A cohort through the revolving door.

In this study a cohort of 1722 patients first admitted to all three psychiatric facilities in a small urban area were followed for up to seven years. Seventy-nine percent had non-psychotic diagnoses. Of the total cohort, 7% (n = 120) were readmitted more than three times--these patients were more likely to be male, young and unemployed, but of similar education level to age-sex matched single admission control. Fifty three per cent (n = 49) of this group were not psychotic (most were alcoholics or personality disorders) and again were more likely to be males, young and unemployed than age-sex matched non-psychotic single admission controls. Efforts to further evaluate the characteristics of the multiplyreadmitted cohort were thwarted for a number of reasons. Information about individuals within a system are essential for further epidemiological study but require a better data base and prospective monitoring of patient careers.