Objective - The aim of this article is to explain the current status of deinstitutionalisation and of community care development in the UK. Design — Literature review of articles and reports on deinstitutionalisation. Setting — Articles included in the review related either to the whole of the UK, to England and Wales, or to a specific area such as London. Main outcome measures — The review was carried out pertaining to the question of the extent to which community care can or should take over the functions of the asylum, with particular reference to those at risk of behaving violently. These functions include those that are manifest, or explicit, and those that are latent, or unintended but implicit (Bachrach, 1976). Results — The example of patients at risk of violent behaviour is one that highlights the continuing relevance of both these sets of functions, which are argued to be exerting a powerful influence on the processes of asylum closure and community care development. This influence is seen in delayed asylum closure, transinstitutionalisation (the shift of some patients from asylums to other institutions), and the institutionalisation of aspects of community care. Conclusions — Both the manifest and the latent functions of asylums must be acknowledged by those involved in planning community care; where it is felt desirable that community care does not take over certain functions, the consequences of this must be anticipated so that they can be prevented or dealt with in other ways.

Objective — The medical interview has important diagnostic and therapeutic functions and requires the integration of doctor-centred and patient-centred interviewing techniques to collect accurate and complete biopsychosocial data from the patient. Analysis of the interaction between patient and doctors which occur during the medical interview allow to evaluate physicians' interview techniques and to eventually improve them. Objective — 1. To review different Interaction Analysis Systems (IAS) used to describe doctor-patient communication in terms of clinical relevance, observational strategy, reliability and behavioural and verbal contents. 2. To critically evaluate these IASs on the basis of their relevant research outcomes. Method — Previous reviews on interaction and keywords for Medline research (HealthGate) listed above were utilised to collect the relevant literature. Results — Seventeen classification systems were identified and ten were discussed in a chronological order. Starting from a general sociological or psycholinguistic approach, the IASs over the years have became more specific and detailed, focusing more and more on the medical interview and on specific topics, such as cancer or hospital medical consultations. Conclusions — When studying interactions in general practice medicine, it is important to define the significant units of interaction which allow to identify a «patient-centred approach», since this is relevant not only for obtaining reliable and complete medical and social data, but also for the recognition of patients with emotional disorders and their correct diagnosis. Listening to the patient and facilitating the expression of emotions is an important aspect of patient education too, as patients learn that talking about psychological problems to their physician is appropriate and may be therapeutic.