Lappin, Julia M. and Heslin, Margaret and Jones, Peter B. and Doody, Gillian A. and Reininghaus ...

22 Pages·2017·0.38 MB·English

by  Christine Middleton

TITLEPAGE
Title:Outcomesfollowingfirstepisodepsychosis–whyweshouldinterveneearlyinallages,not
onlyinyouth
RunningTitle:Agelimitsforearlyinterventionservices:timetorethink?
Authors:JMLappin&MHeslin,PBJones,GADoody,UAReininghaus,ADemjaha,TCroudace,T
Craig,KDonoghue,BLomas,PFearon,RMMurray,PDazzan&CMorgan.
CorrespondingAuthor: JuliaMLappin,SeniorLecturer,SchoolofPsychiatry,UniversityofNew
SouthWales,Sydney,Australia.Email:[email protected]:++61293668610
MargaretHeslin,HonoraryLecturer,InstituteofPsychiatry,PsychologyandNeuroscience,King’s
CollegeLondon,UK.
PeterBJones,ProfessorofPsychiatry,UniversityofCambridge,Cambridge,UK.
GillianADoody,ProfessorofPsychiatry,UniversityofNottingham,Nottingham,UK.
UlrichAReininghaus,PostdoctoralResearchFellow,InstituteofPsychiatry,Psychologyand
Neuroscience,King’sCollegeLondon,UK.
ArsimeDemjaha,HonoraryLecturer,InstituteofPsychiatry,PsychologyandNeuroscience,King’s
CollegeLondon,UK.
TimothyCroudace,ProfessorofAppliedHealthResearch,UniversityofDundee,Dundee,UK.
ThomasJamieson-Craig,ProfessorofPsychiatry. InstituteofPsychiatry,Psychologyand
Neuroscience,King’sCollegeLondon,UK.
KimDonoghue,PostdoctoralResearchWorker.InstituteofPsychiatry,PsychologyandNeuroscience,
King’sCollegeLondon,UK.
BenLomas,ConsultantPsychiatrist,NottinghamUniversityHospitalsNHSTrust,Universityof
Nottingham,Nottingham,UK.
PaulFearon,ProfessorofPsychiatry,TrinityCollegeDublin,Ireland.
RobinMMurray,ProfessorofPsychiatry,InstituteofPsychiatry,PsychologyandNeuroscience,
King’sCollegeLondon,UK.
PaolaDazzan,ReaderofPsychiatry,InstituteofPsychiatry,PsychologyandNeuroscience,King’s
CollegeLondon,UK.
CraigMorgan,ProfessorofSocialEpidemiology,InstituteofPsychiatry,Psychologyand
Neuroscience,King’sCollegeLondon,UK.
Departmentinwhichworkwasdone:DepartmentofPsychosisStudies,InstituteofPsychiatry,
PsychologyandNeuroscience,King’sCollegeLondon,UK.
Abstract
Objective:Tocomparebaselinedemographicsandten-yearoutcomesofafirstepisodepsychosis
patientincidencecohortinordertoestablishwhethercurrentyouth-focusedage-basedcriteriafor
EarlyIntervention(EI)servicesarejustifiedbypatientneeds.
Methods:DataonfirstepisodepsychosispatientsfromtheAESOP-10longitudinalfollow-upstudy
wereusedtocomparebaselinecharacteristics,andten-yearclinical,functional,andserviceuse
outcomesbetweenthosepatientswhowouldandwouldnothavemetage-basedcriteriaforEarly
InterventionServices,inAustraliaorintheUK.
Results:58%menand71%womenwithfirstepisodepsychosisweretoooldtomeetcurrent
Australian-EarlyInterventionageentry-criteria(χ2=9.1,p=0.003);while21%menand34%women
weretoooldforUK-EarlyInterventionage-entrycriteria(χ2=11.1,p=0.001).Ten-yearclinicaland
functionaloutcomesdidnotdiffersignificantlybetweengroupsbyeitherAustralian-orUK-Early
Interventionage-entrycriteria.Serviceusewassignificantlygreateramongthepatientsyoung
enoughtomeetEarlyInterventionage-criteria[Australia:IRR=1.35(1.19-1.52)p<0.001;UK:
IRR=1.65(1.41-1.93)p<0.001].
Conclusions:CurrentEarlyInterventionservicesaregender-andage-inequitable.Largenumbersof
patientswithfirstepisodepsychosiswillnotreceiveEarlyInterventioncareundercurrentservice
provision. Illnessoutcomesatten-yearswerenoworseinfirstepisodepsychosispatientswho
presentedwithintheagerangeforwhomEarlyInterventionhasbeenprioritised,thoughthese
patientshadgreaterserviceuse.ThesedataprovidearationaletoconsiderextensionofEarly
Interventiontoall,ratherthanjusttoyouth.
