Prediction of outcome of patients admitted to a geriatric department

The reliability of a clinical prognosis of outcome at six weeks has been investigated in a series of 1092 consecutive admissions to a geriatric department. The prognosis was made within a day of admission and made use only of the history and clinical findings. The results for discharge were good but categorization of patients who died or who stayed longer than six weeks was less reliable. Unpredicted deaths were often sudden or related to disease with developed after assessment. Other errors were due to over-optimism in forecasting response to medical treatment or rehabilitation. In other cases progressive deterioration had not been clearly recognized at our assessment based on a single point in time.