Identifying Factors That Predict for Unplanned Readmissions for Acute Myeloid Leukemia Patients Receiving Consolidation Cytarabine Based Therapies

Background: Younger patients (pts) with acute myeloid leukemia (AML) who enter a remission after intensive induction chemotherapy routinely receive at least one cycle of consolidation therapy with high dose cytarabine (HiDAC). This is commonly administered over a five-day inpatient stay, after which pts are discharged home as their blood counts nadir. It is thus a natural consequence of therapy that readmission for febrile neutropenia (FN) occurs, which can impact measures of quality and value in this population. Precise descriptions of incidence, type, and severity of infection, if identified, are lacking, and thus it is unknown to what standard cancer centers should be held for anticipated readmission. We measured these rates, and attempted to identify predictive factors for readmission.

Methods: Adult AML pts ≥ 18 years of age who received at least one cycle of HiDAC consolidation (1000-3000 mg/m2 for six doses) in 2009-2019 were included. Our primary aim was to identify predictive factors for readmission after the first cycle of consolidation chemotherapy. The following pt characteristics and co-morbid conditions were analyzed: age, gender, body mass index (BMI), smoking status, AML cytogenetic risk status, history of diabetes, peripheral vascular disease, cardiovascular disease, chronic pulmonary disease, hepatic impairment, and other cancers. Secondary aims included: estimating rates of all-cause readmissions among all HiDAC cycles, defining the rate of FN readmissions, estimating rates of intensive care unit (ICU) admissions, clinical (e.g., probable pneumonia per imaging) and microbiologically-documented infections, prophylactic (ppx) medications used, and mortality. Statistical analyses interrogated potential risk factors for evidence of association with hospital readmission after the first cycle of consolidation chemotherapy.

Results: We identified 182 AML pts who fit inclusion criteria. The median age was 50 years (range 19-73); 55% were female and 45% were male. Statistical analyses revealed no association with readmission after cycle 1 for cytogenetic risk (p=0.85), history of heart failure (p= 0.67), chronic pulmonary disease (p=1), connective tissue disease (p=0.53), cerebrovascular accident (p=0.63), diabetes (p=0.63), gender (p=0.07), history of lymphoma (p=0.53), other solid tumors (p=0.53), liver disease (p=1), myocardial infarction (p=0.71), peripheral vascular disease (p=1), or smoking status (p= 0.52). For 480 HiDAC cycles analyzed (88% at 3000 mg/m2), the overall readmission rate was 50% (242/480), of which 85% (205/242) were for FN. Those readmissions which were not FN were for cardiac complications (chest pain, EKG changes), non-neutropenic fevers or infections, neurotoxicity, bleeding or clotting events, or other symptoms associated with chemotherapy (nausea/vomiting, pain, etc.). Median time to FN hospital admission was 18 days (range 6-27) from the start of HiDAC. Of the 205 FN readmissions, 57% had documented infections. Of these infections, 41% were bacteremia, 23% fungal, 16% sepsis, 12% other bacterial, and 8% viral. Of 480 HiDAC cycles, ppx medications prescribed included: 92% fluoroquinolone (442/480), 81% anti-viral (389/480), 30 % anti-fungal (142/480), and 3% colony stimulating factor (14/480). Only 7% (14/205) of FN readmissions resulted in an ICU admission, and 1% (3/205) resulted in death.

Conclusions: Approximately half of patients treated with consolidation therapy following intensive induction therapy can be expected to be readmitted to the hospital. The majority of FN readmissions were associated with clinical or microbiologically documented infections and are not avoidable, however ICU admission and death associated with these complications are rare. Readmission of AML pts following HiDAC is expected, and therefore, should be excluded from measures of value and quality.