| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Department of Anesthesiology, Evanston Northwestern Healthcare, Evanston, Illinois
Address correspondence and reprint requests to Glenn S. Murphy, MD, Evanston Northwestern Healthcare, Department of Anesthesiology, 2650 Ridge Ave., Evanston, IL 60201. Address email to dgmurphy{at}core.com
Abstract
Hemodilution during cardiopulmonary bypass (CPB) is a primary risk factor for blood transfusion in cardiac surgical patients. Priming of the CPB circuit with the patients’ own blood (retrograde autologous priming, RAP) is a technique used to limit hemodilution and reduce transfusion requirements. We designed this study to examine the impact of RAP on perioperative blood product use. Using a retrospective cohort study design, the medical records of all patients undergoing CPB (excluding circulatory arrest cases) by a single surgeon were examined. Data were collected over a 24-mo period when RAP was routinely used as a blood conservation strategy (RAP group, n = 257). This group was compared with a cohort of patients during the 24 mo immediately preceding the introduction of RAP into clinical practice (no RAP group, n = 288). A small, statistically insignificant reduction in the percentage of patients receiving packed red blood cells was observed in the RAP group (44% versus 51% no RAP, P = 0.083). No differences were found between the groups in the number of units of packed red blood cells, platelets, or fresh frozen plasma transfused throughout the perioperative period. These results suggest that overall, RAP does not offer a clinically important benefit as a blood conservation technique.
IMPLICATIONS: Priming of the cardiopulmonary bypass circuit with the patients’ own blood (retrograde autologous priming) resulted in insignificant reductions in blood use in a large, unselected group of patients undergoing cardiac surgical procedures.
A primary determinant of transfusion requirements during cardiac surgery is the inevitable dilution of the red blood cell mass that occurs on initiation of cardiopulmonary bypass (CPB) (1,2). Hemodilution, which is induced by a large fixed crystalloid prime volume, may reduce hematocrit values to a level that necessitates the administration of packed red blood cells (PRBCs). Retrograde autologous priming (RAP) of the CPB circuit is a recently described technique that involves the replacement of the crystalloid prime with the patient’s own blood. After arterial and venous cannulas are inserted, the crystalloid prime is slowly drained into a recirculation bag. Up to 1100 mL of the circuit volume can be displaced by the patient’s own blood immediately before the onset of CPB (3). Data from previous small, prospective, controlled studies suggest that RAP is effective in reducing the number of cardiac patients receiving PRBCs (3–6). The aim of this investigation was to determine the impact of a RAP technique on perioperative transfusion requirements in a large, unselected population of adult patients presenting for cardiac surgery with CPB. We hypothesized that the introduction of RAP into our clinical practice would result in a significant reduction in perioperative transfusion requirements.
Methods
This retrospective cohort study was approved by the IRB of Evanston Northwestern Healthcare. The medical records of all patients who underwent cardiac surgical procedures by a single surgeon (TVV) during the study period were reviewed. Data were collected for a 24-mo period (2000–2002) when RAP was routinely used on adult patients undergoing CPB at our institution (RAP group, n = 257). This group was compared with a similar cohort of cardiac surgical patients operated on by the same surgeon during a 24-mo period immediately before the introduction of RAP into the clinical practice (no RAP group, n = 288). Data from off-pump coronary artery bypass and circulatory arrest procedures were excluded from the analysis. Demographic information was obtained from anesthesia, surgical, and medical preoperative notes. Intraoperative data, including blood product use, were obtained from anesthesia, perfusion, and surgical records. Intensive care unit (ICU) flow sheets and postoperative progress notes were reviewed to assess postoperative transfusions. The time when blood products were administered was noted (during CPB, intraoperatively, in the ICU, or on the surgical ward). Total perioperative blood use was confirmed by reviewing all transfusion records.
