Diagnosis of coronary artery disease: value and limitations of non-invasive methods. (Ecg, thallium perfusion scintigraphy, radionuclide angiography) (author's transl)

In order to compare the three non-invasive exercise tests Ecg, Thallium myocardial perfusion imaging and radionuclide angiography in the diagnosis of coronary artery disease, the results of these tests in a consecutive series of 30 patients and 14 controls were analyzed. In all 88 symptom-limited exercise tests a significantly higher double product (heart rate x systolic blood pressure, mm Hg/min) was reached on a treadmill test (for Ecg and Thallium scintigraphy) as compared to the supine bicycle ergometer exercise (for radionuclide angiography: 243.1 +/- 61.1 vs. 215.2 +/- 46.5 x 10(2) (p less than 0.01). Considering all 132 diagnostic tests the overall sensitivity for rest/exercise Ecg was 67%, for Thallium scans 77%, for both combined 83% and for the ejection fraction response to exercise determined by radionuclide angiography 97%. If only the exercise response was considered, the corresponding sensitivity values were 60% (Ecg), 47% (Thallium scans), 70% (both tests combined) and 97% (radionuclide angiography). The specificity for coronary artery disease was determined to be 79% for Ecg, 86% for Thallium scintigraphy, 64% for Ecg/Thallium scans and 71% for radionuclide angiography. The most common reason for a false-positive result in all tests was found to be the diagnosis of cardiomyopathy, whereas most false-negative results were seen in patients with single vessel right coronary artery disease. Based on these results, the clinical implications of the three non-invasive tests in the diagnosis of coronary artery disease are discussed.     

Noninvasive and angiographic evaluation of coronary artery disease in patients with peripheral vascular disease

The accuracy of exercise testing for detection of coronary artery disease in a population with a high incidence of claudication was evaluated in 58 consecutive patients with abdominal aortic aneurysms or lower extremity occlusive disease. Each patient was evaluated by history and physical examination, symptom-limited testing with exercise treadmill, arm ergometry and exercise radionuclide ventriculography. An algorithm was designed that retrospectively examined the results of each test in a stepwise fashion to simulate a clinical decision-making process. The results of the clinical examination, each of the exercise tests and the noninvasive diagnostic algorithm were compared with the results of coronary arteriography. The predictive accuracy of the clinical evaluation was 36%, treadmill stress testing 57%, treadmill stress plus arm ergometry 74%, exercise radionuclide ventriculography 57% and the noninvasive diagnostic algorithm 89%. When discriminant analysis was applied to all of the exercise variables, no individual test improved the accuracy of the noninvasive diagnostic algorithm. When the analysis considered only individual variables without the algorithm, the model correctly classified only 67% of the patients. Thus, accurate noninvasive evaluation of coronary artery disease is possible in patients with severe peripheral vascular disease when care is taken to design exercise protocols that allow adequate stress on the cardiovascular system.     

Diagnostic and Prognostic Use of Stress Echocardiography and Radionuclide Scintigraphy

Stress echocardiography and radionuclide scintigraphy are effective diagnostic and prognostic techniques in patients with known or suspected coronary artery disease (CAD), myocardial infarction (MI), chronic left ventricular dysfunction (LVD), and those undergoing noncardiac surgery. Both are sensitive and specific for the detection and extent of CAD. Negative tests confer a high negative predictive value for cardiac events irrespective of clinical risk. Positive studies confer a high positive predictive value for ischemic events in patients with intermediate to high clinical risk. Both provide incremental diagnostic and prognostic information relative to clinical, resting echocardiographic, and angiographic data. Meta-analysis studies have shown that the diagnostic and prognostic information provided by stress echocardiography is comparable with radionuclide scintigraphic stress tests. Stress echocardiography may be more specific for the detection and extent of CAD, whereas radionuclide scintigraphy may be more sensitive for single-vessel disease. Sensitivities are similar for the detection and extent of disease in patients with multivessel CAD.     

