Sie sind auf Seite 1von 8

InteftationalJoLrul of So.ial Psrchi4ttr(1999)Vol.45 No.4 276-283.




Subjeclive qualityol life is an imponantcritedonin outcomeevatuation that has
beenwelfresearched in psychiatry.By compason, the therapeutic retationship
whichmay also be subjeclively assessedhas been €tativetynegtected as an
oulcomecrilerionalthoughit has predictive powerin retationto outcome.This
exploratorystudyinvestigaled subjectivequatityof tifeandlhefapeuticretationships
in Inslaomrssion (N 90)andlolg.term(N=1681schizopf,en.a patierrs, eaih
at iwo poinlsol rime.The lottow-up periodwas 9 monthsfor the fiGt-admission
sampleand 1.5 yearslor the tongtermsampte.A signtficant retaiionship was
toundbelveenglobaiassessments of qualityol tifeandtherapeutic rctationships
In long-lelm,bul nolin firsl-admissionpatients. Thisfindingwasconsistent al both
assessmenrs! suggesting thattherapeuticretationships maybecomemorecenlral
to qualilyol lile in long-termcare situationsand that patientsviewsof lhis
are increasingly embedded in theiroverattappraisat
ot tite.


Quality of Life (QoL) has beconc a popular consrrucr in rhe field of psychiarry and an
impoiari outcome criterion in evaluativeresearch.This is in line $ilh policy which slares
that improvenent in QoL is one of rbe major aims of rnenrathealrhcare.Although objeclile
indicatorsofQol are reponed. subjecriveindicaton ?re central ro its assessment. Tbe con-
strucr has been well-researchedin psychial,tr,over the pas( iwo decades.During rhis rime,
vaDouslnstuments have been developedfor measuringQoL, mosr of which addresssatis-
faction with life in generaland with varioustife domains(e.g.,Lehman. t9S3; Otjver. 1991;
Lower, 1999; Hansson,1999).
By contrast. the therapeulic relationship' appears ro be neglecred in psychialrjc research
even though i( is central 1()the practiceof psychiarry.In Reud's terms,ir is ..rhe vehicle of
successin therapy" and it has beenexlensivelystudiedin psycholhe.alJy€ver since Freud
higblighled the speciil rclarionshipbetweenthe clienr and therapisi.A posirjve therapeutic
relaiionship has been consistenrly found to predicl increased treatmenl adherenceand a bener
outcone acrossdifferenr forms of psychorherap)(Horvath & Syrnonds.l99l: Horaarh &
Greenberg,1994;Alexander& Coffey, 1997).Nume.ouspapershave beenpublishedwhich

atlest to rhe significance of rhe therapcotic relationship as a principal p'redicbr and 'n? cental
non+pecific clement in psychotherapy(for a review see Horvath & Luborsky' 1993)'
ln;sychilrric ca.e, a similar linding has been replicatedamong adultswith s€verenental
illnesi across a nurnber of differcnt settings: in complex bospital lreatnent (Brdker et al'
l q q 5 r .i n d a y h o \ p i r d rl e u l m ( n l , P n e b e& C r u y t e r ' . l o ' J 4 ,r n d i n a c o m m J t u qc a r e\ e t l i n E
{Fra;k & G;nderson,1990;Priebe& Gruyters,1993:Gehrs& Goering l99'l: Solomon?'al
1995r Gaston €r al. 1998). As in psychotherapv the therapeutic relationship displavs
prediclive validily in a psychialric context and appeds lo be a very effective element of
ireatment *t'ictr is mosl tikely used as a means of delivering other treament components
(e.g.,pharnacotherapy,seeWeiss el41 1997).
iroweve.. ttrere ar" no estultistred ..rirods for assessing tle the'apeutic relationship in
psychiatric seltings. Psychiarry ha^s,for the most pan, emploved neasure's dev€loped for
psychothenpy, bur rnodelsof psychothempydo not applv to dyadic telationshipsin psv-
;hiatry which are difiercnl fiom, and more complex tha.. those in conventionalpsvcbo-
therajy setrings. Measures thri have been consFucted expticitlv for use in a psvcluatric
context(i.e.,fout are extremelyrcductionisticand shon, withou! anv r€ validation(Clarkin
etdl. 1987: Stark ctdl. 1992i Priebe & Gruvters. 1993; Klittenberg, 1998) Despitethese
methodological limitalions, global assessmentsof the therapeutic relationship have demon-
strated preJictive lali lity among those wilh severe mental illness. While research thus tar on
the role of the tberapeutic relationship has been conducted more from tbe persp€ctive of the
health professional(panicularly irs ulility in predicting individul outcomc)'lhe therapeuttc
rela&rnship is also of considerableimponance on an individuil level, i c , as subjectiYelv
assessed by the pauent.
As far as we are awre. the therapeuticrelationshiph$ nol b€en systematicallystudied
in relalion to QoL. Whcreas qudlity of lif€ has been emploved primdrilv as 'm outcone
variable. the therapeulicrelationshipis viewed more as a mediating iaclor raher tnan an
outcomecriterion in its own righi. Thus, $e questionsthat led to the presentstudy wer€:
L how de ihe therapeulicrelalionsbipand QoL related in schizophJenia palients?
IL is lhe relationship difterent in shorllerm and long_t€rm treatment situations beca'lse the-
therapeutic relationship is supposed to be different (i.e.. in relation to goals and pace of
treatment. adoption of a shon or long'term perspective)?
The present study was an exploratory one tha! inYestigrled one's p€rceptionof the
therapeutic re)ationship and whether il wa5 associated with satisfaclion with other relation-
ships in one's life and ovenll satisfactionwilh life

