a b s t r a c t

Background: The novel analysis of high-frequency QRS components (HF/QRS) has been proposed in patients with chest pain (CP) referred for exercise tolerance test (ex-ECG). We sought to evaluate the prognostic role of exer- cise high-frequency QRS-analysis (ex-HF/QRS) in patients with recent-onset stable CP, in the emergency setting. Methods: patients with CP underwent ex-ECG. A decrease greater than or equal to 50% of the signal of HF/ QRS intensity was considered as index of ischemia as ST-segment depression greater than or equal to 2 mm or greater than or equal to 1 mm associated with CP. Exclusion criteria were QRS duration greater than or equal to 120 milliseconds and inability to exercise. Baseline characteristics were adjusted with the propensity score matching specifying nearest-neighbor matching in Cardiovascular risk factors and risk scores. The primary end point was the composite of coronary stenosis greater than or equal to 70% or acute coronary syndrome, revascularization, and cardiac death on the 6-month follow-up.

Results: Of 589 patients, 22 achieved the end point. On the univariate analysis, known cardiovascular disease, GRACE score, and ex-HF/QRS were predictors of the end point. On the multivariate analysis, only ex-HF/QRS was predictor of the end point (odd ratio, 28; 95% confidence interval [CI], 6-120; P b .001). Overall, the ex-HF/QRS when compared to ex-ECG showed higher sensitivity (91% vs 27%; P = .02), lower specificity (74% vs 86%; P = .09), and comparable negative predictive value (99% vs 97%; P = .78). receiver operating characteristic curve analysis showed the larger area of ex-HF/QRS (0.83; 95% CI, 0.75- 0.90) over ex-ECG (0.57; CI, 0.44-0.70) and GRACE score (0.65; CI, 0.54-0.76); P b .03 on C-statistic. Women showed the largest area (0.89; CI, 0.83-0.95; P b .03) vs the other clinical data.

Conclusions: In patients with CP, the novel ex-HF/QRS analysis has a valuable incremental prognostic role over ex-ECG, especially in women.

(C) 2015

Introduction

Noninvasive cardiovascular testing is used on a daily basis to decision making for chest pain (CP) patient care. In this setting, the prognostic value for stress-induced ischemia with exercise tolerance test (ex-ECG) is poor when compared to exercise stress echocardiography or exercise stress myocardial perfusion imaging or coronary Computed tomography angiography . Indeed, sensitivity of testing for coronary disease diagnosis is reported up to 50% for ex-ECG compared to 85% for exercise stress echocardiography, 92% for exercise stress myocardial perfusion imaging, and 99% for CTA [1-3]. Data on ex-ECG are far worse in women, when compared to the other testings [4].

* Corresponding author at: Emergency Department, North-West Tuscany HealthCare, NOA General Hospital, Via Martelli 1, Massa, Italy. Tel.: +39 0585493299; fax: +39

0585493344.

E-mail address: [email protected] (A. Conti).

Recent-onset stable CP is a common clinical problem that results in a large number of stress tests in first-line evaluation, without diagnosed heart disease. Accordingly, up to one-third in health care spending usually are for cardiovascular imaging [5-7]. As a consequence, ex-ECG could remain a valuable first-line testing in these patients. Support to this hypothesis derives from recent data which have demonstrated that, in symptomatic patients with suspected coronary artery disease (CAD), an initial strategy with anatomical test as CTA was not associated with better clinical out- comes than functional testing, including stress electrocardiogram (ECG) or stress echocardiography or stress myocardial perfusion im- aging [7]. However, the low predictive value of the standard ex-ECG with the ST-segment analysis needs to be improved, eventually adding the high-frequency mid-QRS analysis [8]. Nevertheless, the value of this novel technique in improving sensitivity and specificity has not been validated yet in emergency care and might be of inter- est in selected populations [9,10].

http://dx.doi.org/10.1016/j.ajem.2015.11.044

0735-6757/(C) 2015

Stress testing“>The aim of the present study was to compare the prognostic value, for CAD or adverse cardiac events, of the ex-HF/QRS analysis to the conventional ST-segment analysis, in patients with recent-onset stable CP in the emergency setting.

