Use of palliative radiation and association with end-of-life quality metrics and healthcare utilization for patients with gynecologic malignancies in Ontario, Canada from 2006-2018.

Abstract

e24055 Background: The optimal use of radiotherapy at the end of life (EOL) is undefined in gynecologic oncology. We investigated the use of radiotherapy at the EOL in patients with gynecologic malignancies, and relationship to established EOL healthcare quality indicators. Methods: A population-based, retrospective cohort study of gynecologic cancer decedents in Ontario, Canada from 2006-2018 using ICES-linked administrative health care data. Results: Of 16,237 decedents, 26.7% received palliative intent radiotherapy; 6.7% received late radiation in the last 31-90 days (d) of life, and 5.3% received EOL radiation in the final 30d. Radiation was most frequently received by patients with non-ovarian malignancies, and the most common site was pelvis (44.2%). Late radiation was initiated a median 10 (IQR 8-12) weeks before death, and these patients received median 20Gy (IQR 8-24) over median 6 (IQR 3-10) fractions. EOL radiation was initiated a median 3 (IQR 2-5) weeks before death, and these patients received median 16Gy (IQR 8-20) over median 6 (IQR 2-8) fractions, accounting for 20% of remaining days of life. On multivariable analysis, age <40 years, non-ovarian cancers, stage IV disease at diagnosis were associated with receipt of late and EOL radiation. Patients receiving late radiation had the lowest rates within the last 30 days of life of hospitalization (45.7% vs. 70.6% with EOL radiation and 56.6% with neither), ICU admission (2.8% vs. 5.8% with EOL radiation vs. 4.1% with neither), death in hospital (36.3% vs. 55.3% with EOL radiation vs. 44.7% with neither), and the composite measure of aggressive EOL care (15.6% vs. 27.3% with EOL radiation vs. 20.7% with neither). They also had a higher likelihood of receiving early palliative care ≥3mo before death (68.1% vs. 50.8% with EOL radiation vs. 62.9% with neither). On multivariable logistic regression, late radiation remained significantly associated with less aggressive care, more supportive care, and lower risk of death in hospital, while EOL radiation was significantly associated with more aggressive care and higher risk of death in hospital. Conclusions: One quarter of gynecologic cancer decedents receive palliative radiotherapy, with 12.1% in the final 90 days of life. Dose schedules did not differ between those receiving late vs. EOL radiation; however, late palliative radiation was associated with less aggressive EOL healthcare utilization and more supportive care, while EOL radiation was associated with the highest rates of individual and aggressive EOL care metrics.