Vectorcardiographic and electrocardiographic criteria to distinguish new and old left bundle branch block

Vectorcardiographic and electrocardiographic criteria to distinguish new and old left bundle branch block Alexei Shvilkin, MD, PhD,*‡ Bosko Bojovic, PhD,† Branislav Vajdic, PhD,† Ihor Gussak, MD, PhD,† Kalon K. Ho, MD,‡ Peter Zimetbaum, MD,‡ Mark E. Josephson, MD‡ From the *South Shore Hospital, South Weymouth, Massachusetts, †NewCardio, Inc., Santa Clara, California, and ‡ Department of Medicine/Division of Cardiology, Beth Israel Deaconess Medical Center, Boston, Massachusetts. BACKGROUND There are no established criteria to differentiate new from old left bundle branch block (LBBB). This complicates management of patients with LBBB and suspected acute coronary syndrome. OBJECTIVES The purpose of this study was to develop electrocardiographic (ECG) criteria to differentiate new and old LBBB. METHODS All LBBB tracings (n ⫽ 3,706) in a hospital ECG database were retrieved. New (⬍24 hours, n ⫽ 39) and old (⬎24 hours, n ⫽ 1,760) LBBB tracings were identified. QRS and T-wave amplitudes, directions, and durations were measured digitally. Vectorcardiograms were reconstructed from 12-lead ECGs using inverse Dower transform and analyzed with Cardio3KG software. Receiver operator characteristic (ROC) curves were used to develop decision rules to distinguish new and old LBBB. RESULTS The new LBBB group had larger T-vector magnitude (1.20 ⫾ 0.07 vs. 0.71 ⫾ 0.01 mV), smaller QRS vector magnitude (2.13 ⫾ 0.12 vs. 2.47 ⫾ 0.02 mV), and a lower QRS/T vector magnitude ratio (QRS/T; 1.79 ⫾ 0.03 vs. 3.92 ⫾ 0.04) compared with the old LBBB group (mean ⫾ standard error of the mean,

Introduction There are no established electrocardiographic (ECG) criteria to differentiate new from old left bundle branch block (LBBB), and patients with LBBB presenting to the emergency room with chest pain pose a significant diagnostic challenge.1–3 Current American College of Cardiology/ American Heart Association guidelines for the management of patients with ST elevation myocardial infarction (MI) consider new or presumed new LBBB associated with symptoms that are suggestive of ischemia a class I indication for reperfusion therapy4 based on increased mortality in this patient population.5,6

Drs. Shvilkin, Zimetbaum, and Josephson own stock options of NewCardio, Inc. Drs. Bojovic, Vajdic, and Gussak are employees and Dr. Josephson is the chairman of the Scientific Advisory Board of NewCardio, Inc. Address reprint requests and correspondence: Alexei Shvilkin, M.D., Ph.D., Baker4/Cardiology, Beth Israel Deaconess Medical Center, 185 Pilgrim Road, Boston, Massachusetts 02215. E-mail address: [email protected]. (Received April 29, 2010; accepted May 15, 2010.)

P ⬍.001). The ratio of deepest S to largest T wave in precordial leads (Max S/T) was significantly smaller in the new compared with in the old LBBB group (1.66 ⫾ 0.05 vs. 3.54 ⫾ 0.08; P ⬍.001). A decision rule using QRS/T ⬍2.25 and Max S/T ⬍2.5 had 100% sensitivity and 96%– 68% specificity in diagnosing new LBBB, including subsets of patients with tachycardia and ischemia. CONCLUSIONS QRS/T and Max S/T allow accurate discrimination between new and old LBBB suitable for both computerized and manual analysis. If confirmed in prospective studies, this finding can improve management of patients with chest pain and LBBB. KEYWORDS Left bundle branch block; Acute coronary syndrome; Electrocardiography; Cardiac memory; Vectorcardiography ABBREVIATIONS AUC ⫽ area under the curve; CM ⫽ cardiac memory; ECG ⫽ electrocardiogram; EMR ⫽ electronic medical record; HR ⫽ heart rate; LBBB ⫽ left bundle branch block; MI ⫽ myocardial infarction; ROC ⫽ receiver operator characteristic; TnT ⫽ Troponin T; VCG ⫽ vectorcardiogram (Heart Rhythm 2010;7:1085–1092) © 2010 Heart Rhythm Society. All rights reserved.

