Author + information
- Received September 10, 1998
- Revision received June 8, 1999
- Accepted August 5, 1999
- Published online November 15, 1999.
- Eng-Shiong Tan, MDa,* (, )
- Jan van der Meer, MDa,
- Pieter Jan de Kam, MSca,
- Peter H.J.M Dunselman, MD∗,
- Barbara J.M Mulder, MD†,
- Carl A.P.L Ascoop, MD‡,
- Matthias Pfisterer, MD§,
- Kong I Lie, MD†,
- for CABADAS Research Group of the Interuniversity Cardiology Institute of the Netherlands∥
- ↵*Reprint requests and correspondence: Dr. Eng-Shiong Tan, Department of Cardiology, Thoraxcentre, University Hospital, Hanzeplein 1, P.O. Box 30.001, 9700 RB Groningen, the Netherlands
This retrospective study sought to assess differences in graft patency and clinical outcome between women and men after coronary artery bypass graft surgery (CABG).
A less favorable clinical outcome has been reported in women as compared with men. Its relation to graft patency has not been studied.
We analyzed one-year follow-up data of 912 patients (120 women) who entered a randomized clinical drug trial. All patients received vein grafts; in 494 patients (56 women) internal mammary artery (IMA) grafts were also used. Graft patency was assessed by coronary angiography at one year. Primary clinical end points were myocardial infarction, revascularization procedures and death; secondary clinical end points included recurrent angina, heart failure and arrhythmias.
Occlusion rates of vein grafts were 16.7% in women and 12.4% in men (odds ratio [OR] 1.62, 95% confidence interval [CI] 0.88 to 3.00, p = 0.12); occlusion rates of IMA grafts were 3.4% and 5.7% in women and men, respectively (OR 0.56, 95% CI 0.08 to 3.96, p = 0.56). Primary clinical end points were observed in 16.7% of women and 9.2% of men (OR 1.97, 95% CI 1.10 to 3.34, p = 0.022), and any clinical end point in 41.7% of women and 25.8% of men (OR 2.06, 95% CI 1.39 to 3.04, p = 0.0004). Myocardial infarction (15% vs. 7.6%, OR 2.15, 95% CI 1.24 to 3.75, p = 0.013) and recurrent angina (26.7% vs. 15.4%, OR 2.00, 95% CI 1.28 to 3.11, p = 0.004) occurred most frequently. Multivariate regression analysis did not identify gender as an independent risk factor for graft occlusion or the clinical end points. Graft occlusion was an independent predictor of the composite primary clinical end point (OR 2.75, 95% CI 1.59 to 4.75, p = 0.0003) and each of the secondary clinical end points. The observed differences were due to an imbalance of risk factors at baseline and to surgical and graft characteristics.
One-year occlusion rates of vein and IMA grafts were comparable in women and men. Clinical outcome was related to graft patency and was less favorable in women owing to their uneven distribution of risk factors among both groups.
Several studies have suggested that the results of coronary artery bypass graft surgery (CABG) in women are less favorable than in men. Two studies demonstrated lower graft patency rates in women (1,2). However, coronary angiography was performed in a small proportion of patients in these retrospective studies, in most cases because of recurrent angina. Clinical outcome, particularly operative mortality, has been studied much more extensively. Women showed higher rates of mortality (3), myocardial infarction (MI) (4)and recurrent angina (2,5). Obvious differences between women and men were not found after adjustment for various risk factors (2,3,6). Whether the reported differences in clinical outcome are related to graft patency remains to be clarified. The results of today’s surgical repair in women may have been improved by changes in surgical techniques, arterial grafts and antithrombotic drugs. To compare graft patency and clinical outcome in women versus men and to assess their relation, we performed a retrospective analysis of the data obtained in a randomized clinical trial—prevention of Coronary Artery Bypass graft occlusion by Aspirin, Dipyridamole and Acenocoumarol/phenoprocoumon Study (CABADAS) (7). This trial was designed to compare the effects of low dose aspirin, alone or in combination with dipyridamole, and oral anticoagulant agents on vein graft occlusion during the first year after CABG.
