Tube feeding through a gastrostomy directly into the stomach has become a common procedure. Rates of gastrostomy placement among elderly Americans approximately doubled between 1987 and 1993. This increase corresponded in time to the introduction of new technology allowing the placement of permanent feeding tubes without general anesthesia. These new placement techniques have permitted use of these tubes in medically sicker patients, especially the frail elderly. Approximately half of patients receiving gastrostomy die within 6 months. We propose to examine the expanding utilization of gastrostomy among patients near the end of life at two points in National Center for Health Statistics, is a random sample of 18,000 decedents from each of two years: 1986 and time using a unique dataset; the National Mortality Followback Survey (NMFS). This file, available from the 1995. The survey includes information on cause of death, demographics, care received and functional status in the last year of life. The two surveys bracket the implementation of the new placement techniques. We have examined the 1986 survey and identified an adequate number of gastrostomy procedures: 175. The mean age of these individuals is 73. We will analyze these surveys to: identify how the characteristics of patients receiving gastrostomy changed between 1986 and 1995 describe the health care utilization of older adults who have received gastrostomy examine the clinical and demographic risk factors which are associated with receipt of a gastrostomy through use of case-control studies. We expect that patients receiving gastrostomy in the 1990's will be older, less functional and more likely to have a cognitive impairment. This analysis will shed light on expansion of this technology in the elderly population. Improved knowledge of patterns of use of this technology in older Americans may suggest new strategies for clinical decision-making about feeding tubes.