Methods

Study population

Unselected consecutive patients suitable for enrollment were those who presented with typical CP consistent for myocardial ischemia, between January 2013 and July 2014. The study setting was the emer- gency department (ED) of the tertiary care teaching Careggi Hospital, in Florence, Italy, with a catchment area serving a population of half- million. Each patient was categorized at baseline with physical examina- tion, instrumental evaluation inclusive of ECG and cardiac biomarkers. The department of cardiology was on call for serial evaluations and choosing wisely the diagnostic strategy. Patients with negative first- line evaluation were subjected to a 6- to 12-hour observation period with serial resting ECGs, serial high-sensitivity Troponin tests, and echocardiography [11-13].

Inclusion and exclusion criteria

The inclusion criterion was the presence of visceral CP consistent with angina pectoris lasting less than 24 hours. Chest pain was charac- terized with a validated CP score inclusive of the characteristics of pain (crushing, pressing, heaviness = 3; sticking, pleuritic, pinprick = 1), localization (substernal or precordial = 3; epigastric, left chest, neck, lower jaw = 1), radiation (as either arm, shoulder, back, neck, lower jaw = 1; absence = 0), the associated symptoms (as dyspnea, nausea, diaphoresis = 2; absence = 0), and eventually the recurrence in the previous 48 hours (yes = 3, no = 0) [14]. We considered patients with a pretest likelihood of CAD between 15% and 85% calculated using a model suggested by Diamond et al and with normal resting ECGs and normal High-sensitivity troponin I levels [2,3,11,15]. Only patients with negative high-sensitivity troponin I, resting ECGs, and echocardi- ography were included in the study.

The exclusion criteria were represented by the presence of hemody- namic instability or Killip class 2 or more and the presence of acute coronary syndrome or stroke defined by the European and North American Guidelines [2,16,17]. In these conditions, patients need specif- ic immediate treatment. Because of the high-frequency QRS compo- nents (HF/QRS) analysis during exercise, we excluded patients with inability to exercise and presenting with cardiac pacemaker or QRS duration greater than or equal to 120 milliseconds.

Tourists and inhabitants outside the catchment area of Careggi Hospital were not enrolled in the study because of the probability to lose them at follow-up.

Each patient gave informed consent to participate in the study and publication of personal data. The study was conducted according to good clinical practice and principles of the Declaration of Helsinki. The institutional review board approved the protocol. Departmental sources supported the work, and no contributorship or competing interest existed.

Management of patients

Twelve-lead ECG and high-sensitivity troponin I were obtained on presentation and after 6 hours. Patients were considered as hav- ing normal ECG in the presence of normal ST-segment, measured at 60 milliseconds from J point, and normal T wave. Patients were con- sidered as having nondiagnostic ECG in the presence of mild changes as ST-segment elevation or depression less than 0.05 mV (0.5 mm) or asymmetrical T-wave inversion less than 0.2 mV (2 mm) and no Q wave or Q waves less than 0.03 seconds. Patients showing

ischemic ECG changes or Troponin elevations or new cardiac wall motion abnormalities at echocardiography were considered at high risk for cardiac events; they were admitted. Patients with negative troponins, ECGs, and echocardiography were considered at interme- diate risk for CAD; they were enrolled and underwent the exercise tolerance test program for conventional ST-segment and ex-HF/ QRS analysis [8,11].

Stress testing

A symptom-limited graded exercise tolerance test was performed within 24 hours from presentation using the Bruce protocol, in washout of drugs. High-resolution 12-lead electrocardiogram (HyperQ Stress System; BSP Ltd, Tel Aviv, Israel) was continuously recorded throughout the exercise test and used for off-line quantitative assessment of ST- segment changes and HF/QRS intensity analysis. The ST-segment was measured 60 milliseconds after the J point using commercial software (HyperQ Stress System) and was considered positive if horizontal slop- ing (ST-segment slope >= 0.5 mm) or downsloping (ST-segment depres- sion >= 2 mm) was found in greater than or equal to 2 contiguous leads. The ex-ECG was considered positive also when ST-segment depression greater than or equal to 1 mm was associated with CP. Patients were ex- cluded from analysis in the presence of leads that exhibited changes considered as noisy. The ex-ECG was performed aiming to reach the 85% of the age-adjusted maximal predicted heart rate at least (% of Max- imal Predicted Theoretical Heart Rate (220 - age in years) x 0.85). The HF/QRS intensity was calculated, and a decrease greater than or equal to 50% of the signal recorded in 2 contiguous leads was considered as an index of ischemia. Peak exercise capacity from the graded test was esti- mated in Metabolic Equivalent Task score, using data from standard predicted equations [11].