Recently, the validity of these recommendations has been challenged in several studies that demonstrated a low prevalence of MI and mortality in LBBB patients presenting to the emergency room with chest pain.7,8 One of the possible reasons for this discrepancy is the inclusion of patients with “presumed new LBBB,” which in fact is old, but prior ECGs are lacking. Therefore, the ability to determine whether LBBB is new or old using a single ECG tracing would facilitate the triage of patients with chest pain and LBBB and potentially help to avoid unnecessary cardiac interventions and reduce complications and the overall cost of care. Abnormal ventricular activation such as ventricular pacing or transient LBBB produces electrical remodeling known as cardiac memory (CM),9 which manifests as Twave inversions upon resolution of aberrant conduction.9,10 Recently, using a vectorcardiographic (VCG) approach in a human pacing-induced CM, we demonstrated that the repolarization changes associated with it can be observed not only when normal ventricular activation is restored but even when abnormal activation (e.g., ventricular pacing) contin-

1547-5271/$ -see front matter © 2010 Heart Rhythm Society. All rights reserved.

doi:10.1016/j.hrthm.2010.05.024

1086 ues. These changes manifest as a reduction in T-vector amplitude of the paced beat with increased duration of pacing.11 We hypothesized that similar T-wave changes should occur with increased duration of aberrant conduction in LBBB and that the T-vector magnitude of the old LBBB will be smaller than the new one. Since electrical remodeling of repolarization in LBBB develops largely within the first 24 hours,9 the difference in T-vector magnitude might help differentiate the acute onset LBBB from the more chronic one lasting over 24 hours.

Methods The study was approved by the Institutional Review Board of South Shore Hospital, South Weymouth, Massachusetts. South Shore Hospital is a 266-bed community hospital performing over 200 percutaneous coronary interventions for acute MI annually.

ECG collection and LBBB classification We performed a retrospective search of a hospital digital ECG database (GE MUSE, GE Healthcare, Waukesha, WI) to identify all cases of LBBB for the period 2007–2009 using the MUSE Resting ECG diagnostic statement 460, “LBBB.” The computer diagnosis was confirmed by manual overread using the accepted LBBB criteria (QRS duration ⬎120 ms and slurred R waves in leads I and v5-v6, with the absence of Q waves and R peak time ⬎0.06 seconds).12 Tracings from a patient were excluded if there was evidence of a gradual progression of conduction abnormalities over time (e.g., left ventricular hypertrophy with secondary repolarization changes),13 ventricular pacing on any of the tracings, or atrial rhythm other than sinus. Twenty-four-hour duration was selected as a cutoff between new and old LBBB. LBBB was classified as “new” if a tracing satisfied one of the following conditions: 1. A prior ECG with normal QRS duration (⬍110 ms)12 within 24 hours before the LBBB tracing without Twave abnormalities. 2. Acute-onset illness with LBBB on the admission tracing resolving within 24 hours without T-wave abnormalities on the subsequent narrow QRS tracings (to exclude LBBB lasting more than 24 hours)9 in patients with no history of LBBB. LBBB was classified as “old” if it was known to exist for more than 24 hours (by prior tracings or reports in the electronic medical record (EMR). LBBB on tracings obtained within the first 24 hours in patients with no prior ECG information was classified as “unknown duration.” Clinical data including demographics and troponin T (TnT) values were collected from the EMR. In patients with multiple ECGs, age was reported at the date of the first recording. The highest level of TnT for the 24 hours after the time of ECG recording was used for analysis to account

Heart Rhythm, Vol 7, No 8, August 2010 for the time lag between ECG changes and biochemical signs of myocardial damage.

