Vox Sanguinis (2014) © 2014 International Society of Blood Transfusion DOI: 10.1111/vox.12163
ORIGINAL PAPER
Transfusion indication predictive score: a proposed risk stratification score for perioperative red blood cell transfusion in cardiac surgery M. Al-Khabori,1 A. Z. Al-Riyami,1 M. Mukaddirov2 & H. Al-Sabti2 1
Department of Hematology, Sultan Qaboos University Hospital, Muscat, Oman Cardiothoracic Surgery Division, Department of Surgery, Sultan Qaboos University Hospital, Muscat, Oman
2
Background and Objectives Red blood cell transfusion is known to be associated with increased morbidity and mortality in cardiac surgery. This study was performed to derive a score to predict that risk in our patients. Materials and Methods Clinical details of patients who underwent cardiac surgery at the Sultan Qaboos University Hospital over 5 years were reviewed. We used univariable and multivariable logistic regression to develop the score, the Hosmer–Lemeshow test for calibration, the receiver operator curve for discrimination and the bootstrap procedure for internal validation. Results The sample included 413 patients. The following were found to be statistically significant transfusion predictors (score given): cerebrovascular disease (4), use of aspirin or clopidogril within 7 days of surgery (3), non-elective surgery (2), haematocrit 60 years (1), diabetes mellitus (1) and male gender (-2). We classified the observations into three groups: group 1 with total score of 5. The calculated probabilities of receiving transfusion were 42%, 63% and 91% for groups 1, 2 and 3 respectively.
Received: 7 February 2014, revised 21 April 2014, accepted 7 May 2014
Conclusion We derived a simple score that can be utilized to assess the need of blood transfusion in patients undergoing cardiac surgery. We are the first to report G6PD deficiency and history of cerebrovascular disease as predictors. We recommend prospective external validation of the proposed score on a larger cohort of patients. Key words: red cell components, transfusion strategy, transfusion-surgery.
Introduction Transfusion in cardiac surgery has been reported to be associated with increased postoperative mortality and morbidity with increased risk of infection, renal failure, cardiac events and neurological insults [1–6]. Moreover, it is associated with the risk of transfusion related complications, including transfusion related acute lung injury, Correspondence: Hilal Al-Sabti, Cardiothoracic Surgery Division, Department of Surgery, Sultan Qaboos University Hospital, P.O box 38, Muscat 123, Oman E-mail: [email protected]
transfusion associated circulatory overload and transfusion transmitted infections [7–9]. The current practice in cardiac surgeries is to transfuse with a nadir haemoglobin (Hb) of 80 g/dl or less [10]. The level of 80 g/dl was reported to be safe in eliminating the risk of anaemia and reducing the costs of transfusion, and the resources needed to manage related complications [10]. That said, transfusion should not be dictated by a particular Hb trigger, but rather should be individualized based on the patient’s risk of developing complications of inadequate oxygenation [11]. Despite published guidelines to guide transfusion in this group of patients [12], a wide variation in the transfusion rate
1
2 M. Al-Khabori et al.
remains, ranging from 40% to 90% [13–15]. This is likely to be caused by the lack of high-quality evidence on which physicians can base their decisions [8, 11] and due to the variation in institutional protocols in regard to transfusion threshold, preoperative discontinuation of antiplatelets agents and the use of blood conservation modalities. In view of the above, a number of different scoring models have been developed to predict the need for allogenic packed red blood cell (PRBC) transfusion in cardiac surgeries [16–21]. Both the TRACK and the TRUST scoring systems were found to be easy to use compared with the other complex scores and have better performance in predicting the need for blood transfusion [20]. Factors found to be strong predictors are female gender, low preoperative Hb level and old age, among others [18]. However, these scores can be limited by the fact that they were designed based on preexisting cardiac databases, which limits the factors available for assessment for the models development to what is available in these databases [21]. In addition, these scores may not be generalizable, nor being fully applicable to certain populations due to the differences in the baseline characteristics of treated patients. For instance, certain populations have a high rate of inherited haemolytic anaemias including sicklecell disease. The rule that such conditions, among others, in predicting the risk may have not been assessed in other score models due to their low prevalence in the studied populations. We therefore hypothesized that the perioperative transfusion predictors in our region may differ from what has been previously published. Herein, we aim to develop and validate a score for predicting the need for PRBC transfusion in patients undergoing cardiac surgery at our institution.
