The decision to discontinue aspirin preoperatively in

Transcription

1 Effect of Preoperative Aspirin Use on Mortality in Coronary Artery Bypass Grafting Lawrence J. Dacey, MD, John J. Munoz, MD, Edward R. Johnson, MD, Bruce J. Leavitt, MD, Christopher T. Maloney, MD, Jeremy R. Morton, MD, Elaine M. Olmstead, BA, John D. Birkmeyer, MD, and Gerald T. O Connor, PhD, DSc, for the Northern New England Cardiovascular Disease Study Group Department of Surgery, Center for the Evaluative Clinical Sciences, Community & Family Medicine, and Department of Medicine, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire, Department of Surgery, Eastern Maine Medical Center, Bangor, Maine, Department of Surgery, Fletcher-Allen Health Care, Burlington, Vermont, Department of Surgery, Catholic Medical Center, Manchester, New Hampshire, and Department of Surgery, Maine Medical Center, Portland, Maine Background. Discontinuing aspirin use in patients before coronary artery bypass grafting (CABG) has focused on bleeding risks. The effect of aspirin use on overall mortality with this procedure has not been studied. Methods. We performed a case patient control patient study of the 8,641 consecutive isolated CABG procedures performed between July 1987 and May 1991 in Maine, New Hampshire, and Vermont. included all 368 deaths. Each case patient was paired with approximately two matched survivors (control patients). Aspirin use was defined by identification of ingestion within 7 days before the operation. Results. CABG patients using preoperative aspirin were less likely to experience in-hospital mortality in univariate (odds ratio [OR] 0.73, 95% confidence interval [0.54, 0.97]) and multivariate [OR 0.55, (0.31, 0.98)] analysis compared to nonusers. No significant difference was seen in the amount of chest tube drainage, transfusion of blood products, or need for reexploration for hemorrhage between patients who did and did not receive aspirin. Conclusions. Preoperative aspirin use appears to be associated with a decreased risk of mortality in CABG patients without significant increase in hemorrhage, blood product requirements, or related morbidities. (Ann Thorac Surg 2000;70: ) 2000 by The Society of Thoracic Surgeons The decision to discontinue aspirin preoperatively in patients undergoing coronary artery bypass grafting (CABG) is controversial. Many studies report that patients receiving aspirin before CABG operations have increased blood loss, increased transfusion requirements, and a higher incidence of reoperation for hemorrhage [1 3]. However, the use of aspirin has many potential benefits. Aspirin therapy has proved beneficial in ischemic heart disease [4 5]. In coronary artery bypass operations, multiple studies have shown increased early and late vein graft patency in patients who receive aspirin therapy in the perioperative period [6 8]. Although the benefit of perioperative aspirin use has been demonstrated clearly on graft patency, the net effect on perioperative mortality has not been defined. We therefore performed a study to examine the use of preoperative aspirin in CABG patients with regard to in-hospital mortality and postoperative hemorrhagic complications. We studied 368 in-hospital deaths occurring in 8,641 consecutive patients undergoing CABG in northern New England between 1987 and Using case patient control patient methodology, patients who died were compared to survivors to assess the factors associated with mortality. Accepted for publication May 23, Address reprint requests to Dr Dacey, Cardiothoracic Surgery, Dartmouth-Hitchcock Medical Center, Lebanon, NH ; lawrence.j.dacey@dartmouth.edu. Material and Methods Setting The Northern New England Cardiovascular Disease Study Group is a voluntary research consortium representing all five medical centers in Maine, New Hampshire and Vermont where CABG operations are performed. Since 1987, the Northern New England Cardiovascular Disease Study Group has maintained a prospective registry of all patients undergoing cardiac operations in the region. Study Design This was a multicenter case patients and control patients study of in-hospital survivors and nonsurvivors of CABG between July 1987 and May 1991 in northern New England. who had isolated CABG operations and died during hospital admission were referred to as case patients. selected from the survivors of isolated CABG operations and matched to the case patients (approximately 2:1) by age (within 5 years), sex, priority at time of operation (elective or emergent), and medical center served as control patients. Risk Factors From the prospective registry substantial patient data were available, including the following: patient age and sex, body surface area, degree of left main coronary artery stenosis, total number of significantly diseased vessels, ejection fraction, prior CABG, priority of the 2000 by The Society of Thoracic Surgeons /00/$20.00 Published by Elsevier Science Inc PII S (00)

