The use of preoperative exercise testing

Transcription

1 The use of preoperative exercise testing to predict cardiac complications after arterial reconstruction N. McPhail, M.D., F.R.C.S.(C), J. E. Calvin, M.D., F.R.C.P.(C), A. Shariatmadar, M.Sc., G. G. Barber, M.D., F.R.C.S.(C), and T. K. Scobie, M.D., F.R.C.S.(C), Ottawa, Ontario, Canada To assess the value of exercise testing in the prediction of cardiac risk, 100 patients requiring arterial reconstructive surgery had either treadmill testing or arm ergometry before operation. Thirty-four patients then had abdominal aortic aneurysm repair, 48 had reconstructions for aortoiliac occlusive disease, and 18 had infrainguinal revascularization procedures. Cardiac complications included myocardial infarction in 10%, acute congestive failure in 5%, malignant ventricular arrhythmias in 7%, and cardiac death in 7%. Contingency table analysis showed that patients who achieved less than 85% of their predicted maximum heart rate (PMHR) during exercise testing had a complication rate of 24%, whereas patients who achieved more than 85% of PMHR had a 6% complication rate (p = 0.036). The degree of ST segment depression during exercise testing was not a significant predictor of cardiac complications. However, patients who had a positive stress test (ST depression more than 1 mm) and achieved less than 85% of their PMHR had a complication rate of 33%, whereas patients with a positive stress test who achieved more than 85% of their PMHR had no complications (p = 0.048). Statistical analysis with a logistic regression model showed two factors to be significant. Patients who achieved a high maximum heart rate during exercise testing had a low probability of developing postoperative cardiac complications (p = 0.04), as did patients who achieved high cardiac work load maximal oxygen uptake (p = 0.03). We conclude that preoperative exercise testing is useful to predict cardiac complications after arterial reconstruction. Patients who are able to achieve more than 85% of their PMHR and a high maximal oxygen uptake represent a low-risk group. (J Vase SURG 188;7:60-8.) Cardiac complications are the greatest cause of death and major morbidity after vascular surgical operations. Patients who have vascular disease are at high risk because of their advanced age, the presence of associated coronary atherosclerosis, and the stressful nature of many operative procedures. The problem is likely to increase because larger numbers of more complex operations are performed on a progressively older population. To reduce the incidence of major and potentially fatal cardiac complications, it is necessary to identify high-risk patients before operation. The information can then be used either From the Divisions of Vascular Surgery and Cardiology, University of Ottawa. Presented at the Thirty-fifth Scientific Meeting of the North American Chapter, The International Society for Cardiovascular Sup gery, Toronto, Ontario, Canada, June 8-, 187. Reprint requests: N. McPhail, M.D., C.P.C., Suite 634, 1053 Carling Ave., Ottawa, Ontario, Canada K1Y 4E. to improve cardiac function or, if this is not feasible, to alter the planned procedure. Available methods to predict cardiac risk include clinical indexes, exercise testing, cardiac imaging, and coronary arteriography. This article reports our experience with the use of exercise testing in an effort to predict cardiac complications in patients requiring arterial repair. METHOD Between May 185 and June 186, 110 consecutive patients requiring arterial surgery who had clinical evidence of significant coronary artery disease were referred for cardiac evaluation before operation. All patients were seen by the same consultant cardiologist. Nine patients with recent myocardial infarction (less than 6 months), unstable angina, or evidence of congestive failure were excluded from the study. Eight of these patients had aortocoronary bypass surgery with one postoperative death. One patient had inoperable disease and died within a month. 60