Patients recognized as having positive ex-ECG or positive ex-HF/QRS

were considered at high risk for cardiac events; they were admitted and considered for angiography. Patients with negative stress testing were considered at Very low risk for coronary events; they were discharged home and submitted to the follow-up.

End point

The primary end point was the composite of coronary stenosis great- er than or equal to 70% on angiography or the occurrence of ischemic cardiac events inclusive of acute coronary syndrome, revascularization, and Cardiovascular death on the 6-month follow-up.

Follow-up

Follow-up was performed by reviewing the ED access archives and by telephone after 6 months in patients discharged with negative stress test- ing. Data were collected by means of a physician-directed telephone inter- view using a standardized questionnaire. Each cardiac event was analyzed and confirmed after review of clinical charts, ECGs and laboratory tests.

Statistical analysis

Summary data are expressed as absolute numbers and percentage for categorical variables, whereas mean +- SD was used for continuous values. Statistical comparisons of demographic and clinical features were performed using the ?² test (Fisher exact test) when expected fre- quencies were less than 5% and the Pearson exact test for categorical variables, whereas the Student t test was used for continuous variables (Wilcoxon rank sum test). P b .05, from a 2-sided test, was considered to indicate statistical significance. Comparisons were made using the log- rank test or the stratified log-rank test in the propensity score- matched cohort. Because there were differences in baseline characteris- tics between patients with vs without positive stress testing, propensity score matching was used to adjust for possible confounders [18]. SPSS

software allows estimation of the propensity score using logistic regres- sion and specifying nearest-neighbor matching. Detailed balance statis- tics and graphs are produced by the program. A 5:1 matched analysis based on the propensity score of each patient was conducted. To ensure good matches, a caliper (maximum allowable difference between 1 par- ticipants) of 0.15 was defined. Variables included in the model were age, sex, hypertension, hypercholesterolemia, diabetes mellitus, active smoking, known cardiovascular disease, chronic renal disease, heart rate, systolic arterial pressure, TIMI risk score [19], GRACE score [2,17], CHA2DS2VASc [20], and CHADS2 score [21]. Cox analysis regression model was performed to identify independent predictors for primary end point. Sensitivity analyses using backward logistic regression for all the clinical variables and comorbidities considered in the study were carried out; the clinical variables which were found to have a probability value less than 0.05 were subjected to multivariate back- ward logistic regression analysis. Odd ratios were used to illustrate the probabilities of adverse events. In addition, to estimate the Predictive power of single tests for predicting the presence or absence of disease, sensitivity, specificity, positive (+) and negative (-) predictive values, and likelihood ratios (LR) [(+)LR = sensitivity/(1 - specificity); (-)LR = specificity/(1 - sensitivity)] were calculated considering follow-up data. Receiver operating characteristic curve analysis was obtained for patients who were found as having positive stress testing to detect sensitivity-specificity ratio. Comparison of ROC curves was per- formed using pairwise C-statistic test. Calculations were performed using SPSS version 21 (SPSS, Inc, Chicago, IL) for all analyses.

Results

Study population

Of 662 patients considered, 40 were excluded due to the presence of wide QRS (n = 11) or noisy on ST-segment (n = 29). Thus, 622 patients were enrolled (mean age, 60 +- 15 years; range, 18-87 years). Baseline clinical characteristics after the propensity score matching and the chart of time and outcomes of patients enrolled in the study are shown in Table 1 and Fig. 1.

Factors associated with adverse outcome

Overall, 22 patients achieved the Composite end point. On the uni- variate analysis, the presence of known cardiovascular disease, the

GRACE score, and the positive ex-HF/QRS were predictors of the end point. However, only the ex-HF/QRS (odd ratio [OR], 28; 95% confidence interval [CI], 6-120; P b .001) was predictor of the end point on the mul- tivariate analysis (Table 2).

