Engaging
psychiatrists
in
the
diagnosis
of
psychogenic
nonepileptic
seizures.
What
can
they
contribute?
Massimiliano
Beghi
a,
Giuseppe
Erba
b,
Cesare
Maria
Cornaggia
c,
Giorgia
Giussani
d,
Elisa
Bianchi
d,
Gianni
Porro
e,
Michela
Russo
e,
Ettore
Beghi
d,*
aDepartmentofMentalHealth,AUSLRomagna,Ravenna,Italy b
DepartmentofNeurology,SEC,UniversityofRochester,Rochester,NY,UnitedStates
c
SchoolofMedicineandSurgery,UniversityofMilanoBicocca,Italy
d
LaboratoryofNeurologicalDisorders,DepartmentofNeuroscience,IRCCS-IstitutodiRicercheFarmacologiche“MarioNegri”,Milan,Italy
e
RehabilitationCentreCorberi,SanGerardoHospital,Monza,Italy
ARTICLE INFO
Articlehistory: Received20July2017
Receivedinrevisedform12October2017 Accepted13October2017
Keywords:
Psychogenicnonepilepticseizures Video Monitoring Semiology Epilepticseizures Psychiatrists ABSTRACT
Purpose:Toinvestigateifpsychiatristscouldpredictthediagnosisofpsychogenicnonepilepticseizures (PNES)byreviewingvideosofseizuresofvarioustypesandtocomparetheaccuracyandthecriteria leadingtothediagnosisusedbypsychiatristswiththoseusedbyepileptologists.
Methods:Fourboard-certifiedpsychiatristswereaskedtoreview23videoscapturingrepresentative eventsof21unselectedconsecutivepatientsadmittedtoanepilepsycenterforlong-termvideo-EEG monitoring.AllraterswereblindtoEEGandclinicalinformation.Theywererequestedto(1)ratethe videosforqualityandcontent;(2)chooseamongfourdiagnoses:(a)epilepticseizures;(b)PNES;(c) Othernonepilepticseizures(syncope,movementdisorder,migraine,etc.);(d)“CannotSay”;and(3) explainin theirownwordsthemain reasonsleadingto thediagnosis ofchoice.Theresultswere comparedtothoseoffourblindepileptologistswhoindependentlyreviewedthesamecases.The inter-raterreliabilitywastestedwiththeKappastatistic.
Results:Allpsychiatristswereconcordantandcorrectin3/23video-events,comparedto8/23among epileptologists.Despitewidespreaddisagreementamongthemselvesandfrequentfailuresasagroup, individual psychiatrists scored a comparable number of correct diagnoses as did individual epileptologists.Thecommentsprovidedtojustifythediagnosisofchoicedifferedfromneurologists, variedamongraters,andreflectedconsiderableattentiontobodymovementsandbodylanguage. Conclusion:Psychiatrists,asagroup,arelessreliablethanneurologistsindifferentiatingseizuretypeson videobut,asindividuals,canbequiteaccurateinmakingthecorrectdiagnosisbecausetheyaremore attunedtocapturethesubtletiesofhumanbehaviour,ofsubjectiveexperiences,astheeffectsofhidden internalconflictsandcancontributeanewlexiconindefiningPNES.
©2017BritishEpilepsyAssociation.PublishedbyElsevierLtd.Allrightsreserved.
1.Introduction
Psychogenic nonepileptic seizures (PNES) are episodes of paroxysmal impairment of self-controlassociated with a range of motor, sensory and mental manifestations that resemble epilepsy and which represent an experiential or behavioural responsetoemotionalorsocialdistress[1].Fromthepsychiatrist’s pointofview,sincethetransitionfromtheDSM-II[2]tothe DSM-III[3],theterm“neurosis”disappearedfromthenosography,and
eveninthecurrentclassification(DSM5)[4]itisnotpresent;PNES are included in the conversion disorder within the spectrum “Somaticsymptomsandrelateddisorders”.Thus,symptomswere dissembledintheirexpressionanddeprivedoftheir“meaning”in the patient’s history. On the other hand, the International Classification of Diseases (ICD-10) [5] places PNES in the Dissociative Disorders although, unlike the DSM-5, it merges DissociativeDisordersandConversionDisorders.Intothis dimen-sion,PNES,thatwereclassifiedas“hysteria”,nolongerfoundtheir unique positionin the DSM, remaining in a place of nowhere, expelledbothbypsychiatristsandneurologists.Thiscanexplain how,despite a 20% prevalence ofPNES in a tertiarycenter for epilepsy[6],thedelayinthediagnosisofPNEShasbeenestimated tobe7–8years[7,8].
*Correspondingauthorat:LaboratoryofNeurologicalDisorders,IRCCS–Istituto diRicercheFarmacologiche“MarioNegri”,ViaGiuseppeLaMasa,19,20156Milan, Italy.
E-mailaddress:ettore.beghi@marionegri.it(E.Beghi).
https://doi.org/10.1016/j.seizure.2017.10.014
1059-1311/©2017BritishEpilepsyAssociation.PublishedbyElsevierLtd.Allrightsreserved.