ECG analysis Digital signals were extracted from the database using Magellan ECG Research Workstation Software Ver. 1.1 (GE Marquette, Milwaukee, WI) and analyzed using Cardio3KG software (NewCardio, Inc., Santa Clara, CA).14 Cardio3KG calculates the approximation of the Frank leads X, Y, and Z from the standard 12-lead ECG input using the inverse Dower transformation,15 normalizes the signal for attenuation,14 creates a VCG, and provides tools for three-dimensional vector loop analysis. The following VCG parameters were analyzed (Figure 1): peak QRS and T-vector magnitudes (calculated as 兹X2 ⫹ Y 2 ⫹ Z2) and directions, azimuth (␸) and elevation (␪); peak QRS/T vector magnitude ratio (QRS/T); and peak QRS-T angle. Detailed definitions and illustrations of azimuth and elevation measurements were published previously.11,16 Digitally calculated R-, S-, and T-wave amplitudes in 12 leads and QRS, QT, and JT durations were obtained using the “export parameters” option output of the MUSE database. QT and JT correction were performed using Bazett’s formula.17

Statistics The independent samples t-test was used to compare continuous variables between the groups. The ␹2-square test was used to compare nominal scale data. We used stepwise linear regression modeling to identify factors associated with T-vector magnitude and calculate correlation coefficients. ROC curve analysis was used to construct decision rules to distinguish between new and old LBBB (SPSS ver. 17.0, SPSS, Inc. Chicago, IL). Circular scale data comparison was performed using the WatsonWilliams F test (Oriana 3.0, KCS, Inc., Agnlesey, UK). A double-sided P ⬍.05 was considered statistically significant. Data are presented as mean ⫾ standard error of the mean unless stated otherwise. The authors had full access to the data and take responsibility for its integrity. All authors have read and agree to the manuscript as written.

Results A total of 3,706 tracings were retrieved from the MUSE database. After a manual overread, 759 of them were excluded, leaving 2,947 ECGs to be used for further analysis. There were 39 tracings in the new LBBB group (24 patients), 1,760 in the old LBBB group (448 patients), and 1,148 tracings with LBBB of unknown duration (847 patients).

Baseline data (Table 1) Patients with new LBBB were younger, more likely to be male, and more likely to have positive TnT levels compared with the old LBBB group (corresponding TnT data were available for 93% of tracings).

Shvilkin et al

New and Old LBBB

1087

Figure 1 Orthogonal VCG projections (transverse, frontal, left sagittal planes). VCG measurements demonstrated peak QRS and T vector magnitudes, azimuth (␸), and elevation (␪). Left: new LBBB (⬍6 hours’ duration). Right: old LBBB (same patient 15 days later).

ECG and VCG data (Table 2) T-vector magnitude in the new LBBB group was larger, and QRS vector magnitude smaller compared with the old LBBB group. As a result, the QRS/T in the new LBBB group (1.79 ⫾ 0.03) was significantly lower compared with that in the old LBBB group (3.92 ⫾ 0.04; P ⬍.001). No differences were found in the spatial QRS and Tvector orientation (elevation and azimuth) or QRS-T angle between the groups. The new LBBB group had higher average heart rate (HR; 93 ⫾ 3 min⫺1) compared with the old (77 ⫾ 0.4 min⫺1) LBBB group. QRS and QTc durations in the new LBBB group were longer, but JTc intervals were not significantly Table 1

Baseline data

No. of tracings No. of patients Tracings per patient, range (median) Age Male gender (%) TnT ⬎0.2 ng/mL, n (%) *P ⬍.05 new vs. old LBBB.

New LBBB

Old LBBB

39 24 1–6* (1)

1,760 448 1–44 (2)

66.6 ⫾ 2.7* 16 (66)* 12 (34)*

76.8 ⫾ 0.5 160 (36) 167 (13.4)

different, suggesting that QTc differences were attributed to longer QRS duration in the new LBBB group. Representative ECG examples of the new and old LBBB are shown in Figure 2. Panel A demonstrates the progression in the same patient from the new (⬍6 hours) to old (15 days) LBBB. It is associated with the decrease in T-wave amplitude and increase in QRS amplitude and QRS/T ratio from 1.64 to 3.22 (see corresponding VCGs in Figure 1). Panel B demonstrates resolution of the new LBBB (QRS/T ⫽ 1.63) within less than 12 hours producing no visible T-wave abnormalities upon resumption of normal conduction. In contrast, resolution of an old LBBB, panel C (duration ⬎ 48 hours, QRS/T ⫽ 3.67), results in T-wave inversions consistent with CM.18 These examples highlight the relationship between T-vector magnitude decrease with LBBB progression and T-wave inversion development associated with CM upon normalization of conduction. We used stepwise linear regression modeling to identify factors associated with T-vector magnitude. In the new LBBB group, there was a strong correlation between peak T and QRS vector magnitudes (r ⫽ 0.95; P ⬍.001), which remained the only significant variable in the equation, explaining over 90% of T-vector magnitude variance. In the old LBBB group, this correlation, although still significant,