2 Ann Thorac Surg DACEY ET AL 2000;70: PREOPERATIVE ASPIRIN IN CABG PATIENTS 1987 operation (elective or emergent) and mortality status at discharge. The ejection fraction results were scored using the method described by Pierpont and colleagues [9]. CABG was defined as emergent when medical factors relating to the patient s cardiac disease dictated that the operation was to be performed within hours to prevent morbidity or death. Urgent CABG was defined by the presence of clinical factors requiring that the patient stay in the hospital to have the operation before discharge. The remaining cases were defined as elective. Additional data were obtained by medical record review of preoperative testing and treatment, operative techniques, and methods of postoperative care. Aspirin therapy was assessed by review of medications listed in the admission history and physical exam, preoperative notes, nursing notes, and medication administration record. If any of these noted aspirin use within 7 days of the operation, the patient was considered to be exposed. Mortality and Hemorrhage-Related Outcomes Preoperative aspirin use was evaluated among the case patients and control patients to assess the relation between its use and mortality. To assess the effects of aspirin on postoperative bleeding and its associated adverse outcomes, we compared chest tube drainage, transfusion requirements of red blood cells, platelets, plasma, and cryoprecipitate, and rates of reexploration between case patients and control patients using preoperative aspirin and those using no aspirin. Analysis Case patients were compared to control patients with respect to preoperative aspirin exposure. The exposure odds ratio for the matched pairs and the McNemar 2 test statistic were computed. To adjust for confounding variables we conducted a multivariate analysis using conditional logistic regression techniques. Variables that differed significantly between groups, as well as common known risk factors, were selected for the multivariate analysis. A comparison of means was performed using t tests. All statistical analyses were conducted using the Stata statistical software program (version 5.0, Stata Corp, College Station, TX). Results Patient Characteristics Baseline patient characteristics of the case patients and the matched control patients were evaluated. Case patients and control patients were similar with respect to history of chronic obstructive pulmonary disease, smoking, prior CABG operations, left ventricular end diastolic pressure, and the variables matched on (age, sex, and priority). However, patients who had hypertension, peripheral vascular disease, a higher preoperative creatinine, history of diabetes, higher Charlson comorbidity scores, prior congestive heart failure, or lower ejection fraction were more likely to be in the case patient group (Table 1). There were 368 deaths identified in the study time Table 1. Characteristics of Case and Control Variables Case (Dead) (n 368) Control (Alive) (n 688) p Value Patient demographics Age (mean, years) Gender (% male) BSA (mean, m 2 /kg) Preoperative characteristics HTN (% yes) COPD (% yes) PVD (% yes) Creatinine (mean, mg/dl) Bleeding time (mean, min) Diabetes (% yes) Insulin Oral agents Tobacco history (% yes) Current Previous Charlson comorbidity scale (%) Cardiac profile Surgical priority (%) Emergent Elective Prior CHF (% yes) Ejection fraction (mean, %) LVEDP (mean, mm Hg) Prior CABG (% yes) aspirin use; BSA body surface area; CABG coronary artery bypass grafting; CHF congestive heart failure; COPD chronic obstructive pulmonary disease; HTN hypertension; LVEDP left ventricular end diastolic pressure; PVD peripheral vasuclar disease. period, from which 114 patients (31%) were using preoperative aspirin. From the 688 control patients, 254 patients (37%) were using preoperative aspirin. From the case patient and control patient groups, there were no significant differences in patients using aspirin compared to those not using aspirin, with the following exceptions (Table 2): There was a higher percentage of patients using aspirin in the urgent/emergent priority category in both case patients (p 0.001) and control patients (p 0.001). Also, a greater percentage of patients using aspirin in the case patient group but not in the control patient group were male (p 0.018). There was a lower mean preoperative creatinine among patients using aspirin in the control patient group (1.12 versus 1.21, p 0.006), and

