Institute of Health Services Research, Peninsula College of Medicine and Dentristry, Exeter
School of Psychology, Bangor University, Gwynedd, Wales
Department of Epidemiology and Public Health, London School or Hygiene and Tropical Medicine
Department of Child and Adolescent Psychiatry, Kings College London, Institute of Psychiatry, UK
Correspondence: Tamsin Ford, Institute of Health Services Research, Peninsula College of Medicine and Dentistry, St Luke's Campus, Heavitree Road, Exeter EX2 8UT. Email: tamsin.ford{at}pms.ac.uk
R.G. and A.G. are directors and part owners of Youthinmind, which provides the www.sdqinfo.com website as a public service in order to make the SDQ freely available in many languages for non-profit use and to publish SDQ norms and the Added Value Score formula.
The British Child and Adolescent Mental Health Survey 2004 was funded by the Department of Health; the Health Foundation funded the trial of parent training and T.F. wrote this paper while supported on an MRC clinician scientist fellowship. None of these funders had any involvement in the design or analysis of this paper or the construction of the Added Value Score.
Background
Routine outcome monitoring may improve clinical services but remains controversial, partly because the absence of a control group makes interpretation difficult.
Aims
To test a computer algorithm designed to allow practitioners to compare their outcomes with epidemiological data from a population sample against data from a randomised controlled trial, to see if it accurately predicted the trial's outcome.
Method
We developed an `added value' score using epidemiological data on the Strengths and Difficulties Questionnaire (SDQ). We tested whether it correctly predicted the effect size for the control and intervention groups in a randomised controlled trial.
Results
As compared with the a priori expectation of zero, the Added Value Score applied to the control group predicted an effect size of –0.03 (95% CI –0.30 to 0.24, t = 0.2, P = 0.8). As compared with the trial estimate of 0.37, the Added Value Score applied to the intervention group predicted an effect size of 0.36 (95% CI 0.12 to 0.60, t = 0.1, P = 0.9).
Conclusions
Our findings provide preliminary support for the validity of this approach as one tool in the evaluation of interventions with groups of children who have, or are at high risk of developing, significant psychopathology.