RT Journal Article SR Electronic T1 Characteristics of elderly people admitted to hospital, Part III homes, and sheltered housing. JF Epidemiology and Community Health JO Epidemiol Community Health FD BMJ Publishing Group Ltd SP 91 OP 95 DO 10.1136/jech.33.1.91 VO 33 IS 1 A1 Alexander, J R A1 Eldon, A YR 1979 UL http://jech.bmj.com/content/33/1/91.abstract AB Approximately 100 consecutive admissions in Southampton to sheltered housing, to Part III homes, and to long-stay geriatric hospital beds in 1975 and 1976 were selected and data were collected by interview with patients and staff. There was a total of 299 admissions in the final sample. The sheltered housing group were younger and more often married than those admitted to Part III homes, who were more often married but only slightly younger than the new geriatric patients. The three groups were separated on dependency measures with almost no overlap between the sheltered housing and the other groups, but a little more evidence of overlap and interchange between Part III and hospital admissions. The patients in hospital whose dependency scores were more typical of Part III residents were almost all special cases and this was found to be true of other apparently misplaced individuals. Most of the sheltered housing admissions were suffering from heart, chest, and musculoskeletal complaints, while new Part III residents had more psychiatric disturbance; cerebrovascular and psychiatric problems occurred most often in the hospital group. The three groups of admissions form distinct categories and it is arguable that this is a concomitant of underprovision. This also shows that there is no great measure of misplacement on the basis of the current norm, although the suitability of this norm in sheltered housing is open to question.