(Chest. 2003;124:1847-1851.)
© 2003
American College of Chest Physicians
Autotransfusion of Shed Mediastinal Blood*
A Risk Factor for Mediastinitis After Cardiac Surgery? Results of a Cluster Investigation
Sandra Dial, MD, MSc;
Dao Nguyen, MD and
Dick Menzies, MD, MSc
* From the Respiratory Epidemiology Unit (Dr. Menzies), Montreal Chest Institute, McGill University, Montreal; Department of Critical Care (Dr. Dial), Sir Mortimer B. Davis Jewish General Hospital, McGill University, Montreal; and Respiratory Division (Dr. Nguyen), McGill University Health Centre, McGill University, Montreal, Canada.
Correspondence to: Sandra Dial, MD, MSc, Montreal Chest Institute, 3650 St. Urbain, Room K3.02, Montreal, PQ, Canada, H2X 2P4; e-mail: sandra.dial{at}mcgill.ca
 |
Abstract
|
|---|
Rationale: After the introduction of autotransfusion of shed mediastinal blood following cardiac surgery, the incidence of mediastinitis increased. The role of autotransfusion in the increased occurrence of this serious complication was examined.
Methods: Using a case-control design, the preoperative, intraoperative, and postoperative characteristics of 11 patients with mediastinitis were compared to those of 33 randomly selected patients undergoing cardiac surgery between September 1, 2000, and April 15, 2001 (control subjects).
Results: Patients with mediastinitis were significantly more likely to have a body mass index > 30 (unadjusted odds ratio [OR], 9.9; 95% confidence interval [CI], 2.3 to 42.5), to have received antibiotic therapy during the 2 weeks prior to cardiac surgery (OR, 12.0; 95% CI, 1.1 to 131), or to have required re-exploration within 24 h of the original operation (OR, 8.3; 95% CI, 1.8 to 39). Patients with mediastinitis had 3.4 known risk factors for mediastinitis, compared to only 1.4 risk factors per control subject (p = 0.0001), and longer duration of autotransfusion. After adjustment for other risk factors, autotransfusion for > 6 h was significantly associated with the development of mediastinitis (adjusted OR, 11.9; 95% CI, 1.4 to 97.2).
Conclusion: Retransfusion of shed mediastinal blood for > 6 h after cardiac surgery was an independent risk factor for mediastinitis.
Key Words: autotransfusion • cardiac surgery • deep sternal wound infections • mediastinitis • postoperative complications
|
Introduction
|
|---|
Deep sternal wound infections, including mediastinitis, are infrequent complications of cardiac surgery, and are associated with significant morbidity, prolonged hospitalizations, and high mortality.1
In two large series,2
3
this complication occurred in 0.16%, or 2.2% of patients, but rates as high as 8% have been reported.4
Risk factors commonly associated with mediastinitis include obesity,1
2
5
diabetes mellitus,1
6
7
chronic obstructive lung disease,2
use of both internal mammary arteries,6
7
longer cardiopulmonary bypass time,5
and postoperative allogeneic blood transfusion.8
Autotransfusion (or retransfusion) of shed mediastinal blood has been widely used in the last 3 decades in an effort to reduce the need for allogeneic blood transfusion.9
10
11
12
It is still unclear if postoperative infections,13
hemorrhage,9
14
or hematologic abnormalities15
16
are improved or worsened by this practice. In two large cohort studies2
3
of risk factors for mediastinitis following cardiac surgery, the practice of autotransfusion of shed mediastinal blood was not examined. In two small studies,10
17
the shed mediastinal blood collected in the autotransfuser was contaminated in as many as 19% of patients after 6 h; however, these studies did not detect an increased risk of mediastinitis,10
17
although they may have been underpowered to detect this relatively rare complication.
Between January 4, 2001, and April 15, 2001, we observed a cluster of cases of mediastinitis occurring in 11 of 150 patients (7.3%) undergoing cardiac surgery, much more than the usual occurrence of 0.7 to 1.8% per year at this institution. This study was performed to try to investigate the cluster, but as the practice of autotransfusion of shed mediastinal blood had been recently introduced in our institution in September 2000, becoming routine by December 2000, we hypothesized a possible association with this new practice. A case-control study was performed to investigate possible causes, including our practice of autotransfusion of shed mediastinal blood.
|
Materials and Methods
|
|---|
Setting
The study was conducted in a 640-bed, university affiliated, tertiary care hospital. Approximately 500 cardiac surgical procedures, including coronary artery bypass grafting and valve replacements, are performed at the study center annually.
