• Aucun résultat trouvé

The effect of communication skills training on residents' physiological arousal in a breaking bad news simulated task

N/A
N/A
Protected

Academic year: 2021

Partager "The effect of communication skills training on residents' physiological arousal in a breaking bad news simulated task"

Copied!
8
0
0

Texte intégral

(1)

Medical

Education

The

effect

of

communication

skills

training

on

residents’

physiological

arousal

in

a

breaking

bad

news

simulated

task

Julie

Meunier

a,b,c

,

Isabelle

Merckaert

a,c

,

Yves

Libert

c

,

Nicole

Delvaux

a,d

,

Anne-Marie

Etienne

e

,

Aurore

Lie´nard

a,c

,

Isabelle

Bragard

e

,

Serge

Marchal

b

,

Christine

Reynaert

f

,

Jean-Louis

Slachmuylder

b

,

Darius

Razavi

a,b,c,

*

a

Universite´ LibredeBruxelles,Brussels,Belgium

b

Psycho-oncologyCenter,Brussels,Belgium

c

JulesBordetInstitute,Brussels,Belgium

d

HoˆpitalUniversitaireErasme,Brussels,Belgium

e

Universite´ deLie`ge,Lie`ge,Belgium

f

Universite´ CatholiquedeLouvain,Louvain-la-Neuve,Belgium

1. Introduction

Improvingphysicians’breakingbadnews(BBN)skillshasbeen recognizedasessential.PoorBBNmayhaveanegativeimpacton patients’ satisfaction withcare[1],adherence totreatment [2], decisionsabouttreatmentoptions[3]andpsychological adjust-ment[4].ReviewsaboutBBNhavebeenpublishedandthereisnow awideconsensusthatBBNrequiresspecificcommunicationskills. In order to appropriately break bad news, physicians need to mastercommunication skills promoting patients’ expressionof concernsbutalsotobeabletoidentifycuesasregardspatients’ needsandexpectationsinordertotailorinformationtransmission step-by-step.BBNisanonlinear,unscriptedandhighlycomplex

processbothcognitivelyandemotionallyforwhichphysiciansare notsufficientlytrained[5–8].

AlthoughreviewsonBBNhaveinsistedonthestressfulnessof theBBNtask,toourknowledge,onlyfourstudiesuntilnowhave investigated physicians’ physiological and psychological stress responsesduringBBN[9–12].Inthefirststudy,medicalstudents wererandomlyassignedtoabadnewsdeliverytask,agoodnews delivery taskor a controltask(readingmagazines). Thisstudy, involvingalimitednumberofsubjects,showedthatbothbadnews and good news delivery produced significant increases in self-reported distress and cardiovascular responses (heart rate and bloodpressure)comparedwiththecontroltask[10].Thesecond studyfoundananticipatorystressresponseamongsecondyear medical students to a simulated bad news consultation on cardiovascularmeasures(systolicbloodpressureandheartrate) andsubjectivestressmeasures(globallyassessedstressandstate anxiety),comparedwithpost-taskmeasures[12].Thethirdstudy foundhighercardiovascularresponsesinthebadnewsscenario

*Correspondingauthorat:Universite´ LibredeBruxelles,Av.Roosevelt,50-CP 191,B-1050Bruxelles,Belgium.Tel.:+3226504581/2631;fax:+3226502209.

E-mailaddress:drazavi@ulb.ac.be(D.Razavi). ARTICLE INFO

Articlehistory:

Received7February2012

Receivedinrevisedform17April2013 Accepted27April2013

Keywords:

Communicationskills Training

Physiologicalarousal Breakingbadnews Cancer

Psychology

ABSTRACT

Objective:Breakingbadnews(BBN)isacomplextaskwhichinvolvesdealingcognitivelywithdifferent relevantdimensionsandachallengingtaskwhichinvolvesdealingwithintenseemotionalcontents.No studyhoweverhasyetassessedinarandomizedcontrolledtrialdesigntheeffectofacommunication skillstrainingonresidents’physiologicalarousalduringaBBNtask.

Methods:Residents’physiologicalarousalwasmeasured,inarandomizedcontrolledtrialdesign,by heartrateandsalivarycortisolbefore,duringandafteraBBNsimulatedtask.

Results:Ninety-eight residents wereincluded. MANOVAshowed significant group-by-timeeffects. Trainedresidents’meanheartratelevelsremainedelevatedaftertrainingandcortisolareasunderthe curveincreasedaftertrainingcomparedtountrainedresidents.

Conclusion:Communicationskillstraininghasaneffectonresidents’physiologicalarousal.Residents’ self-efficacyandcommunicationskillsimprovementsinaBBNsimulatedtaskareassociatedwithan elevatedphysiologicalarousal,whichbecomesproportionaltothecomplexityofthetaskandreflectsa betterengagementandperformance.

Practiceimplications: Residentsshouldbeinformedthat,toperformatask,theyneedtoengageinthe taskwithaphysiologicalarousalproportionaltothecomplexityofthistask.Communicationskills trainingshouldbeadapted.

ß2013ElsevierIrelandLtd.Allrightsreserved. ContentslistsavailableatSciVerseScienceDirect

Patient

Education

and

Counseling

j ou rna l hom e pa ge :ww w. e l s e v i e r. c om/ l o ca t e / pa t e duc ou

0738-3991/$–seefrontmatterß2013ElsevierIrelandLtd.Allrightsreserved.

(2)

relativetothegoodnewsone[9].Perceivedstress,psychological distress and poor communication were not associated with increased cardiovascular responses in the bad news scenario contrarytodoctors’inexperienceandfatigue.Thefourthstudyina medical student population showed that BBN consultations provoked elevated heart rate responses compared to history takingconsultations[11].