2 Volume 7 Number 1 January 188 Exercise testing to predict cardiac risk 61 Table I. Relationship of maximal rate to cardiac complications % PMHR No complications Complications (%) Total < (24.3) 70 > (6.6) 30 p = Table II. ST depression and cardiac complications ST depression (mm) No complications Complications (%) Total < (17.1) (23.8) 21 >2 7 2 (22) NS The remaining 101 patients were referred for exercise testing. This was carried out in the Heart Institute Laboratory under the supervision of an independent physician who had no prior knowledge of the patients. If possible, a treadmill exercise test was performed. If walking on a treadmill was prevented by severe claudication, arm ergometry was used. Sixty-one patients had standard treadmill exercise testing with the protocol of Bruce et al. 1 In these patients, a control interval of 5 minutes permitted assessment of dysrhythrnias at rest and baseline ST segment abnormalities on 12-lead electrocardiography. Leads V1, Vs, and avf were monitored continuously throughout exercise and recovery until the control heart rate was achieved. Exercise was started at a work load of 4 maximal oxygen uptake (METs), and this was increased every 3 minutes. At each 3-minute interval, cuff blood pressure and 12-lead electrocardiography were recorded. Both the blood pressure and the electrocardiogram were repeated at peak exertion and 3 and 6 minutes into recovery. Forty patients had arm crank ergometry with the protocol of Schwade et al. 2 In this protocol the patients were monitored for 5 minutes before the test, and a 12-lead electrocardiogram was performed at baseline in a sitting position. The patient was seated with the fulcrum at shoulder level and the metronome was set at 50 rpm with a modified bicycle ergometer. The first stage of the protocol was at zero load. Each stage lasted 3 minutes with a 1-minute rest period between stages. Work loads were increased by 150 kilopondmeters per minute per stage. Blood pressures were recorded within 15 seconds of each recovery period, and 12-lead electrocardiograms were taken during the last minutes of each stage. In both protocols, the test was terminated when a patient achieved or experienced one of the following: (1) 85% of the predicted maximal heart rate (PMHR) ibr sex and age, (2) symptoms of shormess of breath, angina, or fatigue sufficient to preclude further exercise, (3) serious ventricular arrhythmia (sustained or nonsustained ventricular tachycardia), and (4) a decrease in systolic blood pressure during exercise (more than 10 mm Hg). A positive exercise test was interpreted when horizontal ST segment depression below the resting baseline in any lead was at least 1 mm in depth. Slow upsloping ST segment depression of 1.5 mm below baseline, 80 msec after the J point, was also judged positive. The test results were interpreted and reported by an independent cardiologist who was not involved in the study. Seven of the 101 patients who were tested had severe angina and ST depression greater than 2 mm during the first stage of the exercise test, and the test was terminated. Coronary arteriography confirmed the presence of severe coronary artery disease and all seven patients had aortocoronary bypass, resulting in one postoperative death. The remaining six patients underwent later vascular reconstruction and were included in the study group. Ninety-four patients proceeded directly to planned arterial repair after their exercise test. This left a study group of 100 patients, who form the sample evaluated for predictors of complications after arterial reconstruction. Forty-eight patients had aortoiliac reconstructions for occlusive disease, 35 had repair of abdominal aortic aneurysms, and 18 had peripheral arterial operations. The postoperative cardiac complications studied were acute myocardial infarction, acute congestive cardiac failure, ventricular tachycardia or fibrillation, and cardiac death. Acute myocardial infarction was diagnosed by electrocardiographic evidence of new ST segment elevation, which was corroborated by the finding of elevated cardiac enzymes (CK-MB greater than 5% of total creatine kinase). Acute congestive failure was diagnosed by clinical evidence of acute pulmonary edema, pulmonary capillary wedge pressure greater than 18 mm and a cardiac index of less than 2.2. The test variables analyzed were exercise time, maximal heart rate, maximal systolic blood pressure, the onset of ST segment changes, the maximum ST depression in millimeters, the development of angina and its time of onset, the presence of ventricular ar-