Two gioups of subjecls meetiDg ICD 10 criteria for a diagnosis of schizophreniawere
cornp-",t. ne fir.t udrnission sample was a group of 90 schizopkenia patients admilted ro
a psichirtric hospitalfor the ftst tine in heir life. 51 of whom were followed up 9 months
afterdischarge(Rdder-Wanner& Pnebe. 1998a& b) The long-tenn sanple was a suogroup
of lhe BerI;Deinstitutionalisation studv (Priebeet al 1996i Hoffmann er al 1997; Kaiser
pr dl. 1998):the key inclusion crirerionfor this gloup was a continuoushospitalisationof a1
lelsr 6 nonths. Th; averagecumulativeduration of hospitrlisationsof this samplewas 9 8
( a 10.3) years.176patientsin the long-ierm group were assessed at baselinewhile in hospital
ilt t1"." ils patieniswtro were assessed, 168 sale clear, unequivocal answenro all questions
. .,- :.--.. -----:--si ;E


of interesthere (i.e-, in relation to QoL and thcrapeuticrelarionships).113 of this long-tem

group werefollowed up on averageone-and-a-halfyears late., 98 of whom gaveunequivocal
answersto all questionsofinlerest and were in someform of rreatmenrso that ihe questions
about the therapeuticrelarionshipapplied.4l ofthese had beendischargedby the follow-up


Subjecdvequality of Life was assessed using the cerman version of rhe Lancashireeuality
of Life Profile which was developedby Oiiver (O]i!er. 1991; Oliver .r al. I 99?i priebe €r al.
I995). The queslionDaire permits an evaluationof rhepaiients objeclivecircurn(ances,rheir
subjective satisfaclion wjth nine specinc life domains and their general lile saiisfaction.
Subjectiveratirgs are taken on 7-point scaiesfor sarisfactionwith life as a whole and with
eighrlife domains(1 = couldn'lbe worsei 7 = couldn i be betier).The meansof satisfaction
with life as a whole and the eight domainswere rakenas indicaiors for subjecrivequalil], of

A modified versjon of the Helping Alliance Scale (HAS: Priebe & Gruyters, 1993) was
usedwhich focussedon the therapeuticrelatiorship(s)peninenl in one's rrearmentsjtuation.
Three itens which pertain to therapeuticrelariorships( Do you believe you are receiving
the nght aeahent/care for you?", "Does your therapis/casenanager,/keyworkerundcr
stand you and is h€/she engaged jn your treatmenvcare?" and "Do t ou feel respecred and
well regarded?")were summedto !'ield d indicatorofone's relnrionshipwirh one s pnmary
therapist.typically a keyworker. Each item \ras raredon an ll-point visual analoguescale.
wher€ 0 - not at all and 10 = yes entirely.
Psychopathologylas obsener rated using rhe l8-item version of the B.ief psychiarric
R a u n r S c a l er B P R S :O v e r a l l& c u r h a m . l o h ) , .


