Health related quality of life of people with non-epileptic seizures: The role of socio-demographic characteristics and stigma

(1)

Health related quality of life of people with non-epileptic seizures:

The role of socio-demographic characteristics and stigma

Catherine Robson

^a,

*, Lorna Myers

^b

, Chrisma Pretorius

^c

, Olaug S. Lian

^a

, Markus Reuber

^d

aDepartmentofCommunityMedicine,FacultyofHealthSciences,UniversityofTromsø–TheArcticUniversityofNorway,N-9037Tromsø,Norway

bTheNortheastRegionalEpilepsyGroup,820SecondAvenue,Suite6C,NewYork,NY10017,UnitedStates

cDepartmentofPsychology,StellenboschUniversity,Stellenbosch7600,SouthAfrica

dAcademicNeurologyUnit,UniversityofShefﬁeld,RoyalHallamshireHospital,GlossopRoad,ShefﬁeldS102JF,UnitedKingdom

ARTICLE INFO Articlehistory:

Received22September2017

Receivedinrevisedform8December2017 Accepted1January2018

Keywords:

Psychogenicnon-epilepticseizures Non-epilepticattackdisorder Qualityoflife

Socio-demographicfactors Stigma

ABSTRACT

Purpose:Peoplewithnon-epilepticseizures(NES)consistentlyreportpoorerHealth-RelatedQualityof Life(HRQoL)thanpeoplewithepilepsy.Yet,unlikeinepilepsy,knowledgeofhowsocialfactorsinﬂuence theHRQoLofadultswithNESislimited.Toaddtotheevidencebase,thisstudyexplorestherelationship betweenHRQoL andperceived stigmaamongadultswithNES, andtherole ofsocio-demographic characteristics.

Methods:Datawasgatheredfromasurveyof115peoplelivingwiththecondition,recruitedfromonline supportgroups.Participantsprovidedsocio-demographicandhealth-relateddataandcompletedaseries ofquestionsinvestigatingtheirHRQoL(QOLIE-31)andstigmaperceptions(10-itemEpilepsyStigma Scale).

Results:Participantswerefoundtoexperiencehighlevelsofperceivedstigma(median5.2,mean4.9).A signiﬁcantand moderateinverse correlationwas observedbetweenHRQoL andstigma(rs 0.474, p=<0.001);suggestinghigherperceptionsofstigmacontributetopoorerHRQoLamongadultswith NES.Stigmaperceptionswerefoundtobemoststronglyassociatedwiththeseizureworry(rs= 0.479), emotionalwellbeing(rs= 0.421),andsocialfunctioning(rs=0.407)HRQoLdomains.Participantswho reportedbeinginemploymentoreducationwerefoundtohavesigniﬁcantlybetterHRQoLthanthose whowerenot(p=<0.001).

Conclusion:More(qualitativeandquantitative)researchisjustiﬁedtounderstandhow–andwhy–those with the conditionexperience stigmatisation, and thefactors thatimpede and helpfacilitate the participationofpeoplewithNESineducationandemployment.

1.Introduction

Therehasbeenamarkedshiftinthinkingaboutwhathealthis and how it is measured; with traditional clinical outcomes increasinglygivingwayto,orusedin conjunctionwith,patient reportedoutcomemeasures(PROMs)[1].HealthrelatedQualityof Life(HRQoL),isamultidimensionalPROMconstructusedtoassess theperceivedimpactofhealthstatusonqualityoflife;comprised of physicalfunctioning,emotional status, and social well-being domains[2].

Peoplewithnon-epilepticseizures(NES),oftenreferredtoas psychogenicnon-epilepticseizures(PNES)ornon-epilepticattack

disorder (NEAD), consistently report poorer HRQoL than those withepilepsy[3–5].Arecentsystematicreviewoftheliterature identiﬁed14studiesarisingfromtenseparate researchprojects (data collections) that have explored associations between independentfactorsandHRQoLinthispatientgroup[6].

The evidence available suggests a strong adverseassociation betweenpsychologicalfactorsandtheHRQoLofadultswithNES.

Severalstudiesshowdepressiontobeastrongpredictorofpoorer HRQoLinthispatientgroup[3–5,7–15].Otherpsychologicalfactors associated with poorer HRQoL in people with NES include the number/severity of mood and emotional complaints [3,9,14,15], illness perceptions [16], dissociative experiences [8,11], somatic symptoms[9,10,15],andescape-avoidancecopingstrategies[8,17].

Condition-related factors,suchas olderage ofonset [15,18] and experiencingtheconditionforashorterperiodoftime[15]havealso beenshowntoadverselyaffectHRQoL.Aswithepilepsypatient groups [19], seizure freedom has been shown to be positively

* Correspondingauthor.

E-mailaddresses:[email protected](C.Robson),

[email protected](L.Myers),[email protected](C.Pretorius), [email protected](O.S. Lian),m.reuber@shefﬁeld.ac.uk(M.Reuber).

https://doi.org/10.1016/j.seizure.2018.01.001

creativecommons.org/licenses/by/4.0/).

ContentslistsavailableatScienceDirect

Seizure

j o u r n a lh o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / y s e i z

(2)

associatedwithHRQoLinpatientswithNES[20].However,whereas systematic reviewsofthe literature have found seizure frequency to a (modest)predictorofHRQoLinadultswithepilepsy[19],thesame wasnotfoundtobetrueforadultswithNES[6].

Yet, as Mitchell and colleagues point out [11], studies that attempt to produce a model to explain the factors that are associatedwithHRQoLinadultswithNESonlyaccountfor65%of thevarianceatbest[3].Ourlimitedunderstandingofhowsocial factors affect the HRQoL of those living with NES probably contributestothisshortfall.Therearesigniﬁcantknowledgegaps inrelationtodomains suchasstigma,employmentstatus, and socialandfamilyrelations[6].

