Episode II: Understanding
and preventing relapse after
first episode psychosis
Assoc. Professor John Gleeson
Department of Psychology, The University of Melbourne &
Northwestern Mental Health
Overview
The onset phase of psychosis
Why psychological treatments in FEP?
Treatment targets in psychosis?
What are the foundations of
psychological treatment in FEP?
Some outstanding clinical and research
issues for EI
The onset phase of psychosis
The confusing, fragmenting, and traumatic
nature of symptoms
Highly developmentally dynamic and
sensitive period
High risk phase
Life interrupted….
The onset phase of psychosis
First
treatment
First-episode acute
phase
Untreated
psychosis
Symptom severity
Prodrome Early recovery
Late
recovery
Time
First treatment
Duration of
Untreated
psychosis (DUP)
Response time
Symptom severity
Remission
Time
Why psychological
treatments?
Are not biological models and
treatments the critical factors?
Limitations of biological treatments
10-30% fail to remit by 12 months (Emsley et al.
2007))
Up to 40% may only partially recover
The problem of relapse
Adherence rates (Lieberman et al., 2005; Meltzer 2006)
Side-effects
Secondary morbidity and co-morbidity
Unemployment and functioning (Killackey et al.)
01_Kraepelin.gif
Emil Kraepelin Eugen Bleuler Kurt Schneider
Stress Vulnerability Model of
Psychosis
Genetic & Early Risk Factors
Genetic mechanisms
Neurodevelopmental disorder
Environmental risk factors
Substance Vulnerability Stress
Abuse breakdown in automatic cognitive processes
Life Events
anomalous experiences
related to onset of psychosis
Psychosis
Integrated model of psychosis: Garety et al. 2001
A specific pathway from adversity
to paranoia? (Bentall 2008)
Social anxiety as secondary morbidity
(Birchwood et al., 2006)
79 people with FEP assessed for social anxiety
23 socially anxious (29%) and 56 non-anxious
– Not predicted by psychotic symptoms, or premorbid
functioning
Assessed on measures of shame/stigma of
psychosis and perceived social status
– controlled for depression, psychotic symptoms and
general psychopathology.
Participants with social anxiety experienced
greater shame with their diagnosis
Saw themselves as socially marginalized with
low social status
Post-psychotic depression
Birchwood et al. 2000 (n = 78)
36% developed post psychotic depression
(PPD) 12 months post acute phase
54% of patients with PPD suicidal
18% had made suicidal plans
4% had undertaken a suicide attempt
36% of those with PPD had persistent
sense of hopelessness
50% of FEP developed PPD
n = 26
n= 13
From Birchwood et al., 2000 (N = 78)
35 FEP patients
– 80% described being traumatized by
psychosis
– 38% symptomatic for PTSD
Associated with involuntary hospitalization
– 31% said they had attempted suicide
May be linked with PTSD symptoms
Implications for psychological
treatments
Developmentally sensitive phase
The role of emotion, cognitive biases and
appraisal in mediating acute symptoms
The high risks of secondary psychological
morbidity
– The role of appraisal in mediating secondary
morbidity
The risks of further setbacks or relapses
The possibilities for prevention…
What are the treatment
targets in psychosis?
What are the treatment foci of CBT for psychosis
Prevention or delay of onset
Acute phase
Treatment of persistent positive symptoms
Psychological recovery
Cognitive Remediation
Adherence with medication
Cannabis and psychosis
Relapse prevention
Vocational recovery
The evidence for CBT for psychosis
(Wykes et al., 2007)
34 studies included
Overall beneficial effects for the target
symptom (effect size = 0.400)
Including effects for:
– Positive symptoms (n = 32)
– Negative symptoms (n = 23)
– Functioning ( n = 15)
– Mood (n = 13)
– Social anxiety (n = 2)
But a cautionary note –
methodological rigour does matters!
Trials with raters aware of group
allocation had an inflated effect size
Some doubt therefore about
secondary benefits
(Tarrier et al. 2005; Wykes et al. 2007)
Matching of psychological interventions to
stage of psychosis
Integrated
therapy
CBT for acute
phase Relapse
prevention?