3 1988 DACEY ET AL Ann Thorac Surg PREOPERATIVE ASPIRIN IN CABG PATIENTS 2000;70: Table 2. Characteristics of Case and Control Stratified by Aspirin Use Variables 368 Case (Dead) 688 Control (Alive) (n 114) No No (n 434) Patient demographics Age (mean, years) Gender (% male) 66 a BSA (mean, m 2 /kg) Preoperative characteristics HTN (% yes) COPD (% yes) PVD (% yes) Creatinine (mean, mg/dl) a Bleeding time (mean, min) a a Diabetes (% yes) Insulin Oral agents Tobacco history (% yes) Current Previous Charlson comorbidity scale (%) Cardiac profile Surgical priority (%) Emergent 88 a a 67 Elective 12 a a 33 Prior CHF (% yes) Ejection fraction (mean, %) a LVEDP (mean, mm Hg) Prior CABG (% yes) a Difference between aspirin users vs. nonusers within case and control groups is statistically significant (p 0.05). aspirin use; BSA body surface area; CABG coronary artery bypass grafting; CHF congestive heart failure; COPD chronic obstructive pulmonary disease; HTN hypertension; LVEDP left ventricular end diastolic pressure; PVD peripheral vasuclar disease. there was a higher mean ejection fraction in the patients using aspirin in the control patient group (56% versus 53%, p 0.007). Finally, there were longer bleeding times in patients using aspirin in both the case patient group (p 0.05) and the control patient group (p ). Mortality Preoperative aspirin users were 27% less likely to experience a fatal outcome than nonusers by univariate analysis (OR 0.73, 95% confidence interval [0.54, 0.97]; p 0.03) (Table 3). A stratified analysis incorporating preoperative risk factors including sex, surgical priority, diabetes, and ejection fraction was also performed. In the stratified analysis the protective effect of aspirin appeared more pronounced in women (OR 0.31, 95% confidence interval [0.11, 0.87], p 0.027) than in men (OR 0.83, 95% confidence interval [0.39, 1.72], p 0.620). However, further analysis showed no significant interaction of sex and the protective effect of aspirin. We also performed multivariate analysis adjusting for variables that differed significantly between case patients and control patients. These variables were body surface area, hypertension, history of peripheral vascular disease, preoperative creatinine, diabetes, history of congestive heart failure, and ejection fraction. Results revealed a 45% reduction in fatal outcomes with preoperative aspirin users versus non-aspirin users (OR 0.55, [0.31,0.98]; p 0.04). Hemorrhage and Transfusion There was little difference in complications related to hemorrhage between patients using aspirin and those Table 3. Association of Preoperative Aspirin Use and In-Hospital Mortality After CABG Mode of Analysis Odds Ratio 95% Confidence Interval p Value Univariate analysis 0.73 (0.54, 0.97) 0.03 Multivariate analysis a 0.55 (0.31, 0.98) 0.04 a Adjusted for body surface area, hypertension, history of peripheral vascular disease, preoperative creatinine, diabetes, history of congestive heart failure, and ejection fraction.

4 Ann Thorac Surg DACEY ET AL 2000;70: PREOPERATIVE ASPIRIN IN CABG PATIENTS 1989 Table 4. Effects of Aspirin on Blood Loss, Transfusion, and Reexploration Among Case and Control Outcome (n 114) Case (Dead) No p Value Control (Alive) No (n 434) p Value Chest tube drainage (ml) 1,318 2,078 1,255 1, Transfusion (units) Red blood cells Platelets Plasma Cryoprecipitate Reexploration for bleeding (% yes) aspirin use. not using aspirin (Table 4). using aspirin had higher chest tube drainage than nonaspirin users in both the case patient (1318 ml versus 1255 ml) and the control patient (822 ml versus 799 ml) groups, although this was not of statistical significance. There was no significant difference in blood products transfused among aspirin users and nonaspirin users in both the case patient and control patient groups. Likewise, there was no difference in the rates of reexploration for hemorrhage among aspirin users compared to nonaspirin users in both case patient and control patient groups. Comment In this study, preoperative aspirin use by patients undergoing isolated CABG was associated with a 27% reduction in in-hospital mortality. This reduction in mortality persisted (45%) when adjusted for severity of disease and comorbid factors, and appeared to have a greater effect in women than in men. A similar reduction in mortality was reported in the results of The Society of Thoracic Surgeons National Database [10]. Using multivariate analysis, the authors found a 30% reduction (OR 0.70, p 0.002) in mortality of patients who had preoperative antiplatelet therapy among 78,927 patients who underwent operations between 1980 and Reductions in the rate of perioperative myocardial infarction have also been reported in patients receiving preoperative aspirin [11]. There were no significant differences in hemorrhagic complications between patients who did and did not receive aspirin. There was higher chest tube drainage among patients receiving aspirin but this did not achieve statistical significance. receiving aspirin