Download Accuracy of manual QRS duration assessment

Europace (2009) 11, 638–642
doi:10.1093/europace/eup001
CLINICAL RESEARCH
Non-invasive Electrophysiology and Risk Stratification
Accuracy of manual QRS duration assessment: its
importance in patient selection for cardiac
resynchronization and implantable cardioverter
defibrillator therapy
David R. Tomlinson*, Yaver Bashir, Timothy R. Betts, and Kim Rajappan*
Department of Cardiology, John Radcliffe Hospital, Headley Way, Headington, Oxford OX3 9DU, UK
Received 25 September 2008; accepted after revision 30 December 2008; online publish-ahead-of-print 26 January 2009
Aims
Patients with left ventricular systolic dysfunction and electrocardiographic QRS duration (QRSd) 120 ms may
obtain symptomatic and prognostic benefits from cardiac resynchronization therapy (CRT). However, clinical trials
do not describe the methods used to measure QRSd. We investigated the effect of electrocardiogram (ECG)
display format and paper speed on the accuracy of manual QRSd assessment and concordance of manual QRSd
with computer-calculated mean and maximal QRSd.
.....................................................................................................................................................................................
Methods
Six cardiologists undertook QRSd measurements on ECGs, with computer-calculated mean QRSd close to 120 ms.
Display formats were 12-lead, 6-limb, and 6-precordial leads, each at 25 and 50 mm/s. When the computer-calculated
and results
mean was used to define QRSd, manual assessment demonstrated 97 and 83% concordance at categorizing QRSd as
, and 120 ms, respectively. Using the computer-calculated maximal QRSd, manual assessment demonstrated 83%
concordance when QRSd was ,120 ms and 19% concordance when QRSd was 120 ms. The six-precordial lead
format demonstrated significantly less intra and inter-observer variabilities than the 12-lead, but this did not improve
concordance rates.
.....................................................................................................................................................................................
Conclusion
Manual QRSd assessments demonstrate significant variability, and concordance with computer-calculated measurement depends on whether QRSd is defined as the mean or maximal value. Consensus is required both on the most
appropriate definition of QRSd and its measurement.
----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords
Cardiac resynchronization therapy † Implantable cardioverter defibrillator † Electrocardiogram † QRS duration
Introduction
In patients with left ventricular systolic dysfunction (LVSD), prolongation of the electrocardiographic QRS duration (QRSd) is an
independent predictor of all-cause mortality1,2 and possibly
sudden cardiac death.3 In such patients, an electrocardiographic
QRS interval of .120 ms is also commonly associated with
delayed ventricular activation, resulting in intra-ventricular dyssynchrony.4 In this setting, clinical trials of cardiac resynchronization
therapy (CRT)—principally conducted in patients with a left
bundle branch pattern of prolonged ventricular conduction—
have demonstrated important symptomatic and prognostic
benefits.5,6 Implantable cardioverter defibrillators (ICDs) also
confer prognostic benefit as secondary prevention following
aborted or resuscitated sudden cardiac death7 – 9 and as primary
prevention in those patients with severe LVSD but without prior
life-threatening ventricular arrhythmias.10,11 Furthermore, the
three most recently published primary prevention ICD trials
each demonstrate a trend towards a greater absolute benefit in
patients with QRSd 120 ms.10 – 12 These data have now been
incorporated into UK national guidelines, which recommend consideration of CRT in patients with New York Heart Association
(NYHA) class III heart failure despite optimal medical therapy,
left ventricular ejection fraction (LVEF) ,35%, and QRSd
* Corresponding authors. Tel: þ44 1865 220256, Fax: þ44 1865 221194, Email: [email protected] (D.R.T.) or [email protected] (K.R.)
Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2009. For permissions please email: [email protected].