Study design and patients
Study design, selection criteria and review procedures for assessment of end points have been described previously (7). In brief, 948 patients who underwent elective CABG with saphenous vein grafts for disabling angina were entered into the trial by 10 participating centers between July 1987 and August 1990. Of these, 494 (56 women and 438 men) received internal mammary artery (IMA) grafts in addition to vein grafts. Exclusion criteria included age >70 years, unstable angina <2 days or MI 7 days before the operation, previous or concurrent cardiac surgery, impaired renal or hepatic function, concomitant severe disease and the inability to repeat coronary angiography owing to allergy to a contrast agent. Written, informed consent from all patients was obtained. The protocol was approved by the Ethics Committee of each participating hospital. Patients were assigned randomly to treatment with aspirin, aspirin plus dipyridamole or oral anticoagulant agents (acenocoumarol or phenoprocoumon) for at least one year after the operation. They were stratified per center. Saphenous vein and IMA grafts were implanted according to the routine techniques of each participating hospital. The decision to use IMA grafts was made by the surgeon. The lumen of the grafted coronary artery was measured by calibrated probes at the arteriotomy site and as distal as possible. Coronary endarterectomy was performed in diffusely diseased native arteries at the discretion of the surgeon. Total postoperative blood loss through chest tubes, required red blood cell transfusions, cardiopulmonary bypass time, early reoperation rates (within 24 h) and operative mortality (within 30 days) were recorded. After discharge from the hospital, follow-up visits were scheduled at three-month intervals and coronary angiography was performed one year after the operation. At follow-up visits clinical and laboratory data were collected and an electrocardiogram (ECG) was recorded. A questionnaire was addressed to the cardiologist one year after the operation to complete clinical follow-up data if a patient had been withdrawn from the trial. Coronary angiography was not repeated if it had been obtained for medical reasons >9 months after the operation or if an earlier angiogram already showed graft occlusion according to the protocol of the original vein graft patency study.
Angiographic end points
Grafts were visualized by selective injection. If the origin of a vein graft could not be engaged selectively, aortic root angiography was performed. A vein graft was defined as occluded if either 1) the occluded origin was visualized selectively; 2) the origin could not be visualized selectively and the contrast agent failed to flow through the graft into the grafted artery on aortic root angiography; or 3) one or more distal anastomoses appeared to be occluded. An arterial graft was defined as occluded if occlusion was visualized at selective injection. A distal anastomosis was defined as occluded if the contrast agent failed to flow from the graft into the grafted artery. If the graft was occluded at its origin, all associated distal anastomoses were considered occluded, unless a retrograde flow of contrast agent from the grafted artery into the graft was demonstrated. An arterial graft was classified undefined if selective injection failed for technical reasons and no retrograde filling of the graft with contrast agent at coronary angiography was demonstrated. Angiograms were reviewed by experienced independent cardiologists and members of the angiographic classification committee, and a consensus was reached.
Clinical end points
Primary clinical end points were MI, the need for revascularization procedures (percutaneous transluminal coronary angioplasty or repeat CABG) and death. Myocardial infarction was classified according to strict ECG criteria, including the development of significant Q waves in anterior, lateral or inferior regions, loss of the R wave in lateral leads or a new tall R wave pattern for posterior infarction (8). Enzyme criteria were not used in the immediate postoperative phase. A 12-lead ECG was made in all patients on the day before the operation, the first postoperative day, the day before hospital discharge and at each of the three-month follow-up visits. Additional ECGs were available if an MI had been suspected. Electrocardiograms of all patients were reviewed by an independent blinded classification committee, whether MI was reported by the patient’s cardiologist or not. Death was classified as cardiac or noncardiac according to clinical evidence or necropsy findings. Secondary clinical end points were residual or recurrent angina, heart failure and symptomatic arrhythmias. All reported events were reviewed by the classification committee.