1088 Table 2

Heart Rhythm, Vol 7, No 8, August 2010 Electrocardiographic and VCG findings New LBBB (n ⫽ 39)

Old LBBB (n ⫽ 1,760)

HR, min⫺1 (range) 93 ⫾ 3* (52–134) 77 ⫾ 0.4 (50–138) QRS vector magnitude, 2.13 ⫾ 0.12* 2.47 ⫾ 0.02 mV T-vector magnitude, mV 1.20 ⫾ 0.07† 0.71 ⫾ 0.01 QRS/T 1.79 ⫾ 0.03† 3.92 ⫾ 0.04 83.0 ⫾ 2.3 80.8 ⫾ 0.2 QRS elevation (␪), degrees QRS azimuth (␺), ⫺81.9 ⫾ 3.4 ⫺77.7 ⫾ 0.4 degrees T elevation (␪), degrees 86.2 ⫾ 2.7 87.2 ⫾ 0.3 T azimuth (␺), degrees 82.7 ⫾ 3.4 92.7 ⫾ 0.7 QRS ⫺ T angle, degrees 158.1 ⫾ 2.6 156.2 ⫾ 0.6 12-lead Max precordial 2.56 ⫾ 0.17 2.76 ⫾ 0.01 S-wave amplitude, mV 12 lead Max precordial 1.57 ⫾ 0.01† 0.85 ⫾ 0.01 T-wave amplitude, mV Max S/T 1.66 ⫾ 0.05† 3.54⫾ 0.08 QRS duration, ms 158 ⫾ 2* 147 ⫾ 0.3 QT duration, ms 415 ⫾ 7* 445 ⫾ 1 QTc duration Bazett 511 ⫾ 6* 497 ⫾ 1 JTc duration 353 ⫾ 6 350 ⫾ 1 *P ⬍.05. †P ⬍.001.

was weaker (r ⫽ 0.73; P ⬍.01). The regression slope for the QRS/T ratio in the new LBBB group was steeper than in the old LBBB group (P ⬍.01), resulting in almost complete visual separation of the two groups (Figure 3). In the old LBBB group, we found weak univariate correlations between T-vector magnitude and QRS duration (r ⫽ 0.34) and HR (r ⫽ 0.26, both P ⬍.01), which remained significant in the multivariate model. There was no correlation between T-vector magnitude and QT, QTc, JTc, and TnT levels. To classify LBBB in the new and old categories, we used the QRS/T ratio as the criterion to construct a ROC curve. Using the cutoff QRS/T ratio of ⬍2.25 to achieve 100% sensitivity in detecting new LBBB, 88 out of 1,799 ECGs (4.3%) were classified as “new” (39 true and 49 falsepositives, specificity 96%), with an area under the curve (AUC) of 0.99 (Figure 4). To rule out selection bias, we applied the VCG rule to the LBBB group of “unknown duration.” Using the same QRS/T ratio cutoff, 43 out of 1,148 tracings (3.7%) were classified as “new” (P ⫽ .14 compared with known duration LBBB). The average QRS/T ratio in the “unknown duration” LBBB group (3.79 ⫾ 0.04) was not significantly different from the combined LBBB groups of know duration (3.88 ⫾ 0.04; P ⫽ .14). Next we adapted the VCG decision rule to the 12-lead ECG by approximation of the maximal QRS and T-vector magnitudes using S- and T-wave amplitudes. On the basis of the predominant QRS and T-vector directions, horizontal and parallel to the sagittal plane (Figure 1, Table 2), precordial leads were chosen. The best discrimination was