Autotransfusion Protocol
Autotranfusion of shed mediastinal blood was introduced and was the only form of postoperative blood salvage procedure utilized in September 2000. Its use increased steadily over the next few months, becoming routine after December 2000. The decision to reinfuse the shed mediastinal blood was made by the medical team. In general, autotransfusion was not instituted if < 250 mL were shed in the first 3 h postoperatively, or if a coagulopathy was suspected. The system used was the Pleur-Evac Sahara Plus Continuous Reinfusion Autotransfusion System (Genzyme Surgical Products; Boston, MA). The shed mediastinal blood, which was nonwashed, was collected continuously, and the blood collected was retransfused in the hour following collection. Autotransfusion was continued until there was no mediastinal drainage for 2 consecutive h, with no time limit on the duration.
Patients and Control Subjects
Patients were those who acquired mediastinitis within 45 days after cardiac surgery, as this allowed us to include all of those who appeared to have acquired mediastinitis within the time period of the cluster. This was defined as the presence of pus at the sternal incision site with sternal dehiscence, and/or positive blood culture findings, wound swab or sternal soft tissue, and a clinical diagnosis of mediastinitis. Thirty-three control subjects were randomly selected from all patients who underwent cardiac surgery, during the period from September 1, 2000, to April 15, 2001, but did not acquire mediastinitis. In order to assess all possible factors, and also to ensure that control subjects would have differed in their likelihood of being autotransfused by chance only, control subjects were not matched to patients with mediastinitis on any clinical parameters except that they would have had their surgery during the period in which autotransfusion was being practiced.
Data Collection
From the medical records of the patients with mediastinitis and control subjects, the following data were extracted using standardized data collection sheets: age, gender, body mass index (BMI), diabetes mellitus, chronic obstructive lung disease, hypertension, left ventricular function, preoperative hemoglobin, hospital days, and antibiotic therapy in the 2 weeks prior to cardiac surgery. Intraoperative variables recorded included procedure performed, cardiopulmonary bypass pump time, total operation time, use of both internal mammary arteries, need for re-exploration, prophylactic antibiotic use, operating room in which surgery was performed, and personnel (surgeon, nurses, anesthetists, assistants, and perfusionist). Postoperative variables recorded included duration of mechanical ventilation, ICUs length of stay, blood loss, transfusion requirements, postoperative vasopressor requirements, death, and cause of death. Data on the autotransfusion included use of the autotransfusion, number of hours used, total mediastinal drainage, and volume retransfused. Information regarding mediastinitis included date of diagnosis and bacterial culture results.
Data Analysis
Patients with mediastinitis were compared to control subjects with respect to preoperative, intraoperative, and postoperative variables. Differences between patients with mediastinitis and control subjects were tested for significance using an unmatched Student t test for continuous variables and 
2 tests for categorical variables.
Multivariate logistic regression was performed to obtain an estimate of the effect of prolonged autotransfusion after adjustment for other risk factors. Because of the limited number of patients with mediastinitis, it was recognized that we would only be able to evaluate two to three variables in the model. In order to try to adjust for this, each patient received a numeric score representing the arithmetic sum of the presence or absence of eight factors. They received a score of 1 for each risk factor present, or a score of 0 if that risk factor was absent. Seven risk factors were included that had been identified in previous studies (diabetes, obesity defined as a BMI > 30, COPD, need for re-exploration, use of both internal mammary arteries, cardiopulmonary bypass time > 2 h, and transfusion of > 2 U of packed RBCs (at the time of the study, the transfused units were not leukoreduced). One additional risk factor was included because it was significantly associated, in univariate analysis, with mediastinitis in this study. This was antibiotic therapy for a presumed infection in the 2 weeks prior to undergoing cardiac surgery.