We aimed to compare the incidence of graft occlusion and clinical end points between women and men. Occlusion rates were analyzed by logistic regression with random effects, considering that patients represent clusters of distal anastomoses and the dependence of grafts with respect to occlusion within the same patient (9). Clinical outcome was analyzed by comparison of 1) the incidence of individual clinical end points; 2) the proportion of patients who experienced at least one primary clinical end point; and 3) the proportion of patients with any primary or secondary clinical end point. A difference was quantified by the odds ratio (OR) and its 95% confidence interval (CI). If the 95% CI did not encompass 1, the difference was statistically significant at a 5% level. The groups were compared with respect to baseline characteristics, cardiopulmonary bypass time, blood loss and transfusion requirements by the chi-square test or the Fisher exact test for qualitative variables. The Student ttest or the Mann-Whitney Utest was used for quantitative variables. Continuous variables, unless indicated, are expressed as the mean value ± SD. All tests were two-tailed. A p value <0.05 was considered to indicate statistical significance.
The random-effects logistic regression analysis was applied to assess independent predictors of graft occlusion (10), and multivariate logistic regression analysis was used to evaluate clinical end points. Analysis included all patients conforming to the protocol of the original vein graft patency study. Patients who withdrew their consent before the start of the trial medication or in whom the surgeon decided peroperatively to use only IMA grafts were excluded accordingly.
Of 948 patients who entered the original trial, 912 (120 women and 792 men) were eligible for analysis of clinical outcome at one-year follow-up. Exclusion was due to withdrawal of consent before the start of assigned drug therapy in seven patients and to only arterial grafting in 29 patients. Analysis of graft patency comprised 786 patients (95 women [79%] and 691 men [87%]). Repeat coronary angiography was not available in 126 patients owing to death, withdrawal from the trial or failure of the angiographic procedure. The IMA was used as an additional bypass conduit in 494 patients (56 women [47%] and 438 men [63%]). Baseline clinical characteristics are summarized in Table 1. Compared with men, women were older, had a higher body mass index, systolic blood pressure and serum total cholesterol level. They had more severe angina and more frequently used nitrates. Furthermore, more women had diabetes mellitus. Men were more frequently smokers.
Distal anastomoses averaged 3.6 (vein 3.0, IMA 0.6) in women and 3.8 (vein 3.0, IMA 0.8) in men (Table 2). Single and sequential grafts were equally distributed among the groups. There were no differences between women and men with respect to the location of vein grafts on coronary arteries. Vein grafts were more frequently placed onto small vessels in women. The IMA grafts were somewhat more frequently placed onto the left anterior descending coronary artery in men, whereas there were no differences in the size of coronary arteries receiving IMA grafts.
Occlusion rates by distal anastomosis are shown in Table 3. Of vein grafts 16.7% were occluded in women and 12.4% in men (OR 1.62, 95% CI 0.88 to 3.00, p = 0.12). Occlusion rates of IMA grafts were 3.4% in women and 5.7% in men (OR 0.56, 95% CI 0.08 to 3.96, p = 0.56). Subgroup analysis showed no differences between women and men (Table 4). The highest occlusion rates were observed for vein grafts onto small-sized coronary arteries (lumen diameter ≤1.0 mm). Endarterectomy was performed in 70 vessels with vein grafts (2.6% of all distal sites) and eight vessels with IMA grafts (1.2% of all distal sites) and did not affect the results. Random-effects logistic regression analysis identified the next risk factors for graft occlusion: diabetes mellitus, baseline serum high density lipoprotein cholesterol level, type of distal anastomosis (end-to-side versus side-to-side), location of distal anastomosis (compared with left anterior descending coronary artery), lumen diameter (≤1 mm) of the recipient coronary artery, cardiopulmonary bypass time (>90 min), no administration of protamine sulfate at the end of the operation and postoperative values of hematocrit (>35%) and number of platelets. Gender was not selected as an independent predictor of graft occlusion.