639
Accuracy of manual QRSd assessment
120 ms,13 as well as prophylactic ICD implantation in patients
with ischaemic cardiomyopathy, an LVEF of ,30%, and again
QRSd 120 ms.14
However, the specific methods for QRSd measurement were
not described in these publications, and international guidelines
do not specify a preferred measurement technique.13 – 16 This is
of concern because when the QRSd is close to 120 ms, the
decision to recommend de novo ICD implantation or undertake
further risk stratification investigations is critically dependent on
the ECG interpretation. If QRSd is ,120 ms, this may include
Holter monitoring and electrophysiological testing.17 Thus, inaccurate QRSd classification may increase costs by exposing patients
to otherwise unnecessary invasive procedures or deny some
patients the beneficial effects of CRT and prophylactic ICD.
Finally, the lack of a clear guideline on methods for QRSd measurement may create intra- and inter-individual inconsistencies in
device prescription.
Methods
Study design
The effect of ECG display format and paper speed on manual QRSd
measurement was investigated in a selection of ECGs, each with a
computer-calculated mean QRSd close to 120 ms. Chosen paper
formats were standard 12-lead, 6-limb, or 6-precordial lead tracings,
all separately recorded at 25 and 50 mm/s paper speeds. FX-4101U
or FX-4010 ECG machines (Fukuda Denshi Co., Japan) were used in
all subjects, and ‘on line’ computer-extended analysis calculated both
the mean and individual lead QRSd.
Electrocardiograms
The study ECG series comprised the following: (i) five subjects’ ECGs,
each with a mean calculated QRSd ,120 ms, but with maximal QRSd
120 ms in at least one lead; (ii) one subject’s ECG with both mean
and maximal calculated QRSd 120 ms, ‘positive control’; (iii) one
subject’s ECG with both mean and maximal calculated QRSd
,120 ms, ‘negative control’.
Electrocardiogram analysis
Six cardiologists undertook manual QRSd analysis. All study ECGs
were anonymized, printed on standard ECG paper, and presented in
a random order. Clinicians were asked to undertake measurement
of the QRSd using their usual technique and also to record the lead
used for final measurement. None was asked to specifically record
whether the QRSd was ,120 or 120 ms, nor were they aware
that this particular interval was the subject of the study. ECG measurements were then classified according to this interval for study purposes, following a complete return of analysed ECGs. Intra- and
inter-observer variabilities for QRSd measurement were calculated
using an additional subject’s ECG with mean QRSd ,120 ms, but
maximal calculated QRSd 120 ms; four copies in each different
format were randomly distributed throughout the study ECG series.
Study outcomes
The primary study outcome was concordance between manually
assessed and computer-calculated QRSd, with the latter defined in
two ways—both as the mean and maximum value. Secondary outcomes were the intra- and inter-observer variabilities of QRSd
measurement and concordance in categorizing QRSd as ,120 or
120 ms, according to the ECG format. None of the clinicians was
aware of the primary or secondary study outcomes.
Statistical analysis
The primary outcome was expressed as raw concordance rates. Intraand inter-observer variabilities in absolute QRSd measurement
between ECG formats were analysed using the Kruskal – Wallis statistic
with Dunn’s multiple comparison test. Intra- and inter-observer concordance at defining QRSd as ,120 or 120 ms according to the
ECG format was analysed using x2 test for trend. A P-value less than
0.05 was considered to represent statistical significance.
Results
Electrocardiogram characteristics
Details of study ECGs with computer-calculated mean and range
of QRSd are shown in Figure 1. The lead with maximal calculated
QRSd was precordial leads V1– V4 in all cases. Manual QRSd
measurements were undertaken with a ruler in one study participant and by directly measuring from the paper visually (without
calipers) in the remaining five.
Concordance between manual
and calculated QRSd
When the computer-calculated QRSd was defined according to
the mean value and was ,120 ms, manual assessment demonstrated good concordance with no significant effect of changing
ECG format (Table 1). When the computer-calculated mean
QRSd was 120 ms, manual assessment demonstrated lower concordance and was dependent on ECG format. Concordance for
the 12-lead format at 25 mm/s was 83%; however, both six-limb
and six-precordial formats had poorer concordance (0–67%),
with the greatest effect evident at 50 mm/s paper speed.
When the computer-calculated QRSd was defined according to
the maximal value and was ,120 ms, manual assessment demonstrated concordance rates of 83– 100% for all ECG formats.
