Development and validation of a logistic regression-derived algorithm for estimating the incremental probability of coronary artery disease before and after exercise testing.

OBJECTIVES: Our goals were to develop and validate a multivariate algorithm for estimating the incremental probability of the presence of coronary artery disease. BACKGROUND: Multivariate methods, including logistic regression analysis, have been extensively applied to diagnostic exercise testing. However, few previous studies have included both an incremental design and external validation. METHODS: A retrospective collection of clinical, exercise test and catheterization data was performed involving four U.S. referral medical centers. All patients had no prior history of coronary disease and had undergone coronary angiography < or = 3 months after exercise stress testing. An algorithm was developed in one center (590 patients with a 41% prevalence of coronary artery disease) with the use of logistic regression analysis and was validated in the other three centers (1,234 patients, 70% prevalence). The algorithm incorporated pretest variables (age, gender, symptoms, diabetes, cholesterol), exercise electrocardiographic (ECG) variables (mm of ST segment depression, ST slope, peak heart rate, metabolic equivalents [METs], exercise angina) and one thallium variable. Discrimination was measured with receiver operating characteristic curve analysis. Calibration (that is, reliability) was assessed from a comparison of probability estimates and the actual prevalence of disease. RESULTS: The overall incremental receiver operating characteristic curve areas for the validation group were pretest, -0.738 +/- 0.016; postexercise ECG, 0.78 (SE 0.017); and postthallium, 0.82 (SE 0.016); p < 0.01 for both increments. Within the three validation institutions, the institution with a disease prevalence closest to that of the derivation institution had the best incremental receiver operating characteristic curve areas. There was a stepwise incremental improvement in calibration especially from exercise ECG to thallium testing. CONCLUSIONS: An incremental multivariate algorithm derived in one center reliably estimated disease probability in patients from three other centers. The incremental value of testing was best demonstrated when the derivation and validation groups had a similar disease prevalence. This algorithm may be useful in decision making that relates to the diagnosis of coronary disease.

Development and validation of a clinical score to estimate the probability of coronary artery disease in men and women presenting with suspected coronary disease.

PURPOSE: Guidelines for the management of patients with suspected coronary disease have emphasized stratification into groups with low, intermediate, and high probability of significant coronary disease. Previously derived clinical prediction rules have been difficult to apply in clinical settings. The purpose of this study was to develop and validate a clinical score that facilitates this stratification process. PATIENTS AND METHODS: We performed a retrospective analysis of prospectively acquired data from 915 patients with suspected coronary disease and normal resting electrocardiograms who presented for exercise testing at a university hospital. All patients subsequently underwent coronary angiography. Analysis included logistic regression with significant coronary disease (> or = 1 vessel with a > or = 50% lesion) presence as the dependent variable and clinical variables as independent variables. From this analysis, a coronary disease score was developed to estimate prevalence of coronary disease from clinical variables. Validation of this score was performed in a separate prospectively acquired cohort of 348 patients. RESULTS: For the entire validation group, the prevalence of significant coronary disease was 16% (10/63) in the low probability group, 44% (86/195) in the intermediate probability group, and 69% (62/90) in the high probability group. Both men and women were stratified equally well into the 3 probability groups. CONCLUSION: The clinical score is an easily memorized and accurate method for categorizing patients with suspected but not proven coronary disease and normal resting electrocardiograms into clinically meaningful probability groups upon which decisions concerning appropriate diagnostic test selection could potentially be based.

Diagnostic accuracy of heart rate-adjusted ST segments compared with standard ST-segment criteria.