CBT for
ultra high
risk state
Cannabis Family based
interventions
COPE for
Time recovery Relapse prevention for
schizophrenia and bipolar
The SoCRATES trial
Tarrier et al., 2004
CBT in early psychosis
Far fewer trials
Sample sizes an issue
Effects of smaller magnitude
Difficulties showing significant treatment
effects compared to control interventions
Difficulties showing sustained effects of
specific interventions
Treatment as usual often of a high
standard
The importance of treatment
context
Can specific psychological
interventions improve upon
specialist first episode care?
The example of relapse…
Episode II: Prevention of relapse
following early psychosis
John Gleeson
Darryl Wade
Sue Cotton
Mario Alvarez
Donna Gee
Tracey Pearce
David Castle
Belinda Newman
Daniela Siliotacopolous
Pat McGorry
A Lilly Mac Initiative of the ‘Psychosocial Domain’
Supported by an independent research grant from Eli Lilly
Rationale
1. Relapse is common:
– 70-82% of FEP patients relapse by 5 years
(Robinson et al., 2005)
– Distressing for patient and family
– Relapse and risk of persistent symptoms (Wiersma
et al., 1998)
– Treatment costs (Almond et al., 2004)
2. Prevention may reshape the trajectory of the
illness
– Comparisons with contemporary specialist FEP care
needed
Aims
To develop and evaluate the effectiveness
of a combined individual and family
psychosocial treatment designed to
minimize the rate of, and maximize time
to, psychotic relapse (positive symptoms),
following a first episode of psychosis in
young people aged 15-25 years.
Hypothesis
Primary:
– Remitted FEP patients randomized to a multi-modal targeted
RPT + TAU will have:
– 1) a lower rate of relapse
– 2) a longer time to relapse,
– compared with remitted FEP patients who randomized to TAU in
a specialist FEP program at 7, 12, 18, 24 and 30 months follow-
up
Secondary:
– Remitted FEP patients randomized to RPT + TAU will show
improved:
– 1) medication adherence, psychosocial functioning, and quality
of life compared to TAU group
– 2) Families receiving RPT would have improved: appraisals of
stressors related to caregiving; expressed emotion; and
psychological morbidity
Design
A randomized trial, with independent rater
blind to treatment condition, of a targeted,
multi-modal relapse prevention treatment
versus treatment as usual in a specialist
first-episode service.
30-month follow-up, post baseline
Recruitment and randomization
Recruitment Nov. 2003 to June 2005
Baseline completed before randomization
RPT
– Change in case-manager to research therapists
– Therapy within a 7-month window
– Frequency matched to standard EPPIC guidelines
Treatment as usual
– Guidelines documented
– Continue case-management as usual
– Psychosocial interventions available
Measures at Baseline and Follow-up Time-points
Domain Time 1 Time 2 Tiime 3
T me 3 Tiime 4
T me 4 Tiime 5
T me 5 Tiime 6
T me 6
Baseline 7 Months 12 Montths 18 Montths 24 Montths 30 Montths
12 Mon hs 18 Mon hs 24 Mon hs 30 Mon hs
Diagnosis SCID I &
II
Pre-morbid IQ
Estimate
DUP
Premorbid
adjustment
Psychopathology
Psychotic UCLA Relapse criteria
Depressive
Insight and
Adherence
Treatment and side-
effects
Alcohol and
Substance Use
Functioning and
Quality of Life
Family Measures
Individual therapy 12-13 sessions Family therapy 10-15 sessions
Phase 1: Phase 1
Engaging the patient, assessing recovery and risk for Engagement and assessment
relapse. Phase 2
Phase 2: Psychoeducation regarding psychosis
Agenda setting - summarized in a letter and relapse
Phase 3:
Psychoeducation focused on the risk of relapse its
prevention
Early warning signs and relapse plan
Phase 5: Phase 4
Optional modules for non-adherence, substance Role of family in recovery
abuse, coping with stress, co-morbid anxiety and
depression.
