The Impact of Medicare's Prospective Payment System on Wisconsin Nursing Homes

Abstract

<h3>To the Editor.—</h3> In the April 3, 1987, issue ofThe Journal, Dr Davis¹raises questions about our study of Wisconsin's nursing homes under Medicare's prospective payment system, which showed increased hospitalization rates for elderly Medicaid residents and a shift in location of death from hospitals to nursing homes after implementation of the PPS. Dr Davis concludes that the findings represent "flawed methodology" and "utilization of inappropriate data sources" and that "it may be fallacious to attribute the shift in location of death to the initiation of PPS." Dr Davis' major criticism concerns use of the Medicaid-Medicare hospitalization database. This Medicaid cross-over file represents the most reliable source of hospitalization data available. No data source is error free; the question is whether errors occur consistently across time, represent bias, or appear simultaneously with implementation of the PPS. Dr Davis notes that the reported Medicaid hospitalization rates are lower than