Available online at www.sciencedirect.com

Journal of Electrocardiology 44 (2011) 544–554www.jecgonline.com

A detailed guide for quantification of myocardial scar with the SelvesterQRS score in the presence of electrocardiogram confoundersZak Loring, BS,a Sreetharan Chelliah, MBChB,b Ronald H. Selvester, MD,c

Galen Wagner, MD,b David G. Strauss, MD, PhDd,⁎aDuke University School of Medicine, Durham, NC, USAbDuke Clinical Research Institute, Durham, NC, USA

cMemorial Hospital Research Center, Long Beach, CA, USAdOffice of Science and Engineering Laboratories, Center for Devices and Radiological Health, US Food and Drug Administration, Silver Spring, MD, USA

Received 10 May 2011

Abstract The Selvester QRS score translates subtle changes in ventricular depolarization measured by the

⁎ CorrespondingNew Hampshire Aven

E-mail address: d

0022-0736/$ – see frodoi:10.1016/j.jelectroc

electrocardiogram into information about myocardial scar location and size. This estimated scar hasbeen shown to have a high degree of correlation with autopsy-measured myocardial infarct size. Inaddition, multiple studies have demonstrated the value of the QRS score in post–myocardial infarctpatients to provide prognostic information. Recent studies have demonstrated that increasing QRSscore is predictive of increased implantable defibrillator shocks for ventricular tachycardia andfibrillation as well as decreased response to cardiac resynchronization therapy. Although QRSscoring has never achieved widespread clinical use, increased interest in patient selection and risk-stratification techniques for implantable defibrillators and cardiac resynchronization therapy has ledto renewed interest in QRS scoring and its potential to identify which patients will benefit fromdevice therapy. The QRS score criteria were updated in 2009 to expand their use to a broaderpopulation by accounting for the different ventricular depolarization sequences in patients withbundle-branch/fascicular blocks or ventricular hypertrophy. However, these changes also introducedadditional complexity and nuance to the scoring procedure. This article provides detailed instructionsand examples on how to apply the QRS score criteria in the presence of confounding conductiontypes to facilitate understanding and enable development and application of automated QRS scoring.Published by Elsevier Inc.

Keywords: Electrocardiology; Electrophysiology; QRS complex; ECG criteria; Myocardial infarction; Fibrosis; Multimodalcardiac imaging; Risk stratification

Introduction

In an era of multimodal cardiac imaging, understandingthe information contained in the 12-lead electrocardiogram(ECG) is important yet frequently overlooked. The SelvesterQRS score, which was described in preliminary form 1972,1

translates subtle changes in cardiac electrical activity intoinformation about myocardial scar location and size.2,3

Although QRS scoring was rigorously validated in compar-ison with autopsy-measured myocardial infarct size in the1980s4-7 and was shown to have prognostic value,3,8-14 thisECG method did not achieve widespread clinical use.However, the need to improve patient selection forimplantable defibrillators (ICDs) and cardiac resynchroniza-

author. US Food and Drug Administration, 10903ue, WO62-1126, Silver Spring, MD, 20993, USA.avid.strauss@fda.hhs.gov

nt matter. Published by Elsevier Inc.ard.2011.06.008

tion therapy (CRT) has led to a renewed interest in QRSscoring because characterizing myocardial scar may betterpredict which patients will benefit from ICDs and CRT.15,16

The older versions ofQRS scoring could only be used in theabsence of ECG confounders (ie, bundle-branch/fascicularblocks and ventricular hypertrophy), which makes QRSscoring impractical for ICD/CRT patient selection becausegreater than 50% of potential ICD patients and nearly all CRTpatients have ECG confounders.2,15,17 In 2009, new QRSscore criteria for use in the presence of ECG confounders werepublished and shown to be able to identify and quantifymyocardial scar in comparison with contrast-enhancedmagnetic resonance imaging (MRI) in patients with ischemicand nonischemic cardiomyopathy,2,8 including patients withChaga's heart disease.3 Subsequent studies have shown thatthese QRS score criteria for use in the presence of ECGconfounders can enhance prediction of which patients will

545Z. Loring et al. / Journal of Electrocardiology 44 (2011) 544–554

experience appropriate ICD shocks and respond to CRT.15,16

The purpose of this report is to provide more detailedinstructions beyond the 2009 publication for classifying thetype of ECG confounder and applying theQRS score criteria tobetter facilitate manual and automated application.

Review of modifications to QRS scoring

The QRS score has been updated and amended multipletimes since its original description in 1972.1 The most recentchanges came in 2009 with a publication describing how toapply the QRS score in the setting of hypertrophy andconduction defects.8 Before this study, fascicular blocks,bundle-branch blocks, and hypertrophy were consideredconfounding factors that prevented infarction evaluation viaQRS score. The updated criteria in 2009 account for thedifferent activation sequences and conduction propertiesassociated with these defects and are able to distinguish theadditional electrical changes associated with infarction. Thiswas made possible by using adjusted criteria for each of thefollowing abnormalities: left and right ventricular hypertrophy(LVH and RVH, respectively), left anterior fascicular block(LAFB), left bundle-branch block (LBBB), right bundle-branchblock (RBBB), RBBB + LAFB, and no confounders. Thescoring criteria sets differ slightly among all of theseabnormalities, with LBBB having the most dissimilar criteria,but they still follow the same principle of a score with each pointrepresenting 3% infarct of the left ventricular (LV)myocardium.

