Estimating effectiveness of school-based counselling: using data from controlled trials to predict improvement over non-intervention change

Background: There is a growing body of data to show that participation in school-based counselling is associated with significant reductions in psychological distress. However, this cannot be taken as evidence that school-based counselling is effective, as improvements may have happened without the intervention. Aims: The purpose of this study was to develop a method of estimating the amount of ‘natural’ change that might be expected in young people who would attend school-based counselling, such that the effects of the intervention over and above this amount could be identified. Method: Young Person's CORE (YP-CORE) scores from 74 participants allocated to waiting list control conditions in four pilot trials of school-based counselling in the UK were re-analysed using regression models, and a formula was found for estimating the outcomes for young people were they not to receive counselling. This was termed their Estimated Non-intervention Outcome (ENO), and could then be compared against the young person's Actual Outcome (AO), to give an estimated intervention effect (EIE). Results: The formula for the ENO was 4.17 + 0.64 × baseline score. Using this, we calculated a mean EIE for 256 young people in a cohort evaluation study of school-based counselling, which showed that the counselling was associated with large and significantly greater change than would be expected without the intervention (Cohen's d = 0.91). Discussion: The method presented in this paper is a simple means for improving the accuracy of estimations of treatment effectiveness, helping to adjust for changes due to spontaneous recovery and other non-treatment effects.