The findings in this explorative-observational study showed a high prevalence of clinically relevant, patient-bound risk factors in elderly people in a general hospital. Some risk-factors were of prognostic interest for long hospital stay, although the explained variance was relatively small. This indicates that a more comprehensive study should be designed and conducted to include other patient-bound risk factors like co-morbidity, caregiver issues and social environment. Moreover, non-patient-bound factors should be addressed like intrinsic and logistic factors within the hospital, and the quality of recuperation programmes. Understanding of these factors contributes to timely identification of elderly patients, who are at high risk of a long hospital stay. Future policy is to perform specific treatment programmes for elderly patients identified as being patients at risk. Multidisciplinary person-oriented interventions and case management focussed on risk factors and functional recovery will be provided parallel and after hospital treatment period. Comprehensive scientific research on the cost-effectiveness of such a programme has started at the end of 200oo9 in Vlietland Hospital, Schiedam.

The dimensional structure of the Index condition highly corresponded to the Reference condition; that is to say that the PFD appeared to be relatively independent of cognition and ADL. High prevalences of PFD (NPI), the broad variance of NPI-symptoms and the limited prognostic importance of MMSE, BI and general details for total NPI-score as well as individual NPI-symptoms were confirmed in both conditions. The dimension 'Psychiatric function disorder' was relative independent of the dimension 'Cognition'. As a result it is of clinical interest - in case of referral to clinical and transmural programmes - to distinguish the psychiatric dimension from the cognitive dimension.