![[ Visit Client Website ]](images/banner.gif)
Motivational Interviewing (MI) is a counselling style aimed at supporting behaviour change. There is a strong evidence base for its effectiveness, and studies suggesting it may be helpful in child protection. This randomized controlled trial was carried out in a UK local authority with workers who carry out statutory child protection functions of assessment and intervention. The trial tests the impact of a skills development package of training and supervision in MI on the skills of social workers, engagement of parents and key family outcome measures. The primary hypotheses were that the package would improve the MI skills of workers, and that this would result in increased parental engagement. The impact on the use of authority and on other parent reported outcomes are also reported.
Methods:
The package involved 4 days of training and 15 supervision sessions. All social workers and line managers in the local authority were cluster randomized to receive the package prior to or after the study (n=59). For 7 months all allocated families were randomized to MI or control conditions (n=610). Exclusions – primarily due to limited contact – resulted in 284 families entering the study sample. Interviews between worker and family were recorded for 131 families (46%), and research interviews shortly after allocation (T1; 132; 46%) and 20 weeks later (T2;104; 37%). Outcomes reported are worker skill in MI (Motivational Interviewing Treatment Integrity code (MITI) – a 5 point scale, with 3.5 or higher indicating MI skill), additional codes relating to use of authority (purposefulness, focus on child and clarity about concerns), and parent reported measures of: engagement (Working Alliance Inventory (WAI-C)), achievement of goals in work (Goal Attainment Scaling (GAS)), rating of family life and child welfare. Analysis used multilevel modelling to allow for grouping at the level of worker.
Results:
The MI skills of trained workers were higher (2.91 vs 2.49; p<0.05), as was their empathy (3.03 vs 2.34; p<0.05) though on neither measure was skillful MI the norm. Measures of authority showed no statistical differences. There were no significant between-group differences in engagement, goal achievement, parent rating of family life or child welfare at either T1 or T2. In addition the relationship between worker skill and outcome across the whole sample was studied. Weak relationships with few statistically significant relationships were found between any skill and outcomes. However, skills had moderate correlations with outcomes for families with 3-9 visits and strong correlations where there were 10+ visits.
Conclusions and Implications:
The skills development programme demonstrated limited improvement in MI skills. The impact on outcomes was nugatory. We draw three conclusions. First, the relationship between skills and outcomes needs further study. Second, delivering evidence based interventions to a high standard is difficult and requires more than training: whole agency commitment is required. Finally, further work is required on how to combine an approach such as MI with risk management and appropriate use of authority.