PF253 SAFETY AND RESOURCE UTILIZATION FOLLOWING EARLY HOSPITAL DISCHARGE AFTER INDUCTION CHEMOTHERAPY FOR ACUTE MYELOID LEUKEMIA (AML): A 4‐YEAR RETROSPECTIVE ANALYSIS

Background: While adults with AML typically remain hospitalized after completion of intensive induction therapy until blood count recovery, early hospital discharge (EHD) has become routine at our center after institutional studies suggested safety and reduction of resource utilization. These studies used strict medical and logistic criteria to identify appropriate EHD patients but this practice has since been applied to a broader patient population. Aims: To investigate the safety and resource utilization of routine EHD after intensive AML therapy over a 4‐year period since completion of our prospective EHD studies. Methods: We identified all adults with untreated AML/high‐grade myeloid neoplasms (≥10% blasts in blood/marrow) who began induction chemotherapy from 8/1/14–7/31/18. Patients were assigned to the EHD group if discharged <72 h from completing chemotherapy; otherwise, they were considered inpatient controls. The study period began on the day of initial discharge (EHD group) or completion of chemotherapy (control group) and ended with count recovery, hospital discharge (control group), death, receipt of additional chemotherapy, transfer of care, or after 42 days. Variables included baseline characteristics, logistical information and treatment given. Outcomes included readmission dates/reasons, intensive care unit (ICU) admission and resource utilization: number of emergency department visits, lab and provider visits, units of transfused red blood cells (RBC) and platelets. Results: We identified 236 EHD patients and 140 patients who remained inpatient (for logistical or medical reasons). EHD patients had better performance status, lower white blood cell count and lower treatment‐related mortality scores than controls at baseline (Table 1). EHD patients spent a mean of 17 (range: 1–42) days or 75% of their study period as outpatients. 81% of EHD patients required readmission, 87% due to febrile neutropenia or infection. 7% were admitted to the intensive care unit (ICU) compared to 11% of the controls (p = 0.26). EHD patients spent fewer study days on IV antimicrobials and used less mean RBC and platelet units per study day than controls (Table 1). Rates of early death were similar (5% in EHD vs 7% in controls, p = 0.36). We then restricted analyses to the 173 EHD patients meeting logistical and medical criteria put forth in our phase 2 study (ECOG ≤1, bilirubin ≤3 x upper limit normal, glomerular filtration rate ≥25% x lower limit normal, no uncontrolled bleeding or heart failure, availability of a caregiver and housing <1 h from clinic), and 38 medically‐matched control patients. Their characteristics were well balanced. These “eligible” EHD patients spent a mean of 76% of their study time as outpatients. While the use of IV antimicrobials was similar (35% vs 37% of study days, p = 1.0), EHD required fewer average RBCs per study day (0.3 vs 0.5, p  < 0.001) and fewer platelets per study day (0.2 vs 0.6, p  < 0.001) than controls. Early death rates did not differ between these 2 medically matched groups (2% vs 0%, p = 1.0). Notably, early death rates were higher among EHD patients who were not eligible per our phase 2 criteria compared to those who were (10% vs. 3%, p = 0.03), as were death rates among not‐eligible/eligible control patients (11% vs 0%, p = 0.02). Summary/Conclusion: An EHD care strategy following AML‐like induction chemotherapy is safe and decreases healthcare resource utilization. The medical/logistical criteria we used in our Phase 2 trial to denote EHD “eligibility” accurately identifies EHD patients with low risk for early death.