We compared the accuracy of ST segment/heart rate (ST/HR) index with that of standard criteria (> or = 0.1 mV horizontal/downsloping ST depression 80 ms after the J point) in 121 patients who had undergone angiography (49 with > or = 1 lesion with > or = 50% stenosis) and 50 clinically normal subjects. All exercise tests used the Cornell protocol and computer measurements of maximal ST depression 80 ms after the J point. Thresholds with equal specificity to standard criteria were determined for ST/HR index using each of the 2 normal groups (those who were normal by angiography and those who were clinically normal). In using only patients who underwent angiography, we found that the ST/HR index had a sensitivity that was not significantly greater than that of standard criteria (standard criteria 51%, ST/HR index 59%; p = 0.21). However, the receiver-operating characteristic curve area increased from 64 +/- 4 to 68 +/- 4 (p < 0.02). When clinically normal subjects were used instead of patients without angiographic disease, there was a clearly discernible improvement in sensitivity of ST/HR index over standard criteria (standard criteria 51%, ST/HR index 69%; p < 0.05). The associated curve areas were 69 +/- 4 and 79 +/- 3 (p < 0.001). Therefore, accuracy of the ST/HR index was marginally better than standard criteria only in patients who underwent angiography. When clinically normal subjects were used, the accuracy of the ST/HR index was definitely better than standard criteria. We conclude that the demonstration of improved accuracy of the ST/HR index depends on the population being tested.

Comparison of three Bayesian methods to estimate posttest probability in patients undergoing exercise stress testing.

To determine whether recent refinements in Bayesian methods have led to improved diagnostic ability, 3 methods using Bayes' theorem and the independence assumption for estimating posttest probability after exercise stress testing were compared. Each method differed in the number of variables considered in the posttest probability estimate (method A = 5, method B = 6 and method C = 15). Method C is better known as CADENZA. There were 436 patients (250 men and 186 women) who underwent stress testing (135 had concurrent thallium scintigraphy) followed within 2 months by coronary arteriography. Coronary artery disease ([CAD], at least 1 vessel with greater than or equal to 50% diameter narrowing) was seen in 169 (38%). Mean pretest probabilities using each method were not different. However, the mean posttest probabilities for CADENZA were significantly greater than those for method A or B (p less than 0.0001). Each decile of posttest probability was compared to the actual prevalence of CAD in that decile. At posttest probabilities less than or equal to 20%, there was underestimation of CAD. However, at posttest probabilities greater than or equal to 60%, there was overestimation of CAD by all methods, especially CADENZA. Comparison of sensitivity and specificity at every fifth percentile of posttest probability revealed that CADENZA was significantly more sensitive and less specific than methods A and B. Therefore, at lower probability thresholds, CADENZA was a better screening method. However, methods A or B still had merit as a means to confirm higher probabilities generated by CADENZA (especially greater than or equal to 60%).(ABSTRACT TRUNCATED AT 250 WORDS)

Accuracy of ST/heart rate index in the diagnosis of coronary artery disease.

The accuracy of ST/heart rate (ST HR) index was evaluated in patients presenting for exercise electrocardiography with suspected coronary disease. In all, 420 patients (235 men and 185 women) with normal electrocardiograms at rest underwent exercise testing, followed within 3 months by coronary angiography. The sensitivity and specificity for standard ST criteria (greater than or equal to 1 mm horizontal or downsloping depression) were 48% (78 of 162) and 81% (208 of 258), respectively. An ST HR-index threshold of 1.86 microV/beta/min had the exact same specificity with a sensitivity of 44% (71 of 162; p = not significant). Consideration of greater than or equal to 1.5 mm upsloping depression had no significant impact on the aforementioned results. Using multivariate logistic regression analysis, age, sex, symptoms, cigarette smoking, diabetes mellitus, qualitative ST slope, rate-pressure product, METs achieved and exercise angina were evaluated with and without ST HR index and ST depression. According to this analysis, age, sex, symptoms and ST slope were good predictors of presence or absence of disease. Neither ST HR index nor ST depression had significance in the multivariate analysis. However, when a separate analysis was performed in men and women, the 2 quantitative ST variables showed significance in men, but not in women. Comparisons of discriminative accuracy using receiver-operating characteristic curves demonstrated differences between men and women, but no difference between ST HR index and ST depression. Therefore, concerning questions of coronary disease diagnosis, consideration of ST HR index was not better than standard ST criteria, and added nothing to multivariate analysis of other available variables.

Are the American College of Cardiology/American Heart Association guidelines for exercise testing for suspected coronary artery disease correct?

