Postoperative 12-lead ECG predicts peri-operative myocardial ischaemia associated with myocardial cell damage

Transcription

1 Anaesthesia, 2004, 59, pages Postoperative 12-lead ECG predicts peri-operative myocardial ischaemia associated with myocardial cell damage B. W. Böttiger, 1,4 J. Motsch, 1 P. Teschendorf, 1 G. C. Rehmert, 2 R. Gust, 1 M. Zorn, 3 M. Schweizer, 3 E. L. Layug, 4 S. A. Snyder-Ramos, 1,4 D. T. Mangano 4 and E. Martin 1 1 Departments of Anaesthesiology and 3 Internal Medicine, University of Heidelberg, D Heidelberg, Germany 2 Department of Anaesthesiology and Intensive Medicine, Helios Clinics Schwerin, Schwerin, Germany 4 Ischemia Research and Education Foundation, San Francisco, CA, USA Summary Peri-operative myocardial ischaemia is the single most important risk factor for an adverse cardiac outcome after non-cardiac surgery. The present study examines whether intermittent 12-lead ECG recordings can be used as an early warning tool to identify patients suffering from perioperative myocardial ischaemia and subsequent myocardial cell damage. Fifty-five vascular surgery patients at risk for or with a history of coronary artery disease were monitored for peri-operative myocardial ischaemia using intermittent 12-lead ECG recordings taken pre-operatively and at 15 min, 20 h, 48 h, 72 h and 84 h postoperatively. The effectiveness of the 12-lead ECG was gauged by examining concordance with continuous 3-channel Holter monitoring and capturing peri-operative myocardial ischaemia by serial analyses of creatine kinase myocardial band isoenzyme and cardiac troponin T and I. The incidence of peri-operative myocardial ischaemia detected by 12-lead ECG was 44% and was identifiable in most patients (88%) 15 min after surgery. The incidence of peri-operative myocardial ischaemia detected by continuous monitoring was 53%, with the most severe episodes occurring intra-operatively and during emergence from anaesthesia. The concordance of the 12-lead method with continuous monitoring was 72%. The concordance of creatine kinase myocardial band isoenzyme activity with the 12-lead method was 71% and with Holter monitoring 57%. The concordance of mass concentration of creatine kinase myocardial band with 12-lead ECG recordings was 75%, and the corresponding value for Holter monitoring was 68%. The concordance of cardiac troponin T and I levels with the 12-lead method was 85% and 87%, respectively, and concordance with Holter monitoring was 72% and 66%, respectively. The postoperative 12-lead ECG identified peri-operative myocardial ischaemia associated with subsequent myocardial cell damage in most patients undergoing vascular surgery. Keywords Intraoperative complications; myocardial ischaemia. Biochemical markers; Troponin I, troponin T, creatine kinase. Electrocardiography, Holter.... Correspondence to: Bernd W. Böttiger bernd_boettiger@med.uni-heidelberg.de Accepted: 30 July 2004 Peri-operative morbidity and mortality due to cardiovascular disease is a major health-care challenge, occurring in more than 3 million of the 60 million patients who undergo non-cardiac surgery annually in the United States, Canada, Europe, Australia, New Zealand and Japan [1 4]. In the 9 million patients who have or are at risk for coronary artery disease, the most significant risks are peri-operative myocardial ischaemia and non-fatal myocardial infarction [2, 5]. Early postoperative myocardial ischaemia (i.e. within 48 h) is associated with a ninefold increase in the odds of suffering cardiac death, non-fatal myocardial infarction or unstable angina [5]. Methods that can detect peri-operative myocardial ischaemia are especially relevant because early treatment can mitigate or completely eliminate subsequent Ó 2004 Blackwell Publishing Ltd 1083

2 B. W. Böttiger et al. Æ Peri-operative myocardial ischaemia Anaesthesia, 2004, 59, pages devastating outcomes [6 10]. Recent outcome studies suggest that patients with increased cardiac troponins after non-cardiac surgery often suffer from myocardial infarction and other cardiac complications [11 16]. This suggests that peri-operative myocardial cell damage is occurring and may be responsible for later cardiac dysfunction. Ideally, perioperative myocardial ischaemia should be detected and treated before irreversible cardiac damage occurs [7 9]. The unfortunate reality is, however, that symptoms of myocardial cell damage linked to ischaemia and infarction are uncommon in the postoperative setting, where 