were no more likely to receive blood products (red blood cells, platelets, frozen plasma, or cryoprecipitate) or undergo reexploration for hemorrhage than were patients who did not receive aspirin. The reexploration rate in control patients was high (7% to 8%) compared to current standards (2% as of 1997 in northern New England). The reexploration rate in the case patient group was exceedingly high, nearly 33%. However, all of the patients in this group died in the hospital. Reexploration was likely a desperate attempt to save them. Other studies have also shown no significant increase in bleeding morbidities in patients receiving aspirin preoperatively [12, 13]. Aspirin is thought to inhibit thrombosis of coronary vessels by decreasing platelet aggregation. Aspirin, by irreversible acylation of cyclooxygenase, prevents the conversion of arachidonic acid to cyclic endoperoxide, a prostaglandin precursor. In platelets, this results in decreased production and release of thromboxane A 2, which is a strong vasoconstrictor and platelet aggregator. In areas of turbulent flow (eg, atherosclerotic coronary vessel), shear forces may lead to platelet activation with subsequent thrombin generation and thrombosis. High dosages ( 300 mg/day) of aspirin have been shown to inhibit this activation of platelets [14]. Also, cardiopulmonary bypass results in adverse effects on platelet number, function, and morphology [3, 15]. The transient decrease in platelet function can lead to coagulation abnormalities, and platelet activation by the bypass circuit can result in a release of thromboxane, which can contribute to platelet microaggregate formation and compromised capillary perfusion. Aspirin has been shown to counter this latter effect by limiting thromboxane formation. In experimental studies, thromboxane A 2 receptor-specific blockade has resulted in increased preservation of ventricular function in severe injury models after cardiopulmonary bypass [16, 17]. Therefore, aspirin may have cardiac protective effects that may relate to improved mortality. This study has several limitations. First, the results are vulnerable to the inherent limitation of the case patient control patient method, specifically imperfect matching of case patients and control patients. Selection of an appropriate control patient group is crucial in this type of study and requires that the group be comparable to the source group of the case patients. Though matched for age, sex, surgical priority and center, case patients and control patients were not matched on other factors commonly associated with CABG mortality (eg, ejection fraction or body surface area). We attempted to control for these factors through multivariate analysis, but the existence of unknown confounding variables is possible. Second, our measure of aspirin use (yes or no within 7 days) may be imprecise, leading to inadequate classification of exposure. We did not collect information on dosage or timing of aspirin use within the 7-day period, leading to a possible range of exposure for the aspirin s antiplatelet action. Aspirin use one to two days before the operation is likely to have a much greater effect than aspirin taken a

5 1990 DACEY ET AL Ann Thorac Surg PREOPERATIVE ASPIRIN IN CABG PATIENTS 2000;70: week before. We specifically chose the 7-day period as the more conservative position. The aspirin s action on platelets is irreversible, and it can have a lasting effect for a number of days. Choosing this longer time period results in an underestimate of the benefit of aspirin on mortality, but also may underestimate the aspirin s possible morbidities. However, this seems preferable to using a shorter time period, which would overestimate the beneficial effects of aspirin on mortality, but would mask the problems associated with bleeding (if a time period of 2 days were used there would potentially be more hemorrhage in patients who had taken aspirin three days before the operation than those who had taken none at all, yet they would all be contained in the non-aspirin users group). Reliance on the medical record may introduce recall bias, including elective patients who may not have reported taking aspirincontaining medications within 7 days of their operation. However, the fact that the aspirin group of both case patients and control patients had significantly longer preoperative bleeding times lends support to a correct designation of exposure status. The ability to obtain exposure information recorded before the occurrence of the outcome, in this case in-hospital death, limits the recall bias. Knowing the outcome of death is unlikely to affect the recording of aspirin exposure in the preoperative period. Misclassification, another problem with case patient control patient studies, refers to errors in categorizing either