Stress testing

The final analysis covers 165 patients who were recognized as having 1 in 2 positive stress testing; 22 patients eventually achieved the end point (Fig. 1). The ex-HF/QRS showed higher sensitivity (95% CI, 91% vs 27%; P = .02), lower specificity (95% CI, 74% vs 86%; P =

.09) and comparable negative predictive value (95% CI, 99% vs 97%; P = .78) when compared to ex-ECG. The number of true-positive, false-positive, true-negative, and false-negative patients for each stress testing is shown in the Table 3. Receiver operating characteris- tic curve analysis demonstrated the incremental prognostic value of the ex-HF/QRS over clinical risk scores and conventional ex-ECG (Fig. 2). Indeed, ROC analysis showed the larger area of the HF/QRS (0.83; 95% CI, 0.75-0.90) over the ex-ECG (0.57; CI, 0.44-0.70; P =

.003), the GRACE score (0.65; CI, 0.54-0.76; P = .03), the CHA2DS2VASc score (0.57; CI, 0.45-0.69; P = .03), and the TIMI

score (0.56; CI, 0.43-0.69; P = .03). The areas of the ROC curves were compared with the pairwise C-statistic: the ex-HF/QRS showed a statistically significant larger area than others (P b .03) (Fig. 2). In women, the ROC analysis of ex-HF/QRS showed a larger area (0.89; CI, 0.83-0.95) than men (0.025) but not in other selected subsets of patients including age older than 65 years (P = .165), hypertension (P = .082), and dysglycemia (P = .770), inclusive of diabetes mellitus and Metabolic syndrome (Table 4).

Discussion

At the best of our knowledge, the present study is the first that inves- tigates, in the emergency setting, the relationship between the presence of occlusive CAD and the prognostic value of the novel stress testing mo- dality with the high-frequency QRS mid-components analysis. Because of the poor diagnostic value of traditional exercise tolerance test, when compared to imaging, this novel testing modality could adds to the body of evidence in CP diagnostics in the ED, eventually in special subset of patients with metabolic syndrome, female sex, and hyperten- sion. Consecutive patients were enrolled in the present study and repre- sent a nonselected cohort of patients evaluated in the emergency

Table 1

Baseline characteristics of patients with recent-onset stable CP enrolled in the study and submitted to the propensity score matching (n = 589) according to the results of ex-HF/QRS and outcome

Parameters Propensity-matched 1:5 cohort

	Total	Negative end	Positive end	P	Total	Negative testing	Positive testing	P
	(n = 589)	point (n = 567)	point (n = 22)		(n = 589)	(n = 424)	(n = 165)
Mean age (years +- SD)	61 +- 15	61 +- 15	65 +- 11	.140	61 +- 15	61 +- 15	61 +- 14	.881
Female sex, n (%)	242 (41)	236 (42)	6 (27)	.180	242 (41)	183 (43)	59 (36)	.101
Hypertension, n (%)	306 (52)	291 (51)	15 (68)	.120	306 (52)	207 (49)	99 (60)	.015
Diabetes mellitus, n (%)	63 (11)	59 (10)	4 (18)	.247	63 (11)	43 (10)	20 (12)	.485
Hypercholesterolemia, n (%)	204 (35)	196 (35)	8 (36)	.964	204 (35)	146 (34)	57 (35)	.823
Active smokers, n (%)	131 (22)	127 (22)	4 (18)	.641	131 (22)	92 (22)	39 (24)	.611
Known cardiovascular disease, n (%)	89 (15)	80 (14)	9 (41)	.086	89 (15)	55 (13)	34 (21)	.020
Known coronary artery disease, n (%)	86 (15)	80 (14)	6 (27)	.086	86 (15)	55 (13)	31 (19)	.073
Familiarity for coronary artery disease, n (%)	134 (23)	129 (23)	5 (23)	.998	134 (23)	90 (21)	44 (27)	.157
Metabolic syndrome, n (%)	45 (8)	44 (8)	1 (5)	.578	45 (8)	35 (8)	10 (6)	.368
Max heart rate, beat per minute (mean +- SD)	141 +- 19	141 +- 19	136 +- 20	.199	141 +- 19	141 +- 19	139 +- 18	.136
% of theoretical maximal heart rate per minute (mean +- SD)	88 +- 10	88 +- 10	87 +- 12	.635	88 +- 10	88 +- 9	87 +- 10	.189
METs, MET (mean +- SD)	6.6 +- 1.4	6.6 +- 1.4	6.2 +- 1.3	.203	6.6 +- 1.4	6.6 +- 1.5	6.2 +- 1.3	.728
CHA2DS2VASc score N 1, n (mean +- SD)	351 (60)	335 (59)	16 (73)	.201	351 (60)	242 (57)	109 (66)	.046
TIMI score N 2, n (%)	116 (20)	109 (19)	7 (32)	.523	116 (20)	71 (17)	45 (27)	.004
Chest pain score N 6, n (%)	178 (30)	170 (30)	8 (36)	.523	178 (30)	123 (29)	55 (33)	.305
Max systolic arterial pressure, mm Hg (mean +- SD)	169 +- 22	169 +- 21	175 +- 28	.232	169 +- 22	169 +- 21	170 +- 22	.560
Body mass index, n (mean +- SD)	26 +- 5	26 +- 4	28 +- 6	.150	26 +- 5	26 +- 5	27 +- 4	.150
GRACE score, n (mean +- SD)	72 +- 26	72 +- 26	86 +- 24	.013	72 +- 26	72 +- 26	70 +- 28	.823

