We were alarmed to read the editorial on religion and mental health (Reference KOENIGKoenig, 2008). Some of the assertions are highly contentious, and we believe some of the recommendations for clinical practice are inappropriate. The invited commentary by the (former) President of the Royal College of Psychiatrists (Reference HOLLINSHollins, 2008) is cautious, but none the less seems to endorse Koenig's point of view. In doing so, Hollins lends certain credibility to Koenig's recommendations. Closer integration of religion and psychiatric practice is a key aspiration of an element within the Spirituality and Psychiatry Special Interest Group of the College. We believe that there is an urgent need for a serious debate on the implications of such attempts to shift the boundaries of psychiatry and the other mental health professions.

Koenig uses some statistics that are questionable. For example, the World Chrisitian Database may say that 1.4% of the British population are atheist, but the British Humanist Association website cites recent figures from the national census, a Home Office survey and a Market and Opinion Research International (MORI) poll ranging from 15.5% to 36% (www.humanism.org.uk/site/cms/contentChapterView.asp?chapter=309). However, it is Koenig's fundamental argument that is seriously flawed.

Koenig uses the rhetorical ploy of suggesting that religion is denigrated and under attack by psychiatrists. He states that psychiatry has traditionally regarded religion and spirituality as intrinsically pathological. We have been involved in mental healthcare in the UK since 1978 and none of us has ever known this to be suggested by a mental health professional. Koenig further states that there is a widespread psychiatric prejudice against religious faith and that psychiatrists commonly do not understand the role of religion in service users’ lives. However, the research that he cites can be interpreted as suggesting that psychiatrists, by and large, believe that religion can be both helpful and problematic to service users and that they enquire about religious matters when these are relevant. As the salience of religious issues will vary between service users, this seems to us to be the appropriate approach.

Our major concern about Koenig's paper is his suggestions for practice. No one could seriously challenge some of his assertions, for example that we should always respect people's religious or spiritual beliefs and that we should sometimes make referrals to or consult with appropriate priests or religious elders. However, these are well-established parts of routine practice. They are within the limits of existing codes of professional behaviour. Some of his other suggestions, however, constitute serious breaches of professional boundaries, for example:

1. • Psychiatrists should routinely take a detailed ‘spiritual history’, even from non-believers; Koening recommends that when the person resists this, the clinician should return to the task later. This seems to us to be intrusive and excessive. The insistence that even non-believers have a spiritual life shows a lack of respect for those who find meaning within beliefs that reject the transcendent and the supernatural.

2. • Some spiritual or religious beliefs should be supported and others challenged. This involves the application of the clinician's values, which is incompatible with the maintenance of an appropriate degree of therapeutic neutrality. It is unnecessary and inappropriate for clinicians to take a position on highly sensitive matters of personal conviction, such as the existence and nature of evil, the meaning of unanswered prayer and doctrinal intolerance of homosexuality.

3. • It is sometimes appropriate to pray with service users even when service user and psychiatrist do not share a faith. The introduction of a completely non-clinical activity carries a grave danger of blurring of therapeutic boundaries and creates ambiguity over the nature of the relationship.

We have personal experience of dealing with the adverse consequences of religious breaches of therapeutic boundaries. For the most part, these have been well-intentioned but ill-advised; for example, individuals who want to pray with psychiatrists at one point in their treatment can become persistently distressed over having done so when their mental state changes. We have encountered more worrying breaches of boundaries where clinicians have proselytised in the consulting room. Occasionally, we have encountered frankly narcissistic practice, where clinicians have been emboldened by the certainties of a charismatic faith and take the position that their personal beliefs and practices cannot be challenged because they are supported by a higher authority than secular professional ethics.

The problem with blurring the boundaries by inviting an apparently benign spirituality into the consulting room is that it makes it more difficult to prevent these abuses. Having moved the old boundary it is then very difficult to set a new one.

Psychiatrists will always have to understand service users who are of different gender, class, ethnicity, political beliefs and religious faith. Understanding their lives, the contexts they exist in and the resources that give them strength is a key skill in psychiatric practice (Reference POOLE and HIGGOPoole & Higgo, 2006). Religion can be an important source of comfort and healing, though it can also be a source of distress. Of course, it can be intertwined with psychotic symptoms. Spiritual matters, however, exist in a different domain from psychiatric practice. There are others in our communities who have a proper role in helping individuals spiritually and who can be an important source of advice to us. Quite apart from the obvious dangers inherent in confusing these roles, it is completely unnecessary to do so.

Psychiatry has done much to improve the lot of people with a mental illness, though it has also been guilty of some major historical errors. Our professional roles and professionalism are under sustained attack from a variety of sources (Reference POOLE and BHUGRAPoole & Bhugra, 2008). In order to resist these attacks, we need to be clear about our important and distinctive roles in helping those with a mental illness. Psychiatrists are essentially applied biopsychosocial scientists, who work within a clear set of humanitarian values and ethical principles in order to get alongside service users and facilitate their recovery from a mental illness. Psychiatry does not hold all the answers and other professions, agencies and individuals have different distinctive roles. Within psychiatry, we have to struggle with the internal threat of crude biological reductionism. Equally, if we break the boundaries of our legitimate expertise and become generic healers, we will have lost all usefulness and legitimacy.

Declaration of interest

The authors have a range of personal convictions including atheist, Buddhist, Methodist, Roman Catholic and non-denominational faith.