Distinguishing between health and social aspects of institutional care for frail elderly people remains confusing and contentious in the UK, even though health and community care reforms specifically sought to avoid such problems. Nursing home care in the UK has developed rapidly to 170 000 nursing home beds. Nursing home care now costs significantly more than the total committed to primary care. Some health authorities maintain a large commitment to long-term care provision in hospitals and contract nursing home beds; others have divested themselves of providing long-term care, becoming dependent on means-tested nursing home care provided by social services. Where 'free' health service contract nursing home beds are adjacent to social services means-tested placements the inconsistencies become even more perverse. Distinguishing between health and social patients/clients/supplicants on the basis of health or social care need is often impossible. Health Service responsibilities to support care home residents will become more onerous as cost pressures define the limits of social care for individuals and social services producers alike. There is, in addition, an unresolved tension between the professions associated with the care of elderly patients, who are seeking clear standards, and the hostile purchasing environment which fails to value the cost benefit, let alone the health gain of planned care management in deference to the Mammonism of treatment episodes.

Reports on the death of the district general hospital (DGH) are premature. Despite frequent predictions of doom these hospitals are in good shape and over the years have shown a continuing readiness to adapt to changes in medical practice, technological advances and political fashion. The development of community services is not a threat to the DGH, but rather an opportunity to evaluate carefully what the DGH does best.

Hospitals are disappearing fast and have been doing so for many years almost without our noticing. Figure 1 shows the number of hospital beds and the number of discharges and deaths from hospital in the UK over the past 35 years.

Homeostasis is the ability of a living organism to control its internal environment despite fluctuations in the external environment. Claude Bernard proposed the concept of 'milieu interieur'. The hypothesis rests on the belief that cells need a precisely defined local environment in which to function optimally. The maintenance of the micro-environment is dependent on integration of the systems of the organism (body), including behaviour, to protect the cell(s) from external stressors. When the environment changes, the body needs to sense the alteration, then set in motion mechanisms which mitigate the change and restore the previous balance. If it fails to do this, the organism will ultimately die as systems fail.

Urinary incontinence has been defined by the International Continence Society as 'the involuntary loss of urine which is a social and hygienic problem and can be objectively demonstrated'. This is a useful definition, although the subjective component leaves it open to variable interpretation, with consequent difficulty in establishing its prevalence. This is further contributed to by prevalence differences between populations and inconsistency of questions used. Thus, for example, studies have used such diverse questions as 'do you have difficulty in controlling your water?', 'do you have involuntary loss of urine on two or more occasions per month?', 'have you had involuntary loss of urine at any time in the previous year?' or 'have you ever suffered from bladder problems?'.

Incontinence is a common problem and the incidence increases with age. Approximately 85% of all cases of incontinence occur in women, essentially because bladder neck competence and pelvic floor/urethral sphincter activity is more readily compromised than in the male. Incontinence is an important cause of admission to and a common finding in nursing homes for the elderly but is not an inevitable sequela of aging.

In recent years, psychiatrists have improvedtheir accuracy and efficiency in diagnosingdelirium, particularly in hospitalized elderly patients. Early and accurate diagnosis is essential as unrecognized and untreated delirium isassociated with longer length of hospital stay and greater expense, accelerated functional decline, an increase in nursing home placements, persistent cognitive deficits, and elevated rates of mortality.

Continence promotion includes all nonsurgical therapies that enable a patient to recover continence. Examples include bladder training, pelvic floor exercises, anticholinergic drugs, oestrogen, electrical therapies, vaginal cones and intermittent self-catheterization. Implicit in the use of all of these therapies is a detailed history and examination to diagnose any cause for the symptoms. Continence promotion is most closely identified with the role of the nurse, in co-operation with the doctor, in recognizing medical and surgical causes of incontinence.

Dying is the one common and certain event in life but it is a difficult process to study. Official statistics give information about the point of death; studies covering random samples of deaths are inevitably retrospective and based on the accounts of those who knew the person who died; other studies are selective and generally focused on people dying from cancer, so they include a relatively small proportion of old people. Yet, in developed countries, deaths are concentrated among older people. In England and Wales in 1994 83.6% of all deaths were of people aged 65 or over; 59.0% were 75 or more.

The term 'graduate' (hereafter without quotation marks) has been used to describe the following groups of patients:

1) All those patients with enduring psychiatric illness who have graduated to old age (usually defined as age over 65 years);

2) 'Patients who entered hospitals for the mentally ill before modern methods of treatment were available and have grown old in them after many years there.'

3) Specifically, long-stay patients with schizophrenia over the age of 65 years who came into continuous psychiatric contact well before that age.