Demographicdatafor the two sampleswas collated(seeTable 1). The long,term samplewas

sigdficandy older than rhe first-adnission sample (t: 10.7, p < 0.001) while rhere werc
s i g n r l i c a n dm y o r ef e m r l e . r n r h ef i r \ r . a d m i s \ r osna m p t e , l ? = 1 9 . 0 . p< 0 0 0 2 t .M e a nB p R S
tolal scoreat initial assessment was comparableacrossrbe two groupsbur was sipificantty
lowe. ir the tusr-admissiongroup at follow-up (1 = 6.6, p < 0.001). Mean subjecrjvequality
of life scoreswere also comparedacrossthe two groups:there was no significaDtdifference
beiween tbe groups al eirher poinr of dme.
Pearso, coneiations were calcularedto examinethe relalionshipbetweenthe rheraDeuric
r e l a l i o n ' h r pa n d r h e r e l e \ d r l i r e d o m d r n \ s d l i \ f a c u o nw i d r r r i e n d .a n d I t e a s d s h o r ; , a n d
overall satisfactionscore of dE LQLP. Following re.ommendationsby Kaiser er al. 099?l
and Piebe et al. (1999), partial conelations were obtained ro conlrol for rhe influence of
In the first-adrnitted group, as may be seer from Table 2, only one colrelalion was
significanr thal betweenlife as a wbole and the therapeuiicrelationshipal baseljne.lf we
Chamcr€nstio of tht 6a1.dftision .nd loqlrm snple

FiHl-admisrion smple Lotr8-te.n smple Stalistics (dl)

(N=90) (N= 168)

t lU6) = lA1
< 0.002
CDnulatiYehosp (ted)

! (91)= 6.6 <0.001

focus on the QoL sum score.itmay be seenthat therewas no significantassociationbetween

the therapeuticrelationshipand QoL in this sanple at eitber assessmenl.
In the long-tenn group, quite a differen! patt€m emerged. As may be seen from Table 3.
there were sig ficant correlations between tberapeutic relationships and the 2 domalns
and |he overall score at blselin€. At follow-up. aI of the conelationsremainedsubstantial
and stathtically significant.Tuming our attention !o the QoL sum score. the reladonship
between the theraoeuticrehtionshiD and overall satisfactionshowed an increasefrorn the
inidal to |he follow-up assessnent(the lbllow-up period \ras one and a half years).Panial
conelanons controllins for psychopathologyate shown in bracketsin Trble L The colTe-
ladons eitherremain€dthe sameor were slightly lower but remainedstatisticallysignilicad.
A facbr analysiswas also conductedusing the QoL domain scoresand the surn scoreof
therapeuticrelatiooships.wltle the lherapeuticrelationshjpwas a separatefactor at two
poims in the lirst-admissiongroup, this wa5 'ot a consistentresull in the long-term group.


The nain finding of this sludy was a signiticant associationbetween global assessmen$
of qualiiy of life and the herapeutic relationship in long-lerm schizophreniapatients.
indicaling lhat there .re generalised faclors influencing appraisats of both constructs. This
was an exploratorystudyand ir is not lnown to what exlent thesefindingscan be generalised
!o other samples.It is perhapsusefulto keep in rnjnd so'ne methodologicrl limitntionsof this

Ttbk 2
Co.rclaiioro lEieen $€mp€uli. rclrrionshipd md etisfac-
tion *iLh lr'€nd!,lite 6 r shole and ore..ll €Lisfacton itr
fiBt.dd,ftd pldctrtr

Friends Lil€ as a whole Sm score

0.09 0.12


Tlble 3
CoffelatioN bet*&n th€.ap€uli. relationshipa rnd
stisfaction *ith bends, lift s a whole rnd oremll
satis{s.lion in lona'1€rm patien6 .nd (in bEclerg
alte. conlrotlin8lor p.y.hopathologlt

ld Arsessme.r 0.40*+ 033'*

(0.13!.) (0.33!)

(.0.21r) (0.38**) (0.55'+)


study.Firstiy. asthiswas a cross-sectionalstudywith repeatedmeasures,lherewas no contrel

orer eventsthat occuned in betweenthe lwo atsessmentsHoweler, the resuks seem sub
stantialtaking into accounttha( the corrctalionswere consistcntlydifferenl at two pomts ol
rime in the 1wo samples.
Secondly,the netbod used to assessthe therapeuticrelationshipwas nol as elaborateas
that employcdlo assessQoL. Although the numberofi|ems was similar in boib assessmenls.
ftey w;re more lentatively applied with respect to the therapeutic relalionship. Inaddition.
only thc patients' appraisatof QoL and the therapeuticrela{ionshipwere assessedan'i not
the therdpisl-/observerperspectives. Intcresiingly, in psvchotherapv research the pcrspective
of tbe p;tient has the strongest predictive power, foltowed bl the observer and lastl] the
therapist. Il is not kno*n whether this finding also bolds in the'apeulic rntcmctronsrn

