作者: Keyvan Karkouti , Terrence M. Yau , W. Scott Beattie
DOI:
关键词:
摘要: Purpose: In cardiac surgery with cardiopulmonary bypass (CPB), excessive blood loss requiring the transfusion of multiple red cell (RBC) units is a common complication that associated significant morbidity and mortality. The objective this study was to develop prediction rule for massive (MBT) could be used optimize management of, research on, at-risk patients. Methods: Data were collected prospectively over period from 2000 2005, on patients who underwent CPB at one hospital. Patients received ≥ five RBC within day classified as MBT. Logistic regression appropriately select weigh perioperative variables in rule, which developed initial 60% sample validated remaining 40%. Results: Of 10,667 included, 925 (8.7%) had clinical included 12 (listed order predictive value: duration, preoperative hemoglobin concentration, body surface area, nadir hematocrit, previous sternotomy, shock, platelet count, urgency surgery, age, surgeon, deep hypothermic circulatory arrest, type procedure) highly discriminative (c-index = 0.88). validation set, those low-, moderate-, high-risk by simple risk score derived 5%, 27%, 58% chance MBT, respectively. Conclusion: A accurately identified low-risk or Studies are needed determine external generalizability utility rule.