QT Interval: How to Measure It and What Is Normal

4 Pages·0.207 MB·English

by  Goldenberg I., Moss J.A., Zareba W.,

333
TECHNIQUES AND TECHNOLOGY
Editor:HughCalkins,M.D.
QT Interval: How to Measure It and What Is “Normal”
ILANGOLDENBERG,M.D.,ARTHURJ.MOSS,M.D.,andWOJCIECHZAREBA,M.D.,PH.D.
FromtheCardiologyDivision,DepartmentofMedicine,UniversityofRochesterMedicalCenter,Rochester,NewYork,USA
AbnormallylongandshortQTintervalshavebeenshown Theaccuracylevelsofmanualdeterminationwithacaliper
to be associated with an increased risk for life-threatening is20–40ms.Astandard12-leadECGtracingat25mm/spa-
ventriculararrhythmiasandsuddencardiacdeath.Inrecent perspeedat10mm/mVamplitudeisgenerallyadequatefor
years, various methods for QT-interval measurement have accuratemeasurementofQT-intervalduration.Higherspeeds
been developed, including individual-based corrections for (e.g., 50 mm/sec) may lead to distortion of low-amplitude
repolarization duration, quantitative assessment of repolar- waves such as U waves. The QT interval should be deter-
ization morphology, correction for repolarization dynamic- minedasameanvaluederivedfromatleast3–5cardiaccy-
ity,andanalysisofrepolarizationvariability.However,these cles(heartbeats),andismeasuredfromthebeginningofthe
methods require computer-processed digital-signal analysis earliestonsetoftheQRScomplextotheendoftheTwave.
ofelectronicallystoredECGdataandhavebeenusedmost TheQTmeasurementshouldbemadeinleadsIIandV5or
frequentlyintheassessmentofrepolarizationchangesindrug V6, with the longest value being used. The main difficulty
trials.Inthepresentreview,wewillfocusonmethodsforclin- lies in identifying correctly the point where the descending
icallyrelevantvisualandmanualassessmentofQT-interval limboftheTwaveintersectstheisoelectricline,particularly
durationfroma12-leadECG,whichcanbeutilizedinday-to- when there are T and U waves that are close together. We
daypracticeforthediagnosisoflongQTsyndrome(LQTS) identifytheendoftheTwavewhenitsdescendinglimbre-
andotherrepolarizationdisorders. turnstotheTPbaselinewhenitisnotfollowedbyaUwave
(Fig.1A)orifitisdistinctfromthefollowingUwave(Fig.
1B).WhenT-wavedeflectionsofequalornear-equalampli-
ECGAssessment
tuderesultinabiphasicTwave,theQTintervalismeasuredto
The 12-lead ECG is the most frequently used technique thetimeoffinalreturntobaseline(Fig.1C).Ifasecondlow-
forobtainingthesurfaceelectrocardiographicsignalforeval- amplituderepolarizationwaveinterruptstheterminalportion
uation of ventricular repolarization. Manual ECG readings oftheTwave(Fig.1D),itisdifficulttodeterminewhetherthe
areperformedusingvisualdeterminations(“eyeball”/caliper seconddeflectionisabiphasicTwaveoranearly-occurring
techniques),digitizingmethods,and/oron-screencomputer- U wave. In such cases, it is best to record both the QT
izedmethods.Theaccuracyoftheautomaticmeasurements interval (T-wave offset measured as the nadir between the
of the corrected QT (QTc) interval is questionable in many T and U wave) and the QTU interval (repolarization off-
casesandshouldbesupplementedbymanualreading.Incon- set measured at the end of the second wave). In general,
sistency between manufacturers in terms of the algorithms biphasic T waves are frequently present in multiple leads,
used for calculation of the intervals is another problem in whereas discrete and separate low-amplitude U waves are
the interpretation of computerized readings. Some digitiz- bestseeninthelateralprecordialleads.TheendoftheUwave
ingmethodsemployadigitizingpad,magnifyinglamp,and is defined as the intersection point of the descending limb
pointingdevicetoidentifythebeginningandendoftheQT of the U wave and the isoelectric baseline. This method
interval, with an accuracy level of 5 ms. A more techno- reflects accurately the real duration of ventricular repolar-
logically advanced option is to display digitally recorded ization, but it introduces a large degree of subjectivity, par-
ECGsonacomputerscreen,wheretheycanbemeasuredus- ticularly when biphasic T waves are present or when large
ingcomputer-driven,on-screencalipers.Thislatterapproach U waves interrupt the return of the T wave to the base-
provideshigh-qualityECGdataandisrecommendedatcore line.Themethodcanbeeffectivelyappliedformanualmea-
laboratoriesperformingcentralizedanalysesofalargeECG surements, but is less suitable for computer analysis be-
database. Scanned paper-recorded ECGs can also be sub- cause it requires the definition of a given threshold for the
jectedtoon-screenmeasurements. amplitude below which T or U wave potentials return to
baseline.