ContentslistsavailableatScienceDirect
Seizure
Atpresent,exceptforthecombinationofvideo-EEGmonitoring (VEM), which is reliable only in the ictal phase, and the sophisticated services available only in specialized centers, an instrumentforthediagnosisofPNESislacking.TheInternational LeagueAgainstEpilepsy (ILAE)NonepilepticSeizuresTaskForce publishedrecommendationsindicatingthatincertaincasesitis possibletoreachthediagnosisofPNESonclinicalgroundsinthe absenceofconcomitantVEM[9].Arecentstudycarriedoutbyour group[10]triedtoinvestigateifvisualinformationcontainedin video-recorded events allowed experienced epileptologists to predictthediagnosisofPNESwithouttheaidof electroencepha-lography(EEG)andotherclinicalinformation.Theywereaskedto review23videoscapturingrepresentativeeventsof21unselected consecutivepatientswithamixofepilepticseizures(ES),PNESand physiologicseizures(otherNES).ThefourratersblindtoEEGand clinicalinformationpredictedthediagnosis,confirmedby long-termVEM,in8of23videos(34.8%).Thecorrectdiagnoseswereall ESorPNESpresentingwithclearmotormanifestations.Predictive abilityvariedintheremainingvideos.Interrateragreementwas “moderate”fortheoverallgroup;“moderate”forES;“substantial” forPNES;only“fair”forotherNES.Theseresults,basedexclusively onvideoinformation,weresuperimposabletothoseobtainedina comparable trial conducted totest theinter-rater reliability of epileptologistsinterpretingboth video andEEGdata combined, alsowithoutanyotherpatientinformation[11].
Theaimsofthepresentstudyweretoinvestigateif,how,andto what extent a group of four psychiatrists could predict the diagnosisonpurevisualinformation,reviewingblindlythesame videossubmittedtoepileptologistsinthepreviousstudy[10],and tocomparetheaccuracyandthecriterialeadingtothediagnosisof thepsychiatristsvis-à-viswiththeepileptologists.
Based on the results of previous trials challenging various categories of medical providers in comparison to fully trained epileptologists [12–16], our expectation was that psychiatrists wouldfail,mainlybecauselargelyunfamiliarwiththesemiology of ESand becausethe characteristic features currentlyused to distinguishESfromPNESreflectneurologicalmeasures predomi-nantlyreportedbyepileptologists.
2.Methods
Thisstudyrepresentedanextensionofthefeasibilitytrialofa largerprojectcurrentlyinprogressattheUniversityofRochester (UR). The study protocol was reviewed and approved by the Research Subject Review Board (RSRB) of the UR where the patientswererecruitedandthevideosrecorded.
2.1.Population
Patients18yearsorolderconsecutivelyadmittedbetweenJuly 1andSeptember10,2014,wereaskedtoparticipate.Thepatients’ cohortwasthesameutilizedinthepreviousstudyandthedetails are described elsewhere [10]. For each subject, at the time of discharge, audio-video segments representative of the clinical events were selected and, after removal of the EEG tracing, submittedtotheindependentratersforreview.
2.2.Ratersandprocedure
Unlikethepreviousstudy,thefourraterswereboardcertified psychiatrists,eachwithdifferentpsychiatricbackground,varying degreeofseniority,ofknowledgeaboutepilepsyandexposureto patientswithseizuredisorders(Table1).Eachraterwasblindto theEEGfindings,tothepatient’shistoryandcomorbidities,and unawareofthefinaldiagnosisestablishedbytheclinicalteam.The taskwastoreviewthesamevideossubmittedtoepileptologistsin thepreviousstudy[10]andrenderadiagnosisoutofthefollowing options:
ES,definedaccordingtothe2017ILAEclassification[17]; PNES, classified according to the six categories proposed by
Seneviratneet al. [18]: 1. Rhythmicmotor, 2.Hypermotor, 3. Complexmotor,4.Dialeptic,5.Nonepilepticauras,and6.Mixed; Other nonepileptic seizures (NES), due to paroxysmal non-epileptic events other than psychogenic (syncope or other dysautonomic manifestations, migraine, movement disorder, panicattacks,etc.);
“CannotSay.”
Inaddition,psychiatristshadtospecifythereasonsleadingto thediagnosisofchoiceanddescribeanybehavioralobservations thatmostcontributedtotheirdiagnosticdecision.
Aspreviouslydonebyepileptologists,eachpsychiatristworked independentlyandfiledthedatadirectlyintoadatabasesetupat the IRCCS-Pharmacological Research Institute “Mario Negri” in Milano, Italy, for statistical analysis. We evaluated diagnostic accuracyastheabilityofeachindividualratertocorrectlypredict the“goldstandard”(GS)diagnosis,basedonaudiovisualevidence alone. The GS diagnosis was the result of a comprehensive evaluation ofmultiplefactors.These includedthepatient’s risk factors, comorbidities and psychosocial status; neurological, neuroimaging, interictal EEG findings and the characterization oftheevents(whenrecorded).Thiswasbasedonvideosemiology, ictal EEGfindings(including purelyelectricalseizures), andthe results of monitoring other physiologic parameters such as electrocardiography, blood pressure, orthostatic testing, blood sugar,andsoonasappropriate.InthetwocaseswherebyGSno diagnosiswas possible(NDP), therater’s response“Cannotsay” was considered correct. Raters’ accuracy in predicting the GS diagnosiswaspresentedastheproportionofratersthatcorrectly predictedtheGS.
