Hospice care in Medicare patients with primary liver cancer: the impact on resource utilisation and mortality.

Background: Few studies have assessed the impact of hospice care in patients with primary liver cancer. Aim: To examine the determinants of hospice care and its effects on resource utilisation and survival among Medicare beneficiaries with primary liver cancer. Methods: We utilised the Surveillance, Epidemiology and End result Registry (SEER) database from 2002 to 2009 for this cross-sectional study. A total of 3385 patients with primary liver cancer were included. We used logistic regression to discern variables associated with hospice and Cox proportional hazards models to evaluate one-year mortality risk. Results: Compared to patients who enrolled in a hospice, those patients who did not, were younger, non-White and sicker (P <.05 for all). Half of all patients with primary liver cancer died within six months of diagnosis, and one-year mortality was similar in both groups (P =.413). After adjusting for baseline characteristics [age at diagnosis, race, disease severity, tumour stage and treatment], shorter time to hospice care was associated with reduced mortality (HR per day: 0.99 [95% CI, 0.98-0.99]). Older age, decompensated cirrhosis and advanced tumours stage were associated with decreased time to hospice, while Asian/Pacific Islander race and history of radiosurgery were associated with increased time to hospice (all P <.05). Hospitalisations were more costly for those who never enrolled in a hospice compared to hospice enrollees (median $31 607 [$ 18 394-$54 254] vs $ 22 316 [$ 13 741-$ 36 170], P <.0001). Conclusions: Hospice enrolment of patients with primary liver cancer provides survival and resource utilisation benefits. Some clinical and demographic factors may represent barriers to hospice enrolment. Further studies are needed to fully understand these barriers in patients with primary liver cancer.