After premedication with midazolam and fentanyl, radial and pulmonary artery catheters were placed in all patients. Anesthesia was induced with fentanyl, midazolam, and thiopental and maintained with fentanyl (up to 15 µg/kg) and isoflurane. The use of inotropic drugs was at the discretion of the anesthesiologist and surgeon.
The extracorporeal circuit consisted of roller pumps, a membrane oxygenator, an arterial filter (Intersept Custom Pack; Medtronic, Minneapolis, MN), and a filtered cardiotomy reservoir (BCR-3500; Jostra Bentley, Irvine, CA). The circuit (uncoated) was primed with approximately 1300 mL of crystalloid solution (Normosol), 25 g of mannitol, 50 meq of bicarbonate, and heparin (10,000 U). Patients were not actively cooled during CPB; core temperatures were allowed to drift passively to 33°C–34°C. Pump flows were maintained between 2.0–2.4 L·min–1·m–2 and mean arterial blood pressures between 50–70 mm Hg. Patients were warmed to a bladder temperature of 37°C before separation from CPB. Perfusion equipment and techniques remained constant throughout the study period.
Perfusionists were introduced to the RAP technique during a 1-mo training period. At the end of this training period, all patients undergoing CPB received RAP. After heparin administration (300 U/kg) and obtaining an activated clotting time >400 s, RAP was performed as described by Rosengart et al. (3). Priming of the CPB circuit was conducted in three steps. First, crystalloid was displaced from the arterial line. A 1000-mL blood transfer bag was connected to the venous line. Blood was allowed to flow (by pressure gradients) from the aorta and through the arterial line and filter, displacing crystalloid prime into the blood transfer bag. Next, the crystalloid in the venous reservoir and oxygenator was similarly displaced using pressure gradients from the aorta. Finally, all of the prime from the venous line was displaced into the blood transfer bag at the onset of CPB. Approximately 300 mL of volume in the CPB circuit was replaced with the patients’ own blood during each of these steps. During the entire RAP process, which was completed in 5–8 min, systolic blood pressure was maintained >100 mm Hg using small bolus doses of phenylephrine (80–400 µg total) and crystalloid (100–300 mL). Sequestered prime in the blood transfer bag was infused into the CPB circuit after separation from CPB for processing via a cell saver system.
All patients were admitted to the ICU and managed according to standard protocols. Sedation was maintained in all patients with propofol infusions. Weaning and tracheal extubation were accomplished using standard criteria.
Similar blood conservation strategies were used during the four-year study period. All decisions relating to transfusions in the operating room (OR), ICU, and surgical wards were made by a single surgeon (TVV) using the criteria defined below. Approximately 500 mL of autologous blood was collected after induction of anesthesia from patients with initial hematocrits >30%. This blood was transfused after separation from CPB and protamine administration. Antifibrinolytic drugs were administered only to patients undergoing reoperative procedures; these patients received aprotinin (2 x 106 KIU pre-CPB, 2 x 106 KIU in prime, and 500,000 KIU/h during surgery). Aprotinin, aminocaproic acid, or tranexamic acid was not administered to patients undergoing primary cardiac surgery. A cell saver system was used to process and concentrate red blood cells for reinfusion after CPB. The threshold for transfusion of PRBCs was a hematocrit of 18% on CPB or a hematocrit of 22% post-CPB and postoperatively. PRBCs were administered at a higher hematocrit value if clinically symptomatic anemia developed (evidence of myocardial ischemia, heart failure, or arrhythmias). Use of platelets and fresh-frozen plasma (FFP) was based on evidence of nonsurgical bleeding (diffuse microvascular bleeding) in conjunction with appropriate abnormalities in coagulation variables (platelet count < 80,000/µL or prothrombin time/partial thromboplastin time more than 1.5 times control). Initial treatment of microvascular bleeding consisted of 1 U of platelets (obtained from a single donor by plasmapheresis) or 2 U of FFP.