However, when the computer-calculated maximal QRSd was
120 ms, manual assessment demonstrated concordance rates
of 0–19%, and greatest concordance occurred using the 12-lead
format at 25 mm/s paper speed.
Intra-observer variability
There was no significant difference in intra-observer variability of
manual QRSd assessment between ECG formats, although there
was slightly greater variability for the 12-lead, 25 mm/s format
(Table 2). Furthermore, in this format, only four of six clinicians
provided manual QRSd measurements, resulting in 100% concordance at categorizing ECGs as either ,120 or 120 ms. In contrast, when using a six-precordial ECG format, all manual QRSd
measurements had 100% concordance, regardless of the paper
speed (x2 test for trend, P ¼ 0.03).
Inter-observer variability
There was significantly greater inter-observer variability in
measured QRSd from the 12-lead format compared with the
precordial lead format at 25 mm/s paper speed (Table 3).
Furthermore, the inter-observer agreement in categorizing ECGs
640
D.R. Tomlinson et al.
Figure 1 Study electrocardiograms (ECGS) with mean calculated and ranges of QRS duration for each display format (ms, milliseconds); four
copies of ECG 1 were used to determine intra- and inter-observer observer variabilities, whereas ECGs 7 and 8 were positive and negative
controls, respectively.
Table 1 Raw concordance between measured and
calculated QRS duration (QRSd) in categorizing the
QRSd as either <120 or 120 ms, when the ‘gold
standard’ QRSd was defined as either the mean or
maximum calculated value
ECG format
Median
intra-observer
variability (range)
(ms)
Clinicians with
concordance for
QRSd
categorization: n (%)
.........................................................
12-lead 25 mm/s
25 (10–50)
4 (67)*
Calculated mean
QRSd
Calculated
maximum QRSd
12-lead 50 mm/s
Limb 25 mm/s
10 (0– 25)
20 (5– 40)
5 (83)
5 (83)
<120 ms
<120 ms
120 ms
Limb 50 mm/s
20 (10–30)
6 (100)
V1– V6 25 mm/s
V1– V6 50 mm/s
20 (0– 25)
12.5 (10–35)
6 (100)
6 (100)
120 ms
..........................
................................................................................
12-lead 25 mm/s
92
83
83
19
12-lead 50 mm/s
Limb 25 mm/s
100
97
67
0
100
100
6
6
97
0
96
0
100
100
67
33
100
100
11
6
V1 –V6 25 mm/s
V1 –V6 50 mm/s
ECG format
Raw concordance between measured and
calculated QRSd (%)
..........................
Limb 50 mm/s
Table 2 Intra-observer variability of manually assessed
QRS duration
according to the QRSd was also dependent on the format, with a
significant trend observed against a 12-lead, 25 mm/s display.
Overall, complete inter-observer concordance for QRSd categorization as ,120 or 120 ms was only achieved in two of four
ECGs in a 12-lead format compared with four of four ECGs
using the six-precordial lead format at 25 mm/s paper speed.
................................................................................
The number and percentage of clinicians with manual assessments resulting in
100% concordance for categorizing QRSd as either ,120 or 120 ms are also
shown, according to the ECG format.
*P ¼ 0.03, for trend.
Variation in choice of electrocardiogram
leads for manual QRSd assessment
For analysis of ECGs in a 12-lead format, no study participant consistently measured QRSd from the same lead; however, the
majority of measurements were taken from leads V1–V4. Using
the four identical ECGs from which intra- and inter-observer variabilities were determined and at 25 and 50 mm/s paper speed,
641
Accuracy of manual QRSd assessment
Table 3 Inter-observer variability of manually assessed QRS duration
ECG format
Median inter-observer
variability (range) (ms)
QRSd assessments: n (%)
Number of ECGs with concordance
12-lead 25 mm/s
35 (20–50)
21 (88)
3 (12)
2 (50)†
12-lead 50 mm/s
22.5 (20–40)
23 (96)
1 (4)
3 (75)
Limb 25 mm/s
Limb 50 mm/s
35 (20–40)
40 (30–40)
23 (96)
24 (100)
1 (4)
0
3 (75)
4 (100)
V1 –V6 25 mm/s
12.5 (10–20)*
24 (100)
0
4 (100)
V1 –V6 50 mm/s
12.5 (0 –30)
24 (100)
0
4 (100)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . for QRSd categorization: n (%)
<120 ms
120 ms
...............................................................................................................................................................................
