Page 1
...where“treatment-resistant”pa1entsbecomepeopletakingchargeoftheirlives. www.austenriggs.org
StatesofMindPrecedingaNearLethalSuicideA7empt:AMixedMethodsStudy
JaneG.Tillman,PhD,ABPPJenniferL.Stevens,PhD,ABPP
TheEriksonIns=tuteoftheAustenRiggsCenter
1
Page 2
ResearchTeam• PrincipalInvesAgator:JaneG.Tillman,PhD
• Co-InvesAgators:A.JillClemence,PhD;JenniferL.Stevens,PhD;Ka=eC.Lewis,PhD
• Consultants:DavidE.Reiss,MD;ChristopherJ.Hopwood,PhD;RobynA.Cree,PhD;JohnT.Maltsberger,MD;HerbertHendin,MD
2
Page 3
Funding• Interna=onalPsychoanaly=cAssocia=on
• AmericanPsychoanaly=cAssocia=on
• JeffreyGu=nFundoftheNewHampshireCharitableFounda=on
3
Page 4
SuicideintheUSA• SuicideintheUSAincreasedby24%between1999-2014.
• Bostwicketal.,(2016)foundthat59.3%ofpeopledyingbysuicidediedonthefirsta_empt.
• Over60%offirstsuicidea_emptsoccurwithinthefirstyearoftheonsetofsuicidalidea=on(Kessleretal.,1999).
• Followingasuicidea_empt,thelife=meriskofcompletedsuicideiselevatedforaslongas37yearsfollowingtheindexa_empt(Suominen,2004).
4
Page 6
Suicidevs.BreastCancer:IncidenceandFunding
6
494169
231840
43298 40290
0
100000
200000
300000
400000
500000
600000
Suicide BreastCancer
SuicidevsBreastCancerOccurrence/DeathsChart(2015)
Occurrence Deaths
$46,000,000
$674,000,000
0
100000000
200000000
300000000
400000000
500000000
600000000
700000000
800000000
Suicide BreastCancer
NIHFunding(2015)
Page 7
SuicideResearch• Popula=onbasedepidemiologicalstudiestoiden=fygroupsatelevatedriskforsuicide
• Clinicalstudiesbydiagnosis(depression,PTSD,etc..)
• Lessisknownaboutprotec=vefactors
• Li_lea_en=ontotheinterac=onofriskandprotec=vefactorsinvulnerablepopula=ons
• Beyondriskandprotec=vefactors:Needtounderstandthepsychologyofthepersonwhoa_emptsorcompletessuicideandthestatesofmindprecedingasuicidea_empt
7
Page 8
TheoreAcalModelsofSuicideOveradozenpsychologicalmodelsofsuicidewithempiricalsupportbutli_leintegra=on
• Stress-DiathesisModel(Mannetal.,1999)
• InterpersonalTheoryofSuicide(Joiner,2005)
• Cogni=veTheory(WenzelandBeck,2008)
• 3STTheory;“idea=on-to-ac=on”(KlonskyandMay,2014)
• Psychodynamictheoryofsuicide(Maltsberger,2004)
8
Page 9
CommonFindingsAcrossTheories• Suicideisacomplexbehaviorwithmul=factorialcauses
• Thereareproximalfactorsanddistalfactorsinvolved(state-trait)
• Stressisinvolvedincrea=ngthecondi=onsforsuicide,withacutestressasaprecipitantforsuicidalbehavior
• Cannotexplainwhyextremestressisassociatedwithsuicidebehaviorinsomebutnotallexposedindividuals
9
Page 10
RiskFactors• DemographicRisk:age,gender,raceandethnicity,familyhistoryofsuicide
• NeuropsychiatricRisk:diagnoses,neurosignaturesofrisk
• PsychologicalRisk:impulsiveness,hos=lity,aggression,“psychache.”
• DevelopmentalRisk:earlyadversity,sexualabuse,physicalabuse
• Nega=velifeevents:trauma,violence,loss,betrayal,bullying,poverty
10
Page 11
ProtecAveFactorsWhattheyare:
• Psychologicalprotec=vefactors:reasonsforliving,resilience,hardiness,meaninginlife,hope,self-efficacy
• Accesstoadequatetreatment
• Socialsupport
Whytheyareimportant:
Inthecontextofanincreasingnumberofriskfactors,someofwhicharefixed,protec=vefactorsbecomeextremelyimportant.
