Mentor Name: Alix Seif

Pediatric patients with newly diagnosed acute myeloid leukemia (AML) receive 3-5 consecutive courses of intensive chemotherapy, each followed by prolonged severe neutropenia (35-45 days/course), when they are at high risk for life-threatening infections. Despite ample comparative clinical effectiveness (CCE) data showing outpatient recovery can be as safe as inpatient, current Children’s Oncology Group (COG) guidelines strongly recommend inpatient-only care for neutropenia. Patients spend up to 6 months in the hospital over their treatment, creating significant burdens for patients and caregivers. For patients meeting strict clinical criteria, we showed outpatient care had similar bacteremia, severe infection, and treatment-related mortality (TRM) rates to inpatient. Outpatient recovery was associated with improved sleep quality and parental distress and had high patient and caregiver satisfaction. Higher ICU-level care needs with readmissions suggest that, while outpatient recovery can be safe, current clinical criteria are insufficient to fully personalize care setting decisions. COG recommended mandatory inpatient recovery for every chemotherapy course after a trial in the 1990s observed excess TRM. Despite improved supportive care, clinicians are resistant to outpatient management: only 20% of centers discharge patients for recovery, without personalization by clinical, social, or logistical factors. We collaborated with patients, caregivers, and multi-disciplinary healthcare professionals to develop the “AML Care at Home” toolkit to personalize the care setting and facilitate implementation. Based on our CCE and adult outpatient AML data, the core components are: 1) a tool to identify candidates for outpatient care and 2) a manual of operations to guide care. We piloted toolkit-guided care setting personalization (TCSP) at the Children’s Hospital of Philadelphia from November 2022 — May 2024, with clinical outcomes aligned with prior CCE data and excellent reach, fidelity, and acceptability. These results support broad testing of TCSP for pediatric AML. We hypothesize implementation of AML Care at Home TCSP will 1) result in =60% reach, and 2) reduce total inpatient days per treatment course. We developed a stepped wedge cluster-randomized hybrid implementation-effectiveness trial of TCSP that will open at 16 US sites. To measure the CCE of TCSP, we will leverage chart abstraction data from the Real-World Epidemiology of Acute Leukemia-Acute Myeloid Leukemia cohort (REAL-AML) led by Kelly Getz, PhD, and Richard Aplenc, MD, PhD. Ms. Salwan will collect data for REAL-AML from June 1st — July 24th under the mentorship of Alix Seif, MD, MPH.