Development of a statistical approach to classifying treatment response in individual children with ADHD

Abstract

The use of cognitive tests as measures of treatment response in individual children with ADHD has not been adequately evaluated or commonly applied by clinicians. This is most likely due to a lack of suitable assessment tasks as well as clinicians' limited awareness of the appropriate statistical techniques for analysing cognitive change in individuals. This study investigated the application of statistical decision rules to the cognitive and behavioural measures of individual children with ADHD in order to classify a significant, positive response to medication. The previously reported data of 14 children with ADHD (combined type; ADHD‐CT) were re‐analysed to investigate changes in function following a low‐ and high‐dose of stimulant medication (2.5 mg and 7.5 mg dexamphetamine, respectively). The performances of 14 age‐, gender‐ and IQ‐matched controls was also analysed to provide an estimate of the false‐positive classification rate. Overall, the decision rule yielded a high sensitivity and high specificity to treatment response. In the high‐dose condition, 71% of children with ADHD demonstrated improvements in both cognitive and behavioural function. This study demonstrates an evidence‐based approach to evaluating concurrent cognitive and behavioural improvement in individual children with ADHD following treatment with stimulant medication. Copyright © 2004 John Wiley & Sons, Ltd.