Multivariate logistic regression was also performed to obtain an estimate of the effect of prolonged autotransfusion after adjustment for the two factors that were most important on univariate analysis: obesity and re-exploration in the first 24 h. Prolonged was defined as > 6 h, because 6 h was a duration that had been used in prior studies,17
18
and we had found that the frequency also appeared to rise at 6 h in the study population (Fig 1
). All analysis was performed using SAS software (SAS Institute; Cary, NC). All statistical tests were two tailed, and differences were considered significant if p < 0.05; 95% confidence intervals (CIs) were calculated for the odds ratios (ORs).
|
Results
|
|---|
Mediastinitis was diagnosed a median of 16 days postoperatively (range, 6 to 44 days). The causative organism was Staphylococcus aureus in seven patients, Staphylococcus epidermidis in two patients, and not identified in two patients. As shown in Table 1 , patients with mediastinitis were more likely to be obese, to have received a course of antibiotics for a presumed infection in the 2 weeks prior to cardiac surgery, and to have undergone re-exploration within 24 h of their initial surgical procedure. None of the other intraoperative or postoperative variables were significantly different between the patients with mediastinitis and control subjects.
View this table:
[in this window]
[in a new window]
|
Table 1.. Comparison of Preoperative, Intraoperative, and Postoperative Characteristics of Patients With Mediastinitis and Control Subjects, Univariate Analysis*
|
|
Patients with mediastinitis were more likely than the control subjects to have received > 2 U of allogenic packed RBCs (Table 2
), and had significantly longer duration of autotransfusion, although any use of autotransfusion was not significantly different. Longer duration of autotransfusion was associated with larger volumes retransfused, but total hemorrhage postoperatively was similar in patients with mediastinitis and control subjects. This was because some patients who bled significantly did not receive autotransfusion, because of concerns regarding a coagulopathy, potentially worsened with autotransfusion.
View this table:
[in this window]
[in a new window]
|
Table 2.. Description of Bleeding and Autotransfusion Variables in Patients With Mediastinitis and Control Subjects*
|
|
To provide a summary estimate of all other risk factors for mediastinitis, the arithmetic sum of eight possible risk factors for mediastinitis was calculated for the patients with mediastinitis and control subjects. As seen in Table 1
, patients with mediastinitis had significantly more risk factors than control subjects; after adjusting for this difference using multivariate regression (Table 3
, model 1), autotransfusion for > 6 h was independently associated with mediastinitis. In a model that adjusted only for the two most important risk factors found on univariate analysis (Table 3
, model 2), prolonged autotransfusion was still significantly associated with mediastinitis.
|
Discussion
|
|---|
It is generally believed that multiple factors contribute to the occurrence of mediastinitis.2
In this study, the 11 patients who acquired mediastinitis were more likely than the 33 control subjects to have several previously recognized risk factors. These included obesity,2
5
COPD,2
more allogenic transfusions,8
re-exploration within 24 h,2
6
as well as antibiotic therapy in the 2 weeks prior to surgery. After adjusting for the presence of these risk factors, autotransfusion of shed mediastinal blood for > 6 h was a significant independent risk factor for the development of mediastinitis.
Since the introduction of autotransfusion of shed mediastinal blood in 1978,10
reports of its safety9
13
and efficacy in reducing the need for postoperative blood transfusion9
14
18
have been inconsistent. Patients who receive retransfused shed mediastinal blood have been reported to have an increased incidence of early postoperative fever, band neutrophilia, and plasma-free hemoglobin,19
as well as wound infections.13
The occurrence of positive bacterial culture findings of shed mediastinal blood has been shown to be as high as 19% after 6 h.17
Possible explanations for the correlation with prolonged use and larger volumes rather than any use of the shed mediastinal blood could be as a result of a time-dependent risk of bacterial contamination. The association with larger volumes retransfused implies that increased exposure to other factors, such as cytokines or other inflammatory mediators that may be present in the autotransfused blood, may also be a factor. Our data do not suggest that excessive mediastinal bleeding is responsible, since the total volume drained was not larger in the patients with mediastinitis as compared to the control subjects. The key risk factor was longer duration, especially > 6 h.