2.3.Statisticalanalysis
Wecalculatedinterrateragreementamongallraters,between pairs ofraters,andbetweeneachraterandtheGSusing Fleiss’ Kappa[19]with95%confidenceintervals(CIs).TheKappastatistic isameasureofinterrateragreementadjustedbytheamountofthe agreementexpectedtooccurbychancealone.Kappavalueswere usedtoassessoverallagreementacrossalldiagnosticcategories (PNES,ES,OtherNES,CannotSay),andagreementin differentiat-ingbetweenthediagnosisofES,PNES,OtherNES,andCannotSay. Kappavalueswereclassifiedaspoor(<0.00),slight(0.01–0.20),
Table1
Individualprofileofraters.
Rater Yearsinpractice Formaleducationinepilepsy(Yes/No) N.ofpatientswithseizuredisordersseenduringclinicalpractice Degree Specialtytraining
PS1 30 YES Hundreds MD Psychiatry
PS2 12 YES 6/year MD Psychiatry
PS3 30 NO 15/year MD Psychiatry
fair(0.21–0.40),moderate(0.41–0.60),substantial(0.61–0.80),or almostperfect(0.81–1.00)[20].
DatawereanalyzedusingtheSASstatisticalpackage(version 9.2;SASInstituteInc,Cary,NC,U.S.A.).
3.Results
Table2correlatestheclinicalcharacteristicsofthe23events submitted for review with the accuracy of the four blind psychiatrists vs. four blindepileptologists in predicting the GS diagnosisasagroupandasindividualraters.
Allfourpsychiatristswereconcordantandcorrect(4/4)in3/23 video-events, compared to 8/23 when raters were trained epileptologists.Thesuperiorityof theepileptologistsasagroup isalsoapparentwhentheconcordanceinaccuratelypredictingthe diagnosisforeachindividualvideowas<4/4(i.e.:3/4;2/4;1/4;0/ 4). Comparison between the two groups shows 12 points (+) advantageinfavoroftheepileptologistscomparedto4pointsin favorofthepsychiatrists.
Kappavaluesconfirmthediscrepancybetweenthetwogroups. While overall concordance among the four epileptologists was 0.50,itwas0.18amongthefourpsychiatrists,similardifferences arefoundintheKappavaluesbytypeofseizures,varyingfrom0.20 to0.66(epileptologists) and from 0.03 to0.29 (psychiatrists) (Table 3). Likewise, agreement within pairs of epileptologists showedKappavaluesvaryingfrom0.34to0.73(Table4)whereas agreementamongpairsofpsychiatristswasmuchlower,ranging from 0.2to0.37(Table5).Surprisingly,however,agreementof eachindividualraterwiththeGSyieldedsuperimposableKvalues inthetwogroups,rangingfrom0.30to0.56amongepileptologists (Table 4)and from0.01 to 0.45 among psychiatrists (Table 5). Therefore,despitewidespread disagreement among themselves and frequent failures as a group, our results indicate that the individualpsychiatristswhoparticipatedtothisstudywerealmost as accurate as the epileptologists in predicting the correct diagnosisafterreviewingsingleeventsrecordedonvideo.
Successorfailurewasnotcorrelatedtoanyparticulartypeof event (epilepsy, psychogenic,other physiologic or cannot say). However,raters PS3 and PS4chose thediagnosis“Cannot say”
moreoften(7/23)than PS1and PS2(3/23)whilethe epileptol-ogistswerelessvariable(6/23,R1andR3;4/23,R-2;5/23,R-4).
Thecommentsprovidedbythefourpsychiatriststojustifythe diagnosisofchoicevariedconsiderablyinformatanddetailfrom ratertorater,withPS1beingthemostattentiveandarticulate.In essence, liketheepileptologists, psychiatrists paidconsiderable attentiontobodymovements,ostensiblythemostobvioussignsof theevents’semiologyexhibitedonvideo.Startingfromthebody parts involved,theyconsidered head/eyes/mouthdeviationand lateralized limb posturing as clear indicators of ES versus predominantinvolvementof trunk/hips/fingers assuggestiveof PNES.Similarly,theyremarkedonthefollowing:resistancetoeye opening,‘onesingleshake’,suddenonsetorabruptinterruption, ‘on/off’and“highfrequency”movementsasindicatorsofPNES,but alsoemphasised moresubtlebehaviouralaspects suchas‘slow motion’,or“irregular,unpredictable”shaking.Likewise,theyoften opposedtheprogressionofmovementsfromtonictoclonic,from fasttoslow,frompartialtogeneralizedtothemoredisorganized motionsdefinedas“non-epileptic”or“pseudo-myoclonic”.Finally, whiledirectlymentioningspecifictypesofmovementas ‘side-to-side’ or ‘out of synchrony’, instead of pelvic thrusting they preferreddescriptions as ‘arc de circle’or ‘hips more involved’ or‘bodymotionswithsexualconnotations’.
