Multinational transfusion practices and outcomes in haematology patients admitted to the intensive care unit.
Schaap CM., Oomen LA., Raasveld SJ., Schenk J., de Bruin S., Reuland MC., van den Oord C., Bakker J., Cecconi M., Feldheiser A., Meier J., McQuilten Z., Scheeren TWL., Aubron C., Flint AWJ., Hamid T., Piagnerelli M., Mahečić TT., Benes J., Russell L., Aguirre-Bermeo H., Triantafyllopoulou K., Chantziara V., Gurjar M., Myatra SN., Pota V., Elhadi M., Gawda R., Mourisco M., Lance M., Neskovic V., Podbregar M., Llau JV., Quintana-Diaz M., Cronhjort M., Pfortmueller CA., Yapici N., Nielsen N., de Grooth H-J., Vlaar APJ., Biemond BJ., Shah A., Müller MCA., InPUT Study Group .
BACKGROUND AND OBJECTIVES: The number of critically ill patients with haematological conditions is increasing, yet transfusion practices in this population remain poorly defined. This study aimed to compare transfusion strategies in critically ill patients with versus without haematological conditions. STUDY DESIGN AND METHODS: This international, prospective observational substudy of the International Point Prevalence Study of Intensive Care Unit [ICU] Transfusion Practices (InPUT) evaluated transfusion use in ICU patients with and without haematological conditions, including benign or malignant diseases or a history of stem cell transplantation. Outcomes included use of red blood cells (RBCs), platelets, plasma, haemostatic interventions, transfusion indications and thresholds. RESULTS: Of 3643 ICU patients, 131 (3.6%) had a haematological condition. These patients were more likely to receive RBC (odds ratio [OR] 1.58, 95% confidence interval [CI] 1.09-2.29) and platelet transfusions (OR 8.32, 95% CI 5.09-13.6), primarily due to low haemoglobin rather than physiological triggers. Platelet thresholds were lower (median 23 × 109/L vs. 64 × 109/L) compared to non-haematology patients. Both platelet and plasma transfusions were more frequently administered prophylactically rather than for active bleeding. Haemostatic interventions were more often used in haematology patients, at higher doses and typically without viscoelastic testing. Transfused haematology patients had higher 28-day mortality and longer ICU stays. CONCLUSION: ICU patients with haematological conditions receive transfusions differently, particularly regarding platelet and plasma use. These findings underscore the need for prospective studies to define optimal transfusion thresholds in this growing and vulnerable patient population, although the study's limited sample size and lack of diagnostic granularity may affect interpretation.