TheQRSintervalcanbemodifiedbyseveralfactors(such
JCardiovascElectrophysiol,Vol.17,pp.333-336,March2006. as bundle branch block, Class 1c antiarrhythmic drugs, or
preexcitation); these changes in depolarization can alter re-
Address for correspondence: Ilan Goldenberg, M.D., Heart Research polarization in unexpected ways, and thus the QT interval
Follow-up Program, Box 653, University of Rochester Medical Cen-
maynotbeanaccuratereflectionofrepolarizationduration.
ter,Rochester,NY14642.Fax:[phone];E-mail:Ilan.Goldenberg@
Inthesepatients,themeasureoftheJTfromtheS-waveoff-
heart.rochester.edu.
settoT-waveendmaybeused,butnormalstandardsforthe
doi:10.1111/j.1540-8167.2006.00408.x JTintervalarenotwellestablished.
334 JournalofCardiovascularElectrophysiology Vol.17,No.3,March2006
Figure1. (A)WhentheT-wavemorphologyisnormal,theT-waveoffsetisidentifiedwhenthedescendinglimbreturnstotheTPbaseline;(B)whentheT
waveisfollowedbyadistinctUwave,theT-waveoffsetisidentifiedwhenthedescendinglimboftheTwavereturnstotheTPbaselinebeforetheonsetofthe
Uwave;(C)whentheTwaveisbiphasicwithT1andT2wavesofsimilaramplitude,theT-waveoffsetisidentifiedatthetimewhenT2returnstobaseline;
and(D)whenasecondlow-amplituderepolarizationwaveinterruptstheterminalportionofthelargerTwave(?whetheritshouldbecategorizedasT2
waveoraUwave),theT-waveoffsetshouldbemeasuredbothatthenadirofthetwowaves(1)andatthefinalreturntobaseline(2).
AdjustmentforHeartRate RRintervalof1.0second).Theseformulaehavebeenderived
mainlyfromrestingECGs,andthereforerequireastablesi-
The time–duration intervals are influenced by heart rate nusrhythmwithoutsuddenchangesintheRRinterval.Ex-
(R–Rcyclelength),soheartratecorrectionisrequiredinthe ponential,logarithmic,andlinearformulaehavebeenused1
analysis of repolarization duration. Various heart rate cor- (Table1).Toassesstheperformanceofaparticularheartrate
rectionformulaehavebeendevelopedinordertodetermine correctionformula,thecorrelationbetweenthecorrectedQT
whethertheQTintervalisprolongedincomparisontoitspre- (QTc)intervalscalculatedusingtheformulaandtheRRin-
dictedvalueatareferenceheartrateof60beats/min(i.e.,an tervalscanbeassessed.Ifitdiffersfromzero,asisthecase
TABLE1
QT-HeartRateCorrectionFormulas1
Method Formula Comment
Exponential
Bazett QTc=QT/RR1/2 Widelyused;maygiveerroneousresultsatbothslowandfastheartrates
Fridericia QTc=QT/RR1/3 Widelyused;maygivemoreconsistentresultsatfastheartrates
Linear
Framingham QTc=QT+0.154(1-RR) Mayhavemoreuniformratecorrectionoverawiderangeofheartrates
Hodges QTc=QT+1.75(HR-60)
Rautaharju Mayhavemoreuniformratecorrectionoverawiderangeofheartrates
Femalesandmales<15 QTI=(QT[HR+100])/656
and>50years QTI=(QT[HR+100])/656
Males15–50years QTI=100(QT)/([656/(1+0.01HR])+0.4–25)
Logarithmic
Ashman QT=K1×log(10×[RR+K2]) Atlowheartrates,thevaluesaretoolow
Adultmen K2=0.07,andK1=0.380
Adultwomen K2=0.07,andK1=0.390
Goldenbergetal. TechniquesandTechnology 335
with most of the above-described formulae, the correction
formulaisnottrulysuccessful.Wemostcommonlycorrect
the interval by using Bazett’s square root formula. QTc is
equaltoQTintervalinsecondsdividedbythesquarerootof
theprecedingRRintervalinseconds.Whenheartrateispar-
ticularly fast or slow, the Bazett’s formula may overcorrect
orundercorrect,respectively,butitremainsthestandardfor
clinical use. The cube root Fridericia formula has the same
limitations at slow heart rates, but is considered to reflect a
moreaccuratecorrectionfactorinsubjectswithtachycardia.