Table2
Accuracyofpsychiatristsvs.epileptologistsinpredictingthegoldstandarddiagnosis.
Video Semiology Goldstandard PS1 PS2 PS3 PS4 Psychiatrists’accuracy Epileptogists’accuracy
1 Motor ES ES PNES ES ES 3/4 3/4
2 Non-motor Other ES Cannotsay PNES ES 0/4 0/4
3a Motor PNES PNES PNES PNES PNES 4/4 4/4
3b Motor PNES PNES PNES ES PNES 3/4 3/4
3c Non-motor PNES PNES Cannotsay Cannotsay Other 1/4+
0/4
4 Motor PNES PNES PNES PNES Cannotsay 3/4 4/4+
5 Motor ES ES ES ES PNES 3/4 3/4
6 Motor PNES Cannotsay PNES Other Cannotsay 1/4 4/4+++
7 Non-motor NDP PNES Other Cannotsay PNES 1/4+
0/4
8 Motor ES ES PNES ES ES 3/4 4/4+
9 Motor PNES ES PNES PNES ES 2/4 3/4+
10 Motor ES ES ES ES PNES 3/4 4/4+
11 Motor ES ES ES ES ES 4/4 4/4
12 Non-motor Other Cannotsay ES Cannotsay Cannotsay 0/4 0/4
13 Motor ES ES ES ES Cannotsay 3/4+
2/4 14 Non-motor ES Other ES Cannotsay Cannotsay 1/4 1/4
15 Motor ES ES ES ES PNES 3/4+ 2/4
16 Non-motor NDP Other ES Cannotsay PNES 1/4 2/4+
17 Non-motor PNES Other Cannotsay Cannotsay PNES 1/4 3/4++
18 Non-motor Other Other Other Cannotsay Cannotsay 2/4 2/4
19 Motor PNES Other PNES PNES ES 2/4 4/4++
20 Non-motor Other Cannotsay ES ES Cannotsay 0/4 0/4
21 Non-motor PNES PNES PNES PNES PNES 4/4 4/4
4+ 12+
PS:psychiatrist;NDP:nodiagnosispossible;ES:epilepticseizure;PNES:psychogenicnon-epilepticseizure. Table3
Agreementamongfourpsychiatristsandamongfourepileptologists. Kappa 95%CI Psychiatrists Overall 0.18 0.08 0.28 PNES 0.21 0.03 0.39 ES 0.29 0.11 0.47 Other 0.03 0.21 0.15 Epileptologists Overall 0.50 0.32 0.68 PNES 0.66 0.54 0.78 ES 0.48 0.36 0.60 Other 0.20 0.08 0.32
Specialattentionwasgiventothepresenceof“automatisms”, distinguishingstereotyped,outofcontext,complexmotoractivity (such as aimlessly looking around), typical of complex partial seizures, from comparable motor manifestations that, at close scrutiny,appearedmorepurposefulordeliberatesuchas‘subject puts herself at the center of attention’ (case #17) or ‘slow movementsofonehandonly’(#3b)orbothhands(#21)andmore “incontext”suchas‘movementsofposturaladjustment’or‘mirror movements’imitatingthe examiner(#3a) or‘partiallyin touch withthecontextbutdistant,asifconfusedorwaitingtogaintime’ (#14).Likewise,certaingestures suchas ‘bringinghands tothe chest’ortothehead‘asifinpain’or‘holdingarmsbythebreasts’ (#3b, #4) were interpretedby psychiatrists as indicative of an inner conflictor suffering. With few exceptions, mostof these observationswerementionedtosupportthediagnosisofPNES.
Finally,psychiatristspointedoutanumberofmotorsystem’s inconsistenciessuchas‘holdingupaseeminglyhypotonicarm’, ‘falling without body hypotonia’ or discrepancies such as the incompatible association between level of consciousness and myoclonicmovements’(#3b),aswellasbehavioralinconsistencies such as the subject’s calm, ‘almost placid’ appearance ‘while holdinghandwiththenearestperson’duringmotorarrest(#21).
Table6showsaseriesofrepresentativecommentstypicallymade by psychiatrists correlating each observation with the video’s number,therater,therater’sdiagnosisandtheGSdiagnosis. 4.Discussion
Bymoststatisticalmeasures,experiencedepileptologistsasa groupweremoreskilfulthanagroupofpracticingpsychiatristsin blindly predicting the GS diagnosis, based exclusively on the physical semiology of a “seizure”. This result was expected
consideringthat,contrarytoepileptologists,psychiatristsseldom have the opportunity to directly scrutinize events on video. However,whereasdegreeofinterrateragreementwithingroups wasquitedifferent,successrateinthetwogroupswasclosewhen the accuracyof individual raters examining a singlevideo was compared. The comments presented to justify thediagnosis of choicemayexplainthisapparentcontradiction.Theyindicatethat psychiatrists,encouragedbytraining,duringtheunfoldingofan event,detectsubtlepsycho-dynamicindicatorsthatcanbeutilized asdiagnostictoolsin additiontopure semiology.Suchsigns or manifestations are implicitly part of the currently accepted definitionofPNESbutcanbeeasilyignoredbyprofessionalswith lesspsychodynamicexperience.