Implications of probability analysis on the strategy used for noninvasive detection of coronary artery disease: Role of single or combined use of exercise electrocardiographic testing,radionuclide Cineangiography and myocardial perfusion imaging

The accuracy of any less than perfect noninvasive test in detecting coronary artery disease is critically dependent not only oh its sensitivity and specificity, but also on the prevalence or pretest likelihood of disease in the population under study. Thus, an abnormal test result in a patient with a low pretest probability of disease (for example, an asymptomatic subject) is considerably more likely to be falsely indicative of disease than an identical result in a patient with a high pretest probability (for example, a patient with angina). On the basis of both theoretical considerations and clinical studies, it is now clear that diagnostic information derived from electrocardiographic exercise testing, radionuclide cineangiography or thalium perfusion imaging of persons with a low pretest probability of the disease is limited. However, by applying easily obtained clinical information, one can change estimates of the pretest likelihood of coronary artery disease and thereby greatly improve diagnostic information derived from a normal or an abnormal test result. Thus, the probability of coronary artery disease in a middle-aged man with hypertension and hypercho-lesterolemia who has an abnormal noninvasive test result is much greater than the probability of disease when the same abnormal result occurs in a younger man without such risk factors. Probability analysis can further enhance diagnostic accuracy if one applies the principle that when the results of two test procedures are independent of one another, the post-test likelihood of disease derived from the first test can be used as the pretest likelihood of disease for the second. Thus, if an electrocardiographic exercise test result is abnormal and the resulting probability of coronary artery disease is estimated at 70 percent, the diagnosis is still uncertain. However, the postelectrocardiographic test probability of 70 percent can be used as the pretest likelihood of coronary artery disease for radionuclide testing. If the radionuclide study result is abnormal, the probability of disease increases to nearly 100 percent. Thus, (1) nonin-vasive testing procedures yield probability estimates of disease, rather than simple “yes or no” diagnostic statements; (2) a working knowledge of probability analysis will make it easier for the physician to decide which patients might benefit from noninvasive diagnostic studies and whether more than one test should be employed; (3) this approach can provide reliable estimates of the probability that coronary artery disease is present or absent in an individual patient.     

Noninvasive diagnostic testing of coronary artery disease in women

Noninvasive diagnostic testing of coronary artery disease (CAD) is widely recognized as an area that is less studied and less accurate with regard to women than to men. Accurate and safe diagnostic testing constitutes the crucial link between early detection and optimal management of CAD. Many noninvasive diagnostic modalities are available to the clinician, including traditional electrocardiography, the relatively novel imaging of echocardiography, the emerging nuclear perfusion technology of electron beam computed tomography, exercise testing, and pharmacologic testing. The most accurate and cost-effective diagnostic method for patients depends on the patients' pretest likelihood of the disease as determined by factors such as sex, age, and cardiovascular risk factors. Noninvasive tests are most useful in the diagnosis of CAD in patients with intermediate pretest likelihood of CAD. Patients with low pretest likelihood of CAD with normal electrocardiograms may benefit from noninvasive tests or a watchful waiting strategy. Patients with a high pretest likelihood of CAD may benefit greatly from direct referral to coronary angiography. Among the noninvasive diagnostic methods, exercise electrocardiography is the most studied and least accurate with regard to women patients. Electrocardiography improves in accuracy when combined with imaging techniques such as echocardiography or nuclear single photon emission computed tomography. Combining data from all studies has shown that exercise echocardiography yields the highest diagnostic accuracy in women among all of the exercise stress tests. Patients who are unable to achieve maximal exercise capacity may undergo pharmacologic testing using dipyridamole or adenosine radionuclide perfusion or dobutamine echocardiography. Recent development of electron beam computed tomography accurately detects coronary artery calcium but has not been validated yet as a standard diagnostic test for CAD.     