95% of the episodes are silent [1, 4, 5, 10, 17, 18]. Accurate and reliable diagnostic procedures that are currently used to detect peri-operative myocardial ischaemia, such as continuous Holter monitoring and biochemical markers, are both labour intensive, costly and do not provide timely results [1, 19 22]. Thus, the use of simple, less expensive procedures that can detect peri-operative myocardial ischaemia shortly after its onset may improve outcome. The postoperative 12-lead ECG is sensitive and specific for the detection of peri-operative myocardial ischaemia. Therefore, the present study examined the concordance of 12-lead ECG with continuous Holter monitoring and its ability to capture peri-operative myocardial ischaemia sufficient to release serum biochemical markers. Methods Patients The study was approved by the Institutional Review Board. All of the patients received general anaesthesia for at least 1 h and either had a documented history of coronary artery disease or had two or more of the following risk factors: age over 65 years, treatment for arterial hypertension, diabetes mellitus, current cigarette smoking or cholesterol concentration over 240 mg.dl )1. Patients were excluded from the study if they had conduction defects that precluded electrocardiographic analysis of the ST segment, were in cardiogenic shock and had clinical signs of congestive heart failure or required chronic inotropic support, had an unstable pattern of angina or acute or recent myocardial infarction within 6 weeks of surgery or had uncontrolled hypertension. Peri-operative patient management Chronically administered anti-anginal medications including nitrates, b-adrenoreceptor blockers and calciumchannel blockers were given until the day of surgery. Anaesthesia was induced with fentanyl and thiopental sodium and was subsequently maintained with isoflurane and fentanyl. During the intra-operative period, systolic arterial blood pressure and heart rate were maintained within 20% of the pre-operative value (i.e. the mean of three measurements obtained within 48 h of surgery) by using prespecified changes of anaesthetic agents and cardiovascular drugs [23]. All patients were continuously monitored using 3-lead ECG for up to 96 h after surgery. If signs of perioperative myocardial ischaemia were observed on any ECG and or the patients reported chest pain, they were treated immediately. If peri-operative myocardial ischaemia persisted, a cardiologist was consulted to consider other interventions. Twelve-lead ECG criteria for peri-operative myocardial ischaemia and myocardial infarction Standard 12-lead ECG recordings were obtained preoperatively and at 15 min, 20 h, 48 h, 72 h and 84 h postoperatively. All ECGs were evaluated by an experienced cardiologist blinded to all other data. The R axis, PQ, QRS, QT and RR intervals were determined for each recording. A new negative T wave, ST segment depression or elevation of more than 0.2 mv in one or more leads was considered electrocardiographic evidence of myocardial ischaemia. For each recording, the cardiologist made a yes no decision concerning myocardial ischaemia. Peri-operative myocardial infarction was diagnosed by the presence of a new Q wave on a postoperative 12-lead ECG, according to the Minnesota Code criteria (Appendix A) [24]. Holter criteria for peri-operative myocardial ischaemia Three-channel continuous electrocardiographic AM Holter ECG recorders (bipolar leads CC 5,CM 5, ML, Marquette Electronics, Milwaukee, WI) were used for 8 h before induction of anaesthesia and until 96 h after surgery. Analysis of Holter data was performed using standard procedures [23]. ST segment deviation was assessed 60 ms after the J point in each of the three channels. The measurement period was shortened if this point occurred within the T wave to a minimum of 40 ms after the J point. Ischaemia was diagnosed by reversible horizontal or downsloping ST segment depression from baseline of 1 mm lasting at least 1 min and by ST segment elevation from baseline 2 mm, with episodes separated by at least 1 min. Each ischaemic episode was assessed for magnitude (maximum ST depression), duration and severity (area under the curve), as well as for ischaemic burden (i.e. minutes of ischaemia per minutes monitored). All electrocardiographic investigators were blinded to all other data. Holter data were analysed separately for different time periods: pre-operative, intra-operative emergence (defined as 30 min prior to leaving the operating room), postoperative 0 24 h, postoperative h, postoperative > 72 h Ó 2004 Blackwell Publishing Ltd

