Clinical outcomes and length of stay of a co-located psychogeriatric and geriatric unit.

To examine the effect of co-location of psychogeriatric and geriatric services on length of stay and changes in patients' psychosocial characteristics. A retrospective analysis of the performance indices of psychogeriatric patients aged > or =50 years admitted to a co-located psychogeriatric and geriatric unit at Bankstown-Lidcombe Hospital, New South Wales (NSW), Australia from April 2004 to June 2006. Comparisons were made between the performance of the Bankstown-Lidcombe's unit and the NSW state average (consisting of traditional solitary models of care) with respect to patients' length of stay (LOS) and changes in psychosocial indices. Bankstown's patients had a higher burden of psychosocial impairments. The mean LOS for psychogeriatric episodes was significantly shorter at Bankstown-Lidcombe Hospital than the NSW state average (28.3+/-19.6 days vs. 33.4+/-22.7 days, p<0.001). The overall improvement in aspects of mental state and social behaviors for psychogeriatric admissions at Bankstown-Lidcombe Hospital was significantly better than the NSW state average. Co-location of psychogeriatric and geriatric services reduced patients' LOS and improved psychosocial performance compared to traditional models of care. However, more robust studies are required to fully examine the benefits of this type of service.