achieved by using the ratio between maximal precordial Swave amplitude and maximal precordial T-wave amplitude (Max S/T) with an AUC of 0.98 (Figure 4 and Table 3). Using the cutoff value of MaxS/T⬍2.37 to achieve 100% sensitivity, the rule had 90% specificity in detecting the new LBBB. Both VCG and ECG rules were applied to the subset of patients with HR ⬎100 min⫺1 to assess whether the differences in average HR between the groups affected the results of classification. At identical mean HR (112 min⫺1 in both groups), the sensitivity of the rules decreased while specificity remained 100% (Table 3). To assess the possible confounding effect of ischemia on LBBB classification, we studied a subset of tracings (12 new and 155 old LBBB) with corresponding TnT levels exceeding twice the upper normal limit (⬎0.2 ng/mL). The performance of both VCG and ECG rules remained virtually unchanged (Table 3). Both rules were applied to 11 LBBB patients who underwent urgent cardiac catheterization for presumed acute MI. In nine of those, the MI was confirmed by enzymatic criteria, and six patients underwent coronary intervention. Four of them were positive for Sgarbossa criteria.19 Three patients developed new LBBB that resolved after reperfusion, and eight had old LBBB at baseline. Both VCG and ECG rules correctly categorized LBBB in all 11 patients, including those with significant ST segment shifts (Figure 5).

Discussion Main study findings and presumed mechanisms In our study, we present an approach to distinguish between new and old LBBB based on the difference in the QRS/T vector magnitude ratio. The premise for the study came from the observation that T-vector magnitude during continuous ventricular pacing decreases with time, the change that was quantitatively linked to the development of CM11 and therefore expected to be present in any form of aberrant conduction. Consistent with this premise, we found that T-vector magnitude in the old LBBB group was significantly lower than in the new LBBB group. A very strong correlation between QRS and T-vector magnitudes (r ⫽ 0.95) resulted in a remarkably uniform QRS/T ratio in the new LBBB group (1.79 ⫾ 0.03), similar in magnitude to that of new-onset right ventricular apical pacing (1.59 ⫾ 0.17),11 the condition that is physiologically similar to LBBB.20 It suggests that the QRS/T ratio at the onset of aberrant conduction might be predetermined physiologically by the anatomical site of abnormal activation. In addition, QRS vector magnitude in the chronic LBBB group was significantly larger than in the new LBBB group. This likely reflects the LBBB-induced structural left ventricular remodeling including eccentric hypertrophy and cavity dilation20 –22 or progression of concurrent cardiac disease. The difference in the QRS vector magnitude further widens the differences in QRS/T vector ratios between the groups, therefore enhancing the performance of discriminating rules. Using the cutoff value of 2.25 for QRS/T VCG designed to achieve 100% sensitivity, the rule had over 92%

Shvilkin et al

New and Old LBBB

1089

Figure 2 Representative ECGs in new and old LBBB. A: Development of a new (⬍6 hours) and progression to old (15 days) LBBB (same patient as in Figure 1). B: Resolution of a new LBBB (⬍8 hours) produces no T-wave abnormalities. C: Resolution of an old LBBB (⬎3 days) produces T-wave inversions consistent with CM. D: Development and resolution of a new LBBB in a patient with acute MI (proximal left anterior descending artery occlusion treated by stenting).

specificity in the new LBBB diagnosis in the whole population as well as subsets of tracings with tachycardia and evidence of myocardial injury. We adapted the rule to the 12-lead ECG by using Max S/T, achieving 100% sensitivity and 68%– 89% specificity for the new LBBB diagnosis at the cutoff value of 2.5, which can be used for manual analysis.

“New” and “old” LBBB definitions There is no accepted definition of the term “new LBBB.” Previous studies considered LBBB new at its first documentation even when the prior ECG was recorded 2–5 years earlier.23,24 We chose the 24-hour cutoff between the new and old LBBB based on the time course of T-wave electrical remodeling in LBBB that largely occurs within the first 24 hours.9 This is also a practically relevant time period when

the clinical decisions regarding reperfusion therapy in acute MI have to be made. Neither new nor old LBBB by our definition had to be permanent, and in fact several patients demonstrated its resolution. Invariably intermittent LBBB lasting longer than 24 hours demonstrated a high QRS/T ratio, and its resolution was accompanied by T-wave inversions consistent with CM, whereas resolution of the new LBBB produced no visible T-wave abnormalities (Figures 2B and 2C).