Phase 5
Needs-based phase which included
additional CBT interventions for
specific problems
Review and termination, booster sessions
Structure of RPT
Results
Demographics (n = 81)
Variable Descriptor Statistics
Gender Males 63.0% (n = 51)
Age M = 20.11 years (SD = 3.05)
Marital Status Never Married 95.1% (n = 77)
Married/Defacto 4.9% (n = 4)
Living Arrangement Parents 76.5% (n = 62)
Siblings 60.5% (n = 49)
Country of Birth Australia 88.9% (n = 72)
Still Attending School 29.6% (n = 24)
Employment Status Unemployed 43.2% (n = 35)
Full-time Paid Work 12.3% (n = 10)
Part-time Paid Work 4.9% (n = 4)
Casual Work 14.8% (n = 12)
DSM-IV Diagnoses (N = 81)
SCID Diagnosis Axis I - Schizophreniform Disorder 9 (11%)
Psychotic Disorders Schizophrenia 27 (33%)
Bipolar Disorder 4 (5%)
Schizoaffective 4 (5%)
Major Depressive Disorder 9 (11%)
with psychotic features
Psychotic Disorder NOS 24 (30)%
Other 4 (5%)
Comorbid Depression Depression/Dysthymia 38 (50%)
SCID Axis I Diagnosis Cannabis 42 (52%)
of Substance Abuse Alcohol 22 (27%)
and/or Dependencea Amphetamines 16 (20%)
Hallucinogens 14 (17%)
Polydrug dependence 2 (2%)
Other 13 (16%)
SCID Axis II Diagnosis Borderline Personality 6 (7%)
Disorder Antisocial + (8 subthreshold)
Personality Disorder 8 (19%)
Psychopathlogy (N = 81)
BPRS Total Score M = 34.78 (SD = 7.40)
Psychotic Subscale 1 M = 5.53 (SD = 1.66)
Psychotic Subscale 2 M = 3.96 (SD = 1.29)
SANS Summary score M = 4.61 (SD = 3.51)
M = 13.59 (SD =
Composite score
11.46)
Affective Flattening or Blunting M = 5.25 (SD = 7.15)
Alogia M = 2.35 (SD = 2.91)
Avolition or Apathy M = 3.59 (SD = 4.37)
Anhedonia or Asociality M = 5.00 (SD = 5.77)
Attention M = 13.00 (SD = 2.72)
MADRS Total Score M = 10.42 (SD = 9.13)
SOFAS Total Score M = 63.17 (SD =
15.89)
Results: Time 2
The relapse rate was significantly lower in the
therapy condition (5.3%) compared to treatment
as usual (21.8%) (p = 0.042)
Time to relapse was significantly longer for the
relapse therapy condition (p = 0.03).
The number needed to treat was 6 to prevent 1
relapse over 7 months.
No differences on any secondary outcome
measures
Gleeson et al., 2009 Journal of Clinical Psychiatry
Family outcome
The RPT group demonstrated significantly
greater reductions in appraisal of Negative
Symptoms compared with the TAU group (ECI)
The RPT group had significantly higher mean
scores on the Positive Personal Experiences
subscale and Total Positive Score compared to
the TAU group (ECI)
Main effects for time on EE (FQ)
No effects for psychological morbidity (GHQ 28)
Gleeson et al., in press Journal of Clinical Psychiatry
Foundations of CBT interventions FEP
1. Comprehensive psychosocial, diagnostic, and risk assessment
2. Collaborative, flexible engagement process
3. Development of individualized formulation for target problems
4. Sensitivity to development stage
5. Sensitivity to phase of disorder
6. Normalizing and interactive approach to psychoeducation
7. The responsivity principle
8. Flexibility with respect to involvement of family
9. Selective use of specific, targeted interventions
10. Embedded within comprehensive, team-based youth friendly
service
Some important questions for CBT
in FEP
How can we improve upon:
– Relapse prevention
– Substance abuse
– Depression and suicide
– Anxiety
– Functional impairments & quality of life
– Engagement over the long-term?
– Outcomes for high risk sub-groups (e.g., co-
morbid personality disorders)
Study Limitations and strenths
Limitations
– Long-term follow-up data not yet analyzed
– Cannot separate the contributions of
individual and family components
– Diagnosis breakdown
Strengths
– High quality comparison
– Randomization successful
– Management of fidelity
Psychosocial treatment, antipsychotic postponement, and
low does medication strategies in first-episode psychosis…
Bola, Lehtinen, Culberg & Ciompi, 2009
Soteria, USA and Switzerland
Parachute Project
Need-Adapted Treatment
Future Directions
When (and for whom) might psychological
treatments be enough to minimize
relapse?
Using interactive technology as a
supplementary engagement tool.