Application of tissue Doppler to interpretation of exercise echocardiography: diagnostics of ischemia localization in patients with ischemic heart disease.

Development of optimal methods for the objective non-invasive diagnosis of coronary artery disease remains a challenge for imaging techniques in stress tests. AIM: The aim of this study was to obtain quantitative diagnostic criteria TDI which could detect significant coronary artery disease during exercise echocardiography. METHODS AND RESULTS: We evaluated regional systolic and diastolic myocardial functions of 123 patients by pulsed wave tissue Doppler imaging (TDI) in eight segments of left ventricle during exercise stress testing. Diagnostic criteria were obtained by comparing TDI and coronary angiography data. Best cut-points of velocity parameters allowed developing two diagnostic models for the detection of left anterior descending (LAD) and circumflex (LCx) artery diseases. The accuracy of the TDI diagnostic model for LAD-disease was 86.2% and for LCx-disease 78.3%. There were no criteria for the detection of RCA disease in this study. CONCLUSION: So TDI is a very accurate method for the detection of LAD- and LCx-disease during exercise stress echocardiography.     

Identification of patients with left main and three vessel coronary disease with clinical and exercise test variables

To determine whether clinical and exercise test variables either separately or in combination could reliably detect the presence of left main or three vessel coronary disease, 5 clinical and 11 exercise test variables were compared with the findings of coronary arterlography in 436 patients. Patients with left main coronary artery disease (n = 35) had an earlier onset of S-T segment depression (2.1 ± 1.4 versus 2.8 ± 1.7 min, p < 0.05), which was more prolonged (8.7 ± 3.6 versus 6.9 ± 3.3 min, p < 0.05) and appeared in a greater number of electrocardiographic leads (6.4 ± 2.2 versus 5.0 ± 2.2 leads, p < 0.001), than did patients with three vessel coronary disease (n = 89). Individual clinical or exercise test variables were unable to detect left main coronary disease because of their low sensitivity or predictive values. The pattern of 2 mm or greater downsloping S-T segment depression—which starts in stage 1, lasts at least 6 minutes into recovery and is displayed in at least five electrocardiographic leads—was highly predictive (74 percent) and reasonably sensitive (49 percent) for the detection of either left main or three vessel coronary disease. These criteria have a sensitivity of 74 percent and predictive value of 32 percent for the detection of isolated left main coronary artery disease.It is concluded that combining several exercise test variables facilitates the detection of severe coronary disease. The specific presence of left main coronary artery disease nevertheless remains largely unpredictable even with this approach.     

Noninvasive myocardial imaging with potassium-43 and rubidium-81 in patients with left bundle branch block

Noninvasive myocardial imaging with potassium-43 and rubidium-81 has been used successfully to identify areas of infarction and exercise-induced ischemia as regions of decreased radioactivity. The image defects observed are believed to be due to a decreased radionuclide uptake in regions of myocardial scar or to heterogeneous myocardial accumulation of tracer as a result of regional ischemia. Of 27 patients with left bundle branch block studied with noninvasive imaging at rest and during exercise, 25 manifested at rest reduced radioactivity in the region of the interventricular septum. This pattern is similar to that seen in patients with anteroseptal myocardial infarction. Sixteen of the 27 patients underwent diagnostic coronary arteriography and left ventriculography. Only five of these patients had evidence of either previous infarction or significant obstructive coronary artery disease as assessed with clinical or anglographic criteria, or both. Although the image defect was routinely demonstrated at rest in patients with left bundle branch block, this defect was generally normalized or less distinct with exercise in patients with no anatomic heart disease. In contrast, a larger, more distinct or new image defect with exercise correctly identified the presence of significant obstructive coronary artery disease in patients with left bundle branch block. In the clinical application of noninvasive myocardial imaging, these image defects observed at rest can lead to the false positive radionuclide interpretation of anteroseptal myocardial infarction.     

Exercise left ventriculography utilizing intravenous digital angiography

Exercise left ventriculography has been shown to be a sensitive and specific tool for the detection of coronary artery disease. At the present time, such studies require radionuclide-base methods. Computer-based techniques recently have been shown to provide high resolution images of the left ventricle when the levophase of an intravenous injection of radiopaque contrast medium is imaged with fluoroscopy. To evaluate the possible efficacy of using "intravenous digital subtraction left ventriculograms" in exercise ventriculography, such ventriculograms were performed at rest and during maximal supine bicycle exercise in 31 patients. Studies that could be analyzed were obtained in 29 patients. In 21 patients with coronary artery disease, ejection fraction was 58% at rest and 45% with exercise (p less than 0.001 vs. rest). In contrast, in seven patients with no coronary artery disease, ejection fraction was 65% at rest and 69% with exercise (difference not significant). In a subgroup of 8 patients with "severe" coronary obstruction, the change in ejection fraction from rest to exercise was -18%, while in the remaining 13 patients with less severe disease, it was -9% (p less than 0.001). All patients with coronary artery disease manifested new or worsening segmental wall abnormality with exercise, compared with two of seven patients without coronary disease (p less than 0.01). Sixteen patients underwent rest and exercise radionuclide cineangiography in addition to digital subtraction angiography. There was a strong correlation between the two techniques for ejection fraction at rest (r = 0.78, p less than 0.001), ejection fraction and with exercise (r = 0.83, p less than 0.001) and change in ejection fraction from rest to exercise (r = 0.88, p less than 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)     

Comparison of exercise stress testing with dobutamine stress echocardiography and radionuclide ventriculography for diagnosis of coronary artery disease

Dobutamine stress echocardiography, Tc-99m radionuclide ventriculography (RNVG), and exercise stress testing were performed prospectively in 63 patients with suspected coronary artery disease to compare the values of exercise testing, dobutamine stress echocardiography and RNVG in the non-invasive diagnosis of coronary artery disease. The sensitivities of dobutamine stress echocardiography and RNVG were found to be higher than that of exercise testing (93-62%, p < 0.001; 83-62%, p < 0.05). The sensitivities of dobutamine stress echocardiography and RNVG were similar (p > 0.05). There were no differences between the sensitivities of the three techniques in multiple vessel disease (p > 0.05). The specificities of dobutamine stress echocardiography and RNVG were higher than that of exercise testing (for both of the tests 86-62%, p < 0.05). The diagnostic accuracy of dobutamine stress echocardiography and RNVG were similar (p > 0.05). The results of dobutamine stress echocardiography RNVG were concordant with each other in 46 patients (76%, kappa = 65%) in sectional analysis. Dobutamine stress echocardiography and RNVG tests were comparable with each other in 85% of the 189 segments (kappa = 64%). The expected 5% decrease at peak doses of dobutamine was not detected in stress echocardiography in 25 patients and in RNVG in 26 of the patients. Dobutamine stress echocardiography and RNVG are superior to exercise testing in the diagnosis of single vessel disease and there is no significant difference between the two techniques. When the ejection fraction is considered in dobutamine stress echocardiography and RNVG, it does not make an additional contribution to the diagnosis of coronary artery disease.     

How useful are the cold pressor test and sustained isometric handgrip exercise with radionuclide ventriculography in the evaluation of patients with coronary artery disease?

The feasibility of using the cold pressor test and the sustained isometric handgrip test as alternatives to dynamic exercise for stressing the heart was investigated. Serial changes in heart rate, blood pressure, and left ventricular performance induced by these tests were studied by radionuclide ventriculography in patients with coronary artery disease and in normal volunteers. Both tests significantly increased heart rate and blood pressure. The reproducibility of serial evaluation of ejection fraction response to cold pressor and isometric handgrip stresses was satisfactory but the sensitivity for detecting coronary artery disease was not. Both stress tests are valuable interventions for the serial evaluation of left ventricular function by radionuclide ventriculography, but they should not be used to detect coronary artery disease.     

Exercise-induced regional wall motion abnormalities on radionuclide angiography: Lack of reliability for detection of coronary artery disease in the presence of valvular heart disease

Exercise-induced regional wall motion abnormalities on radionuclide angiography have been thought to be a reliable indicator of coronary artery disease. To evaluate their reliability, particularly in patients with valvular heart disease, exercise radionuclide angiography was performed in 12 normal subjects, 35 patients with coronary artery disease and 19 patients with valvular heart disease and normal coronary arteries. Exercise-induced regional wall motion abnormalities were found in none of the normal subjects, 63 percent of the patients with coronary artery disease and 42 percent of those with valvular heart disease and were predominantly inferoapical in location in the group with valvular heart disease. We conclude that exercise-induced regional wall motion abnormalities are not reliable for the detection of coronary artery disease in patients with valvular heart disease.     

How useful are the cold pressor test and sustained isometric handgrip exercise with radionuclide ventriculography in the evaluation of patients with coronary artery disease?

The feasibility of using the cold pressor test and the sustained isometric handgrip test as alternatives to dynamic exercise for stressing the heart was investigated. Serial changes in heart rate, blood pressure, and left ventricular performance induced by these tests were studied by radionuclide ventriculography in patients with coronary artery disease and in normal volunteers. Both tests significantly increased heart rate and blood pressure. The reproducibility of serial evaluation of ejection fraction response to cold pressor and isometric handgrip stresses was satisfactory but the sensitivity for detecting coronary artery disease was not. Both stress tests are valuable interventions for the serial evaluation of left ventricular function by radionuclide ventriculography, but they should not be used to detect coronary artery disease.     

Diagnostic and prognostic use of stress echocardiography in stable patients

Stress echocardiography is an effective diagnostic and prognostic technique in stable patients with known or suspected coronary artery disease (CAD), myocardial infarction, or chronic left ventricular dysfunction and those undergoing noncardiac surgery. Stress echocardiography is sensitive and specific for the detection and extent of CAD. Negative tests confer a high negative predictive value for cardiac events regardless of the clinical risk. Positive studies confer a high positive predictive value for ischemic events in patients with intermediate to high clinical risk. Stress echocardiography provides incremental prognostic information relative to clinical, resting echocardiographic, and angiographic data. Meta-analysis studies have shown that the diagnostic and prognostic information provided by stress echocardiography is comparable to that from radionuclide scintigraphic stress tests. Stress echocardiography may be more specific for the detection and extent of CAD, whereas radionuclide scintigraphy may be more sensitive for one-vessel disease. Sensitivities are similar for the detection and extent of disease in patients with multivessel CAD.     

Assessment of single vessel coronary artery disease: results of exercise electrocardiography, thallium-201 myocardial perfusion imaging and radionuclide angiography

The sensitivity of the commonly used stress tests for the diagnosis of coronary artery disease was analyzed in 46 patients with significant occlusion (greater than or equal to 70% luminal diameter obstruction) of only one major coronary artery and no prior myocardial infarction. In all patients, thallium-201 perfusion imaging (both planar and seven-pinhole tomographic) and 12 lead electrocardiography were performed during the same graded treadmill exercise test and radionuclide angiography was performed during upright bicycle exercise. Exercise rate-pressure (double) product was 22,307 +/- 6,750 on the treadmill compared with 22,995 +/- 5,622 on the bicycle (p = NS). Exercise electrocardiograms were unequivocally abnormal in 24 patients (52%). Qualitative planar thallium images were abnormal in 42 patients (91%). Quantitative analysis of the tomographic thallium images were abnormal in 41 patients (89%). An exercise ejection fraction of less than 0.56 or a new wall motion abnormality was seen in 30 patients (65%). Results were similar for the right (n = 11) and left anterior descending (n = 28) coronary arteries while all tests but the planar thallium imaging showed a lower sensitivity for isolated circumflex artery disease (n = 7). The specificity of the tests was 72, 83, 89 and 72% for electrocardiography, planar thallium imaging, tomographic thallium imaging and radionuclide angiography, respectively. The results suggest that exercise thallium-201 perfusion imaging is the most sensitive noninvasive stress test for the diagnosis of single vessel coronary artery disease.     

