Impact of transfusion timing on clinical outcomes across hematologic malignancies.

Abstract

e18613 Background: Transfusion support is essential in the inpatient management of hematologic malignancies; however, disease-specific differences in transfusion timing, blood bank processing delays and their association with clinical outcomes remain poorly studied. We sought to evaluate whether these factors are independently associated with hospital length of stay (LOS) and in-hospital mortality. Methods: We performed a retrospective cohort study of 2,415 red blood cell (RBC) and platelet transfusion events among hospitalized patients with acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL) and multiple myeloma (MM) at a tertiary care center from October 2022 to October 2025. Transfusion timing variables included day versus night administration (7:00 AM–6:59 PM vs 7:00 PM–6:59 AM), weekday versus weekend administration, STAT priority orders and blood bank issue lag (time from order placement to product release). Outcomes were LOS and in-hospital mortality. Diagnosis-specific differences were evaluated using chi-square testing and ANOVA. Multivariable linear regression and logistic regression models were incorporated to test independent associations between transfusion-related variables and outcomes, and adjusted for diagnosis. Results: RBCs accounted for 42.5% (674/1,585), 45.7% (252/552) and 76.6% (213/278) of transfusions in AML, ALL and MM, respectively, with platelets comprising 911, 300 and 65 transfusions (p<0.001). Night transfusions differed modestly (18.3% AML, 14.3% ALL, 21.9% MM; p=0.017), whereas weekend transfusions varied more substantially (24.9% AML, 18.3% ALL, 30.9% MM; p<0.001). STAT utilization also differed by diagnosis (31.5% AML, 21.6% ALL, 23.7% MM; p<0.001). Blood bank issue lag differed significantly (p<0.001), longest in MM (2.97 hours) compared with ALL (2.50 hours) and AML (2.26 hours). Mean LOS was 36.7 days for AML, 21.3 days for ALL and 17.0 days for MM (p<0.001). In-hospital mortality occurred in 52.6% of AML, 22.1% of ALL and 48.2% of MM patients (p<0.001). After adjustment, weekend transfusions were independently associated with longer LOS (β=0.154, p=0.033), while STAT priority orders were associated with shorter LOS (β=−0.087, p=0.028). STAT priority was independently associated with increased odds of in-hospital mortality (OR 1.46, 95% CI 1.20–1.77; p<0.001). Time of day and issue lag were not independently associated with outcomes. Conclusions: Transfusion timing, issue lag and clinical outcomes differed significantly by diagnosis among patients with AML, ALL and MM. STAT priority was the strongest independent predictor of outcomes, reflecting greater illness severity. Weekend transfusions were associated with longer LOS, potentially related to system-level factors affecting care coordination. Chronic leukemias were excluded due to limited sample size.