We welcome the debate initiated by Craddock et al ^{Reference Craddock, Antebi, Attenburrow, Bailey, Carson, Cowen, Craddock, Eagles, Ebmeier, Farmer, Fazel, Ferrier, Geddes, Goodwin, Harrison, Hawton, Hunter, Jacoby, Jones, Keedwell, Kerr, Mackin, McGuffin, MacIntyre, McConville, Mountain, O'Donovan, Owen, Oyebode, Phillips, Price, Shah, Smith, Walters, Woodruff, Young and Zammit1} and agree that the role of the psychiatrist is key to the delivery of high-quality services, and may be currently threatened. However, we believe that their proposals would be restrictive and counterproductive. If the psychiatrist has to assess all those referred to secondary services, access to such care would be restricted increasing the burden of unmet need. To deploy services effectively the psychiatrist should assess only those who require their direct input, freeing-up the psychiatrist to have an overview of the clinical work of all the team members: from allocation, initial assessment and management through to discharge as well as a training and development role. This was the ambition of New Ways of Working, ² although not realised in its implementation, partly due to the lack of training of the other team members for their extended role and the development of teams without adequate medical input for them to work effectively. These issues should be addressed directly. To return to a position of the consultant taking full clinical responsibility for all the team's case-load is not only retrogressive, but unworkable. Allowing staff to take the personal responsibility that they already have improves the quality of care delivered and works best when the consultant is readily available for consultation and review rather than running over-booked out-patient clinics as occurred hitherto.

The authors, in focusing on the importance of biomedical methods, appear to underestimate the important contribution of other approaches, psychological and social, to psychiatry, which have been shown to lead to effective interventions. The profession of medicine is changing, with our physician colleagues taking up many of the challenges of a psychosocial approach. We appreciate that psychiatry is a medical specialty and that psychiatrists are physicians who have an expertise in psycho- and socio-dynamics in their broadest forms. In reconsidering our roles and values on the 200th anniversary of our specialty we should consider what we should be doing in the 21st century and how we can adapt to this. The mental health services have far to go to improve standards, quality and the delivery of evidence-based practice. The users of these services should expect to encounter experts in the field of mental disorders, but these experts need a wide range of skills and knowledge to guide assessment (including diagnosis) and management (including treatment). But, in addition, they need to utilise the ideas of recovery ^{3,Reference Shepherd, Boardman and Slade4} (a term regrettably omitted from Cradock et al's paper) to negotiate and facilitate the types of goals and outcomes valued by service users and to allow people with mental disorders to participate more fully in their communities and society.

It is important not to polarise this crucial debate, nor to retreat into restrictive medical modes of thinking. To meet the challenges of the 21st century will mean an important shift in our ways of working, which can be of enormous value to our professional roles and to the service users that we work with.