From rhe findings reported herein, il would appear that the tbetapeutic relationship. aiier a
while. is embeddedin an overall appraisalolone's whole iife situation.This was.l'owever'
onty the casein the long te.m and not in the first'admissionsample lt i s conceilahle thal the
therapeuticrelationshipbccomesan imponant part of dav'to day life for thosein long-letn
care ;itualions and is not separateany morc as it appearsit is in a first-admi ssion sanple lt is
plausible that the therapeuticrelationshipmoves inlo the quality of life arena and that x
isviewe<lby rhe sameglobal tendenciesas one'slife generallv on the other haod,n miJht be
to discriminatc between djfferent
argued that long-term schimphrenia palients lose the abiliry--.
domainsin their appraisalof life circumstancesduc b cognitiverigjditv' an arguncnt thal rs
not supported by the QoL ratings obtained in this study which do differentiate bclween lite

Al€matiyely, cognitive dissonance theory mav account for the differences in how o'e
p€rceivesthe therapeulicrelatiotship and ote's QoL depe'ding o. siageof illness-Accord-
ing to this theory,peopledo not tend ro hold inconsistentfeelings,betiefsor attitudesfor very
Iong. Rather.they lend to harmonisetheir liews aboutthe world so thai they are consstent
ln the presenr context. we would expect thai a negative assessment of the therapeutic
relationshipwould not exist alongsidea positiveappraisalofone's life and 'tce wrsa Hence.
when the iherapeutic relationship be.omes more imponant or features mo'e in one's life
(cither positively or negatively).il will not be possibleto make very difierent appraisals
for the iwo areas.Holvever,this dissonancebetweenappraisalswouid be more lik€l) if lhe
ercnts in questjon ue onl-r relevanl in the shon-retm as $r!h the nrsl_admissrcnsamlle

because,according1olhis theory,it is lessproblematicto have dissonanc€betweenatrituder

in the shon-krm lhan in the long-term.
Finllly, one could speculalethat the therapcuticrelationshipdevelopsin a similar way to
otber relationshipsin one s lile (e.g.,with friendt. Underlying rhis explanarionis the notion
that similar paliems ofbehaviour characterise differcnt relationships that one has, an ide! rhat
has its roots in psychodynanxctheory. Although thcre may be sone divergencefrorn one's
lypical posidon in a relatiorship when a new reladonsbipis fonned, ir rnay be that these
rehlionships are subsequendy subject to the same patterns as previous other relationships.
Enpirical researchfrom a systenic p€rspective(Priebe,1989;Priebe& Haug. 1992;Priebe
& Pornmerien.1992) p.ovides some suppon for the idea that therapeuticpolential may be
assessed, in pad, by exploring how the the.apeuticrelalionshipis similar to or differs from
relationshipswith significut olhers,an areaof researchthar wananrsfurthe. atrenrion.
In conclusbn, both qualiiy of life and the thenpeutic relalionship ue imponanl conslrucrs
which may overlap dependingon the sampleand the trealmentsirualion,an associationthar
is not auributableto the infiuenceof psychoparholosy.With referencero possiblepractical
implicationsof ihese findings,it is possiblerhat,in long-r€nn samples.interventionsin QoL
or changesin the therapeuticrelarionshipwill have an influenceon eachother.Moreoyer, it
may be thar neitherwiil be as nexible al1eryearsofillness and trearmenlas fiey were earlier
in ihe illnesstrajeclory.Conceivably,lf interventbns to improve th€ rherapculicrel.rtionship
are inlroduced, one's perceptionof the therapeulicrelationshipmight nol change if it is
viewed predonjnantly in fie context of one's life overall.
The therapeudc relationship is probably linked to how menral heallh ser.r'ices are
perceived.On a speculatjvenote,jf it is more flenble early on in lreahent (when a patienl
f|st presents),rhrs would be the tilne to iniiuence i1 in a positive dnection. ff it is viewed
negatively and this perception remains for many ye[s, it may be rnuch more difficult to
changewhich clearly will affect therapeutic€ilbciiveDess.However, theseare na$ralistic
studjes and the ways in which the therapeuticrelationsbip can be inltuenced are not yet