Bacterial contamination may occur due to several reasons. Shed mediastinal blood may pool in the collecting system for several hours, since many autotransfusion protocols require that a minimum volume is collected before transfusion is initiated. The RBCs in the collected blood may hemolyze, releasing free hemoglobin that suppresses macrophage function.20
The variable incidence of positive bacterial culture findings reported in previous studies21
21
22
23
could be due to differences in culture techniques, the amount of blood cultured, use of antibiotic prophylaxis (which can inhibit bacterial growth in culture), or surveillance protocols.21
In two large cohort studies2
24
of risk factors for mediastinitis, autotransfusion of shed mediastinal blood was not examined. In three controlled trials17
21
25
in which patients were randomized to autotransfusion, or not, following cardiac surgery occurrence of mediastinitis was not different, but these studies were too small to detect uncommon events such as mediastinitis. This study utilized a case-control design, which is more powerful and efficient, given the infrequent occurrence of mediastinitis. Notwithstanding this design, the present study is still limited by the small number of patients with mediastinitis and control subjects. This limited the power to detect associations with many known risk factors, and particularly limited the ability to adjust for these risk factors in multivariate analysis. The approach taken, to use a summary score of the total number of other risk factors present, may have oversimplified the importance of these other risk factors, since each risk factor was given equivalent weight. Another model that allowed adjustment for two of the most important risk factors also showed a statistically significant association with prolonged autotransfusion. Matching for some of the known risk factors for mediastinitis could have increased our ability to evaluate the effect of autotransfusion, but as the study was designed to try to explain the cluster of cases, including whether the newly introduced practice was a factor, we did not wish to limit our ability to evaluate any possible contributors.
Following preliminary analysis of this study, the autotransfusion protocol at our hospital was modified in June 2001. Autotransfusion of shed mediastinal blood was discontinued if the drainage was < 50 mL/h for 2 consecutive h, with an absolute maximum duration of 6 h. Without any other significant interventions or modification to the perioperative and postoperative management of the patients, the incidence of mediastinitis between June 2001 and June 2002 decreased to 1.9%.
In conclusion, obesity, re-exploration within the first 24 h after surgery, antibiotic therapy in the 2 weeks prior to surgery, and allogenic transfusion of > 2 U packed RBCs were associated with mediastinitis. Duration and amount of autotransfusion of shed mediastinal blood were also associated with this serious complication, although the use of autotransfusion and total mediastinal drainage were not. Based on these findings, we suggest that autotransfusion of shed mediastinal blood should be limited to a maximum of 6 h.
|
Footnotes
|
|---|
Abbreviations: BMI = body mass index; CI = confidence interval; OR = odds ratio
Received for publication December 11, 2002.
Accepted for publication April 2, 2003.
|
References
|
|---|
- Milano, CA, Kesler, K, Archibald, N, et al (1995) Mediastinitis after coronary artery bypass graft surgery: risk factors and long-term survival. Circulation 92,2245-2251[Abstract/Free Full Text]
- Baskett, RJ, MacDougall, CE, Ross, DB Is mediastinitis a preventable complication? A 10-year review. Ann Thorac Surg 1999;67,462-465[Abstract/Free Full Text]
- Munoz, P, Menasalvas, A, Bernaldo deq Uiros, JC, et al Postsurgical mediastinitis: a case-control study. Clin Infect Dis 1997;25,1060-1064[ISI][Medline]
- Brown, IW, Jr, Moor, GF, Hummel, BW, et al Toward further reducing wound infections in cardiac operations. Ann Thorac Surg 1996;62,1783-1789[Abstract/Free Full Text]