HRQoL in this patient grouphas beennegatively associated withfamilyrolesandaffectivefamilyinvolvementsubscalesusing theFamilyAssessmentDevice(FAD)[5],suggestingtherolesand influence of significant others to be a potentially important predictor of HRQoL for people with NES. There is also some evidence that concerns about relationships with the main caregiverseem tocause more distress in those with NESthan patientswithepilepsy[21].Weknowthatthestigmaassociated withepilepsyisconsiderableandthatithasnegativeeffects on HRQoL[22]–infact,itmayaccountformoreHRQoLvariancethan clinicaloutcomes(such asseizurefrequencyandside-effectsof antiepilepticdrugs)[23].However,whereasthereisawealthof researchtosupporttheviewthatthesocialprognosisofepilepsyis often less good than the clinicalone [22], comparativelylittle researchhasexploredthesocialimpactofNES[24],andnonehas explored the relationship between stigma and HRQoL in this patientgrouptodate.Theonlystudytohaveexaminedtheroleof socio-demographic variables found no significant correlation between employment status, marital status, having children, religiousinvolvement,andproximitytofamilyandHRQoL[15],but moreresearchisneededtosubstantiatethesefindings.

To addtotheevidencebase,thisstudyseeks toexplorethe relationshipbetweenHRQoLandperceivedstigmaamongadults with NES, and the role of participants’ socio-demographic characteristics. Findings (‘statistical pointers’) will inform an upcomingqualitativestudyexploring thestigmaperceptions of peoplewiththecondition,which willinclude exploring participants writtentexts abouttheirfamily relations and the social impactofNES.Takentogether,wehopetoidentifysocialdynamics thatwillcontributetolarger(multipleregression)studiesaiming toproduceamodeltoexplainfactorsaffectingtheHRQoLofadults withNES.

2.Methods

Alinktoanin-depth(86-item,233-question)surveycomprised of polar, frequency, Likert scales, and open questions was advertisedtomembersof20patientandpractitioner-ledonline supportgroupsandwebsitesforpeoplewithNES(basedintheUK and US; not disclosed for reasons of conﬁdentiality, details availableonrequest). Thesurveywas pilotedamong 25people livingwiththecondition.Finalsurveydatawereorganizedaround four key themes: 1) the diagnostic journey 2) access to and experienceof treatment 3)interactions withhealthcareprofes- sionals and 4) social support and social stigma. Advertising commencedMay2016and ﬁnaldatacollectedfrom1Julyto1 October2016.

To include as many people with NES as possible, the only inclusioncriteriawerethatparticipantshadtobeover18yearsof ageandhadreceivedadiagnosisofNESbyahealthprofessional.

ParticipantswereadvisedthatweusedthetermNESthroughout the survey to describe diagnoses of psychogenic non-epileptic seizures(PNES),non-epilepticattackdisorder(NEAD),andother diagnosticterms sometimesusedtodescribetheconditionand

symptoms; such as, dissociative, conversion, functional, and pseudoseizures.Theywerealsoinformedthatweusedtheterm seizure throughout the survey, whilst recognising that some people experience non-epileptic events in which they do not exhibitmovements,onlybrieﬂyloseconsciousness,orexperience analteredstateofconsciousness,oramixtureofthesebehaviours andsensations.Participantswereadvisedthat,unlessotherwise stated,toconsider theterm“seizure” toinclude such“events”. Thosewithadual-diagnosisofepilepsyandNESwereaskedtoonly commentonnon-epilepticseizuresandeventswhereverpossible.

Participants wereable tosave theiranswers and return tothe surveyviaasecureandautomatedemaillink.Typically,open(free- text)questionswereoptionalandallothersmandatory.Thesmart- logicsurveyformathelpedtoprotectagainstparticipantsgiving conﬂictinganswers,andto‘re-check’andcorrectresponseswhen theydidso.

This study uses a subset of thefull survey data to explore associations betweenthesocio-demographicand health-related characteristicsofparticipants,theirHRQoL,andlevelsofperceived stigmatisation;usingthemeasureslistedbelow.

2.1.Measures

Participants were asked a range of socio-demographic and health-relatedquestions,asindicatedinTable1.

The31-itemQualityofLifeinEpilepsyinventory(QOLIE-31)[ 25]wasusedtomeasureHRQoL.Theinventory,designedforadults with epilepsy aged 18 years and older, is divided into seven subscales that explore various aspects of patients’ health and wellbeing: emotional well-being, social functioning, energy/

fatigue,cognitivefunctioning,seizureworry,medicationeffects, and overall quality of life (a single-item subscale). A weighted averageof themulti-itemscale scoresisused toobtainatotal score. Although speciﬁcally designed for people withepilepsy, there are important clinical similarities and shared concerns betweenNESandepilepsypatientpopulations.Areviewofhealth status measuresdidnot produceany bettertoolstoassess the constructofHRQoLinthispatientpopulation[26];andarecent systematic reviewidentiﬁedtheQOLIE-31 asthe mostpopular measureinstudiesexploringtheHRQoLofthispatientgroup[6].

StigmawasmeasuredusingtheEpilepsyStigmaScaledevel- opedbyDilorioandcolleagues[27].Theten-itemscaleassesses thedegreetowhichapersonbelievesthattheirseizurecondition is perceived as negativeand interferes with relationshipswith others, rated on a 7-point scale from strongly disagree (1) to stronglyagree (7). Itemresponsesare summed toyielda total score.Inthisstudy,overallmedianscores(1–7)werecalculated.