Exercise ToleranceTest n = 662

Table 3

The number of true-positive, false-positive, true-negative, and false-negative patients for each stress testing

Endpoint Endpoint n = 14 n = 2

Endpoint Endpoint n = 6 n = 2

ex-HFQRS n = 132

ex-ECG n = 33

Negative: ex-ECG and ex-HFQRS n = 424

Positive: ex-ECG or ex-HFQRS n = 165

Positive: ex-ECG and ex-HFQRS

n = 83

Stress testing	True negative	False negative	True positive	False positive
ex-HF/QRS	422	2	145	20
ex-ECG	537	20	30	2

ex-HF/QRS noisy n = 29

ex-QRS?120 ms n = 11

After Propensity Score Matching n = 589

Fig. 1. The chart of time to management and outcome of patients with recent-onset stable CP enrolled in the study (n = 589) according to the occurrence of positive or negative test- ing and incidence of the primary end point, inclusive of myocardial infarction, revascular- ization, and cardiac death.

settings, with CP as the major complain associated with negative serial resting ECGs, serial high-sensitivity troponin, and echocardiography.

Main findings

The present study evaluates the novel HF/QRS analysis in exercise- induced QRS changes as a marker of myocardial ischemia. In patients with CP at intermediate risk, who eventually may have a hidden

Table 2

univariate and multivariate analyses in patients with recent-onset stable CP enrolled in the study and submitted to 6-month follow-up (n = 589)

propensity-matched cohort (n = 589) Univariate analysis Multivariate analysis OR 95% CI P OR 95% CI P

Positive ex-HF/QRS Known cardiovascular disease	29.1 4.2	6.7-126.0 1.7-10.2	b.001 .001	27.6 2.5	6.3-120.4 0.9-6.6	b.001 .064	that the decrease in high-frequency QRS intensity precedes and is more sensitive than ST-segment changes in animal models of coro-
GRACE	1.0	1.0-1.0	.015	1.0	1.0-1.0	.062	nary occlusion [9,10] and in patients with acute myocardial infarc-

coronary atherosclerosis and will show a substantial proportion of ad- verse cardiac events on the follow-up [22], improvements in sensitivity and specificity have been reported when ex-HF/QRS analysis was added to ex-ECG, although in elective evaluations [8]. Indeed, the results of the present study provide information to support the evidence that the ex- HF/QRS more likely allows the detection of high-risk patients for ad- verse cardiac events when compared to ex-ECG alone. In our cohort of patients, the sensitivity of ex-HF/QRS was 3-fold higher than ex-ECG (P = .02), whereas the specificity and predictive value were compara- ble. In addition, the ROC analysis showed a significant larger area of the ex-HF/QRS (0.83) over ex-ECG (0.57; P = .003) and over the validat- ed risk scores as GRACE score (0.65), CHA2DS2VASc score (0.57), and TIMI score (0.56) (P b .03). Interestingly, in women vs men, the ROC analysis showed a larger area (0.89) of the ex-HF/QRS than ex-ECG (P = .025).