3 Anaesthesia, 2004, 59, pages B. W. Böttiger et al. Æ Peri-operative myocardial ischaemia Blood sampling and quantitative serum biochemical marker assays Blood for biochemical markers was drawn between 12 and 24 h pre-operatively and at 15 min, 4 h, 8 h, 12 h, 16 h, 20 h, 24 h, 32 h, 40 h, 48 h, 60 h, 72 h and 84 h postoperatively. Serum was frozen and stored as aliquots at )70 C until assays were performed in a blinded manner. Total creatine kinase activity and creatine kinase myocardial band isoenzymes activity were determined by immunoelectrophoresis (Electrophoresis Systems TM CK Isoenzyme system; Ciba Corning Diagnostics Corp., San Antonio, CA). The index of creatine kinase myocardial band isoenzymes activity (creatine kinase myocardial band isoenzymes activity total creatine kinase activity 100) was calculated for each measurement. The upper limit of the reference range for the total creatine kinase activity was 75 IU.l )1. The upper limit of the reference range of creatine kinase myocardial band isoenzymes activity was 5 IU.l )1 and 5% of total creatine kinase activity. In addition, the mass concentration of creatine kinase MB (i.e. the quantity of protein per millilitre of serum) was measured using an automated fluorescence immunoassay (Stratus CKMB Fluoreszenz Immunoassay TM ; Dade International Inc., Miami, FL). The cut-off value of mass concentration of creatine kinase MB was determined by the receiver operating characteristic (ROC) curve, which is a statistical method to obtain an optimal balance between sensitivity and specificity with regard to the detection of peri-operative myocardial ischaemia as assessed by the 12-lead or continuous monitoring. Measurements of cardiac troponin T were determined by enzyme-linked immunosorbent assay (ELISA troponin T TM ; Boehringer Mannheim, Mannheim, Germany). All serum levels 0.1 lg.l )1 were considered positive. Levels of cardiac troponin I were determined by automated fluorescence immunoassay (Stratus Cardiac Troponin-I Fluorometric Enzyme Immunoassay TM ; Dade International Inc.). The cut-off value for cardiac troponin I was determined by the ROC curve. Statistical analysis The association between myocardial cell damage as assessed by serum biochemical markers and electrical markers of peri-operative myocardial ischaemia was determined using sensitivity, specificity, concordance and ROC curve analyses. The concentrations of biochemical markers in patients with and without perioperative myocardial ischaemia were compared using the non-parametric Wilcoxon test, because cardiac troponin I data were not normally distributed. Holter data were assessed using the Kruskal Wallis test followed by the Wilcoxon test. The relationship between these data and cardiac troponin release was assessed using Spearman correlation analysis. A p-value of < 0.05 was considered significant (b = 0.1). Results Fifty-five patients were enrolled in the study after informed consent (Table 1). In 19 patients, postoperative antianginal treatment was initiated following bed-side diagnosis of peri-operative myocardial ischaemia. In two patients, peri-operative myocardial ischaemia (T-wave inversion) could not be resolved by anti-anginal drug intervention and both underwent acute coronary angiography, followed by percutaneous transluminal coronary angioplasty. In both patients, increased concentrations of serum biochemical markers were found within 4 8 h postoperatively. No other major cardiac adverse events were observed. Twelve-lead ECG and Holter monitoring ECG evidence of peri-operative myocardial ischaemia occurred in 24 (44%) patients; in 21 (88%) of these patients ischaemia occurred during the immediate postoperative phase (within 15 min) (Fig. 1). Seventy-five percent of patients with peri-operative myocardial ischaemia had ST segment depression with or without T-wave inversion and the remaining (25%) developed T-wave inversion only. Neither ST segment elevation nor new Q waves were observed on any postoperative 12-lead ECG. Peri-operative myocardial ischaemia was detected more than once in 13 patients and only once in 11 patients. Fifty-three patients had interpretable Holter recordings. Two patients with uninterpretable recordings had no signs of peri-operative myocardial ischaemia and myocardial cell damage as assessed by 12-lead ECG and cardiac troponins. Overall, peri-operative