Thesefourstudiesreportedincreasedcardiovascularresponses ofdoctorswhenfacedwithBBN.Theseheightenedcardiovascular responses however were not systematically associated with subjectivestressmeasures[9].Itisthereforeessentialtoconsider thesubjectivequalityofresponseswhenstudyingtheimpactofa taskon physiological measures[13,14]. Physiological measures alonedonotallowdistinguishingphysiologicalarousalasasignof cognitiveandemotionaleffort(investmentofresourcesinorderto perform a task) or of cognitive and emotional overload (and thereforeofstress)[15].ThisgoesinlinewithGaillarddistinction betweenmentalloadandstresswherementalloadmanifestsitself asatemporarynormalmentaleffort(ahealthycopingstrategy) whereas stress is seen as an enhanced activation that fails to improveperformanceandtofacilitaterecovery[16].Hulsmanetal. studysuggeststhatpartoftheobservedphysiologicalresponse couldalsobeattributedtothenoveltyofthetask[11].Beyondthat, the impact of a stressor is also modulated by biological predispositions,personalitypatterns,learninghistory and avail-ablecopingresources[17,18].

In the last decades, communication skills training research programshavebeenconducted.Theseprogramshavebeenshown toimprovenotonlyphysicians’andnurses’self-efficacy (subjec-tive performance) [19] but also their communication skills (objective performance) [20–22]. No study however has yet assessed,ina randomizedcontrolledtrialdesign,theeffectofa communicationskillstrainingonresidents’physiologicalarousal inaBBNsimulatedtask.

YerkesandDodsondescribedaninverted-Urelationbetween arousal and performance for numerous tasks (letter-detection, mood priming manipulation, public speaking, etc.). Moderate physiological arousal levels may therefore result in optimal performance,whereastoolittleortoomucharousalmayresult insub-optimalperformances[23–26].Physiologicalarousallevels havealsobeenshowntoberelatedtoindividuals’appraisaloftheir abilitytoperformagiventask:whenindividualsperceivethatthey areunabletoperformacomplextask(threatappraisal),theymay experiencedifficultiesinengagingthemselvesinthetaskandtheir physiologicalarousallevelsremainlow,whereaswhenindividuals perceive that they are able to perform a complex task, they experiencelessdifficultiesinengagingthemselvesinthetaskand their physiological arousal levels remain elevated (challenge appraisal) [26,27]. Yeo and Neal [28] moreover examined the relationshipbetween motivation and performanceduring skills acquisition and reported that the relationship between effort intensityand performance increasedwith practicein the early phases of skills acquisition for tasks that involve complex information-processingdemands.

The study objective was thus to assess training effect on residents’physiological arousal(Fig.1). Theresponse measures choseninthisstudy(heartrateandsalivarycortisol)aredifferent in terms of source systems, pattern of response, latency and potential impact or correlation with central mechanisms [29]. Giventhecomplexityanddurationofthetask,itwasconsidered thatheartrateasanelectrophysiologicalmechanismandcortisol asanHPAaxisstresshormonewouldbemediatingphysiological arousal. Heart rate changes are usually reported to reflect attentional aspects of a task, such as cognitive processing of task-related information or cognitive appraisal of stressful situations[30],whereassalivarycortisolisameasureofaffective

responses to a task that is reported to be independent of the cognitivedemandsofataskandtaskengagement[31,32].Ina non-experimental task as a BBN task, it is however impossible to distinguish both responses as both cognitive processing and affectiveresponsesaresimultaneous.

Itshouldbeunderlinedthattheefficacyofthetrainingprogram assessedinthisstudyhasalreadybeenshownonresidents’ self-efficacyabout their communication,on their abilityto manage their stresstocommunicate (residents’subjectiveperformance)

[33],oncommunicationskills(residents’objectiveperformance)

[22]butnotontheirburnoutlevels[33].Itwashypothesizedthat theimprovementinsubjectiveandobjectiveperformancewould beassociatedwithanincreasedmentaleffortinvestedintheBBN simulatedtaskand consequentlywithanelevatedphysiological arousal. Trained subjects were expected to show an elevated physiologicalarousal,whichisanindicatoroftheirengagementto respondadequatelytothetaskusingnewlylearned communica-tionskills while maintainingstep-by-step attentiontothetask challenges.

2. Methods 2.1. Ethicsstatement

TheJulesBordetInstitute’sethicscommitteeapprovedofthe study.Residentshadtogivetheirwritteninformedconsent. 2.2. Subjectsandstudydesign

AllBelgianFrench-speakinghospitalswerecontactedwithan internalletterof invitationtotheir residentsworkingincancer care (n=2160). Because of the low response rate (n=41), attending physicians and heads of department (n=117) were contacted by phone. Six hundred and twenty-six residents, includingthe41potentiallyinterestedresidents,werecontacted byphone,17wereindividuallymetand24informationsessions wereorganized.

Tobeincludedinthisstudy,residentshadtoworkwithcancer patients,tospeakFrench,andtobewillingtoparticipateinthe training program and its assessment procedure. Residents participating in another communication skills training were excluded.

After the first assessment time, residents were randomly allocatedtoa40-htraining(trainedresidents)ortoawaiting-list (untrainedresidents)(Fig.2).Assessmentswerescheduledbefore

Fig.1.Trainingeffectsonresidents’performanceandphysiologicalarousalinaBBN simulatedtask.

(3)

randomizationforallresidentsand8monthslaterforbothgroups (aftertrainingfortrainedresidentsand8monthsafterbaselinefor untrained residents).Residents in thewaiting-list weretrained afterthepost-testassessment. Thesameassessment procedure wasusedatbaselineandatpost-testforbothgroups.