Blood loss, transfusion requirements and reoperation
The mean ± SD volume of blood loss by chest tube drainage was 874 ± 609 ml in women versus 1,025 ± 669 ml in men (p = 0.024). The median transfusion requirement for red blood cells was two (donor) units (range 0 to 9) in women versus one unit (range 0 to 13) in men (p = 0.001). Cardiopulmonary bypass time did not differ between women (99 ± 33 min) and men (96 ± 32 min) (p = 0.53). Early reoperation rates (within 24 h) were 8.3% in women and 3.8% in men (p = 0.031). Operative mortality (within 30 days) was 0.8% in women and 0.6% in men (p = 0.57).
Rates of clinical end points are shown in Table 5. Myocardial infarction occurred in 15.0% of women and 7.6% of men (OR 2.15, 95% CI 1.24 to 3.75, p = 0.013). Of the 78 MIs, 67 (86%) were observed during the perioperative period; 17 (94%) of 18 in women and 50 (83%) of 60 in men. Revascularization procedures were performed during follow-up in 0.8% of women and 1.0% of men. Of 16 patients who died, a cardiac cause was found in 10 (2 women, 8 men), bleeding in 4 (1 woman, 3 men) and thrombosis in 2 (0 women, 2 men). Bleeding was due to a ruptured aortic aneurysm (n = 2) and massive gastric and cerebral bleeding, and the patients with thrombosis had ischemic stroke and pulmonary embolism. Cardiac causes of death were perioperative MI (n = 4) or its late complications (n = 3), sudden death (n = 2) and ventricular arrhythmia (n = 1).
Angina at any time during the one-year follow-up period was reported by 26.7% of women and 15.4% of men (OR 2.00, 95% CI 1.28 to 3.11, p = 0.004). At the end of the first year, 85.0% of women and 91.7% of men were free of angina (p = 0.005). The overall primary event rate was higher in women than in men (16.7% vs. 9.2%, OR 1.97, 95% CI 1.16 to 3.34, p = 0.02), as was the overall event rate, including primary and secondary clinical end points (41.7% vs. 25.8%, OR 2.06, 95% CI 1.39 to 3.04, p = 0.0005). Multivariate regression analysis did not select female gender as an independent predictor of either MI or angina (Table 6). Severity of angina, systolic blood pressure, height and serum triglyceride level at baseline, as well as required transfusions of red blood cells at the operation were identified as independent risk factors for MI. Risk factors for recurrent angina were age, height, nonsmoking and treatment with calcium antagonists before the operation, postoperative hematocrit and graft occlusion. None of the remaining clinical end points was related to gender. In contrast, graft occlusion showed to be an independent risk factor for the need for revascularization procedures (OR 13.21, 95% CI 1.56 to 111.71, p = 0.018), heart failure (OR 4.32, 95% CI 1.54 to 12.11, p = 0.005) and arrhythmia (OR 3.03, 95% CI 1.28 to 7.16, p = 0.012). Severity of angina at baseline, required transfusions of red blood cells and graft occlusion were predictors of the composite primary clinical end point (Table 6).
This study showed that gender is not a risk factor for graft occlusion and cardiovascular events in the first year after CABG, although the opposite was suggested by the observed differences between women and men.
Vein graft occlusion rates tended to be higher in women, probably because of an uneven distribution of grafts to small-sized (≤1 mm) coronary arteries. However, the observed differences did not achieve statistical significance. These grafts, which are more prone to occlude, were more frequently found in women, as expected because of their smaller stature (1). Occlusion rates of IMA grafts were similar in women and men, but apparently lower than those of vein grafts. As we demonstrated previously, this difference is related to an imbalance of graft characteristics, like location of distal anastomosis and diameter of the recipient coronary artery, rather than graft material (11). The only two previous studies that reported on gender-related differences in graft patency showed higher occlusion rates of vein and IMA grafts in women at an average follow-up of two years (1,2). The results of these studies may have been biased by selection. Coronary angiography was performed in 16% and 19% of women and 18% and 23% of men with vein and IMA grafts, respectively, and in a majority of cases because of recurrent angina. In our study, 79% of women and 87% of men underwent coronary angiography, whether they experienced angina or not; asymptomatic women refused coronary angiography more frequently than did men.