Introduction
Specialistearlyinterventionservices(EIS)provideintensivesupportandmanagementforyounger
individualsintheearlyyearsfollowingtheirfirstpsychoticillness.Thereisnotablevariability
internationallyintheupperagethresholdselectedforEIserviceprovision:inAustralia,servicesare
typicallyoffereduptoage25;inSingaporetoage40.IntheUK,35yearshasbeenthe
recommendedupperagecut-offforreferrals(DoH,2001)butrecentNICEguidelinesrecommend
thatearlyinterventionshouldbeavailabletoall,regardlessofage(NICE,2014).
Historically,EIserviceshavebeenyouth-focussedbasedonseveralprinciples:first,thezeitgeistthat
themajorityofpsychosispresentsearlierinlife.Second, priortoEISdevelopment,evidencethat
therewasdelayamongyoungpeoplewithemergingpsychosisobtainingearlytreatment(Lincoln
andMcGorry,1995),andfinally,thetheorythatthosewhodeveloppsychosisatayoungerage
suffergreaterlong-termimpairmentbecausetheillnessinterruptstheirsocial,personaland
scholastic/occupationaldevelopment(DoH,2001).
UsingdatafromtheUKAESOP-10study-alongitudinalfollow-upofanincidencecohortoffirst
episodepsychosispatients-thisstudyexaminedfirst,thebaselinecharacteristicsoffirstepisode
psychosis(FEP)individualswhowouldandwouldnotmeetcurrentage-basedcriteriaforEISin
Australia,orintheUK.Second,ittestedthequestionwhetherten-yearclinical,functionaland
serviceuseoutcomeswereworseinthosewhodevelopFEPatanageyoungenoughtomeetcriteria
forEIprovision.Itisimportanttoemphasisethatthecohortstudiedwastreatedinanerapriorto
theestablishmentofEIservices;thusthisisnotanexaminationoftheeffectivenessofEIcare.
Rather,theseanalysescomparebaselinecharacteristicsandten-yearoutcomesofallfirstepisode
psychosispatientsinordertoestablishanevidencebaseforEIprovisionbytestingthetheorythat
thosewhodeveloppsychosisatayoungeragehaveworseoutcomesthanthosewhodevelop
psychosisatanolderage.
Methods
Setting
ThispaperisbasedondatafromÆSOP(Kirkbrideetal,2006)andÆSOP-10(Morganetal,2014),
whichareincidenceandten-yearfollow-upstudies,respectively,ofallindividualswithafirst
episodeofpsychosispresentingforthefirsttimetospecialistmentalhealthservicesindefined
catchmentareasintheUKbetween1997and1999.RecruitmentofÆSOPcasesendedbeforeEI
serviceswereestablishedintheseareas.
Cases
WithintightlydefinedgeographicalareasinLondonandNottingham,allcaseswithfirstepisodeof
psychosis(codesF20–29andF30–33inICD–10(WHO,1993))whopresentedtospecialistservices
wereincludedintheincidencestudy.TheScreeningScheduleforPsychosis(Jablenskyetal,1992)
wasusedtoscreencaseswhopresentedtotheseservicesforeligibilityandcompletedbasedon
informationfromclinicalnotes,corroborationfrommentalhealthstaffand,wherepossible,by
interviewwiththeparticipant.
Inclusioncriteriaforcaseswere:agedbetween16and64yearswithafirstepisodeofpsychosisand
residentwithinthestudycatchmentareas.Exclusioncriteriawere:evidenceofpsychoticsymptoms
precipitatedbyanorganiccause;transientpsychoticsymptomsresultingfromacuteintoxicationas
definedbyICD–10(WHO,1993);previouscontactswithmentalhealthservicesforpsychosis;and
moderateorseverelearningdifficulties,oranIQoflessthan50(WHO,1993).
Follow-up
Caseswerefollowed-uptenyearsafterfirstcontactwithmentalhealthservices(detailedinMorgan
etal(2014).Atbaseline,532incidencecaseswereidentified.Ofthose,387hadfollow-upoutcome
dataandthusmadeupthecoreanalyticsampleforoutcomeanalyses(excludingthosewhohad
died,emigratedorbeenexcluded,plusthosewhodidnothaveuseableinformationonclinical
courseandoutcomeforatleasteightyearsoffollow-up).Withinouranalysesofoutcomes,we
excludedsixfurthercasesastheypresentedtoserviceslessthanthreeweeksbeforeEISwere
launchedintheLondoncatchmentarea.