BACKGROUND: Recently published American College of Cardiology (ACC)/American Heart Association (AHA) guidelines state that patients with suspected coronary disease and an intermediate pretest probability are appropriate candidates for exercise ECG, while those with low or high pretest probability are not. METHODS: From 5,103 consecutive patients with symptoms of suspected coronary disease, we evaluated 872 patients who underwent coronary angiography following exercise ECG. Differences in test performance were determined using receiver operating characteristic curve area analysis. A score using age, gender, symptoms, and risk factors was used to classify patients into low, intermediate, and high pretest probability groups. RESULTS: When patients with inadequate exercise tests were excluded, overall sensitivity and specificity were 70% and 66%, respectively. Only the intermediate pretest probability group demonstrated significant incremental value: pretest vs posttest intermediate, 70 +/- 3 vs 79 +/- 3 (p < 0.0001); low, 71 +/- 6 vs 76 +/- 7 (p = 0.39); and high, 69 +/- 8 vs 75 +/- 7 (p = 0.12). From the low- to the high-probability groups, there was a progressive increase in positive predictive value (21%, 62%, and 92%) and decrease in negative predictive value (94%, 72%, and 28%), respectively. The frequencies of abnormal exercise ECGs were lower in the unselected groups compared with the angiography groups (low, 13% vs 36%; intermediate, 22% vs 53%; high, 36% vs 63%). CONCLUSIONS: Based on the information added by exercise testing to clinical data, these results confirm the ACC/AHA guideline assignments for test selection. However, despite these guidelines, patients with a low pretest probability can be selected for exercise testing with the knowledge that a positive result is infrequent and a negative result carries a very high negative predictive value. Intermediate-probability patients on average carry a significant false-negative rate, suggesting that exercise ECG alone may not be a sufficient screening test in all intermediate-probability patients. Because of poor negative predictive value and a large percentage of negative tests, high-probability patients should undergo coronary angiography as the initial strategy, unless the goal of exercise testing is to assess prognosis.

The effect of amyl nitrite on the mitral valve echocardiogram in presumably healthy young adults.

We analyzed the use of amyl nitrite as a provocative factor in the diagnosis of mitral valve prolapse in a population of healthy young adults. Sixty-five men and 11 women underwent continuous M-mode echocardiographic and phonocardiographic monitoring before, during and after the administration of inhaled amyl nitrite. All of the 76 subjects had normal baseline echocardiograms, and all had a satisfactory hemodynamic response to amyl nitrite. Mitral valve prolapse, defined by echocardiography and phonocardiography, was not provoked in any of the subjects. Therefore, we concluded that, although this technique may be difficult, significant false-positive results should not occur if adherence to strict diagnostic criteria takes place.

The specificity of exercise electrocardiography in women grouped by estrogen status.

We compared the specificity of exercise electrocardiography in 1880 men and 1818 women with women grouped by menopausal and estrogen replacement status. Specificity for > or = 1 mm horizontal or downsloping ST-segment depression was determined using angiography in 781 patients and using two other nonangiography-based methods (a pretest probability-based method and a predictive accuracy-based method) in all patients. Using angiography, the specificities+/-SE were 84 +/- 2 for men, 79 +/- 3 for women, 81 +/- 5 for premenopausal women, 81 +/- 4 for postmenopausal women without estrogen replacement, and 77 +/- 5 for women on estrogen replacement. None of these were significantly different. For all patients, the respective specificities using the probability and predictive accuracy-based methods were 97 +/- 1 and 94 +/- 1 for men, 90 +/- 1 and 88 +/- 1 for women, 97 +/- 1 and 92 +/- 2 for premenopausal women, 92 +/- 4 and 88 +/- 3 for postmenopausal women without estrogen replacement, and 85 +/- 4 and 81 +/- 3 for women on estrogen replacement. (Men vs. all women groups except premenopausal women-P < 0.05). Therefore, the premenopausal women had significantly greater specificity than women on estrogen replacement (P < 0.001) and no difference in specificity with men. Women on estrogen replacement had a significantly lower specificity than postmenopausal women not on estrogen replacement (P < 0.05). These results suggest that estrogen replacement therapy and not naturally occurring estrogen has a role in producing false positive exercise electrocardiograms in women.

