Computerized Misinterpretation of QT Interval in 12-Lead Electrocardiogram and Its Clinical Consequences:ACase of Recurrent Syncope

1. Abstract The diagnosis of long QT syndrome may be delayed due to absenceofelectrocardiogram(ECG)examinationand/orQTinterval misdiagnosis. We present the case of a patient who took 8 years before being diagnosed with long QT syndrome. Despite the presence of marked QT prolongation and bimodal Twaves, automated ECG recording misdiagnosed the patienttobenormal. The patient hadrecurrentsyn cope; however, repeat ECG examination was not performed because her initial ECG was considered normal. It is crucial in cases of syncope that multiple ECG examinations and physician assessment of findings are performed, without relying exclusively on an automated ECG diagnosis

Keywords: Automateddiagnosis; QTinterval; Syncope; Long QT syndrome

2.Abbreviations: ECG: Electrocardiogram; QTc: Corrected QT Interval; TdP: Torsades De Pointes

3. Introduction Patientswithcardiacsyncopehaveincreasedriskofdeath[1],and the occurrence of “syncope” is an important indicator of arrhyth- mic risk [1-3]. Patents with unexplained syncope may have hiddenlongQT syndrome (LQTS) [2].Althoughelectrocardiogram (ECG) isimportant for the diagnosisof LQTS, the diagnosis may be delayed due to absence of ECG examination and/or difficulties of QT interval evaluation [2]. Automated ECG diagnosis has the benefit to reduce the effort of physician for ECG evaluation [4]. However, the automated ECG diagnosis are not always correct,inparticular,QTinterval[5].Wepresentthecase of apatient for whom LQTS diagnosis took8 years. Despite the presence of marked QT prolongationand bimodal T waves, automated ECG diagnosis misdiagnosed the patient to be normal at the first synco- pe. The patient had recurrent syncope; however, repeat ECG ex- amination was not performed.

4. CaseReport A27-year-old female visited our emergency room in the middle of the night because she experienced syncope several times through- out the preceding 2 days upon waking up and when attempting walking at night. At the time of the visit, her consciousness lev- el was clear and plasma potassium concentration was 4.0 mEq/L (normal range: 3.6–4.8 mEq/L). Her 12-lead electrocardiogram(ECG)revealed sinusrhythm(65 beats/min) and QT prolongation with low amplitude T waves (QT/QTc [intervals], 600/640 ms) (Figure 1A). Echocardiography, chest X-ray, and brain magnetic resonance imaging (MRI) findings were normal. She was admitted for follow-up due to the presence of marked QT prolongation on12- leadECG.Fivehoursaftertheadmission,torsadesdepointes (TdP) spontaneously occurred with typical initiating “short- longshort” sequences (Figure 1B) when she felt faintness. Therefore, the patient was clinically diagnosed with long QT syndrome. She had no abnormal ECG findings in past school physical examina- tion. She had no family history of syncope, sudden death, or long QT syndrome (Figure 2A). ECGs of her mother and son showed no QT prolongation (Figure 2A).At the age of 19, the patient experiencednightmaresduringsleep, which werefollowedbyconvulsionsandgroaning.Shefaintedtwiceafterwakinginthemorning. The patient visited a neurology clinic and underwent brain MRI, but the underlying cause remained unknown. At the age of 21,sheexperienceddizziness while walkingortryingto get upinthe morning. She experienced syncope and visited a hospital, where automated ECG diagnosis was within the normal range (Figures 3A, 3B). Brain computed tomography and electroencephalogram revealed no obvious abnormality. Therefore, the physician found no apparent cause for her syncope and she returned her home.The patient continued to have episodes of syncope several timesa year until she the age of 27. Throughout this period, she visited several hospitals and clinics. However, episodes of syncope due to long QT syndrome was not suspected, and the patient did not undergo another ECG examination nor was referred to a cardi- ologist. After admission to our hospital, the ECG recorded at the age of 21 was retrieved and re-evaluated. Her ECG revealed QT prolongationwithnotchedTwaves(Figure3A),afindingthatwas clearlyabnormal(QT/QTc,580/608ms).Geneticanalysisforlong QT syndromeincludingKCNQ1,KCNH2,SCN5A,KCNE1,and KCNE2 was performed after obtaining written informed consent fromthepatient.Anovelframeshiftvariantthatresultedfromduplication (c. 526dupC, p.R176fsX331) in KCNH2 was identified (Figure 2B). Her mother and son, who did not have QT prolongation, did not carry the variant (Figure 2A). The patient was diagnosed with type 2 long QT syndrome. Carvedilol was initiated from 5.0 mg per day, and the dose was gradually increased. One monthlater,herQTintervalhadshortened(QT/QTc,480/488ms), and notched T waves had disappeared (Figure 4A). Two months later,whentaking7.5mgcarvedilolperday,herQTintervalshortenedfurther(QT/QTc,460/460ms)(Figure4B).Aftersixmonths ofreceiving10mgcarvedilolperday,herQTintervalcompletely normalized (QT/QTc,450/442), andT-wave shape was no longer abnormal(Figure4C).Thepatientrecoveredwell,andnosyncope recurrence was observed throughout a 3-year follow-up period.