exposure or disease status, and can either be random or differential. Differential misclassification is often a concern when the patient s experience of the outcome event influences their recall of exposure. Before exposure and outcome here are well defined and unlikely to be subjected to differential misclassification, the main problem would be random misclassification. However, this always results in an underestimate of the true relative risk and would bias the findings toward the null hypothesis. Though we cannot rule out the possibility of systemic bias (eg, if aspirin use was missed more often in sicker patients, or if beneficial effects, such as increased internal thoracic artery use, were clustered in aspirin patients); this would lead to a bias of overestimating the beneficial effects of aspirin. Another limitation of the study is that the database is from 1987 to At that time, aspirin was one of the few platelet inhibitors available. The variety of new antiplatelet agents (ticlodinpine, clopidogrel, abciximab, tirofiban, eptifibatide, etc) that patients may be exposed to in addition to aspirin could have an impact on the effects of aspirin we observed. The complexities of these issues strongly suggest that CABG registries prospectively collect detailed information on aspirin and other platelet inhibitor use before the operation. Conclusion We found that the use of preoperative aspirin in patients undergoing isolated CABG is associated with a marked reduction of in-hospital mortality. There were no significant hemorrhagic complications associated with aspirin use. These results, along with the proven benefits of preoperative aspirin on vein graft patency, suggest that it is beneficial for patients to continue aspirin use before CABG. References 1. Taggart DP, Siddiqui A, Wheatley DJ. Low-dose preoperative aspirin therapy, postoperative blood loss, and transfusion requirements. Ann Thorac Surg 1990;50: Sethi GK, Copeland JG, Goldman S, Moritz T, Zadina K, Henderson WG. Implications of preoperative administration of aspirin in patients undergoing coronary artery bypass grafting: department of Veterans Affairs cooperative study on antiplatelet therapy. J Am Coll Cardiol 1990;15: Kallis P, Tooze JA, Talbot S, Cowans D, Bevan DH, Treasure T. Preoperative aspirin decreases platelet aggregation and increases post-operative blood loss: a prospective, randomized, placebo controlled, double-blind clinical trial in 100 patients with chronic stable angina. Eur J Cardiothorac Surg 1994;8: Fuster V, Cohen M, Halperin J. Aspirin in the prevention of coronary disease. New Engl J Med 1989;321: Steering Committee of the Physicians Health Study Research Group. Final report on the aspirin component of the ongoing physicians health study. New Engl J Med 1989;321: Schonberger JP, Bredee JJ, van Oeveren W, et al. Preoperative therapy of low-dose aspirin in internal mammary artery bypass operations with and without low-dose aprotinin. J Thorac Cardiovasc Surg 1993;106: Chesebro JH, Fuster V, Elveback LR, et al. Effect of dipyridamole and aspirin on late vein-graft patency after coronary bypass operations. New Engl J Med 1984;310: Goldman S, Copeland J, Moritz T, et al. Saphenous vein graft patency one year after coronary artery bypass surgery and effects of antiplatelet therapy: results of a Veterans Administration Cooperative Study. Circulation 1989;80: Pierpont GL, Kruse M, Ewald S, Weir EK. Practical problems in assessing risk for coronary artery bypass grafting. J Thorac Cardiovasc Surg 1985;89: Edwards FH, Clark RE, Schwartz M. Coronary artery bypass grafting: the society of thoracic surgeons national database experience. Ann Thorac Surg 1994;57: Klein M, Keith PR, Dauben HP, et al. Aprotinin counterbalances an increased risk of peri-operative hemorrhage in CABG patients pre-treated with aspirin. Eur J Cardiothorac Surg 1998;14: Reich DL, Patel GC, Vela-Cantos F, Bodian C, Lansman S. Aspirin does not increase homologous blood requirements in elective coronary bypass surgery. Anesth Analg 1994;79: Vuylsteke A, Oduro A, Cardan E, Latimer RD. Effect of aspirin in coronary artery bypass grafting. J Cardiothorac Vasc Surg 1997;11: Ratnatunga CP, Edmondson SF, Rees GM, Kovacs IB. Highdose aspirin inhibits shear-induced platelet reaction involving thrombin generation. Circulation 1992;85: Kirklin JK, Westaby S, Blackstone EH, Kirklin JW, Chenoweth DE, Pacifico AD. Complement and the damaging effects of cardiopulmonary bypass. J Thorac Cardiovasc Surg 1983;86: Mendeloff EN, Liang IY, Swain JA, Clark RE. Thromboxane A2 receptor-specific antagonism in hypothermic cardiopulmonary bypass. Ann Thorac Surg 1994;57: Kulatilake N, Gonzalez-Lavin L, Grover GJ. Thromboxane A2 receptor blockade improves contractile function following cardiopulmonary bypass in dogs and pigs. J Surg Res 1991;51: