International Journal of Cardiology 157 (2012) 169–173
Contents lists available at ScienceDirect
International Journal of Cardiology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i j c a r d
Non-red blood cell transfusion as a risk factor for mortality following percutaneous coronary intervention Simon D. Robinson a,⁎, Christian Janssen a,b, Eric B. Fretz a, Alex J. Chase c, Anthony Della Siega a, Ronald G. Carere d, Anthony Fung e, Gerald Simkus f, J. David Hilton a, Brian Berry a,g, W. Peter Klinke a a
Victoria Heart Institute Foundation, Victoria BC, Canada University of Alberta, Edmonton AB, Canada Morriston Cardiac Centre, Swansea, Wales, UK d St Paul's Hospital, Vancouver BC, Canada e Vancouver General Hospital, Vancouver BC, Canada f Royal Columbian Hospital, New Westminster BC, Canada g Hematopathology, Royal Jubilee Hospital, Victoria BC, Canada b c
a r t i c l e
i n f o
Article history: Received 21 June 2010 Received in revised form 28 October 2010 Accepted 4 December 2010 Available online 7 January 2011 Keywords: Mortality Percutaneous coronary intervention Transfusion
a b s t r a c t Background: Bleeding following percutaneous coronary intervention (PCI) is common and may lead to transfusion and death. Although previous work has examined the effect of red blood cell (RBC) transfusion in patients with coronary disease, no study had investigated whether transfusion of non-RBC components was associated with mortality following PCI. Methods: All subjects transfused in the 10 days following PCI were identified using the British Columbia Cardiac and Central Transfusion Registries. Patients undergoing cardiac surgery following PCI were excluded as transfusion was assumed to be due to surgical related bleeding. Transfusion products were categorised as RBC and non-RBC comprising platelets, plasma and cryoprecipitate. Blood product use was compared according to thirty day mortality using multivariate regression and propensity adjustment for confounding variables. Results: From a total of 32,580 patients who underwent PCI, 952 patients received at least 1 blood product within 10 days of PCI. Non-RBC transfusion occurred more commonly in the cohort of transfused patients dying within 30 days (p b 0.001). After adjustment for baseline risk, transfusion of plasma/cryoprecipitate (HR 5.17; 95% C.I. 2.87–9.32, p b 0.001) and platelets (HR 2.13; 95% C.I. 1.10–4.13, p = 0.03) was associated with increased 30 day mortality. In a propensity risk adjusted model, transfusion of plasma/cryoprecipitate and RBC transfusion volume remained as significant predictors of 30-day mortality (p b 0.001). Conclusions: Transfusion following PCI appears to be associated with an increased risk of death within 30 days. We now report that transfusion with plasma rich non-RBC products may confer an additional mortality risk to patients undergoing PCI. © 2010 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Platelet activation, thrombus formation and fibrin deposition are central to the pathogenesis of coronary artery disease and are optimally treated with a combination of anti-platelet and anticoagulant therapy [1,2]. These therapies combined with percutaneous coronary intervention (PCI) improve outcomes for patients with CAD but have also led to increasing rates of treatment related major bleeding [3]. Furthermore, major bleeding following PCI has been linked to increased mortality [4,5] and correcting blood loss with transfusion may not reduce this risk [6]. ⁎ Corresponding author. Victoria Heart Institute Foundation, 200-1900 Richmond Avenue, Victoria, BC, Canada, V8R 4R2. Tel.: +1 250 595 1884; fax: +1 250 595 5367. E-mail address:
[email protected] (S.D. Robinson). 0167-5273/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijcard.2010.12.013
To date, most studies of transfusion in patients with cardiac disease have evaluated the effect of red blood cell (RBC) transfusion. Red blood cell transfusion has been associated with increased mortality following PCI [4,7] and coronary artery bypass grafting (CABG) [8]. Much less is known about the potential risks to patients with cardiac disease from transfusion of non-RBC products such as platelets and plasma derivatives. As key factors in the maintenance of vascular hemostasis and regulation of thrombus formation, transfusion of these products might influence the risk of arterial occlusion, myocardial infarction and death. Transfusion of non-RBC products has been shown to be harmful to critically ill patients [9] though controversy exists on whether they affect outcome following cardiac surgery [10–12]. To date, no study has investigated whether non-RBC transfusion influences mortality following PCI. We therefore studied the effect of individual blood
170
S.D. Robinson et al. / International Journal of Cardiology 157 (2012) 169–173
components in patients transfused following PCI and hypothesized that use of platelets and plasma products would be associated with an increased risk of death within 30 days.
Table 1 Baseline characteristics of patients according to transfusion product received. Transfusion product received RBC only n = 800 (84%)
Non-RBC n = 152 (16%)
Age BMI, kg/m2 Male Current smoker
70.9 ± 11.1 26.8 ± 5.7 51.1 17.4
67.7 ± 11.6 25.9 ± 5.0 64.6 21.5
Past medical history Hypertension PVD CVD Diabetes mellitus Prior malignancy Dialysis Prior MI/CHF Prior PCI/CABG Chronic lung disease GI/Liver disease Creatinine, μmol/L Mean RBC transfusion volume, units
67.6 19.7 12.3 36.2 11.1 8.4 46.8 35.0 17.7 13.1 156 ± 141 2.7 ± 1.8
59.4 14.0 13.0 32.1 8.8 8.6 38.2 29.6 9.6 10.3 147 ± 104 4.8 ± 3.7†
2. Materials and methods We identified subjects receiving a transfusion following PCI using an electronic linkage between the British Columbia Cardiac (BCCR) and BC Central Transfusion Registries (CTR) [7]. The CTR has collected data on all transfusions since 1999 within the province of BC, Canada. Mortality was determined from the Ministry of Health Vital Statistics database and this was cross-referenced with the BCCR and CTR. The study was approved by the research and ethics committee of the hospitals involved and the University of British Columbia. All patients undergoing PCI in British Columbia between 1st January 1999 and 31st December 2005 were included in this study. Thereafter we identified those who were transfused with at least one blood product within 10 days of PCI. Patients undergoing cardiac surgery were excluded as transfusion events were assumed to be due to surgical bleeding. The BC cardiac registry data fields are compiled prospectively by catheter laboratory staff at the time of PCI and held by a dedicated data management team at the Provincial Health Services Authority. Procedures undertaken via brachial access accounted for 0.1% of the entire cohort and were excluded from multivariate analysis. Transfusion components were classified as RBC or non-RBC products which comprised platelets, plasma and cryoprecipitate. To investigate the effect of restoring levels of plasma coagulation factors we grouped plasma and cryoprecipitate transfusions together (Plas/Cryo) within the outcome model for 30 day mortality. Platelet and plasma concentrates included those obtained from centrifugation of whole blood and those from apheretic donation. All blood products were leuko-reduced at source prior to separation into component parts.
2.1. Statistical analysis Means of continuous variables were compared using the independent samples t-test (with equal or unequal variances determined with Levine's test). Pearson Chi-square test was used to analyze frequencies between groups. A Cox proportional hazard regression model incorporating baseline demographics, co-morbid conditions, procedural data and transfusion variables was used to estimate hazard ratios (together with 95% confidence intervals) for 30-day mortality. Transfusion variables were platelets, Plas/Cryo and RBC transfusion volume. Explanatory variables were entered stepwise with forward selection (complemented with backward elimination) specifying p-in = 0.05 and p-out = 0.10 under the criterion of Maximum Likelihood. We undertook further analysis using propensity scoring [13,14] in an attempt to reduce the effect of confounding between transfusions and mortality. Binary logistic regression was used to estimate the propensity for transfusion based upon baseline demographics, co-existing illnesses and PCI data (treatment model). The estimated conditional probability was then used with the transfusion variables (platelets, Plas/ Cryo, and RBC volume) in a propensity adjusted Cox proportional hazards model (outcome model). Statistical calculations and analyses were performed using SPSS 15 software (SPSS Inc., Chicago, Illinois). A two-sided P value of less than 0.05 was considered to indicate statistical significance.