Materials and methods Study design Ethical approval was obtained from the Institution’s Ethics Board, and the need for informed consent from the patients was waived. Retrospective review of the cardiothoracic database of all adult patients who underwent cardiac surgery at the Cardiothorasic Surgery Division at the Sultan Qaboos University Hospital (SQUH) between 2008 and 2012 was performed. All patients who had undergone cardiac procedures, including those not requiring cardiopulmonary bypass (CPB), were included. Procedures included were on- and off-pump coronary artery bypass graft surgery (CABG), valve surgery, as well as other major cardiac surgeries (e.g. aortic dissection procedures). All procedures were included regardless of their urgency. The primary outcome was intraoperative and
postoperative exposure to allogeneic blood transfusion throughout the admission.
Operative protocol The majority of patients underwent surgery using cardioplegia under moderate hypothermia. Deep hypothermic circulatory arrest at 20°C is applied in aortic arch surgeries. The CPB circuit is primed with crystalloid solutions at a volume of 30 ml/kg body weight. The perfusion system (centrifugal pump or roller) is used with standard polyvinyl chloride tubing and membrane oxygenators. Heparin 300 IU/kg is given, aiming for a target ACT of 480s. Antegrade/retrograde warm or cold blood cardioplegia is used. When CPB is used, shed blood is aspirated into a cardiotomy reservoir from the intracardiac chambers as well as the pleural and pericardial cavities before it is reinfused. A high-potassium solution is used to arrest the heart promptly during induction of cardioplegia, after which it is changed to a low-potassium solution to limit hyperkalemia. After termination of the CPB, heparin is neutralized by protamine sulphate at 3 mg/kg. Aprotinin has never been used, neither other antifibrinolytics. Pre-, intra- and postoperative transfusion of PRBC, fresh-frozen plasma (FFP), platelet concentrates and cryoprecipitate follow the local transfusion protocol. Patients are transfused preoperatively if the Hb level is 40% vs. 40% or less), use of aspirin or clopidogrel within 7 days of surgery (yes vs. no), type of surgery (CABG, value surgery, CABG and value combined, and other types of surgeries), non-elective surgery (yes vs. no), preoperative Hct level (≥35% vs.
ORIGINAL PAPER
Transfusion indication predictive score: a proposed risk stratification score for perioperative red blood cell transfusion in cardiac surgery M. Al-Khabori,1 A. Z. Al-Riyami,1 M. Mukaddirov2 & H. Al-Sabti2 1
Department of Hematology, Sultan Qaboos University Hospital, Muscat, Oman Cardiothoracic Surgery Division, Department of Surgery, Sultan Qaboos University Hospital, Muscat, Oman
2
Background and Objectives Red blood cell transfusion is known to be associated with increased morbidity and mortality in cardiac surgery. This study was performed to derive a score to predict that risk in our patients. Materials and Methods Clinical details of patients who underwent cardiac surgery at the Sultan Qaboos University Hospital over 5 years were reviewed. We used univariable and multivariable logistic regression to develop the score, the Hosmer–Lemeshow test for calibration, the receiver operator curve for discrimination and the bootstrap procedure for internal validation. Results The sample included 413 patients. The following were found to be statistically significant transfusion predictors (score given): cerebrovascular disease (4), use of aspirin or clopidogril within 7 days of surgery (3), non-elective surgery (2), haematocrit 60 years (1), diabetes mellitus (1) and male gender (-2). We classified the observations into three groups: group 1 with total score of 5. The calculated probabilities of receiving transfusion were 42%, 63% and 91% for groups 1, 2 and 3 respectively.
Received: 7 February 2014, revised 21 April 2014, accepted 7 May 2014
Conclusion We derived a simple score that can be utilized to assess the need of blood transfusion in patients undergoing cardiac surgery. We are the first to report G6PD deficiency and history of cerebrovascular disease as predictors. We recommend prospective external validation of the proposed score on a larger cohort of patients. Key words: red cell components, transfusion strategy, transfusion-surgery.
