New evidence on Medicare's prospective payment system : a survival analysis based on the NHANES I epidemiologic followup study

Medicare’s prospective payment system (PPS), introduced in 1983, pays hospitals a fixed price for each stay rather than reimbursing costs. Previous studies evaluated its first few years using endogenous measures to control for heterogeneity in patients’ health. We examine PPS over a full decade using competing risks Cox survival models and a national longitudinal survey with independent information on patients’ health. New findings include: risk of death in hospital increased; risk of discharge to a nursing home continued to increase as PPS matured; and risk of nursing home admission from the community following hospital discharge rose. HMOs may have contributed to these outcomes.