Higherscoresareassociatedwithgreaterperceptionsofstigma.To our knowledge, the measure has not been validated in a NES patientpopulation.Weassessedthescaleforinternalconsistency and found

a

^coefﬁcient for the responses of our (n=115) participantstobe0.89.

2.2.Statisticalanalysis

AnalysisofthedatawasperformedusingSPSS,version24.To guard against assumptions of normality and homogeneity of variance, and because measures include ordinal data, non- parametric tests of signiﬁcance and correlation were used. In somecases,meanscoresarepresentedordiscussedforcompara- tive purposes. The primary outcome measure was QOLIE-31 (weighted) total score. The Mann-WhitneyU Test was used to comparequantitativevariablesbetweentwoindependentgroups.

Spearman'srankcorrelationcoefﬁcient(r_s)wasusedtocompare continuousandordinaldatavariables.Thestrengthofcorrelations were deﬁned as: 0–0.39 weak, 0.4–0.69 moderate, and 0.7–1

(3)

strong.The coefficient of determination (rs2) was calculated to establish the proportion of shared variance between HRQoL domains and total stigma score. Holm’s Sequential Bonferroni Procedure[28]wasperformedformultipleteststoprotectagainst inflationofType1error.Statisticallysignificantresults(p<0.05) notrejectedfollowingtheHolm’sSequentialBonferronimethod areshowninbold.

3.Results 3.1.Participants

289peoplebeganthesurvey.Ofthese,141(49%)completedall mandatoryquestionsandsubmittedtheirresponsesforinclusion in the study. Six people reported a diagnosis other than NES (functional movement disorder) and their responses were excludedfromfurtheranalysis.Oftheremaining135participants, wereporton115participantswhodescribedreceiving“aformal (highlylikelyorcertain)diagnosisofNES”byahealthprofessional.

20participantswhoreportedreceiving“a tentative(possibleor likely)diagnosisof NES”(or whoindicated theywereawaiting furthertests)wereexcludedfromfurtheranalysis.

Thesocio-demographicand health characteristicsof the115 participantsincludedinthisstudyareshowninTable1.

3.2.Health-relatedqualityoflife

Asagroup,participantsdemonstratedatotalmedian(weighted) QOLIE-31 score of 31.7 (mean 33.8, 95%CI=31.0–36.7). No signiﬁcant differences in HRQoL (QOLIE-31 total scores) were observed betweenthose who self-reporteda dualdiagnosis of epilepsyandNES(median31.1,range13.7–63.5),andthosewith

NES alone (median 31.7, range 3.4–87.8) (p=0.800); nor were significant differences in individual HRQoL domain scores ob- servedbetweenthetwogroups.Nosignificantassociationswere foundbetweenHRQoLtotal scoresand time fromonset ofNES (rs= 0.111,p=0.239)ortimefromonsettodiagnosis(rs=0.066, p=0.486). No significant differences in HRQoL were observed betweenthosewhohadbeenerroneouslydiagnosedwithepilepsy inthepastandthosewhohadnot(p=0.502).Seizurefrequency wasshowntobesignificantly,butweaklycorrelatedwithHRQoL (rs= 0.382,p=<0.001).HRQoLwasnotsignificantlycorrelated withparticipants’age(rs=0.062,p=0.512).FollowingtheHolm– Bonferroni method, participantsin work or educationreported significantlybetterHRQoLthanthosewhowerenot.Asshownin Table2,noothersocio-demographicvariablestestedreturneda significantresult.

3.3.Stigma

Themedianstigmascoreacrossthewholegroupofparticipants was5.2(mean4.9,95%CI=4.7–5.2).Nosignificantdifferencesin perceived stigma (total stigmascores) wereobserved between those who self-reporteda dual diagnosis of epilepsyand NES (median5.2,range2.7–7),andthosewithNESalone(median5.2, range 1–7) (p=0.718). No significant associations were found betweentotalstigmascoresandtimefromonsetofNES(r_s=0.070, p=0.456)ortimefromonsettodiagnosis(rs=0.049,p=0.604).No significantdifferences in stigmascoreswereobservedbetween thosewhohadbeenerroneouslydiagnosedwithepilepsyinthe past and those whohad not (p=0.561).Seizure frequencywas showntobesignificantly,but weaklycorrelated withperceived stigma scores (rs= 0.252, p=0.007). After Holm-Bonferroni correction,nosignificantdifferenceswereobservedbetweenthe

Table1

Socio-demographicandhealthcharacteristics.

Numberormedian Proportionorrange

Country UK(63)andIreland(3) 66 57%

US(42)andCanada(2) 44 38%

Restoftheworld(AustraliaandNorway) 5 4%

Age 37years 18–75years

Gender Female 102 89%

Male 11 10%

Transgender 2 2%

Relationshipstatus Married/CivilUnion(54)orpartnered(17) 71 62%

Single(34)orseparatedordivorced(10) 44 38%

Livingarrangements Livingalone 13 11%

Livingwithothers 102 89%

Employmentstatus Infull-time(14)orpart-timeemployment

(14)oreducation(9)

37 32%

Unabletowork(72)orhome-maker(3) 75 65%

Retired 3 3%

Disabilitybeneﬁts Inreceiptofdisabilitybeneﬁts 62 54%

Notinreceiptofdisabilitybeneﬁts 53 46%

Timefromonset(ofNES) 4–5years <1yearto20+years

Diagnosedby(multipleanswerspossible) Aneurologistwhospecialisesinseizures 81 70%

Aneurologistwhodoesnotspecialiseinseizures 27 23%

Apsychiatristorclinical/neuropsychologist 28 24%

Timetodiagnosis(ofNES) 3–4years <1yearto20+years

TestsusedtodiagnoseNES(multipleanswers possible)