Computed-assisted high-frequency ECG analysis

The activation of myocardial fibers is obtained via myocyte-to- myocyte conduction, the signal spreading away from the ends of the Purkinje fibers. The depolarization wavefront, which creates the QRS complex, is fragmented on the microscopic level and brings about low-amplitude, high-frequency notches on the QRS. Ischemia slows down myocyte-to-myocyte conduction and reduces the fragmentation of the depolarization wavefront and, consequently, the high-frequency QRS intensity [8-10]. The novel electrocardiographic technique allows the detection of stress-induced ischemia with the examination of the high-frequency QRS intensity in the 150- to 250-Hz frequency band. Because the high-frequency QRS signals are very low in amplitude, eventually, advanced signal processing is required to extract and inter- pret the value of physiological information embedded in them. Thus, the detection of depolarization changes during exercise using an automated analysis of high-frequency QRS components allows detection of ische- mia than identification of ST-segment deviations. Indeed, traditional exercise ECG testing is based on ST-segment analysis, yet with limited diagnostic accuracy.

Literature

Usually, previous conventional techniques for ECG detection of exercise-induced myocardial ischemia have relied on identification of Repolarization abnormalities as ST-segment changes in the 0.05- to 100-Hz frequency band [23,24]. Recent studies have suggested

tion [25]. In the most important and large available study [8], the HF/QRS analysis was compared to conventional ST-segment analysis

Positive ex-ECG testing	1.9	0.8-4.5	.140
TIMI N 2	2.0	1.0-1.7	.152
CHA2DS2VASc N 1	1.8	0.7-1.8	.207

FPR N 3	0.9	0.4-2.2	.857	in patients with greater than 10% ischaemic myocardium detected by
CP score N 6	1.3	0.6-3.2	.524	the nuclear scan imaging. In the study of Sharir et al [8], the HF/QRS
Age Hypertension Female sex Diabetes mellitus Hypercholesterolemia	1.0 2.0 0.5 1.9 1.1	1.0-1.1 0.8-5.1 0.2-1.4 0.6-5.8 0.4-2.6	.142 .128 .186 .255 .877	analysis was more sensitive (69% vs 39%; P b .005) and more specific (86% vs 82%; P b .05) than the conventional ex-ECG testing. Conversely, in the present study, we have considered HF/QRS with- out any thresholds for abnormality, and then we have compared
Active smoker	0.8	0.3-2.3	.642	ex-HF/QRS to ex-ECG, with conventional ST-segment analysis. Only

a hard end point was considered, including the presence of occlusive CAD and Major adverse cardiac events. The HF/QRS analysis was

known CAD	2.3	1.9-6.0	.095
Familiarity for CAD	1.0	0.4-2.8	.998

4.5. Strengths and limitations of the study

The present study enrolled a large cohort of consecutive patients with recent-onset stable CP and intermediate pretest likelihood of CAD, representing the population usually referred for ex-ECG and the substantial portion of patients presenting to the ED with CP. The retro- spective data were managed in a prospective fashion due to the propen- sity score matching for possible confounders in baseline characteristics. Indeed, SPSS software allowed estimation of the propensity score using logistic regression and specifying nearest-neighbor matching in cardio- vascular risk factors, CP score, and TIMI score, GRACE score, CHA2DS2VASc score.

However, the present study shows several limitations, the most important of which is represented by the single-center design of enrol- ment. In addition, nevertheless the propensity score elaboration, the need of a properly designed study with randomization to diagnostic strategy could be considered a strong limitation. Eventually, the out- come evaluation based on dichotomy (normal/abnormal tests) may be a limitation of any screening workup in patients with CP.

Fig. 2. Comparison among the areas under the curve of ROC of the ex-HF/QRS, the ex-ECG, and the GRACE score, in 589 CP patients submitted to exercise tolerance test. Exercise-HF/ QRS vs ex-ECG: P = .003; ex-HF/QRS vs GRACE score: P b .03; and GRACE score vs ex-ECG: P = nonsignificant.

superior to traditional ST-segment analysis and Clinical scores, espe- cially in women. Of note, the predictive values of ST-segment analy- sis in this study were substantially within the range reported in previous studies.

Future clinical application of ex-HF/QRS analysis

Our study provides a picture update of the presence of CAD and the risk of short-term cardiac events in patients with recent-onset stable CP on presentation and confirms that masked CAD plays a pivotal role in a substantial proportion of these patients. Recent-onset stable CP can be related eventually to mild-to-moderate or even moderate-to-severe myocardial ischemia, nevertheless in patients presenting with negative serial resting ECGs, serial troponins, and echocardiography. Indeed, in our cohort of patients, the positive stress testing was detected in 165 pa- tients, and eventually, the hard end point was reached in 22 patients. Thus, in acute cardiac care, the computed-assisted stress testing HF/ QRS analysis might aid physicians in stratifying CP patients overcoming the poor predictive values of traditional ST-segment analysis during ex- ercise tolerance test. Finally, the ex-HF/QRS analysis should drive the threshold approach to clinical decision making over the ex-ECG alone.

