Long-term institutional residents: does the environment affect outcomes?

Quality in institutional care is frequently questioned, yet there has been little research on the effects of different environments on important outcomes such as disability and quality of life. These outcomes are difficult to measure, and randomised trials difficult to perform. An alternative to formal trials is to assess the effects of 'natural experiments' arising from service reforms. We have studied the relocation of 95 residents from the continuing-care wards of a hospital which closed to two different ward environments: one new and purpose built, the other refurbished maternity wards. Deaths and changes in disability over 12 months from pre-relocation measurements were recorded, using the Barthel Index and Clifton Assessment Procedure for the Elderly (CAPE) Behavioural Rating and Cognitive Assessment Scales. No advantage could be shown for the new wards in terms of changes in disability; in fact, Barthel scores improved more on the refurbished wards (the difference being 2.5 points at 4 months, p less than 0.0005). Changes in CAPE scores did not differ between the wards. Mortality at 1 year was 44%, with no significant difference between the ward types. Logistic regression analysis and multivariate analysis of variance showed this effect to be independent of initial differences between the groups. Efforts to improve quality in long-term care for the elderly are not wholly dependent on the provision of expensive new facilities.