Data are reported as the number of patients or median and range. Nominal data were compared between treatment groups using the 
2 or Fisher’s exact probability test. Ordinal data were compared using the Mann-Whitney rank sum test. Because nearly all interval data were determined not to be normally distributed by the Kolmogorov-Smirnov test for normality of the underlying population, these data were also reported as median and range and compared using the Mann-Whitney rank sum test. The criterion for rejection of the null hypothesis was P < 0.05. When a test was applied multiple times to the same set of data, the criterion for rejection of the null hypothesis was adjusted using the Bonferroni correction.
Multiple logistic regression analysis was performed using SigmaStat 2.03 (SPSS, Inc.; Chicago, IL) to determine predictors of the need for PRBC transfusion. Variables included in the initial multiple logistic regression analysis were gender, age, height, weight, ASA physical status, CPB duration, cross-clamp duration, baseline hematocrit, hematocrit on CPB, crystalloid infused in OR, and RAP. In the intermediate analysis, only those variables having low P values in the initial analysis were included with variables for the type of surgery, including coronary artery bypass graft surgery (CABG), valve, CABG and valve, redo CABG, redo valve, and redo CABG and valve. In the final analyses, variables with low P values were removed from the model one at a time and were excluded from the final model if their removal either did not diminish the fit of the model or actually improved it, as determined by the Pearson 2 statistic, the likelihood ratio test statistic, the Hosmer-Lemeshow statistic, and the correct prediction of both positive and reference responses. The sensitivity and specificity of the logistic model were calculated from the model-predicted reference and model-predicted positive responses (using the default threshold probability for positive classification of 0.5) and the actual reference and actual positive responses.
Relative risk (RR) and its 95% confidence interval (CI) were calculated for the risk for PRBC transfusion in no RAP patients versus the risk for PRBC transfusion in RAP patients using NCSS 2001 (Number Cruncher Statistical System, Kaysville, UT). Subset analysis was performed on the data to determine if a particular group of patients predicted a priori to be at high or low risk for perioperative transfusions benefited from RAP. Patients predicted to be at high risk for transfusions included those who were female, elderly (
70 yr), or small (body surface area [BSA] <1.7 m2, weight <70 kg) or had low initial hematocrit (<32%), or long CPB times (120 min). Low risk patients included those who were male, younger (<70 yr), or larger (BSA 1.7 m2, weight 70 kg) or had higher initial hematocrit (32%), or shorter CPB times (<120 min).
Because the diverse nature of our patient population might have explained our largely negative results, subset analysis was also performed on patients undergoing first-time CABG.
Results
During the 2 24-mo study intervals, data were collected and analyzed on 257 patients in the RAP group and 288 patients in the no RAP group. Medical records were not located for 14 patients. Complete data were identified in the 585 study patients. Table 1 presents demographic data. The two groups were comparable with respect to age, gender, height, weight, BSA, history of tobacco use, ASA physical status, and preexisting medical conditions. The two groups were similar with respect to the types of surgical procedures performed. The number of reoperative and emergency procedures did not differ between groups. Intraoperative data are presented in Table 2. Although baseline hematocrits (obtained after induction of anesthesia) were similar, hematocrits on CPB (measured 15 minutes after initiation of CPB) were significantly higher in the RAP group (26% versus 22%, P < 0.001). Patients in the RAP group also received significantly more crystalloid from the anesthesia team pre- and post-CPB than patients in the no RAP group (3000 mL versus 2700 mL, P = 0.012).
|
|
|
2 test with continuity correction and with a significance level of 0.05.
|
|
70 yr, BSA <1.7 m2, weight <70 kg, hematocrit <32%, CPB 120 min), no significant difference in the number of patients transfused PRBCs or in the number of units of PRBCs administered was observed between the RAP and no RAP groups. Use of FFP and platelets were also similar in each of the subgroups examined throughout the perioperative period. In the analysis of low-risk patients (males, age <70, BSA 1.7 m2, weight 70 kg, hematocrit 32%, CPB <120 min), we observed differences in PRBC transfusion only in patients with hematocrits 32%. In the no RAP group, a larger percentage of patients were administered PRBCs (44% versus 32%, P = 0.014) and the RR of receiving a transfusion was increased (RR, 1.38; 95% CI, 1.06–1.80). Because the diverse nature of our patient population might have explained our largely negative results, subset analysis was also performed on patients undergoing first-time CABG. However, no significant difference in the number of patients transfused PRBCs or in the number of units of PRBCs administered was observed between the RAP and no RAP groups. In addition, the RR of receiving a transfusion was not different between the no RAP and the RAP patients undergoing first-time CABG (Table 3).