Pronounced differences between women and men were observed. Women revealed an approximately twofold higher risk of MI, recurrent or residual angina and composite clinical end points. Analysis of the most common events (i.e., MI and angina) identified the patient’s height as an independent risk factor, whereas gender did not contribute to the observed differences. Consequently, women are at higher risk, as they commonly are smaller than men. Myocardial infarction mainly occurred within the first postoperative days (women 14.2% and men 6.3%), whereas its rate was low and equal in women and men for the remaining follow-up period (0.8% vs. 1.3%). The perioperative difference between both groups probably reflects technical difficulties of grafting small coronary arteries and an a priori higher risk of occlusion for these grafts. Recurrent or residual angina, though reported more frequently by women, did not influence the need for repeat revascularization procedures. As angina is subjective, it might have been more readily reported by women. However, this explanation is unlikely, because the difference could not be attributed to gender. Moreover, angina was found to be related to graft occlusion. Our findings are in agreement with the results of the few previous studies that addressed gender-related differences in clinical outcome, other than mortality (2,4,5,12).
One-year mortality was somewhat higher in women than in men (2.5% vs. 1.6%), but operative mortality was low and equal in the two groups (0.8% vs. 0.6%). Since 1975, 23 studies consistently reported an approximately twofold higher hospital mortality rate in women (3). This difference has been explained by exposure to risk factors, more frequently found in women (13,14). The low mortality and absence of a difference between women and men in our study, despite an imbalance of risk factors similar to previous studies, may be due to exclusion of patients who required an urgent operation.
Relation of graft occlusion and clinical outcome
Graft occlusion was not identified as a risk factor for MI and death, in contrast with the other clinical end points. This lack of a relation might be explained by missing angiographic data. Coronary angiography was not performed in 31% of patients with MI, as compared with 12% of patients without MI. Only one of 16 patients who died had undergone angiography. Autopsy in five patients showed graft occlusion in one of three with cardiac death and in one of two with noncardiac death. These numbers were too small for analysis. Furthermore, graft occlusion at time of MI, mostly in the first postoperative days, may not have been detected by angiography at one year owing to restored patency.
Surgical bleeding as a risk factor
Multivariate analysis identified determinants of surgical bleeding as risk factors for graft occlusion and recurrent angina (hematocrit >35% at discharge) and MI (number of transfusions). This finding seems to agree with the results of a recent study that showed an increased perioperative mortality in women owing to major bleeding (3). The higher rate of early reoperations in women in our study reflected the higher risk of surgical bleeding, probably related with technical difficulties in smaller patients, but it did not influence operative mortality. Remarkably, women had less blood loss, even after correction for their smaller body size, while they received more transfusions. An excess of transfusions in women, and hence a relatively higher postoperative hematocrit and higher blood viscosity, may explain this discrepancy, as well as the relation between transfusions, postoperative hematocrit, graft occlusion and clinical outcome. Women already have a physiologically lower hematocrit before operation and tolerate a lower postoperative level than do men. They will show a stronger decline of hematocrit by hemodilution during cardiopulmonary bypass owing to their smaller body size. As transfusions are commonly given to women and men at the same hematocrit level, women thus may receive an excess of transfusions. The resultant higher blood viscosity, together with the smaller size of coronary arteries, may have contributed to early graft occlusion and perioperative MI in women. Our observation is supported by the results of another study that demonstrated an increased risk of MI, more severe left ventricular dysfunction and a higher mortality rate at a postoperative hematocrit value ≥34% (15).