For example, in LBBB, the QRS scoring criteria for thelimb leads (I, II, aVL, and aVF) are similar to those in all otherECGs; however, the precordial lead scoring criteria aredramatically different. In normal conduction, anteroseptalscar causes the rest of the LV to activate largely unopposed,producing net posteriorly directed electrical activity andcreating a Q wave in leads V1 to V3. Conversely, in LBBB, inthe absence of scar, the LV is activated via electricaldepolarization from the right ventricle (RV) through theseptum and finally out to the posterolateral LV wall, creatinga small or absent initial R wave and deep Q or S wave in V1

and V2 (Fig. 1A). Anteroseptal scar in the presence of LBBBcauses depolarization from the RV to the LV through theseptum to be impeded, and the initial electrical forces aredirected anteriorly because of the unopposed activation of theRV free wall, resulting in large R waves in V1 and V2 (Fig.1B).8 Thus, anteroseptal scar results in Q wave formation innormal conduction but causes large R wave formation inLBBB. Posterolateral scar disrupts the late forces of the QRScomplex as depolarization spreads from the septum out to theposterolateral wall, causing notching of the terminal portionof the S wave in V1 and V2 and, if the scar extends toward theapex, of the R wave in V5 and V6 (Fig. 2C). Inferior scarbehaves similarly in LBBB as other conduction types,causing a Q wave in aVF because of unopposed superioractivation early in ventricular activation (Fig. 2D).

In RBBB, the RV activates via depolarization from the LVthrough the septum with the RV free wall activating last,producing broadR or R' waves inV1 andV2.

18 This conductionpattern confounds the diagnosis of posterolateral infarcts thatcan also produce large R waves because of unopposed anterior

forces. To account for this, in QRS scoring, the R wave criteriainV1 andV2 for posterolateral scar are increased in the setting ofRBBB, and only the initial part of the R wave that peaks withinthe first 50 milliseconds is considered (see Appendix Table 1).Furthermore, there are additional possible anterior points in thesetting of RBBB to increase the sensitivity to large infarcts thatare the most common cause of RBBB.8

Performing QRS scoring

Classification of conduction/hypertrophy type

Figs. 2 and 3 show flow diagrams that graphicallyrepresent the process of determining the QRS conduction/hypertrophy type “bin” in which to evaluate the ECG: LBBB,LAFB, LVH, RBBB + LAFB, or no confounders. Lead V1

QRS morphology is considered to determine if there is a leftvs right conduction/hypertrophy abnormality. If there is aterminal negative deflection in lead V1 (rS or Qmorphology),then analysis proceeds according to the diagram in Fig. 2, andif there is a terminal positive deflection in lead V1 (R or R'),then analysis proceeds according to the diagram in Fig. 3.

As reported previously,2,8 the definitions for classifyingLBBB and LAFB for QRS score purposes differ slightlyfrom other reported definitions in the literature.19 Thedefinitions used with QRS scoring correlate with theobservations made from computer simulations of thedifferent conduction types from which the QRS scorecriteria were derived.18,20 The critical difference withLBBB (recently described in detail21) is that the QRSscoring definition for LBBB requires the presence of mid-QRS notching and a minimum duration of 140 millisecondsfor men and 130 milliseconds for women. To satisfy LAFBclassification for QRS scoring, we have used a minimumQRS duration of 100 milliseconds because a block in the leftanterior fascicle causes a prolongation of the QRS complexby 20 milliseconds.22 However, because women haveshorter baseline QRS duration than men,23 we propose aminimum QRS duration of 90 milliseconds in women and100 milliseconds in men for defining LAFB. With regard tothe lack of an upper boundary for QRS duration, LAFB maycoexist with other QRS, prolonging conduction defects suchas LVH, a dilated ventricle, or diffuse interstitial fibrosisleading to delayed intramural conduction. Patients meetingLAFB criteria with QRS duration greater than 120milliseconds undoubtedly have multiple pathologies (eg,LAFB + LVH); however, these patients are most accuratelyevaluated by the LAFB QRS score criteria. For the purposesof QRS scoring, we recommend using these definitions thatwere used during the development of QRS scoring,20 thevalidation in comparison with cardiac MRI,2,3 and to predictoutcomes from CRT and ICD therapy.15,16

For scoring purposes, it does not matter if LVHaccompanies one of the abnormal conduction types. Theprincipal difference between QRS scores of LVH and noconfounders is that, in LVH, Q waves in V1 through V3 onlyreceive QRS score points if 4 QRS score points or more arealso present in leads I, aVL, V4, V5, or V6. This is because inLVH, Q waves may be present in leads V1, V2, and V3 in the

546 Z. Loring et al. / Journal of Electrocardiology 44 (2011) 544–554

absence of infarction; however, evidence of infarction inother leads makes it more likely that the formation of Qwaves was caused by an infarction. In addition, smooth Qwaves can be normal in V1 to V3 in the presence of LVH, butnotches in the initial 40 milliseconds (NtchInit40) areconsidered a sign of infarction. An ECG is considered todemonstrate LVH if either the Sokolow-Lyon or the Cornellcriteria are satisfied (Figs. 2 and 3).

