Evaluating hemoglobin thresholds for blood transfusions in oncology patients admitted in the intensive care unit.

person Abijha Biju Boban Northeast Georgia Medical Center, Gainesville, GA info_outline Abijha Biju Boban, Andrew Strike, Dhaval Patel, Ania Rynarzewska

Authors person Abijha Biju Boban Northeast Georgia Medical Center, Gainesville, GA info_outline Abijha Biju Boban, Andrew Strike, Dhaval Patel, Ania Rynarzewska Organizations Northeast Georgia Medical Center, Gainesville, GA, Georgia College & State University, Milledgeville, GA Abstract Disclosures Research Funding No funding sources reported Background: Current guidelines recommend transfusing at a hemoglobin (Hb) threshold of 7g/dl. There is a lack of evidence to support a specific threshold for transfusion in patients with active malignancy admitted to the Intensive Care Unit (ICU). This study attempts to determine an evidence-based threshold for blood transfusions in cancer. Methods: A retrospective study was done to study 561 patients with an active cancer diagnosis admitted to Northeast Georgia Medical Center ICU between 2017 and 2023 that received at least one unit of blood transfusion. Chi-Square test of association with Bonferoni correction was applied to the top five common cancers seen in ICU: Leukemia, Lymphoma, Brain tumors, Lung cancers, and GI cancers. It compared three thresholds of transfusion and looked for any change in mortality risk, readmissions, need for multiple transfusions, prevalence of new onset or recurrent heart failure, development of altered mental status, or cardiac arrest. The three thresholds for transfusions that were studied were < 7g/dL, < 8 g/dL, and > 8 g/dL. The studied cancer types were further divided into either hematologic or solid tumors. Results: There was a decrease in overall cancer readmissions compared to expected over 30 days by 64%, 60 days by 44%, and 90 days by 47.5% when transfused for a Hb threshold of <8 g/dL compared to an increase of 22.8%, 18.2%, and 18.8% in 30-, 60-, and 90-day readmissions when transfused for a Hb threshold of <7 g/dL. When divided by cancer types, there was a significant decrease from expected by 30%, 47.5%, and 56% in 30-, 60-, and 90-day readmissions for patients with solid tumors for a Hb threshold of < 8 g/dL. Mortality risk decreased by 50% and the need for multiple transfusions decreased by 21% when transfused for a Hb threshold of < 8 g/dl. Of note, there was no significant association between transfusion levels and associated reactions. There was a decrease of 82% in altered mental status when transfused for threshold of Hb <8 g/dL. Finally, significant association was not found with AKI, cardiac arrest, or heart failure and any Hb transfusion thresholds. Conclusions: Based on these results, there is a significant mortality benefit and decreased length of stay for cancer patients admitted to ICU when blood transfusion is undertaken at a hemoglobin threshold of <8g/dL. These results should be further investigated in a randomized controlled trial.