3 62 McPhail et al. Journal of VASCULAR SURGERY Table III. Relationship of PMHR to complications in patients with positive exercise test (~ ST ~> 1 ram) ST depression > 1 mm No complications Complications (%) Total PMHR < 85% 14 7 (33.3) 21 PMHR > 85% 0 (0) p = rhythmias, and METs. The study was prospective, and computer-coded sheets were uscd to create a data file. The information was subjected to statistical evaluation, first by contingency table analysis, and then by logistic regression analysis. RESULTS Major cardiac complications observed in 100 patients were acute myocardial infarction in 10, acute congestive failure in five, malignant ventricular arrhythmias in seven, and cardiac death in seven. Contingency table analysis showed that maximum heart rate achieved during exercise was a signfficant predictor of complications (Table I). Of 70 patients who achieved less than 85 % of their PMHR, 17 developed complications (24.3%). Only two (6.6%) of 30 patients who achieved more than 85% of their PMHR had complications (p = 0.036). The degree of ST segment depression that occurred with exercise was not significant in predicting cardiac complications (Table II). Twelve of 70 patients (17.7%) with ST depression less than 1 mm developed complications, as did five of 21 patients (23.8%) with ST depression of 1 to 2 mm and two of nine patients (22%) with ST depression of more than 2 mm. However, if patients with ST segment depression of 1 mm or greater (a positive stress test result) were compared on the basis of whether they achieved 85% of their PMHR, the result became significant (p = 0.048). Of 21 patients with a positive stress test (1 rnm or greater ST depression) who attained less than 85% of their PMHR, seven (33.3%) developed cardiac complications. In contrast, no complications occurred among nine patients with ST depression of i mm or more who were ablc to achieve 85% of their PMHR (Table III). There were no other significant test variables. The data were then analyzed by means of a logistic regression model. Each test variable was entered into the model separately, and the probability of complications was estimated. The analysis indicatcs that patients who achieved a high maximal heart rate during exercise had a low probability of developing cardiac complications (p = 0.040). The correlation coefficient calculated from the model (R = ) also indicates that the variation caused by maximum heart rate is 15.5%. Fig. 1 illustrates the relationship between maximal heart rate and the probability of cardiac complications, which is derived from estimation of the coefficient for maximal heart rate. A similar result was observed when a high METs was present (p = 0.033). The correlation coefficient r = indicates that the variation in complications from METs was 17.3%. Fig. 2 illustrates the relationship between METs and the probability of cardiac complications and was estimated by the same method. No other test variables were significant in this model. DISCUSSION Myocardial infarction is responsiblc for 50% of the deaths and major morbidity after vascular reconstructive operations, aa Among 273 consecutive patients reported by Hertzer s who had lower extremity revascularization and were followed up for as long as 11 years, acute myocardial infarction caused 52% of the early postoperative deaths and 55% of the late deaths. Hertzer et al. 6 found that the use of routine coronary arteriography in 1000 patients who had elective peripheral vascular reconstruction revealed severe correctable coronary artery disease in 25%. Several reports demonstrated that patients tolerate major arterial reconstructions well after successful aortocoronary bypass3'4; however, it has not as yet been conclusively shown that the overall mortality rate is reduced by the routine use of coronary arteriography and where indicated, aortocoronary bypass before peripheral arterial repair. The risk of aortocoronary bypass is substantial, particularly in patients more than 70 years old and has to be included in the overall mortality and morbidity rates. This observation is supported by the operative mortality rate of the recent Coronary Artery Surgery Study, which demonstrated a mortality rate of 5.8% in men and 12.3% in women (combined mortality rate of 7.%) ages 70 years or more. 7 This mortality rate must be included in any discussion of the po-

4 . Volume 7 Number 1 January 188 Exercise testing to predict cardiac risk o f c 0.25: I o c o m P 0.20 E 0.15 d Q 0.I0 d g g O. 05- O. O0 d B II m~xre~e Fig. 1. Relationship of maximal heart rate to probability of cardiac complications (logistic regression). tential benefit to be derived from prophylactic bypass in older patients with coronary disease. Moreover, coronary artcriography is an expensive screening technique and provides an anatomic rather than a functional evaluation of the heart. ' In I77, Goldman et al.8 published the first multivariate statistical analysis of clinical cardiac risk fac- tors in a study of 1001 patients undergoing noncardiac procedures. They developed a nine-variable risk index and showed that it had good predictive value for the data from which it was generated; however, relatively few of the patients in their study ( 80/1001) had arterial reconstructive operations. A recent modification of the Goldman risk index has been reported