3. Results A total of 32,580 patients undergoing PCI during the study period were identified. Of these 952 patients were transfused with at least 1 blood product; 800 patients received only RBCs whilst 152 patients received non-RBC products with (n = 109) or without RBCs (n = 43). Baseline characteristics in transfused patients receiving RBC and nonRBC blood products are shown in Table 1. Those receiving non-RBC products were on average younger, were more likely to be male and had a lower burden of chronic lung disease. Mean RBC transfusion volumes were higher in the patients who also received platelets or Plas/Cryo compared to those transfused with RBC alone (4.8 ± 3.7 vs 2.7 ± 1.8 units, p b 0.001). 3.1. Individual transfusion components and 30 day mortality Mortality was significantly higher in patients receiving non-RBC products with or without additional transfusion of RBCs (Fig. 1). A total of 3979 units of blood products were transfused to the 952 patients. Platelet, plasma and cryoprecipitate transfusions were more frequently administered to the cohort of transfused patients dying within 30 days (p b 0.001, Table 2).
P
0.002⁎ 0.09 0.003⁎ 0.56
0.07 0.12 0.80 0.35 0.43 0.94 0.05 0.20 0.02⁎ 0.36 0.52 b0.001⁎
Values are % or mean ± SD. P = Pearson Chi-square tests for frequencies and independent samples t-test of means between groups. BMI—body mass index, PVD—peripheral vascular disease, CVD—cerebrovascular disease, MI—myocardial infarction, CHF—congestive heart failure, PCI—percutaneous coronary intervention, CABG—coronary artery bypass graft, GI—gastrointestinal. Non-RBC products = plasma, platelets or cryoprecipitate. ⁎ P b 0.05. † Excluding the 43 patients who did not receive red cells.
3.2. Procedural complexity and transfusion details There appeared to be no difference in the procedural indication or vascular access site used between patients transfused with non-RBC products compared to those receiving only RBCs (Table 3). Fluoroscopy time (for procedural duration), contrast volume and total stent length were also similar in both groups. 3.3. Estimated hazard ratios for 30-day mortality A Cox proportional hazards model for 30-day mortality was created to adjust for the effect of subject demographics, procedural details and individual transfusion components. Multivariate adjusted hazard ratios for the entire cohort are shown in Table 4. Platelet (Hazard ratio 2.13; 95% confidence interval 1.10–4.13, p = 0.03) and Plas/Cryo transfusion (HR 5.17; 95% C.I. 2.87–9.32, p b 0.001) were associated with mortality in a Cox proportional hazards model for mortality. However after adjustment for the effect of potential confounders on the propensity for transfusion, use of Plas/Cryo but not platelets was a significant predictor of 30-day mortality (Table 5). Larger RBC transfusion volumes were associated with a greater risk of death in the propensity adjusted risk model (Fig. 2). 4. Discussion We have found that transfusion following PCI is associated with increased 30-day mortality. We have extended this observation and now report that use of non-RBC products may confer an excess risk of death to patients transfused following PCI. As far as we know, our study is the first to report on clinical outcome with receipt of specific blood products following PCI. Within this large cohort of patients undergoing PCI, mortality was two-fold greater in patients transfused with platelets, plasma or cryoprecipitate whilst patients who also received red cells had almost a four-fold higher rate of death (Fig. 1). The association between mortality and use of Plas/Cryo persisted after
S.D. Robinson et al. / International Journal of Cardiology 157 (2012) 169–173
171
Fig. 1. Summary of transfusion products and mortality. Patients receiving red blood cells (RBCs) alone; patients receiving non-RBCs alone and patients transfused with both RBC and non-RBC products. P b 0.001 for 30-day mortality in RBC only group vs RBC + non-RBC group. P = 0.04 for 30-day mortality in RBC only group vs non-RBC only group.
propensity risk adjusted multivariate analysis which included information on concomitant RBC transfusion volume. To date, studies examining the effects of transfusion in patients undergoing PCI have largely focussed on the use of RBCs with transfusion appearing to predict increased mortality following PCI [4,7,15]. Furthermore despite correcting anemia, larger transfusion volumes may be associated with greater risk than adopting a restrictive approach to transfusion [16]. Our results are in agreement with these earlier observations suggesting that higher RBC transfusion volumes appear to confer a greater risk of death. Disruption of the atheromatous plaque within coronary vessels following balloon angioplasty or stent deployment during PCI is a potent stimulus for thrombus formation and platelet activation. Clinical trials have shown reduced ischemic events and death with greater inhibition of platelet function [17] and thrombin activity [18]. Correspondingly, increased residual platelet function despite dual anti-platelet therapy [19] and higher plasma levels of thrombotic factors [20] are associated with a greater risk of adverse events following PCI. We had therefore questioned whether patients transfused with plasma rich non-RBC products following PCI might be at increased risk perhaps due to the development of a prothrombotic state. From a cohort of over 32,000 patients undergoing PCI and almost 1000 transfused patients, we have found that transfusion with nonRBC products is associated with a greater risk of death. To the best of our knowledge, our study is the first to link mortality to the use of specific blood products in patients undergoing PCI. Within our study,
Table 2 Frequency of individual transfusion products according to mortality at 30 days.