AccordingtotheILAErecommendations[9],thediagnosisof PNESis essentiallybased onthefollowingcriteria: inconsistent semiologywithclinicalmanifestationsthatdonotconformtoa coherentneurologicalschemeasESdo;lackoftherequired neuro-physiological substrate (ictal EEG discharges); evidence of risk factorsthatmayleadtothe“episodicimpairmentofself-control” as defined by Reuber [1]. The ILAE definition, likethe DSM-III definition,failstounraveltheunderlyingmechanismsofPNESalso describedas“experientialorbehaviouralresponsestoemotional or social distress” [1]. The experiment reported here was an attempttodeterminehowdifferentandcomplementarywouldbe theobservationsoffullyboardedpsychiatristscomparedtothose ofexperiencedepileptologists.Theresultshavebeensomewhat perplexingbutencouraging.Despitelimitedtrainingand unfamil-iarity withthetypeof videomaterial submitted,theindividual psychiatrists in our panel, blind to patient’s history and EEG findings,provedtobeveryclosetoexperiencedepileptologistsin predictingtheGSdiagnosis.However,theywereclearlyinferior andingreaterdisagreementwitheachotherwhenchallengedasa
Table4
Agreementwithinpairsofepileptologistsandbetweeneachepileptologistandgoldstandard.
Overall PNES ES Other
Pair Kappa 95%CI Kappa 95%CI Kappa 95%CI Kappa 95%CI
R1vs.R2 0.73 0.44–1.00 0.82 0.41–1.00 0.63 0.22–1.00 –a – R1vs.R3 0.48 0.21–0.75 0.56 0.15–0.97 0.40 0.01–0.81 0.05 0.46–0.36 R1vs.R4 0.54 0.27–0.81 0.82 0.41–1.00 0.62 0.21–1.00 0.02 0.43–0.39 R2vs.R3 0.34 0.07–0.61 0.56 0.15–0.97 0.25 0.16–0.66 0.05 0.46–0.36 R2vs.R4 0.40 0.11–0.69 0.63 0.22–1.00 0.45 0.04–0.86 0.02 0.43–0.39 R3vs.R4 0.48 0.23–0.73 0.56 0.15–0.97 0.40 0.01–0.81 0.64 0.23–1.00 R1vs.GS 0.49 0.24–0.74 0.82 0.41–1.00 0.81 0.40–1.00 0.10 0.51–0.31 R2vs.GS 0.35 0.08–0.62 0.63 0.22–1.00 0.45 0.04–0.86 0.10 0.51–0.31 R3vs.GS 0.30 0.05–0.55 0.39 0.02–0.80 0.40 0.01–0.81 0.23 0.18–0.64 R4vs.GS 0.56 0.31–0.81 0.63 0.22–1.00 0.81 0.40–1.00 0.32 0.09–0.73 ES:epilepticseizure;PNES:psychogenicnon-epilepticseizure:R:rater;GS:goldstandard;CI:confidenceinterval.
a
Noreviewersgivetheresponse“Other”.
Table5
Agreementwithinpairsofpsychiatristsandbetweeneachpsychiatristandgoldstandard.
Overall PNES ES Other
Pair Kappa 95%CI Kappa 95%CI Kappa 95%CI Kappa 95%CI
group.Itispossiblethatsuchdiscrepancyininterrateragreement asagroupandasindividualpairs(highforepileptologists,lowfor psychiatrists)reflectsthedifferentapproachintheinterpretation ofthevideomaterialadoptedbythetwogroups.Epileptologists, bytraining, tend tostrictlyadhere to pre-setcriteria based on semiological features validated in published material. Such disciplinedapproachconfersconsiderableuniformitytotheraters asindividualsandasagroups.Conversely,psychiatrists,though payingdueattentiontothesameindicators,arenotexclusively boundtoevidence-basedcriteria,displayinggreatersensitivityto nuances and to the significance of subtle behavioral features.
Table6showsexamplesofhowpsychiatristscanreadintobody language and interpret subtle behavioral manifestations or subjectiveexperiencesastheeffectsofhiddeninternalconflicts. Thiswillingnesstoexplorebeyondthemerefactsandtocapture crypticsignalsotherwise ignoredinamoreorthodox approach, probablyexplainsthesuccessofthepsychiatricratersinpredicting thecorrectdiagnosiswhenconsideringaspecificcase.
