| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1996;61:8-9
© 1996 The Society of Thoracic Surgeons
Department of Medicine and Pathology, Washington University School of Medicine, Barnes Hospital, St. Louis, Missouri
It has been estimated that 11% of the national red cell resource was used for transfusion support of patients undergoing coronary artery bypass grafting (CABG) in the United States in 1990 [1]. The magnitude of this need, along with the realization that blood safety was compromised by viral agents such as human immunodeficiency virus and hepatitis C virus, has led to studies that have served as descriptive analyses of transfusion outcomes in patients undergoing CABG. For example, determinants of blood use during myocardial revascularization have been identified to include such preoperative factors as female sex, age, preoperative hematocrit level, and preoperative red blood cell mass [2, 3]. Later, a multicenter study of transfusion outcomes in patients undergoing primary, elective CABG demonstrated a marked variability in the transfusion of red cells, plasma, and platelets among institutions, despite apparent homogeneity in these preoperative patient factors that were known to determine blood use [4]. The variability in transfusion outcomes independent of patient factors has been attributed to differences in transfusion triggers, surgical technique, blood conservation strategies, inappropriate transfusions, or a combination of these. What has been lacking, until now, is a validated scoring system that can assess the relevant patient factors, stratify patients at risk for transfusion, and predict transfusion outcomes prospectively.
In this issue of The Annals of Thoracic Surgery, Magovern and colleagues [5] have developed a model for predicting transfusion after CABG, using a retrospective analysis of outcomes in more than 2,000 patients to develop a scoring system, along with performing prospective validation of their scoring system in more than 400 additional patients. The observed rates of the validation group fell within the 95% confidence intervals of the predicted rates. Some of the preoperative factors that affected the need for blood include the usual suspects: female sex, age, low body mass, and low red cell mass. It is also not surprising that catheter-induced coronary occlusion, emergent or urgent operation, and redo operation were additional important preoperative factors. However, the evidence that preoperative comorbid conditions and diseases such as insulin-dependent diabetes, peripheral vascular disease, renal insufficiency, and poor left ventricular function contributed independently to the need for transfusion are important new findings. These preoperative comorbidities most likely also explain the operative and postoperative predictors of transfusion listed in their Table 4, because preexisting comorbidities have been identified to be directly related to morbid and mortal outcomes in CABG patients postoperatively [6].
What is unclear is whether these comorbid factors, and their effect on transfusion outcomes, represent the cart or the horse. Did ``sick'' patients require more transfusions, or did they receive more transfusions because they were ``sick''? Although Magovern and associates list general hemoglobin levels as parameters for transfusion, they state that the decision to transfuse blood was made individually. These decisions were almost certainly affected by the clinicians' knowledge of comorbid diseases. Magovern and associates' report does not include data regarding hemoglobin levels at the time of transfusion or at discharge among the groups stratified by their transfusion risk score, which would prove this point. Nevertheless, their analysis does document something that is understood by all clinicians, and that is that patients who are ``sick'' are transfused differently than patients who are otherwise ``healthy.''
What should we do with this information, and how can this model be used in the clinical practice of surgery for coronary revascularization? The first issue might be the process of informed consent. In the same way that Higgins and colleagues' [6] classification can be used to estimate the morbidity and mortality of a CABG procedure for an individual patient, and presumably be used as part of the informed consent process for the surgical procedure, Magovern and associates' transfusion risk score can be used to estimate, for the individual patient, the likelihood of transfusion. This information is important to patients.
The second (and related) issue is the possible use of intervention(s) designed to reduce the anticipated likelihood of transfusion. Although Magovern and associates speculate that pharmacologic agents to reduce blood loss might be useful, they have really given us no data to support this, because blood loss was not analyzed as a factor related to transfusion. They did find that duration on bypass was only a minor predictor of transfusion (which is contrary to the literature and most clinician's experience), but this is probably due to the low percentage of redos and absence of combined procedures in both series. A more relevent issue is the procurement of autologous blood in patients with high transfusion scores. Most of the predictors listed in Table 3 would make preoperative autologous blood donation impractical, and even undesirable. However, the patients in Magovern and associates' series did not have reinfusion of postoperative shed blood and apparently did not undergo ``blood pooling'' as a hemodilution technique immediately before bypass. Although the value of these techniques is poorly defined, the controversy regarding their value arises in large part because previous studies have not used a transfusion scoring system to stratify for patient-related risk factors and have not used comparable or even ``standardized'' transfusion triggers.
This last point brings us to the third issue of consequence for a transfusion scoring model. Now that we can predict transfusion outcomes related to patient factors, why not use blood transfusion and blood conservation algorithms to minimize the variability of transfusion outcomes related to institutional (procedural) and physician (transfusion practices) factors? Such algorithms could take into account patient heterogeneity by using the Higgins and Magovern stratification of ``standard'' and ``increased'' risk patients. Such an approach would enable physicians to analyze transfusion outcomes, and the relationship of these outcomes to transfusion triggers, autologous blood procurement, pharmacologic interventions, and even emerging blood substitutes. Perhaps the time has arrived when blood transfusions and blood conservations should be administered according to algorithms that are incorporated into the daily practice of coronary revascularization [7]. The knowledge that we learn from the standardization of these practices, and the comparison of transfusion outcomes between different institutions or different therapeutic approaches, would form an important database that could serve as a reference for the continuous improvement of care in this important surgical setting [8].
Footnotes
Address reprint requests to Dr Goodnough, Division of Laboratory Medicine, Department of Medicine and Pathology, Washington University School of Medicine, 660 S Euclid Ave, Box 8118, St. Louis, MO 63110.
References
Related Article
This article has been cited by other articles:
|
|
 |
|
  H. Reinecke, T. Fetsch, N. Roeder, C. Schmid, A. Winter, M. Ribbing, E. Berendes, M. Block, H. H. Scheld, G. Breithardt, et al. Emergency coronary artery bypass grafting after failed coronary angioplasty: what has changed in a decade? Ann. Thorac. Surg., December 1, 2000; 70(6): 1997 - 2003. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. T. Goodnough, G. J. Despotis, and C. A. Parvin Erythropoietin Therapy in Patients Undergoing Cardiac Operations Ann. Thorac. Surg., December 1, 1997; 64(6): 1579 - 1580. [Full Text]
|
|
|
|
 |
|
 D. M. Wingerchuk, K. N. Krecke, and J. R. Fulgham Multifocal brain MRI artifacts secondary to embolic metal fragments Neurology, November 1, 1997; 49(5): 1451 - 1453. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. K. Rosengart, KarlH. Krieger, and J. A. Magovern Predicting Transfusion Ann. Thorac. Surg., December 1, 1996; 62(6): 1885 - 1886. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
