Beattie, W Scott MD, PhD, FRCPC; Abdelnaem, Esam MD; Wijeysundera, Duminda N. MD, FRCPC; Buckley, D Norman MD, FRCPC
Preoperative cardiac assessment is designed to identify patients at risk for major cardiovascular complications. Preoperative risk indices (1–3) perform better when coupled with the demonstration of ischemia on a stress test (4). Ergometric preoperative exercise stress testing is limited as a result of disease status (e.g., arthritis, claudication) and/or cardiac status. Pharmacologic tests were developed to avoid these problems. There have been 5 meta-analyses assessing preoperative pharmacologic stress testing (5–9). Two of these analyses are now outdated (5,6). A third meta-analysis was limited to stress testing before transplant surgery (7). A recent comparative meta-analysis evaluated only vascular patients (10) while excluding patients sent to angiography. In the final and eloquent meta-analysis of quantitative thallium imaging (TI), Etchells et al. (8) demonstrated 3 findings: first, large perfusion defects of more than 30% were found to increase risk, although this was only seen in 12% of patients. Second, smaller perfusion defects were not found to confer any added risk. Finally, a negative test did not significantly decrease the probability of a postoperative cardiac event. To restate these findings, the only stress result that changed the overall risk of a cardiac event was the finding of a moderate-to-large perfusion defect. A similar analysis of stress echocardiography (SE) is lacking. It is not clear if there are differences in the accuracy of these tests.
The need for accurate risk assessment of perioperative risk seems obvious. Therapies are not without adverse effects. Beta-blockade has been demonstrated to reduce perioperative cardiac events only in higher risk patients (10) but carries serious side effects (10). Lower risk patients may actually be harmed with therapy (11). Moreover, the Coronary Artery Revascularization Before Elective Major Vascular Surgery trial (CARP) (12) suggests that patients with moderate-to-large perfusion defects may benefit from revascularization. We are now left to determine how to identify those populations that may benefit from interventions. The purpose of this meta-analysis was to assess and compare the reliability of TI with SE in predicting postoperative cardiac events.
We designed our search to identify all studies assessing cardiac risk for any type of noncardiac surgery. Separate sensitivity analyses were planned a priori to assess vascular and nonvascular surgery.
We conducted 2 MEDLINE searches (one for TI and the second for SE). First, we searched the terms dipyridamole (exploded) (6319 citations), thallium (exploded) (3761 citations), sestamibi (exploded) (3303 citations), or adenosine (exploded) (3409 citations); these were combined with the OR function (12,973 citations). In a second step we then searched under postoperative complications (26,943 citations) and these 2 searches were combined using the AND function (439 citations) and limited to human studies (381 citations). The cardiac surgery citations were then eliminated, leaving 79 citations. Second, we searched dobutamine (exploded) (4268 citations), adenosine (exploded) (34,090 citations), dipyridamole (exploded) (6319 citations), and these were combined using the OR function for 43,530 citations. We then searched the term echocardiography (exploded) (60,840 citations). These searches were combined using the AND function, resulting in 2143 citations. We then combined this with the postoperative complication search using the AND function for 91 citations. After limiting this search to human studies and eliminating cardiac surgery, we identified 32 citations. The sum of the 2 separate searches left 111 citations. The search was last conducted on March 7, 2005. All 111 articles were then obtained in hard copy and read in full by 2 authors. There was no language restriction used. After identifying relevant studies each bibliography was hand-searched for further articles, then each article was searched in PubMed under “related articles.” To reduce the chance of duplicate inclusion of patients resulting from groups of authors having multiple publications, we recorded the dates of enrollment for each study; where there was overlap, only the study with the largest population was included in the analysis.
Data abstraction initially involved assessment of methodology noting prospective data collection, consecutive patient enrollment, routine screening for postoperative myocardial infarction (MI), selection of patients for coronary angiography, and the number of patients revascularized. Next, we recorded the number of patients with risk factors for coronary disease, history of revascularization, and the use of β-adrenergic antagonists. Third, we abstracted the stress test results: no defect, a fixed defect, a reversible defect, and, where possible, recorded the size of any defects. The results were then correlated to the number of MIs or death in each study. Patients who did not have surgery were excluded from the analysis. Sensitivity and specificity were calculated for each study. The likelihood ratio (LR) was chosen a priori as our primary outcome measure (13). The LR is calculated as follows: LR = Sensitivity/(1 − Specificity).
This equation uses the same parameters describing a receiver operating curve. The Receiver Operating Characteristic (ROC) curve is composed of the LR at differing positive levels of the stress test. A ROC curve provides a common scale for comparing tests even though they measure different parameters and use different units. The prevalence of the target disorder has little influence on the ROC. Finally, the product of pretest probability and LR yields posttest probability of the target disorder. The ROC curve and 95% confidence interval (CI) were calculated for each quantitative study using SAS software (SAS, Cary, NC). The quantitative data included the number of negative tests, the number of tests with baseline abnormalities, a positive test, and the number of tests with moderate-to-large defects and/or 2 or more areas of perfusion defect (13). Two authors (WSB and EA) abstracted the data set. Discrepancies were resolved by discussion and consensus.
The standard principles of meta-analysis were applied to combine studies. In this review we focused on the differences between imaging techniques; we did not subdivide the analysis to consider the stressor (dobutamine, dipyridamole, atropine) separately. A summary ROC (SROC) was calculated for TI and SE, respectively (14,15). All data were entered in REVMAN 4.2 (Cochrane Collaboration), ROC curves from the quantitative studies were combined meta-analytically using the random-effects model (16). Sensitivity analysis was planned a priori for the effect of study quality and in patients having vascular procedures. Heterogeneity, defined as the variation among the results of individual trials beyond that expected by chance, was evaluated using the I2 test (17).
The results of this meta-analysis represent data from 68 studies in 10,278 patients. Forty-five studies were excluded for the following reasons: 31 publications were clinical management papers lacking new data, 10 were re-publication of previously published series, and in four cases we could not determine a sensitivity or specificity. The PubMed “Related Article Search” and bibliographic search found 2 new studies. There were 25 studies assessing SE (18–42 and 50 reports (31,34,35,37,41–46,46–87) of TI. Seven studies were direct comparisons examining both SE and TI (18,30,31,34,35,37,42). Characteristics of the studies are seen in Table 1. The following aspects of the studies were similar: average number of patients, number of vascular surgeries, number of retrospective studies, and number of blinded studies. Routine screening for MI was more frequent in SE studies (48% versus 21% P = 0.008) whereas results of the test were used to refer to angiography more often in TI studies (71% versus 46%; P = 0.027).
The populations are compared in Table 2. The numbers of patients with cardiac disease, diabetes, or history of congestive failure were similar. This analysis found that patients evaluated with SE had more previous coronary revascularization and that twice as many patients were using β-adrenergic antagonists, but these differences did not achieve statistical significance.
Comparisons of the outcome data for each stress test are seen in Table 3. The unadjusted MI or death rates were 8.1% for TI and 7.5% for SE. Figure 1 illustrates the SROC and shows the LR of each study in this meta-analysis. A positive SE results in a LR that is 2 times more predictive than a positive TI. SE was found to be superior after the planned sensitivity analysis (Table 3). In post hoc sensitivity analyses we excluded all studies conducted on or before 1995, matched the studies by the year and again the difference was maintained. We also performed a regression analysis of the year of study against both the sensitivity and LR (see online appendix at www.anesthesia-analgesia.org). In the 7 studies, where direct comparisons between SE and TI were made, the difference in LR was not statistically different. However, in this sensitivity analysis the LR for the TI studies was not statistically significant i.e., did not improve the prediction of postoperative MI or death.
There was significant heterogeneity in all these analyses. In an attempt to explain the heterogeneity, 2 separate post hoc analyses were conducted. Heterogeneity was not explained on the basis of the stressor (dobutamine, dipyridamole, or atropine) or the year the study was conducted. We used the more conservative random effects model in all calculations of summary LR.
There were 20 studies (9 SE and 13 TI with 2 common studies) where a ROC curve could be constructed. The cumulative ROC for SE was 0.80 (95% CI, 0.76–0.84) and 0.75 (95% CI, 0.70–0.80; not significant) for TI (Fig. 2). In a further analysis of quantitative SE the LR of a negative SE was a better predictor of an uneventful operation than TI (LR, 0.23; 95% CI, 0.17-0.32 versus 0.44; 95% CI, 0.36–0.54; P < 0.02). The finding of a moderate-to-large abnormality occurred in 16.2% of SE tests and 14.6% of TI tests. These findings were associated with a LR 8.35 (95% CI, 5.6–12.45) with no difference between the two techniques. This evaluation of cumulative ROC again revealed mild heterogeneity. The heterogeneity was no longer significant when we eliminated the 2 studies with ROC more than 0.96 (I2 = 0%).
The results of stress testing were used, in 37 studies, to refer patients for coronary angiography. The rate of referral to angiography was more than 2 times more frequent in patients screened with TI. The percentage of patients who were revascularized was the same whether they were screened with TI or SE.
This meta-analysis shows SE has better predictive powers compared to TI. The LR of a positive SE was twice that of positive TI. A negative SE reduces the probability of MI or death. Sensitivity analysis confirmed these findings. There were fewer false negative SE results after analysis of vascular patients, including only higher quality studies, and after eliminating all studies conducted before 1996. We also found fewer false negative SEs in the quantitative studies where ROC was calculated. The second major finding of this study revealed, in quantitative screening tests, that a moderate-to-large defect, detected by either SE or TI, is highly predictive of postoperative events.
A negative TI does not reliably reduce the probability of a postoperative cardiac event. This meta-analysis has found that more than one third of the cardiac events occurred in patients with a negative test. This finding has important clinical implications because careful patient selection is required to make safe interventions. Medical therapy is not without significant side effects. Beta-adrenergic antagonists started de novo after cardiac surgery increase length of hospital stay (88). In noncardiac surgery, β-adrenergic blockade increases the need to treat hypotension and bradycardia and may increase the incidence of heart failure (10). In lower risk patients β-adrenergic antagonists may cause harm (11). A finding of a moderate-to-larger defect also has important implications.
The recently completed CARP trial has been used by some to advocate for no preoperative testing. We disagree with this assessment of the results (89). First, many of the patients in the trial did not conform to the current America Heart Association/American College of Cardiology guidelines for testing. Second, the trial did not have the power to assess in-hospital cardiac events. There was a trend toward a 20%–25% reduction in MI. Third, in subgroup analysis, the patients who would most benefit from testing (moderate-to-high risk Eagle criteria and Revised Cardiac Risk Index patients) seemed to benefit from revascularization if a moderate or large defect was demonstrated. Regrettably, this landmark trial was underpowered to demonstrate an effect in this subgroup of patients. The present meta-analysis shows that patients with moderate-to-large defects, by either test, have an almost ninefold increase in the risk of MI or death after noncardiac surgery. In a surgical population with a perioperative MI incidence of 5%, the finding of a moderate perfusion defect carries a 40%–50% chance of a perioperative MI. Based on the CARP trial and this meta-analysis, any patient in our institution with moderate-to-large perfusion defect continues to be referred for coronary angiography. The false negative rate we describe suggests that fewer patients will be missed if SE is used. The results of this meta-analysis extend the findings of Etchells et al. (8) by showing that moderate or multiple defects on SE are at least as accurate as the demonstration of a large perfusion defect on TI. A negative test does not reliably confirm less risk of perioperative cardiac event, although a positive SE is 2 times more predictive than a positive TI. We continue to support the contention that an accurate and quantitative ischemic assessment is required in all moderate-to-high risk patients.
There are important differences in the way these studies were conducted. Postoperative MI is often clinically silent, and many postoperative MIs go undetected. Our finding that routine screening protocols exist more frequently in SE studies would have been expected to inflate the sensitivity measurement of SE. We have attempted to control for this effect by using a sensitivity analysis entering only those trials in either group that used routine screening and blinding. The sensitivity analysis showed that the comparative LRs were not changed and the statistical significance of the differences between the groups was maintained. The sensitivity and specificity of SE and TI may have been artificially changed by postoperative care. It is natural for physicians to alter the care of patients with positive preoperative screening tests to try to diminish postoperative morbidity. Knowledge of increased risk may increase monitoring. A variety of measures are thought to decrease perioperative MI, including β-blockade (90,91), administration of α-adrenergic agonists (92), calcium channel blockers (93), and thoracic epidural analgesia (94). We contend that therapeutic interventions do not explain the results of this meta-analysis. First, in studies that supply these data, we could not detect a difference in perioperative treatment regimens. Second, the rates of preoperative revascularization are similar. Finally, the crude MI and death rates are not different when the 2 study groups are compared.
We have chosen the LR and ROC as our primary outcome measurements. These measures have been used in the last 2 meta-analyses on this subject (8,9). Odds ratios and relative risk have been criticized and can be inaccurate for classifying or predicting risk (13). The LR incorporates elements of both the sensitivity and specificity and expresses the odds that a given level of test result (a perfusion defect or regional wall motion abnormality) would be expected in the target disorder, in this case the odds of a postoperative MI or death.
The shortcomings of meta-analysis are well recognized and we have documented them (92–94). These shortcomings are even more pronounced when evaluating diagnostic testing because a variety of definitions may be used as end-points. We have tried to control for this in our analysis by using a standardized end-point from which to calculate sensitivity and specificity. Several other deficiencies must be addressed.
First, studies in this meta-analysis were conducted over more than 2 decades. In a post hoc sensitivity analysis we could not demonstrate a difference based on the date of study or a change in accuracy over the span of these studies (see online appendix at www.anesthesia-analgesia.org). Second, as was noted by others (8), the general quality of the publications is poor. The sensitivity analyses show that the differential between SE and TI is largest in the studies where blinding was incorporated. The analysis is also limited in that there has been no control of the planned interventions after a positive test. Third, it is possible that the underlying patient characteristics are different. We tried to control for this by evaluating the baseline characteristic where possible, but little information was actually given. The overall morbidities among studies were not different nor was there a difference in the percentage of patients who were revascularized as a result of these investigations. Fourth, we noted a large amount of heterogeneity. Accordingly we used the conservative random effects model. Our attempt to explain the heterogeneity was not successful. We note that there was little heterogeneity in the analysis of quantitative studies. The results of the quantitative studies mirror the finding of the whole study. Finally, the method used to combine the ROC curves uses the inverse of the variance to weight studies. In this case, more weighting is given to studies with an ROC approaching 1. The analysis found 2 TI studies with an ROC of 0.5 and 2 studies with values at 0.6, whereas the lowest SE ROC was 0.73. These low values are underweighted, and this meta-analysis may have minimized the difference between SE and TI studies. A strength of this analysis is the number of studies and the number of subjects evaluated. Funnel plots do not suggest a publication bias. Furthermore, it is unlikely that any missed studies, should they exist, would influence the major findings of this study. Systematic reviews and meta-analysis are best for hypothesis generation but not for testing. To demonstrate a superior negative predictive test, in a randomized controlled trial based on an event rate of 6% and a 25% reduction, more than 10,000 patients would be required. This would necessitate incorporation of a standardized management algorithm for positive study results and a routine screening protocol for perioperative events.
In conclusion, in this meta-analysis we adjusted for the known problems in combining diagnostic tests and used the identical diagnostic criteria for each study. The preoperative risks of both SE and TI appear to be similar. We used several sensitivity analyses; the results show that SE has superior negative predictive ability. Second, moderately large defects, detected by either method, are highly predictive of subsequent postoperative cardiac events. In considering the clinical utility of our analysis, we would suggest that a negative TI should result in little change in perioperative management. All patients with a positive test should be considered at increased risk for an event and managed with maximal medical therapy. Patients with moderately large defects should be referred for angiography. SE, as a screening tool in patients with suspected cardiac disease before noncardiac surgery, has many positive features, including better negative predictive power, and we conclude that it is superior to TI in predicting postoperative cardiac events.
1. Detsky AS, Abrams HB, Forbath N, et al. Cardiac assessment for patients undergoing noncardiac surgery: a multifactorial clinical risk index. Arch Intern Med 1986;146:2131–4.
2. Lee TH, Marcantonio ER, Mangione CM, et al. Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation 1999;100:1043–9.
3. Goldman L, Caldera DL, Nussbaum SR, et al. Multifactorial index of cardiac risk in noncardiac surgical procedures. N Engl J Med 1977;297:845–50.
4. Eagle KA, Coley CM, Newell JB, et al. Combining clinical and thallium data optimizes preoperative assessment of cardiac risk before major vascular surgery. Ann Intern Med 1989;110:859–66.
5. Mantha S, Roizen MF, Barnard J, et al. Relative effectiveness of four preoperative tests for predicting adverse cardiac outcomes after vascular surgery: a meta-analysis. Anesth Analg 1994;79:422–33.
6. Shaw LJ, Eagle KA, Gersh BJ, Miller DD. Meta-analysis of intravenous dipyridamole-thallium-201 imaging (1985 to 1994) and dobutamine echocardiography (1991 to 1994) for risk stratification before vascular surgery. J Am Coll Cardiol 1996;27:787–98.
7. Rabbat CG, Treleaven DJ, Russell JD, et al. Prognostic value of myocardial perfusion studies in patients with end-stage renal disease assessed for kidney or kidney-pancreas transplantation: a meta-analysis. J Am Soc Nephrol 2003;14:431–9.
8. Etchells E, Meade M, Tomlinson G, Cook D. Semiquantitative dipyridamole myocardial stress perfusion imaging for cardiac risk assessment before noncardiac vascular surgery: a meta-analysis. J Vasc Surg 2002;36:534–40.
9. Kertai MD, Boersma E, Bax JJ, et al. A meta-analysis comparing the prognostic accuracy of six diagnostic tests for predicting perioperative cardiac risk in patients undergoing major vascular surgery. Heart 2003;89:1327–34.
10. Devereaux PJ, Beattie WS, Choi P, et al. How strong is the evidence for the use of perioperative beta-blockers in non-cardiac surgery? Systematic review and meta-analysis of randomized controlled trials. BMJ 2005;1136:313–21.
11. Lindenauer PK, Pekow P, Wang K, et al. Perioperative beta-blocker therapy and mortality after major noncardiac surgery. N Engl J Med 2005;353:349–61.
12. McFalls EO, Ward HB, Moritz TE, et al. Coronary artery revascularization before elective major vascular surgery. The CARP Trial. N Engl J Med 2005;351:2795–804.
13. Pepe MS, Janes H, Longton G, et al. Limitations in odds ratio in gauging the performance of a diagnostic, prognostic or screening marker. Am J Epidemiol 2000;159:882–90.
14. Irwig L, Macaskill P, Glasziou P, Fahey M. Meta-analytic methods for diagnostic test accuracy. J Clin Epidemiol 1995;48:119–30.
15. Walter SD, Irwig L, Glasziou PP. Meta-analysis of diagnostic tests with imperfect reference standards. J Clin Epidemiol 1999;52:943–51.
16. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88.
17. Higgins J, Thompson SG. Quantifying heterogeneity in meta-analysis. Stat Med 2002;21:1539–58.
18. Van Damme H, Pierard L, Gillain D, et al. Cardiac risk assessment before vascular surgery: a prospective study comparing clinical evaluation, dobutamine stress echocardiography, and dobutamine Tc-99m sestamibi tomoscintigraphy. Cardiovasc Surg 1997;5:54–64.
19. Davila-Roman VG, Waggoner AD, Sicard GA, et al. Dobutamine stress echocardiography predicts surgical outcome in patients with an aortic aneurysm and peripheral vascular disease. J Am Coll Cardiol 1993;21:957–63.
20. Ballal RS, Kapadia S, Secknus MA, et al. Prognosis of patients with vascular disease after clinical evaluation and dobutamine stress echocardiography. Am Heart J 1999;137:469–75.
21. Tischler MD, Lee TH, Hirsch AT, et al. Prediction of major cardiac events after peripheral vascular surgery using dipyridamole echocardiography. Am J Cardiol 1991;68:593–7.
22. Plotkin JS, Johnson LB, Rustgi VK, et al. Dobutamine stress echocardiography for orthotopic liver transplant evaluation. Transplantation 2001;71:818.
23. Bossone E, Martinez FJ, Whyte RI, et al. Dobutamine stress echocardiography for the preoperative evaluation of patients undergoing lung volume reduction surgery. J Thorac Cardiovasc Surg 1999;118:542–6.
24. Torres MR, Short L, Baglin T, et al. Usefulness of clinical risk markers and ischemic threshold to stratify risk in patients undergoing major noncardiac surgery. Am J Cardiol 2002;90:238–42.
25. Langan EM, Youkey JR, Franklin DP, et al. Dobutamine stress echocardiography for cardiac risk assessment before aortic surgery. J Vasc Surg 1993;18:905–11.
26. Day SM, Younger JG, Karavite D, et al. Usefulness of hypotension during dobutamine echocardiography in predicting perioperative cardiac events. Am J Cardiol 2000;85:478–83.
27. Das MK, Pellikka PA, Mahoney DW, et al. Assessment of cardiac risk before nonvascular surgery: dobutamine stress echocardiography in 530 patients. J Am Coll Cardiol 2000;35:1647–53.
28. Lalka SG, Sawada SG, Dalsing MC, et al. Dobutamine stress echocardiography as a predictor of cardiac events associated with aortic surgery. J Vasc Surg 1992;15:831–40.
29. Sicari R, Ripoli A, Picano E, et al. Perioperative prognostic value of dipyridamole echocardiography in vascular surgery: a large-scale multicenter study in 509 patients. EPIC (Echo Persantine International Cooperative) Study Group. Circulation 1999;100:II269–74.
30. Pasquet A, D’hondt AM, Verhelst R, et al. Comparison of dipyridamole stress echocardiography and perfusion scintigraphy for cardiac risk stratification in vascular surgery patients. Am J Cardiol 1998;82:1468–74.
31. Mocini D, Uguccioni M, Galli C, et al. Dipyridamole echocardiography and 99mTc-MIBI spect dipyridamole scintigraphy for cardiac evaluation prior to peripheral vascular surgery. Minerva Cardioangiol 1995;43:185–90.
32. Rossi E, Citterio F, Vescio MF, et al. Risk stratification of patients undergoing peripheral vascular revascularization by combined resting and dipyridamole echocardiography. Am J Cardiol 1998;82:306–10.
33. Shafritz R, Ciocca RG, Gosin JS, et al. The utility of dobutamine echocardiography in preoperative evaluation for elective aortic surgery. Am J Surg 1997;174:121–5.
34. Mondillo S, Ballo P, Agricola E, et al. Noninvasive tests for risk stratification in major vascular surgery. Vasa 2002;31:195–201.
35. Kontos MC, Akosah KO, Brath LK, et al. Cardiac complications in noncardiac surgery: value of dobutamine stress echocardiography versus dipyridamole thallium imaging. J Cardiothorac Vasc Anesth 1996;10:329–35.
36. Zamorano J, Duque A, Baquero M, et al. Stress echocardiography in the pre-operative evaluation of patients undergoing major vascular surgery: are results comparable with dipiridamole versus dobutamine stress echo? Rev Esp Cardiol 2002;55:121–6.
37. Lacroix H, Herregod MC, Ector H, et al. The value of dipyridamole thallium scintigraphy and dobutamine stress echocardiography as predictors of cardiac complications following reconstruction of the abdominal aorta. Int Angiol 2000;19:231–6.
38. Eichelberger JP, Schwarz KQ, Black ER, et al. Predictive value of dobutamine echocardiography just before noncardiac vascular surgery. Am J Cardiol 1993;72:602–7.
39. Poldermans D, Arnese M, Fioretti PM, et al. Sustained prognostic value of dobutamine stress echocardiography for late cardiac events after major noncardiac vascular surgery. Circulation 1997;95:53–8.
40. Boersma E, Poldermans D, Bax JJ, et al. Predictors of cardiac events after major vascular surgery: role of clinical characteristics, dobutamine echocardiography, and beta-blocker therapy. JAMA 2001;285:1865–73.
41. Lane RT, Sawada SG, Segar DS, et al. Dobutamine stress echocardiography for assessment of cardiac risk before noncardiac surgery. Am J Cardiol 1991;68:976–7.
42. Lin K, Stewart D, Cooper S, Davis CL. Pre-transplant cardiac testing for kidney-pancreas transplant candidates and association with cardiac outcomes. Clin Transplant 2001;15:269–75.
43. Boucher CA, Brewster DC, Darling RC, et al. Determination of cardiac risk by dipyridamole-thallium imaging before peripheral vascular surgery. N Engl J Med 1985;312:389–94.
44. Levinson JR, Boucher CA, Coley CM, et al. Usefulness of semiquantitative analysis of dipyridamole-thallium-201 redistribution for improving risk stratification before vascular surgery. Am J Cardiol 1990;66:406–10.
45. Marshall ES, Raichlen JS, Forman S, et al. Adenosine radionuclide perfusion imaging in the preoperative evaluation of patients undergoing peripheral vascular surgery. Am J Cardiol 1995;76:817–21.
46. Younis LT, Aguirre F, Byers S, et al. Perioperative and long-term prognostic value of intravenous dipyridamole thallium scintigraphy in patients with peripheral vascular disease. Am Heart J 1990;119:1287–92.
47. Coley CM, Field TS, Abraham SA, et al. Usefulness of dipyridamole-thallium scanning for preoperative evaluation of cardiac risk for nonvascular surgery. Am J Cardiol 1992;69:1280–5.
48. Marwick TH, Shan K, Go RT, et al. Use of positron emission tomography for prediction of perioperative and late cardiac events before vascular surgery. Am Heart J 1995;130:1196–202.
49. Shaw L, Miller DD, Kong BA, et al. Determination of perioperative cardiac risk by adenosine thallium-201 myocardial imaging. Am Heart J 1992;124:861–9.
50. Kresowik TF, Bower TR, Garner SA, et al. Dipyridamole thallium imaging in patients being considered for vascular procedures. Arch Surg 1993;128:299–302.
51. McEnroe CS, O’Donnel TF, Yeager A, et al. Comparison of ejection fraction and Goldman risk factor analysis to dipyridamole-thallium 201 studies in the evaluation of cardiac morbidity after aortic aneurysm surgery. J Vasc Surg 1990;11:497–504.
52. Hendel RC, Chen MH, L’Italien GJ, et al. Sex differences in perioperative and long-term cardiac event-free survival in vascular surgery patients: an analysis of clinical and scintigraphic variables. Circulation 1995;91:1044–51.
53. Huang Z, Komori S, Sawanobori T, et al. Dipyridamole thallium-201 single-photon emission computed tomography for prediction of perioperative cardiac events in patients with arteriosclerosis obliterans undergoing vascular surgery. Jpn Circ J 1998;62:274–8.
54. Baron JF, Mundler O, Bertrand M, et al. Dipyridamole-thallium scintigraphy and gated radionuclide angiography to assess cardiac risk before abdominal aortic surgery. N Engl J Med 1994;330:663–9.
55. Fletcher JP, Antico VF, Gruenewald S, Kershaw LZ. Dipyridamole-thallium scan for screening of coronary artery disease prior to vascular surgery. J Cardiovasc Surg (Torino) 1988;29:666–9.
56. Kontos MC, Brath LK, Akosah KO, Mohanty PK. Cardiac complications in noncardiac surgery: relative value of resting two-dimensional echocardiography and dipyridamole thallium imaging. Am Heart J 1996;132:559–66.
57. Ombrellaro MP, Dieter RA, Freeman M, et al. Role of dipyridamole myocardial scintigraphy in carotid artery surgery. J Am Coll Surg 1995;181:451–8.
58. Seeger JM, Rosenthal GR, Self SB, et al. Does routine stress-thallium cardiac scanning reduce postoperative cardiac complications? Ann Surg 1994;219:654–61.
59. Stratmann HG, Younis LT, Wittry MD, et al. Dipyridamole technetium-99m sestamibi myocardial tomography in patients evaluated for elective vascular surgery: prognostic value for perioperative and late cardiac events. Am Heart J 1996;131:923–9.
60. Nguyen TT, Amsterdam EA, Schaefer S. Risk stratification prior to vascular surgery: does the location of a dipyridamole thallium scintigram defect provide prognostic information? Cardiology 1997;88:569–75.
61. Strawn DJ, Guernsey JM. Dipyridamole thallium scanning in the evaluation of coronary artery disease in elective abdominal aortic surgery. Arch Surg 1991;126:880–4.
62. Vandenberg BF, Rossen JD, Grover-McKay M, et al. Evaluation of diabetic patients for renal and pancreas transplantation: noninvasive screening for coronary artery disease using radionuclide methods. Transplantation 1996;62:1230–5.
63. Sachs RN, Tellier P, Larmignat P, et al. Assessment by dipyridamole-thallium-201 myocardial scintigraphy of coronary risk before peripheral vascular surgery. Surgery 1988;103:584–7.
64. Zarich SW, Cohen MC, Lane SE, et al. Routine perioperative dipyridamole 201Tl imaging in diabetic patients undergoing vascular surgery. Diabetes Care 1996;19:355–60.
65. Bry JD, Belkin M, O’Donnell TF, et al. An assessment of the positive predictive value and cost-effectiveness of dipyridamole myocardial scintigraphy in patients undergoing vascular surgery. J Vasc Surg 1994;19:112–21.
66. Mangano DT, London MJ, Tubau JF, et al. Dipyridamole thallium-201 scintigraphy as a preoperative screening test; a reexamination of its predictive potential. Study of Perioperative Ischemia Research Group. Circulation 1991;84:493–502.
67. DeVirgillio C, Pak S, Arnell T, et al. Cardiac assessment prior to vascular surgery: is dipyridamole-sestamibi necessary? Ann Vasc Surg 1996;10:325–9.
68. DeVirgillio C, Toosie K, Elbassir M, et al. Dipyridamole-thallium/sestamibi before vascular surgery: a prospective blinded study in moderate-risk patients. J Vasc Surg 2000;32:77–89.
69. Vanzetto G, Machecourt J, Blendea D, et al. Additive value of thallium single-photon emission computed tomography myocardial imaging for prediction of perioperative events in clinically selected high cardiac risk patients having abdominal aortic surgery. Am J Cardiol 1996;77:143–8.
70. Vanzetto G, Sessa C, Magne JL, et al. Evaluation of a clinical and scintigraphic management strategy for cardiac risk before abdominal aorta surgery. Apropos of 982 surgical patients [in French]. Arch Mal Coeur Vaiss 1999;92:211–8.
71. Vaquette B, Clergues F, Kalangos A, et al. Prognostic value of thallium 201 myocardial scintigraphy with dipyridamole before peripheral arterial surgery [in French]. Arch Mal Coeur Vaiss 2003;96:281–7.
72. Watters TA, Botvinick EH, Dae MW, et al. Comparison of the findings on preoperative dipyridamole perfusion scintigraphy and intraoperative transesophageal echocardiography: implications regarding the identification of myocardium at ischemic risk. J Am Coll Cardiol 1991;18:93–100.
73. Erickson CA, Carballo RE, Freischlag JA, et al. Using dipyridamole-thallium imaging to reduce cardiac risk in aortic reconstruction. J Surg Res 1996;60:422–8.
74. Roghi A, Palmieri B, Crivellaro W, et al. Preoperative assessment of cardiac risk in noncardiac major vascular surgery. Am J Cardiol 1999;83:169–74.
75. Hashimoto J, Suzuki T, Nakahara T, et al. Preoperative risk stratification using stress myocardial perfusion scintigraphy with electrocardiographic gating. J Nucl Med 2003;44:385–90.
76. Chen T, Kuwabara Y, Tsutsui H, et al. The usefulness of dipyridamole thallium-201 single photon emission computed tomography for predicting perioperative cardiac events in patients undergoing non-cardiac vascular surgery. Ann Nucl Med 2002;16:45–53.
77. Antalffy J, Bajnok L, Kozlovszky B, et al. Estimation of perioperative cardiac risk by means of dipyridamole myocardial scintigraphy in patients undergoing vascular surgery on the lower limbs [in Hungarian]. Orv Hetil 1995;136:703–7.
78. Fleisher LA, Nelson AH, Rosenbaum SH. Failure of negative dipyridamole thallium scans to predict perioperative myocardial ischaemia and infarction. Can J Anaesth 1992;39:179–83.
79. Cutler BS, Leppo JA. Dipyridamole thallium 201 scintigraphy to detect coronary artery disease before abdominal aortic surgery. J Vasc Surg 1987;5:91–100.
80. Patel AD, bo-Aud WS, Davis JM, et al. Prognostic value of myocardial perfusion imaging in predicting outcome after renal transplantation. Am J Cardiol 2003;92:146–51.
81. Mistry BM, Bastani B, Solomon H, et al. Prognostic value of dipyridamole thallium-201 screening to minimize perioperative cardiac complications in diabetics undergoing kidney or kidney-pancreas transplantation. Clin Transplant 1998;12:130–5.
82. Iqbal A, Gibbons RJ, McGoon MD, et al. Noninvasive assessment of cardiac risk in insulin-dependent diabetic patients being evaluated for pancreatic transplantation using thallium-201 myocardial perfusion scintigraphy. Transplant Proc 1991;23:1690–1.
83. Klonaris CN, Bastounis EA, Xiromeritis NC, Balas PE. The predictive value of dipyridamole-thallium scintigraphy for cardiac risk assessment before major vascular surgery. Int Angiol 1998;17:171–8.
84. Lette J, Waters D, Lassonde J, et al. Postoperative myocardial infarction and cardiac death: predictive value of dipyridamole-thallium imaging and five clinical scoring systems based on multifactorial analysis. Ann Surg 1990;211:84–90.
85. Madsen PV, Vissing M, Munck O, Kelbaek H. A comparison of dipyridamole thallium 201 scintigraphy and clinical examination in the determination of cardiac risk before arterial reconstruction. Angiology 1992;43:306–11.
86. Stratmann H, Younis L, Wittry MD, et al. Dipyridamole technetium 99m sestamibi myocardial tomography for preoperative risk stratification before major or minor nonvascular surgery. Am Heart J 1996;536–41.
87. Lane SE, Lewis SM, Pipppin JJ et al. Predictive value of quantitative dipyridamole-thallium scintigraphy in assessing cardiovascular risk after vascular surgery in diabetes mellitus. Am J Cardiol 1989;64:1275–9.
88. Crystal E, Thorpe KE, Connolly SJ, et al. Metoprolol prophylaxis against postoperative atrial fibrillation increases length of hospital stay in patients not on pre-operative beta blocker: Beta blocker Length Of Stay (BLOS) trial. Heart 2004;90:941–2.
89. Rubin BB, Beattie WS. Does coronary artery revascularization before major vascular surgery benefit patients with coronary artery disease? Nature Clin Pract Cardiovasc Med 2005;2:190–1.
90. Stevens RD, Burri H, Tramer MR. Pharmacologic myocardial protection in patients undergoing noncardiac surgery: a quantitative systematic review. Anesth Analg 2003;97:623–33.
91. Auerbach AD, Goldman L Beta-blockers and reduction of cardiac events in noncardiac surgery: scientific review. JAMA 2002;287:1435–44.
92. Wijeysundera DN, Naik JS, Beattie WS. Alpha-2 adrenergic agonists to prevent perioperative cardiovascular complications: a meta-analysis. Am J Med 2003;114:742–52.
93. Wijeysundera DN, Beattie WS. Calcium channel blockers for reducing cardiac morbidity after noncardiac surgery: a meta-analysis. Anesth Analg 2003;97:634–41.
94. Beattie WS, Badner NH, Choi P. Epidural analgesia reduces postoperative myocardial infarction: a meta-analysis. Anesth Analg 2001;93:853–8.