Improved detection of coronary artery disease
in women USING ST/HR hysteresis METHOD

J. Viik, R. Lehtinen and J. Malmivuo
Ragnar Granit Institute, Tampere University of Technology,
P.O. Box 692, FIN-33101 Tampere, FINLAND

Abstract: ST/HR hysteresis, the method that integrates the heart rate-adjusted ST-segment depression of the exercise and recovery phases, has been suggested to improve the diagnostic accuracy of exercise ECG in the identification of coronary artery disease (CAD) compared with other traditional methods. However, the comparisons have been made in the male-dominated populations. This study compares the diagnostic performance of the ST/HR hysteresis, ST/HR index, and ST-segment depression at end-exercise using a female study population. The comparison of the variables was made with the receiver-operating characteristic curve analysis. The diagnostic performance of the ST/HR hysteresis was significantly better than standard end exercise ST segment depression. Furthermore the sensitivities at the high specificities were considerably higher than that of others. ST/HR hysteresis significantly increases the diagnostic performance of exercise ECG in the detection of CAD in women.

INTRODUCTION

The detection of coronary artery disease (CAD) by exercise electrocardiography (ECG) testing is considerably delicate among women. Use of the standard criterion (0.10 mV ST-segment depression for test positivity) with women causes more false-positive responses than with men [1] . A new method, ST-segment depression/heart rate (ST/HR) hysteresis, which integrates the ST/HR analysis of the exercise and recovery phases of the exercise ECG test has recently been introduced and has been shown to improve CAD detection [2-4] . The efficacy of the ST/HR hysteresis has been demonstrated in the male-dominated populations. The objective of this study was to evaluate the diagnostic capabilities of ST/HR hysteresis in the detection of CAD among the women. Comparison of the ST/HR hysteresis was made with ST/HR index and end-exercise ST depression.

MATERIAL AND METHODS

Study population: Subjects were selected from a group of patients who had been referred for exercise testing at Tampere University Hospital, Finland. Study population comprising of 115 female patients. Of those patients 26 had a stenosis of ³50% in at least one of the major coronary arteries verified with coronary angiography. The reference group consists of 89 women who have no history of any cardiac disease, a normal resting ECG, and no anginal type chest pain or cardiac medication. With the probabilistic assessment, the reference group can be assumed to have a very low likelihood (p<0.05) of CAD. There were no patients with left or right bundle branch block and recent myocardial infarction (< 8 weeks) in CAD of reference groups. The characteristics of the study population are given in Table I. Because of the different exclusion criteria used in the selection of the groups, highly significant differences (p < 0.0001) were achieved with all clinical characteristics, with the exception of digitalis.

Exercise electrocardiography: The exercise test was performed on a bicycle ergometer using a computerized recording system. The graded protocol followed standard clinical routine with an initial 40 W workload and an increment of 40 W every 4 minutes. The exercise tests were sign- and symptom-limited maximal tests using recommended criteria for termination; fatigue and/or chest pain were the reasons for termination in most cases. The lead system used was the Mason-Likar modification of the standard 12-lead system. Computer-determined ST-segment amplitudes measured to the nearest 10 µV were obtained at a point 60 ms after the J-junction considering the end of the PR-segment as the isoelectric line.

Diagnostic variables: The ST/HR hysteresis, ST/HR index, and ST segment depression at end exercise (ST_end) were determined for each lead from the digitally stored data with a computer program developed for comprehensive ST/HR analysis [5] . The maximum values derived from 12-lead system (leads V₁, aVR and aVL were excluded) were used as diagnostic variables for each method.

Calculation of the ST/HR index was performed by dividing the overall ST-segment deviation at the end-exercise by the exercise-induced change of HR [6] . ST/HR hysteresis was calculated by integrating the difference in ST depression between the exercise and recovery phases over the HR from the minimum HR of recovery to the maximum HR of the exercise test. The integral was divided by the HR difference over the integration interval in order to normalize the ST/HR hysteresis with respect to the recovery HR decrement. This variable represents the average difference of the ST depression between the recovery phase and the exercise phase [2] .

