Authors' reply

We would like to thank Dr Mushtaq and Dr Helal for their letter, and welcome the opportunity to clarify the points they have raised. With regard to the study design, since 2002, data on emergency presentations have been prospectively collected at the time of presentation and recorded on a secure database within the hospital network. Access to this information is regulated, and in 2008 we sought and received ethical approval to access and analyse these data retrospectively for the purpose of this study. No data other than those recorded at the time of presentation were included in the study.

During the study period there were no direct admissions from the emergency department to specialist child and adolescent psychiatric in-patient units. This finding most likely reflects the significant lack of capacity within such units as discussed in the paper. Of the subset from 2006 for which data on onward referral were collected (n = 278), 20 were referred onwards for in-patient psychiatric assessment. Presenting complaints for those referred were self-harm, suicidal ideation and psychosis.

We agree on the many benefits of interdisciplinary liaison and acknowledge the clinical and resource implications. Indeed, the need to review the efficacy and value for money of services we deliver was a significant factor in our decision to conduct this study. We have presented the findings to all the involved service providers, to encourage awareness of the demand and the rationale for ongoing service provision. Although a cost–benefit analysis was outside our study design, possible cost savings attributable to the model of service provision have been considered in the study discussion. Finally, within a national context in Ireland, improving child and adolescent mental health and reducing suicide are both key performance indicators for our health services, thereby supporting the ongoing provision of services.

We would like to acknowledge the study of Hillen & Szaniecki, and that this study also addresses many aspects of the service model and demand for out-of-hours services. This paper's publication coincided with the timing of our original submission, and the lead author apologises that this study was not located at the time of revision of the paper.