ALEXANDER. L.B. & COFFEY. DS (1997) U.derskndiq rhe rhdp€uric Elario.ship Curtdt Opinion i^
Psrchiatn, l0: 233-234
BRoKjR, M., RoIIRTCHT, Ii. & PRIEBE, S. (r95) rnirar alse$ment oI hospitd rcaime.r by patients wilh
puanoid shircphrcnia: A Iredictr of outcone. Pq.rilrD Rueatch, Sa:T-81.
CLARKIN. J F , IIIIRT, S W & CRILLY. J.L. ( 1937)The.apeuricaltimce dd hospikt reahenr odcome. Haspital
.tu:l Connunit! Prrchiltlr, 38: 811-815.
FRANK. A F. & CTNDERSON, I G ( 1990)Th€ .ole or the rhe.apcuticauirce in fie lreatme.t oi schizophEniz
Archier oJAe^ PqchiotrJ, 47: 224-236
techque, ed rhcn interacrionsin pr€dicdry outcone ol behavioural,cogninrc, md b.icf dyomic ftelopy
Psr.hoth.rapy, a 194 2n9.
GEHRS,M & COERING,P. (1994)Tbe r.lrtionshipb€tweenthe eorkin8 rllirnce dd rthrbilirabonourcones
of shllzophntua. Pi).hdrr.@l RchdbiLnotioaJortul, la: 43 54.
HANSSON, L. (l99) Adrancinglheconcepbof qualityof lile ln s. Priette,
t P.r.Oliver& W. Kaisq (E.L\.)
ol LrJeMlI Mentul Health Crre (pp. 19-23) Pcten6.ld, lI(: Wri8haon Bionedicrl PublisbinS.
IIoFFMANN. K., PRI9BE,S.,ISERMANN,M. & TCATSER, W (r99?) Quarityor tif.. needsand dsessnent

of ftarmcni in long bft hosPitalisedpatien6: Par. Il or the Be.lin Dei itotionalnaliun studv (in Geman)
Pry^.hiatril.he Prck,a 27.1 226.
HoRv;TH, a o. & sYMoNDs. B.D. (r 991)Relarionbei*eenwor*r! allianceandoulcom€rn tsvchoth'Fpv:
A nch-analysj. .lautul of Cohtultiq PqchaLoqr.34 139-]19
HoRVATH- AO. & LUBORSKY.L (1993)The rolc ol the iheaPeuricallian& in PsvcholbqaryJoumt 'J
CotL\uhi4ond ClinicolPq.holrS). 6l: 561_5?3
HORIATH, A.O & GREENBERC,L S ( 1994)7IE Wo.t&n4Alliec.: Th.oa Rex'z'T' 'n? Pta'ltt Ne* York:


HtxLEY. P (1997) Proliles .f subj{rive qualig of lde in schizoPhEnc rn and oul_patrenlsanptes
PrlcAizh 153 166
riAIsEi. w., HoFFNIANN.L, ISERMAMi M & PRIEBE.s (1998)Prt,enoses of crrtsrvss anddishdles
ovd t$o yes: P.n l1l of iheBdlm dcinslituriomlis,tion sbdv (in Gcmran)Ps)'n 'ir'l'l'n;r,25:6?-?l
KLNI(INBEiG, w.D. O98) Thc bclpinsaltirn.oincasenanagene for honelossperonswrh seveEnentil
illness. Cann^i1! Mental H.altt JoltMl, 34: 569 513
LEHMAN, A.F ( r9$j nmewell beingof chrcnic menralPanenL: Archi.s ol cene'at Pstthiatry. '440: 369 3?5
LO'iER, G. ( 1999)Concepl\oi qua1iryof lifd in mcnt.l heaxh cm ln S Pnebc,l P J Oli v€r & w Kaiser (Eds)
Quali\ oJlifc old Mlnbl Aealr, Cur€ (pp r - I 8) PeEslield. UK: wrishcon Roncdical Puhlhhrg
OLnfuR Lrj tj9gll Tn" .o"ial c,re didi; tlel€lopnmt of a quahv of lile pronle for use in connunn]
swices ior the nenbnt il. So.iol Wo* .tu1 So.ial Sden . R.rien. 3: 5-45
oLNER. ., P.J..HLrXL€Y.P . PRIEBEs. & KAISER.w ( 199?)Measun4ihequahl of lile ol sereElvncnralJv
itl peopleusins tbe Lhcashnc Qualily df Life Pro6le. Sd.i'l PJ-lclvtry & Psfchiai' Elia'hiologl, 3l:

O\ERALL, t. & CORHAM, D.R. (1962) Tbc bnef psychiaric raring scale Prl./',logi"l RdPort, 10 ?99-312
PRIEBE.S (l9s9)\'vEwsoi a dreraFulicslslcn ptedicrourcone? An enpincll studywiihdcpr6sive
Darient. l'anrlr Pr,..rr. 28: 349 355
PRIEBE. S. & HAiG, H.J. {1992) Inreactional panemin slep deptirarim tbenpv: An enpiriql studt Jou'ur
oJ lienau a Menal Dk.ak. 140: 59-60
PRIEBE-S & POMIiEREN. W (1992) Tn€ lnerapeuiicsveen iewed bv delEsilc in parisls a'd oulcom€
An expandedstudy FamilJ Prdess, 3l: 431439.
PRIEBE.S & ORTJYTERS, T ( r99l) The rclc of fte neldng alliancein pslchiadc conmuniiv ce: A Fosp'cuve
s$Ay l.urul oJNenousant Mtndl DNdtsP,131:5s2_557
PRIEBE. s. & GRIrYTERS, T. (1994) Paliens and aresven inidal assessmcnls oi dav hospihl teatn€nr
dd coune of synPtom ConPr?hzLir. Pslchiatry, 35: 234 214
PRIERE.S., GRUYiF,RS.T , i{En\lzE, M., HOFFMANN. C. & JriKEL. A. ( 1t95) Subj€lire cnbia lor cvalualio'
ofpsychialiccde:Methodslorassessmehtin Esearchandrcltinecd ainGemas) PtlrhiatnschePru;l 22

PRTEBE- S.- HoFFMANN. K., rsERL4ANN,M. & KAISER w (1996) chdaclc.isticsor lotsrem
hospitallsd parienl\: Pd I of the Berljn deinslitutionalisalionstudv ain Geman) /t:v'ltidfir'l1d Prdn 23
PRIEBE.S.-I{(IXLEY. P.. (NICm. S & EVANS,S (1999)Amlicaton andresullsol rhe MzncheecrShon
A$eshent ol Ouality of Lire {M'IJ]'|SA).tntPmtionl Jormt .f so.ial l'!t.hizt.^ 45: 1 12
RODER-WANI\ -U. & PRJEBE,S. (1993a)hrsr admnd scbizoplnenicpaticnts The burdcnof the dieaso
ai. Cemanl. Pr.rtid,rir.r. Prdi!, 25: t9-24
RbDiR WAN\ER. U.-U & PRIEBE s. ( 1998b)obicctve and strbc.rire qualnv of lfc of ri^t admil!€dromen
and nen *i$ schjzophrmra.E! rcp?u Archive\ .J Pn.hi.t. t- anA Clinicat Neuros.i.nce. Aa 254 528
SOLoMoN, P., DRAINE, J. & DILANEY, M.A (1995) The workins alliancc and consuncr casenanasdnent
htffil ol Me^tat H.aLth Adninifiotun, 22: 126 134
STARK. F.M.. LErIJANDOWSKI. C & BUCHXREMER, G. (1992)ThcrapisrPatienttelarionshrFas a prcdicror
of thecoDseof schizophrenic illne$ Erryea Pry.rurry. 7: 161-169
wErss, M., cAsTor_,L., FRoPsT.A. '''IISEBORD. S & ZICHERMAN.v. ( 199?)Th€roleof lbe alt'bcc in $.
Dhamatolic rEahenr of delr.sstor. Jormol af Psr.hidtry. 58: 196 204

RosDarie Mcc.b.. Ph.D : Resemh Feuos, & Sbfan Priebe,MD: Proicssoroi sdral & Connuniiy Psvchrab'1,
Depatuent of Psychiaq, St Bdholorcw's & rhe Rord kndon schml oi MediciDe& Denlisbt, Acadenjc Unn,
EastHan Menorial H.sp'hl, t .don E? 8Q . UK

*od"-"**. "-r.hiatric clidc, EnsrvonBelsnmn Ktidknn, rn.tcr Auc59,61,D,14430pdsdm.

**"..., "f seid Pslcniar.!,FEi€ univ€nitiit Berrin,prarmcnarcele. D l4o5oBe.ljn,
to RosemrieMccab€