5 64 A/IcPhail et al. Journal of VASCULAR SURGERY O. 4O ~ I0 Q O. 00- ' i l l.... I.... I.... I.... I ' ' "~F~-' " 'l....!.... I ' ~ - ' - v - r " r ' F ~ ' ' ' I.... i ' l l ' r ' ~ ' ' l ' ' ' ~ 0 I fo l! t m~t8 Fig. 2. Relationship of METs to probabili~ of cardiac complications (logistic regression). by Detsky et al., who added additional variables and altered the scoring scheme. This modified risk index has been recently applied by Johnston and Scobie (personal communication) to a large group of panents requiring aortic aneurysm repair with encouraging results. We have previously studied clinical cardiac risk factors in a group of 353 patients who had arterial reconstructive operations during a 15-month period. 1 Fifty-six patients had postoperative cardiac complications that resulted in 1 deaths. With the usc of both contingency table and logistic regression analysis, the following four factors were significant predictors of cardiac risk: (1) age more than 70 years, (2) symptom severity according to the New York Heart Association classification, (3) electrocardiographic evidence of a previous myocardial infarct, and (4) nonspecific ST segment changes on the resting electrocardiogram. The logistic regression model indicated that these criteria were able to predict only 30% of the observed cardiac complications. In a subsequent small group of patients who dcvelopcd unexpected cardiogenic shock after major arterial repair and had triple-vessel coronary artery diseasc and subendocardial infarction, clinical risk

6 Volume 7 Number 1 Januaq, 188 Exercise testing to predict cardiac risk 65 evaluation failed to identify the high risk.n Accordingly, we believe that clinical risk indexes may be difficult to apply successfully to individual patients. We chose to study exercise stress testing because the technique is simple, widely available, and has become a well-standardized screening technique in patients with coronary artery disease since the initial studies of Bruce et al. ~ They found that the overall mortality rate could be predicted in patients with coronary artery disease and that it increased if the exercise duration was less than 3 minutes and the maximal systolic blood pressure achieved during exercise was less than 130 mm Hg. n Exercise testing can also be used to predict the outcome after acute myocardial infarction. Williams et al. I3 tested 205 patients 12 days after myocardial infarction and found that during the first year of follow-up, the incidence of recurrent myocardial infarction and cardiac death was increased if exercise duration was less than 3 minutes. Cutler et al)4 demonstrated that preoperative exercise testing could be used to predict the risk of postoperative myocardial infarction in patients requiring vascular surgery. They found that those patients who had more than 1 mm ST depression and who achieved less than 75% of their PMHR were at highest risk. They also noted that the usefulness of the test was limited by the inability of one third of their patients to achieve adequate exercise levels because of claudication, and they used arm ergometry in an effort to improve the test results. The management implications of exercise test resuits were studied by Arous et al) s in 135 patients with vascular disease who had a positive exercise test. In 56 patients who had their originally planned operation irrespective of the test result, 15 myocardial infarctions occurred after surgery, which resulted in 11 deaths. Twenty-three patients had a less stressful operation, which resulted in four postoperative myocardial infarctions and one death. Ten patients had coronary arteriography and aortocoronary bypass, which was followed later by planned vascular surgery, with no myocardial infarctions and no deaths. In our patient population, nine patients were excluded from stress testing because of severe unstable angina, a recent myocardial infarct, or evidence of congestive cardiac failure. Eight of these patients had aortocoronary bypass, with one postoperative death. One patient had inoperable disease and died within a month. None of these patients ever had the planned vascular operation. In the group of 100 patients subjected to exercise