RESUMO
BACKGROUND AND OBJECTIVES: Excessive preoperative blood orders frequently occur during the preoperative planning of resections of sarcomas. We aimed to develop a prediction score model that would be able to identify a patient cohort in which the cross-matching could be safely evaded. PATIENTS AND METHODS: We retrospectively analyzed data of 309 consecutive patients with extra-abdominal soft tissue sarcomas treated between September 2012 and December 2014. Scorecard scores for variables were calculated and summarized to a total score that can be used for risk stratification. The score was used in a logistic regression model. Results of the optimized model were described as a receiver operating characteristic curve. RESULTS: Preoperative units of red blood cells were requested for 206 (66.7%) patients, of which only 31 (10%) received them. Five parameters were identified with high predictive power. In the visualized barplot, there was an increased risk of blood transfusion with a higher score of TRANSAR. CONCLUSION: A TRANSAR score is a new tool that can predict the probability of transfusion for patients with sarcoma. This may reduce the number of preoperative cross-matching and blood product ordering and associated costs without compromising patient care.