The role of the decision rules in patients with chest pain The most important potential application of the LBBB age determination is in the setting of a suspected MI.

1090

Heart Rhythm, Vol 7, No 8, August 2010

Limitations of the study

Figure 3 Relationship between QRS and T-vector magnitude in the new (n ⫽ 39) and old (n ⫽ 1760) LBBB. Open blue circles: old LBBB; solid red circles: new LBBB.

The current guidelines4 advocate for equivalency of new (or presumed new) LBBB with symptoms suggestive of ischemia to ST elevation MI. However, as many as 20%– 35% of these patients develop no infarct by enzymatic criteria.25–27 The ability to reliably diagnose LBBB as old without prior ECG tracings might help avoid invasive cardiac interventions in a subset of patients with borderline symptoms, reducing complications and lowering the overall cost of care. On the other hand, confirmation of LBBB as new will reinforce aggressive management. The LBBB age determination does not address the problem of diagnosing ischemia in the setting of LBBB for which established criteria exist.19 Development of a new LBBB during chest pain by itself does not prove ischemia as well. Conditions such as rate-dependent LBBB in patients with noncardiac chest pain and a syndrome of “painful LBBB”28,29 will continue to contribute to false-positive MI diagnosis. To address the concern that ischemic repolarization changes can affect the morphology of LBBB30 and therefore invalidate age classification, we performed a subgroup analyses in a setting of biochemical and ECG/clinical evidence of myocardial injury. In both subgroups (elevated TnT and patients undergoing emergent cardiac catheterization for acute MI including those with positive Sgarbossa criteria), the decision rules held accuracy.

The proposed decision rules were derived based on group comparisons rather than ECG progression in individual patients. Therefore, the proposed physiologic mechanism of the observed findings remains to some extent speculative, although in patients for whom follow-up tracings were available, individual ECG changes were consistent with it (Figure 1). The study was performed retrospectively and therefore is subject to selection bias. Although very few tracings of the new LBBB (mostly exercise or rate related) are available in the literature,9,28,31 they demonstrate an S/T ratio in the presented leads in accord with our findings. Since there was no systematic ECG follow-up, the exact time course of LBBB maturation is unknown. It is likely that in the new LBBB group its duration was significantly shorter, whereas in the old LBBB group it was much longer than 24 hours. Therefore, LBBB of intermediate duration can potentially be misclassified. The extreme scarcity of documented new (⬍24 hours’ duration) digitally recorded LBBB tracings precluded us from breaking down our observations into derivation and validation data sets, and our findings need to be confirmed prospectively and cutoff thresholds for the decision rules might need to be adjusted based on the follow-up data. The study can also be subject to oversampling bias owing to an allowance of more than one tracing per patient. However, the decision rules were developed based solely on the ECG properties of the new LBBB group where oversampling was minimal (18/24 patients with a single tracing). We chose to use ECG tracing rather than a patient as a unit of observation to decrease the risk of selection bias (all LBBB ECGs in the hospital database were considered for inclusion). Patients with baseline QRS widening and significant T-wave abnormalities (such as progressive left ventricular

Practical application of the decision rule Based on the ROC curve analysis, we designed the decision rules to achieve 100% sensitivity for the new LBBB detection. Using the VCG rule, this was accomplished with the QRS/T cutoff point of ⬍2.25. The ECG rule had 100% new LBBB detection at the MaxS/T cutoff of ⬍2.37. For the purposes of manual analysis, we advocate a slightly less specific but more practical for manual measurement MaxS/T cutoff of ⬍2.5.

Figure 4 ROC curves for VCG QRS/T (AUC ⫽ 0.99) and ECG Max S/T (AUC ⫽ 0.975) rules to distinguish old and new LBBB.