Red blood cells Plasma (Plas) Platelets Cryoprecipitate (Cryo) Total
% (n) of transfusion patients alive 30 days
% (n) of transfusion patients dead 30 days
72.9 8.7 16.6 1.8 100
44.6 (403) 23.6 (2130) 26.0 (235) 5.8 (52) 100 (903)
(2241) (268) (511) (56) (3076)
909 patients received RBCs, 96 received Plasma, 91 received Platelets and 12 received Cryoprecipitate. 831 transfused patients survived to 30 days; 121 transfused patients died within 30 days. P b 0.001 for transfusion product use between groups according to 30-day mortality.
patients receiving non-RBC products were on average younger but did have higher mean RBC transfusion volumes. In a Cox proportional hazards model which included baseline demographics and PCI procedural variables, transfusion of platelets and Plas/Cryo was associated with the risk of death. Using propensity techniques to adjust for the a priori risk of transfusion, RBC transfusion volume, use of Plas/Cryo but not platelets remained as a significant predictor of 30 day mortality. We combined plasma and cryoprecipitate transfusions in our analysis as relatively few patients received cryoprecipitate (only 1 patient received cryoprecipitate alone) and we wished to assess the effect of restoring plasma coagulation factors. Furthermore, despite statistical modelling to estimate the effect of receiving platelet or plasma transfusion, patients were not randomly allocated to the receipt of a specific blood product and platelet concentrates also contain plasma. Until further studies are completed, physicians should consider the possibility that transfusion with any of these non-RBC products may have adverse consequences in patients undergoing PCI. Transfusion is recognised as potentially hazardous due to the risk of transmissible infection, acute allergic and haemolytic reactions Table 3 Procedural details and transfusion products. Transfusion product
P
RBC only n = 800
Non-RBC n = 152
Procedural indication Elective, % Urgent, % Emergency, %
9.0 62.7 28.2
11.3 52.3 36.4
0.06
Vascular access site Radial, % Femoral, %
11.7 87.8
11.2 87.5
0.51
Procedural data Bare metal stent, % Drug eluting stent, % Total stent length, mm Contrast volume, mL Fluoroscopy time, min
79.4 8.9 32 ± 25 240 ± 120 18.3 ± 13.8
80.3 5.3 30 ± 27 249 ± 136 19.5 ± 15.0
0.82 0.51 0.45 0.41 0.34
Values are % or mean ± SD. Non-RBC = platelets, plasma or cryoprecipitate. P = RBC only vs non-RBC products (Pearson Chi-square tests for frequencies and independent samples t-test of means).
172
S.D. Robinson et al. / International Journal of Cardiology 157 (2012) 169–173
Table 4 Cox regression estimated hazard ratios for 30-day mortality.
Age N70 years BMI b 18.5 kg/m2 Diabetes mellitus CVD Prior pulmonary disease Prior MI/CHF Prior malignancy Creatinine N 90 μmol/L Stable angina Urgent PCI Emergent PCI RBC transfusion 1–2 units RBC N2 units RBC Platelet transfusion Plas/Cryo transfusion Fluoroscopy time, min
Hazard ratio
95% confidence interval
P
2.59 2.57 1.48 1.77 1.55 1.37 0.57 1.33 0.28 1.95 10.28
2.07–3.23 1.57–4.21 1.17–1.87 1.31–2.39 1.15–2.07 1.10–1.70 0.35–0.94 1.07–1.65 0.15–0.53 1.05–3.63 5.43–19.48
b 0.001 b 0.001 0.001 b 0.001 0.004 0.006 0.03 0.01 b 0.001 0.04 b 0.001
1.75–3.77 1.92–4.86 1.10–4.13 2.87–9.32 1.0–1.004
b 0.001 b 0.001 0.03 b 0.001 0.02
2.57 3.06 2.13 5.17 1.002
Hazard ratios (HR) as predictors of 30 day mortality after adjustment for baseline variables. Variables permitted in the model: age N 70 years, body mass index (BMI) b18.5 kg/m2, male gender, smoking history, history of stable angina, acute coronary syndrome (ACS), hypertension, peripheral vascular disease (PVD), cerebrovascular disease (CVD), diabetes mellitus, prior malignancy, prior myocardial infarction/congestive heart failure (MI/CHF), documented chronic lung disease, documented gastrointestinal/ liver (GI/liver) disease, previous percutaneous coronary intervention/coronary artery bypass grafting (PCI/CABG), arterial access site, creatinine N90 mmol/L, indication for procedure (elective, urgent or emergency), red blood cell transfusion volume [0, 1–2 or N 2 RBC units], platelet transfusion or plasma/cryoprecipitate (Plas/Cryo) transfusion, fluoroscopy time.