Electroencephalography(EEG) 96 83%

AmbulatoryElectroencephalography(Amb-EEG) 23 20%

Video-Electroencephalographymonitoring(vEEG) 60 52%

Electrocardiography(ECGorEKG) 69 60%

Magneticresonanceimaging(MRI) 88 77%

ComputedTomography(CT/CATscan) 75 65%

Tilttabletest 15 13%

Priorerroneousdiagnosisofepilepsy Yes 28 24%

No 84 73%

Self-reportedseizurediagnosis NESalone 100 87%

NESandEpilepsy 15 13%

NESfrequencypastmonthpriortotesting 15 0-309

(4)

stigma perceptions of socio-demographic groups detailed in Table2,norwasstigmasigniﬁcantlycorrelatedwithparticipant age(rs= 0.007,p=0.942).

A significant and moderate inverse correlation was found betweenperceivedstigmascaleandHRQoL(QOLIE-31)totalscores (rs= 0.474,p<0.001).AsdetailedinTable3,analysisofQOLIE-31 subscales (post Holm–Bonferroni method) shows that seizure worryâ,emotionalwellbeing^b,socialfunctioning^candstigmascale scoresweresignificantlyandmoderatelycorrelated;thepropor- tionofsharedvarianceforthesesubscaleswas â23%,^b18%, and

c17%. Energy/fatigue and cognitive subscales were found tobe signiﬁcantly, but weakly correlated with stigma. Medicat ion effectsand(thesingle-item)OverallQoLsubscaleswerenotfound tobesigniﬁcantlycorrelated.

4.Discussion

Thisstudysoughttoexploretherelationshipbetweensocial factors (socio-demographic characteristics and stigma perceptions)andtheHRQoLofadultswithNES.Participantswerefound toexperiencehighlevelsofperceivedstigmawhichwasinversely correlatedwithHRQoL. Stigma perceptions weremoststrongly associated with the HRQoL domains seizure worry, emotional wellbeing, and social functioning. HRQoL was better amongst thoseinemploymentoreducationthanthosewhowerenot.

The levels of perceived stigma reportedby our participants (mean4.9)areconsiderablyhigherthantypicallyfoundinepilepsy patientpopulations.Astudyof314peoplewithepilepsyusingthe samemeasurereportsameanscoreof3.7[27].Similarly,arecent studyusingasinglefour-pointLikertscalequestiontakenfromthe NEWQOL-6D(Howmuchdoyoufeelpeopletreatyouasaninferior person?), found that perceived stigmawas signiﬁcantly higher amongindividualswithNEScomparedtothosewithepilepsy[29].

Theseﬁndingsﬁtwiththewiderliterature,which suggeststhat people with functional somatic syndromes experience greater perceived stigmatisation than those with comparable organic disease[30].

Thestigmaofepilepsyiswidelyreported,andisconsistently linkedtoreducedHRQoL[22,31].Toourknowledge,oursistheﬁrst

studytoexploreassociationsbetweenHRQoLandstigmaamong adultswithNES.Asignificantandmoderateinverse correlation wasobserved;suggestinghigherperceptionsofstigmacontribute to poorer HRQoL among those with the condition. Stigma perceptions were found to be most strongly associated with seizureworry,emotionalwellbeing,andsocialfunctioningHRQoL domains; withover one-half of the variability related tothese features.Thereisadearthofresearchexploringthesocialstigmaof NES, but peripheral findings from previous studies broadly corroborateourfindings.StudiesshowthatpeoplewithNEScan experience feelings of shame [32], blame and stigmatisation [ 33,34];andmightconcealtheconditionandisolatethemselvesto avoidpotentialadversesocialreactionstoseizuresandfeelingsof embarrassment[34,35].On-goingsupportfromfamily,friendsand colleagueshasbeen describedextremely importantin counter- actingthesocialisolationassociatedwithNES[36].

ForpeoplewithNES,theirstigmaperceptionsareprobablynot withoutfoundation.InWesternnationsderogatoryviewsofNES may be linked to the disparaging use of terms such as

‘psychosomatic’inthemedia,whichmightbetakentomeanan illnessthatisfeigned,malingeredorrepresentativeofacharacter ﬂaw[37].Unfortunately,thesepejorativeopinionsarealsofound inmedicalcircles[38].Forthosewiththecondition,stigmatising interactions withhealth professionals are notuncommon [39].

People with NES often report their symptoms are met with disbelief,nottakenseriously,andthatthelegitimacyoftheillness is sometimes questioned by clinicians [32,33,35,38–41]; and research exploring health professionals’ views supports these assessments[42–48].

Contrary toprevious HRQoL findings [15], we foundpartic- ipantswhoreportedbeinginemploymentoreducation(part-time orfull-time)tohavesignificantlybetterHRQoLthanthosewho werenot.Thisdiscrepancymightbeexplainedtheclassificationof thoseineducationas‘employed’inouranalysis.Beforeapplying theHolm–Bonferronimethod,wealsofoundreceiptofdisability benefitstobeadifferentiatingfactor.Aspreviouslyobserved[15], we did not find relationship status or participants’ age to be discriminatingfactors.Norweresignificantdifferencesobserved inrelationtoparticipants’countryofresidence,gender,orliving arrangementsandtheirHRQoL.Toourknowledge,thesearenovel findingsandrequiresubstantiation.

Seizurefrequencywas showntobesignificantly,but weakly correlatedwithHRQoL. Thisfinding isin contrasttothose ofa systematicreviewwhichconcludedthatseizurefrequencyisnota predictorofHRQoLinthispatientgroup[6];butisconsistentwith astudyof96patientswithNES,whichfoundseizurefrequencyto besignificantlyassociatedwithlowerHRQoLsummaryscores(SF- 36)[10].

