Rural-Urban Variations in Medicare Live Discharge Patterns From Hospice, 2012-2013

For the last decade-and-a-half, the proportion of patients discharged from the Medicare hospice program prior to death – known as a "live discharge" – has increased across the country with significant hospice-level geographic variations. Despite clear geographic variations in live discharge rates and known rural-urban disparities (e.g., patients of rural hospices have higher satisfaction), previous studies have not explored differences in live discharging patterns between rural and urban areas in depth. Accordingly, this brief first provides an overview of the geographic distribution of "freestanding" (i.e., rather than those co-located in a hospital, home health agency, or skilled nursing facility) rural and urban hospices and, second, explores live discharge rates for hospices operating in rural versus urban areas. We merged data from the 2012-2013 Medicare claims (Provider of Services (POS) file, the Hospice Research Identifiable File, and the Master Beneficiary Summary File). Due to data use agreement restrictions, hospices with 10 or fewer observed discharges (either alive or deceased) were excluded. Rural-urban status (i.e., metro versus non-metro) was determined at the county level based on whether the hospice was located in a Core-Based Statistical Area. In 2012-2013, 21.5 percent of freestanding hospices were located in rural areas. Rural hospices were more likely to be: located in the South Census region (63.4 percent); non-chain-owned (56.8 percent); or for-profit (58.3 percent) in 2012-2013. The average rate of live discharge for rural hospices was 16.0 percent, compared to 12.9 percent for hospices located in urban areas. Across all types of hospices, rural for-profit hospices in the South Census region had the highest average rate of live discharge (22.8 percent).
Rapid Response to Requests for Rural Data Analysis
Rachel Dolin, Denise Kirk, Mark Holmes