11
Page 12
AimsofStudy• Understandthe“career”ofsuicideinAustenRiggspa=ents:psychiatrichistory,suicidehistory,psychologicalcharacteris=cs
• Howdoriskandprotec=vefactorsinteractinthisgroupofhigh-riskpa=entswithsuicidestatus(i.e.non-a_empters,ideators,a_empters,near-lethala_empters,numberoflife=mea_empts)?
• Whatcanpa=entstellusabouttheirstateofmindimmediatelyprecedinganear-lethalsuicidea_empt?
• Canweheara“deepstory”abouttheprocessofsuicidethatimprovesourcapacitytorecognizethetransi=onfromchronictoacuterisk?
12
Page 13
MixedMethodsApproach• Researchques=oncallsforbothquan=ta=veandqualita=vedata
• Onriskandprotec=vefactorsmeasures,whatdoestheRiggspa=entpopula=onlooklikeincomparisontoothersamples?Howdoriskandprotec=vemeasuresinteract?(quanAtaAve)
• Whatisthe“deepstory”thatsurvivorsofnear-lethalsuicidea_emptstellabouttheirstateofmindpriortothesuicidea_empt?(qualitaAve)
13
Page 14
ResearchContext• AustenRiggsCenter:Approximately45%ofpa=entsadmi_edduringtheperiodofthestudyhadahistoryofapriorsuicidea_empt
• Fromthefollow-alongstudyweknowourpa=entshavecomplexpsychiatricdifficul=es
• Extensiveco-morbidity;difficultyengaginginoutpa=entpsychotherapy;mul=plemedica=ontrialsorotherbiologicalapproachesthathavenotbeeneffec=ve
14
Page 15
HumanSubjectsResearchatRiggs
• Anxietyandconcerninclinicalstaffaboutaskingpar=cipantstoreconstructtheirstateofmindpriortothesuicidea_empt
• Concernabouthowtalkingsoexplicitlyanddirectlyaboutsuicideinaresearchcontextmightaffectclinicalcareandmilieu
• Almostyear-longdiscussionandnego=a=onwithclinicalleadershipaboutethicsanddesignofresearchproject
• Substrate:Trauma=zedorganiza=onbasedonpa=entsuicides
15
Page 17
ObjecAve
• Todeterminetheincrementalassocia=onofreasonsforlivingtothelifeAmenumberofsuicidea7emptsinrela=ontootherknownriskandprotec=vefactors
17
Page 18
Hypotheses1.ResilienceandReasonsForLivingwillbeposi=velycorrelated,demonstra=ngconvergentvalidity
2.Higherscoresonprotec=vefactorsmeasures(CDRISCandRFLI)willbeassociatedwithfewerlife=mesuicidea_empts,eveninthepresenceofriskfactorsofimpulsiveness,generalpsychiatricseverity(GAFscores),andnega=velifeeventsofsexualand/orphysicalabuse
18
Page 19
MeasuresProtec=veFactorMeasures
• ReasonsforLivingInventory(RFLI)
• A48-itemself-reportmeasureasking“whataresomereasonsyouwouldNOTkillyourself?”InLinehan’sfactoranalysistherearesixfactors.