Participants’socio-demographiccharacteristicswerenotfound todeterminestigmaperceptions.However,signiﬁcantdifferences Table3

CorrelationsbetweenHRQoL(QOLIE-31)subscalesandStigmaScalescores.

QOLIE-31subscalescores(weighted) rs p

Seizureworry 0.479 <0.001

Emotionalwellbeing 0.421 <0.001

Socialfunctioning 0.407 <0.001

Cognitive 0.314 0.001

Energyandfatigue 0.252 0.007

Medicationeffects -0.146 0.120

OverallQoL -0.132 0.161

Table2

DifferencesinHRQoLandStigmabetweensocio-demographicgroups.

QOLIE-31totalscore StigmaScaletotalscore

Groupingvariable Groups Median(andrange) Sig. Median(andrange) Sig.

Countries UKandIreland 33.3(13.4–87.8) 0.534 5.1(1.3–7.0) 0.250

USandCanada 30.6(3.4–60.2) 5.3(1.0–7.0)

Gender Male 32.0(11.7–57.0) 0.532 5.0(3.6–6.4) 0.934

Female 31.7(3.4–87.2) 5.2(1.0–7.0)

Relationshipstatus Single,separated,divorced 32.5(12.6–87.8) 0.929 5.5(1.3–7.0) 0.022

Married,CivilUnion,partnered 31.1(3.4–66.5) 5.0(1.0–6.8)

Livingarrangements Livingalone 39.3(15.7–87.8) 0.126 5.5(1.3–6.5) 0.463

Livingwithothers 30.4(3.4–66.5) 5.2(1.0–7.0)

Employmentstatus Notinworkoreducation 28.0(3.4–66.5) <0.001 5.3(1.0–7.0) 0.017

Inworkoreducation 41.5(22.2–87.8) 4.8(1.3–6.6)

Disabilitybeneﬁts Inreceipt 29.6(9.7–66.5) 0.025 5.3(1.7–7.0) 0.057

Notinreceipt 37.9(3.4–87.8) 4.9(1.0–6.5)

(5)

inlevelsoffeltstigmaaccordingtorelationshipandemployment statuswerenotedpriortoapplyingtheHolm–Bonferronimethod;

and seizure frequency was shown to be significantly (albeit weakly)correlatedwithperceivedstigmascores.Thesefindings areconsistentwithstudiesofepilepsypatientpopulations[27,49], butrequireverificationinNESpatientpopulations.

4.1.Limitations

Whileourﬁndingsofferanovelcontributiontotheliterature,it isimportantthattheyareinterpretedwithinthecontextoftheir limitations.

Perhapsthegreatestconcernwiththecollectionofinternet- basedpatientinformationisthereliabilityandvalidityofthedata obtained.Yet,recentreviewssuggesthealthdatacanbecollected withequalorevenbetterreliabilityinWeb-basedquestionnaires comparedwithtraditionalapproaches[50].Participantswereable tocompletethesurveyoveranextendedperiodiftheysowished, andsurveymetricsshowthat97%ofrespondentscompletedthe surveywithin82h(around3.5days).Theaddedbenefitoftimefor reflection,theabilitytoconsiderandcorrectinformation,andthe useofvalidationchecks(asusedinoursurvey)hasbeenshownto improvedataquality[50].Therearealsostrongindicationsthat web-basedquestionnairesarelesspronetosocialdesirabilitybias[ 50,51].Studiesshowthatperceivedhealth statusdata[52] and HRQoL measures [53] can be reliably collected using online methods.However,itisimportanttonotethatthestandardised measuresusedinthisstudyhavenotbeentestedininternet-based studies,andresearchisneededtoconfirmtheironlinereliability.

Duetothedesignofourstudy,thereisnowaytoassessresponse rate. Using number of surveys started as a proxy denominator suggestsacompletionrateof49%.

PreviousstudiesexploringtheHRQoLofadultswithNEShave recruited participantsfrom inpatient epilepsymonitoringunits (EMUs),outpatientneurologysettings,psychotherapeuticcentres, or a combination of these; and most report diagnoses were establishedusingvideo-electroencephalographymonitoring(video-EEG) [6]. Video-EEG is the best-practice (‘gold standard’) diagnosticmethod [54]; however, it is expensive and resource limited[55]andmaynotbefeasiblebecauseofthelowfrequency ofseizures[56].

Oftheparticipantsinthisstudyonlyhalfdescribeundergoing video-EEG monitoring,withtheremainder reportingelectroen- cephalography (EEG) or ambulatory-electroencephalography (Amb-EEG)testing. Ourapproach meansthat wecannot sayto what extent participants met the diagnostic criteria for NES proposed by the PNES Task Force of the International League AgainstEpilepsy Non-epileptic Seizures Task Forceguidelines [ 54,57]. We must also consider that the diagnosis of NES is notoriouslycomplexanddifﬁcult,andsomeparticipantsmayhave been misdiagnosed. In view of the uncertainties about the diagnosisinherent inour recruitmentmethod,theinclusion of people witha dual diagnosis of epilepsy and NES (13% of the sample) could also be considered a limitation of this study.

However, given that this study was intended to explore the sociologicaldimensionofNES,wethoughtitwasimportantnotto exclude any subgroup of the whole NES patient population.