Table 4 Areas under the curve of ROC according to sex, age, clinical characteristics, or risk scores, in 589 CP patients enrolled in the study: comparison of the prognostic results of the ex-HF/ QRS vs the ex-ECG

Area under ROC

	Ex-HF/QRS	Ex-ECG		P
Total	0.83	0.75-0.90	0.57	0.44-0.70	.003
Women	0.89	0.83-0.95	0.49	0.26-0.73	.025
Age >= 65 y	0.83	0.73-0.92	0.49	0.33-0.65	.156
Hypertension	0.82	0.74-0.91	0.49	034-0.64	.082
Dysglycemia	0.77	0.56-0.98	0.50	0.24-0.75	.770
GRACE N 108	0.65	0.54-0.76	0.45	0.07-0.83	.272
CHA2DS2VASc N 1	0.65	0.54-0.77	0.52	0.37-0.67	.018
TIMI N 2	0.63	0.52-0.74	0.55	0.32-0.79	.350
CP score N 6	0.62	0.53-0.72	0.67	0.46-0.88	.501

Conclusions

In patients with recent-onset stable CP and nondiagnostic first-line workup including serial ECGs, high-sensitivity troponins, and echocar- diography submitted to exercise tolerance test, the novel ex-HF/QRS analysis showed a substantial incremental prognostic value over the ex-ECG with the conventional ST-segment analysis and over the validat- ed clinical risk scores. The results were of valuable interest especially in women.