Linear formulae may have more uniform correction over a
wide range of heart rate. The most commonly used linear
formuladerivesfromtheFraminghamstudy.Thelatterfor-
mulae may give QT values that are too low at slow heart
rates. There is no general consensus on the best formula to
beutilizedinclinicalpractice.Ofnote,inrestingconditions
with heart rates in the 60–90 beats/min range, most formu-
laeprovidealmostequivalentresultsforthediagnosisofQT
Figure 2. Lower and upper limits of QT interval for different RR cycle
prolongation.EveniftheratedependenceoftheQTinterval lengthsbasedonQT-intervalmeasurementsinnormalhealthysubjects.
isprobablybestdescribedbyanexponentialrelation,inthe
normalheartraterangetheQT–RRrelationisapproximately
linear. Asuggestedthree-levelclassificationbasedonthisanalysis
ispresentedinTable2.
HeartRateCorrectioninPatientswithSinus AsimplegraphicaldisplayoflowerandupperlimitsofQT
Arrhythmias interval for different RR cycle lengths based on population
studies has been developed in our laboratory (Fig. 2). The
If a stable sinus rhythm cannot be obtained in a patient,
datacanbeusedbycliniciansforreferencepurposes.
more advanced methods evaluating repolarization dynam-
icity may be required. The QT interval adapts to heart rate
RepolarizationMorphology
changeswithadelayknownasQThysteresisorQTlag.When
thechangeintheheartratepersistsforseveralminutes,the Recentadvanceshavebeenmadeinquantitatingrepolar-
QT lag is visible on the trend of QT and RR intervals. The ization using such measurements as the symmetry of the T
QTadaptsmoreslowlytodecelerationsthantoaccelerations wave, T-wave area, or the interval between the end of the
oftheheartrate.TheplotofQTversusRRintervalsduring S wave and the maximum amplitude of the T wave. How-
dynamic adaptation of repolarization to heart rate changes ever, quantitative analysis of the T-wave shape and pattern
formsaloopknownashysteresis.QT–RRhysteresispattern requires computer software and electronically stored ECG
is highly individual and, therefore, methods that take into data.
accountindividualprofilesarerequired. Themorphologyorconfigurationofrepolarizationcanbe
described from visual inspection of the T wave and placed
NormalValuesoftheQTInterval(12-LeadECG) intoprespecifiedcategories.Inourassessmentofventricular
repolarization,weincorporatethisinformationintothedata
Bazett’sformulahasbeenmorefrequentlyusedinmed-
obtainedfromQT-intervaldurationmeasurement.Distinctive
ical publications than Fridericia’s formula. Therefore, most
T-wavepatternshavebeenobservedinpatientswitheachof
reported criteria for normal and abnormal values for QTc
thethreemajorLQTSgenotypes(Fig.3).3InLQT1asingle,
are derived from Bazett’s formula. Our research group has
smooth, broad-based T wave is common, as well as a late-
utilizeddigitizeddatafileforQT-andRR-intervalmeasure-
onsetnormal-appearingTwave;inLQT2,bifidTwavesare
ments on 581 healthy individuals: 158 children aged 1–15
ahallmarkECGfeature;inLQT3,theTwavesaretypically
years (80 boys and 78 girls) and 423 adults aged 16-81
late-onset,prominent,andusuallypeaked.