Potentially confounding factors
It has been suggested that clinical outcome in women is impaired because they are referred later in the course of their disease, thereby more frequently revealing unstable angina before the operation (16,17). Furthermore, revascularization might be more complex and less complete in women owing to their smaller stature and, consequently, smaller coronary arteries (1,4,18). In our study, patients who experienced unstable angina <2 days or MI <7 days before the operation were excluded. Revascularization was extensive and comparable in women and men (3.6 and 3.8 distal anastomoses, respectively). The smaller proportion of women (47%), as compared with men (63%), in whom IMA grafts were used might have favored clinical outcome in men, as arterial grafting has been associated with improved morbidity and mortality (19,20). This assumption, however, was not supported by the present analysis. Previously, we demonstrated no difference in clinical outcome between patients who received both IMA and vein grafts and those who received only vein grafts (11). Finally, we considered the possibility that clinical outcome in women had been influenced by aspirin, as it has been suggested that this drug is less effective in the prevention of cardiovascular events in women as compared with men (21,22). We found no differences in graft patency and clinical outcome between women and men, comparing treatment with aspirin versus oral anticoagulants (data not shown).
One might speculate that the worse clinical outcome in women is related to progression of atherosclerosis in native coronary arteries rather than graft occlusion, considering the discrepancy of a pronounced difference in clinical outcome and a statistically nonsignificant difference in graft occlusion rates between women and men. Although antithrombotic drugs largely maintain graft patency, they may be insufficient to delay progression of atherosclerosis, especially in women, who are more extensively exposed to risk factors. Previous studies showed a similar maldistribution of risk factors (6,13,14). It should be noticed, however, that MI, a major determinant of clinical outcome, mostly occurred within the first postoperative days.
Assuming that clinical outcome after CABG in women depends on progression of atherosclerosis in native coronary arteries, further attempts to improve the results of the operation should be aimed at reducing the risk factors for atherosclerosis. Women, as well as men, may benefit from such an approach, as has been demonstrated for lipid-lowering drugs (23)and discontinued smoking (24). An important risk factor for graft occlusion is the lumen diameter of the grafted coronary artery. As small-sized coronary arteries are more common in women owing to their smaller stature, improvement by modification of this risk factor can hardly be obtained. This emphasizes the need for an appropriate drug regimen to maintain graft patency. Considering the technical difficulties of operating on small patients, improved techniques may still contribute to a reduction of early graft occlusion, as well as a critical reappraisal of current criteria for transfusions.
This study was performed retrospectively in patients <70 years old who underwent elective CABG. It is possible that the observed lack of a statistically significant difference in vein graft occlusion rates between men and women was related to an insufficient number of women. However, it is one of the largest studies in which consecutive patients were scheduled prospectively for repeat coronary angiography at one year postoperatively. Moreover, it is one of the few studies that assessed both patency of vein and IMA grafts and clinical outcome. Thereby, this study enabled us to analyze the relation between clinical outcome and graft patency. Another limitation is its short follow-up of one year. This was appropriate to assess the effects of antithrombotic drugs in the prevention of vein graft occlusion, the primary aim in the original trial. Extrapolation of the results to a longer term is not allowed.
One might wonder whether the results of this study are applicable to current practice. The IMA grafts were used in ∼50% of patients because of the design of the original trial, which primarily addressed vein grafting. Moreover, only 13% of patients were women, although this represented the national rates at that time. However, neither gender nor graft material was identified as a determinant of graft patency and clinical outcome at one year. Therefore, it is not likely that our analysis of risk factors has lost its validity only by changes in the proportion of women or patients with IMA grafts.
The present study showed a relation between graft patency and clinical outcome. Occlusion rates of vein and arterial grafts at one year after CABG were comparable in women and men. A pronounced less favorable clinical outcome in women was not related to gender, but to an uneven distribution of risk factors among both groups.
The authors contributed by conception and design (J. van der Meer), statistical analysis (P. J. de Kam), interpretation of data (all), drafting of the manuscript (E. S. Tan) or revising it critically (all).