2.3. BBNsimulatedtask

Residents’skillsandphysiologicalarousalwereassessedina standardizedBBN simulatedtaskassimulatedtasksare avalid method to study communication style [34] and physiological arousal [10]. The simulated task consisted of a first medical encounter withan actress in which residents had todeliver a breastcancerdiagnosisandtodiscusstreatment.Thescenariowas constructed to be complex (highly difficult cognitively and emotionally).Thecomplexityofthetaskresultedfromthemedical situation(size ofthe tumorrequiring mastectomyand chemo-therapy)andtheemotionalconsequencesofthenewsonthepatient. Threeactresses were used during the study. The actresseswere trainedtoexhibitahighlevelofdistressatbothassessmentpoints when bringing up concerns about the medical and marital consequencesofthedisease.Trainingincludedpracticingthe role-playandparticipatinginregularfeedbacksessionsledbythestudy coordinators.Eachassessmentprocedurelastedfrom10amto12pm andconsistedoffourperiods(Fig.3).BeforetheBBNsimulatedtask, residentshadtocompletequestionnaires(restperiod)andtolearn thecasedescriptionofthetask(10-minpreparationperiod).TheBBN simulatedtasklasted20minandwasaudiotaped.Residentswere askedtoremainseatedatthedeskinordertominimizemovement artefacts.Afterwards,residentshadtocompletequestionnaires (30-minrecuperationperiod).Thesameassessmentprocedurewasused atbaselineandatpost-testforbothgroups.

2.4. Communicationskillstraining

The communicationskills training program includeda 30-h communication skills training module and a 10-h stress

management module. The training program was based on a previouscommunicationskillstrainingprogramwhichhasbeen tested for its efficacy[35,36]. Sessionswere spread over an 8-monthperiodandwereorganizedbimonthlyinsmallgroups(up to7participants).The30-hcommunicationskillstrainingprogram consistedofa17-hcommunicationskillstrainingfocusingon two-personinterviews,a10-hcommunicationskillstrainingfocusing on three-person interviews, and a last 3-h session promoting integration of learned skills. Among these 30h, a 1-h session focusedontheoreticalinformation.Intheothersessions,residents wereinvitedtopracticecommunicationskillsthroughpredefined role-plays focusingon BBNand role-playsbasedon theclinical problems brought up by the residents. Residents were given immediatefeedbackonthecommunicationskillsperformed.The threephasesoftheBBNprocesswereintroducedgradually[22,37]. The choiceof theskills taughtwasbased on resultsof studies havingshownthepositiveimpactofusingspecificcommunication skills on patients’ disclosure of concerns [38]. The 10-h stress management skills training included four 2.5-h sessions. This programhasbeendescribedindetailsinBragardetal.[37]. 2.5. Residents’subjectiveperformance

Self-efficacyaboutcommunicationwasassessed,duringtherest period, with a 13-item self-report questionnaire adapted from Parleetal.’sscale[39].Adaptationofthescaleconsistedintheuse ofa5-pointLikertscale(fromfeelingnotatallableto...tototally able to...) and in the addition of items. This scale assesses residents’perceptionoftheirownabilitytocommunicatewitha cancerpatientandtomanagestressduringcommunication.The adapted questionnaireshowedlow toadequateinternal consis-tencyreliabilities(Cronbach’salphascoresrangedfrom.61to.79). Stateanxietywasassessed,immediatelybeforeandaftertheBBN task,with the20-itemself-reportStateTraitAnxietyInventory– StateformYa(STAI-Ya)[40]assessingstate-anxietyatthetimeof completion;theFrench-languageversionhasbeenvalidated[41].

Satisfactionaboutperformancewasrated,duringthe recupera-tion period, on a 10-cm visual analogue scale (VAS) assessing residents’satisfactionabouttheircommunicationduringtheBBN simulatedtask.

2.6. Residents’objectiveperformance

CommunicationskillswererecordedduringtheBBNsimulated task.Audiotapesweretranscribed.Transcriptswereanalyzedby LaComm. LaComm is a Frenchcommunication contentanalysis software.Thissoftwareusesontheonehandawordcountstrategy basedoncategoriesofwordslikeProtan[42]orLinguisticInquiry Word Count [43]and on the other hand a word combination strategyliketheGeneralInquirer[44].Theaimofthissoftwareis toanalyze,utterancebyutterance,verbalcommunicationusedby identifyingutterances types and contents. Thisstudy considers only residents’ utterances types. Utterances are categorized in threemaintypes:assessment(openandopendirectivequestions), support (acknowledgement and empathy) and information (proceduralinformation,negotiationandothertypesof informa-tion) types.AmoreprecisedescriptioncanbefoundinLienard etal.[22].

2.7. Residents’physiologicalarousal

Residents’physiologicalarousalwasassessedbyheartrateand salivarycortisolduringthe2-hassessment.Thismethodhasbeen validated inotherstudies [27,45].Residentswereinstructed to abstainfromfood,alcohol,caffeine,nicotineforhalfanhourbefore theassessment,andfromexercisefor24hbeforeit.

Fig.2.Studydesign.

(4)

Heart ratewasmonitoredcontinuouslyusingan ambulatory digital holter recorder (Lifecard CF Holter Recorder, Delmar Reynolds).Therecordedheartratewastransformedininterbeat intervals(IBI)andautomaticallycorrectedforartifactsbyasoftware (HRVTools,HeartRateVariabilitySoftware,DelmarReynolds)and thenhand-corrected.Therecordingwas dividedinnineperiods: rest,preparation,BBNsimulatedtaskdividedinto4periods(0–5,5– 10,10–15and15–20min)andrecuperation(0–10,10–20and20– 30min).Meanheartratelevelswerecalculatedforeachperiodand changeswerecomputedbetweenthedifferentperiods.

Salivarycortisolwascollectedbysalivasamplesatfive time-points: 10min before the task (rest), just before the task (preparation)and0,10and30minafterthetask(recuperation). Areasundertheresponsecurve(AUCs)werecalculatedbetween thedifferenttime-pointsandfortheentire2-hassessmentusing thetrapezoidalmethodasanindicatorfortheintegratedcortisol response in the BBN simulated task [46]. As cortisol is a very sensitive responsemeasure, all potentialconfounding variables (medication,food,alcohol,caffeinetakenduringthe24hbefore theassessmentprocedureandnumberofhoursofphysicalactivity andsleepduringthe24hbeforetheassessmentprocedure)were collectedcarefullyandcontrolledfor.

