Randomised controlled trials are rare in social work so I was pleased to review one for my second blog for the Social Care Elf this week. However, the study’s limitations mean that its findings must be
Randomised controlled trials are rare in social work so I was pleased to review one for my second blog for the Social Care Elf this week. However, the study’s limitations mean that its findings must be applied with caution. Unfortunately this is all too common, highlighting the importance of improving the rigour of social work evaluation research. Here are my thoughts on Coleman et al.’s (2014) evaluation of their training programme which aims to help Sure Start Children’s Centre staff talk to parents about their relationship difficulties.
Introduction
Relationships are not always smooth, particularly when faced with the pressures of modern life. It is often children who suffer the most when parents argue, though relationship problems can impact on parental mental health too. Verbal aggression in relationships, in particular, can lead to problems such as emotional and behavioural difficulties in children, problems settling and achieving at school, or poor physical health (Reynolds et al, 2014).
There is some evidence that parenting programmes (e.g. Webster-Stratton & Reid, 2003) support parents to resolve some of their difficulties, but most people do not seek help or only do so as a last resort. One possible solution is to train practitioners working with children and families to routinely offer support with relationships.
Coleman et al (2014) evaluated a training programme named ‘Relationship support: An early intervention’ with Sure Start Children’s Centre (SSCC) staff. The training used a blended model of learning which involved an online introduction to the Brief Encounters® model followed by a one-day workshop facilitated by two skilled practitioners. The aims of the programme are to recognise relationship difficulties, respond using active listening and solution-focused techniques, and review the need for further support. The training is designed to be implemented alongside practitioners’ usual work.
Methods
The researchers recruited 237 SSCCs and divided them at random into an intervention group (119 SSCCs and 351 practitioners) and a control group (118 SSCCs and 356 practitioners). The intervention group received the training and the control group did not receive it until the completion of the study six months later.
To evaluate the effect of the training on practitioners, the researchers used an outcome questionnaire about three months after the intervention group received their training, but before the control group received theirs. They asked different questions than in their pre-training registration questionnaire as they did not want to prime respondents about the questions which may make it difficult to ascertain real changes. The self-completed outcome questionnaire focused on practitioners reporting about their:
- ability to recognise relationship difficulties;
- ability to talk to parents about their relationship difficulties;
- confidence in referring parents for further support;
- confidence in helping the parent;
- likelihood that they would offer similar support in the future.
Findings
The researchers found significantly greater confidence amongst practitioners in the intervention group in each of the five outcome domains (except no. 1 above) than those in the control group. Additional analysis found that length of prior experience of working with children and families, types of practitioner and initial level of confidence was not associated with outcomes, though no data was presented on this. Although this was a cluster randomised controlled trial, the data was analysed at the level of an individual and it was unclear how clustering by SSCC was dealt with in the analysis.
Conclusion
This study found that SSCC practitioners’ confidence in supporting people with relationship problems is higher for those who have received training than those who have not. It is possible that the training could be extended to GPs or volunteers to increase the availability of relationship support.
Limitations
The findings of this study need to be treated with some caution.
Firstly, the outcomes measured were self-reported by practitioners and there is no evidence that this translated into their practice. Those who received training may have reported an increase in their confidence and skills, but we do not know if this impacted on their actual practice.
Secondly, we do not know if parents perceived that practitioners were more supportive or actually helped them to resolve their relationship difficulties. More research is required to investigate this.
Thirdly, as there were no baseline measures taken, it is difficult to know for certain that those in the intervention group did not already have greater confidence than those in the control group. The researchers argue that a large sample size, their process of randomisation and the lack of differences between the groups at baseline (though they did not present their data on this) suggest that the outcomes at baseline were similar, but this was not measured.
Fourthly, the outcome measures were piloted in a small group of practitioners, but their reliability and validity was not established. Most randomised controlled trials use outcome measures which have been proven to reliably measure what they set out to measure so it must be noted as a limitation that this one did not.
Finally, the researchers did not use a standard ‘intention to treat’ analysis which includes data from all participants, irrespective of whether they completed the study or not. As they did not have baseline outcome data of participants they were unable to do this and were only able to make comparisons using the data they received.
Randomised controlled trials are rare in social care and they have the potential to produce evidence of the effectiveness of interventions with minimal bias. This study has found that practitioners’ confidence in providing relationship support appears to grow with training, but more research is required to investigate the outcomes of this for parents and their children.
References
Coleman, L., Houlston, C., Casey, P., Purdon, S. &Bryson, C. (2014) A randomised controlled trial of a relationship support training programme for frontline practitioners working with families, Families, Relationships and Societies, doi.org/10.1332/204674313X13872824892033
Reynolds, J., Houlston, C., Coleman, L. & Harold, G. (2014) Parental conflict: Outcomes and interventions for children and families, Bristol: Policy Press
Webster-Stratton, C. & Reid, M. (2003) The incredible years, parent, teacher and child intervention: targeting multiple areas of risk for a young child with pervasive conduct problems using a flexible, manualized treatment program, Cognitive and Behavioral Practice 8, 4, 377–86