Pragmatic measures in paediatric psychopharmacology--are we getting it right?

Recent changes in legislation have stimulated a new wave of interest and activity in paediatric psychopharmacology. This increased activity has coincided with a recognition that pragmatic measures of outcome such as those that tap into impairment and health related quality of life (HRQOL) have the potential to add considerably to the traditional symptom based measures of outcome. There are however considerable methodological issues associated with these types of measure, and these are made more complex when they are applied in mental health and paediatric settings. There is a clear need for the continued development of valid and reliable measures of HRQOL and impairment that are fit for purpose for use in clinical trials. Other more specific issues that need to be considered, and which all require further investigation include those relating to; age, self versus proxy ratings, contextual issues, generic versus disorder specific measures and definitions of clinically meaningful change. Most of the child and adolescent mental health trials that have included pragmatic measures have been conducted in ADHD samples, have been industry sponsored, use only parent ratings and focus on one drug (atomoxetine). Taken together they do however suggest that pharmacological treatments can impact positively on impairment and HRQOL, although with smaller effect sizes than is seen for symptom reduction. Further studies, across a wider range of disorders and treatments with multiple measures and multiple raters, are to be encouraged. In addition to reporting the basic outcomes from these studies researchers should use these data to improve the measurement models and refine both the measures and the trial designs.