Inter-observer variation in QRSd categorization as ,120 or 120 ms is also shown, according to the ECG format (six observers each analysed four identical ECGs in each format,
hence 24 observations were made).
*P ¼ 0.04, inter-observer variability V1 –V6 compared with 12-lead, at 25 mm/s.
†
P ¼ 0.02, for trend.
QRSd was assessed from one of leads V1–V4 in 58 and 75% of
measurements, a limb lead in 29 and 21%, and lead V5 or V6 in
13 and 4%, respectively. Notably, inter-observer variability in the
choice of ECG leads for QRSd assessment resulted in no instances
where all participants used the same lead for any individual ECG.
Discussion
These data demonstrate important differences in the concordance
rates between manual and computer-calculated QRSd, depending
on whether QRSd was defined according to the mean or
maximal computer-calculated value. Manual assessment showed
good concordance with computer-calculated measurements
(.90%) when the QRSd was ,120 ms. However, when the
computer-calculated QRSd was 120 ms, manual assessment
demonstrated lower concordance, particularly when QRSd was
defined according to the maximal value.
Furthermore, there was also significant intra- and inter-observer
variabilities in absolute QRSd measurement. This finding may partly
be explained by the variation in the choice of ECG lead used for
manual assessments, as changing to a six-precordial lead format
improved both intra- and inter-observer variabilities and concordance rates. In this regard, it is worth noting the significant variation in the computer-calculated QRSd measurements between
individual ECG leads. Although the mean QRSd was ,120 ms,
the computer-calculated QRSd was 120 ms in at least one lead
(usually from leads V1–V4). Thus, manual assessment of QRSd
from a standard 12-lead ECG format may be subject to significant
variability, simply due to a lack of standardization of the approach
to measuring this interval.
Clinical implications
The findings of the present study are of great relevance to current
guidelines on patient selection for both CRT and ICD. Recent published data on CRT for patients with narrow QRS (,120 ms) have
reported conflicting results. Some case series have demonstrated
improvement in both LV dimensions and symptoms in unselected
patients18 or those selected on the basis of echocardiographic
dyssynchrony.19,20 However, in the RethinQ study,21 subgroup
analysis of patients with narrow QRS demonstrated no difference
between controls and those receiving CRT in the ‘hard’ endpoints
of peak oxygen consumption, LVEF, and dimensions. Consequently,
present guidelines define an ECG ‘QRS interval’ of 120 ms as
the lower cut-off criterion for undertaking echocardiographic
dyssynchrony assessments. In those patients with a QRSd
between 120 and 150 ms demonstrating dyssynchrony, significant
LVSD, and with NYHA III/IV symptoms, CRT implantation is
recommended.13,16,22 For prophylactic ICD implantation, ACC/
AHA/ESC guidelines do not specify an ECG QRS interval.23
However, UK National Institute for Health and Clinical Excellence
(NICE) guidelines for patients with ischaemic heart disease reserve
ICD for those patients with LVEF ,30% and ‘QRS duration
120 ms’.14 Notably, no guideline explicitly recommends a specific
method for QRSd assessment.
Importantly, clinical trials of both the ICD and CRT have utilized
entry criteria of a ‘measured QRS interval’5 or ‘any two leads of a
12-lead ECG with a QRSd 120 ms’.6 Although there is evidence
for QRSd assessment to be based on the maximal QRSd in one or
two leads, this was precisely the area where manual assessment
demonstrated poorest concordance with the computer-calculated
QRSd. In fact, apart from the ‘negative control’, all ECGs in this
present study were eligible for entry into either COMPANION6
or CARE-HF;5 however, most manual assessments resulted in an
ECG QRSd classification of ,120 ms.
