Improvement in identification of multivessel disease after acute myocardial infarction following stress-recovery analysis of ST depression in the heart rate domain during exercise

Impact of advanced exercise ECG analysis on cost of coronary artery disease management

BACKGROUND

Conventional analysis of exercise electrocardiogram (EX-ECG) has limited accuracy. This study aims to evaluate the potential impact of improving EX-ECG accuracy on costs of diagnosis and number of misdiagnoses of coronary artery disease (CAD).

METHODS

A decision-tree model was simulated including sequential application of diagnostic procedures for suspected CAD. The model was structured in two main branches (presence or absence of CAD). A probabilistic sensitivity analysis was then performed for several combinations of improvement in test sensitivity and specificity.

RESULTS

A clear trend in cost reduction was observed at improving EX-ECG specificity (about 8-8.5 million dollars, corresponding to a 5.6-7.6% reduction according to the prevalence level). Wrong diagnoses counted for 9-13% of test. Improvements in test parameters lead to reductions in wrong diagnoses, especially when increasing specificity (8.8-12.5%).

CONCLUSIONS

A proper improvement in EX-ECG sensitivity and specificity would have a relevant impact on the costs of CAD management, while reducing the number of misdiagnoses.

The Stress-Recovery Index for the risk stratification of women with typical chest pain

AIM

To prospectively assess the prognostic value of the Stress-Recovery Index (SRI) in women with typical chest pain.

METHODS

165 women without known coronary artery disease, who complained of typical chest pain, were exercise tested and prospectively followed-up for the occurrence of cardiac death and nonfatal myocardial infarction. SRI, defined as the difference in absolute values between the area of heart rate-adjusted ST-segment depression during exercise and recovery, was derived in all. Clinical data, resting ejection fraction, and exercise testing data were entered into a sequential Cox's model; SRI was entered last. Model validation was performed by bootstrap adjusted by the degree of optimism in estimates. Survival curves were set up using Kaplan-Meier method and compared by the log-rank test.

RESULTS

During a median follow-up time of 42 months, 19 events (14 cardiac deaths and 5 nonfatal myocardial infarction) were observed. Age (hazard ratio 3.58, 95% CI 0.87-15) and SRI (hazard ratio 0.62, 95% CI 0.42-0.92) were multivariate predictors of outcome. However, the addition of SRI increased the prognostic power of the model on top of clinical and exercise testing variables, as demonstrated by the significant (p=0.003) increase of the area under the ROC curve of the risk function. Survival analysis showed ascending SRI quartiles to identify a significant (p=0.005) increase in event-free survival.

CONCLUSIONS

SRI is of value in predicting outcome of women with typical chest pain and provides additional prognostic information on the top of clinical and standard exercise testing data.

Exercise electrocardiography for diagnosis of coronary artery disease: impact of sampling rate on the diagnostic performance of ST/HR-loop based on data from early recovery phase

Quantitative analysis of electrocardiographic ST-segment/heart rate relationship (ST/HR loop) during early recovery phase of exercise stress test provides a sensitive tool for the detection of coronary artery disease (CAD). This study evaluates the effect of data sampling frequency on the diagnostic performance of ST/HR loop in 1876 patients undergoing a routine exercise test on a bicycle ergometer. CAD was verified angiographically in 668 patients and excluded by coronary angiography (n = 119), myocardial scintigraphy (n = 250) and on clinical grounds (n = 839) in 1208 patients. The normalized ST/HR loop area was calculated in all cases by integration of ST-segment amplitude difference from the end of exercise to the end of the first 3 min of recovery period over HR and dividing the integral by the HR difference over the integration period. The effect of different sampling rates (one, two and five samples per minute) on the CAD discrimination ability of ST/HR loop area was subsequently evaluated using receiver operating characteristic (ROC) curves. Reduction in ST/HR data sampling frequency from two to one sample per minute resulted in a significantly decreased diagnostic performance of the ST/HR loop whereas no differences in CAD discrimination capacity were observed between sampling frequencies of two and five samples per minute. The choice of ST/HR data sampling frequency may have a significant impact on the CAD diagnostic ability of the ST/HR loop. The use of sampling frequency below two samples per minute results in a significantly diminished diagnostic performance, a fact that should be taken into consideration when employing ST/HR diagnostic procedures.