Epilepsyisanimportantcomorbidityof NESandthe13%ﬁgure actually places our study well within the prevalence range of comorbidepilepsywhichhaspreviouslybeenreportedinHRQoL studiesofNESpatientpopulations(6–22%)[9,11,20].Participants withmixedseizuredisorderswereencouragedtothinkabouttheir NESwhen respondingtoquestions abouttheirseizures,butwe acknowledgethatwecannotbecertainthatallrespondentswere abletodistinguishaccuratelybetweentheirepilepticand nonepilepticseizures.

Despite differences in recruitment methods, other socio- demographicandhealth-relatedcharacteristicsofourparticipants are alsowithintherangeof those reportedinpreviousstudies exploring theHRQoL of peoplewith NES.In terms of age and gender (mean 31–42 years, 69–100% female) [6]; relationship status(58%marriedorpartnered)[4,9];proportionineducationor employment(45–67%)[4,8];timefromonsetofNES(median3–4 years)[10,13](mean4.7–8.9years)[3,5,8,18];timetodiagnosisof NES (median3.5 years)[11]; andfrequency of NESin thefour weekspriortotesting(median6–15seizures)[4,5,10]mean(10.9 to23.7)[3,12,20].Data pertaining tophysicalandpsychological comorbiditieswasnotwithinthescopeofouranalysis,andour sample might differ from those previously described in these respects.

Theremightalsobeimportantdifferencesbetweenpeoplewith NESwho haveaccess totheInternetand participatein patient supportgroups,andthosewhodonot.Itisalsoaweaknessofthe studythattherecruitmentmethoddidnotallow ustorecruita comparison group,and thestudyiscross-sectional andcorrela- tional,whichmeansthatresultscanbebidirectionalandshouldbe interpreted with caution. It is possible that changes in social circumstancesorstatusaremorerelevanttoHRQoLand/orstigma thancurrentcircumstances–somethingbestexploredlongitudi- nally. The correlational nature of our ﬁndings means that we cannotsayanythingaboutcausalities.

5.Conclusion

Despite these limitations, theresearch improves ourunder- standingofhowsocialfactorsanddynamicsmightinfluencethe HRQoL of adults withNES.To ourknowledge,the studyis the largestHRQoLsurveyofpeoplewithNEStodate,andthefirstto exploretherelationshipbetweenHRQoLandstigmainthispatient group.Animportantfindingisthatparticipantsexperiencehigh levelsofperceivedstigma,whichnegativelyaffectstheirHRQoL.

Ourdatasuggeststhatnotbeinginemploymentoreducationis detrimentaltotheHRQoLofpeoplewithNES.

Theseexploratory ﬁndings servea heuristicfunction,in that they identify several issues for further (qualitative) research.

Qualitative analysiscanhelp achieve fullerand more complete descriptionsofphenomena,helpcorrectinterpretationofquanti- tative results, and providetriangulation [58]. Perceived stigma couldbeatreatmenttarget,andresearchisneededtounderstand how–andwhy–thosewiththeconditionexperiencestigmatisa- tion;whichinourdatawasmoststronglyassociatedwithseizure worry, emotional wellbeing, and social functioning HRQoL domains.Moreresearchisalsoneededtounderstandfactorsthat impedeandhelpfacilitatetheparticipationofpeoplewithNESin education and employment. These studies could be usefully followed-up by a projectthat looks speciﬁcallyat the enacted stigmafacedbythispatientgroup.

Ethicalapproval

EthicalclearancewasapprovedbyNelsonMandelaUniversity Human ResearchEthics Committee(NMU-HREC) on22nd April 2016(REF:H16-RTI-RCD-002),wheretheﬁrstauthorwasavisiting researcheratthetimeofdatacollection.Thesurveywashostedby aUK-EUdata-protectioncompliantprovider.Potentialparticipants were guided through an online study protocol detailing the purposeofthestudy,anddataprotection,ethicalcomplianceand complaint procedures.Allparticipantsgaveinformedelectronic consent to participate, and did so again to conﬁrm survey completionandauthorisesubmissionoftheirdataforuseinthe researchproject.

(6)

Competinginterests

Nocompetinginterestsaredisclosed.

Conﬂictofintereststatement

The authors certify that they have NO affiliations with or involvement in any organization or entity with any financial interest(suchashonoraria; educationalgrants; participationin speakers’bureaus;membership,employment,consultancies,stock ownership, or other equity interest; and expert testimony or patent-licensingarrangements),ornon-financialinterest(suchas personalorprofessionalrelationships,affiliations,knowledgeor beliefs) in the subject matter or materials discussed in this manuscript.

Grantinformation

FundedbyWellcomeTrustGrantREF:grantnumber:200923.

Acknowledgements

Authors thanktheonlinesupportgroupsfor advertisingthis studytotheirmembers,and participantsfor contributing their timetoparticipateinthisstudy.

References

[1]DevlinNJ,ApplebyJ.Gettingthemostoutofproms.Puttinghealthoutcomes attheheartofNHSdecision-making.London:TheKing'sFund;2010.

[2]TheWhoqolG.TheWorldHealthOrganizationqualityoflifeassessment (WHOQOL):developmentandgeneralpsychometricproperties1.SocialSci Med1998;46:1569–85.

[3]TestaSM,SchefftBK,SzaﬂarskiJP,YehH-S,PriviteraMD.Mood,personality, and health-related quality of life in epileptic and psychogenic seizure disorders.Epilepsia2007;48:973–82.

[4]KarakisI,MontourisGD,PiperidouC,LucianoMS,MeadorKJ,ColeAJ.Patient andcaregiverqualityoflifeinpsychogenicnon-epilepticseizurescomparedto epilepticseizures.Seizure2014;23:47–54.

[5]LaFranceJWC,AloscoML,DavisJD,TremontG,RyanCE,KeitnerGI,etal.