The effect of disease-prevalence adjustments on the accuracy of a logistic prediction model.

The accuracy of a logistic prediction model is degraded when it is transported to populations with outcome prevalences different from that of the population used to derive the model. The resultant errors can have major clinical implications. Accordingly, the authors developed a logistic prediction model with respect to the noninvasive diagnosis of coronary disease based on 1,824 patients who underwent exercise testing and coronary angiography, varied the prevalence of disease in various "test" populations by random sampling of the original "derivation" population, and determined the accuracy of the logistic prediction model before and after the application of a mathematical algorithm designed to adjust only for these differences in prevalence. The accuracy of each prediction model was quantified in terms of receiver operating characteristic (ROC) curve area (discrimination) and chi-square goodness-of-fit (calibration). As the prevalence of the test population diverged from the prevalence of the derivation population, discrimination improved (ROC-curve areas increased from 0.82 +/- 0.02 to 0.87 +/- 0.03; p < 0.05), and calibration deteriorated (chi-square goodness-of-fit statistics increased from 9 to 154; p < 0.05). Following adjustment of the logistic intercept for differences in prevalence, discrimination was unchanged and calibration improved (maximum chi-square goodness-of-fit fell from 154 to 16). When the adjusted algorithm was applied to three geographically remote populations with prevalences that differed from that of the derivation population, calibration improved 87%, while discrimination fell by 1%. Thus, prevalence differences produce statistically significant and potentially clinically important errors in the accuracy of logistic prediction models. These errors can potentially be mitigated by use of a relatively simple mathematical correction algorithm.

The prediction of cardiac risk in patients undergoing vascular surgery.

In an attempt to determine whether noninvasive cardiac testing could be used to assess cardiac risk in patients undergoing surgery for vascular disease, the authors studied 96 patients. Seventy-seven patients eventually underwent major vascular surgery with 11 (14%) experiencing a significant cardiac complication. Thallium imaging was much more likely to be positive (p less than 0.01) in patients with a cardiac complication; however, there was a significant number of patients with cardiac complications who had a positive history or electrocardiogram for myocardial infarction. When grouped by complication and history of infarction, thallium imaging, if negative, correctly predicted low cardiac risk in the group with a history of infarction. Thallium imaging, however, did not provide a clear separation of risk in those without a history of infarction. Age and coronary angiography, on the other hand, did reveal significant differences within the group without a history of infarction. The resting radionuclide ejection fraction followed a similar pattern to thallium imaging. It is concluded that a positive history of myocardial infarction at any time in the past is the strongest risk predictor in this population and that the predictive value of noninvasive testing is dependent on this factor. Considering these findings, a proposed scheme for assessing risk that will require further validation is presented.

A new formula for the measurement of diastolic mitral valve orifice area.

This study evaluated the accuracy of a new formula for the calculation of mitral valve area. Fifty-two patients with mitral stenosis who underwent cardiac catheterization were evaluated by the standard Gorlin and the new formulas. The correlation between the two formulas was excellent (r = 0.89) for valve areas of 0-1.5 cm2. When the new valve area formula yielded an area greater than 1.5 cm2, there was no correlation with the Gorlin formula. However, the likelihood of the Gorlin mitral valve area being less than 1.0 cm2 was low (10%). The results of the new formula do not appear to be affected by atrial fibrillation and are probably subject to the same limitations when used in the presence of a regurgitant lesion. Therefore, with moderate to severe mitral stenosis, this new formula shows a good correlation. However, with mild stenosis, further work is needed to determine the accuracy and limitations of the new formula.

Left atrial abnormality by electrocardiogram predicts left ventricular hypertrophy by echocardiography in the presence of right bundle-branch block.