Prognostic evaluation by clinical exercise test scores in patients treated with primary percutaneous coronary intervention or fibrinolysis for acute myocardial infarction (a Danish Trial in Acute Myocardial Infarction-2 Sub-Study)

The prognostic accuracy of exercise testing after myocardial infarction is low, and different models have been proposed to enhance the predictive value for subsequent mortality. This study tested a simple score against 3 established scores. Patients with ST-elevation myocardial infarctions were randomized in the Danish Trial in Acute Myocardial Infarction-2 (DANAMI-2) to either primary percutaneous coronary intervention or fibrinolysis with predischarge exercise testing. Clinical and exercise test data were collected prospectively and were available for 1,115 patients. A simple score was derived, awarding 1 point for history or new signs of heart failure, 1 point for a left ventricular ejection fraction <40%, 1 point for age >65 years in men and age >70 years in women, and 1 point for exercise capacity <5 METs in men and exercise capacity <4 METs in women. This DANAMI score was compared with the Veterans Affairs Medical Center score, the Duke treadmill score, and the Gruppo Italiano per lo Studio Della Sopravvivenza nell'Infarto Miocardico-2 (GISSI-2) score in multivariate Cox models and receiver-operating characteristic plots. All scoring systems were predictive of adverse outcomes. The DANAMI score performed better, with greater chi-square values (142 vs 53 to 88 for the prediction of death). Areas under the receiver-operating characteristic curves were compared and were larger for the DANAMI score (C-statistic 0.79 vs 0.71 to 0.74 for the other tests regarding mortality). The DANAMI score stratified patients into a small high-risk group (8% of the population with 43% mortality in 6 years), an intermediate-risk group (13% with 16% mortality in 6 years), and a low-risk group (79% with 4% mortality in 6 years). In conclusion, a simple exercise test score composed of age, METs, heart failure, and a left ventricular ejection fraction <40% seems to outperform the Duke treadmill score, Veterans Affairs Medical Center score, and GISSI-2 score in risk stratifying patients after myocardial infarction and deserves further evaluation.

Comparison of the prognostic value of the stress-recovery index versus standard electrocardiographic criteria in patients with a negative exercise electrocardiogram

To verify whether the stress recovery index (SRI) improves risk stratification in patients with a negative exercise electrocardiogram (ECG) using standard criteria, the SRI was derived in 708 consecutive patients with a negative exercise ECG. All-cause mortality and the combination of death or nonfatal myocardial infarction were target end points. The individual effect of clinical and exercise testing data on outcome was evaluated using Cox regression analysis with separate models for each group of variables. Model validation was performed using bootstrap adjusted by degree of optimism in estimates. Survival analysis was performed using a product-limit Kaplan-Meier method. During a 37-month follow-up, 22 deaths and 40 nonfatal acute coronary syndromes occurred. After adjusting for confounding variables, age (hazard ratio 1.62, 95% confidence interval [CI] 1.14 to 2.31 for interquartile difference), hypertension (hazard ratio 1.74, 95% CI 1.04 to 2.89), and SRI (hazard ratio 0.75, 95% CI 0.65 to 0.86 for interquartile difference) were predictive of death or nonfatal myocardial infarction. Moreover, SRI increased the prognostic power of the model on top of clinical and exercise testing variables and provided significant discrimination of survival. In conclusion, the SRI may help refine the prognostic stratification of patients with a negative exercise test result using standard electrocardiographic criteria.

Stress Recovery Index for Risk Stratification of Asymptomatic Patients Following Coronary Bypass Surgery

OBJECTIVE

To prospectively assess the prognostic value of the stress recovery index (SRI) following coronary bypass surgery.

DESIGN AND PATIENTS

Two hundred seventy-eight patients who had undergone coronary bypass surgery and participated in a secondary prevention program were exercise tested and prospectively followed up for a median of 36 months. Cardiac death, nonfatal infarction, and need for further revascularization were target end points. SRI, defined as the difference in absolute values between the area of heart rate-adjusted ST-segment depression during exercise and recovery, was derived in all. Clinical data, resting ejection fraction, and exercise testing data of patients were entered into a sequential Cox model; SRI was entered last. Model validation was performed by bootstrap adjusted by the degree of optimism in estimates. Survival curves were set up using Kaplan-Meier method and compared by the log-rank test.

RESULTS

SRI was the only significant and independent prognostic indicator (hazard ratio, 0.68; 95% confidence interval, 0.53 to 0.89) and increased the prognostic power of the model on top of clinical and exercise testing variables, as demonstrated by the significant (p = 0.01) increase of the area under the receiver operating characteristic curve of the risk function. Survival analysis showed ascending SRI quartiles to identify a significant (p = 0.001) increase in event-free survival.

CONCLUSIONS

SRI is of value in predicting outcome after coronary bypass surgery and provides additional prognostic information over clinical and exercise testing data.

Exercise versus recovery electrocardiography for predicting outcome in hypertensive patients with chest pain

BACKGROUND

Exercise electrocardiography has limited prognostic accuracy in hypertensives because of unsatisfactory specificity. We prospectively used comparative stress-recovery heart rate-adjusted ST (ST/HR) analysis to predict mortality in a consecutive population of hypertensives with chest pain.

METHODS

The stress-recovery index (SRI), defined as the difference between ST/HR areas during exercise and recovery, was derived in 460 hypertensive with known (n=360, 78%) or suspected (n=100, 22%) coronary artery disease. To assess whether it added prognostic information to routinely obtained information, clinical data, the resting ejection fraction, and exercise testing data were entered into a sequential Cox's model; the SRI was entered last. Model validation was performed by bootstrap adjusted by the degree of optimism in estimates. Survival analysis was performed using the product-limit Kaplan-Meier method.