Therearenodatain theliteraturethat provethediagnostic validity and reliability of this approach. Our preliminary data indicatethat out of 14videos where atleast one of theraters includedpsychodynamicobservationsinhiscommentstojustify thediagnosticchoice,thediagnosiswascorrectin9(7PNES;2ES) andincorrectintheremaining5 videos(4ESand 1NDP).This suggeststhatpsychiatristsaremoresuccessfulindiagnosingPNES thanothertypesofseizures.Thoughthesamplewas small,our resultsclearlyindicatesthatthepsychodynamicinterpretationof certainbehavioralsignscanbehelpfulindiagnosingPNESbutcan alsobemisleading,especiallyindifferentiatingnon-motorPNES from complex partial seizures. This represents a diagnostic dilemmaalsoforepileptologists(10).Otherpossibleexplanations forthedifferencesbetweenthetwogroupscanbefoundwhen comparingthediagnosticattitudesofpsychiatristsand neurolo-gists.PsychiatristslostthehabitofdiagnosingPNES,aspatients withthis clinical condition directlygoto theneurologist, who makes the first differential diagnosis. On the other hand, the
neurologistdealswiththe“physical”body,thatcorrespondstothe “homunculus”(sensoryand/ormotor)intheCNS,while psychia-tristsdealwiththesymbolicbody [21],thatcorrespondstothe representative language of an original “traumatic” event and whosesemiologydoesnotcorrespondtoany“homunculus”ofthe CNS.The expertisethat thepsychiatristcanputin thefield,in addition tothe epileptologist, consistsin reading the symbolic bodylanguage,forinstancetheostensible“indifference”,originally describedbyCharcot[22].ContrarytopatientswithES,patients withPNESremainpartiallyintouchwiththecontextduringthe event.Thus,theyhesitate,asiftheywantedtogaintime,andtend toputthemselvesatthecenterofattention.Inthisperspective, psychiatristsconsiderPNESauniqueentityonitsown,“something thatis”ratherthan“somethingthatisnot”.
Indeed,psychiatristsseemtohavepoorskillsinnon-PNESand non-ESattacks,wheresymbolicbodylanguageisvirtuallyabsent, while they addelementsto thediagnosisofPNESwherethesymbolic bodylanguageispresent[23].Conversely,epileptologistsseemto havegoodskillsinmotorES,wheretopodiagnosisappearsmore straightforwardandsimple,whiletheyappeartohavelessskillin seizures wherethe relationshipbetween seizure semiologyand “homunculus”is lessobvious.Forthiskindofseizures,asimplevideo documentation seems unsatisfactory. This is true for both the epileptologist and the psychiatrist, because of the imbrication betweenthesomaticand thesymbolicbody.In theseinstances, video-EEGorlinguisticanalysis[24–26]assumeanessentialrole.
Inrecentyears,psychiatristshavebeendeterredfromdirectly participatinginthediagnosticprocessofidentifyingPNESbyseveral factors.First,thereplacementintheDSMIIIoftheterm“hysteria” withaphenomenologicalclassificationofsymptomsand manifes-tationsdevoidfrometiologicalcontent.That,initself,hascurbed their interest in the condition. Second, the realization that diagnosingPNEShas become, bydefault,a responsibilityof the epileptologistswho,bynecessity,workwithphysicalevidenceand physiological parameters. Third, there is a widespread trendin medicineofrelyingprimarilyonevidence-baseddataignoringany
Table6
Originalobservationsbypsychiatrists.
Video# Originalobservations Diagnosisofchoice DiagnosisGS 3a Apparentlossofcontactafterprolongedphoticstimulation(PS1) PNES PNES-Motor
Purposefulposturaladjustmentsduringeyesopeningandclosing(PS1) PNES PNES-Motor Absenceofagitatedbehaviour+calmbreathing(PS3) PNES PNES-Motor Eventinducedbystress(prolongedphoticstimulation)(PS4) PNES PNES-Motor 3b Subtlyregainscontactbringinghandstochest(PS1) PNES PNES-Motor Apparentthoracicpain+slowhandmovements(PS4) PNES PNES-Motor 3c Bilateralsensorymisperceptions(PS1) PNES PNES-Nonmotor
Mantainscontactduringpsychomotorslowing(PS1) PNES PNES-Nonmotor Tendencytodisengagefromcontext(PS1) PNES PNES-Nonmotor 4 Nopelvictrustingbutmovementswithsexualconnotation(PS1) PNES PNES-Motor
Apparentconfusionduringperceptionofpain(PS1) PNES PNES-Motor "Non-epileptic"movementsofarmsandlegs(PS3) PNES PNES-Motor Indifferenttowhathashappened(PS3) PNES PNES-Motor Alertbehaviourduringseizure(PS4) PNES PNES-Motor 7 Looksastonished,slowmotionsanddiffusemalaise(PS1) PNES NDP-Nonmotor
"Dissociationsymptoms"(feelslikeshakingevenifnotapparent)(PS2) Other NDP-Nonmotor
Emotionalbehavior(PS2) Other NDP-Nonmotor
8 Twilightstate,withpartialdetachment(PS1) EPILEPSY EPILEPSY-Motor
9 Arcdecircle(PS2) PNES PNES-Motor
Questionableimpairmentofconsciousness/seizureonlyinpresenceofwitness(PS3) PNES PNES-Motor
10 Indifference(PS4) PNES EPILEPSY-Motor
13 "Morpheic"event(PS2) EPILEPSY EPILEPSY-Motor
14 Appearspartiallyintouchwithcontextbutdistant,asifconfusedorwantingtogaintime(PS1) Other EPILEPSY-Nonmotor
15 Indifferentattitude(PS4) PNES EPILEPSY-Motor
17 Duringtheevent,subjectputsherselfatthecentreofattention(PS1) Other PNES-Nonmotor
19 Seizureonlywhenpeoplepresent(PS3) PNES PNES-Motor
intuitiveapproachbecauselessreliable.Asaresult,wehavebeen payinggreatattentiontoexternalmanifestationandlesstowhatthe patienthastoexpressorcommunicatesubliminally.Afirstsignof renewed interest in the hidden signals contained in patients’ behaviorhasbeena seriesof publications onthedifferencesin linguisticexpressionbetweensubjectswithEScomparedtoPNES
[24–26].