Introduction Transfusion in cardiac surgery has been reported to be associated with increased postoperative mortality and morbidity with increased risk of infection, renal failure, cardiac events and neurological insults [1–6]. Moreover, it is associated with the risk of transfusion related complications, including transfusion related acute lung injury, Correspondence: Hilal Al-Sabti, Cardiothoracic Surgery Division, Department of Surgery, Sultan Qaboos University Hospital, P.O box 38, Muscat 123, Oman E-mail: [email protected]
transfusion associated circulatory overload and transfusion transmitted infections [7–9]. The current practice in cardiac surgeries is to transfuse with a nadir haemoglobin (Hb) of 80 g/dl or less [10]. The level of 80 g/dl was reported to be safe in eliminating the risk of anaemia and reducing the costs of transfusion, and the resources needed to manage related complications [10]. That said, transfusion should not be dictated by a particular Hb trigger, but rather should be individualized based on the patient’s risk of developing complications of inadequate oxygenation [11]. Despite published guidelines to guide transfusion in this group of patients [12], a wide variation in the transfusion rate
1
2 M. Al-Khabori et al.
remains, ranging from 40% to 90% [13–15]. This is likely to be caused by the lack of high-quality evidence on which physicians can base their decisions [8, 11] and due to the variation in institutional protocols in regard to transfusion threshold, preoperative discontinuation of antiplatelets agents and the use of blood conservation modalities. In view of the above, a number of different scoring models have been developed to predict the need for allogenic packed red blood cell (PRBC) transfusion in cardiac surgeries [16–21]. Both the TRACK and the TRUST scoring systems were found to be easy to use compared with the other complex scores and have better performance in predicting the need for blood transfusion [20]. Factors found to be strong predictors are female gender, low preoperative Hb level and old age, among others [18]. However, these scores can be limited by the fact that they were designed based on preexisting cardiac databases, which limits the factors available for assessment for the models development to what is available in these databases [21]. In addition, these scores may not be generalizable, nor being fully applicable to certain populations due to the differences in the baseline characteristics of treated patients. For instance, certain populations have a high rate of inherited haemolytic anaemias including sicklecell disease. The rule that such conditions, among others, in predicting the risk may have not been assessed in other score models due to their low prevalence in the studied populations. We therefore hypothesized that the perioperative transfusion predictors in our region may differ from what has been previously published. Herein, we aim to develop and validate a score for predicting the need for PRBC transfusion in patients undergoing cardiac surgery at our institution.
Materials and methods Study design Ethical approval was obtained from the Institution’s Ethics Board, and the need for informed consent from the patients was waived. Retrospective review of the cardiothoracic database of all adult patients who underwent cardiac surgery at the Cardiothorasic Surgery Division at the Sultan Qaboos University Hospital (SQUH) between 2008 and 2012 was performed. All patients who had undergone cardiac procedures, including those not requiring cardiopulmonary bypass (CPB), were included. Procedures included were on- and off-pump coronary artery bypass graft surgery (CABG), valve surgery, as well as other major cardiac surgeries (e.g. aortic dissection procedures). All procedures were included regardless of their urgency. The primary outcome was intraoperative and
postoperative exposure to allogeneic blood transfusion throughout the admission.
Operative protocol The majority of patients underwent surgery using cardioplegia under moderate hypothermia. Deep hypothermic circulatory arrest at 20°C is applied in aortic arch surgeries. The CPB circuit is primed with crystalloid solutions at a volume of 30 ml/kg body weight. The perfusion system (centrifugal pump or roller) is used with standard polyvinyl chloride tubing and membrane oxygenators. Heparin 300 IU/kg is given, aiming for a target ACT of 480s. Antegrade/retrograde warm or cold blood cardioplegia is used. When CPB is used, shed blood is aspirated into a cardiotomy reservoir from the intracardiac chambers as well as the pleural and pericardial cavities before it is reinfused. A high-potassium solution is used to arrest the heart promptly during induction of cardioplegia, after which it is changed to a low-potassium solution to limit hyperkalemia. After termination of the CPB, heparin is neutralized by protamine sulphate at 3 mg/kg. Aprotinin has never been used, neither other antifibrinolytics. Pre-, intra- and postoperative transfusion of PRBC, fresh-frozen plasma (FFP), platelet concentrates and cryoprecipitate follow the local transfusion protocol. Patients are transfused preoperatively if the Hb level is 40% vs. 40% or less), use of aspirin or clopidogrel within 7 days of surgery (yes vs. no), type of surgery (CABG, value surgery, CABG and value combined, and other types of surgeries), non-elective surgery (yes vs. no), preoperative Hct level (≥35% vs.