 Data analysis and statistical methods: Continuous variables are described as mean values ± standard deviations (SD). The diagnostic capabilities of the variables were compared using receiver-operating characteristic (ROC) curve analysis, which allows comparison without fixed partition values. Statistical differences between the areas under two ROC curves were compared using a non-parametric analysis of correlated ROC curves. Comparisons of the different methods at a fixed 85% specificity (equal to standard -0.10 mV of ST_end) were performed by means of McNemar's modification of the chi-square method for paired proportions.

TABLE I.
Clinical characteristics of the study population.

RESULTS

Figure 1 presents areas under the ROC curves for every method. The ST/HR hysteresis provided the best diagnostic performance whilst the ST_end gave the poorest results. Statistical comparison of the ROC curves revealed significant differences between the ST/HR hysteresis and ST_end (p = 0.0103) and between ST/HR index and ST_end (p = 0.0446). No significant difference was achieved in comparison of ST/HR hysteresis and ST/HR index (p = 0.0863).

The sensitivity values at a fixed 85% specificity for ST/HR hysteresis, ST/HR index and ST_end were 85%, 69% and 50%, respectively. Statistical differences in sensitivities were noted when comparing ST/HR hysteresis and ST/HR index with ST_end (p = 0.0010 and p = 0.0313, respectively). Statistical comparison between ST/HR hysteresis and ST/HR index showed no significance difference (p = 0.0625).

Figure 1. Receiver operating characteristics (ROC) curves for ST/HR hysteresis, ST/HR index and end-exercise ST depression using maximum values derived from 12 leads (V1, aVR and aVL excluded). Black marks in curves indicate the nearest cut-off criterion at 85% specificity. Values for ROC areas are percentages of total ROC space.

DISCUSSION

In previous study we have showed the superiority of ST/HR hysteresis over other variables. However the predominance of men has been marked in the study materials. The results of this study suggest that the use of ST/HR hysteresis increases the diagnostic performance of the exercise ECG test in the detection of CAD also in women. The sensitivity of ST/HR hysteresis at a high specificity (>80%), which is the most important portion of the ROC curve for a diagnostic test, was higher than that of the others. ST/HR hysteresis and ST/HR index were significantly more competent than the traditional end-exercise ST-segment depression in the diagnosis of CAD in female patients. However, the number of women with CAD was small, thus the results requires validation in a larger women population.

REFERENCES

[1] G. F. Fletcher, G. Balady, V. F. Froelicher, et al., "Exercise standards. A statement for healthcare professionals from the American Heart Association. Writing Group", Circulation, vol. 91, pp. 580-615, 1995.

[2] R. Lehtinen, H. Sievänen, J. Viik, et al., "Accurate detection of coronary artery disease by integrated analysis of the ST-segment depression/heart rate patterns during the exercise and recovery phases of the exercise electrocardiography test", American Journal of Cardiology, vol. 78, pp. 1002-1006, 1996.

[3] J. Viik, Diagnostic properties of exercise electrocardio-graphic leads and variables in the detection of coronary artery disease, Tampere University of Technology, Tampere, 2000.

[4] J. Viik, R. Lehtinen, V. Turjanmaa, et al., "The effect of lead selection on traditional and heart rate-adjusted ST segment analysis in the detection of coronary artery disease during exercise testing", American Heart Journal, vol. 134, pp. 488-494, 1997.

[5] R. Lehtinen, H. Vänttinen, H. Sievänen, et al., "A computer program for comprehensive ST-segment depression/heart rate analysis of the exercise ECG test", Computer Methods and Programs in Biomedicine, vol. 50, pp. 63-71, 1996.

[6] R. Detrano, E. Salcedo, M. Passalacqua, et al., "Exercise electrocardiographic variables: a critical appraisal", Journal of the American College of Cardiology, vol. 8, pp. 836-847, 1986.