testing who had subsequent vascular reconstruction, the most important test factor was the maximal heart rate achieved during exercise, and ST depression became significant only when combined with this variable. Both maximal heart rate and METs, which is related to maximal rate, were significant predictors with the use of the logistic regression statistical model. Although the incidence of complications was low (6.6%) among patients who achieved more than 85% of their PMHR, the usefulness of this observation was limited by the fact that 70% of our patients could not achieve this exercise level with either treadmill testing or arm ergometry. The surgeon, cardiologist, and anesthetist were aware of the test results. Although the exercise test information was not used to influence the type of arterial repair eventually performed, seven patients had severe angina and ST depression greater than 2 mm on the first stage of the exercise test. It was believed that it would be unsafe to continue with the test, and all patients had coronary arteriography. This showed triple-vessel disease in four and left main coronary artery stenosis in three. All seven patients had aortocoronary bypass surgery resulting in one postoperative death. The remaining six patients had planned peripheral vascular reconstruction and were kept in the study group because we thought it was not known whether previous coronary artery bypass grafting would influence total cardiac complications. Although all six patients survived the vascular operation, one had acute pulmonary edema after operation. Our present view is that exercise testing is a useful screening method in patients who require arterial reconstruction. If they can achieve a high maximum heart rate and a high cardiac work load METs, the risk of complications is low, and they can proceed to have the operation using the test in conjunction with clinical evaluation. If they are unable to achieve 85 % of their PMHR, further investigation is needed, including dipyridamole-thallium imaging, 16,17 and cor: onary arteriography 6 where appropriate. We acknowledge the assistance of Mr. Alan Skanes, medical student, and Mrs. Prudy White, R.N., in the preparation of the data base for this article. REFERENCES 1. Bruce RA, Kasumi F, Hosmer J. Maximal oxygen uptake and nomographic assessment of functional aerobic impairment in cardiovascular disease. Am Heart J 173;85: Schwade J, Blomqvist AG, Shapiro W. Comparison of the response to arm and leg work in patients with ischemic heart disease. Am Heart J 177;4: Thompson JE, Hollier LH, Patman RD, Person A. Surgical management of abdominal aortic aneurysms: factors influ-

7 66 A/icPhail et al. Journal of VASCULAR SURGERY encing mortality and morbidity. A 20 year experience. Ann Surg 175;181: Crawford DS, Bomberger RA, Glaser DH, Saleh SA, Russell WL. Aortoiliac occlusive disease: factors influencing survival and fi.mction following reconstructive operation over a 25 year period. Surgery 181;0: Hertzer NR. Fatal myocardial infarction following lower extremity revascularization. 273 patients followed 6 to 11 postoperative years. Ann Surg 181;13: Hertzer NR, Young GR, Kraemer JR, et al. Coronary artery disease in peripheral vascular patients. A classification of 1,000 coronary angiograms and results of surgical management. Ann Surg 184;1: Kennedy JW, Kaiser GC, Fisher LD, et al. Clinical and angiographic predictors of operative mortality from the collaborative study in coronary artery surgery. Circulation 181;63: Goldman L, Caldera DL, Nussman SR, et al. Multifactorial index of cardiac risk in non-cardiac surgical procedures. N Engl J Med 177;27: Detsky AS, Abrams HB, Forbath N, Scott JG, Hilliard JR. Cardiac assessment for patients undergoing non-cardiac surgery: a multifactorial risk index. Arch Intern Meal 186; 146:2131-4, 10. McPhail N, Menkis A, Shariatmadar A, et al. Statistical pre- diction of cardiac risk m patients who undergo vascular surgery. Can J Surg 185;28: Calvin J, Kieser T, Walley V, McPhail N, Barber G, Scobie TK. Cardiac mortality and morbidity after vascular surged,. Can J Surg 186;2: Bruce RA. Exercise testing for evaluation ofventricular function. N Engl J Med 177;26: Williams WL, Nair R, Higginson L, Baird M, Allan K, Beanlands D. Comparison of clinical and treadmill variables for the prediction of outcome after myocardial infarction. J Am Coil Cardiol 184;4: Cutler B& Wheeler HB, Paraskos JA, Cardullo PA. Applicability and interpretation of electrocardiographic stress testing in patients with peripheral