Background: Patients with hematologic malignancies often receive aggressive care at the end of life (EOL), leading to lower quality of life. Access to early palliative care may improve EOL care outcomes, its benefits are less well established in hematologic malignancies than in solid tumors. We sought to describe the use of palliative care services among Medicare beneficiaries with hematologic malignancies, and associated EOL quality measures.
Methods: Using the linked Surveillance, Epidemiology, and End Results (SEER)-Medicare registry, we studied fee-for-service Medicare beneficiaries diagnosed with acute or chronic leukemias, lymphomas, myeloma, myelodysplastic syndrome, or myeloproliferative neoplasms, who died in 2001-2015. We described trends in the use of billed palliative care services (BPCS, identified by codes in clinician encounter claims: ICD-9 V66.7 or ICD-10 Z51.5). Among patients surviving >30 days from diagnosis, we compared baseline characteristics and EOL care quality metrics for patients with and without "early" BPCS (defined as services initiated >30 days before death), as well as Medicare spending in the last 30 days of life. Multivariable models were fitted as appropriate according to outcome variable (robust Poisson, negative binomial, or log-gamma) and adjusting for hematologic malignancy histology, patients' age, sex, race, marital status, Medicaid co-insurance, comorbidity index and performance status indicator (calculated from claims within 1 year before death), and year of death.
Results: Among the 139,191 decedents, median age at death was 82 years and 46.4% were women. The proportion with any BPCS was 5.2% overall during the study period, and it increased from 0.4% in 2001 to 13.3% in 2015 (Fig. A). Median time from the first BPCS encounter to death was 10 days (interquartile range, 3 to 39), and it increased from 6 days in 2001 to 12 days in 2015. Most (84.3%) BPCS encounters occurred during hospital admissions (Fig. B), and this proportion did not significantly change over time. Although the number of BPCS claims increased over time for any specialty, there was a relative increase in claims billed by nurse practitioners (from 7.9% in 2001/05, to 29.7% in 2011/15) or palliative care specialists (from 0% to 15.6%, respectively). Use of early BPCS remained rare, but increased from 0.2% in 2001 to 4.3% in 2015. Overall, early BPCS constituted 28.5% of all first BPCS. A relatively higher proportion of early BPCS occurred in the ambulatory setting (15.0%).
In the comparative cohort of patients who survived >30 days from diagnosis (N=120,741, Table), the use of early BPCS (1.7% overall) was more frequent in acute leukemia than in other histologies, adjusting for other factors. It was also significantly more frequent among Black patients, those with higher comorbidity indices or poor performance statuses, and those who received active chemotherapy at any point. Presence of early BPCS was associated with significantly improved EOL care quality metrics, including higher rates of hospice use, longer hospice length of stay, and lower use of aggressive measures, like repeated hospitalizations, admissions to the intensive care unit, and receipt of chemotherapy in the last 14 days of life (see Table). Early BPCS were also associated with significantly lower average Medicare spending in the last 30 days of life (marginal means $21,380 with and $23,651 without early BPCS, P<.001).
Conclusion: Use of BPCS among Medicare beneficiaries with hematologic malignancies has increased steeply in recent years, but most encounters still occur within days of death in the inpatient setting. This pattern potentially limits the benefits that could be achieved for patients and their caregivers with earlier institution of palliative care. Early BPCS are associated with better EOL care quality metrics similar to those observed in solid tumors, but causation remains uncertain in retrospective claims data, especially given known underutilization of palliative care billing codes in non-terminal patients. Our results support the need for prospective trials of early palliative care for patients with hematologic malignancies, and for research about the barriers to early access to palliative care that may be specific to this patient population.
Disclosures
LeBlanc: Pfizer Inc: Consultancy; Heron: Membership on an entity's Board of Directors or advisory committees; Daiichi-Sankyo: Membership on an entity's Board of Directors or advisory committees; CareVive: Consultancy; Agios: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Medtronic: Membership on an entity's Board of Directors or advisory committees; Otsuka: Consultancy, Membership on an entity's Board of Directors or advisory committees; Helsinn: Consultancy; Astra Zeneca: Consultancy, Research Funding; Amgen: Membership on an entity's Board of Directors or advisory committees; AbbVie: Membership on an entity's Board of Directors or advisory committees; Seattle Genetics: Consultancy, Research Funding; American Cancer Society: Research Funding; Duke University: Research Funding; Jazz Pharmaceuticals: Research Funding; NINR/NIH: Research Funding; Flatiron: Consultancy; Celgene: Honoraria. Olszewski:Genentech: Research Funding; TG Therapeutics: Research Funding; Adaptive Biotechnologies: Research Funding; Spectrum Pharmaceuticals: Research Funding.
Early Palliative Care Services and End-of-Life Care in Medicare Beneficiaries with Hematologic Malignancies: A Population-Based Retrospective Cohort Study