• Connor-DavidsonResilienceScale(CDRISC)
• 25-itemself-reportmeasure,developedforuseinpopula=onswithPTSD
RiskFactorMeasures
• Barra7ImpulsivenessScale-11(BIS)
• 30-itemself-reportmeasure.Threesubscales:cogni=ve,behavioral,motoricimpulsiveness
19
Page 20
ParAcipants298 patients admitted during study period (38 months)
131 (53%) participated
119 questionnaires only 12 questionnaires + psychodynamic research interviews
1 excluded from analysis
50 met exclusion criteria
248 eligible participants
117 (47%) refused or did not respond
20
Page 21
ParAcipantCharacterisAcsTable 1
Demographic, Suicide, and Psychiatric History Descriptive Variables
Demographics N Overall Min/Max Gender 131
Male
49 (37.4%) Female
82 (62.6%)
Age (years), mean ± SD 131 33.06 ± 12.74 18-71 Psychiatric and suicide history Suicide Attempters
71 (54.2 %)
Lifetime number of attempts, mean ± SD
2.65 ± 2.02 1-12 Non-Attempters
60 (45.8%)
History of physical abuse 128 40 (30.5%) History of sexual abuse 130 43 (32.8%) Age of first psychiatric contact, mean ± SD 131 15.53 ± 9.94 2-70
Age of first onset of suicidal ideation, mean ± SD 107 19.97 ± 13.44 5-70 Age of first suicide attempt, mean ± SD 71 24.52 ±13.70 8-70 Prior Psychiatric Hospital Admissions, mean ± SD 129 3.75 ± 5.27 0-35 Presence of Axis I Diagnosis 131 131 (100%)
Presence of Axis II Diagnosis 131 122 (93.0%) Current Severity (GAF score), mean ± SD 131 45.85 ± 6.92 25-70
21
Page 22
FactoranalysisofRFLI
• Weiden=fied5andnot6factorsontheRFLI
• Linehangroupiden=fiedfearofconsequencesofsuicideandfearofsuicideactasseparatefactors.
• Wefoundasingle“fearfactor”
22
Page 23
CorrelaAonsbetweenReasonsforLivingCompositesandCriterionVariables
RFLI Composite CDRISC LSA BIS-11 GAF SA PA
Survival and Coping Beliefs .60* -.29* .02 .07 -.01 -.07
Responsibility to Family .08 -.08 .08 -.03 -.12 -.22
Child-Related Concerns .33* .04 -.12 -.11 .05 .11
Fear -.13 -.33* .14 .18 -.14 -.20
Moral Objections .17 -.23 .07 .09 .11 -.04
CDRISC = Connor Davidson Resilience Scale; LSA = Lifetime number of suicide attempts; BIS-11 = Barratt Impulsiveness Scale; GAF = Global Assessment of Functioning; SA = History of sexual abuse; PA = History of physical abuse
* p< .01
23
Page 24
BetacoefficientsfromamodelwithriskfactorsandselectedReasonsforLivingcompositespredicAng
lifeAmesuicidea7empts
Variable 𝛽 p
History of Sexual Abuse .34 .001*
History of Physical Abuse -.06 .53
Impulsiveness -.01 .94
Current Severity (GAF) -.09 .26
Survival and Coping Beliefs -.23 .007*
Fear -.22 .01*p<.01
24
Page 25
Conclusion
25
•Theprotec=vefactorofSurvivalandCopingBeliefsmaybeanimportantbarriertorepeatedsuicidea_empts
•RFLImaybeasuicidespecificresiliencemeasure
•Weneedabe_erunderstandingofhowprotec=vefactorsaredevelopedinearlylifeandwhatsupportsorerodesprotec=vefactors
Page 27
ObjecAvePa=entswithahistoryofsuicidalidea=onora_empts,especiallyiftheyhaveseriouspsychopathologywithrepeatedhospitaliza=ons,areburdenedbyongoingriskforsuicide.Westudiedthishigh-riskgrouptoassesstheirpsychologicalstatusfollowingtheirmostrecentsuicidea_empt,incontrasttoequallyillpa=entswithoutasuicidehistory.Further,amongsuicidalpa=ents,wecomparedthosewithonlyidea=on,withanon-medicallyserioussuicidea_emptandwithmedicallyserioussuicidea_empts.Wealsoreportonthedevelopmentofanewmeasureofpsychicpain.
27
Page 28
Hypotheses• Groupsofpar=cipantswithdifferentsuicidehistorieswillbesignificantlydifferentfromoneanotheronriskandprotec=vefactormeasures
• non-a_empters<ideators<a_empters<medicallyseriousa_emptersonriskfactormeasures
• non-a_empters>ideators>a_empters>medicallyseriousa_emptersonprotec=vefactormeasures
28
Page 29
MeasuresProtec=veFactorMeasures
• ReasonsforLivingInventory(RFLI)
• A48-itemself-reportmeasureasking“whataresomereasonsyouwouldNOTkillyourself?”InLinehan’sfactoranalysistherearesixfactors.