2.8. Statisticalanalyses

Statistical analyses includeda comparative analysis of both groups of residents at baseline using parametric tests and nonparametric tests as appropriate (Student’s t test, Mann– Whitneytest and X2 test). To be considered for data analysis, residentshadtoattendatleast1hofstressmanagementand1hof communicationskillstraining.Thiswasdoneinordertolimitthe riskofbiasassociatedwithnon-randomlossofparticipants.We excludedresidentswhodidnotatallparticipateinthetraining programastheirdrop-outcouldnotbeattributedtothetrainingin itself.Changesinresidents’performanceandphysiologicalarousal were assessed using group-by-time Multivariate Analysis of Variance (MANOVA), General Linear Model or Generalized EstimatingEquation Poisson Regression Modelsas appropriate. Alltestsweretwo-tailed,andalphawassetat.05.Analyseswere performedwithSPSSVersion17.0forPC(SPSSInc.,Chicago,IL). 3. Results

3.1. Residents’characteristics

Fig. 4 describes the recruitment and assessment process. Principalbarrierstoparticipationwerepersonalandinstitutional

reasons,timelimitations,trainingduration,andtime-consuming assessmentprocedures.Ninety-eightresidentswereincludedin the statistical analyses. Comparison of included and excluded residentsshowednostatisticallysignificantdifferencesfor socio-demographic and socio-professional characteristics. For these characteristics,nostatisticallysignificantdifferenceswerefound at baselinebetween trained and untrainedresidents exceptfor specialty: untrained residents were more often residents in oncology (p=.028). No statistically significant group-by-time difference was found regarding the number of cancer patients treatedintheweekbeforetheassessments.Trainedresidentswere ameanof28yearsold(SD=3years)and68%werefemale.They wereonaverageintheirthirdyear(SD=1.3years)ofresidency.Six percentwereresidentsinoncology,32%ingynaecologyand62%in otherspecialties.Untrainedresidentswereameanof28yearsold (SD=2.1years) and60%werefemale.Theywereon averagein

Fig.4.Recruitmentprocedure,studydesign,trainingandassessmentprocedures.

Table1

Trainingeffectsonresidents’performanceinabreakingbadnewssimulatedtask(n=98).

Traininggroup(n=50) Controlgroup(n=48) Statisticalanalysesa

Atbaseline Aftertraining Atbaseline 8monthsafter baseline

Groupbytimeeffects

Mean SD Mean SD Mean SD Mean SD F RR p Subjectiveperformance

Self-efficacyaboutcommunication(beforethesimulatedtask) 3.0 0.4 3.4 0.5 3.1 0.5 3.2 0.6 13.29 – <0.001 Stateanxiety(beforethesimulatedtask) 45.7 7.7 43.2 6.6 44.5 9.4 42.4 8.4 0.08 – 0.782 Stateanxiety(afterthesimulatedtask) 44.8 9.4 38.1 8.1 43.2 10.2 38.7 8.4 1.34 – 0.251 Satisfactionaboutperformance(afterthesimulatedtask) 32.4 19.0 52.5 22.1 39.4 20.0 46.7 21.9 5.96 – 0.017 Objectiveperformance

Assessment(openquestionsandopendirectivequestions) 3.2 2.0 5.2 3.5 3.3 2.7 2.8 2.5 – 1.92 <0.001 Support(acknowledgementandempathy) 23.3 14.4 27.1 15.2 24.2 17.3 22.3 14.0 – 1.26 0.055 Information(proceduralinformation,

negotiationandotherinformation)

63.4 22.5 45.4 24.2 64.8 29.0 64.9 28.5 – 0.72 <0.001 Thereisonemissingvalueforsubjectiveperformancemeasures(n=97).

a

MANOVA(MultivariateAnalysisofVariance)wereusedforsubjectiveperformancemeasuresandGeneralizedEstimatingEquationPoissonRegressionModelswereused forobjectiveperformance.

(5)

theirthirdyear(SD=1.2years)ofresidency.Twenty-fivepercent wereinoncology,21%ingynaecologyand54%inotherspecialties. Trained residents took part on average in 25h of training (SD=8.1).Theyparticipated to8hof stressmanagement skills training(SD=2.4) andto17hofcommunicationskills training (SD=6.8).

3.2. Trainingeffectsonresidents’performance

AsshowninTable1andasithasalreadybeenshownelsewhere

[33], MANOVA showed significant or marginal group-by-time effectsonresidents’subjective(self-efficacyaboutcommunication and satisfaction about performance)andobjective performance

Table2

Trainingeffectsonresidents’physiologicalarousalinabreakingbadnewssimulatedtask(n=98).

Traininggroup(n=50) Controlgroup(n=48) MANOVA Atbaseline Aftertraining Atbaseline 8 months after

baseline

Group by time effects Mean SD Mean SD Mean SD Mean SD F p Meanheartrate

Levels

Rest 82.1 8.7 81.2 11.2 81.1 10.0 79.1 10.5 0.59 0.445 Preparation 84.7 9.3 83.4 11.3 83.1 11.9 81.1 12.7 0.12 0.728 Simulatedtask0–5min 95.6 13.1 92.8 16.2 93.3 17.2 83.1 14.0 9.89 0.002 Simulatedtask5–10min 85.7 10.8 85.6 13.5 85.9 13.0 79.6 11.8 11.55 0.001 Simulatedtask10–15min 83.4 9.3 82.4 11.7 83.1 10.8 78.8 10.4 4.08 0.046 Simulatedtask15–20min 82.0 9.8 82.2 11.0 81.8 10.6 77.9 10.0 7.53 0.007 Recuperation0–10min 78.2 8.2 77.9 9.8 77.4 9.2 76.0 9.6 0.73 0.396 Recuperation10–20min 76.1 7.7 75.8 10.1 75.9 8.9 74.0 9.9 1.60 0.208 Recuperation20–30min 76.1 7.6 75.9 9.9 75.8 9.2 74.4 10.0 0.73 0.395 Changes