After determining conduction type, all ECGs should beevaluated for the presence of right atrial overload (RAO) by

Fig. 1. Left bundle-branch block activation patterns. All panels demonstrate the vefrontal plane (top), horizontal plane (middle), and sagittal plane (bottom) in LBBBposterolateral infarction (C), and LBBB with inferior infarction (D). Colored lineperiod (isochrones). Numbers represent milliseconds since beginning of activationwith anteroseptal infarction (B), increased R/R' amplitude ratios in V5 and V6 wposterolateral infarction (C), and Q waves and decreased R/Q or R/S in aVF amp

analyzing the P waves in both leads V1 and aVF as shown inFig. 4. Right atrial overload commonly occurs in parallel withRVH.18,20 Large R waves in V1 and V2 are consideredevidence of unopposed anterior electrical forces and areassigned points indicating infarction in the posterolateral wallof the LV or, in the case of LBBB, in the anteroseptal the wall;however, these R waves may also be due to increased anteriorelectrical force from RVH. If a patient has RAO, then large Rwaves in V1 and V2 are likely to be because of RVH ratherthan infarction.8

ntricular activation pattern in LBBB and ECG wave forms as seen from thewithout infarction (A), LBBB with anteroseptal infarction (B), LBBB withs represent areas of myocardium activated within the same 10-millisecond. Key ECG changes include the development of large R waves in V1 and V2

ith apical infarction (B), increased S/S' amplitude ratios in V1 and V2 withlitude ratios with inferior infarction (D).

Fig. 1. (continued).

547Z. Loring et al. / Journal of Electrocardiology 44 (2011) 544–554

Leads V1 and V2 have different criteria for measuringanteroseptal and posterolateral scar. The QRS scoringsheet (Fig. 5) lists the anteroseptal criteria as V1 anteriorand V2 anterior in LBBB and simply as V1 and V2 in allother conduction types. Posterolateral criteria are listed as V1

posterior and V2 posterior in all conduction types. AllECGs should be analyzed for the presence of bothanteroseptal and posterolateral scar by evaluating bothsets of criteria for leads V1 and V2. In the presence ofRAO, we have excluded the V1 and V2 posterolateral QRSscore criteria (anteroseptal criteria in LBBB) to preventallocating false-positive points due to RVH; however, theycan be included in situations where sensitivity to detectscar is more valuable than specificity.

Application of QRS scoring

False-positive points in QRS scoring were common inyounger men who have increased voltage and older womenwho have lower voltages.8 Before scoring, all absoluteamplitude criteria (not ratios) are corrected to the age of 55years in the score sheet (Fig. 5) by increasing them 1% peryear for those aged 54 years and younger and decreasingthem 1% per year for those aged 56 years and older. Forwomen, both duration and absolute amplitude (not ratio)criteria are further decreased by 10%.

After ECGs have been placed in conduction/hypertrophybins and amplitude and duration are corrected, the QRS scorefor that conduction/hypertrophy type is selected and applied.

Sokolow-Lyon OR Cornell CriteriaSokolow-Lyon

(S in V1) + (R in V5 or V6) ≥ 3.50mV OR

R in V5 or V6 >2.60mVCornell criteria

R in aVL + S in V3 >2.80mV ( )R in aVL + S in V3 >2.00mV ( )

Lead V1

V1 terminal negative deflection (rS or Q wave)

LBBB

QRS ≥ 100 ms ( )QRS ≥ 90 ms ( )

ANDLeft axis

≤ -45° & > -180°

LAFB

QRS ≥140 ms ( )QRS ≥130 ms ( )

ANDmid-QRS (after 40ms)

notching/slurring/slowing in ≥2 of the leads V1, V2,

V5, V6, I or aVL

LVH

Yes

Yes

Yes

No

No

No

No Confounders

Fig. 2. Flow chart for determining conduction/hypertrophy typeabnormality. This chart describes the sequence in which ECGs areevaluated for conduction type when the terminal deflection in V1 isnegative (rS or Q wave).

V1 terminal positive deflection (R or R’ wave)

Lead V1

RBBB+ LAFB

RBBB

QRS ≥120 ms

No

Left Axis≤45° & >180°

No

Yes Yes

Sokolow-Lyon OR Cornell CriteriaSokolow-Lyon

(S in V1) + (R in V5 or V6) ≥ 3.50mV OR

R in V5 or V6 >2.60mVCornell criteria

R in aVL + S in V3 >2.80mV ( )R in aVL + S in V3 >2.00mV ( )

LAFB

LVH

Yes

Yes

No

No

No Confounders

QRS ≥ 100 ms ( )QRS ≥ 90 ms ( )

ANDLeft axis

≤ -45° & > -180°

Fig. 3. Flow chart for determining conduction/hypertrophy typeabnormality. This chart describes the sequence in which ECGs areevaluated for conduction type when the terminal deflection in V1 ispositive (R or R' wave).

P-wave in,V1 ≥0.1 mV

ORaVF ≥0.175 mV

(only positive deflections of P waves are measured)

All Conduction types

No Right Atrial Overload

No

Right Atrial Overload (RAO)

(suspect accompanying RVH)In RBBB, LAFB, LAFB +

RBBB, LVH and No confounders, exclude V1,

V2 posterior criteria.In LBBB, exclude V1, V2 anterior “R criteria points”.

