Title: WHAT ARE THE BEST WAYS OF DEFINING REMISSION IN ANXIETY DISORDERS
1WHAT ARE THE BEST WAYS OF DEFINING REMISSION IN
ANXIETY DISORDERS?
- Borwin Bandelow
- University of Göttingen, Germany
- Dept. of Psychiatry and Psychotherapy
Dr Bandelow has been a consultant for, received
grant/research support from, and been on the
speakers/advisory board for AstraZeneca,
Bristol-Myers-Squibb, GlaxoSmithKline,
Janssen-Cilag, Lilly, Lundbeck, Pfizer, Roche,
Sanofi-Aventis, Solvay, Wyeth
2Göttingen
- 45 Nobel Prize Winners
- Nuclear Fission (Otto Hahn)
- Bell Curve (Gauss)
3Definition of clinically meaningful change
- More than 1.96 standard deviations under the mean
of the ill population
Less than 1.96 standard deviations over the mean
of the healthy population
Healthy population
Ill population
Nearer to the mean of the healthy population than
to the ill population
Requires normative data!
Healthy population
Ill population
4RELIABLE CHANGE INDEX(Jacobson Truax, 1991)
- x1 pre-treatment score
- x2 pre-treatment score
- s1 pre-treatment standard deviation
- rxx test-retest reliability of the scale
When RC 1.96 then it is probable that there was
an actual change When RC 1.96 then the
difference occured through unprecise measurement
5Remission Response - CGI
- Response CGI Improvement (CGI-I)
- much improved or
- very much improved
- Remission CGI Severity (CGI-S)
- not at all ill or
- borderline mentally ill
62 Methods for Measuring Improvement
7Response Conventional definition
- 50 reduction on a standard rating scale, e.g.
- Depression MADRAS, HAMD
- GAD HAMA
- Panic Disorder PA, PDSS
- SAD LSAS
8What is the threshold for symptomatic response
and remission?
- Randomized controlled studies
- Major depression (5 studies)
- Panic disorder (1 study)
- GAD (4 studies)
- SAD (2 studies)
- Comparison of effect sizes of CGI and gold
standard scale for these disorders (MADRS, PA,
HAM-A, and LSAS).
Bandelow, Baldwin, Dolberg, Friis Andersen,
Stein. JCP, submitted
9Corresponding CGI-I values for mean changes ()
from baseline of the total scores of the standard
scales
10Response
- 50 reduction on a standard scale may be too
conservative - Or Patients who are much improved on the CGI-I
still have substantial symptomatology
11Remission Corresponding CGI-S values for mean
total scores of the standard scales
Proposals from the literature
12Comprehensive Definition of Remission (Ballenger
et al., 1999)
- FOR SAD
- LSAS 30
- HAMA 7-10
- Sheehan Disability 1
- HAMD 7
13Comprehensive Definition of Remission (Ballenger
et al. 2001)
- For Panic Disorder
- LSAS 30
- HAMA 7-10
- Sheehan Disability 1
- HAMD 7
- Or
- PDSS 3
- HAMD 7
14Detection of Differences between Active Drug and
Placeboin Panic Disorder Trials in of all
Studies
n Number of Studies
Bandelow et al., 1995)
15Why do we need a special panic scale?
- Other scales, such as the HAMA, are not designed
for DSM/ICD panic disorder - Measuring panic attacks is not sufficient, as in
some patients severity is determined by
agoraphobic avoidance, anticipatory anxiety etc. - Panic attack frequency is not a sensitive measure
to separate active drug from placebo - One main efficacy measure needed, not a battery
of scales
16Panic and Agoraphobia Scale (PA)
- Specifically designed to monitor treatment
efficacy - 2 versions observer-rated, self-rated
- 13 item scale (0-4 Likert scale)
- 5 sub-scales
- Compatible with DSM-IV or ICD-10 Panic Disorder
With or Without Agoraphobia - Inter-rater reliability 0.78 test-retest
0.73 - Correlation with CGI 0.90
- 21 translations available
www.hogrefe.de
17PA subscales
- 1. Panic attacks
- Frequency
- Severity
- Duration
- 2. Agoraphobia
- Frequency
- Number of situations
- Importance of situations)
- 3. Anticipatory anxiety
- Frequency
- Severity
- 4. Disability
- Family
- Social
- Work
- 5. Worries about health
- Fear of having bodily disorder
- Fear of health damage during panic attacks
18Treatment Effect Sizes (Rosenthalss r)
Bandelow B et al (1998) The use of the Panic and
Agoraphobia Scale in a clinical trial. Psychiatry
Res 7743-9.
19Panic Disorder Escitalopram vs. Citalopram vs.
PlaceboBandelow Panic Agoraphobia Scale (P A)
Stahl et al., 2003
20Panic attack frequency at week 10 (LOCF)
21Difference from placebo at week 10 for PA
subscales (LOCF)
plt.05 plt.01 plt.001 plt.05 ESC vs CIT
22Sertraline vs. Paroxetine PAS Total and
Subscale Scoresat Week 12 / Endpoint (Per
Protocol)
LS means reduction based on ANCOVA model at Week
12 / Endpoint
Bandelow et al., 2004. J Clin Psychiatry
65405-413
23Escitalopram Effect Sizes in Depression and 3
Anxiety Disorders
24Panic Scales vs Panic Freqency
25CPMP Guidelines for Panic DisorderDraft
- In PD not only the recording of the frequency
and severity of panic attacks (full and limited
attacks) should be assessed but also the severity
of agoraphobic avoidance and anticipatory anxiety
should be evaluated. Therefore several scales
have been developed for measurement of specific
symptoms according to the diagnostic criteria of
PD, which may be more appropriate for use in
clinical trials, e.g. the Panic Disorder Severity
Scale (PDSS)1 and the Panic and Agoraphobia Scale
(PAS) 2, than focussing only on frequency and
severity of panic attacks. - 1Shear et al. 1997
- 2Bandelow et al. 1995
26PA (1995) vs. PDSS (1997)
- PA first ever panic scale
- PA has 13 items, PDSS has 7 items
- All PDSS items are also in the PA
- PA Higher number of items higher reliability
- PA better psychometric properties
- PA Translations into gt 20 languages
27GADSS (Shear)
- Worries
- Frequency
- Distress
- Associated Symptoms
- Frequency
- Severity/Distress
- Impairment
- Work
- Social Functioning
28Conclusions
- Defintions of response and remission should not
be based on CGI - Should be based on cut-off score on comprehensive
scale - Scales should not be a list of symptoms, but
should include measures of all domains that are
dysfunctional in a certain disorder, including
QoL measures (e.g. PA, PDSS, GADSS, Y-BOCS) - Cut-off scores should be determined in empirical
trials
29Ortrud Schweigert after her visit to the Munch
Museum