- Bitkover, CY, Gardlund, B Mediastinitis after cardiovascular operations: a case-control study of risk factors. Ann Thorac Surg 1998;65,36-40[Abstract/Free Full Text]
- Stahle, E, Tammelin, A, Bergstrom, R, et al Sternal wound complications: incidence, microbiology and risk factors. Eur J Cardiothorac Surg 1997;11,1146-1153[Abstract]
- Grossi, EA, Esposito, R, Harris, LJ, et al Sternal wound infections and use of internal mammary artery grafts. J Thorac Cardiovasc Surg 1991;102,342-346[Abstract]
- Leal-Noval, SR, Rincon-Ferrari, MD, Garcia-Curiel, A, et al Transfusion of blood components and postoperative infection in patients undergoing cardiac surgery. Chest 2001;119,1461-1468[Abstract/Free Full Text]
- Schaff, HV, Hauer, J, Gardner, TJ, et al Routine use of autotransfusion following cardiac surgery: experience in 700 patients. Ann Thorac Surg 1979;27,493-499[Abstract]
- Schaff, HV, Hauer, JM, Bell, WR, et al Autotransfusion of shed mediastinal blood after cardiac surgery: a prospective study. J Thorac Cardiovasc Surg 1978;75,632-641[Abstract]
- Adan, A, Brutel de la Riviere, A, Haas, F, et al Autotransfusion of drained mediastinal blood after cardiac surgery: a reappraisal. Thorac Cardiovasc Surg 1988;36,10-14[ISI][Medline]
- Bouboulis, N, Kardara, M, Kesteven, PJ, et al Autotransfusion after coronary artery bypass surgery: is there any benefit? J Card Surg 1994;9,314-321[ISI][Medline]
- Body, SC, Birmingham, J, Parks, R, et al Safety and efficacy of shed mediastinal blood transfusion after cardiac surgery: a multicenter observational study; Multicenter Study of Perioperative Ischemia Research Group. J Cardiothorac Vasc Anesth 1999;13,410-416[CrossRef][ISI][Medline]
- Roberts, SR, Early, GL, Brown, B, et al Autotransfusion of unwashed mediastinal shed blood fails to decrease banked blood requirements in patients undergoing aortocoronary bypass surgery. Am J Surg 1991;162,477-480[CrossRef][ISI][Medline]
- de Haan, J, Schonberger, J, Haan, J, et al Tissue-type plasminogen activator and fibrin monomers synergistically cause platelet dysfunction during retransfusion of shed blood after cardiopulmonary bypass. J Thorac Cardiovasc Surg 1993;106,1017-1023[Abstract]
- de Haan, J, Boonstra, PW, Monnink, SH, et al Retransfusion of suctioned blood during cardiopulmonary bypass impairs hemostasis. Ann Thorac Surg 1995;59,901-907[Abstract/Free Full Text]
- Lepore, V, Radegran, K Autotransfusion of mediastinal blood in cardiac surgery. Scand J Thorac Cardiovasc Surg 1989;23,47-49[ISI][Medline]
- Thurer, RL, Lytle, BW, Cosgrove, DM, et al Autotransfusion following cardiac operations: a randomized, prospective study. Ann Thorac Surg 1979;27,500-507[Abstract]
- Vertrees, RA, Conti, VR, Lick, SD, et al Adverse effects of post-operative infusion of shed mediastinal blood. Ann Thorac Surg 1996;62,717-723[Abstract/Free Full Text]
- Weiss, G, Widner, B, Zoller, H, et al Immune response and iron metabolism. Br J Anaesth 1998;81(Suppl 1),6-9
- Andreasen, AS, Schmidt, H, Jarlov, JO, et al Autologous transfusion of shed mediastinal blood after coronary artery bypass grafting and bacterial contamination. Ann Thorac Surg 2001;72,1327-1330[Abstract/Free Full Text]
- Bland, LA, Villarino, ME, Arduino, MJ, et al Bacteriologic and endotoxin analysis of salvaged blood used in autologous transfusions during cardiac operations. J Thorac Cardiovasc Surg 1992;103,582-588[Abstract]
- Schwieger, IM, Gallagher, CJ, Finlayson, DC, et al Incidence of cell-saver contamination during cardiopulmonary bypass. Ann Thorac Surg 1989;48,51-53[Abstract]
- Gummert, JF, Barten, MJ, Hans, C, et al Mediastinitis and cardiac surgery: an updated risk factor analysis in 10,373 consecutive adult patients. Thorac Cardiovasc Surg 2002;50,87-91[CrossRef][ISI][Medline]
- Grinda, JM, Zegdi, R, Couetil, JP, et al Coronary reoperations: indications, techniques and operative results: retrospective study of 240 coronary reoperations. J Cardiovasc Surg (Torino) 2000;41,703-708[Medline]
TOP
Abstract
Introduction
