Not so long ago I attended a conference with colleagues in old age psychiatry. Shortly before lunch the Chairman of the morning session stood up, ostensibly to praise the preceding speaker for his lecture. His words went something like: ‘Thank you Dr X. I'm sure we'd agree that your excellent talk has given us all thought for food.’

Such startling betrayals of the unconscious are, of course, rare. Freudian analysis of dreams is also an uncommon pastime in our everyday clinical practice. Nevertheless, in old age psychiatry evidence for unconscious activity is to be found, for example, in psychotic processes, in the fragmented behaviour of people with dementia and in interactions between patients and staff.

Old age psychiatrists are drawn into all aspects of ageing. A detailed consideration of each patient's predicament requires time. Garner (2002, this issue) argues that psychodynamic work is time well spent, particularly when it acknowledges the multiplicity of our practice. Hardly any patients of mine will be suited to 50-minute sessions on a psychoanalytic couch. Many cannot see or hear properly; others do not know who we or their loved ones are. Few will ask for psychological treatments of any sort. And yet, all of our patients are searching for something, usually in the form of a relationship. We therefore have to adapt psychoanalysis to fit the modern world and the old people who inhabit it.

We know that, by and large, psychotherapists are reluctant to take on elderly clients (Reference MurphyMurphy, 2000). Garner refers to Freud's somewhat dismal view of old age. His own was one of considerable suffering, both from physical pain and self-reproach. Sixty years on from Freud's death we still have much to learn about the normal psychological sequelae of ageing. Old age psychiatrists encounter the unsuccessful adaptations to later life in our anxious and depressed patients. Psychodynamic thinking strives to apply meaning to our patients' symptoms, thereby potentially supplementing our understanding.

In work with older people a fundamental concern is the patient's fear of losing independence. All around are threats to existing autonomy. Whether the patient can successfully negotiate these will be determined by several factors. One is to be found in the patient's early developmental history. A second is the fact that young therapists are likely to overvalue the desirability of independence. Any signs of a patient developing an ‘unhealthy’ attachment to a younger therapist may be vigorously resisted. A successful acceptance of old age is the tolerance of dependence. The National Service Framework for Older People (Department of Health, 2001), reflecting a society in which independence is prized and dependence regarded with contempt (see Reference BellBell, 1996), unwittingly reinforces the ‘dependence is bad, independence is good’ paradigm. Sadly, the dependability of National Health Services ‘from cradle to grave’ is now open to serious doubt. Anxieties in our elderly patients may be fuelled by real fears that they will indeed be dealt with mindlessly.

The road to psychological well-being will be smoother if staff providing care for the elderly are of high morale. Those who find themselves in the forefront of caring are likely to be better placed to contain the projections of disturbed patients if they in turn feel cared for (Reference Zagier Roberts, Obholzer and Zagier RobertsZagier Roberts, 1994). One aspect of this is the opportunity for supervision. Garner emphasises the importance of supervision in psychodynamic work with older people. To use a battlefield analogy, troops who have a clear belief and understanding of their task, and who are suitably equipped, are better able to perform than those who are blindly sent over the top. Supervision demands a reflective space away from the front line. Mr B., whom Garner offers us, poses a familiar problem of his narcissism alienating the home care services he apparently needs. In supervision the reasons why he was so phobic of these were unravelled.

A cynical patient once philosophised to me that ‘Life is a terminal illness.’ This raised the question of what is outcome in psychotherapy – longevity, symptomatic relief or more nebulous improvement in quality of life? Psychodynamic therapy is essentially a preventive treatment. Just as the cigarette smoker who develops lung cancer may need surgery, so the anxious patient who is bereaved and becomes dangerously depressed may need electroconvulsive therapy. Once better, however, he may be interested in learning more about the reasons for his depression through psychotherapy. In the same way that we cannot predict which cigarette smoker will develop bronchial carcinoma, we do not know which psychologically vulnerable patient will become depressed in later life.

Attempts to compare the efficacy of psychodynamic therapy with other psychological or drug treatments may be no more sensible than comparing the usefulness of a health promotion campaign with a thoracic surgeon's ability to remove tumours. Furthermore, we do not know when to measure outcome, since the internal processing and subsequent ‘benefits’ might not become evident for years. Self-reporting has some merits, but those who spend time (and perhaps money) in psychotherapy are likely to proclaim that therapy works. Psychotherapy can be harmful and a wise move for some patients might be to steer clear of having their fragile defences demolished. Despite the fact that older people are referred much less often than their younger counterparts, there is no evidence that they are less likely to benefit. So the conclusion must be that Freud's prejudice is still with us.

Meanwhile in old age psychiatry, the best way of employing a psychodynamic therapist may be to enhance the skills of the multi-disciplinary team. Here Garner suggests that a valid outcome measure might be that of staff turnover. In overstretched services, deeper psychotherapeutic work will be indicated for only few patients.

Ong et al's (1987) successful attempt at group therapy to prevent relapse in older patients with depression, which Garner quotes, has been supplemented by the findings of Blanchard (1997). In both instances the anticipated reliability of human contact, rather than psychoanalytic expertise, appeared to be important. We now need a major study in this field to determine what psychological intervention (in the out-patient clinic, day hospital or by way of home visiting) is most useful in protecting known vulnerable patients from relapse and readmission.