myocardial ischaemia was detected Table 1 Clinical characteristics of the 55 patients. Age; year (mean (SD)) 65 (9) Gender Female 11 (20) Male 44 (80) Medical history and inclusion criteria Documented coronary artery disease 25 (45) Two or more risk factors for 30 (55) coronary artery disease Age greater than 65 years 32 (58) Treated hypertension 39 (71) Treated diabetes mellitus 13 (24) Current smoking 18 (33) Cholesterol > 240 mg dl )1 33 (60) ASA physical status III 49 (89) Type of surgery Peripheral vascular surgery 32 (58) Carotid endarterectomy 23 (42) Data given in number (percentage) unless otherwise stated. ASA, American Society of Anesthesiologists. Ó 2004 Blackwell Publishing Ltd 1085

4 B. W. Böttiger et al. Æ Peri-operative myocardial ischaemia Anaesthesia, 2004, 59, pages (n) min Postoperative time (hr) Figure 1 Number of patients with peri-operative myocardial ischaemia as assessed by 12-lead electrocardiography (ECG) at different time points (the incidence of the first detection of peri-operative myocardial ischaemia in an individual patient is given in black). The overall incidence of 12-lead ECGdocumented peri-operative myocardial ischaemia was 44% (n = 24). In patients (88%), peri-operative myocardial ischaemia was detected in the immediate postoperative phase (15 min postoperatively). during 104 episodes in 29 patients (53%) (Table 2). Intra-operative peri-operative myocardial ischaemia was observed in 21 patients. Compared with pre-operative episodes, intra-operative episodes were more severe in duration, area under the curve and ischaemic burden. All but two episodes of peri-operative myocardial ischaemia persisted until arrival on the intensive care unit (ICU). In 13 (62%) of 21 patients with an intra-operative onset of perioperative myocardial ischaemia the ischaemic episode was diagnosed by intermittent 12-lead ECG recordings after arrival in the ICU. Furthermore, 12-lead ECG recordings at this time detected additional ischaemic episodes in eight (25%) of 32 patients who did not show signs of perioperative myocardial ischaemia on Holter recordings. Of the 29 patients with peri-operative myocardial ischaemia detected by Holter, 19 (66%) revealed signs of perioperative myocardial ischaemia in at least one 12-lead ECG. In the other 10, no peri-operative myocardial ischaemia could be detected by 12-lead ECG. Of the 24 patients without peri-operative myocardial ischaemia detected by the Holter, no signs of peri-operative myocardial ischaemia in 12-lead ECG recordings were detected in 19 patients (79%). Biochemical markers In nine of the 10 patients with peri-operative myocardial ischaemia detected by Holter but not by 12-lead ECG, cardiac troponin I and T were not increased. Of the five patients in whom peri-operative myocardial ischaemia was not detected by Holter but was detected by 12-lead ECG, cardiac troponins I and T increased in four patients, and three showed new T-wave inversion as the only sign of peri-operative myocardial ischaemia (a feature that is not used in the interpretation of Holter recordings). There was a moderate postoperative increase in total creatine kinase, which did not differ in patients with and those without peri-operative myocardial ischaemia. There was no increase in creatine kinase myocardial band isoenzyme activity concentrations in any patient preoperatively. Creatine kinase myocardial band isoenzyme activity concentrations were increased in at least one blood sample in 46% of patients with peri-operative myocardial ischaemia detected by 12-lead ECG and in 34% of patients detected by Holter (Tables 3 and 4). The concordance of creatine kinase myocardial band isoenzyme activity with the 12-lead method was 71% and with Holter monitoring 57% (Tables 3 and 4). The absolute concentrations of creatine kinase myocardial band isoenzyme activity in patients with and without peri-operative myocardial ischaemia as assessed by 12-lead ECG (Fig. 2) were significantly different at all time points between 8 and 32 h postoperatively. Using ROC curve analysis, the optimum threshold for mass concentration of creatine kinase MB was Period Incidence (onset) No. of episodes ST change; mm Duration; min AUC; mm.min Ischaemic burden; min.min )1 Table 2 Peri-operative myocardial ischaemia as assessed by Holter monitoring. Pre-operative ) Intra-operative ) ) non-emergence Intra-operative ) ) emergence Early postoperative ) (0 24 h) Late postoperative ) (24 72 h) Over 72 h ) Data are median values for episodes per period (ST change, duration and area under the curve (AUC) and patients per period (ischaemic burden)). p < 0.05 vs. pre-operative Ó 2004 Blackwell Publishing Ltd

5 Anaesthesia, 2004, 59, pages B. W. Böttiger et al. Æ Peri-operative myocardial ischaemia Table 3 Peri-operative myocardial ischaemia as assessed by 12-lead electrocardiogram (ECG), Holter monitoring and biochemical markers. Myocardial ischaemia 12-lead ECG (n = 24) Holter (n = 29) No myocardial ischaemia 12-lead ECG (n = 31) CKMBact CKMBact positive CKMBact negative CKMBmass CKMBmass positive CKMBmass negative ctnt ctnt positive ctnt negative ctni ctni positive ctni negative Holter (n = 24) CKMBact, creatine kinase MB isoenzyme activity. CKMBmass, mass concentration of creatine kinase MB; ctni, cardiac troponin I. ctnt, cardiac troponin T. 4.0 ng.ml )1 (Fig. 3). No patient had an increased mass concentration of creatine kinase MB pre-operatively. Overall, concentrations of mass concentration of creatine kinase MB were increased in at least one blood sample in 71% of patients with peri-operative myocardial ischaemia assessed by 12-lead ECG and in 62% of patients with perioperative myocardial ischaemia assessed by Holter (Tables 3 and 4). The concordance of mass concentration of creatine kinase MB with 12-lead ECG recordings was 75%, and the corresponding value for Holter monitoring was 68% (Tables 3 and 4). Mass concentration of creatine kinase MB in patients with and without peri-operarive myocardial ischaemia assessed by 12-lead ECG were significantly different at all time points between 4 and 48 h postoperatively (Fig. 2). The median peak mass concentration of creatine kinase MB in the group with peri-operative myocardial ischaemia in any 12-lead ECG was (range ) ng.ml )1. Preoperatively, cardiac troponin T was not detected in any patient. However, it was detected in at least one blood sample in 71% of patients with peri-operative myocardial ischaemia assessed by 12-lead ECG and in 55% of patients with peri-operative myocardial ischaemia assessed by Holter (Tables 3 and 4). The concordance of cardiac troponin T levels with the 12-lead method was 85% and its concordance with Holter monitoring was 72% (Tables 3 and 4). The concentrations of cardiac troponin T were significantly different in patients with vs. those without peri-operative myocardial ischaemia assessed by 12-lead ECG at all time points between 15 min and 84 h postoperatively (Fig. 2). The median individual peak concentrations of cardiac troponin T in the group with peri-operative myocardial ischaemia in any 12-lead ECG was (range ) lg.l )1. Using ROC curve analysis, a cardiac troponin I threshold of 1.6 ng.ml )1 was determined at the best combination of sensitivity and specificity (Fig. 3). Preoperatively, no patient had cardiac troponin I concentrations exceeding 1.6 ng.ml )1. Serum concentrations exceeding 1.6 ng.ml )1 were found in 79% of patients with perioperative myocardial ischaemia assessed by 12-lead ECG and in 55% of patients with peri-operative myocardial ischaemia assessed by Holter (Tables 3 and 4). The concordance between cardiac troponin I and 12-lead ECG recordings was 87%, and the corresponding value for Holter monitoring was 66% (Tables 3 and 4). The concentrations of cardiac troponin I in patients with and in those without peri-operative myocardial ischaemia assessed by 12-lead ECG were significantly different at all time points (Fig. 2). The median individual peak concentration of cardiac troponin I in the group with perioperative myocardial ischaemia in any 12-lead ECG was 3.5 (range ng.ml )1 ). Discussion The present data demonstrate that postoperative 12-lead ECG recordings are equally effective as continuous ECG Table 4 Relationship of 12-lead electrocardiogram (ECG) and Holter monitoring vs. biochemical markers. Sensitivity, specificity, positive and negative predictive values, and concordance of 12-lead ECG and Holter monitoring for the detection of peri-operative myocardial cell damage as assessed by creatine kinase isoenzyme MB activity (CKMBact), mass concentration of creatine kinase MB (CKMBmass), cardiac troponin T (ctnt) and cardiac troponin I (ctni). Sensitivity; % Specificity; % Positive predictive value; % Negative predictive value; % 12-lead ECG CKMBact CKMBmass ctnt ctni Holter monitoring CKMBact CKMBmass ctnt ctni Concordance; % Ó 2004 Blackwell Publishing Ltd 1087

