Hospice Autopsy Trends Among Patients With Cancer

Abstract

<h3>Purpose/Objective(s)</h3> Given increasing emphasis on end of life quality metrics including respecting patient values and dying with dignity in a patient's choice location, the number of patients with cancer who die in hospice instead of an inpatient facility is steadily rising. When a terminal patient is transferred to a hospice facility, the assumption is made that both the patient and the family understand the disease course and prognosis. However, each year there are patients with cancer who die in a hospice facility who receive an autopsy. We sought to evaluate the characteristics of those that undergo an autopsy after hospice death. <h3>Materials/Methods</h3> CDC WONDER contains death certificate data for all US counties and is maintained by the National Center for Health Statistics. All cancer deaths from 2003- 2019 were evaluated and demographic data was captured. Hospice death as well as autopsy status were noted. Chi-square tests compared variables and linear regression assessed for changes in trends. <h3>Results</h3> From 2003 through 2019, a total of 12,042,519 cancer deaths occurred. Of these deaths, 888,194 were patients who died in a hospice facility with yearly deaths increasing from 3,209 in 2003 (0.57% of cancer deaths) to 84,868 in 2019 (14.2% of cancer deaths) (significant trend <i>P</i> < 0.0001). There were a total 94,061 autopsies among all cancer patients (0.78%) of which 2,002 autopsies occurred in those who died in a hospice facility (2.1% of cancer autopsies and 0.26% of hospice deaths). Since 2003, the absolute number of autopsies for patients who died from cancer decreased from 6,589 in 2003 to 4,933 in 2019, with overall autopsy rate decrease (1.50 % to 0.85%, significant trend <i>P</i> < 0.0001). Since 2003, the absolute number of autopsies for patients who died from cancer in hospice increased from 20 in 2003 to 226 in 2019, however the autopsy rate decreased (0.68% to 0.28%, significant trend <i>P</i> < 0.0001). Male patients (OR 1.32) and patients living in a Large Metropolitan region (OR 1.38) were significantly more likely to undergo autopsy after cancer death in hospice (both <i>P</i> < 0.0001). There was no statistically significant difference in number of autopsies found between race, ethnicity, or age (all <i>P</i> > 0.05). Lung cancer was the most common cancer type of hospice autopsy (22.4%), cancer of unknown primary was the second most common cancer type (8.8%). <h3>Conclusion</h3> : Autopsies among patients with cancer who died while in hospice have decreased by more than 50% since 2003, although over 200 still receive autopsies every year. Decreasing autopsy trends likely relate to improvements in the diagnosis and treatment of cancer and are consistent with trends in other diseases. As autopsies after hospice death suggest unresolved questions or burden on patient families, it is important to improve communication, shared decision making, and ease end of life transitions for the entire family unit so that autopsies after hospice death continue to decrease.