Multi-Site Randomized Trial of a Collaborative Palliative and Oncology Care Model for Patients with Acute Myeloid Leukemia (AML) and Myelodysplastic Syndrome (MDS) Receiving Non-Intensive Therapy

Background:

Patients with AML and high-risk MDS receiving non-intensive chemotherapy have substantial quality of life (QOL) impairments and often do not engage in timely discussions with their clinicians about their end-of-life (EOL) care preferences. Yet interventions to optimize EOL care delivery and QOL for this population are lacking.

Methods: We conducted a multi-site randomized clinical trial of a collaborative palliative and oncology care model compared to usual care in 115 adult patients with AML and high-risk MDS receiving non-intensive therapy at two tertiary care academic hospitals. Patients with a new diagnosis or relapse/ refractory disease were eligible to participate within 30 days of initiating therapy. Patients assigned to the intervention met with a palliative care clinician monthly in the outpatient setting and a minimum of twice weekly during every hospital admission. Patients assigned to usual care were seen by palliative care only upon request. We used Natural Language Processing methods to interrogate the Electronic Health Record (EHR) with a validated algorithm to collect documented EOL care preferences. The primary outcome was to compare time from documentation of EOL care preferences to death between the study arms. Secondary outcomes obtained from the EHR include rates of documentation of EOL care preferences, hospitalization, and hospice utilization at the EOL. Patient-reported secondary outcomes include discussions with clinicians about EOL care preferences, QOL (FACT – Leukemia), and psychological distress (Hospital Anxiety and Depression Scale) at 3 months after enrollment.

Result: We enrolled 51.8% (115/222) of eligible patients. The rate of documented EOL care discussions in the EHR was higher among intervention patients vs. usual care (96.5% vs. 68.4%, P<0.001). Overall, 61.7% (71/115) of patients died, and those receiving the intervention were more likely to have their EOL care preferences documented a week or more prior to death (74.3% vs. 40.4%, P<0.001) and had a longer time from documentation of EOL care preferences to death (41 days vs. 1.5 days, P<0.001). Intervention patients were more likely to report discussing their EOL care preferences with their clinicians (56.9% vs. 14.0%, P<0.001), and less likely to be hospitalized in the last 30 days of life (70.6% vs. 91.9%, P=0.031). There was no statistically significant difference in hospice utilization (55.9% vs. 48.6%, P=0.637) or hospice length-of-stay between the two groups at the EOL. Patients receiving the palliative care intervention were also less likely to receive chemotherapy in the last seven days of life compared to those assigned to usual care, but this did not reach statistical significance (8.9% vs. 24.3%, P=0.115). At 3 months, patients assigned to the intervention reported better QOL (138.6 vs. 125.5, P=0.010), but no difference in depression or anxiety symptoms compared to those assigned to usual care.

Conclusions: Palliative care significantly improved rates of discussion and documentation of EOL care preferences, reduced hospitalization at the EOL, and improved QOL in patients with AML and high-risk MDS. Palliative care should be incorporated into the care of patients with AML and MDS receiving non-intensive therapy to optimize their overall quality of life and EOL care.