There were no complications directly attributable to RAP during the study period.
Discussion
RAP of the CPB circuit is a blood conservation modality developed to limit the degree of hemodilution occurring during extracorporeal circulation. This concept was first described by Panico and Neptune in 1960 (7). The technique was revived in the late 1990s after the publication of a clinical trial by Rosengart et al. (3). Subsequent investigations have been published using modifications of the technique described by Rosengart et al. These studies demonstrated clinical efficacy of RAP in reducing the number of patients transfused or in decreasing total blood product usage (3–5). Other investigators observed no effect of RAP on postoperative or total transfusion requirements (6,8). Although most of these clinical trials were randomized, they were all limited by small sample sizes (10–57 patients in the RAP group). In addition, these studies excluded patients at highest risk for perioperative transfusion.
We performed a retrospective analysis on data from a large group of patients presenting for cardiac surgery with CPB. We confined our analysis to a single surgeon because there is significant variability between surgeons in surgical techniques, transfusion triggers, and the use of blood conservation techniques (9,10). The two cohorts of subjects that were studied appeared similar in all preoperative and intraoperative variables. In particular, the 2 groups were similar with respect to the risk factors for perioperative transfusions. In addition, surgical, anesthetic, and perfusion management appeared to remain unchanged during the study period. There are limitations inherent in any retrospective data analysis, however. During the 4 years data were collected, subtle changes in patient demographics or perioperative management may have occurred. Although the groups appeared evenly matched, differences between cohorts that were not identified in our retrospective analysis might have existed.
The introduction of RAP into our clinical practice resulted in a small, statistically insignificant reduction in the percentage of patients transfused PRBCs (44% in RAP group versus 51% in no RAP group). The median number of units of PRBCs administered during the study was small in both groups and did not differ significantly. As in previous studies, we observed that hematocrit on initiation of CPB was significantly higher in the RAP group than in the no RAP group. Unlike other investigators, however, we did not observe a significant reduction in the number of units of PRBCs transfused on CPB or during the remainder of the intraoperative or postoperative periods. A willingness to accept a lower hematocrit on CPB or on separation from CPB may explain these findings. The initial decrease in hematocrit resulting from hemodilution was not significant enough to reach our transfusion threshold (18%) in the majority of patients in the no RAP group or the RAP group. We also observed that more crystalloid was administered to RAP patients in the OR. Hemodilution resulting from this additional crystalloid administration may have decreased post-CPB hematocrit values and negated benefits achieved by RAP. In addition, the number of units of platelets and FFP administered intraoperatively and postoperatively was small and did not differ significantly between our groups.