Preparation–rest 2.8 4.4 2.5 4.6 1.8 4.2 1.7 4.2 0.04 0.835 Simulatedtask0–5–preparation 10.7 10.1 9.4 11.9 11.2 10.6 2.5 6.9 12.82 0.001 Simulatedtask5–10–task0–5 9.9 5.4 7.2 7.7 7.4 6.5 3.5 4.2 0.70 0.406 Simulatedtask10–15–task5–10 2.3 3.5 3.2 5.2 2.8 4.8 0.7 3.3 8.82 0.004 Simulatedtask15–20–task10–15 1.5 3.0 0.2 3.3 1.3 2.8 0.9 2.2 1.62 0.206 Recuperation0–10–task15–20 3.8 4.3 4.3 4.7 4.4 5.0 1.9 4.0 15.89 <0.001 Recuperation10–20–recuperation0–10 2.1 2.2 2.1 3.0 1.5 2.1 2.0 2.6 0.84 0.361 Recuperation20–30–recuperation10–20 0.1 1.8 0.2 2.3 0.3 2.3 0.4 2.8 0.24 0.625 Recuperation20–30–rest 6.1 4.2 5.6 5.1 5.8 4.8 4.8 4.5 0.14 0.713 Salivarycortisol

Areasunderthecurve

Resttoendofpreparation 32.4 22.0 44.9 28.0 31.6 14.0 32.4 18.4 5.12 0.026 Endofpreparationtoendofsimulatedtask 130.5 81.7 166.5 100.8 119.4 57.2 121.2 66.5 4.10 0.046 Endofsimulatedtasktorecuperation10min 64.4 47.5 75.3 47.1 56.4 33.2 54.9 28.6 2.16 0.145 Recuperation10torecuperation30min 105.8 73.2 131.3 77.5 95.4 53.4 99.1 48.2 2.32 0.131 Resttorecuperation30min 346.0 219.0 441.3 247.6 312.1 152.3 307.5 159.4 4.61 0.035 Thereareforseveralperiodssomemissingvalues(max=13).

(6)

(increaseinassessmentandsupportanddecreaseininformation giving).

3.3. Trainingeffectsonresidents’physiologicalarousal

The GeneralLinear Model assessing overtime and between group’schanges(PhasesbyTimebyGroup)ishighlysignificant (p<.0001;partialetasquare=.095;power=.999).Asshownin

Table2andinFig.5,MANOVAshowedsignificantgroup-by-time effectsonresidents’physiologicalarousallevels.Residents’mean heart rate levels remained elevated after training for trained residentscomparedtountrainedresidentswhosemeanheartrate levelsdecreased.Trainedresidents’meanheartratelevelswere higher during the 20-min BBN simulated task than untrained residents’ mean heart rate levels. Moreover, trained residents’ meanheartratechangeswerehigherbetweenpreparationandthe BBNsimulatedtask0–5min,betweentheBBNsimulatedtask5– 10minandtheBBNsimulatedtask10–15minandbetweenthe BBNsimulatedtask15–20minandrecuperation 0–10minthan untrained residents mean heart rate changes. Finally, trained residents’ cortisol AUCs were higher between rest and end of preparation,betweenendofpreparationandendofBBNsimulated task,andbetween restandrecuperation30minthanuntrained residents’cortisolAUCs.Nogroupbytimechangewasobservedin termsoftheconfoundingvariablestestedinthestudy(medication, food,alcohol,caffeinetakenduringthe24hbeforetheassessment procedure and number of hours of physical activity and sleep during the 24h before the assessment procedure) except for nicotineconsumption(p=0.037).Fourtrainedresidentsand2 un-trainedresidentsbegansmokingbetweenT1andT2.

4. Discussionandconclusion 4.1. Discussion

Thisisthefirststudyassessing,inarandomizedcontrolledtrial design,theeffectof communicationskillstrainingon residents’ physiological arousalin a BBNsimulated. Results of this study show that a communication skills training has an effect on residents’ physiological arousal in a BBN simulated task. The hypothesis that trained residents will present an elevated physiologicalarousalwhentheyengageinandperformthetask comparedtountrainedresidentshasbeenconfirmed.

AsshowninFig.5,heartrateandcortisollevelshavedifferent patternsof changeover time. For heart rate, thereis a change (decreasedlevel)inthewaitinglistgroupbetweenpre-and post-test,butnochangeinthetraininggroup.Forcortisollevel,thereis a change(increased level)in thetraining groupbut not in the waitinglistgroup.Thepatternofchangeovertimeofheartrate andcortisollevelsshouldthusbeinterpretedseparately.

Thepatternofchangeofheartratefoundinthisstudysupports theideathatanelevatedheartratemayreflect,eitheranarousalin anunknownandstressfultestcontext(forbothgroupsofresidents at baseline) [16,47],or an arousalwhich is an indicator of an engagementtorespondadequatelytothetaskusingnewlylearned communicationskillswhilemaintainingstep-by-stepattentionto thetaskchallenges(fortrained residents).Itshouldberecalled that,beforetraining,residentshavelimitedcommunicationskills. Asitcouldbeexpected,theresultsofthisstudyshowthat,before training, an elevated heart rate in residents during thetask is relatedtotheexposuretoanunknownandstressfultask.Resultsof thisstudyshowthat,aftertraining,heartrateismoreelevatedin trained residents compared to untrained residents. This higher elevationinheartrateintrainedresidents–whichisassociated withhigherself-efficacyandsatisfactionabouttheirperformance in thetask, with less stress tocommunicate [33] and withan

improvementinresidentscommunicationskills[22]–maybean indicator,ashypothesized,oftheirengagementinperformingthe task. An additional explanation for these higher elevations in trainedresidents’heartratemaybethattrainedresidentsaremore awareofthechallengesofBBNandtheimpactforthepatientof poor communication and maytherefore be moremotivated to engagein thetask.Thelowerlevel ofelevationin heartratein untrainedresidentsmaybeexplainedbythefactthat,beingfora secondtimeexposedtoanalreadyexperimentedtask–thatthey havenotlearnedtoperformbetter–residentsdealwiththetaskas they do ‘‘usually’’. An additional explanation for this lower elevation in untrained residents’ heart rate may be residents’ habituationtothesimulatedtask[11].