Shvilkin et al Table 3

New and Old LBBB

1091

VCG and ECG parameter performance in discriminating new and old LBBB based on ROC curve analysis Sensitivity/specificity for new LBBB detection at cutoff values

All tracings (new LBBB n ⫽ 39; old LBBB n ⫽ 1,760) QRS/T Max S/T HR⬎100/min (new LBBB n ⫽ 14; old LBBB n ⫽ 179) QRS/T Max S/T TnT⬎0.2 (new LBBB n ⫽ 12; old LBBB n ⫽ 155) QRS/T Max S/T

AUC

⬍2.0

⬍2.25

⬍2.5

⬍3.0

0.99 0.975

80/99.5 87/95

100/96 95/90

100/81 100/89

100/59 100/58

0.99 0.98

93/98 93/93

100/92 100/81

100/79 100/68

100/53 100/36

0.99 0.98

83/100 83/96

100/96 92/91

100/85 100/74

100/49 100/47

hypertrophy with secondary repolarization abnormalities) before LBBB development were excluded. The decision rules were not tested at fast ventricular rates (over 135/min) or atrial arrhythmias (atrial fibrillation and flutter).

Conclusion QRS/T vector magnitude ratio allowed accurate discrimination between new and old LBBB in a retrospective study of a hospital ECG database. If confirmed prospectively, the proposed criteria for both computerized and manual ECG diagnosis of the new (⬍24 hours’ duration) LBBB can improve management of patients with chest pain and LBBB.

Acknowledgments The authors thank Christine Dindy, CCT, Stephen L. Feeney, RN, BSN, MS, and Peter Duffy, CVT, of South Shore Hospital for their help in obtaining digital ECG recordings.

References

Figure 5 ECG examples of acute MI patients with LBBB and positive Sgarbossa criteria. Duration of LBBB denoted on the top. Numbers on the bottom represent QRS/T. A: New LBBB in a patient with 95% left main coronary artery stenosis and three-vessel disease. B: New LBBB in a patient with proximal left anterior descending artery occlusion. C: Old LBBB in a patient with left circumflex artery occlusion.

1. Brady WJ, Lentz B, Barlotta K, Harrigan RA, Chan T. ECG patterns confounding the ECG diagnosis of acute coronary syndrome: left bundle branch block, right ventricular paced rhythms, and left ventricular hypertrophy. Emerg Med Clin North Am 2005;23:999 –1025. 2. Kontos MC, McQueen RH, Jesse RL, Tatum JL, Ornato JP. Can myocardial infarction be rapidly identified in emergency department patients who have left bundle-branch block? Ann Emerg Med 2001;37:431– 438. 3. Li SF, Walden PL, Marcilla O, Gallagher EJ. Electrocardiographic diagnosis of myocardial infarction in patients with left bundle branch block. Ann Emerg Med 2000;36:561–565. 4. Antman EM, Anbe DT, Armstrong PW, et al. ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2004;44:E1–E211. 5. Indications for fibrinolytic therapy in suspected acute myocardial infarction: collaborative overview of early mortality and major morbidity results from all randomised trials of more than 1000 patients. Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group. Lancet 1994;343:311–322. 6. Guerrero M, Harjai K, Stone GW, et al. Comparison of the prognostic effect of left versus right versus no bundle branch block on presenting electrocardiogram in acute myocardial infarction patients treated with primary angioplasty in the primary angioplasty in myocardial infarction trials. Am J Cardiol 2005;96:482– 488. 7. Chang AM, Shofer FS, Tabas JA, et al. Lack of association between left bundle-branch block and acute myocardial infarction in symptomatic ED patients. Am J Emerg Med 2009;27:916 –921. 8. Poon K, Frederick PD, French WJ. Abstract 4317: Does a new or presumed new left bundle branch block have equivalent mortality to an acute ST-elevation myocardial infarction? Circulation 2009;120:S935. 9. Rosenbaum MB, Blanco HH, Elizari MV, Lazzari JO, Davidenko JM. Electrotonic modulation of the T wave and cardiac memory. Am J Cardiol 1982;50: 213–222.

