Factors associated with the high prevalence of short hospice stays.

This study's goal was to gain an understanding of the factors associated with hospice stays of 7 days or less (i.e., short hospice stays), and to test the hypothesis that independent of changes in sociodemographics, diagnoses, and site-of-care, the likelihood of a short hospice stay increased over time. We examined hospice stays for 46655 nursing home and 80507 non-nursing home patients admitted between October 1994 and September 1999 to 21 hospices across 7 states, and owned by 1 provider. Logistic regression was used to determine the factors significantly associated with a higher probability of a short stay. Compared to patients admitted in (fiscal year) 1995, and controlling for potential confounders, the probability of a short stay significantly increased in each year after 1995 in nursing homes, and in 1999 in non-nursing home settings. In (fiscal year) 1995, a nursing home resident admitted to hospice had a 26% probability (95% confidence interval [CI] 0.24, 0.28) of a less than 8-day stay and, in (fiscal year) 1999, the probability was 33% (95% CI 0.31, 0.34); a non-nursing home patient had a 32% probability in 1995 (95% CI 0.30, 0.34) and a 36% probability in 1999 (95% CI 0.34, 0.37). The probability of a short hospice stay was greater for patients with noncancer diagnoses, independent of year of hospice admission. In this paper we discuss the possible underlying reasons for the increased probability of short hospice stays and we speculate on what this increase may mean in terms of hospice's ability to provide high-quality end-of-life care.