Study limitations
The calculated QRSd was assessed using a single ECG machine
manufacturer (Fukuda Denshi Co.); so possible differences in
QRSd calculations between different manufacturers’ algorithms
were not investigated. This issue is worthy of further study, as
one possible recommendation from this present study is that all
patients with a QRSd close to 120 ms and under consideration
for CRT should undergo computer-calculated ECG QRSd
assessment. However, this would only be appropriate if there
were no important inter-manufacturer differences in calculated
QRSd measurements.
642
Conclusions
Manual QRSd assessments demonstrate significant variability, and
concordance with computer-calculated measurement depends
on whether QRSd is defined as the mean or maximal value. We
propose that guidelines should be updated to reflect a consensus
view on how best to measure QRSd (manual vs. computercalculated) and whether to use the mean or maximal QRSd to
better standardize decision-making in cardiac rhythm management
device implantation. Without this initiative, our data suggest that—
based on inaccurate and inconsistent manual QRSd measurements—patients may be denied device therapy for which they
may otherwise have been eligible, potentially resulting in adverse
outcomes on both morbidity and mortality.
Acknowledgements
We thank Lyn Barrett, Cardiac Physiologist at the ECG Department, John Radcliffe Hospital, Oxford, UK, for performing the
ECGs used in this study.
Conflict of interest: none declared.
References
1. Shenkman HJ, Pampati V, Khandelwal AK, McKinnon J, Nori D, Kaatz S et al. Congestive heart failure and QRS duration: establishing prognosis study. Chest 2002;
122:528–34.
2. Kalra PR, Sharma R, Shamim W, Doehner W, Wensel R, Bolger AP et al. Clinical
characteristics and survival of patients with chronic heart failure and prolonged
QRS duration. Int J Cardiol 2002;86:225 – 31.
3. Iuliano S, Fisher SG, Karasik PE, Fletcher RD, Singh SN. QRS duration and
mortality in patients with congestive heart failure. Am Heart J 2002;143:1085 – 91.
4. Cazeau S, Leclercq C, Lavergne T, Walker S, Varma C, Linde C et al. Effects of
multisite biventricular pacing in patients with heart failure and intraventricular
conduction delay. N Engl J Med 2001;344:873–80.
5. Cleland JG, Daubert JC, Erdmann E, Freemantle N, Gras D, Kappenberger L et al.
The effect of cardiac resynchronization on morbidity and mortality in heart
failure. N Engl J Med 2005;352:1539 –49.
6. Bristow MR, Saxon LA, Boehmer J, Krueger S, Kass DA, De Marco T et al.
Cardiac-resynchronization therapy with or without an implantable defibrillator
in advanced chronic heart failure. N Engl J Med 2004;350:2140 –50.
7. The Antiarrhythmics Versus Implantable Defibrillators (AVID) Investigators. A
comparison of antiarrhythmic-drug therapy with implantable defibrillators in
patients resuscitated from near-fatal ventricular arrhythmias. N Engl J Med 1997;
337:1576 –83.
8. Kuck KH, Cappato R, Siebels J, Ruppel R. Randomized comparison of antiarrhythmic drug therapy with implantable defibrillators in patients resuscitated
from cardiac arrest: the Cardiac Arrest Study Hamburg (CASH). Circulation
2000;102:748 –54.
9. Connolly SJ, Gent M, Roberts RS, Dorian P, Roy D, Sheldon RS et al. Canadian
Implantable Defibrillator Study (CIDS): a randomized trial of the implantable
cardioverter defibrillator against amiodarone. Circulation 2000;101:1297 –302.
10. Moss AJ, Zareba W, Hall WJ, Klein H, Wilber DJ, Cannom DS et al. Prophylactic
implantation of a defibrillator in patients with myocardial infarction and reduced
ejection fraction. N Engl J Med 2002;346:877 –83.
D.R. Tomlinson et al.
11. Bardy GH, Lee KL, Mark DB, Poole JE, Packer DL, Boineau R et al. Amiodarone or
an implantable cardioverter-defibrillator for congestive heart failure. N Engl J Med
2005;352:225 –37.
12. Kadish A, Dyer A, Daubert JP, Quigg R, Estes NA, Anderson KP et al. Prophylactic
defibrillator implantation in patients with nonischemic dilated cardiomyopathy. N
Engl J Med 2004;350:2151 – 8.