Yes

Fig. 4. Flow chart for determining RAO. After conduction type isdetermined by following Fig. 2 or Fig. 3, all ECGs are evaluated forpresence of RAO, which suggests accompanying RVH.

548 Z. Loring et al. / Journal of Electrocardiology 44 (2011) 544–554

Careful measurements of both amplitudes and durations aremade with up to an 8× magnifier, when necessary.

When scoring, 2 concepts must be considered:“weighting” and “selecting” (Fig. 6). Weighting refers tothe number of points awarded for the criteria satisfied andis provided to the right of the criteria. If the point valueis not listed directly to the right of the criterion, its pointvalue is the same as the criterion directly above it.Selecting is the process of choosing a single criterionfrom a group. When criteria with multiple weights areincluded within a box, they are ordered by decreasingweights. Therefore, proceeding from the top down withina box (delineated by a dotted line), only the first satisfiedcriterion should be selected. Only 1 criterion can beselected for each box, but there may be multiple boxes ineach lead. The total point score for each ECG is the sumof the points accumulated from the criteria satisfied foreach lead.

Glossary and definitions

The glossary section in Appendix Tables 1 and 2 containdefinitions and illustrations of terms and criteria that areencountered when performing QRS scoring. The glossary isdivided into 2 sections: the first is for no confounders, LVH,LAFB, RBBB, and RBBB + LAFB conduction/hypertrophy

Amplitude adjust ( 1%/yr age 20-54; 1%/yr >55 yrs; 10% for females)

Yes/No)*** ,RAO(**Duration adjust( 10% for females)

LBBB

Lead Criteria Pts Criteria Pts Criteria Pts Criteria Pts Criteria Pts Lead Criteria PtsI Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 I any Q 1

R/Q 1 1 R/Q 1 1 R/Q 1 1 R/Q 1 1 R/Q 1 1 R/Q 1 2R 0.2 mV R 0.2 mV R 0.2 mV R 0.2 mV R S/R .2 mV0 1

II Q 40 ms 2 Q 40 ms 2 Q 40 ms 2 Q 40 ms 2 Q 40 ms 2 R/Q 1.5 1Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 R/S 1.5

aVL Q 30 ms 1 Q 40 ms 1 Q 40 ms 1 Q 40 ms 1 Q 30 ms 1 II Q 40 ms 2R/Q 1 1 R/Q 1 1 R/Q 1 1 R/Q 1 1 R/Q 1 1 Q 30 ms 1

aVF Q 50 ms 3 Q 50 ms 3 Q 50 ms 3 Q 60 ms 3 Q 50 ms 3 R/Q 0.5 1Q 40 ms 2 Q 40 ms 2 Q 40 ms 2 Q 50 ms 2 Q 40 ms 2 R/S 0.5Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 Q 40 ms 1 Q 30 ms 1 aVL Q 50 ms 2

R/Q 1 2 R/Q 1 2 R/Q 1 2 R/Q 1 2 R/Q 1 2 Q 40 ms 1R/Q 2 1 R/Q 2 1 R/Q 2 1 R/Q 2 1 R/Q 2 1 R/S 0.5 2

V1 Q Q2sm 05 50 ms 2 any QR R/Q 0.5

Ant. any Q 1 any QR 1 any Q 1 (or any Q if *) 1

1

any Q 1 R/S 1 1Init R 20 ms Q/R04tinIhctN 1

V1 R/S 1 1 1 R/S 1 R/S 1 1 aVF Q 50 ms 2

Post.** Init R 60 ms 2 R 50 ms 2 Init R 60 ms 2 R 50 ms 2 R 50 ms 2 Q 40 ms 1Init R 1.5 mV R 1 mV Init R 1.5 mV R 1 mV R Q/RVm 1 0.5 1Init R 50 ms 1 R 40 ms 1 Init R 50 ms 1 R 40 ms 1 R 40 ms 1 R/S 0.5Init R 1.0 mV R 0.7 mV Init R 1.0 mV R 0.7 mV R 0.7 mV V1 NchInit40 1

Q 0.2&S Q10.2mV 0.2&S 0.2 mV 1 Q 0.2&S 0.2mV 1 Ant.*** R 0.3 mV 2V2 Q Q2sm 05 R2sm 05 30 ms

Ant. any Q 1 any QR 1 any Q 1 1 any Q 1 R 0.2 mV 1R 10 ms R 10 ms R 10 ms (or any Q if *) R R10 ms 20 msR 0.1 mV R 0.1 mV R 0.1 mV NtchInit40 R 0.1 mV V1 S/S' 2.0 3

V2 R/S S/R15.1 1.5 1 R/S 1.5 1 Post S/S' 1.5 2

Post.** Init R 70 ms 2 R 60 ms 2 Init R 70 ms 2 R 60 ms 2 R 60 ms 2 S/S' 1.25 1Init R 2.5 mV R 2 mV Init R 2.5 mV R 2 mV R 2 mV V2 NchInit40 1Init R 50 ms 1 R 50 ms 1 Init R 50 ms 1 R 50 ms 1 R 50 ms 1 Ant.*** R 0.4 mV 2Init R 2.0 mV R 1.5 mV Init R 2.0 mV R 1.5 mV R R1.5 mV 30 ms

Q 0.3&S Q10.3mV 0.3&S 0.3mV 1 Q 0.3&S 0.3mV 1 R 0.3 mV 1V3 Q 30 ms 2 Q 30 ms 2 Q 30 ms 2 QR& (Q 30 ms) 2 Q 30 ms 2 R 20 ms