RESULTS

During a median follow-up of 28 months (interquartile range, 13-44 months), 32 (7%) patients died, 23 (5%) suffered from acute myocardial infarction and 60 (13%) underwent late (> 3 months) revascularization. Male gender (hazard ratio, 1.53; 95% confidence interval, 1.01-2.34), peak double product (hazard ratio, 0.70; 95% confidence interval, 0.54-0.90) and the SRI (hazard ratio, 0.69; 95% confidence interval, 0.59-0.81 for interquartile difference) were independent predictors of outcome. The SRI increased the prognostic power of the model on top of clinical and exercise testing variables (concordance index, + 10%; discrimination index, + 32%) and showed the widest area under the ROC curve to predict outcome as compared with exercise-only ST analysis and the ST/HR index. Moreover, it provided a significant discrimination of survival.

CONCLUSIONS

The SRI predicts all-cause mortality in hypertensive patients with chest pain and provides additional prognostic information over clinical and standard exercise testing data.

Usefulness of a quantitative analysis of the recovery phase patterns of the ST-segment depression in the diagnosis of coronary artery disease

PURPOSE

To assess the diagnostic value of the recovery phase patterns of the ST-segment depression in patients referred for chest pain.

PATIENTS AND METHODS

Continuous plots of ST-segment depression against heart rate during exercise and recovery were constructed within a population of 160 consecutive symptomatic patients who all had undergone catheterization (80 with > or = 1 stenosis > or = 50%). We used a new quantitative method of measurement allowing all kinds of rate recovery loops (even the so-called "intermediate" loops) to be considered for analysis. The measurements of the heart rate (HR)-adjusted ST-segment depression were performed at 20 and 60 ms from the J point, providing two different values of a quantified recovery loop index (RLI): RLI 20 and RLI 60.

RESULTS

Both RLI showed a higher specificity (0.81 +/- 0.04 and 0.74 +/- 0.05, respectively) than did the standard criterion (0.65 +/- 0.10), but the difference was significant regarding RLI 20 only (P = 0.011). As to the sensitivity, no significant differences were found among all of the criteria (0.74 +/- 0.05, 0.80 +/- 0.04, 0.76 +/- 0.05, respectively). The timing of measurements of the RLI within the repolarization phase did not affect their overall accuracy (0.77 +/- 0.03 for both RLI). The values of the receiver-operating characteristic (ROC) curve areas were significantly greater for both RLI (0.83 +/- 0.06 and 0.84 +/- 0.06 respectively) than for the standard criterion (0.75 +/- 0.07; P < 0.02). Finally, both RLI allowed to differentiate accurately the study subjects according to the number of diseased vessels, whereas the standard criterion could only distinguish between CAD patients and subjects with normal angiograms.

CONCLUSION

The quantitative analysis of the rate recovery phase patterns appears to be useful for the diagnosis of coronary heart disease and the assessment of its severity in symptomatic patients.

Clinical evaluation of algorithms for ST measurement during exercise test

HYPOTHESIS

Computer processing of the exercise electrocardiogram (ECG) has many advantages, but the reliability of the analysis algorithms is not easily evaluable. No standard annotated database, nor recommended practice for testing and reporting performance results is available: thus, performance evaluation of such devices can be accomplished only by using a set of unannotated recordings, obtained in clinical practice. We evaluated the accuracy of an original microcomputer-based exercise test analyzer comparing the ST computer output with the measurements obtained by two experienced cardiologists.

METHODS

Six hundred ECG strips were randomly selected from the exercise test recordings of 60 patients. The ST shift (at J + 80 ms) was blindly assessed by two observers (with the aid of a calibrated lens) and compared with computer measurements. Correlation coefficients, linear regression equations, percent of discrepant measurements, and 95% confidence limits of the mean error were calculated for all leads, peripheral leads, precordial leads, and "stress-test" leads (II, III, aVF, V4, V5, V6).

RESULTS

The computer did not analyze five samples on a total of 600 (0.83%) ECG strips because of excessive noise or signal loss, while 51 (8.5%) were considered unreadable by both observers and 67 (11.2%) were rejected by at least one observer. Correlation between the measurements taken by computer and observer(s) measurements was statistically significant (p < 0.001 for all lead groups), no systematic measurement bias was found, and the mean difference was lower than human eye resolution.

CONCLUSIONS

Our algorithms provide results as good as those provided by trained cardiologists in measuring ST changes occurring during exercise test. However, this study did not evaluate whether computer improvement of the signal-to-noise ratio would allow accurate measurements even on cardiologists' uninterpretable ECG. This potential advantage of computer-assisted analysis could be assessed only by using a dedicated exercise test database, in which different patterns of noise are superimposed on noise-free recordings previously annotated for ST level.