6 B. W. Böttiger et al. Æ Peri-operative myocardial ischaemia Anaesthesia, 2004, 59, pages A CKMBact [IU.l -1 ] CKMBmass [ng.ml -1 ] B 0.6 ctnt [µg.l -1 ] ctni [ng.ml -1 ] 5 4 Figure 3 Receiver operating characteristics (ROC) curves for mass concentration of creatine kinase MB (A) revealed a cut-off value of 4.0 ng.ml )1 and ROC for cardiac troponin I (ctni; B) revealed a cut-off value of 1.6 ng.ml )1 yielding an optimal balance between sensitivity and specificity with regard to perioperative myocardial ischaemia as assessed by 12-lead ECG recordings (black lines) and Holter monitoring (grey lines), respectively. 3 Figure 2 Serum concentrations of creatine kinase isoenzyme MB activity (CKMBact), mass concentration of creatine kinase MB (CKMBmass), cardiac troponin T (ctnt) and cardiac troponin I (ctni) pre-operatively (pre) and during the postoperative study period in patients with peri-operative myocardial ischaemia as assessed by 12-lead ECG recordings (black bars, n = 24) and those without peri-operative myocardial ischaemia (white bars, n = 31). Data are given as means and standard deviations p < 0.05 vs. patients without peri-operative myocardial ischaemia. monitoring for detection of peri-operative myocardial ischaemia sufficient to cause release of cardiac markers in non-cardiac vascular surgery patients. Electrocardiography, and in particular continuous Holter monitoring, may represent a gold standard of ischaemia monitoring because peri-operative myocardial ischaemia as assessed by Holter has been shown to be an important surrogate for cardiac outcome [2, 5, 25 27]. In the present study, the Holter monitoring and serum biochemical markers demonstrated the effectiveness of intermittent 12-lead ECG recordings for detecting clinically significant peri-operative myocardial ischaemia. As with previous studies [5, 23], approximately half of our patients developed peri-operative myocardial ischaemia detected by Holter monitoring. Intermittent 12-lead ECG recording was more strongly associated with peri-operative myocardial ischaemia and myocardial cell damage than was continuous Holter monitoring. At least two possible explanations exist. First, 12-lead ECG recordings may have been more sensitive for specific regional changes because more leads were used and, possibly most importantly, specific changes in T wave were included for assessment of peri-operative myocardial ischaemia. T-wave inversion was the sole indicator for myocardial ischaemia in 25% of patients with perioperative myocardial ischaemia and both angioplasty patients; in these patients peri-operative myocardial 1088 Ó 2004 Blackwell Publishing Ltd

7 Anaesthesia, 2004, 59, pages B. W. Böttiger et al. Æ Peri-operative myocardial ischaemia ischaemia could not be detected using the Holter technique. Second, the specificity of 12-lead ECG recordings for the detection of prolonged episodes of peri-operative myocardial ischaemia may be greater, because continuous Holter monitoring due to its greater sensitivity also detects shorter and possibly less severe episodes. Shorter ischaemic episodes may not, however, cause myocardial cell damage and are far less important than prolonged episodes [28, 29]. In contrast to electrocardiography, serum biochemical markers are late markers of myocardial injury [19, 30 33]. Although the most severe episodes of peri-operative myocardial ischaemia were found intra-operatively and during emergence from anaesthesia, the increase in biochemical markers lasted a few hours postoperatively, This, however, is expected, as cardiac biochemical markers increase between 2 and 6 h after the onset of ischaemia [19, 31 33]. Any increase in cardiac troponins is thought to be associated with myocardial cell damage [30 32]. Besides their importance in the diagnosis of myocardial infarction [30 32], cardiac troponin T and I are known to be sensitive markers for myocardial ischaemia [32, 34]. In the present study, there were no patients with evidence of acute Q wave myocardial infarction. Nevertheless, we observed a postoperative increase of serum biochemical markers in most patients with peri-operative myocardial ischaemia as detected by both electrocardiographic methods. The present data suggest that most patients with peri-operative myocardial ischaemia develop myocardial cell damage and necrosis. Our findings agree with recent studies showing an increased concentration of cardiac troponins in patients at risk for coronary artery disease undergoing non-cardiac surgery but who did not meet the criteria for myocardial infarction [12 16]. Accurate and reliable diagnostic procedures for the detection of peri-operative myocardial ischaemia, such as continuous Holter monitoring and serum biochemical markers, are both labour intensive and costly. Moreover, these monitors do not provide timely results and therefore are not practical for routine clinical use [1, 19 22]. Thus, the use of simple, less expensive procedures like the 12-lead ECG that can detect peri-operative myocardial ischaemia shortly after its onset may improve outcome. For clinical practice, intermittent ECG monitoring 15 min after surgery is an important tool in identifying or predicting adverse cardiac outcome. However, the current study did not investigate whether serial 12-lead ECG is equally effective as continuous ECG monitoring in detecting peri-operative myocardial ischaemia in patients at high risk (e.g. identified by dobutamine stress echocardiogram). In conclusion, postoperative 12-lead ECG, particularly 15 min after surgery, is sensitive and specific for detection of peri-operative myocardial ischaemia associated with subsequent myocardial cell damage in patients undergoing vascular surgery. Its routine clinical use may improve the care and treatment of these patients. Acknowledgements B. W. Böttiger was awarded a grant from the Medical Faculty of the University of Heidelberg. The authors would like to thank the staff of the postoperative intensive care unit (Department of Anaesthesiology, University of Heidelberg) and all other colleagues involved in patient recruitment, treatment and care. References 1 Mangano DT. Perioperative cardiac morbidity. Anesthesiology 1990; 72: Mangano DT, Browner WS, Hollenberg M, Li J, Tateo IM. Long-term cardiac prognosis following noncardiac surgery. The Study of Perioperative Ischemia Research Group. Journal of the American Medical Association 1992; 268: Sprung J, Abdelmalak B, Gottlieb A, et al. Analysis of risk factors for myocardial infarction and cardiac mortality after major vascular surgery. Anesthesiology 2000; 93: McCann RL, Clements FM. Silent myocardial ischemia in patients undergoing peripheral vascular surgery: incidence an association with perioperative cardiac morbidity and mortality. Journal of Vascular Surgery 1989; 9: Mangano DT, Browner WS, Hollenberg M, London MJ, Tubau JF, Tateo IM. Association of perioperative myocardial ischemia with cardiac morbidity and mortality in men undergoing noncardiac surgery. The Study of Perioperative Ischemia Research Group. New England Journal of Medicine 1990; 323: Martinez EA, Kim LJ, Faraday N, et al. Sensitivity of routine intensive care unit surveillance for detection of myocardial ischemia. Critical Care Medicine 2003; 31: Mangano DT, Layug EL, Wallace A, Tateo I. Effect of atenolol on mortality and cardiovascular morbidity after noncardiac surgery. Multicenter Study of Perioperative Ischemia Research Group. New England Journal of Medicine 1996; 335: Pasternack PF, Grossi EA, Baumann FG, et al. Beta blockade to decrease silent myocardial ischemia during peripheral vascular surgery. American Journal of Surgery 1989; 158: Stone JG, Foex P, Sear JW, Johnson LL, Khambatta HJ, Triner L. Myocardial ischemia in untreated hypertensive patients: effect of a single small oral dose of a beta-adrenergic blocking agent. Anesthesiology 1988; 68: Badner NH, Knill RL, Brown JE, Novick TV, Gelb AW. Myocardial infarction after noncardiac surgery. Anesthesiology 1998; 88: Ó 2004 Blackwell Publishing Ltd 1089

8 B. W. Böttiger et al. Æ Peri-operative myocardial ischaemia Anaesthesia, 2004, 59, pages Landesberg G, Shatz V, Akopnik I, et al. Association of cardiac troponin, CK-MB, and postoperative myocardial ischemia with long-term survival after major vascular surgery. Journal of the American College of Cardiology 2003; 42: Neill F, Sear JW, French G, et al. Increases in serum concentrations of cardiac proteins and the prediction of early postoperative cardiovascular complications in noncardiac surgery patients. Anaesthesia 2000; 55: Rapp HJ, Rabethge S, Luiz T, Haux P. Perioperative ST-segment depression and troponin T release. Identification of patients with highest risk for myocardial damage. Acta Anaesthesiologica Scandinavica 1999; 43: Lee TH, Thomas EJ, Ludwig LE, et al. Troponin T as a marker for myocardial ischemia in patients undergoing major noncardiac surgery. American Journal of Cardiology 1996; 77: Lopez-Jimenez F, Goldman L, Sacks DB, et al. Prognostic value of cardiac troponin T after