Introduction

My task is to provide a general introduction to research into ways of defining and measuring the needs of people who are mentally ill and into the efficacy of the various methods that society has used to try to meet those needs. I shall not be concerned with the special needs of mentally ill offenders but the same principles of investigation apply. It is convenient to divide the subject matter into three parts: epidemiology, needs assessment and service evaluation.

Before starting I will specify a particular meaning for three key terms: ‘care’, care ‘agents’ and service ‘settings’. Understanding the differences between them is basic to clarity of thought about evaluation. ‘Care’ includes all forms of personal interventions, from medication through counselling to welfare support. ‘Agents’ are the people, professional and informal, who provide the care. ‘Settings’ comprise the various structures in which the care is given. All three terms are sometimes characterized as services, so each should be understood as being in quotation marks in what follows, in order to emphasize their differential significance.

The epidemiology of need

An epidemiological perspective is essential for the understanding of any disorder. David Mechanic pointed out, when he first understood what epidemiology meant, that this was what many sociologists thought they had been doing most of their lives.

In that spirit I will quote some figures about crime, produced by a biostatistician, Morton Kramer (1989), for a book about epidemiology (see Table 7.1). They show the number of people, per 100 000 US population, who had been resident in institutions on decennial census days from 1950 to 1980. The overall ratio has remained remarkably steady but the same cannot be said of its components. While that for mental institutions has dropped to about one quarter of the 1950 figure, the ratio for the elderly and dependent has trebled. In 1950, the ratio for mental hospitals was the highest, in 1980 it was lowest, lower than that for prisons.

The interactional dynamics of these groups cannot be determined from such figures (Kramer provides a sophisticated analysis and commentary) but they suggest interesting questions for our own country, which echo those considered in the papers by Drs Fowles and Bowden.