000) compared the costs and effectiveness of community and hospital settings for group-based parent training for families of children with behavioural disorders. There were no significant differences in either effectiveness or costs between groups (either service-use costs or costs of the intervention). Attendance was poor and the location did not affect this.

Three further economic evaluations of UK-based parenting programmes using RCTs were in progress at the time of the paper being published – the Family Links Nurturing Programme, an RCT and economic evaluation of the Triple P parenting programme for five- to 11-year-olds displaying problem behaviour, and the National Academy for Parenting Research (NAPR) is running a four-armed trial comparing the IY, a literacy parenting programme, the two programmes combined and a signposting control group for five- to seven-year-olds.

The evidence from trials suggests there are short-term benefits of parenting programmes; decision-makers need to decide whether the benefits are worth the cost. Tentative evidence suggests a reduction in use of services following programmes from the few studies that measured use of services other than the intervention in question.

There is evidence from controlled trials of a sustained effect up to one-year post-intervention, but less is known about the longer-term effects of parenting programmes on conduct disorders. A few studies have attempted to estimate longer-term consequences by integrating data from effectiveness studies and other sources, despite arguments that putting a value on reductions in child conduct disorder is too difficult. This may for example, involve evidence from cohort studies on the associations between behaviour problems and long-term outcomes connected with wellbeing such as health, employment and criminal behaviour.

Long-term benefits

Bonin et al. (2011) estimate the cost-savings of providing a parenting intervention to five-year-olds with clinical levels of behaviour problems. This model looks at the impact of preventing persistent conduct disorder to age 30. Drawing on the best estimates available in the existing literature, the model estimates that on average 34 per cent more children receiving the programme will improve to non-clinical levels of behaviour than controls. The model concludes that savings would arise from reduced use of health services, educational support, social care, voluntary agencies, and crimes averted, and that total net savings would be 5.1–11.1 times the intervention cost. The figures imply that every £1 invested in the intervention generates £14 of savings to society, including £5 of savings to the public sector.

The societal figure (£14) includes costs of crime beyond the criminal justice system, including costs to victims of crime due to stolen or damaged property, losses in earnings, and emotional and physical impact. With this perspective, the costs of crime account for about three quarters of the costs of conduct disorder.

A modelling study (McCabe et al. 2005) undertaken in connection with guidance from the National Institute for Health and Clinical Excellence (Nice) recommending the use of parenting programmes for conduct disorders in children aged three to eight years (Nice 2006) assumed that improvements in child behaviour found in a systematic review (Dretzke et al. 2005) were sustained for one year, and calculated net public service costs for this period.

It is calculated that compared with no intervention the incremental mean cost of improving a child’s behaviour from conduct disorder classification to conduct problems, or from conduct problems to normal behaviour, will be £90 for group community-based, £1,380 for individual home-based and £2,400 for individual clinic-based. For group clinic-based programmes, a saving of £70 per family was estimated. The study concluded the majority of cost-savings would be for education and health services.

The high costs of conduct disorders in adolescence and adulthood have been demonstrated, and so an effective, reasonably priced intervention is very likely to lead to large cost-savings in the long term. However, the long-term effectiveness of parenting programmes and the effectiveness of parenting programmes rolled out into routine practice have yet to be proven. In general only a few of the possible economic impacts of parenting programmes are taken into account in modelling studies, and many potential impacts, such as on quality of life, are hard to measure.

Factors influencing cost

The study also assessed the evidence on what factors influenced the cost of delivering parenting programmes. Analysis looking at the delivery costs of more than 100 parenting programmes found the median cost per person, assuming 10 participants per group, was £241. For programmes that could be delivered individually, the median one-to-one cost was £862 per person.

Staff costs represented around 60 per cent of costs but are likely to be underestimated as nearly half of the programmes said only one practitioner was required, whereas the norm for group delivery in practice is two.

The review looks at a number of UK studies that calculate the cost of delivering parenting programmes and conclude that there are so many uncertainties it is hard to accurately estimate savings. The costs of recruiting families may differ greatly depending on the area and the existing service structure, and the costs of recruiting and retaining families in the programme are likely to be different when taking place outside a trial (in which participants have to agree to assessment and possible assignment to a control condition). Supervision costs may also differ greatly; some trials provide relatively high-level supervision, which may be crucial for effectiveness, or travel to a distant location. Estimates from trials may not take account of full implementation costs, including management.

Nevertheless, all estimates suggest relatively low costs compared with the potential savings suggested by modelling studies.

Overall conclusion

The researchers say there is strong evidence that effective early intervention will be cost-saving. Parenting programmes are relatively low cost and should be cost-saving in the long term if short-term improvements last. However, much depends on the way programmes are implemented outside trials.

There is increasing awareness of a need to show how “demonstration” programmes can be made to work effectively in practice.

Qualitative work looking at interrelationships between services and supports, and how positive effects can be maintained, would also help.

In the face of the highest ever rate of care applications and cuts to preventative services, research needs to continue to investigate the case for the cost effectiveness of preventative family intervention, reflecting the complexity of family circumstances, service configurations and child development.

Click here for more in CYP Now's Parenting Programmes Special Report