• Connor-DavidsonResilienceScale(CDRISC)
• 25-itemself-reportmeasure,developedforuseinpopula=onswithPTSD
RiskFactorMeasures
• Barra7ImpulsivenessScale-11(BIS)
• 30-itemself-reportmeasure.Threesubscales:cogni=ve,behavioral,motoricimpulsiveness
• PsychicPainScale(PPS)
• Newlydevelopedaspartofthisstudy,20-itemself-reportmeasurewithitemsconstructedbasedonMaltsberger’stheoryofthesuicideprocessinvolving“affec=vedeluge”andasubjec=vefeelingofa“lossofcontrol.”
29
Page 30
30
Variable N Overall Non-suicidality (N = 24)
Suicidal Ideation(N=36)
Non-Severe Suicide
Attempts (N = 30)
Medically Serious Suicide
Attempts (N = 41)
p†
RFLI, mean ± SD 127 3.24 ± 0.90 4.28 ± 0.85 3.26 ± 0.79 2.72 ± 0.83 2.99 ± 0.56 <0.001
CDRISC, mean ± SD 131 49.95 ± 18.5 59.81 ± 18.23 49.61 ± 17.45 48.01 ± 20.58 45.90 ± 16.43 0.026
PPS, mean ± SD 129 61.31 ± 17.21 45.21 ± 12.73 60.49 ± 15.99 66.12 ± 18.02 67.65 ± 14.14 <0.001
BIS-11, mean ± SD 130 69.40 ± 12.31 69.85 ± 10.64 68.38 ± 14.84 67.15 ± 11.07 71.69 ± 11.59 0.444
Prior Admissions, mean ± SD
129 3.75 ± 5.27 1.69 ± 4.20 2.14 ± 2.28 4.0 ± 4.15 6.14 ± 7.20 <0.001
DescripAonofsamplebysuicidality
† P-value for ANOVA (continuous variables); 𝛘2 or Fisher’s exact test (categorical variables)RFLI, Reasons for Living Inventory; CDRISC, Connor Davidson Resilience Scale; PPS, Psychic Pain Scale, BIS-11, Barratt Impulsiveness Scale-11
Page 31
31
4.35 4.28 4.23 4.18
3.623.26
2.992.72
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
PsychiatricControlGroup
NonAttempters UniversityStudents
AfricanAm.WomenwithSuicideHistory
Suicideattempters
Ideators Near-LethalAttempters
Attempters
Osman,Kopperetal(1999)
Tillman,Clemenceetal.
(2017)
Grewwal-Sandhu(2009)
Street,J.C. Osman,Kopperetal(1999)
Tillman,Clemenceetal.
(2017)
Tillman,Clemenceetal.
(2017)
Tillman,Clemenceetal.
(2017)
ReasonsforLivingInventoryMeanItemScore
Page 32
32
80.4076.70
68.0059.81 59.50
49.80 49.61 48.01 45.90
GeneralPopulation
USVeteransofIraqand
Afghanistan
PsychiatricOutpatients
NonAttempters OEFVeteranswithPTSD
AbstinentSubstance
Abuserswithahistoryofsuicide
attempts
Ideators Attempters Near-LethalAttempters
Connor,Davidson(2003)
Youssef, Greenetal.(2013)
Connor,Davidson(2003)
Tillman,Clemenceetal.
(2017)
Pietzak,Johnsonetal.(2010)
Royetal.(2007) Tillman,Clemenceetal.
(2017)
Tillman,Clemenceetal.
(2017)
Tillman,Clemenceetal.
(2017)
ConnorDavidsonResilienceScale
Page 33
33
59.40 61.21
67.15 68.38 69.85 71.47 71.6975.10
82.10
0
10
20
30
40
50
60
70
80
90
AdultHealthy
Control
Undergraduate
students
Attempters Ideators NonAttempters Femaleswith
BulimiaNervosa
Near-Lethal
Attempters
ManicAdolescents BipolarPatients
withMedically
SevereSuicide
Attempts
Malloy-Diniz,
Fuentes,etal.
