Impact of palliative care on end-of-life outcomes in hematologic malignancies.

Abstract

12026 Background: Patients (pts) with hematologic malignancies (HMs) receive more aggressive end-of-life (EOL) care and often die in the hospital. The impact of palliative care (PC) on EOL quality outcomes in HMs has not been well described. In 2017 we embedded a PC specialist within our inpatient malignant hematology team to facilitate the use of early PC. We sought to determine if this practice was accompanied by a shift in EOL outcomes. Methods: We conducted a retrospective review of pts diagnosed with acute myeloid leukemia (AML) at our institution in the 2 years before (Cohort A) and after (Cohort B) implementation of embedded PC. We identified pts who received PC and if it was early (during initial inpatient stay) or late (sometime after). We then examined EOL quality outcomes: hospitalizations and intensive care (ICU) admissions in the last 30 days of life, chemotherapy use in the last 14 days of life, and use of hospice and death out of hospital (DOH), using Fisher’s exact test to compare proportions. Results: Among 139 AML pts, 46 in Cohort A, 93 in Cohort B, we identified 34 and 47 decedents in each cohort respectively. The use of PC was significantly higher in Cohort B (75% vs 43%, P= 0.0006), with a significant increase in early PC (52% vs 11%, P < 0.0001). There was no significant improvement in EOL quality outcomes between Cohort A and B, or uniquely among pts receiving early PC ( P > 0.05); however, PC use in general across all cohorts was associated with significant increase in hospice use and fewer ICU admissions ( P =0.016 and 0.0043, respectively). Among pts not receiving PC, a numerical improvement was noted in EOL metrics between Cohorts A and B ( P > 0.05; see table). Conclusions: PC for pts with AML was associated with significantly better EOL quality outcomes. We also observed improvement in EOL metrics over time among pts not receiving PC, which may indicate a culture shift with the embedded PC service, whose benefit extended to pts not directly receiving PC. Embedding a PC specialist and early PC in AML, however, was not significantly associated directly with EOL care improvements. The value of these interventions in HMs may be better measured using patient-reported outcomes and quality of life measures rather than strict EOL outcomes. Further research should consider potential differential role of PC among pts with HM undergoing aggressive/curative, or non-intense/palliative therapy. [Table: see text]