though with modern transfusion practices these occur infrequently [21]. Data has accumulated on other adverse effects of transfusing plasma products and platelets including transfusion related acute lung injury (TRALI) [22] as well as an increased risk of multi-organ failure and sepsis [9]. We now report that transfusion of non-RBC products appears to increase the risk of death following PCI. Accumulation of pro-inflammatory products in stored blood products has been implicated in the pathogenesis of TRALI, a condition which is frequently fatal [22]. Although our study cannot provide a mechanism linking non-RBC transfusion and mortality post-PCI, transfusion related activation of inflammatory cytokines might contribute to excess mortality as levels of inflammatory cytokines predict death following PCI [23]. We have found that transfusion of plasma rich products was related to 30-day mortality after adjustment for concomitant RBC transfusion and baseline risk. In contrast to earlier work [10], both Karkouti [11] and more recently McGrath and colleagues [12], have suggested that platelet transfusions do not Table 5 Propensity adjusted outcome model for 30-day mortality.
RBC transfusion 1–2 units RBC transfusion N2 units Plas/Cryo transfusion Propensity to transfusiona
Hazard ratio
95% confidence interval
P
4.49 6.33 3.92 1.007
3.21–6.28 4.37–9.18 2.52–6.11 1.006–1.007
b 0.001 b 0.001 b 0.001 b 0.001
The propensity score for transfusion was estimated using a binary logistic treatment model and included acute coronary syndrome (odds ratio 0.48 [95% confidence interval 0.35–0.66], p b 0.001): stable angina (0.34 [0.23–0.50], p b 0.001), peripheral vascular disease (1.75 [1.40–2.19], p b 0.001), diabetes mellitus (1.71 [1.44–2.04], p b 0.001), urgent procedure (2.50 [1.75–3.58], p b 0.001), emergency procedure (9.01 [6.04–13.45], p b 0.001), female gender (2.12 [1.79–2.53], p b 0.001), prior pulmonary disease (1.65 [1.32–2.07], p b 0.001), prior gastrointestinal/liver disease (1.68 [1.29–2.20], p b 0.001), prior myocardial infarction/congestive heart failure (1.25 [1.06–1.49], p = 0.01), fluoroscopy time (1.007 [1.004–1.011], p b 0.001), femoral access (1.86 [1.45–2.39], p b 0.001), age N 70 years (2.07 [1.74–2.46], p b 0.001), creatinine N 90 mmol/L (1.76 [1.48–2.10], p b 0.001). a Magnitude adjusted conditional probability for transfusion. Transfusion of platelets within the propensity adjusted model had a p value of 0.32.
Fig. 2. Survival function from propensity adjusted Cox regression model. Survival to 30 days according to (a) RBC transfusion volume and (b) receipt of plasma/ cryoprecipitate (yes/no).