R 10 ms R 10 ms R 10 ms NtchInit40 1 R 10 ms V2 S/S' 2.5 3Q 20 ms 1 Q 20 ms 1 Q Q1sm 02 20 ms 1 Post S/S' 2.0 2R 20 ms R 20 ms R Rsm 02 'S/Ssm 02 1.5 1

V4 Q 20 ms 1 Q 20 ms 1 Q 20 ms 1 Q 20 ms 1 Q 20 ms 1 V5 any Q 1R/Q 0.5 2 R/Q 0.5 2 R/Q 0.5 2 R/Q 0.5 2 R/Q 0.5 2 R/R ' 2 2R/S 0.5 R/S 0.5 R/S 0.5 R/S 0.5 R/S 'R/R.50 1 1R/Q 1 1 R/Q 1 1 R/Q 1 1 R/Q 1 1 R/Q 1 1 R/S 2R/S 1 R/S 1 R/S 1 R/S 1 R/S R1 0.5 mV 1

R 0.5 mV R 0.5 mV R 0.5 mV R 0.5 mV R 0.5 mV V6 Q 20 ms 1 'R/R04tinIhctN04tinIhctN04tinIhctN04tinIhctN04tinIhctN 2 2

V5 Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 R/R' 1 1R/Q 1 2 R/Q 1 2 R/Q 1 2 R/Q 1 2 R/Q 1 2 R/S 2R/S 1 R/S 1 R/S 1 R/S 1 R/S R1 0.6 mV 1R/Q 2 1 R/Q 2 1 R/Q 2 1 R/Q 2 1 R/Q 2 1 Total PointsR/S 2 R/S 1.5 R/S 1.5 R/S 2 R/S 2

R 0.6 mV R 0.6 mV R 0.6 mV R 0.6 mV R %LV infarct

Vm 6.0 )stp# * 3(04tinIhctN04tinIhctN04tinIhctN04tinIhctN04tinIhctN

V6 Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 Q 30 ms 1 Q 30 ms 1R/Q 1 2 R/Q 1 2 R/Q 1 2 R/Q 1 2 R/Q 1 2R/S 1 R/S 1 R/S 1 R/S 1 R/S 1R/Q 3 1 R/Q 3 1 R/Q 3 1 R/Q 3 1 R/Q 3 1R/S 3 R/S 2 R/S 2 R/S 3 R/S 3

R 0.6 mV R 0.6 mV R 0.6 mV R 0.6 mV R 0.6 mVNtchInit40 NtchInit40 NtchInit40 NtchInit40 NtchInit40

Total Points Points Points Points Points

%LV infarct % LV infarct %LV infarct % LV infarct % LV infarct(3 x #pts) (3 x #pts) (3 x #pts) (3 x #pts) (3 x #pts)

* (for LVH) if 4 other points in leads I, aVL, V4, V5 or V6 then count QS in V1-V3

** (RAO) if P positive amp in V1 0.1 mV or aVF P .175 mV, then exclude V1-V2 Post criteria

RBBB LAFB

QRS Scoring

No Confounders

Age & gender

Patient ID

*** (RAO) if P positive amp in V1 0.1 mV or aVF P .175 mV, then

exclude V1-V2 R-criteria points

any QR (or any Q if *)

LAFB + RBBB

any QR

LVH

QRS duration

QRS axis

Fig. 5. QRS scoring sheet. Sample QRS scoring sheet with all conduction types and criteria is listed. After demographic information is entered in the top portionconduction type is selected, and analysis proceeds down the appropriate column.

549Z. Loring et al. / Journal of Electrocardiology 44 (2011) 544–554

types (Appendix 1), and the second is only for LBBBconduction type (Appendix 2). This is because of LBBBhaving different criteria definitions compared with the otherconduction/hypertrophy types.

Examples

An example ECG with a completed score sheet isdepicted in Fig. 7. In addition, the online supplementcontains 2 sets of ECGs (1 scanned, 1 digital) with bothblank test sets and completed score sheets. This supplementis included to provide an opportunity to apply the QRSscoring criteria in the various conduction abnormalities onprescored examples.

,

Limitations of QRS scoring

As with all diagnostic modalities, the QRS score haslimitations. Using P wave amplitude evidence of RAO todiagnoseRVH is not awidely usedmethod; however, it does notrely on features of the QRS complex that may be affected byinfarction. Previous studies have found that P wave analysis fordiagnosing RVH has a sensitivity of 30% to 40% and aspecificity of 80% to 91%.24,25 Because the presence of RAO isused to exclude evaluation of posterolateral infarction,specificity is more important than sensitivity. One consequenceof this method is that patients with both RAO and posterolateralinfarction (or anteroseptal infarction in LBBB) may haveunderestimated QRS scores. Another challenge arises when

Fig. 6. Selecting and weighting examples. Criteria for an example lead areshown. Boxes are delineated by dotted lines, check marks and x's indicatewhich criteria are satisfied, and circles indicate criteria that are selected andcontribute to the overall score for the lead. TheQ≥ 20milliseconds criterion ismet, yielding 1 point for the top box in this lead. Proceeding from the top downto the next box, the next criterion that is met is R/S≤0.5, yielding 2 points (nopoint value is listed directly to the right of this criterion; therefore, its pointvalue is equal to that of the criterion directly above it). Although 2 additionalcriteria are met (R/S≤1 and R≤0.5mV), only the single criterion within eachbox that is closest to the top of the box is selected.