References

1. Demir OM, Alfakih K, Plein S. Current international guidelines for the investigation of patients with suspected coronary artery disease. Eur Heart J Cardiovasc Imaging 2014;15(12):1422-4.
2. Montalescot G, Sechtem U, Achenbach S, Andreotti F, Arden C, Budaj A, et al. 2013 ESC guidelines on the management of stable coronary artery disease: the task force on the management of stable coronary artery disease of the European Society of Cardiology. Eur Heart J 2013;34(38):2949-3003.
3. Fihn SD, Gardin JM, Abrams J, Berra K, Blankenship JC, Dallas AP, et al. 2012 ACCF/ AHA/ACP/AATS/PCNA/SCAI/STS Guideline for the diagnosis and management of pa- tients with stable ischemic heart disease: a report of the American College of Cardi- ology Foundation/American Heart Association Task Force on Practice Guidelines, and the American College of Physicians, American Association for Thoracic Surgery, Preventive Cardiovascular Nurses Association, Society for Cardiovascular Angiogra- phy and Interventions, and Society of thoracic surgeons. J Am Coll Cardiol 2012; 60(24):e44-164.
4. Daugherty SL, Magid DJ, Kikla JR, Hokanson JE, Baxter J, Ross CA, et al. Gender differ- ences in the prognostic value of exercise treadmill test characteristics. Am Heart J 2011;161(5):908-14.
5. Carr JJ, Hendel RC, White RD, Patel MR, Wolk MJ, Bettmann MA, et al. 2013 appropri- ate utilization of cardiovascular imaging: a methodology for the development of joint criteria for the appropriate utilization of cardiovascular imaging by the American College of Cardiology Foundation and American College of Radiology. J Am Coll Cardiol 2013;61(21):2199-206.
6. Kramer CM. Cardiovascular imaging and outcomes-PROMISEs to keep. N Engl J Med 2015;372:1366-7.
7. Douglas PS, Hoffmann U, Patel MR, Mark DB, Al-Khalidi HR, Cavanaugh B, et al. Out- comes of anatomical versus functional testing for coronary artery disease. N Engl J Med 2015;372:1291-300.
8. Sharir T, Merzon K, Kruchin I, Bojko A, Toledo E, Asman A, et al. Use of electrocardio- graphic depolarization abnormalities for detection of stress-induced ischemia as de- fined by myocardial perfusion imaging. Am J Cardiol 2012;109(5):642-50.
9. Abboud S, Berenfeld O, Sadeh D. Simulation of high-resolution QRS complex using a ventricular model with a fractal Conduction system-effects of ischemia on high- frequency QRS potentials. Circ Res 1991;68(6):1751-60.
10. Watanabe T, Yamaki M, Tachibana K, Kubota I, Tomoike H. Decrease in the high fre- quency QRS components depending on the local conduction delay. Jpn Circ J (Eng Ed) 1998;62(11):844-8.
11. Amsterdam EA, Kirk JD, Bluemke DA, Diercks D, Farkouh ME, Garvey JL, et al. Testing of low-risk patients presenting to the emergency department with chest pain a sci- entific statement from the American Heart Association. Circulation 2010;122(17): 1756-76.
12. Erhardt L, Herlitz J, Bossaert L, Halinen M, Keltai M, Koster R, et al. Task force on the management of chest. Eur Heart J 2002;23(15):1153-76.
13. Douglas PS, Khandheria B, Stainback RF, Weissman NJ, Brindis RG, Patel MR, et al. ACCF/ASE/ACEP/ASNC/SCAI/SCCT/SCMR 2007 appropriateness criteria for transtho- racic and transesophageal echocardiography: a report of the American College of Cardiology Foundation Quality Strategic Directions Committee Appropriateness Criteria Working Group, American Society of Echocardiography, American College of Emergency Physicians, American Society of Nuclear Cardiology, Society for Car- diovascular Angiography and Interventions, Society of Cardiovascular Computed Tomography, and the Society for Cardiovascular Magnetic Resonance endorsed by the American College of Chest Physicians and the Society of Critical Care Medicine. J Am Coll Cardiol 2007;50(2):187-204.
14. Conti A, Sammicheli L, Gallini C, Costanzo EN, Antoniucci D, Barletta G. Assessment of patients with Low-risk chest pain in the emergency department: head-to-head comparison of exercise stress echocardiography and exercise myocardial SPECT. Am Heart J 2005;149(5):894-901.
15. Diamond GA, Forrester JS, Hirsch M, Staniloff HM, Vas R, Berman DS, et al. Applica- tion of conditional probability analysis to the clinical diagnosis of coronary artery disease. J Clin Invest 1980;65(5):1210-21.
16. Steg G, James SK, Atar D, Badano LP, Blomstrom-Lundqvist C, Borger MA, et al. ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation. Eur Heart J 2012;33(20):2569-619.
17. Hamm CW, Bassand JP, Agewall S, Bax J, Boersma E, Bueno H, et al. ESC Guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J 2011;32(23):2999-3054.
18. Austin PC, Small DS. The use of bootstrapping when using propensity-score matching without replacement: a simulation study. Stat Med 2014;33(24): 4306-19.
19. Antman EM, Cohen M, Bernink PJ, McCabe CH, Horacek T, Papuchis G, et al. The TIMI risk score for unstable angina/non-ST elevation MI: a method for prognostication and therapeutic decision making. JAMA 2000;284(7):835-42.
20. Lip GY, Frison L, Halperin JL, Lane DA. Identifying patients at high risk for stroke de- spite anticoagulation: a comparison of contemporary stroke risk stratification schemes in an anticoagulated atrial fibrillation cohort. Stroke 2010;41(12):2731-8.
21. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: results from the National Reg- istry of Atrial Fibrillation. JAMA 2001;285(22):2864-70.
22. Eagle KA, Ginsburg GS, Musunuru K, Aird WC, Balaban RS, Bennett SK, et al. Identi- fying patients at high risk of a cardiovascular event in the near future current status and future directions: report of a National Heart, Lung, and Blood Institute Working Group. Circulation 2010;121(12):1447-54.
23. Higgins JP, Higgins JA. Electrocardiographic exercise stress testing: an update be- yond the ST segment. Int J Cardiol 2007;116(3):285-99.
24. Kligfield P, Lauer MS. Exercise electrocardiogram testing-beyond the ST segment. Circulation 2006;114(19):2070-82.
25. Pettersson J, Pahlm O, Carro E, Edenbrandt L, Ringborn M, Sornmo L, et al. Changes in high-frequency QRS components are more sensitive than ST-segment deviation for detecting acute coronary artery occlusion. J Am Coll Cardiol 2000;36(6):1827-34.