M Ranucci1, A Jeppsson, E Baryshnikova. 1. Department of Cardiothoracic Anesthesia and Intensive Care, IRCCS Policlinico San Donato, San Donato Milanese (Milan), Italy.
Abstract
BACKGROUND: Pre-operative fibrinogen levels are negatively associated with postoperative bleeding in cardiac surgery patients. The guidelines of the European Society of Anaesthesiology consider the possibility of a prophylactic pre-operative supplementation in patients with fibrinogen levels<`3.8 g/l. The present study is a reanalysis of published data aimed to define the diagnostic accuracy of different values of pre-operative fibrinogen levels in predicting severe post-operative bleeding. METHODS: Data were retrieved for 2154 patients in four different studies. Severe bleeding (SB) was defined as a post-operative chest drain output>1 l/12 h. Diagnostic accuracy for prediction of SB was tested at three cutoff values of pre-operative fibrinogen (2.5 g/l, 3.0 g/l, and 3.8 g/l). RESULTS: At all the three cutoff values, pre-operative fibrinogen levels had an excellent negative predictive value, ranging from 86% to 100%. Conversely, the positive predictive value was poor at all the cutoff levels: 12% (3.8 g/l), 14% (3.0 g/l), and 19% (2.5 g/l). Overall, the accuracy of pre-operative fibrinogen levels for the prediction of SB was poor. A strategy based on pre-operative fibrinogen supplementation would lead to inappropriate treatment in > 80% of the treated patients. Overall, a trigger value of 3.8 g/l would result in an inappropriate treatment in 52% of the patients, of 3.0 g/l in 20% of the patients, and of 2.5 g/l in 4% of the patients. CONCLUSION: Correction of pre-operative fibrinogen levels below 3.8 g/l would lead to an excessive rate of inappropriate interventions. Values below 2.5 g/l could be considered.
BACKGROUND: Pre-operative fibrinogen levels are negatively associated with postoperative bleeding in cardiac surgery patients. The guidelines of the European Society of Anaesthesiology consider the possibility of a prophylactic pre-operative supplementation in patients with fibrinogen levels<`3.8 g/l. The present study is a reanalysis of published data aimed to define the diagnostic accuracy of different values of pre-operative fibrinogen levels in predicting severe post-operative bleeding. METHODS: Data were retrieved for 2154 patients in four different studies. Severe bleeding (SB) was defined as a post-operative chest drain output>1 l/12 h. Diagnostic accuracy for prediction of SB was tested at three cutoff values of pre-operative fibrinogen (2.5 g/l, 3.0 g/l, and 3.8 g/l). RESULTS: At all the three cutoff values, pre-operative fibrinogen levels had an excellent negative predictive value, ranging from 86% to 100%. Conversely, the positive predictive value was poor at all the cutoff levels: 12% (3.8 g/l), 14% (3.0 g/l), and 19% (2.5 g/l). Overall, the accuracy of pre-operative fibrinogen levels for the prediction of SB was poor. A strategy based on pre-operative fibrinogen supplementation would lead to inappropriate treatment in > 80% of the treated patients. Overall, a trigger value of 3.8 g/l would result in an inappropriate treatment in 52% of the patients, of 3.0 g/l in 20% of the patients, and of 2.5 g/l in 4% of the patients. CONCLUSION: Correction of pre-operative fibrinogen levels below 3.8 g/l would lead to an excessive rate of inappropriate interventions. Values below 2.5 g/l could be considered.