influence outcome following cardiac surgery after adjustment for confounders including RBC transfusion. Our analysis has also not investigated the importance of transfusing ABO-compatible but nonidentical platelet products which has been linked to increased mortality following cardiac surgery [24]. Transfusion guidelines suggest that plasma and cryoprecipitate transfusion should be considered in patients with active bleeding and low levels of clotting factors [25,26]. Patients with elevated levels of fibrinogen and clotting factors are however at increased risk of future myocardial infarction and sudden death [27]. Restoring plasma levels of clotting factors with transfusion of plasma or cryoprecipitate might therefore promote the development of a pro-thrombotic environment and risk arterial thrombosis. The likelihood of coronary occlusion and/ or stent thrombosis may be further exacerbated by incomplete endothelial coverage of stent struts [28] within the 10 day time period used to define procedural related transfusions in this study. 4.1. Limitations Despite published guidelines on the use of individual blood products, surveys suggest wide variations in their use [29]. To date, such guidelines have not addressed the use of blood in patients undergoing coronary revascularization who may be at particular risk following transfusion. Our results represent a retrospective study of
S.D. Robinson et al. / International Journal of Cardiology 157 (2012) 169–173
the individual blood products received by patients who required transfusion following PCI. Blood products were chosen according to physician preference and although patients receiving non-RBC products appeared broadly similar to those transfused with only red cells, unmeasured factors including the use of anti-platelet and antithrombotic pharmacotherapy may have influenced our findings. Despite using statistical techniques to control for measured variables known to be associated with adverse outcomes following PCI, we acknowledge that use of non-RBC products may still be a surrogate marker for sicker patients at higher risk of death. This study does not include information on the indication for selection of blood products such as anemia severity or peri-procedural bleeding. Though transfusions do not cause such complications these may have directly contributed to the demise of some of the transfused patients. Finally, although we have suggested that non-RBC products may be potentially harmful to patients undergoing PCI, these may improve survival in other clinical situations [30]. 5. Conclusion We have found that transfusion with blood appears hazardous to patients undergoing PCI and that those patients also receiving plasma rich non-RBC products may be at greater risk of death within 30 days. Given the absence of randomized studies on the risks/benefits of using specific blood products, our data emphasize the importance of therapeutic strategies which minimise bleeding events and the likelihood of needing transfusion following PCI. Acknowledgments We thank Cheryl Lewis and Karin Humphries and their staff at the BC Provincial Blood Coordinating Office and BC Cardiac Registry for their assistance with data extraction. We are grateful to the Victoria Foundation (registered Canadian charity 130650898RR0001) for its grant support of this research. The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology [31]. References [1] Cannon CP, Weintraub WS, Demopoulos LA, et al. Comparison of early invasive and conservative strategies in patients with unstable coronary syndromes treated with the glycoprotein IIb/IIIa inhibitor tirofiban. N Engl J Med 2001;344:1879–87. [2] Wallentin L, Lagerqvist B, Husted S, Kontny F, Stahle E, Swahn E. Outcome at 1 year after an invasive compared with a non-invasive strategy in unstable coronaryartery disease: the FRISC II invasive randomised trial. FRISC II Investigators. Fast Revascularisation during Instability in Coronary artery disease. Lancet 2000;356:9–16. [3] SYNERGY Trial Investigators. Enoxaparin vs unfractionated heparin in high-risk patients with non-ST-segment elevation acute coronary syndromes managed with an intended early invasive strategy: primary results of the SYNERGY randomized trial. JAMA 2004;292:45–54. [4] Kinnaird TD, Stabile E, Mintz GS, et al. Incidence, predictors, and prognostic implications of bleeding and blood transfusion following percutaneous coronary interventions. Am J Cardiol 2003;92:930–5. [5] Ndrepepa G, Berger PB, Mehilli J, et al. Periprocedural bleeding and 1-year outcome after percutaneous coronary interventions: appropriateness of including bleeding as a component of a quadruple end point. J Am Coll Cardiol 2008;51: 690–7.