Fig. 7. QRS score example. The ECG and completed QRS score sheet areshown for a 45-year-old man with a QRS score of 7.

550 Z. Loring et al. / Journal of Electrocardiology 44 (2011) 544–554

individual ECGs straddle the definitions of conduction/hypertrophy types. Although the QRS score criteria weretested when adhering to the previously outlined conduction/hypertrophy type definitions, further research may improvethe classification of borderline ECGs, especially with regardto defining the presence vs absence of LBBB.21

It is possible that beat-to-beat variation may lead todifferent numbers of points to be assigned to a leaddepending on which beat is evaluated. The best way tolimit bias and error in this setting is to score the mostrepresentative QRS complex (ie, isoelectric baseline, nointerceding premature beat, etc) or the computer-generatedmedian beat, if available. Some of the criteria call forrecognition of notches and/or slurs in the QRS complex. Thepresence and/or morphology of slurs and notches can begreatly affected by low ECG sampling rates or poor-qualitytracings. Sampling rates of at least 500 Hz are ideal for QRSscoring but not always available. Many of the previouslycited studies used routinely acquired paper 12-lead ECGs,suggesting that QRS scores derived from lower sample rateECGs still provide clinically useful information. Thesestudies demonstrated good intra- and interobserver agree-ment, and manually applied QRS score criteria alsodemonstrated a strong ability to identify scar in referenceto contrast-enhanced MRI (receiver operating characteristicarea under the curve, 0.91 [95% confidence interval, 0.86-0.95]) and good correlation with MRI scar size (r = 0.74, P b.0001).2,15 Limitations of beat-to-beat variability, poor-quality paper ECGs, and the increasing complexity of theQRS score criteria can be improved by implementation ofautomated QRS scoring algorithms.

Conclusions

The Selvester QRS score is a useful clinical tool thatprovides perspective on the relationship between cardiacelectrical function and structural changes (viz, scar presenceand size). The clinical use of understanding this relationshipcontinues to be explored. By providing this detailed guide,we hope to empower others to continue to study QRSscoring, enable programming of automated QRS scoring,

and help determine how this understanding can best informclinical decisions.

Acknowledgments

Supported, in part, by Duke University's Clinical andTranslational Science Award grant TL1RR024126 fromNCRR/NIH (to Z. Loring) and Food and Drug Administra-tion Critical Path and Office of Women's Health grants (toD. Strauss).

(continued on next page

551Z. Loring et al. / Journal of Electrocardiology 44 (2011) 544–554

Appendix Table 1: Glossary for no confounders, LVH, LAFB, RBBB, and RBBB + LAFB

Definitions

Illustrations

Q-Wave: When the first deflection of the QRS complex is negative, it is termed a “Q-wave”. It can be seen as“smooth Q” or “notched Q.”

Smooth Q: Present if there is no reversal in direction ≥0.05 mV in the first negative deflection (left picture).Notched Q: Present when there is a reversal in direction ≥0.05 mV within the initial negative deflection(right picture).

Q-Wave Duration: Taken from the start of the global QRS complex (ie, if there is an isoelectric segmentpreceding a Q-wave, it contributes to the Q-wave duration (left picture) to the point directly above thepeak of the notch in a notched Q (right picture) or where the tracing recrosses the PQ baseline in asmooth Q.

Q-Wave Amplitude: Measured from the nadir of the negative deflection of the notch in a notched Q (left picture)or from the most negative deflection (right picture), to the PQ-segment baseline in a smooth Q.

QR: R-wave must be present following a Q-wave.

Any Q: Any type of Q-wave is present.In leads V1-V3 for LVH:If QR is present, a point is given.If any Q-wave is present a point is only given if ≥4 other QRS points are present in leads I, aVL, V4,V5, or V6. This is done because LVH can cause a Q-wave in V1-V3 without the presence of infarction;however, if signs of anterior infarct are present in other leads, then the Q-wave in V1-V3 is more likelycaused by infarct/scar.

R-Wave: The first positive deflection of the QRS complex.

R-Wave Duration:No Q-wave: Measured from beginning of global QRS complex to the point where it re-crosses PQ-segmentbaseline or the end of the QRS complex (whichever comes first).

Q-Wave Present: The R wave begins when the Q-wave returns to the PQ-segment baseline and ends when thetracing re-crosses the PQ baseline or the end of the QRS complex (whichever comes first).

R-Wave Amplitude: Measured from the PQ-segment baseline to the most positive deflection before it re-crossesthe baseline (regardless of the presence of a notch). Except RBBB, see below.

⁎⁎In LAFB, lead V2: An R wave must be present to receive ANY points.

In RBBB:

When awarding points for large R waves in V1-V2 for posterolateral infarct, a notch (reversal in direction≥90°that peaks within the first 50 ms should be considered the peak of the initial R wave (“Init R”). The nadir of

the notch should be considered the end of the initial R wave.If the R wave's initial peak occurs after the first 50 ms of the global QRS complex, points are not given.This is done because the later part of the R wave or R' represents the depolarization of the rightventricle.The duration of the initial R wave is measured from where the R wave begins to either the nadir of thenotch following the initial peak or where the tracing re-crosses the PQ baseline. Initial R waves refer tothe initial peak of the R wave; presence of Q waves preceding the R wave does not discount the“InitR” criteria.