vascular disease. Am J Surg!81;141: Arous EJ, Baum P, Cutler BS. The ischemic exercise test in patients with peripheral vascular disease: implications for management. Arch Surg 184; 11: Leppo J, Joaquin P, Gionet M, Tumolo J, Paraskos J, Cutler B. Noninvasive evaluation of cardiac risk before elective vascular surgery. J Am Coil Cardiol 187;: Curler B, Leppo J. Dipyridamole thallium scintigraphy to detect coronary artery disease before abdominal aortic surgery. J VASC SURG 187;5:1-. DISCUSSION Dr. David C. Brewster (Boston, Mass.). This paper by McPhail et al. from Ottawa highlights the continued interest in identifying the subset of patients with preoperative vascular disease truly at high risk of perioperative cardiac complications. Few of us would question the importance and timeliness of this topic. However, the best means of achieving such risk stratification remains controversial. This report focuses on the role of exercise testing as a predictor. The finding of a highly significant difference in cardiac event rate--only 6.6% in patients able to achieve 85% of their predicted maximal heart rate vs a 24% complication rate in patients unable to do so--is clearly quite impressive. I was also interested in their data, which show that the heart rate seems to supersede the importance of the conventional criteria of ST segment depression. The low likelihood of cardiac problems is reassuring if one's patient is capable of achieving this exercise stress level. However, the major limitation of the current study, as the authors acknowledge, is the fact that a full 70% of their patients could not achieve this exercise level. The inability of many patients with vascular disease to achieve adequate degrees of exercise stress because of claudication, pulmonary insufficiency, or simply advanced age and then infirmity is well recognized. Arm ergometry, as used in 40% of the current study group, is often an inadequate substitute for the treadmill. Although cardiac complications are indeed more likely in the large group unable to reach an adequate heart rate, such testing does not appear selective enough as a predictor to warrant more intensive preoperative evaluation, specifically coronary angiography. What then is the clinician to do with this large group of patients? Our own approach to risk stratification emphasizes pharmacologically induced cardiac stress as opposed to exercise. Intravenous dipyridamole (Persantine) is a maximal coronary vasodilator, and should be available within the year with approval from the Food and Drug Administration for use in such circumstances. When combined with thallium 201 myocardial imaging, it provides information equal to exercise-thallium studies. Reversible thallium defects are highly predictive of ischemic myocardial areas at risk for new perioperative ischemia. Our most recent data show that 88% of perioperative cardiac ischemic events occurred in patients with reversible thallium defects. Unlike the authors, we continue to believe that clinical indicators are still useful in risk delineation, separating patients into high- and low-risk groups. Specifically, if a patient has no histol T of angina, myocardial infarction, congestive heart failure or diabetes, and no Q wave abnormalities on an electrocardiogram, the risk of cardiac ischemia is very low (2%), and in our opinion no further evaluation appears necessary. If one or more of these clinical markers is present, screening with dipyridamole thallium is carried out. Our experience has now expanded to approximately 200 patients with similar results. Patients with either nor-

8 Volume 7 Number 1 January 188 Exercise testingto predict cardiac risk 67 mal dipyridamole-thallium scans or fixed defects indicative of stable lesions of scar are at very low risk of cardiac complications. Conversely, patients in the high-risk group with reversible thallium defects have a cardiac event rate of almost 50% with vascular surgical reconstruction. These are the patients in whom more intensive preoperative investigation and scrutiny are needed. However, I emphasize that delineation by risk grouping into high and low risk on the basis of clinical indicators will exclude about one half of patients from further study, and the use of dipyridamole-thallium scanning will reduce the patients truly at risk to approximatdy 20%. In our opinion, such risk stratification is much more discriminatory than is exercise testing, and in our mind, is the current approach of choice. Dr. McPhail, could you indicate the reason that the exercise