Thepatternofchangeovertimeofcortisollevelsfoundinthis studysupportstheideathattheelevatedlevelsofcortisolfound after training reflect the physiological arousal related to the sustainedcognitiveand emotional activation– starting already before thetask– ofresidents and maybeanindicatoroftheir preparationand engagementtorespond adequatelytothetask whilemaintainingstep-by-stepattentiontothetaskchallenges. Thelowercortisollevelsfoundinresidents,atbaselineforboth groupsandatpost-testinuntrainedresidents,mayreflectalower level of cognitive and emotional activation because they are probablydealingwiththetaskastheydo‘‘usually’’.

4.2. Conclusions

Tosummarize,whentrainedresidentsperceivethattheyare moreabletoperformthetask(withhigherlevelofself-efficacy aboutcommunication[33]),theyengageinandperformthetask better(withimprovementsinobservedcommunicationskills[22]) andtheirphysiologicalarousallevelsareconsequentlyelevated. Theirphysiologicalarousalbecomesprobablymoreproportional tothecomplexityofthetaskandreflectscognitiveandemotional activation to engage in the task and to respond with learned communicationskillstotaskchallenges.Theseresultsaresimilar with results of other studies that have shown a link between performanceandarousal[24–27].

Thestrengthofthis studyisthat residents’performanceand physiological arousal were assessed before, during and after a standardizedBBNsimulatedtaskwithreliablemeasureslikeheart rate and salivary cortisol levels. Heart rate and cortisol were selectedbecausetheyareabletoreflectreliablychangesovertime duringagiventaskandforagiventaskrepeatedovertime(heart rateasafastresponsemeasureandcortisolasa slowresponse measure–20–40min–afterastimulation).Inparticularmeasures of cortisol levels were selected as it has been suggested that cortisollevelsarea‘‘usefulindexofsubjectivestress’’[48],allowto ‘‘assess relationship between perceived stress, activation and performance’’ [49], and are related to‘‘cognitive performance’’

[50].

Thestudyhassomelimitations.First,itshouldberecalledthat physicianswerevoluntarilyenrolledand weremainly inexperi-encedclinicians.Thisfactlimitsthegeneralizabilityofourresults. Itcouldbearguedalsothatthemotivationofthosephysicianswas highandthatthiscouldhavehadanimpactonthetrainingeffects observed. Second,this study investigated changes immediately aftertraining.Thearousallevelsfoundarethereforerelatedtoan earlylearningstage,whichstillneedsacontrolledprocessingin theexecutionof thetask.Itmaybehypothesizedthatin more advancedlearningstages–afterafewyearsofpracticeforexample –thisarousallevelwouldnotbefoundanymoregiventhefactthat a constant attentional control would not be needed anymore. Third,the extremelylow response rateto thestudy shouldbe underlined.Thismaybeduetoatoolargeandnon-personalized recruitmentprocedure.Resultofthisstudymaythusreflectresults

(7)

ofmoremotivatedandlessanxiousresidents.Fourth,thisstudy focusedonresidents’verbalcommunication.Futurestudiesshould include the assessment of non-verbal communication. Fifth, it shouldbenotedthatthefactthatresidentsfilledinquestionnaires duringtherecuperationphasemayhaveincreasedtheir autonom-icarousallevels.

4.3. Practiceimplications

Whatcouldbethepracticalimplicationsof theseresultsfor residents?Residentsconsiderusuallythattoperformacomplex task such as a BBN task (highly difficult cognitively and emotionally),theyneedtoacquireafeelingofself-efficacyabout thetaskandtolearnskillstoperformthetaskproperly.Resultsof this study confirm that residents could benefit from being informed that they also need to engage in the task with a physiologicalarousalproportionaltothecomplexityoftheBBN task.Whatcouldbethepracticalimplicationsoftheseresultsfor trainers?Trainersusuallyconsiderthattoperformacomplextask suchasaBBNtask,theyhavetodevelopresidents’feelingof self-efficacyabout thetask and toteach skills to performthe task properly.Giventheresultsofthisstudy,itmaybehypothesized thattheyshouldteachresidentstoengageactivelyinthetaskwith aphysiologicalarousalproportionaltothecomplexityoftheBBN task.Toachievethat,trainersmayuseroleplayingexerciseswith increaseddegreesofcomplexity,debriefresidentsateachstepof theserole-plays and teachresidents specificskills todeal with difficultiesmetateachofthesesteps.Thiswouldbeinteresting areas to address in future research designed to improve communicationskillstrainingefficacy[51,52].

Acknowledgements

Thisresearchprogramwassupportedbythe‘‘FondsNationalde laRechercheScientifique–SectionTe´le´vie’’ofBelgiumandbythe C.P.O.,trainingandresearchgroup(Brussels–Belgium).Thestudy sponsorshadnoroleinstudydesign,datacollection,dataanalysis, ordatainterpretation,orinthepreparation,review,orapprovalof thereport.

Theauthorshavenofinancialandpersonalrelationshipswith otherpeopleororganizationsthatcouldinappropriatelyinfluence theirwork.Principalinvestigatorhadfullaccesstoallthedatain thestudyandtakesresponsibilityfortheintegrityofthedataand theaccuracyofthedataanalysis.

References

[1]BredartA,BouleucC,DolbeaultS.Doctor–patientcommunicationand satis-factionwithcareinoncology.CurrOpinOncol2005;17:351–4.

[2]StavropoulouC.Non-adherencetomedicationanddoctor–patient relation-ship:evidencefromaEuropeansurvey.PatientEducCouns2011;83:7–13.

[3]AlbrechtTL,EgglySS,GleasonME,HarperFW,FosterTS,PetersonAM,etal. Influenceofclinicalcommunicationonpatients’decisionmakingon partici-pationinclinicaltrials.JClinOncol2008;26:2666–73.

[4]BrownVA,ParkerPA,FurberL,ThomasAL.Patientpreferencesforthedelivery ofbadnews–theexperienceofaUKCancerCentre.EurJCancerCare(Engl) 2011;20:56–61.