Of 35 variables, the following were selected by univariate analysis at a p value <0.20 and entered into the logistic regression analysis of risk factors for angiographic and clinical end points. In previous studies body surface area has been used instead of height, although height correlates better with lean body mass and therefore might be more predictive of coronary artery size. We used height because it was selected by univariate analysis in contrast with body surface area.
graft material (IMA or vein), number of distal anastomoses per graft, type of distal anastomosis (end-to-side or side-to-side), location of distal anastomosis, lumen diameter of the recipient coronary artery, gender, height, angina New York Heart Association (NYHA) functional class (baseline), diabetes mellitus (baseline), blood pressure (baseline), serum high density lipoprotein cholesterol (baseline), assigned trial medication, concomitant medication (i.e., lipid-lowering drugs [baseline]), cardiopulmonary bypass time, administration of protamine sulfate at the end of surgery, postoperative hematocrit, postoperative number of platelets and center.
gender, height, angina NYHA functional class (baseline), history of hyperlipidemia, systolic blood pressure (baseline), serum triglyceride (baseline), transfusion requirement at operation, graft occlusion (at one year) and center.
gender, age, height, smoking (baseline), concomitant medication (i.e., calcium antagonists [baseline]), postoperative hematocrit, graft occlusion (at one year) and center.
number of distal anastomoses and graft occlusion.
gender, age, angina NYHA functional class (baseline), history of MI, history of hyperlipidemia, diabetes mellitus (baseline), concomitant medication (i.e., beta-blocking agents [baseline]), transfusion required at operation, MI (follow-up), revascularization procedures (follow-up), graft occlusion (at one year) and center.
age, weight, angina NYHA functional class (baseline), history of MI, history of hyperlipidemia, serum cholesterol (baseline) and center.
Composite primary clinical end point
gender, age, height, weight, angina NYHA functional class (baseline), history of hypertension, history of hyperlipidemia, nonsmoking before operation, systolic blood pressure (baseline), serum high density lipoprotein cholesterol (baseline), cardiopulmonary bypass time, transfusion required at operation, postoperative number of platelets, graft occlusion (at one year) and center. (Appendix)
☆ This study was funded by Grant 86.077 from the Netherlands Heart Foundation and by grants from Dr. Karl Thomae, GMBH, Germany; Boehringer Ingelheim BV, the Netherlands; and the Interuniversity Cardiology Institute of the Netherlands.
- prevention of Coronary Artery Bypass graft occlusion by Aspirin, Dipyridamole and Acenocoumarol/phenoprocoumon Study
- coronary artery bypass graft surgery
- confidence interval
- internal mammary artery
- myocardial infarction
- New York Heart Association
- odds ratio
- Received September 10, 1998.
- Revision received June 8, 1999.
- Accepted August 5, 1999.
- American College of Cardiology
- Loop F.D,
- Golding R.L,
- MacMillan J.P,
- Cosgrove D.M,
- Lytle B.W,
- Sheldon W.C
- Douglas J.S,
- King S.B III.,
- Jones E.L,
- Craver E.M,
- Bradford J.M,
- Hatcher C.R
- Northern New England Cardiovascular Disease Study Group,
- O’Conner G.T,
- Morton J.R,
- Diehl M.J,
- et al.
- Rahimtoola S.H,
- Bennett A.J,
- Fessler C,
- Grunkemeier G.L,
- et al.
- Bär F.W,
- Brugada P,
- Dassen W.R,
- van der Werf T,
- Wellens H.J.J
- Mauritsen R.H
- ↵Van der Meer J, Hillege JL, van Gilst WH, et al. A comparison of internal mammary artery and saphenous vein grafts after coronary artery bypass surgery: no difference in one year occlusion rates and clinical outcome. Circulation 1994;90:2367–1274.
- Sergeant P,
- Lesaffre E,
- Flameng W,
- Suy R,
- Blackstone E
- Institutions of the Multicenter Study of Perioperative Ischemia (McSPI) Research Group,
- Spiess B.D,
- Ley C,
- Body S.C,
- et al.
- Cosgrove D.M,
- Loop F.D,
- Lytle B.W,
- et al.
- Kirklin J.W,
- Naftel D.C,
- Blackstone E.H,
- Pohost G.M
- Weksler B.B
- Vaccarino V,
- Krumholz H.M,
- Berkman L.F,
- Horwitz R.I
- Jukema J.W,
- Bruschke A.V.G,
- van Boven A.J,
- et al.
- Voors A.A,
- van Brussel B.L,
- Plokker H.W.T,
- et al.