Impactoffamilyfunctioningonqualityoflifeinpatientswithpsychogenic nonepilepticseizuresversusepilepsy.Epilepsia2011;52:292–300.

[6]JonesB,ReuberM,NormanP.Correlatesofhealth-relatedqualityoflifein adultswithpsychogenicnonepilepticseizures:asystematicreview.Epilepsia 2016;57:171–81.

[7]BroughJL,MoghaddamNG,GresswellDM,DawsonDL.Theimpactofreceiving adiagnosisofNon-EpilepticAttackDisorder(NEAD):asystematicreview.J PsychosomRes2015;79:420–7.

[8]Kuyk J, Siffels MC,BakvisP, Swinkels WAM.Psychological treatmentof patientswithpsychogenicnon-epilepticseizures:anoutcomestudy.Seizure 2008;17:595–603.

[9]LaFrance WC, Syc S. Depression and symptoms affect quality of lifein psychogenicnonepilepticseizures.Neurology2009;73:366–71.

[10]LawtonG,MayorRJ,HowlettS,ReuberM.Psychogenicnonepilepticseizures andhealth-relatedqualityoflife:therelationshipwithpsychologicaldistress andotherphysicalsymptoms.EpilepsyBehav2009;14:167–71.

[11]MitchellJW,AliF,CavannaAE.Dissociativeexperiencesandqualityoflifein patients with non-epileptic attack disorder. Epilepsy Behav 2012;25:

307–12.

[12]SzaﬂarskiJerzyP,HughesC,SzaﬂarskiM,FickerDavidM,CahillWilliamT,LiM, et al. Quality of life in psychogenic nonepileptic seizures. Epilepsia 2003;44:236–42.

[13]SzaﬂarskiJP,SzaﬂarskiM.Seizuredisorders,depression,andhealth-related qualityoflife.EpilepsyBehav2004;5:50–7.

[14]SzaﬂarskiJP,SzaﬂarskiM,HughesC,FickerDM,CahillWT,PriviteraMD.

Psychopathologyandqualityoflife:psychogenicnon-epilepticseizuresversus epilepsy.MedSciMonit2003;9:113–8.

[15]MyersL,LancmanM,Laban-GrantO,MatznerB,LancmanM.Psychogenicnon- epilepticseizures:predisposingfactorstodiminishedqualityoflife.Epilepsy Behav2012;25:358–62.

[16]RawlingsGH,BrownI,ReuberM.Predictorsofhealth-relatedqualityoflifein patientswithepilepsyandpsychogenicnonepilepticseizures.EpilepsyBehav 2017;68:153–8.

[17]CronjeG,PretoriusC.Thecopingstylesandhealth-relatedqualityoflifeof SouthAfricanpatientswithpsychogenicnonepilepticseizures.EpilepsyBehav 2013;29:581–4.

[18]StruttAM,HillSW,ScottBM,Uber-ZakL,FogelTG.Motivation,psychopathol- ogy, locus of control, and quality of lifein womenwith epilepticand nonepilepticseizures.EpilepsyBehav2011;22:279–84.

[19]TaylorRS,SanderJW,TaylorRJ,BakerGA.Predictorsofhealth-relatedquality oflifeand costsin adults withepilepsy: asystematic review.Epilepsia 2011;52:2168–80.

[20]QuiggM,ArmstrongRF,FaraceE,FountainNB.Qualityoflifeoutcomeis associatedwithcessationratherthanreductionofpsychogenicnonepileptic seizures.EpilepsyBehav2002;3:455–9.

[21]GreenB,NormanP,ReuberM.Attachmentstyle,relationshipquality,and psychologicaldistressinpatientswithpsychogenicnon-epilepticseizures versusepilepsy.EpilepsyBehav2018;66:120–6.

[22]JacobyA,SnapeD,BakerGA.Epilepsyandsocialidentity:thestigmaofa chronicneurologicaldisorder.LancetNeurol2005;4:171–8.

[23]Suurmeijer TPBM, Reuvekamp MF, Aldenkamp BP. Social functioning, psychological functioning, and quality of life in epilepsy. Epilepsia 2001;42:1160–8.

[24]RawlingsGH,ReuberM.Whatpatientssayaboutlivingwithpsychogenic nonepilepticseizures:asystematicsynthesisofqualitativestudies.Seizure 2016;41:100–11.

[25]VickreyBG,HaysPKRD,HermannBP,CramerJA,MeadorKJ,etal.QualityofLife inEpilepsyQOLIE-31(Version1.0):ScoringManualandPatientInventory.

SantaMonica,CA:Rand;1993.

[26]McDowellI.MeasuringHealthAGuidetoRatingScalesandQuestionnaires.

3rded.Oxford:OxfordUniversityPress;2006.

[27]DiIorioC, OsborneShaferP,LetzR,Henry T,SchomerDL, YeagerK.The associationofstigmawithself-managementandperceptionsofhealthcare amongadultswithepilepsy.EpilepsyBehav2003;4:259–67.

[28]HolmS.Asimplesequentiallyrejectivemultipletestprocedure.ScandJStat 1979;6:65–70.

[29]RawlingsGH,BrownI,ReuberM.Deconstructingstigmainpsychogenicnon- epilepticseizures:anexploratorystudy.EpilepsyBehav2017;74:167–72.

[30]LooperKJ,KirmayerLJ.Perceivedstigmainfunctionalsomaticsyndromesand comparablemedicalconditions.JPsychosomRes2004;57:373–8.

[31]JacobyA,AustinJK.Socialstigmaforadults andchildrenwith epilepsy.

Epilepsia2007;48:6–9.