13. Cardiac resynchronisation therapy for the treatment of heart failure: NICE Technology appraisal guidance 120. www.nice.org.uk/TA120 (2007).
14. Implantable Cardioverter Defibrillators for Arrhythmias: NICE Technology
appraisal guidance 95. www.nice.org.uk/TA095 (January 2006).
15. Radford MJ, Arnold JM, Bennett SJ, Cinquegrani MP, Cleland JG, Havranek EP et al.
ACC/AHA key data elements definitions for measuring the clinical management
outcomes of patients with chronic heart failure: a report of the American
College of Cardiology/American Heart Association Task Force on Clinical Data
Standards (Writing Committee to Develop Heart Failure Clinical Data Standards): developed in collaboration with the American College of Chest Physicians
the International Society for Heart Lung Transplantation: endorsed by the Heart
Failure Society of America. Circulation 2005;112:1888 – 916.
16. Epstein AE, DiMarco JP, Ellenbogen KA, Estes NA III, Freedman RA, Gettes LS
et al. ACC/AHA/HRS 2008 Guidelines for device-based therapy of cardiac
rhythm abnormalities: a report of the American College of Cardiology/American
Heart Association Task Force on Practice Guidelines (Writing Committee to
Revise the ACC/AHA/NASPE 2002 Guideline Update for Implantation of
Cardiac Pacemakers and Antiarrhythmia Devices): developed in collaboration
with the American Association for Thoracic Surgery and Society of Thoracic Surgeons. Circulation 2008;117:e350–408.
17. Moss AJ, Hall WJ, Cannom DS, Daubert JP, Higgins SL, Klein H et al. Improved
survival with an implanted defibrillator in patients with coronary disease at high
risk for ventricular arrhythmia. Multicenter Automatic Defibrillator Implantation
Trial Investigators. N Engl J Med 1996;335:1933 –40.
18. Gasparini M, Regoli F, Galimberti P, Ceriotti C, Bonadies M, Mangiavacchi M et al.
Three years of cardiac resynchronization therapy: could superior benefits be
obtained in patients with heart failure and narrow QRS? Pacing Clin Electrophysiol
2007;30:S34 – 9.
19. Bleeker GB, Holman ER, Steendijk P, Boersma E, van der Wall EE, Schalij MJ et al.
Cardiac resynchronization therapy in patients with a narrow QRS complex. J Am
Coll Cardiol 2006;48:2243 – 50.
20. Yu CM, Chan YS, Zhang Q, Yip GW, Chan CK, Kum LC et al. Benefits of cardiac
resynchronization therapy for heart failure patients with narrow QRS complexes
and coexisting systolic asynchrony by echocardiography. J Am Coll Cardiol 2006;48:
2251 –7.
21. Beshai JF, Grimm RA, Nagueh SF, Baker JH III, Beau SL, Greenberg SM et al.
Cardiac-resynchronization therapy in heart failure with narrow QRS complexes.
N Engl J Med 2007;357:2461 –71.
22. Hunt SA, Abraham WT, Chin MH, Feldman AM, Francis GS, Ganiats TG et al.
ACC/AHA 2005 Guideline Update for the Diagnosis and Management of
Chronic Heart Failure in the Adult: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing
Committee to Update the 2001 Guidelines for the Evaluation and
Management of Heart Failure): developed in collaboration with the American
College of Chest Physicians and the International Society for Heart and Lung
Transplantation: endorsed by the Heart Rhythm Society. Circulation 2005;112:
e154 –235.
23. Zipes DP, Camm AJ, Borggrefe M, Buxton AE, Chaitman B, Fromer M et al. ACC/
AHA/ESC 2006 Guidelines for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death: a report of the American
College of Cardiology/American Heart Association Task Force and the European
Society of Cardiology Committee for Practice Guidelines (writing committee to
develop guidelines for management of patients with ventricular arrhythmias and
the prevention of sudden cardiac death): developed in collaboration with the
European Heart Rhythm Association and the Heart Rhythm Society. Circulation
2006;114:e385 –484.