173
[6] Maluenda G, Lemesle G, Ben-Dor I, et al. Value of blood transfusion in patients with a blood hematocrit of 24% to 30% after percutaneous coronary intervention. Am J Cardiol 2009;104:1069–73. [7] Chase AJ, Fretz EB, Warburton WP, et al. Association of the arterial access site at angioplasty with transfusion and mortality: the M.O.R.T.A.L study (Mortality benefit Of Reduced Transfusion after percutaneous coronary intervention via the Arm or Leg). Heart 2008;94:1019–25. [8] Koch CG, Li L, Sessler DI, et al. Duration of red-cell storage and complications after cardiac surgery. N Engl J Med 2008;358:1229–39. [9] Khan H, Belsher J, Yilmaz M, et al. Fresh-frozen plasma and platelet transfusions are associated with development of acute lung injury in critically ill medical patients. Chest 2007;131:1308–14. [10] Spiess BD, Royston D, Levy JH, et al. Platelet transfusions during coronary artery bypass graft surgery are associated with serious adverse outcomes. Transfusion 2004;44:1143–8. [11] Karkouti K, Wijeysundera DN, Yau TM, et al. Platelet transfusions are not associated with increased morbidity or mortality in cardiac surgery. Can J Anaesth 2006;53:279–87. [12] McGrath T, Koch CG, Xu M, et al. Platelet transfusion in cardiac surgery does not confer increased risk for adverse morbid outcomes. Ann Thorac Surg 2008;86: 543–53. [13] D'Agostino RB. Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med 1998;17:2265–81. [14] Blackstone EH. Comparing apples and oranges. J Thorac Cardiovasc Surg 2002;123:8–15. [15] Rao SV, Jollis JG, Harrington RA, et al. Relationship of blood transfusion and clinical outcomes in patients with acute coronary syndromes. JAMA 2004;292:1555–62. [16] Hebert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion Requirements in Critical Care Investigators, Canadian Critical Care Trials Group. N Engl J Med 1999;340:409–17. [17] Karvouni E, Katritsis DG, Ioannidis JPA. Intravenous glycoprotein IIb/IIIa receptor antagonists reduce mortality after percutaneous coronary interventions. J Am Coll Cardiol 2003;41:26–32. [18] Stone GW, Witzenbichler B, Guagliumi G, et al. Bivalirudin during primary PCI in acute myocardial infarction. N Engl J Med 2008;358:2218–30. [19] Gori AM, Marcucci R, Migliorini A, et al. Incidence and clinical impact of dual nonresponsiveness to aspirin and clopidogrel in patients with drug-eluting stents. J Am Coll Cardiol 2008;52:734–9. [20] Marcucci R, Brogi D, Sofi F, et al. PAI-1 and homocysteine, but not lipoprotein (a) and thrombophilic polymorphisms, are independently associated with the occurrence of major adverse cardiac events after successful coronary stenting. Heart 2006;92:377–81. [21] Norda R, Tynell E, Akerblom O. Cumulative risks of early fresh frozen plasma, cryoprecipitate and platelet transfusion in Europe. J Trauma 2006;60:S41–5. [22] Silliman CC, Boshkov LK, Mehdizadehkashi Z, et al. Transfusion-related acute lung injury: epidemiology and a prospective analysis of etiologic factors. Blood 2003;101:454–62. [23] Ridker PM, Rifai N, Pfeffer M, Sacks F, Lepage S, Braunwald E. Elevation of tumor necrosis factor-alpha and increased risk of recurrent coronary events after myocardial infarction. Circulation 2000;101:2149–53. [24] Blumberg N, Heal J, Hicks G, Risher W. Association of ABO-mismatched platelet transfusions with morbidity and mortality in cardiac surgery. Transfusion 2001;41:790–3. [25] O'Shaughnessy DF, Atterbury C, Maggs PB, et al. Guidelines for the use of freshfrozen plasma, cryoprecipitate and cryosupernatant. Br J Haematol 2004;126: 11–28. [26] Practice parameter for the use of fresh-frozen plasma, cryoprecipitate, and platelets. Fresh-frozen plasma, cryoprecipitate, and platelets administration practice guidelines development Task Force of the College of American Pathologists. JAMA 1994;271:777–81. [27] Meade TW, Ruddock V, Stirling Y, Chakrabarti R, Miller GJ. Fibrinolytic activity, clotting factors, and long-term incidence of ischaemic heart disease in the Northwick Park Heart Study. Lancet 1993;342:1076–9. [28] Virmani R, Kolodgie FD, Farb A, Lafont A. Drug eluting stents: are human and animal studies comparable? Heart 2003;89:133–8. [29] Metz J, McGrath KM, Copperchini ML, et al. Appropriateness of transfusions of red cells, platelets and fresh frozen plasma. An audit in a tertiary care teaching hospital. Med J Aust 1995;162(572–573):576–7. [30] Holcomb JB, Wade CE, Michalek JE, et al. Increased plasma and platelet to red blood cell ratios improves outcome in 466 massively transfused civilian trauma patients. Ann Surg 2008;248:447–58. [31] Shewan LG, Coats AJ. Ethics in the authorship and publishing of scientific articles. Int J Cardiol 2010;144:1–2.