noncardiac surgery: 6-month follow-up data. Journal of the American College of Cardiology 1997; 29: Metzler H, Gries M, Rehak P, Lang T, Frühwald S, Toller W. Perioperative myocardial cell injury: the role of troponins. British Journal of Anaesthesia 1997; 78: Haagensen R, Steen PA. Perioperative myocardial infarction. British Journal of Anaesthesia 1988; 61: Ouyang P, Gerstenblith G, Furman WR, Golueke PJ, Gottlieb SO. Frequency and significance of early postoperative silent myocardial ischemia in patients having peripheral vascular surgery. American Journal of Cardiology 1989; 64: Gerhardt W, Katus H, Ravkilde J, et al. S-troponin T in suspected ischemic myocardial injury compared with mass and catalytic concentrations of S-creatine kinase isoenzyme MB. Clinical Chemistry 1991; 37: Eisenberg MJ, London MJ, Leung JM, et al. Monitoring for myocardial ischemia during noncardiac surgery. A technology assessment of transesophageal echocardiography and 12-lead electrocardiography. The Study of Perioperative Ischemia Research Group. Journal of the American Medical Association 1992; 268: Ellis JE, Shah MN, Briller JE, Roizen MF, Aronson S, Feinstein SB. A comparison of methods for the detection of myocardial ischemia during noncardiac surgery: automated ST-segment analysis systems, electrocardiography, and transesophageal echocardiography. Anesthesia and Analgesia 1992; 75: Urban MK, Gordon MA, Harris SN, O Connor T, Barash PG. Intraoperative hemodynamic changes are not good indicators of myocardial ischemia. Anesthesia and Analgesia 1993; 76: McSPI, Europe Research Group. Perioperative sympatholysis. Beneficial effects of the alpha 2-adrenoceptor agonist mivazerol on hemodynamic stability and myocardial ischemia. Anesthesiology 1997; 86: Blackburn H, Keys A, Simonson E, Rautaharju P, Punsar S. The electrocardiogram in population studies: a classification system. Circulation 1960; 21: Landesberg G, Mosseri M, Wolf Y, Vesselov Y, Weissman C. Perioperative myocardial ischemia and infarction: identification from continuous 12-lead electrocardiogram with online ST-segment monitoring. Anesthesiology 2002; 96: Landesberg G, Mosseri M, Zahger D, et al. Myocardial infarction after vascular surgery: the role of prolonged stressinduced, ST depression-type ischemia. Journal of the American College of Cardiology 2001; 37: Klootwijk P, Meij S, von Es GA, Muller EJ, Umans VA, Lenderink T, Simoons ML. Comparison of usefulness of computer assisted continuous 48 h ECG with 12-lead ECG ischaemia monitoring for detection and quantification of ischaemia in patients with unstable angina. European Heart Journal 1997; 18: Fleisher LA, Nelson AH, Rosenbaum SH. Postoperative myocardial ischemia: etiology of cardiac morbidity or manifestation of underlying disease? Journal of Clinical Anesthesia 1995; 7: Landesberg G, Luria MH, Cotev S, et al. Importance of longduration postoperative ST-segment depression in cardiac morbidity after vascular surgery. Lancet 1993; 341: Adams JE 3rd, Bodor GS, Davila-Roman VG, et al. Cardiac troponin I. A marker with high specificity for cardiac injury. Circulation 1993; 88: Adams JE 3rd, Sicard GA, Allen BT et al. Diagnosis of perioperative myocardial infarction with measurement of cardiac troponin I. New England Journal of Medicine 1994; 330: Katus HA, Remppis A, Neumann FJ, et al. Diagnostic efficiency of troponin T measurements in acute myocardial infarction. Circulation 1991; 83: Mair J, Larue C, Mair P, Balogh D, Calzolari C, Puschendorf B. Use of cardiac troponin I to diagnose perioperative myocardial infarction in coronary artery bypass grafting. Clinical Chemistry 1994; 40: Hamm CW, Ravkilde J, Gerhardt W, et al. The prognostic value of serum troponin T in unstable angina. New England Journal of Medicine 1992; 327: Appendix A Minnesota Code Criteria The Minnesota Code was developed in the late 1950s by Blackburn et al. [24] for better comparability and objective measure of electrocardiograms in clinical trials and epidemiological studies. This classification system is based on a defined set of measurement rules to assign specific numerical codes according to severity of ECG findings (e.g. new myocardial infarction, ischaemia, progression or regression of left ventricular hypertrophy, conduction defect). The visual or computer assigned standard Minnesota Code includes amplitude, axis and duration measurements, evolving bundle block, heart rate variability, QT dispersion, and QT or JT intervals Ó 2004 Blackwell Publishing Ltd