Variation in Hospice Patient and Admission Characteristics by Referral Location.

Abstract

Little is known regarding differences between patients referred to hospice from different care locations. The objective this study was to describe the associations between hospice referral locations and hospice patient and admission characteristics. Cross-sectional analysis of hospice administrative data. Adult (age older than 18 y) decedents of a national, for-profit, hospice chain across 19 US states who died between January 1, 2012, and December 31, 2016. Patients' primary hospice diagnosis, hospice length stay, and hospice care site. We also determined the frequency of opioid prescriptions with and without a bowel regimen on hospice admission. Among 78,647 adult decedents, the mean age was 79.2 (SD=13.5) years, 56.4% were female, and 69.9% were a non-Hispanic White race. Most hospice referrals were from the hospital (51.9%), followed by the community (21.9%), nursing homes (17.4%), and assisted living (8.8%). Cancer (33.6%) was the most prevalent primary hospice diagnosis; however, this varied significantly between referral locations (P<0.001). Similarly, home hospice (32.8%) was the most prevalent site; however, this also varied significantly between referral locations (P<0.001). More hospital-referred patients (55.6%) had a hospice length of stay <7 days compared with patients referred from nursing homes (30.3%), the community (28.9%), or assisted living (18.7%), P<0.001. Hospital-referred patients also had the lowest frequency (58.4%) of coprescribed opioids and bowel regimen on hospice admission compared with other referral locations. We observed significant differences in hospice patient and admission characteristics by referral location.