was stopped in most of the patients. Was it due to c!audication, angina, electrocardiographic abnormalities, or simple infirmity? Could you tell us the clinical characteristics of patients with cardiac ischemic events in your group? Did any occur in truly asymptomatic patients? Finally, what do you advise for a large group, namely, 70% of patients in your study, in whom the clinician is still uncertain about the exact magnitude of cardiac risk? Dr. McPhail. Our present policy is that if a patient has severe symptoms, he would proceed directly to coronary arteriography. If he does not have severe symptoms, he would receive exercise testing, and if the test results were negative, he would proceed to operation. If the test were positive or equivocal, then he would have thallium imaging and coronary arteriography if it seemed appropriate. Like Dr. Cutler, we are also actively involved in studying thallium imaging and have data on 77 patients. These data have not as yet been analyzed and are not included in this report. We are also studying ejection fraction in the same group of patients and are trying to compare ejection fraction and thallium imaging with exercise testing in this patient population. We continue to believe that clinical criteria are important and that probably a third of high-risk patients can be identified by purely clinical criteria, a further third by exercise testing, and a further third by these other modalities. President-Elect Perdue. Dave, were your questions answered? Dr. Brewster. I believe, Mr. President, that most of the questions were answered, but I would like to ask Dr. McPhail whether indeed any of the cardiac events occurred in truly symptomatic patients without any history whatsoever of any of these indicators. Dr. McPhail. Yes, they did. Dr. John L. Provan (Toronto, Ontario, Canada). This study confirms our recognition that the assessment of cardiac risk in these patients is one of the major problems facing vascular surgery today. We have looked prospectively at many of the criteria that have been mentioned by the authors. We found that the history was extremely useful, but the greatest sensitivity was achieved by the use of the left ventricular ejection fraction calculated by gated radionuclide angiography. If that was normal, no patient suffered any cardiac complication. If it was less than 30, all patients sustained some cardiac problem after operation. Unfortunately, most of these patients are unsuitable for myocardial revascularization because they usually have diffuse coronary artery disease. The problem then devolves on our anesthetists because the problems with these patients occur during aortic cross-clamping, and the intraoperative management becomes of crucial importance. We also looked at the effect of preoperative volume loading of these patients to obtain a Starling curve for left atrial function. This also gave us a reasonably good indication of the likelihood of a cardiac event after operation, but was not as good prognostically as the left ventricular ejection fraction. I noticed in the patients of McPhail et al. that they included some who underwent infrainguinal reconstruction, and I believe this may falsely modify their figures because the high-risk group comprises the patients undergoing aortic surgery. I have two questions. First, what was the risk of cardiac complications in patients undergoing infrainguinal reconstructions, and second, what do your anesthetists do for the intraoperative management of those patients with poor left ventricular fimction? Dr. McPhail. To answer your questions, the risk of infrainguinal revascularization surprisingly was not less than the risk of aortic reconstruction. As far as the anesthetists are concerned, we have had difficulty obtaining a core group of anesthetists who are specifically interested in the group of patients. We would very much like to achieve this and believe that it would improve our results considerably. Professor Andrew Nicolaides (London, England). We have faced similar problems with achieving target heart rate. We have discovered that if we exercise these patients on a bicycle, their heart rate goes much higher than if they are on a treadmill, presumably because the claudicants use different types of muscles. Probably the thigh muscles come into effect rather than the muscles that come into play on walking. We have found that on a bicycle, a third of the patients would achieve a target heart rate and have a positive test result, a third will achieve a target heart rate and have a negative exercise test, and a third will not achieve a target heart rate and will cause the test to be inconclusive because of the absence of ischemic electrocardiographic changes. We would like this third of patients to have an intravenous thallium-dipyridamole test. We also went a step further. We have used multiple electrodes on the chest and were able to identify whether patients have one-, two-, or three-vessel coronary disease using three criteria: ST segment depression, Q waves at rest, and the appearance of inverted U waves. In this way we find that we can identify the patients who have severe three-vessel coronary disease or left anterior descending

