There was no evidence that the use of MTFC-A resulted in better outcomes than usual care. The intervention may be more beneficial for young people with antisocial behaviour but less beneficial than usual treatment for those without.

The present study investigated the referential communication skills of children with imaginary companions (ICs). Twenty-two children with ICs aged between 4 and 6 years were compared to 22 children without ICs (NICs). The children were matched for age, gender, birth order, number of siblings, and parental education. All children completed the Test of Referential Communication (Camaioni, Ercolani & Lloyd, 1995). The results showed that the children with ICs performed better than the children without ICs on the speaker component of the task. In particular, the IC children were better able to identify a specific referent to their interlocutor than were the NIC children. Furthermore, the IC children described less redundant features of the target picture than did the NIC children. The children did not differ in the listening comprehension component of the task. Overall, the results suggest that the IC children had a better understanding of their interlocutor's information requirements in conversation. The role of pretend play in the development of communicative competence is discussed in light of these results.

The letter from Rodger et al uses our editorial to rehearse the well-worn arguments of the small group of so-called 'critical psychiatrists' who are active and vocal in criticising core aspects of the practice of psychiatry as a medical subspecialty underpinned by science. The views expressed in the letter are mainly tangential to the views we expressed in our editorial and the authors have made assumptions and accusations that are unsupported by our text. We are very keen to encourage informed and constructive debate to advance patient care and mental health. However, it is important to make a distinction between the freedom that is properly enjoyed in academic debate and the responsibilities that come with professional practice. At present, those who work as psychiatrists are expected to practise in accordance with evidence-based standards. The standards we adhere to will of course change over time as the evidence base develops. This is expected by patients and colleagues and required by regulators. We continue to believe that our patients are best served by seeing psychiatrists who are trained to make a thorough assessment, come to a diagnosis and shared formulation with the patient of their problems and use this to draw up an evidence-based management plan. It seems strange to us that this should be surprising, contentious or upsetting to the authors of the letter.