1092 10. Denes P, Pick A, Miller RH, Pietras RJ, Rosen KM. A characteristic precordial repolarization abnormality with intermittent left bundle-branch block. Ann Intern Med 1978;89:55–57. 11. Shvilkin A, Bojovic B, Vajdic B, et al. Vectorcardiographic determinants of cardiac memory during normal ventricular activation and continuous ventricular pacing. Heart Rhythm 2009;6:943–948. 12. Surawicz B, Childers R, Deal BJ, et al. AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram. 3. Intraventricular conduction disturbances. J Am Coll Cardiol 2009;53:976 –981. 13. Hancock EW, Deal BJ, Mirvis DM, et al. AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram. 5. Electrocardiogram changes associated with cardiac chamber hypertrophy. J Am Coll Cardiol 2009;53:992–1002. 14. Bojovic B, Hadzievski L, Vukcevic VD, et al. Visual 3Dx: algorithms for quantitative 3-dimensional analysis of ECG signals. Conf Proc IEEE Eng Med Biol Soc 2009;1:6751– 6754. 15. Edenbrandt L, Pahlm O. Vectorcardiogram synthesized from a 12-lead ECG: superiority of the inverse Dower matrix. J Electrocardiol 1988;21:361–367. 16. Wecke L, Gadler F, Linde C, et al. Temporal characteristics of cardiac memory in humans: vectorcardiographic quantification in a model of cardiac pacing. Heart Rhythm 2005;2:28 –34. 17. Luo S, Michler K, Johnston P, Macfarlane PW. A comparison of commonly used QT correction formulae: the effect of heart rate on the QTc of normal ECGs. J Electrocardiol 2004;37:81–90. 18. Shvilkin A, Ho KK, Rosen MR, Josephson ME. T-vector direction differentiates postpacing from ischemic T-wave inversion in precordial leads. Circulation 2005;111:969 –974. 19. Sgarbossa EB, Pinski SL, Barbagelata A, et al. Electrocardiographic diagnosis of evolving acute myocardial infarction in the presence of left bundle-branch block. GUSTO-1 (Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries) Investigators. N Engl J Med 1996;334: 481– 487. 20. Prinzen FW, Cheriex EC, Delhaas T, et al. Asymmetric thickness of the left ventricular wall resulting from asynchronous electric activation: a study in dogs

Heart Rhythm, Vol 7, No 8, August 2010

21.

22.

23.

24. 25.

26.

27.

28.

29.

30. 31.

with ventricular pacing and in patients with left bundle branch block. Am Heart J 1995;130:1045–1053. Vernooy K, Verbeek XA, Peschar M, et al. Left bundle branch block induces ventricular remodelling and functional septal hypoperfusion. Eur Heart J 2005; 26:91–98. Bavelaar-Croon CD, Wahba FF, Van Hecke MV, et al. Perfusion and functional abnormalities outside the septal region in patients with left bundle branch block assessed with gated SPECT. Q J Nucl Med 2001;45:108 –114. Schneider JF, Thomas HE, Jr., McNamara PM, Kannel WB. Clinical-electrocardiographic correlates of newly acquired left bundle branch block: the Framingham Study. Am J Cardiol 1985;55:1332–1338. Rabkin SW, Mathewson FA, Tate RB. Natural history of left bundle-branch block. Br Heart J 1980;43:164 –169. Al-Faleh H, Fu Y, Wagner G, et al. Unraveling the spectrum of left bundle branch block in acute myocardial infarction: insights from the Assessment of the Safety and Efficacy of a New Thrombolytic (ASSENT 2 and 3) trials. Am Heart J 2006;151:10 –15. Hands ME, Cook EF, Stone PH, et al. Electrocardiographic diagnosis of myocardial infarction in the presence of complete left bundle branch block. Am Heart J 1988;116:23–31. Wong CK, French JK, Aylward PE, et al. Patients with prolonged ischemic chest pain and presumed-new left bundle branch block have heterogeneous outcomes depending on the presence of ST-segment changes. J Am Coll Cardiol 2005; 46:29 –38. Candell Riera J, Oller Martinez G, Vega J, et al. Exercise-induced left bundlebranch block in patients with coronary artery disease versus patients with normal coronary arteries. Rev Esp Cardiol 2002;55:474 – 480. Virtanen KS, Heikkila J, Kala R, Siltanen P. Chest pain and rate-dependent left bundle branch block in patients with normal coronary arteriograms. Chest 1982;81:326 –331. Haywood LJ. Left bundle branch block in acute myocardial infarction: benign or malignant? J Am Coll Cardiol 2005;46:39 – 41. Koito H, Spodick DH. Physiologic differences in rate-related versus exerciseinduced left bundle branch block. Am J Cardiol 1988;62:316 –319.