5. Discussion Figure4.Twelve-leadelectrocardiogramsafterdailyoraladministration ofA)5,B)7.5,andC)10mgcarvedilol. QTcinterval[6].Thephenomenonwasobservedforalgorithmsof This report describes a case of recurrent episodes of syncope that persisted 8 years. When initial 12-lead ECG findings were obtained, the physician did not evaluate data carefully. Instead, the physicianreliedonanautomatedECGdiagnosisthatwasobtained usingECGanalysissoftware.AftertheQTintervalinthepatient’s 12- lead ECG was first misinterpreted via an automated ECG diagnosis, no physicians recommended a repeat ECG examination or reviewed prior findings, despite the recurrence of syncope episodes.Cardiacsyncopeisthesecondmostcommontypeofsyncope[1],andarrhythmias,includingTdP,areitsmostcommoncause [2,3]. Moreover, patients with cardiac syncope have the highest riskofrecurrenceandareatincreasedriskofdeathfromanycause related to syncope [1]. ECG facilitates the diagnosis of cardiac syncope and improves the prognosis of patients. Therefore, to ensure an accurate diagnosis and determine the cause of syncope, multiple ECG examinations and careful analysis of results are crucial[2].AutomatedECGdiagnosisbenefitsphysiciansbecause it reduces the effort needed to evaluate ECG findings and allows physicianstoimmediatelyusefindingsintheclinic.Theautomat- ed ECG diagnoses are widely utilized in medical practices with nonexpert physicians and medical staff [4]. However, automated ECG diagnoses are not always correct. In particular, QT prolongation is underestimated or unreported via diagnostic software [5,6],and measurement differences betweenalgorithms inthe determination of QT interval are larger in patients with congenital [7,8]andacquired[9]longQTintervalthaninhealthysubjects.A previousstudyrevealed that automated ECG diagnosisconcluded thatfindingswere“normal”in42.1%ofpatientswithaprolonged various manufacturers [10]. From these perspectives, automated ECG diagnosis plus over-readingbyphysicians offer the accurate diagnosis,andlongQT syndrome expertsstronglyadvocate manual measurement [5,11,12]. In the present case, apparent and obviousQT prolongationat21 yearsofage(Figure3A) was missed duetoan incorrectautomatedECGdiagnosis. Twaves inleads I, II, III, aVR, aVF, and V2-V6 had prominent notches, and the secondcomponentoftheTwaveshouldhavebeenincludedinthe analysis of the QT interval [11]. The automated ECG diagnosis didnotconsiderthesecondcomponentofthenotchedTwave,and insteadregardedtheendpointofthefirstcomponentofthenotched T wave as the endpoint of the QT interval. If the physician had selfevaluated ECG findings, they would have noticed the mark- edly prolonged QT interval and bimodal T waves, even if the QT intervalwasinaccuratelymeasured.Thepatientwasadministered carvedilol, a non-selective β- and α1-blocker, from the age of 27, whenshewasdiagnosedwithtype2longQTsyndrome.Inaprior report, we showed that carvedilol maybe used to effectively treat type 2 long QT syndrome [14]. It was suggested that α1-adrenergic stimulation acutely reduces Kv 11.1 channel activity via the membranePIP2pathway [15]. Acuteα1-adrenoceptor-mediated reduction in IKr at lower heart rate would act additionally to prolongactionpotentialdurationsandmayenhanceinwardcurrent through Na/Ca exchanger. After the administration of carvedilol, thepatient hadnoepisodesofsyncopeand herQT intervalatrest shortened considerably

6. Conclusions ECG is anessential method that may beused in the examination of patients with recurrent syncope. Moreover, all physicians should confirmECG withtheir owneyes ifthere is QT prolongationand not rely exclusively on automated ECG diagnosis.