Global and regional left ventricular response to bicycle exercise in coronary artery disease. Assessment by quantitative radionuclide angiocardiography.

The left ventricular response to bicycle exercise was evaluated in 60 patients with coronary artery disease and in 13 normal control subjects. Left ventricular ejection fraction, mean normalized ejection rate and regional wall motion were determined using first-pass radionuclide angiocardiograms obtained at rest and again during peak graded bicycle exercise. All normal subjects demonstrated improved left ventricular function with exercise. Left ventricular ejection fraction increased significantly from 67 ± 3 per cent (mean ± SE) at rest to 82 ± 4 per cent with exercise (p < 0.001). Similarly, the left ventricular ejection rate increased significantly from 3.47 ± 0.31 sec^?1 to 6.53 ± 0.42 sec^?1(p < 0.001). In contrast, in 44 of 60 patients with coronary artery disease, the ejection fraction or ejection rate either decreased or remained the same with exercise. New or exaggerated regional wall motion abnormalities were detected in 28 of 60 patients with coronary artery disease. Over-all, global or regional evidence of compromised left ventricular reserve was found in 48 of 60 patients with coronary artery disease.The major determinant of an abnormal left ventricular response to exercise was the presence or absence of electrocardiographic evidence of myocardial ischemia. Left ventricular ejection fraction decreased or remained the same with exercise in all patients with coronary artery disease and electrocardiographic ischemia. New regional wall motion abnormalities were detected in 20 of these patients. In this group, the left ventricular ejection fraction decreased from 66 ± 2 per cent at rest to 58 ± 2 per cent with exercise (p < 0.001), whereas the ejection rate was unchanged by exercise (rest 3.33 ± 0.21 sec^?1; exercise 3.34 ± 0.22 sec^?1, p > 0.05). Of the 30 patients with coronary artery disease who exercised to symptom-limiting fatigue without electrocardiographic ischemia, 18 demonstrated compromised left ventricular reserve with exercise. Twelve of the remaining patients with coronary artery disease had normal left ventricular reserve, in eight of whom ventricular function was completely normal both at rest and during exercise. In this group exercised to fatigue, the left ventricular ejection fraction increased from 53 ± 4 per cent at rest to 58 ± 2 per cent with exercise (p < 0.001). The ejection rate also increased from 2.48 ± 0.24 sec^?1 to 3.67 ± 0.39 sec^?1 (p < 0.001). The direction and magnitude of the left ventricular responses to exercise were not affected by long-term oral propranolol administration in 22 patients. Based upon either abnormal exercise left ventricular reserve or abnormal global and regional left ventricular function at rest, the over-all sensitivity of this radionuclide technic for the detection of coronary artery disease was 87 per cent (52 of 60 patients). These data demonstrate that exercise ventricular performance studies provide important physiologic insights into left ventricular functional reserve as well as a sensitive noninvasive approach for the detection of coronary artery disease.     

Coronary calcification in the diagnosis of coronary artery disease.

Clinical, postmortem and angiographic studies of coronary calcification are reviewed to define the value of fluoroscopy in the diagnosis and management of coronary artery disease. Autopsy studies consistently show a unique association between calcification of the coronary arteries and atherosclerosis. The relation of coronary calcification to the presence of major stenosis is more variable but is strong enough to be of clinical value, particularly in the younger subject. The diagnostic value of fluoroscopy can be improved by attention to the detailed features of calcification observed with the technique. Combined use of fluoroscopy and exercise testing appears to be a valid and as yet unexploited approach to the noninvasive diagnosis of coronary stenosis. Fluoroscopy has been a neglected method of noninvasive diagnosis and is sufficiently promising to warrant greater clinical use.     