[32]WyattC,LarawayA,WeatherheadS.Theexperienceofadjustingtoadiagnosis of non-epilepticattack disorder (NEAD) and the subsequent process of psychologicaltherapy.Seizure–EurJEpilepsy2014;23:799–807.

[33]KarterudHN,KnizekBL,NakkenKO.Changingthediagnosisfromepilepsyto PNES:patientsexperiencesandunderstandingoftheirnewdiagnosis.Seizure 2010;19:40–6.

[34]KarterudHN,HaavetOR,RisørMB.Socialparticipationinyoungpeoplewith nonepilepticseizures(NES):aqualitativestudyofmanaginglegitimacyin everydaylife.EpilepsyBehav2016;57(PartA):23–8.

[35]FaircloughG, Fox J, Mercer G, Reuber M, Brown RJ.Understanding the perceivedtreatmentneedsofpatientswithpsychogenicnonepilepticseizures.

EpilepsyBehav2014;31:295–303.

[36]PretoriusC,SparrowM.Lifeafterbeingdiagnosedwithpsychogenicnon- epilepticseizures(PNES):aSouthAfricanperspective.Seizure2015;30:32–41.

[37]StoneJ, Colyer M, Feltbower S, Carson A, SharpeM. Psychosomatic: a systematicreview of its meaningin newspaper articles. Psychosomatics 2004;45:287–90.

[38]TolchinB,BasletG,DworetzkyB.Psychogenicseizuresandmedicalhumor:

jokesasadamagingdefense.EpilepsyBehav2016;64(PartA):26–8.

[39]Robson C, Lian OS.Blaming, shaming,humiliation: stigmatising medical interactionsamongpeoplewithnon-epilepticseizures.WellcomeTrustOpen Res2017;2:1–17.

[40]Dickinson P,Looper KJ, GroleauD. Patientsdiagnosedwith nonepileptic seizures:theirperspectivesandexperiences.EpilepsyBehav2011;20:454–61.

[41]KarterudHN,RisørMB,HaavetOR.Theimpactofconveyingthediagnosis when using a biopsychosocial approach: a qualitative study among adolescentsandyoungadultswithNES(non-epilepticseizures).Seizure EurJEpilepsy2015;24:107–13.

[42]AhernL,StoneJ,SharpeMC.Attitudesofneurosciencenursestowardpatients withconversionsymptoms.Psychosomatics2009;50:336–9.

[43]KanaanR,ArmstrongD,BarnesP,WesselyS.Inthepsychiatrist'schair:how neurologistsunderstandconversiondisorder.Brain:JNeurol2009;132:2889–

96.

[44]KanaanR, Armstrong D, WesselyS.Limits to truth-telling: neurologists' communicationinconversiondisorder.PatientEducCouns2009;77:296–301.

[45]O'Sullivan SS, Sweeney BJ, McNamara B. The opinion of the general practitioner toward clinical management of patients with psychogenic nonepilepticseizures.EpilepsyBehav2006;8:256–60.

[46]ShnekerBF,ElliottJO.Primarycareandemergencyphysicianattitudesand beliefsrelatedto patients withpsychogenic nonepilepticspells. Epilepsy Behav2008;13:243–7.

[47]WhiteheadK,ReuberM.Illnessperceptionsofneurologistsandpsychiatrists inrelationtoepilepsyandnonepilepticattackdisorder.Seizure2012;21:104–

9.

[48]WorselyC,WhiteheadK,KandlerR,ReuberM.Illnessperceptionsofhealth careworkersinrelationtoepilepticandpsychogenicnonepilepticseizures.

EpilepsyBehav2011;20:668–73.

[49]Jacoby A. Stigma, epilepsy, and quality of life. Epilepsy Behav 2002;3:

10–20.

(7)

[50]vanGelderMMHJ,BretveldRW,RoeleveldN.Web-basedquestionnairesthe futureinepidemiology?AmJEpidemiol2010;172:1292–8.

[51]Booth-Kewley S, Larson GE, Miyoshi DK. Social desirability effects on computerizedand paper-and-pencil questionnaires.Comput Hum Behav 2007;23:463–77.

[52]GrahamAL,PapandonatosGD,BockBC,CobbNK,Baskin-SommersA,NiauraR, etal.Internet-vs.telephone-administeredquestionnairesinarandomized trialofsmokingcessation.NicotineTobRes:OffJSocResNicotineTob2006;8:

S49–57.

[53]BrentDB,StanleyEK,SpertusJA.Electroniccollectionofhealth-relatedquality oflifedata:validity, timebeneﬁts, andpatientpreference.QualLife Res 2001;10:15–21.

[54]LaFranceWC,ReuberM,GoldsteinLH.Managementofpsychogenic nonepilepticseizures.Epilepsia2013;54:53–67.

[55]McGonigalA,OtoM,RussellAJC,GreeneJ,DuncanR.OutpatientvideoEEG recordinginthediagnosisofnon-epilepticseizures:arandomisedcontrolled trial of simple suggestion techniques. J Neurol Neurosurg Psychiatry 2002;72:549–51.

[56]MalmgrenK,ReuberM,AppletonR.Differentialdiagnosisofepilepsy.In:

ShorvonM,GuerriniR,LhatooS,editors.Oxfordtextbookofclinicalneurology.

Oxford:OxfordUniversityPress;2012.p.81–94.

[57]LaFrance WC, Baker GA, Duncan R, Goldstein LH, Reuber M. Minimum requirementsforthediagnosisofpsychogenicnonepilepticseizures:astaged approach.Epilepsia2013;54:2005–18.

[58]KalinowskiP,LaiJ,FidlerF,CummingG.Qualitativeresearch:anessentialpart ofstatisticalcognition.StatEducResJ2010;9:22–34.