References 1. Soteriades ES, Evans JC, Larson MG, Chen MH, Chen L, Benjamin EJ, et al. Incidence and prognosis of syncope. N Engl J Med.2002;347(12):878-885.

2. PrioriSG,WildeAA,HorieM,ChoY,BehrER,BerulC,BlomN,et al. HRS/EHRA/APHRS expert consensusstatement onthediag-nosis and management of patients with inherited primary arrhyth-mia syndromes. J Arrhythm 2014; 30: 1-28.

3. Saklani P, Krahn A, Klein G. Syncope. Circulation. 2013;127(12):1330-9.

4. KatohT,YashimaM,TakahashiN,WatanabeE,IkedaT,Kasama-ki Y, et al. Expert consensus document on automated diagnosis oftheelectrocardiogram:Thetaskforceonautomateddiagnosisoftheele ctrocardiograminJapan.Part1:Nomenclaturefordiagnosisandabnorm al findings. J Arrhythm. 2021; 37: 871-876.

5. IndraratnaP,Tardo D,Delves M,SzirtR, NgB. Measurement andManagementofQTIntervalProlongationforGeneralPhysicians.J Gen Intern Med. 2020; 35(3): 865-873.

6. Garg A, Lehmann MH. Prolonged QT interval diagnosis suppressionbyawidelyusedcomputerizedECGanalysissystem.CircArrhythm Electrophysiol. 2013; 6(1): 76-83.

7. KligfieldP,BadiliniF,DenjoyI,BabaeizadehS,ClarkE,DeBieJ,et al. Comparison of automated interval measurements by widelyused algorithms in digital electrocardiographs. Am Heart J. 2018;200: 110.

8. Miller MD, Porter Cb, Ackerman MJ. Diagnostic accuracy ofscreening electrocardiograms in long QT syndrome I. Pediatrics.2001; 108(1) :8-12.

9. TalebiS,AzhirA,ZuberS,SomanS,ViscoF,Totouom-TanghoH,et al. Underestimated and unreported prolonged QTcbyautomatedECG analysis in patients on methadone: can we rely on computerreading? Acta Cardiol. 2015; 70(2): 211-216.

10. SchläpferJ,WellensHJ.ComputerInterpretedElectrocardiograms:Benefits and Limitations. J Am CollCardiol. 2017; 70(9): 1183-1192.

11. VinkAS,NeumannB,LieveKVV,SinnerMF,HofmanN,ElKadiS,etal. DeterminationandInterpretationoftheQTInterval.Circu-lation. 2018; 138(21): 2345-2358.

12. PostemaPG,WildeAA.ThemeasurementoftheQTinterval.CurrCardi ol Rev. 2014; 10(3): 287-294.

13. PostemaPG,DeJongJS,VanderBiltIA,WildeAA.Accurateelectrocardiographic assessment of the QT interval: teach the tangent.Heart Rhythm. 2008; 5(7): 1015-1018.

14. KimuraH,MizusawaY,ItohH,MiyamotoA,KawamuraM,KawaguchiT,etal.Carvedilol,anon-selectiveβ-withα1-blockerIeffec-tive in long QT syndrome type 2. J Arrhythm 2011; 27: 324-331.

15.ZankovDP,YoshidaH,TsujiK,TsujiK,ToyodaF,DingWG,etal.Adren ergic regulation of the rapid component of delayed rectifierK+ current: Implications for arrhythmogenesis in LQT2 patients.Heart Rhythm 2009; 6(7): 1038-104

Tada. Computerized Misinterpretation of QT Interval in 12-Lead Electrocardiogram and Its Clinical Consequences:ACase of Recurrent Syncope . Annals of Clinical and Medical Case Reports 2022