)

552 Z. Loring et al. / Journal of Electrocardiology 44 (2011) 544–554

Appendix Table 1 (continued)

Definitions

Illustrations

S-Wave: The first negative deflection after an R wave.

S-Wave Amplitude: Measured from the PQ-segment baseline to the most negative part of the S wave (regardlessof presence of notch).

NchInit40: Any reversal of direction≥90° that begins within the first 40 ms of the global QRS complex yields 1point in V4, V5, and V6 in all conduction types (except LBBB) and also 1 point in V1, V2, and V3 in LVH.Smooth R or Q-waves do not satisfy “NchInit40” criteria.

R/Q or R/S Amplitude Ratios:

Ratio of the R-wave amplitude to the Q- or S-wave amplitude.If no R wave is present, then this is considered an R/S or R/Q of 0, which receives the maximum points for theratio criteria.For QRS complexes with multiple R and/or R' waves, only the first R wave is considered for scoring.Similarly, for QRS complexes with multiple S or S' waves, only the first S wave is considered for scoring.

Q ≤ 0.2 & S ≤ 0.2 mV and Q ≤ 0.3 & S ≤ 0.3 mV:

Points will be awarded for this criteria only if BOTH the Q and the S amplitudes of the lead are ≤0.x mV(where x = 2 for V1 posterior and x = 3 for V2 posterior).If either the Q or S wave is absent, the wave is regarded as having an amplitude of 0.00 mV.

QR and (Q ≥ 30 ms):

Points will be awarded only if BOTH QR morphology is present AND Q-Wave duration is ≥30 ms.

Appendix Table 2: Glossary for LBBB

Definitions

Illustrations

Q-Wave: When the first deflection of the QRS complex is negative, it is termed a “Q-Wave”. It can be seen as“smooth Q” or “notched Q.”In contrast to QRS scoring in other conduction/hypertrophy types, notches are ignored for the purpose ofdefining the end of the Q-wave.

Q-Wave Duration: Taken from the start of the global QRS complex (ie, if there is an isoelectric segmentpreceding a Q-wave, it contributes to the Q-wave duration) and proceeds until the tracing recrosses thePQ-segment baseline (regardless of the presence of a notch).

Q-Wave Amplitude: Measured from PQ baseline to the most negative deflection (regardless of the presenceof a notch).

553Z. Loring et al. / Journal of Electrocardiology 44 (2011) 544–554

Appendix Table 2 (continued)

Definitions

Illustrations

R-Wave Amplitude and Amplitude Ratios (R/Q and R/S):

When considering absolute R-wave amplitude or R/Q or R/S amplitude ratios, the most positive point(R or R') should be considered the R wave (regardless of notches).The most negative point should be considered the Q or S wave amplitude (regardless of notches).

R-Wave Duration: R-wave duration is measured from the start of the global QRS complex to the point whereit recrosses the PQ baseline (regardless of presence of notches) or the end of the QRS complex (whichevercomes first).

NchInIt40: A notch (change in direction ≥90°) that begins its change in direction within the initial 40 ms.

A notch on the R wave or a notch on the S wave within the first 40 ms does count as a NchInIt40.A smooth R wave does not count as NchInIt40.In LBBB, qR waves count as NchInIt40.

R/R' and S/S' Amplitude Ratios:

These points can be obtained in V1-V2 (S/S' amplitude ratio) and V5-V6 (R/R' amplitude ratio).The R or S is defined as the first peak/nadir, notch or beginning of the slur.If no R wave is present, for this criteria, treat Q-wave as S wave.The R' or S' is defined as the second peak/nadir, notch, or end of the slur after the initial 40 ms of theQRS complex.If there is no identifiable notch/slur, then these criteria should be ignored.In leads V5 and V6, both peaks/notches/slurs must be above the PQ baseline to receive R/R' points.

Appendix A. Supplementary data

Two sets of ECGs are included that can be used topractice QRS scoring. ECG set 1 consists of 26 scannedECGs along with a key with annotated ECGs and completedQRS scoring sheets. ECG set 2 contains 12 ECGs of higherquality along with a key with completed QRS scoring sheets.Supplementary data to this article can be found online atdoi:10.1016/j.jelectrocard.2011.06.008.

References

1. Selvester R, Wagner J, Rubin H. Quantitation of myocardial infarct sizeand location by electrocardiogram and vectorcardiogram. Quant Cardiol1972:31.

2. Strauss DG, Selvester RH, Lima JA, et al. ECG quantification ofmyocardial scar in cardiomyopathy patients with or without conduction

defects: correlation with cardiac magnetic resonance and arrhythmo-genesis. Circ Arrhythm Electrophysiol 2008;1:327.

3. Strauss DG, Cardoso S, Lima JA, Rochitte CE, Wu KC. ECG scarquantification correlates with cardiac magnetic resonance scar size andprognostic factors in Chagas' disease. Heart 2011;97:357.

4. Wagner GS, Freye CJ, Palmeri ST, et al. Evaluation of a QRS scoringsystem for estimating myocardial infarct size. I. Specificity and observeragreement. Circulation 1982;65:342.

