Transcript Insight & Adherence in Schizophrenia

Predictors of adherence and
outcome in schizophrenia
Richard Drake,
Senior Lecturer in Adult Psychiatry,
University of Manchester
What I’m Going to Tell You
• Non-adherence predicts poor outcome
• Determinants of adherence and
concordance
– Insight relates to degree of concordance
• Poor insight predicts poor outcomes
– it may predict poor outcome better than
adherence
• Insight and adherence can be changed
Predictors of Outcome in
Schizophrenia
• Demographic
– Sex, age
• Historical
– Premorbid adjustment
– DUP, Course of illness
– Symptoms
• Behavioural
– EE, Substance Misuse
– Adherence, Engagement
Predictors of Outcome in
Schizophrenia
• Demographic
– Sex, age
• Historical
– Premorbid adjustment
– DUP, Course of illness
– Symptoms
• Behavioural
– EE, Substance Misuse
– Adherence, Engagement
Drug
Placebo
16.2%
57.6%
Davis & Andrukaitis 1986
J Clin Psychopharmacol
Davis & Andrukaitis 1986
J Clin Psychopharmacol
Adherence and Chronic Illness
• Stopping APS suddenly:
46% relapse in 6/12 if stable
56% in two years
• Stopping gradually:
50% over 2 years
Viguera et al 1997
100
90
80
70
60
50
Survival (%)
40
30
20
10
0
m
o
6m
o
12
m
o
18
m
o
24
m
o
30
m
o
36
m
o
42
m
o
48
m
o
54
m
o
60
m
o
0
Robinson D et al, Arch Gen Psych, 1999
Gaebel W et al 2002 Sz Res
Gaebel W et al 2002 Sz Res
MESIFOS STUDY
Wunderink L et al 2005, 2007
Determinants of Poor Adherence
• Maybe
– isolation, youth and being male
– substance abuse
– more side effects or their subjective
experience
– more frequent doses
– Better or worse cognitive function
• Or not, for most of the above
More recent studies
• Internal rather than external locus of
control
• High EE family
• Poor therapeutic alliance
• Negative attitudes
• Poor insight
Influence
of Others
Belief in
Prevention
Medication
Affinity
Vauth R et al, 2004 Psychiatry Res
Length of
Illness
Neurocognition
No
symptoms
Meaningful
Work
Influence
of Others
Age
Medication
Affinity
Belief in
Prevention
Vauth R et al, 2004 Psychiatry Res
Insight and medication attitudes
Day, J., et al. 2005 Arch Gen Psych 62 717-24
Insight and medication attitudes
Day, J., et al. 2005 Arch Gen Psych 62 717-24
The Construct of Insight
• Insight said to have different dimensions
– E.g. recognising sx, illness, its social
consequences, & need for Rx; attributing sx
to illness; “hypothetical contradiction”
• Insight appears to have cultural,
symptomatic and neuropsychological
determinants
Insight and Outcome
• In chronic samples predicts relapse,
readmission, symptoms, objective QoL,
adherence, other outcomes
– perhaps including violence (esp. in short term
or in forensic populations).
– Perhaps not including engagement?
• Insight during the process of relapse
predicts readmission.
• However, these samples select for poor IS
Why does RLS predict relapse?
• Re-Labelling Symptoms predicts relapse
– Unlike accepting NFT, awareness of illness
•
•
•
•
Chance finding?
Related to identifying relapse?
Related to substance misuse, EE?
Related to adherence?
Recent FE naturalistic studies
• Poor adherence in FE
– definitions vary but most involve stopping
>7/7 as a minimum
– Many of these will stop altogether for some
time
– 33% over any 6 months
– 42-59% at some point over 1-5years
Recent FE naturalistic studies
• All multivariate analyses of predictors of
adherence have global insight as
significant
– Except Coldham et al, 2002
• IS predicts irregular or non-adherence
• Disappears when adjust for PM function, age,
cannabis at 1y, family support
HBM and Adherence in a Trial
• Perkins et al. 2006: FE trial of APS in 254
• Small effect of objective SE predicted nonadherence
• Lack of benefit too
• Negative attitudes to medication did not
• Awareness of benefits of medication in
reducing sx. & NFT both predicted good
adherence
IS and Adherence in a Trial
• McEvoy et al. 2006: FE trial of 251 OLZ v
HPL
• ITAQ scale predicted time to nonadherence
• Type of medication did not
FE cohorts and trials
• Global IS predicts adherence even after
attitudes to medication
• In detail, awareness of past and future
symptom reduction predicts continued
adherence
• Negativity about medication or perceived
SE not predictive; possibly objective SE
• Other aspects of insight inconsistent
Insight may be improved
• Specific IS-focussed CBT may
improve it (Turkington et al, 2002, 2006; Rathod et
al 2005)
– This form of CBT protected against
depression
• Focussed CBT may alter attitudes to
illness and self (Gumley et al, 2005)
• General CBT for psychosis may not
improve IS or suicidality (Tarrier et al, 2006)
Adherence may be improved
• Psychoeducation about medication almost
never effective
• Multi-modal interventions appear more
likely to be effective
– certainly more fashionable
Insight and medication attitudes
Day, J., et al. 2005 Arch Gen Psych 62 717-24
Adherence may be improved
• Combine education about meds & disorder
with:
– Behavioural approaches (techniques and
skills, reminders, reinforcement)
– Work on therapeutic relationship
– Use other interpersonal relationships, feelings
of loyalty etc. (“affective approaches”)
• May be effective individually, in groups or
via services (e.g. supported housing)
Summary
• Certain attitudes may be more
predictive of future adherence and
relapse than current behaviour
– There maybe a very disadvantaged
subgroup
– Sealing over may predict
disengagement
Summary
• Insight is associated with suicidality
– If you then become depressed and hopeless
because you’re ashamed and overwhelmed
– It doesn’t help if you abuse substances
• Insight leads to adherence and
improvement, which prevents the above
• Specific CBT might improve IS without
depression
• Multimodal interventions improve
adherence
NFM
Hosp
Well
SE
AIS
P
+ve correlation
D
DSH
-ve correlation