(2007)
Snorrason,Smariet
al.(2011)
Tillman,Clemence
etal.(2017)
Tillman,Clemence
etal.(2017)
Tillman,Clemence
etal.(2017)
Merlotti,Mucciet
al.(2013)
Tillman,Clemence
etal.(2017)
Nandagopal,Fleck
etal.(2011)
Swann,Dougherty,
Pazzaglia,etal.
(2005)
BarrattImpulsivenessScale-11
Page 34
34
38.5445.21
60.4966.12 67.65
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
Undergraduatestudents NonAttempters Ideators Attempters Near-LethalAttempters
Hopwood&Good(2016)
PsychicPainScale
Page 35
35
1 2 3 4 5 6 7 8 9
1. PPS
2. RFLI -0.5***
3. CDRISC -0.73*** 0.46***
4. BIS-11 0.21* 0.07 -0.25**
5. Psychiatric admissions 0.27** -0.17 -0.13 0.05
6. Child abuse 0.21* -0.18* -0.06 -0.04 0.20*
7. Adolescent abuse 0.13 -0.13 0.05 -0.02 0.26** 0.51***
8. Adult abuse 0.13 -0.08 0.11 0.00 0.35** 0.23* 0.40***
9. Age 0.02 -0.13 -0.07 -0.20* 0.11 0.09 0.06 0.12
10. Gender -0.24** 0.13 0.11 0.01 -0.27** -0.21* -0.37*** -0.31*** -0.01
CorrelaAonsofStudyVariables
* p <0.05, ** p < 0.01, *** p <0.001
Page 36
36
Variable Unadjusted OR (95% CI)
Unadjusted p-value
Adjusted OR (95% CI)
Adjusted p-value
PPS 1.06 (1.03,1.09) <0.001 1.05 (1.01,1.10) 0.009
RFLI 0.24 (0.15,0.40 <0.001 0.28 (0.15,0.52) <0.001
CDRISC 0.97 (0.95,0.99) 0.009 1.03 (0.99,1.07) 0.157
Admissions 1.24 (1.08,1.42) 0.002 1.08 (0.95,1.23) 0.247
Childhood abuse 2.14 (1.08,1.42) 0.048 0.71 (0.25,1.99) 0.514
Adolescent abuse 3.53 (1.63,7.66) 0.001 2.40 (0.79,7.30) 0.122
Adult abuse 3.92 (1.47, 10.44) 0.006 2.27 (0.65,7.95) 0.202
Age 1.03 (1.00,1.06) 0.068 1.03 (0.99,1.07) 0.087
Male 0.44 (0.21,0.91) 0.026 1.14 (0.45, 2.87) 0.782
Page 37
Results
Ahistoryofsuicidalidea=onorsuicidea_emptsisassociatedwithpropor=onallymorepsychicpainandfewercurrentreasonsforliving.Priorhistoryofabuse,impulsiveness,andgeneralresiliencewerenotsignificantlyassociatedwithsuicidalseverity.
37
Page 38
Conclusions•Itisconceivablethatpsychicpainandotherriskfactorswalka
persontotheedgeofaproverbialcliff,butreasonsforlivingmaystopapersonfromjumping.
•Forpa=entswhohavesuicidalidea=on,orhavea_emptedsuicide,andalsohaveaddi=onalriskfactorsincludingpasthospitaliza=on,treatmentsshouldincludebothunderstandingthesourcesofpsychicpainandpromo=ngindividualdiscoveryofreasonsforliving.
38
Page 39
39
PTSD and a foreshortened sense of the future and/or hopelessness about the future
Increased Psychic Pain
Childhood Abuse
Suicidal Ideation and Number of Lifetime Attempts
Decreased Reasons for Living
Page 41
QualitaAveResearchInterviews:Amorecomplex
story
41
Page 42
Semi-StructuredDynamicInterview
Semi-structuredpsychodynamicinterviewforpar=cipantswhohavesurvivedanear-lethala_emptinthetwoyearspriortoadmission.