[5]EgglyS,PennerL,AlbrechtTL,ClineRJ,FosterT,NaughtonM,etal.Discussing badnewsintheoutpatientoncologyclinic:rethinkingcurrentcommunication guidelines.JClinOncol2006;24:716–9.

[6]FallowfieldL, Jenkins V. Communicating sad,bad, and difficultnews in medicine.Lancet2004;363:312–9.

[7]PtacekJT,EberhardtTL.Breakingbadnews.Areviewoftheliterature.JAmer MedAssoc1996;276:496–502.

[8]SweenyK,ShepperdJA.Beingthebestbearerofbadtidings.RevGenPsychol 2007;11:235–57.

[9]BrownR,DunnS,ByrnesK,Morris R,HeinrichP,ShawJ. Doctors’stress responses and poorcommunication performance in simulatedbad-news consultations.AcadMed2009;84:1595–602.

[10]CohenL, BaileWF,HenningerE,AgarwalSK,KudelkaAP,LenziR, etal. Physiologicalandpsychologicaleffectsofdeliveringmedicalnewsusinga simulatedphysician–patientscenario.JBehavMed2003;26:459–71.

[11]HulsmanRL,PrangerS,KootS,FabriekM,KaremakerJM,SmetsEM.How stressfulisdoctor–patientcommunication?Physiologicalandpsychological stress of medical students in simulated history taking and bad-news consultations. IntJPsychophysiol2010;77:26–34.

[12]vanDulmenS,TrompF,GrosfeldF,tenCateO,BensingJ.Theimpactofassessing simulatedbadnewsconsultationsonmedicalstudents’stressresponseand communicationperformance.Psychoneuroendocrinology2007;32:943–50.

[13]TomakaJ,BlascovichJ,KiblerJ,ErnstJM.Cognitiveandphysiological ante-cedentsofthreatandchallengeappraisal.JPersSocPsychol1997;73:63–72.

[14]BlascovichJ,MendesWB.Challengeandthreatappraisals:theroleofeffective cues. In:ForgasJ,editor.Feelingandthinking:theroleofaffectinsocial cognition.NewYork:CambridgeUniversityPress;2000.p.59–82.

[15]StaalMA.Stress,cognition,andhumanperformance:alitteraturereviewand conceptualframework.MoffettField,CA:NASA/TM;2004.

[16]GaillardAWK.Stress,workload,andfatigueasthreebiobehavioralstates:a generaloverview.In:HancockPA,DesmondPA,editors.Stress,workloadand fatigue.Mahwah(NJ),London:LawrenceErlbaumAssociates;2001.p.623–49.

[17]LazarusRS,FolkmanS.Stress, appraisal,and coping.NewYork:Springer PublishingCompany;1984.

[18]EverlyGS,LatingJM.Aclinicalguidetothetreatmentofthehumanstress response.In:MeichenbaumD,editor.ThePlenumseriesonstressandcoping. NewYork:KluwerAcademic/PlenumPublishers;2002.

[19]AmmentorpJ,SabroeS,KofoedPE,MainzJ.Theeffectoftrainingin commu-nicationskillsonmedicaldoctors’andnurses’self-efficacy.Arandomized controlledtrial.PatientEducCouns2007;66:270–7.

[20]BarthJ,LannenP.Efficacyofcommunicationskillstrainingcoursesin oncolo-gy:asystematicreviewandmeta-analysis.AnnOncol2011;22:1030–40.

[21]LienardA,MerckaertI,LibertY,BragardI,DelvauxN,EtienneAM,etal. Transferofcommunicationskillstotheworkplaceduringclinicalrounds: impactofaprogramforresidents.PLoSONE2010;5:e12426.

[22]LienardA,MerckaertI,LibertY,BragardI,DelvauxN,EtienneAM,etal.Isit possibletoimproveresidentsbreakingbadnewsskills?Arandomisedstudy assessingtheefficacyofacommunicationskills trainingprogram. BritJ Cancer2010;103:171–7.

[23]YerkesRM,DodsonJD.Therelationofstrengthofstimulustorapidityof habit-formation.JCompNeurolPsychol1908;18:459–82.

[24]CalabreseEJ.Convergingconcepts:adaptiveresponse,preconditioning,andthe Yerkes–DodsonLawaremanifestationsofhormesis.AgeingResRev2008;7:8–20.

[25]DiamondDM,CampbellAM,ParkCR,HalonenJ,ZoladzPR.Thetemporal dynamicsmodelofemotionalmemoryprocessing:asynthesisonthe neuro-biologicalbasisofstress-inducedamnesia,flashbulbandtraumaticmemories, andtheYerkes–Dodsonlaw.NeuralPlast2007;60803.33p..

[26]GendollaGH,RichterM,SilviaPJ.Self-focusandtaskdifficultyeffectson effort-relatedcardiovascularreactivity.Psychophysiology2008;45:653–62.

[27]MathesonK,GillR,KellyO,AnismanH.Cortisolandcardiacreactivityinthe contextofsexdiscrimination:themoderatingeffectsofmoodandperceived control.OpenPsycholJ2008;1:1–10.

[28]YeoGB,NealA.Amultilevelanalysisofeffort,practice,andperformance: effects; ofability, conscientiousness,andgoalorientation. JApplPsychol 2004;89:231–47.

[29]ClowA.Thephysiologyofstress. In:JonesF,BrightJ,editors.Stress:myth, theoryandresearch.Harlow:PearsonEducation;2001.p.47–61.

[30]LaceyBC, Lacey JL. Studies of heart rateand other bodilyprocesses in sensorimotorbehavior. In:ObristPA,BlackAH,BrenerJ,DiCaraLV,editors. Cardiovascularpsychophysiology.Chicago:AldinePublishingCo.;1974.p. 538–64.

[31]CacioppoJT,SandmanCA.Physiologicaldifferentiationofsensoryand cogni-tivetasksasafunctionofwarning,processingdemands,andreported un-pleasantness.BiolPsychol1978;6:181–92.

