Many patients depend on the care and support provided by relatives and friends as much as or more than on the help of services. In order to investigate this role and the burden that it might entail, interviews were sought with those who actually played a practical caring role. Members of the research team asked people in the series and staff whether (a) the patient shared a household with someone, or planned to do so in the near future if living in a residential unit; or (b) the patient regularly visited the relative; or (c) someone regularly helped with tasks such as laundry, washing up or cooking. There were 79 such helpers, all but two of whom turned out to be relatives. One exception was a landlady who had for years provided a very disabled patient with extensive motherly support. The other was a friend who visited the patient every evening in hospital. He used to visit her every day when she was living at home, and helped her round the house when she could not manage. For convenience the term ‘relative’ is used to include these two.

There is a considerable variation in the amount and type of care and supervision provided for people in the survey population. At one end of the spectrum, the in-patient ward provides skilled 24-hour medical and nursing care; at the other extreme, some homes are simply providing sheltered housing for ex-patients who would find it very difficult to obtain private rented accommodation in London. Although the sample is defined in terms of service utilization, it covers a broad range of types of treatment, care and social support.

There is a well-established tradition in standardized evaluation of the symptomatology and the social performance of psychiatric patients (Wing et al. 1974; WHO, 1983b; Platt, 1983) together with instruments for describing the pattern of contacts with services, like Psychiatric Case Registers (Wing & Hailey, 1972; ten Horn et al. 1986). Interest in a systematic assessment procedure for recording which action should be taken by services in the presence of a problem is more recent. Instruments for these evaluations are still experimental.