Measures
Baseline:Clinicalanddemographicdatawerecollectedfromclinicalrecordsand,wherepossible,
frominterview.AshortenedversionoftheSchedulesforClinicalAssessmentinNeuropsychiatry
(SCANversion2;WHO,1994)wasusedtoassesssymptompresenceandseverity.Thiswasusedin
conjunctionwithotherclinicalinformation(excludingdiagnosis)toassignICD-10(WHO,1993)
psychoticdiagnoseswithinresearch-teamconsensusmeetings.Diagnosesweremadeblindto
ethnicityanddiagnosisfromtheclinicalnotes,assoonaspossibleafterfirstcontact.ThePersonal
andPsychiatricHistorySchedule(WHO,1996)wasusedtodeterminedurationofuntreated
psychosis(DUP),definedastheperiodfromonsetofpsychosistofirstcontactwithstatutorymental
healthservices.Onsetofpsychosiswasdefinedasthepresenceforoneweekormoreofpsychotic
symptoms,furtherdetailedinMorganetal(2006).
Follow-up:TheWHOLifeChartSchedule(Harrisonetal,2001;Sartoriusetal,1996;Susseretal,
2000)-designedtoassessthelong-termcourseofschizophrenia-wasusedtocollateinformationat
follow-up.Itcomprisesfourmainareas:symptoms;treatment;residence;andwork.Itwasadapted
toincludeadditionalinformationonserviceuse,includinguseofprescribedmedicationoverfollow-
up.InformationwasderivedfortheLifeChartfrommultiplesources:casenotes,interviewswith
cases,andinformantinformation.Keyvariableshavebeendefinedpreviouslyelsewhere(Morganet
al,2014).Courseofillnesswascategorisedasfollows:remissionwithinsixmonths;episodic(no
episodelongerthansixmonths’duration);continuous(noremissionlongerthansixmonths’
duration);ornoneoftheabove.Coursetype“noneoftheabove”referstoanintermediateillness
coursewhichwasneitherepisodicnotcontinuous;theseindividualsexperiencedbothanepisode
thatlastedlongerthan6monthsandaperiodofremissionthatlastedlongerthan6monthsduring
thefollow-up.
TheLifeChartwasalsousedtorecordnumberofinpatientdays;mentalstateatfollow-up(inthe
last30days;psychoticornotpsychotic);historyofself-harmoverfollow-up;lifetimesubstance
misuse(present/absent);andpercentageoftimeemployedoverfollow-up(dichotomisedintounder
andover25%).TheLifeChartwaspresentedatconsensusmeetingsalongwithcasenote
informationsothatdecisionsaboutallaspectsoftheLifeChartcouldbedecideduponbyconsensus.
TheGlobalAssessmentofFunction(GAF)disabilityscale(Endicottetal,1976)wasusedtoassess
functionatfollow-up;higherscoreindicatesbetterlevelofgeneralfunctioning.Treatment
resistancewasdefinedinlinewithmodifiedKanecriteriafortreatmentresistance(ConleyandKelly,
2001).
Ethics
Fullethicalapprovalforallaspectsofthefollow-upwasprovidedbythelocalresearchethics
committeesinSouthEastLondonandNottingham.Allresearchershadsubstantiveorhonorary
contractswitheithertheSouthLondonandMaudsleyNationalHealthService(NHS)Foundation
TrustortheNottinghamHealthcareNHSTrust,theprimaryparticipatingserviceproviders.
Analyses
AlldatawereanalysedusingSTATA11(StataCorp,2009).Ageatfirstpresentationtospecialist
serviceswasusedtoassignsubjectstoFEPgroupsbyage.Usingchi-squaretests,wecompared
clinical and sociodemographic characteristics between those ≤35years and those ≥ 36years (UK-EI
age-criteria); these analyses were repeated for those ≤25years and those ≥ 26years (Australia-EI age-
criteria).
We compared outcomes in those ≤35years and those ≥ 36years (UK-EI age-criteria) using, as
appropriate,logistic(binaryormultinomial),Poisson,orlinearregressionanalyses.Weadjusted
analysesforgender,ethnicity,centre,anddiagnosistoassesswhethervariationsinoutcomebyage-
groupstatuswereaccountedforbythesevariables.Non-normallydistributedcontinuousdatawere
analysedusingnon-parametricbootstrapregressions(ordinaryleastsquares).Bootstrapregressions
wereusedastheyproducethesamecoefficientsaslinearregression(andsoareinterpretedinthe
sameway)butgivemorerobustconfidenceintervalsandthereforeamorerobustestimateof
statistical significance. Analyses were repeated for those ≤25years and those ≥ 26years (Australia-EI
age-criteria).
