The point raised by Mr Lawton-Smith is of the utmost importance, but we found no published randomised studies that considered patients' perspectives as an entry point for interventions to improve compliance. To fill this gap, our group has recently completed a qualitative study involving patients, families and therapists to identify their concerns with adherence and to design effective interventions. We agree with Mr Lawton-Smith that much remains to be done to adapt research methodologies and clinical practices to the needs expressed by people with mental illnesses.

Vergouwen & Bakker incorrectly attribute our statement, ‘the important relationship between adherence and outcome of treatment has been evaluated only in one study’ to randomised interventions, when it referred to descriptive studies, both in the Results and Discussion sections. Out of the 14 randomised interventions we reviewed, only five reported data on response which could be extracted. In addition, the design applied by the five studies made it impossible in our review to explore the relationship between intervention and response.