There have been relatively few studies of hospital hostels (HMSO, 1991) and the present publication, by a nurse, is therefore welcome. Emerson's study reports on the progress of 65 residents of four contrasting hospital hostels, using a quasiexperimental design, i.e. using each patient as his/her own control. Fifty of the subjects were studied on four occasions over 12 months, the assumption being that any changes were owing to the effects of living in the hostel. The study's aims were to identify the causes of any improvements observed over time, to compare outcomes in terms of discharges and transfers and to obtain the views of staff, carers and residents about the hostels.

The four hostels differed in their internal environment, their admission policies, the stated degree of dependency of the patients and the degree of restrictiveness of the nursing policies adopted. Two mainly had residents engaged in gainful occupations outside the hostel, one encouraged voluntary light work around the house rewarded with ‘incentive payments’, while the fourth arranged few activities and residents were left with lots of time on their hands. In two hostels residents cooked their own food under the supervision of the staff, another had food cooked in a nearby hospital and served by a catering assistant and one had meals cooked by the staff.

Few details are given of the clinical status of the patients, and although about 70% had hospital diagnoses of schizophrenia, one cannot tell the length of their illnesses or their hospital age. The Present State Examination is mentioned in the ‘Methods’ chapter [p. 25] as an instrument for obtaining “the client's own judgements on their outcome”, but it is not mentioned again, and one cannot tell whether or not it was administered. Since one of the hostels allowed admissions from the community, one was for new long-stay and two were predominantly for old long-stay, it is probably safe to assume that the patients were not comparable. Thus, it is not clear whether they would all have been suitable for the same kind of nursing care. However, they are probably a much less disabled group of patients than those cared for in Douglas House (Reference Hyde, Bridges and GoldbergHyde et al, 1987), and we were able to allow our patients a greater degree of autonomy than the most restrictive hostel described here.

Emerson's conclusions are that improvement in the residents — as measured by the Social Behaviour Schedule and the Social Role Performance Scale — was greatest in the hostel with the least restrictive policies. This hostel was designated a ‘rehabilitation hostel’, but it was one where most patients were engaged in outside gainful activities, where each resident cooked a substantial meal every week and where staff carried out domestic work themselves if the residents did not volunteer for it. If these patients were less severely ill than those in the other hostels their progressive improvement would be easier to understand. While I have no problems with his conclusion, it is not clear to me that it follows from the results. It is a great pity that the hostel studied in the pilot study was not included in the follow-up, since in my view only that hostel had really desirable policies.

This is a difficult book to read, as facts are presented in a rather diffuse way, but the conclusions are really very modest. My main impression was disappointment that the hostels studied had many features of mini-institutions and that nursing still has a way to go in adjusting its practices to the new system. The statistical treatment of the results leaves something to be desired, but the author's heart is in the right place.