[32]PetersML,GodaertGLR,BallieuxRE,vanVlietM,WillemsenJJ,SweepFCGJ, etal.Cardiovascularandendocrineresponsestoexperimentalstress:effectsof mentaleffortandcontrollability.Psychoneuroendocrinology1998;23:1–17.

[33]BragardI,EtienneAM,MerckaertI,LibertY,RazaviD.Efficacyofa communi-cationandstressmanagementtrainingonmedicalresidents’self-efficacy, stresstocommunicateandburnout:arandomizedcontrolledstudy.JHealth Psychol2010;15:1075–81.

[34]RoterDL,HallJA,KernDE,BarkerLR,ColeKA,RocaRP.Improvingphysicians interviewingskillsandreducingpatientsemotionaldistress–arandomized clinical-trial.ArchInternMed1995;155:1877–84.

[35]Delvaux N,MerckaertI, Marchal S,LibertY, Conradt S, BoniverJ, etal. Physicians’communicationwithacancerpatientandarelative:arandomized study assessing the efficacy of consolidation workshops. Cancer 2005;103:2397–411.

[36]RazaviD,MerckaertI,MarchalS,LibertY,ConradtS,BoniverJ,etal.Howto optimizephysicians’communicationskillsincancercare:resultsofa ran-domizedstudyassessingtheusefulnessofposttrainingconsolidation work-shops.JClinOncol2003;21:3141–9.

[37]BragardI,RazaviD,MarchalS,MerckaertI,DelvauxN,LibertY,etal.Teaching communicationandstressmanagementskillstojuniorphysiciansdealing with cancerpatients:aBelgianInteruniversity Curriculum. SupportCare Cancer2006;14:1–8.

[38]MaguireP,FaulknerA,BoothK,ElliottC,HillierV.Helpingcancerpatients disclosetheirconcerns.EurJCancer1996;32A:78–81.

[39]ParleM,MaguireP,HeavenC.Thedevelopmentofatrainingmodeltoimprove health professionals’ skills,self-efficacy andoutcome expectancieswhen communicatingwithcancerpatients.SocSciMed1997;44:231–40.

(8)

[40]SpielbergerCD,GoruschRR,LusheneRE,VaggPR.ManualfortheState-Trait AnxietyInventorySTAI-formY.PaoloAlto:ConsultingPsychologistsPress Inc.;1983.

[41]SpielbergerCD,Bruchon-SchweitzerM,PaulhanI.Inventaired’anxie´te´ Etat-TraitFormeY(STAY-Y).Paris:LesEditionsduCentredePsychologie Appli-que´e;1993.

[42]HogenraadR,DaubiesC,BestgenY.Unethe´orieetuneme´thodege´ne´rale d’analysetextuelleassiste´eparordinateur:leSyste`mePROTAN(PROTocol ANalyser).Louvain-la-Neuve:Universite´ CatholiquedeLouvain,Psychology Department;1995.

[43]PennebackerJW,BoothRE,FrancisME.Linguisticinquiryandwordcount: LIWC2007–operator’smanual.Austin,TX:LIWC.net;2007.

[44]StoneP,DunphyD,SmithM,OgilvieD.Thegeneralinquirer:acomputer approachtocontentanalysis.Cambridge,MA:MITPress;1966.

[45]RichterM,FriedrichA,GendollaGH.Taskdifficultyeffectsoncardiacactivity. Psychophysiology2008;45:869–75.

[46]PruessnerJC,KirschbaumC,MeinlschmidG,HellhammerDH.Twoformulas forcomputationoftheareaunderthecurve representmeasuresoftotal hormoneconcentrationversustime-dependentchange. Psychoneuroendocri-nology2003;28:916–31.

[47]LazarusRS.Theorybasedstressmeasurement.PsycholInquiry1990;1:3–13.

[48]BohnenN,NicolsonN,SulonJ,JollesJ.Copingstyle,traitanxietyandcortisol reactivityduringmentalstress.JPsychosomRes1991;35:141–7.

[49]al’AbsiM,HugdahlK,LovalloWR.Adrenocorticalstressresponsesandaltered workingmemoryperformance.Psychophysiology2002;39:95–9.

[50]VedharaK,HydeJ,Gilchrist ID,TytherleighM,PlummerS.Acutestress, memory,attentionandcortisol.Psychoneuroendocrinology2000;25:535–49.

[51]MerckaertI,LibertY,RazaviD.Communicationskillstrainingincancercare: whereareweandwherearewegoing?CurrOpinOncol2005;17:319–30.

[52]UitterhoeveRJ,BensingJM,GrolRP,DemulderPHM,VanachterbergT.The effectofcommunicationskillstrainingonpatientoutcomesincancercare:a systematicreviewoftheliterature.EurJCancerCare(Engl)2010;19:442–57.

Figure

Fig. 2. Study design.
Fig. 4 describes the recruitment and assessment process.
Fig. 5. Heart rate levels (mean heart beats per minute) and salivary cortisol levels (areas under the response curve).

Références

Documents relatifs

Pour organiser un atelier philosophique en classe, à partir de ces grandes images, découvrez des outils pédagogiques sur www.pommedapi.com?. À ton avis, un ami ça sert

Par ailleurs, même si un programme de 8 semaines d’exercice physique aérobie n’a pas permis de faire varier la concentration plasmatique des MPs de ces sujets obèses, nous avons

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des

Unwed mothers who decide to place their child for adoption were compared, both pre- and post-parturition, with those who chose to keep their child.. Dependent variables included

Related to the previous result, another finding of the present research unveils that, in order to move the change project forward, the managers observed juggle with the

Dans ce court paragraphe, non seulement les diverses reprises permettent d’éviter la répétition, mais aussi l’emploi de termes plus spécifiques contribue à ajouter des

Ainsi, après l’analyse des coûts, les entreprises mettent en place des budgets prévisionnels et réels pour contrôler les réalisations et mesurer les écarts ; c’est pourquoi

An assessment of the autonomic nervous system in the electrohypersensitive population : a heart rate variability and skin conductance study... An assessment of the autonomic