Weundertookthisexperimenttoforcepsychiatristsbackinthe diagnosticfieldandexploretheobservationstheyhadtooffer.We decidedtochallengethem withmaterialostensiblymore suitablefor epileptologistsforthesakeofcomparingthetwogroups.Moretrials willbenecessaryusingmoreappropriatematerialsuchasrecorded patient’s interviews, rather than, or in addition to, the events recordedonvideo.Thoughonlytentative,someoftheobservations reportedhereshouldmakeusreflectontheopportunitiesweare missing.Psychiatrists,bytraining,aremoreattunedtocapturethe subtletiesofhumanbehaviourthanneurologistsandcancontribute anewlexiconindefiningPNES.Thus,theycanplayanimportant complementaryrolenotonlyinestablishingthediagnosisbutalso, byofferingaglimpseontothepossiblepathophysiological mecha-nismsofthisdisorder,pavingthewaytoeffectivetreatment.Thus, theissueisworthpursuingfurtherbycontinuingthedialogueand fostering more active collaboration between epileptologistsand psychiatristsinthemanagementofpatientswithPNES.Itiswell known fromtheepilepsy literaturethat,so far,no singleindicator has provedpathognomonicforthediagnosisofeitherESorPNES.Rather, aconstellationofsignsorsymptomsmaybemoreindicative[12].It appearsthatassessingthediagnosticweightofanysinglefeature mentionedbythepsychiatristsmaybeequallyproblematic.Our resultsindicatethatcertainobservationscanbemisleadingevenfor experiencedpsychiatrists.Signsmustbeinterpretedcarefully,in context,andgainsignificanceifsupportedbyadditionalevidence.In this respect,the convergence of multidisciplinary observations made byepileptologistsincollaboration withpsychiatristsshouldbeideal. Giventhehighdegreeofdiscordanceamong psychiatristsin ourstudy,itseemsunlikelythat,inthecurrentstateofmedical practice,theymaycontributemuch todiagnosticaccuracyin a populationthatisalreadydifficulttodiagnose.However,greater familiarity with seizure-like events may enable them to add valuable diagnostic observations for screening patients with nonepilepticevents.Conversely, afterappropriateexposure and cross training, epileptologists may learn how to read into the symbolicmeaningsofbodylanguageandfurtherrefinetheirskills inthedifferentialdiagnosis.
Thisstudyhaslimitations.First,thenumberofraterswhotook partinthestudyisfairlysmall.This,initself,canaffecttheresults. Second, wetried toinvolve individuals withvarying degreeof seniorityand diverse knowledge and experienceabout seizure disorders, who were representative of practicing psychiatrists. Thus, the participating psychiatrists may not reflect the back-groundandexperienceofallpsychiatristsin clinicalpracticein Italyand,perhapsevenmoreimportant,inothercountries.Most importantly,ourfindingthatindividualpsychiatrists maybeas accurate as individual epileptologists even without specific trainingmustbeinterpretedwithcaution,keepinginmindthat thelevelofexpertizeofthesingleparticipatingratersiscrutial. Clearly,theadditionofonebadrater,orofoneexcellentrater,to eithersidecouldsubstantiallychangetheaccuracyratiobetween thetwogroups. However,ourresults,though farfromdefinite, provideinsightinafairlyunexploredfieldandcanbeusedasthe backgroundtostimulatenewresearch.
Conflictofinterest
DrBeghiservesontheeditorialboardsofAmyotrophicLateral Sclerosis, Clinical Neurology & Neurosurgery, and
Neuroepidemiology;hasbeenanassociateeditorofEpilepsia;has received money forboard membershipfromVIROPHARMA and EISAI;hasreceivedfundingfortravelandspeakerhonorariafrom UCB-Pharma, Sanofi-Aventis, GSK; has received funding for educational presentations from GSK; reports grants from the ItalianDrugAgencyandfromtheItalianMinistryofHealth.
Theremainingauthorshavenothingtodisclose. References
[1]ReuberM.Psychogenicnonepilepticseizures:answersandquestions.Epilepsy Behav2008;12(4):622–35.
[2]AmericanPychiatricAssociation.DiagnosticandStatisticalManualofMental Disorders(DSM-II).secondeditionAmericanPychiatricAssociation;1968. [3]AmericanPychiatricAssociation.DiagnosticandStatisticalManualofMental
Disorders(DSM-III).thirdeditionAmericanPychiatricAssociation;1980. [4]AmericanPychiatricAssociation.DiagnosticandStatisticalManualofMental
Disorders(DSM-5).fiftheditionWhashington,DCLondon,England:American PychiatricAssociation;2013.
