Possible transfusion-related acute lung injury (TRALI) in cardiac surgery patients

Aim To determine the incidence of possible transfusion-related acute lung injury (TRALI) and related risk factors in cardiac surgery patients. Methods A single-center prospective cohort study was conducted from January 2009 to March 2010 at the Zagreb University Hospital Center, Croatia. Patient-, transfusion-, and surgery-related data were collected. The study included 262 patients who were observed for respiratory worsening including measurements of arterial oxygen saturation (SaO(2)), fraction of inspired oxygen (FiO 2), and partial pressure of arterial oxygen (PaO2). Possible TRALI was defined according to the Toronto Consensus Conference definition broadened for 24-hour post-transfusion. This cohort was divided in two groups. TRALI group included 32 participants with diagnosis of TRALI and the control group included 220 patients with or without respiratory worsening, but with no signs of ALI. Results Possible TRALI was observed in 32 (12.2%) patients. Compared with the control group, possible TRALI patients had higher American Association of Anesthesiology scores, higher rate of respiratory comorbidity (43.8% vs 15.5%), and required more red blood cells (median 4, range [2.5-6] vs 2 [1-3]), plasma (5 [0-6] vs 0 [0-2]), and platelet units (0 [0-8] vs 0 [0-0]) (P < 0.001 all). Risk factors for possible TRALI were total number of transfused blood units (odds ratio [OR] 1.23; 95% confidence interval [CI] 1.10-1.37) and duration of cardiopulmonary bypass (OR 1.08; 95% CI 1.05-1.11). Posttransfusion PaO2/FiO(2) ratio was significantly decreased in possible TRALI patients and significantly increased in transfused controls without acute lung injury. Conclusion We observed a higher rate of possible TRALI cases than in other studies on cardiac surgery patients. Serial monitoring of PaO2/FiO(2) ratio and detection of its posttransfusion worsening aids in identification of possible TRALI cases.