Computer-enhanced thallium scintigrams in asymptomatic men with abnormal exercise tests

The use of treadmill testing in asymptomatic patients and those with an atypical chest pain syndrome is increasing, yet the proportion of false positive stress electrocardiograms increases as the prevalence of disease decreases. To determine the diagnostic accuracy of computer-enhanced thallium perfusion scintigraphy in this subgroup of patients, multigated thallium scans were obtained after peak exercise and 3 or 4 hours after exercise and the raw images enhanced by a computer before interpretations were made. The patient group consisted of 191 asymptomatic U.S. Air Force aircrewmen who had an abnormal exercise electrocardiogram. Of these, 135 had normal coronary angiographic findings, 15 had sub-critical coronary stenosis (less than 50 percent diameter narrowing) and 41 had significant coronary artery disease.

Use of computer-enhancement resulted in only four false positive and two false negative scintigrams. The small subgroup with subcritical coronary disease had equivocal results on thallium scintigraphy, 10 men having abnormal scans and 5 showing no defects. The clinical significance of such subcritical disease is unclear, but it can be detected with thallium scintigraphy.

Thallium scintigrams that have been enhanced by readily available computer techniques are an accurate diagnostic tool even in asymptomatic patients with an easily interpretable abnormal maximal stress electrocardiogram. Thallium scans can be effectively used in counseling asymptomatic patients on the likelihood of their having coronary artery disease.     

Radionuclide studies in the evaluation of ischemic heart disease

In recent years, radionuclide studies have gained an important place in the evaluation of ischemic heart disease, be it as diagnostic procedures, as predictors of prognosis or for evaluation of therapy. For diagnostic purposes, myocardial perfusion studies using thallium-201 or newer technetium-99m bound perfusion agents have been used as well as radionuclide angiocardiography both at rest and during exercise/stress. Used in a Bayesian approach, these methods yield the highest diagnostic accuracy in patients with a 30% to 70% pre-test likelihood of disease, i.e. in the clinically difficult patients with atypical chest pain and/or non-specific ECG changes. In addition, scintigraphic studies have proved valuable in the setting of silent ischemia and acute myocardial infarction. These methods provide not only a yes/no answer to our diagnostic questions but allow one to assess severity, extent and localization of coronary artery disease. Portable devices are now being constructed which allow continuous ambulatory monitoring of left ventricular function by scintigraphic techniques.     

Diagnostic accuracy of cardiologists compared with probability calculations using Bayes'' rule

Probability analysis has provided insights into the use of diagnostic tests in coronary artery disease, and recent developments may permit clinical application to individual patients. To validate independently two available methods of probability calculation, their diagnostic accuracy was compared with that of cardiologists. Ninety-one cardiologists participated in the study; each evaluated the clinical summaries of eight randomly selected patients. For each patient, the cardiologist assessed the probability of coronary artery disease after reviewing the clinical history, physical examination and laboratory data, including complete results from a treadmill exercise test. The probability of coronary artery disease was also obtained for each patient, using the identical information, from two methods employing Bayes' rule: (1) from a published table of data based on the patient's age, sex, symptoms and degree of S-T segment change during exercise; and (2) from a computer program using the age, sex, risk factors, resting electrocardiogram and multiple exercise measurements. Diagnostic accuracy was assessed on a scale from 0 to 100 with the coronary angiogram as the diagnostic standard. The average diagnostic accuracy on this scale was: 80.2 for the cardiologists' estimates, 78.0 for the estimates based on tables (difference from cardiologists' estimates p < 0.05) and 83.1 for the estimates based on computer calculations (p < 0.01). Thus probability analysis incorporating sufficient detail can achieve a diagnostic accuracy comparable with that of cardiologists. Studies of the efficacy of probability analysis in patient care are warranted.