A subset analysis of the data was performed to determine if there was a group of patients that benefited from RAP. Data from Rosengart et al. (3) and Balachandran et al. (5) suggested that RAP reduced the risk of exposure to a PRBC transfusion to a greater degree in patients predetermined to be high-risk for perioperative transfusions. Several preoperative and intraoperative variables have been identified as highly significant predictors of red blood cell use in cardiac surgery. Factors associated with increased transfusion risk include female gender, increased age (>70–75 years), small size (BSA <1.7 m2, weight <70 kg), low initial hematocrit (<32%–34%), and long CPB times (>120 minutes) (11–15). However, patients at highest risk for requiring PRBC were excluded from previous trials. We specifically examined transfusion data of patients with any of these risk factors to assess efficacy of RAP. We observed no significant benefits of RAP on any transfusion outcomes in this high-risk group of patients undergoing CPB or in patients undergoing first-time CABG. In the analysis of patients determined to be low-risk for perioperative PRBC use, one group appeared to benefit from RAP (Table 5). In the RAP group, patients with preoperative hematocrits >32% had a significantly less risk of receiving PRBCs. This finding suggests that patients with a larger initial red blood cell mass may derive a greater benefit from this technique. The use of pre-CPB autologous donation in patients with initial hematocrits >30%, in addition to RAP, may have accounted for the observed benefit in this group. RAP had little effect on PRBC use in small patients with low initial hematocrits. This result is not surprising, as there is a relatively limited ability of RAP to prevent excessive hemodilution in patients with a small red cell mass. In both the RAP and no RAP groups, the majority of patients with an initial hematocrit <32% (76% versus 84%), a weight <70 kilograms (67% versus 72%), or a BSA <1.7 m2 (72% versus 81%) were administered red cells.
Hemodynamic instability can develop when blood is drained from the patient into the arterial and venous lines, and may be worsened if pre-CPB autologous blood donation has been performed. Reduced arterial blood pressure (systolic blood pressure <100 mm Hg) was treated in a standardized manner using crystalloid (100–300 mL) and phenylephrine bolus doses (80–400 µg). If these measures were unsuccessful, the RAP process was terminated. The incidence of hypotension during RAP was not determined in this investigation. However, hypotension not responsive to these measures occurred in <1% of patients. Although induced hypovolemia may potentially adversely affect organ perfusion, the time from the start of the RAP until onset of CPB was typically brief (<5–8 minutes).
In most previous prospective clinical trials, retrograde autologous priming resulted in significant reductions (15%–32%) in the number of patients receiving PRBCs (3–5). Differences in our study design may have accounted for the lack of significant benefit of RAP on transfusion outcomes observed in our study. Unlike most previous RAP investigators, we did not routinely use antifibrinolytic drugs. The administration of antifibrinolytic drugs, by reducing excessive mediastinal bleeding, may have produced further reductions in blood product use in the RAP group. Finally, our study was underpowered to detect the observed difference in the requirement for PRBCs between the groups with an 
of 0.05 and 80% power. Forty-four percent of patients in the RAP group received packed red cells versus 51% of patients in the no RAP group. Group sample sizes of 657 (each) would have been required to achieve 80% power to detect the observed difference of 0.07 between the null hypothesis that both group proportions of patients requiring PRBC transfusions are 0.51 and the alternative hypothesis that the proportion in the RAP group is 0.44, using a one sided 2 test with continuity correction and an of 0.05.
In conclusion, we observed only a minimal reduction in PRBC use when RAP was routinely applied as a blood conservation modality. Patients with a larger initial red cell mass appear to derive a greater benefit from this technique. Further studies are needed to define the role of RAP in patients undergoing CPB.
Acknowledgments
Supported, in part, by the Department of Anesthesiology, Evanston Northwestern Healthcare.
References
This article has been cited by other articles:
|
|
 |
|
  F. Pappalardo, C. Corno, A. Franco, G. Giardina, A.M. Scandroglio, G. Landoni, G. Crescenzi, and A. Zangrillo Reduction of hemodilution in small adults undergoing open heart surgery: a prospective, randomized trial Perfusion, September 1, 2007; 22(5): 317 - 322. [Abstract] [PDF]
|
|
|
|
 |
|
 The Society of Thoracic Surgeons Blood Conservatio, V. A. Ferraris, S. P. Ferraris, S. P. Saha, E. A. Hessel II, C. K. Haan, B. D. Royston, C. R. Bridges, R. S.D. Higgins, G. Despotis, et al. Perioperative Blood Transfusion and Blood Conservation in Cardiac Surgery: The Society of Thoracic Surgeons and The Society of Cardiovascular Anesthesiologists Clinical Practice Guideline Ann. Thorac. Surg., May 1, 2007; 83(5_Supplement): S27 - S86. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