9 68 McPhail et al. Journal of VASCULAR SURGERY circumflex disease, which could mean left main stem, so we finally use coronary angiography in not more than 15% of our patients. I have three questions to ask: Have you considered bicycle ergometry? What do you think about intravenous dipyridamole-thallium tests? Finally, what is the feeling in the United States about multiple chest electrodes? Dr. McPhail (closing). To answer your questions, all of these were treadmill tests. We have used intravenous dipyridamole-thallium scanning in 77 patients already, as I mentioned, and arc using it more all of the time. All of these patients have a 12-lead electrocardiogram continuously throughout the test. In closing, I emphasizc again that this is a screening test to rule out the need for further investigation in patients who are at low risk. It is not a highly sensitive and specific indicator of severe coronary artery disease. THE E. J. WYLIE TRAVELING FELLOWSHIP OF THE EDUCATIONAL FOUNDATION OF THE SOCIETY FOR VASCULAR SURGERY The Educational Foundation of the Society for Vascular Surgery (with financial assistance from W. L. Gore & Associates, Inc.) has established an E. J. Wylie Traveling Fellowship. The purpose of the Fellowship is to enable young surgeons to visit centers of excellence in vascular surgery in the United States and abroad. The benefits of educational travel for the maintenance and enhancement of excellence in the practice Of vascular surgery are obvious. To be considered for selection a candidate must: 1. Be younger than 40 years of age at the time the traveling fellowship is awarded 2. Have completed a postgraduate vascular training program or have considerable experience in vascular surgery supplemental to general surgical training 3. Be committed to an academic career in vascular surgery and have obtained an academic appointment in a medical school or freestanding clinic devoted to excellence in medical education 4. Have a demonstrated record of success in pursuing clinical or basic science research sufficient to achieve academic excellence in his or her pursuit of a career in vascular surgery Selection will be made without regard to the candidate's geographic location. A candidate submitting documentation for consideration for selection must furnish an upto-date curriculum vitae and a list of publications, research projects, current research support, and a list of the centers that he or she wishes to visit. Three letters of recommendation are required, including one from the Division Head and another from the Chairman of the Department of Surgery of the institution in which the candidate holds a faculty appointment. A 500-word essay describing the objectives of the candidate's travel plans and linking these to his or her career goals must be appended. The first Selection Committee has been appointed and consists of John Bergan, Chairman, and members Henry Bahnson, Ben Eiseman, Robert Hobson, Ronald Stoney, and James Yao. The incumbent Recorder of the Society for Vascular Surgery will serve as an ex officio member of the Committee. The length of service of each of the members of the Committee and future members of the Committee will be determined by members of the Council of the Society for Vascular Surgery. The Travel Fellowship Award is $10,000, granted to one person for use during a time limit and for an itinerary to be arranged by agreement between the awardee and the Committee. Application for the Fellowship award shall be made in a letter containing the information and documents as detailed. The deadline for receiving applications is March 1, 188. Letter of nomination or intent should be directed to: Ronald J. Stoney, M.D., F.A.C.S. Chairman, E. J. Wylie Traveling Fellowship Committee Division of Vascular Surgery University of California Medical Center 505 Parnassus Ave., M-488 San Francisco, CA 4143