InterviewStructure:
• Generalthoughtsandfeelingsatthe=meofthea_empt
• Acutecontextfora_emptwitha_en=ontointernalandexternalcircumstances(fantasies,dreams,stress)
• Interpersonalcontext
• Detailsofmethod
• Rescue,damage,lethalityinforma=on
• Survival
42
Page 43
InterpretaAvePhenomenologicalAnalysis(Smith,Flowers,&Larkin,2009)
• IPAisatheore=callybasedqualita=veapproachtoanalyzinginterviewdataattheleveloftheindividualinordertounderstandhowpeoplemakesenseofadefinedlifeexperience.
• Helpsresearchersgainanunderstandingofthepsychologicalprocessesandthemeslinkedtoaspecificsitua=on.
• “movingfromthepar=culartotheshared,andfromthedescrip=vetotheinterpreta=ve”(p.79).
43
Page 44
IPAProcedure1. Interviewdigitallyrecordedandtranscribed
2. Eachtranscriptreadforini=alfamiliarityandtocorrecterrorsintranscrip=on
3. Transcriptsreadasecond=meandini=althemesanddetailswerenotedalongwithdescrip=vesummarycommentsabouteachtranscript
4. Transcriptsandnotesfromstep3usedinathirdreadingtoconstructalistofemergingthemes
5. Eachresearcheddevelopsalistofthemesforeachpar=cipant’sinterviewandlooksforrecurringthemesacrossinterviews
6. Researcherscomparedalistofallthemesthateachhadindependentlyiden=fiedineachinterview.Placeofdiscordancediscussedandfurtherelaborated.Elabora=onandconsolida=onofthemeswithineachinterviewoccurred.
7. Researchersworktogethertoreconcile61thema=cunitsiden=fiedacrossalleleveninterviews,sor=ngtheunitsintocategories.Finalsortyieldedseventhema=ccontentandprocessunitsassociatedwiththestateofmindleadinguptoanearlethalsuicidea_empt.
44
Page 45
DevelopingtheDeepStoriesofSuicide
45
Page 46
46
Theme DescriptionDevelopmental
conflictsDevelopmental crises, difficulty tolerating separation and loss, conflicts with authority played out in separation struggles, childhood trauma
Character traits and vulnerabilities
Perfectionism, grandiosity, rigid character and cognition, nihilistic philosophy of life, sensitivity and reactivity to betrayal and rejection, malignant narcissism, fixed ideas, perverse relationship to death and others
Interpersonal and object relations
paradigms
Misuse of persons, self and other seen as in perverse engagement, fixed ideas about the interpersonal world, spitting with the self as particularly bad in relation to good other or other as particularly bad in relation to victimized and misunderstood good self.
Page 47
47
Theme Description
Thinking and affect Unbearable, difficult to symbolize pain that is believed will never end; nihilistic philosophy used as a defense against painful feelings; thinking problems such as subtle thought disorder, poverty of understanding, and strained logic or reality testing about the finality of death; shame and humiliation; unconscious and/or denied rage; feelings of helplessness and of being trapped.
Fantasies of death Death as a means to peaceful relief from suffering; relational fantasies of death as a means of reunion with a dead parent, death as a means of revenge or a way of evoking guilt or longed for recognition in others.
Page 48
48
Theme Description
Paradoxical nature of the immediate moment of the
suicide attempt
Dissociation, attempt paradoxically both planned and impulsive, ineffable quality of the immediate moment, paradoxical focused attention and confused thinking, internal world simultaneously numb while unbearable pain imposes the scene
Reactions to survival Loss of a fear of death, regret, relief, ambivalence about surviving; effect on relationships with family members.
Page 49
Conclusions• Mixedmethodsresearchishardtowriteup!
• Thereappearstobeagapbetweenwhatwediscoverwithourquan=ta=vemeasuresandtherichtextureofourinterviews
• Iseachsuicideasingularity,withidiosyncra=canddeeplyheldpersonalmeaningsthatmustbediscoveredandaddressed?Ifso,whataretheimplica=onsforquan=ta=veresearch?
49
Page 50
NextSteps• Preparingtodoafollow-upwithourpar=cipantstofindoutaboutsuicidalidea=onora_emptssincedischargeandtorepeattheReasonsforLivingandPsychicPainScale
• Weknowthat6ofour131par=cipantshavedied,5bysuicide(3.8-4.6%)
• Inaddi=on5ofournon-par=cipantshavediedbysuicide
50