BACKGROUND: Left ventricular hypertrophy (LVH) on the electrocardiogram (ECG) may be masked in the presence of complete right bundle-branch block (RBBB). Left bundle-branch block on the ECG is associated with LVH at autopsy in 93% of hearts studied. However, RBBB does not predict LVH and the usual ECG criteria applied for LVH may not be reliable in the presence of RBBB. HYPOTHESIS: The study was undertaken to evaluate left atrial (LA) abnormality as a criterion for the diagnosis of LVH in the presence of RBBB. METHODS: Left atrial abnormality in the ECG was assessed by two independent observers as a criterion of LVH in the presence of RBBB in 100 patients, and data were compared with those of 50 patients without LA abnormality. RESULTS: Left ventricular hypertrophy was confirmed by echocardiographic determination of left ventricular (LV) mass in both groups. Observers reliably differentiated between hypertrophied and normal-sized LV in the presence of RBBB by using LA abnormality as an ECG criterion when correlated with LV mass determined by echocardiography. Observer 1 correctly detected LVH in 88% and Observer 2 in 82% of patients. False positive diagnosis was made in 12 and 18% of patients by Observers 1 and 2, respectively. Observers' performance of recognition of LA abnormality in the present study was 94%. Results showed sensitivity of 76 and 70% and specificity of 84 and 92% for Observers 1 and 2, respectively. Left ventricular mass increased significantly and was diagnostic of LVH in 92% of patients with LA abnormality. Left ventricular mass was high in 84% of patients when corrected by body surface area. LVH in the presence of RBBB by the ECG was found in only seven patients (5%) when six commonly used conventional criteria of diagnosis of LVH by ECG were employed. Regression analysis found LA abnormality to be a strong independent predictor of increased LV mass. Multiple regression analysis revealed that age, body mass index, body surface area, and frontal axis are also significant predictors of LV mass. CONCLUSION: The results obtained by the correlation of LA abnormality by ECG and LVH by echocardiography conclude that LA abnormality by ECG was significantly diagnostic of LV hypertrophy in the presence of RBBB.

Prediction of single versus multivessel disease following myocardial infarction using 201thallium scintigraphy and electrocardiographic stress testing.

We evaluated 50 patients who suffered a single myocardial infarction with graded electrocardiographic stress testing, 201thallium myocardial perfusion imaging and coronary angiography to assess the role of noninvasive indices as predictors of single versus multivessel coronary artery disease. Multivessel involvement was defined angiographically as the presence of two or more major coronary arteries with at least a 70% intraluminal diameter narrowing. Multivessel disease was defined scintigraphically as the presence of stress and/or redistribution perfusion defects in the distribution of more than one coronary artery. The results of stress electrocardiography were not useful in differentiating patients with single (9/16 positive) versus multivessel (22/34 positive) disease. The degree of exercise-induced ST-segment depression was also not helpful. Stress 201thallium imaging did offer limited additional information with correct predictions of multivessel disease in 21 of 26 patients. Predictions of single-vessel disease were accurate in 11 of 24 patients. Eleven of these 13 incorrect predictions of single-vessel disease were due to the relative insensitivity of the thallium stress image to perceive defect in the anterior wall when the left anterior descending artery had significant obstruction at catheterization. Further refinements of stress perfusion imaging are needed before this method can be used to reliably separate patients with single and multivessel disease after myocardial infarction.

Exercise testing for the primary care physician.

In the primary care setting, exercise testing is frequently used to evaluate patients with known or suspected coronary artery disease. Appropriate utilization requires knowledge of the indications as well as the contraindications and interpretation of the variables generated by the test. While ST segment changes are widely regarded as the most important variable generated by the exercise test, other variables including ventricular arrhythmias, blood pressure response, ischemic symptoms, workload, and exercise capacity are also important and useful in interpretation. Consideration of the question being addressed, the clinical context, and the test adequacy are all important for proper interpretation. Interpretation concerning both diagnostic and prognostic questions should employ principles of both the categorical and probability approaches.

Primary angioplasty in acute myocardial infarction: West Virginia University experience.

Primary angioplasty is a useful alternate treatment to open the occluded coronary arteries during the acute phase of a myocardial infarction associated with ST segment elevation. We describe our experience with 50 patients seen at West Virginia University Hospital in Morgantown over a period of one year.