years (223 men and 200 women). Using this data set, we
evaluated the range of Bazett QTc values by age and gen-
OtherECGRecordingTechniques
der. The QTc values were stable for children, with no gen-
der difference, while there was a significant difference be-
Holter and exercise testing have also been used to eval-
tween adult men and women in this healthy population.2
uate the QT interval. Holter monitoring is not sufficiently
wellstandardizedtoserveintheprimaryassessmentforven-
tricularrepolarizationanalysis.However,wesometimesem-
TABLE2 ploy this method for the detection of extreme QT-interval
SuggestedBazett-CorrectedQTcValuesforDiagnosingQTProlongation events that occur infrequently during the day. Since QT in-
tervals measured using the Holter methodology do not cor-
1–15years AdultMale AdultFemale respondquantitativelytothoseforstandardECGs,dataob-
Rating (msec) (msec) (msec)
tainedfromthetwomethodologiesarenotsuitablefordirect
Normal <440 <430 <450 comparison.Exercisetestingwithastandardactivityproto-
Borderline 440–460 430–450 450–470 col can be used for evaluation of QT prolongation during
Prolonged >460 >450 >470
exerciseandrecoveryperiods.Intermittent12-leadECGsor
336 JournalofCardiovascularElectrophysiology Vol.17,No.3,March2006
Figure 3. Specific LQTS T-wave patterns:
broad-based T-wave pattern in LQT1, bi-
fid T waves in LQT2, and late-onset
peaked/biphasicTwavesinLQT3.Reprinted
with permission from Moss AJ, Zareba W,
BenhorinJ,etal.Circulation1995;92:2929-
2934.
continuousmultichannnelECGrecordingscanbeused.How- and RR intervals was achieved by 96% of QT experts and
ever, the adaptation of QT-interval duration to heart rate is 62% of arrhythmia experts, but by only <25% of cardiolo-
notinstantaneous,andsubstantialerrorsmaybeintroduced gists and noncardiologists. Clearly, experience and training
ifnonstationaryepisodesareanalyzed. play an important role in the accurate measurement of the
Recent analysis from the International LQTS Registry QTcinterval.
demonstrates that there is individual subject variability in
QTc duration on repeat ECGs during long-term follow-up
extendingoverseveralyears.Therefore,wesuggestthatsev- References
eralECGsrecordedovertimeshouldbemoreusefuliniden-
tifying subjects with abnormally long or short QT intervals 1. AhnveS:CorrectionoftheQTintervalforheartrate:Reviewofdifferent
formulasandtheuseofBazett’sformulainmyocardialinfarction.Am
thansimplyonebaselineECGrecording.
HeartJ1985;109:568-574.
2. MossAJ:MeasurementoftheQTintervalandtheriskassociatedwith
ConclusionsandRecommendations QTcintervalprolongation:Areview.AmJCardiol1993;72:23B-25B.
3. ArthurJM,WojciechZ,JesaiaB,EmanuelaHL,W.JacksonH,Jen-
niferLR,PeterJS,JeffreyAT,G.MichaelV,MichaelHL,MarkTK,
The simple measurement of the QT interval is valuable
JeanWM, KatherineWT:ECGT-wavepatternsingeneticallydistinct
for the diagnosis of abnormal QTc intervals. However, the
formsofthehereditarylongQTsyndrome.Circulation1995;92:2929-
routine measurement of the QT interval requires the use of 2934.
uniform criteria for the determination of T-wave offset (es- 4. ViskinS,RosovskiU,SandsAJ,ChenE,KistlerPM,KalmanJM,Ro-
peciallywhenthereispartialsuperimpositionoftheTandU driguezChavezL,IturraldeTorresP,CruzFFE,CenturionOA,Fujiki
A,MauryP,ChenX,KrahnAD,RoithingerF,ZhangL,VincentGM,
wave),adjustmentforheartrate,andT-wavemorphology.
ZeltserD:InaccurateelectrocardiographicinterpretationoflongQT:
Inarecentstudy,4correctclassificationofQTintervals(ei- ThemajorityofphysicianscannotrecognizealongQTwhentheysee
ther“long”or“normal”)usingonlymanuallymeasuredQT one.HeartRhythm2005;2:569-574.