[5]WorldHealthOrganization.InternationalStatisticalClassificationofDiseases andRelated HealthProblems(ICD-10).10threvision. fifth editionWorld HealthOrganization;2016.
[6]LesserRP.Psychogenicseizure.Neurology1996;46:1499–507.
[7]BoddeNM,BrooksJL,BakerGA,BoonPA,HendriksenJG,MulderOG,etal. Psychogenic non-epileptic seizures–definition, etiology, treatment and prognosticissues:acriticalreview.Seizure2009;18(8):543–53.
[8]KerrWT,JanioEA,LeJM,HoriJM,PatelAB,GallardoNL,etal.Diagnosticdelay inpsychogenicseizuresand theassociationwithanti-seizuremedication trials.Seizure2016;40:123–6.
[9]LaFranceJr.WC,BakerGA,DuncanR,GoldsteinLH,ReuberM.Minimum requirementsforthediagnosisofpsychogenicnonepilepticseizures:astaged approach: areportfromthe InternationalLeagueAgainst Epilepsy Non-epilepticSeizuresTaskForce.Epilepsia2013;54(11):2005–18.
[10]ErbaG,GiussaniG,JuersivichA,MagauddaA,ChiesaV,LaganàA,etal.The semiologyofpsychogenicnonepilepticseizuresrevisited:canvideoalone predictthediagnosis?Preliminarydatafromaprospectivefeasibilitystudy. Epilepsia2016;57(5):777–85.
[11]BenbadisSR,LaFranceJr.WC,PapandonatosGD,KorabathinaK,LinK,Kraemer HC.NEStreatmentworkshop:interraterreliabilityofEEG-videomonitoring. Neurology2009;73(11):843–6.
[12]SyedTU,LaFranceJr.WC,KahrimanES,HasanSN,RajasekaranV,GulatiD,etal. Can semiology predictpsychogenic nonepileptic seizures?A prospective study.AnnNeurol2011;69(6):997–1004.
[13]SamuelM,DuncanJS.Useofhand-heldvideocamcorderintheevaluationof seizures.JNeurolNeurosurgPsychiatry1994;57:1417–8.
[14]Ristic AJ, Draškovic M,Bukumiric Z,SokicD. Reliability of thewitness descriptionsofepilepticseizuresandpsychogenicnon-epilepticattacks:a comparativeanalysis.NeurolRes2015;37(6):560–2.
[15]SeneviratneU,DingC,BowerS,CraigS,LeechM,PhanTG.Video-based training improvestheaccuracy ofseizure diagnosis. JNeurol Neurosurg Psychiatry2014;85(4):466–70.
[16]MacDonaldT,HillA,PhanT,FitzgeraldP,SeneviratneU.Psychiatryversus generalphysicians:whoisbetteratdifferentiatingepilepticfrompsychogenic non-epilepticseizures?AustralasPsychiatry2012;20(5):379–83.
[17]Fisher RS,Cross JH, FrenchJA, Higurashi N,Hirsch E, Jansen FE, etal. Operationalclassificationofseizuretypesbytheinternationalleagueagainst epilepsy: position paper of the ILAE commission for classification and terminology.Epilepsia2017;58(4):522–30.
[18]SeneviratneU,ReutensD,D'SouzaW.Stereotypyofpsychogenicnonepileptic seizures:insightsfromvideo-EEGmonitoring.Epilepsia2010;51(7):1159–68. [19]FleissJ.Measuringnominalscaleagreementamongmanyraters.PsycholBull
1971;76:378–82.
[20]LandisJR,KochGG.Anapplicationofhierarchicalkappa-typestatisticsinthe assessment ofmajorityagreement among multipleobservers.Biometrics 1977;33:363–74.
[21]FoucaultM.PsychiatricPower:LecturesattheCollègedeFrance,1973–1974. TranslationbyGrahamBurchell.NewYork:Picador;2006.
[22]CharcotJM.LeçonsdumardiàlaSalpetriere.Policliniques1887–1889.Tuesday lessonsattheSalpetriere.Paris:BureauxduProgrèsMedical;1889. [23]CornaggiaCM,BeghiM,BeghiE.Whytalkingaboutpsychiatriccomorbidityin
PNES?EpilepsyBehav2017;75:276.
[24]Cornaggia CM, Gugliotta SC, Magaudda A, Alfa R, Beghi M, Polita M. ConversationanalysisinthedifferentialdiagnosisofItalianpatientswith epilepticorpsychogenicnon-epilepticseizures:ablindprospectivestudy. EpilepsyBehav2012;25(4):598–604.
[25]ReuberM,MonzoniC,SharrackB,PlugL.Usinginteractionalandlinguistic analysis to distinguish between epileptic and psychogenic nonepileptic seizures:aprospective,blindedmultiraterstudy.EpilepsyBehav2009;16 (1):139–44.