Results
532incidencecaseswereidentifiedatbaselineandcomprisedthesamplefordeterminingwho
wouldhavebeeneligibleforEIserviceprovision.DemographiccharacteristicsaredetailedinTable
1.Table2showsthenumberandpercentageofcaseswhowouldandwouldnothavemetage-entry
criteriaforEISinAustralia(<=25years);andintheUK(<=35years);bygender,DUP,diagnosis,
treatmentresistanceandsubstanceuse.
Baselineillnessprofilesandcharacteristics
Australia-EISage-criteria.Of532cases,196(36.8%)wouldhavemetageentrycriteriaforEISin
Australia(χ2 =9.1, df1, p=0.003) (Table 2). 42% of men were ≤25years, compared with only 29% of
women aged ≤25 (χ2 =11.0, df1, p=0.001). There was a greater proportion of patients aged≥ 26years
who had a DUP>2years (19.4% compared to 7.7% in those ≤25years) (χ2=13.2,df1,p=0.001). There
wasanon-significanttrendforgreaterproportionofdepressivepsychosesinthe26+yearsgroup
(χ2=5.8, df2, p=0.06). Treatment resistant illness was significantly more common in the ≤25years
group (30.5% compared to 16.7% in those≤26years) (χ2=6.7,df1,p=0.009);aswassubstancemisuse
group (31.5% compared to 17.8% in those≥26years) (χ2=10.4,df1,p=0.001)(Table2).
UK-EISage-criteria.Of532cases,391(73.5%)wouldhavemetageentrycriteriaforEISinUK. 62%
of men were ≤35years, compared with only 38% of women aged ≤35 (χ2=11.0,df1,p=0.001)(Table
2).Atotalof80FEPcasespresentedwithaDUP>2yearsandinsomeserviceswouldnothavebeen
accepted by EIS. Of those ≤ 35years, 11.5% had a DUP over two years, compared to 24.8% of
those≥36years (χ2=14.4,df1,p<0.001).Diagnostically,thereweresignificantlyhigherproportionsof
manic psychoses in those ≤35years, and of depressive psychoses in those≥36years (χ2=11.5,df2,
p=0.003).Thereweresignificantlymorecasesoftreatmentresistantillness(χ2=8.3,df1,p=0.004)
andofsubstancemisuse(χ2 =26.56, df1, p<0.001) in those ≤35years (Table 2).
Outcomes
Thecoreanalyticsampleforanalysesofoutcomesatfollow-upcomprised387individuals(Morgan
etal,2014).Fortheanalysesdetailedhere,sixwereexcludedwhopresentedlessthanthreeweeks
beforetheintroductionofEISintheLondoncatchmentareaforAESOP,givingatotalsampleof381.
Clinical,functionalandserviceuseoutcomesareshowninTable3[Australia-basedanalyses]and
Table4[UK-basedanalyses]. Adjustedanalyseshavebeenincludedtoprovideinformationabout
howkeyvariables(gender;centre;diagnosisandethnicity)impactonoutcomes;unadjusted
analysesarereportedbecausetheyreflecttheserviceuseandoutcomesofthepopulationsasthey
wouldbepresentingtoEISorotherservices,butitisimportanttonotethatforanygivenservice,
theseoutcomeswoulddifferaccordingtothedemographicsofthepopulationbeingserved.
ClinicalOutcomes
Australia-EISage-criteria.Therewerenodifferencesbetweengroupsincourseofillness(OR=1.18,
CI:0.67-2.08);mentalstateatfollow-up(OR=1.00,CI:0.61-1.62);orself-harm(OR=1.17,CI:0.63-
2.16)(Table3).
UK-EISage-criteria.Therewasnodifferencebetweengroupsincourseofillnessoverfollow-up,with
thehighestproportioninbothgroupshavingacoursethatwasneitherepisodicnorcontinuous(OR=
0.96,CI:0.51-1.81),norinmentalstateatfollow-up(OR=1.28,CI:0.75-2.20).Morepatients
≤35years engaged in self- harming behaviour over follow-up than those≥36years (OR=2.88, CI: 1.18-
7.04,p=0.02)(Table4).
FunctionalOutcomes
Australia-EISage-criteria.NeitherGAFdisabilityatfollow-up(BMD=-1.19,CI:-5.63-3.24)nor
employmentoverfollow-up(OR=1.20,CI:0.70-2.04)significantlydifferedbetweengroups(Table3).
UK-EISage-criteria.Again,neitherGAFdisabilityatfollow-up(BMD=-1.85,CI:-6.81-3.11)nor
employmentoverfollow-up(OR=1.87,CI:0.96-3.65)significantlydifferedbetweengroups(Table4).