5. Ideker RE, Wagner GS, Ruth WK, et al. Evaluation of a QRS scoringsystem for estimating myocardial infarct size. II. Correlation withquantitative anatomic findings for anterior infarcts. Am J Cardiol1982;49:1604.

6. Roark SF, Ideker RE, Wagner GS, et al. Evaluation of a QRS scoringsystem for estimating myocardial infarct size. III. Correlation withquantitative anatomic findings for inferior infarcts. Am J Cardiol1983;51:382.

7. Ward RM, White RD, Ideker RE, et al. Evaluation of a QRS scoringsystem for estimating myocardial infarct size. IV. Correlation with

554 Z. Loring et al. / Journal of Electrocardiology 44 (2011) 544–554

quantitative anatomic findings for posterolateral infarcts. Am J Cardiol1984;53:706.

8. Strauss DG, Selvester RH. The QRS complex—a biomarker that“images” the heart: QRS scores to quantify myocardial scar in thepresence of normal and abnormal ventricular conduction. J Electro-cardiol 2009;42:85.

9. Roubin GS, ShenWF, Kelly DT, Harris PJ. The QRS scoring system forestimating myocardial infarct size: clinical, angiographic and prognosticcorrelations. J Am Coll Cardiol 1983;2:38.

10. Bounous Jr EP, Califf RM, Harrell Jr FE, et al. Prognostic value of thesimplified Selvester QRS score in patients with coronary artery disease.J Am Coll Cardiol 1988;11:35.

11. Fioretti P, Tijssen JG, Azar AJ, et al. Prognostic value of predischarge12 lead electrocardiogram after myocardial infarction compared withother routine clinical variables. Br Heart J 1987;57:306.

12. Tjandrawidjaja MC, Fu Y, Westerhout CM, Wagner GS, Granger CB,Armstrong PW. Usefulness of the QRS score as a strong prognosticmarker in patients discharged after undergoing primary percutaneouscoronary intervention for ST-segment elevation myocardial infarction.Am J Cardiol 2010;106:630.

13. Richardson K, Engel G, Yamazaki T, Chun S, Froelicher VF.Electrocardiographic damage scores and cardiovascular mortality. AmHeart J 2005;149:458.

14. Barbagelata A, Califf RM, Sgarbossa EB, et al. Prognostic value ofpredischarge electrocardiographic measurement of infarct size afterthrombolysis: insights from GUSTO I Economics and Quality of Lifesubstudy. Am Heart J 2004;148:795.

15. Strauss DG, Poole JE, Wagner GS, et al. An ECG index of myocardialscar enhances prediction of defibrillator shocks: an analysis of theSudden Cardiac Death in Heart Failure Trial. Heart Rhythm 2011;8:38.

16. Sweeney MO, van Bommel RJ, Schalij MJ, Borleffs CJ, Hellkamp AS,Bax JJ. Analysis of ventricular activation using surface electrocardiog-raphy to predict left ventricular reverse volumetric remodeling duringcardiac resynchronization therapy. Circulation 2010;121:626.

17. Bristow MR, Saxon LA, Boehmer J, et al. Cardiac-resynchronizationtherapy with or without an implantable defibrillator in advanced chronicheart failure. N Engl J Med 2004;350:2140.

18. Selvester RH, Wagner GS, Ideker RE. Myocardial infarction. In:Macfarlane PW, Lawrie TDV, editors. Comprehensive electrocardiol-ogy: theory and pracice in health and disease. New York: PergammonPress; 1989. p. 565.

19. Surawicz B, Childers R, Deal BJ, Gettes LS. AHA/ACCF/HRSrecommendations for the standardization and interpretation of theelectrocardiogram: part III: intraventricular conduction disturbances ascientific statement from the American Heart Association Electrocar-diography and Arrhythmias Committee, Council on Clinical Cardiol-ogy; the American College of Cardiology Foundation; and the HeartRhythm Society endorsed by the International Society for Computer-ized Electrocardiology. J Am Coll Cardiol 2009;53:976.

20. Selvester RH, Strauss DG, Wagner GS. Myocardial infarction. In:Macfarlane PW, Van Oosterom A, Pahlm O, Kligfield P, Janse M,Camm J, editors. Comprehensive Electrocardiology. London: Springer-Verlag; 2011:651-746.

21. Strauss DG, Selvester RH, Wagner GS. Defining left bundle branchblock in the era of cardiac resynchronization therapy. Am J Cardiol2011;107:927.

22. Rosenbaum MB, Elizari MV, Levi RJ, et al. Five cases of intermittentleft anterior hemiblock. Am J Cardiol 1969;24:1.

23. Macfarlane P, Lawrie TDV, editors. Comprehensive electrocardiology.NewYork: Pergamon; 2011, Appendix 1: Adult Normal Limits 2057-2125.

24. Lewczuk J, Ajlan AW, Piszko P, et al. Electrocardiographic signsof right ventricular overload useful in improving diagnosis ofchronic thromboembolic pulmonary hypertension (CTE-PH) inpatients with chronic exertional dyspnea. Pol Arch Med Wewn2002;108:1049.

25. Henkens IR, Mouchaers KT, Vonk-Noordegraaf A, et al. ImprovedECG detection of presence and severity of right ventricular pressureload validated with cardiac magnetic resonance imaging. Am J PhysiolHeart Circ Physiol 2008;294:H2150.