Changes in health service use and mortality among U.S. elderly in 1980-1986.

Concern has emerged about the impact on quality of care of recent changes in Medicare reimbursement for acute hospital episodes (i.e., the introduction of the Prospective Payment System based on the Diagnosis Related Groups Reimbursement methodology). One aspect of those concerns is that very sick patients would be prematurely discharged to nursing homes in which the high level of medical care required would not be available. This was recently studied in terms of changes between 1981 and 1985 in the location of deaths (e.g., hospital, institutional) reported on U.S. death certificates. We analyzed the death certificate data for a longer period of time (1980 to 1986) and stratified the analysis by both age and cause of death--which we felt were important determinants of location of death. We also examined data from the 1982 and 1984 National Long Term Care Surveys linked to data on Medicare service use. In those analyses we could explicitly identify chronically disabled and institutionalized populations and study death rates in different locations within those populations. Our analyses showed little evidence of increased mortality rates due to premature hospital discharge. There were, however, significant changes in the patterns of service use. More home health agency (HHA) and skilled nursing facility (SNF) services were consumed though the rates of death per episode in those venues declined.