James Mitchell, M.D.
Steve Wonderlich, Ph.D.
Ross Crosby, Ph.D.
James Roerig, Pharm. D.
Martina de Zwaan, M.D.
Neuropsychiatric Research Institute
University of North Dakota School of Medicine
Project Successful? Example
Multicenter Efficacy ++ BN
Multicenter +/- BN
Effectiveness
Multicenter Naturalistic AN, BN
Baseline +
Prospective -
Long Term Follow-up ++ BN, GBP
Multicenter Efficacy
Bulimic Nervosa CBT Non-responders
Background
1. CBT best established psychotherapy
for BN
2. IPT may also work (2 studies)
3. Antidepressants benefit many patients
- Fluoxetine best established
Questions
1. What therapy works best for CBT
non-responders: medication or IPT?
2. Do additional visits prevent relapse in
responders?
Assess CBT Assess Secondary Assess F/U F/U
| | | | | | | |
16 wk 1 wk
-2 0 16 17 33 34 60 80
Measures
EATING PSYCHOPATHOLOGY/
OTHER
BE/V Recall SCID I
EDE (Eating Disorders SCID II
Examination)
BES (Body Esteem Scale) Beck
BTQ (Bulimic Thoughts I.I.P. (Inventory of Interpersonal
Questionnaire) Problems)
EDE-Q (EDE-Questionnaire) MPQ – Impul.
TFEQ (Three Factor Eating Self-Efficacy
Questionnaire)
Rosenberg
HRQ (Helping Relationship
Questionnaire)
SAS (Social Adjustment Scale)
YBC (Yale-Brown-Cornell)
Centers
Clinical - Univ. of Minnesota - J. Mitchell, M.D.,
S. Crow, M.D.
Clinical - Cornell University - K. Halmi, M.D.
Clinical - Rutgers University - T. Wilson, Ph.D.
Data - Stanford University - S. Agras, M.D.,
H. Kraemer, Ph.D.
Rapid Interv. F/U
| |
Remit
F/U F/U
| |
CBT
|
IPT F/U
| |
Symp.
Fluox. Desip. F/U
| |
Week Week Week
17 34 60
Recruitment
Number screened 847
Number screened out 591
Number interviewed 258
Number interviewed out 66
Number started CBT 194
Enter CBT N = 194
Dropout CBT N = 54 (28%)
Abstinent after CBT = 76 (54%)
Symptomatic after CBT = 64 (46%)
35 32
30 IPT
25
20
15 10
10
4
5
0
IPT Drop- Remission
out/Withdrawn
35 32
30 Med Mgmt.
25
20
15
15
10
3
5
0
Meds Drop- Remission
out/Withdrawn
Dropouts
IPT Meds Total
Site 1 9% 55% 31%
Site 2 36% 29% 32%
Site 3 50% 83% 67%
Total 29% 48% 39%
Conclusions
1. With sequencing, attrition unacceptably
high: first intervention 28%
second intervention 39%
2. Response to secondary treatments
unacceptably low.
•
3. • • Sequencing of little/no utility.
Conclusions
1. Priorities for further research
a. improve initial treatment
“packages”
b. identify likely non-responders
early in treatment and shift/
enhance therapy then.
Signal Detection Analysis
Post 4 weeks (6 sessions)
Purging < 70% Non-responders
Sensitivity 86%
Specificity 69%
74% get correct treatment
4% unnecessary treatment
22% not get needed treatment (add later)
Multicenter Effectiveness
Bulimia Nervosa
Self-Help
CBT MEDS
CBT
Long-Term Follow-Up
• Bulimia Nervosa 10-15 years
• Gastric Bypass 13-15 years
Participants
• Females
• Evaluated 1981 - 1987
• Participated in:
– Imipramine/CBT Treatment Trial
– 2 - 5 Year Follow-Up
Assessments
• SCID I
• Ham-D
• Eating Disorders Questionnaire - Version 6
(EDQ-VI)
• Body Shape Questionnaire (BSQ)
• Weissman Social Adjustment (WSAS-SR)
• Multidimensional Personality
Questionnaire - Scale 8 (MPQ-8)
• Reproductive History
• N = 222
• N = 200 (90.1%) located
• N = 179 (80.6%) agreed to participate
• N = 19 (8.5%) refused participation
• N=1 (0.5%) deceased
• N=1 (0.5%) severely disabled
N = 176
N = 20 (11.6% ) Full ED
N = 81 (46.8%) Partial ED
N = 72 (41.5%) Full Remission
Prediction
Baseline Treatment c CBT predicted
long-term outcome
Follow-up Gastric Bypass Patients
18 months Kolanowski, 1997
18 months Choman, et al, 1999
5 years Howard et al, 1995
7 years Avinoch, et al, 1992
7 years Smith, et al, 1995
10 years Wolfe, et al, 1994
Follow-up Gastric Bypass Patients
• Gastric bypass roux-en-y
• 100-125 cm roux limb
• Two applications of TA 90 B stapler 20-
30 cc pouch
• Anastamosis of 12-14 mm diameter + O-
Deklene suture
Follow-up Gastric Bypass Patients
M-FED Interview Eating behavior--
Psychopathology
SF-36 Health Status, Q.O.L.
AUDIT ETOH Use
MeritCare Gastric Health Status/Medical
Bypass History
Questionnaire
Follow-up Gastric Bypass Patients
N = 100
70 Agreed to interview
16 Initially refused (8 agreed later)
8 Deceased
6 Not located
78 Interviewed
+ 8 Deceased
86 Outcome Data
Follow-up Gastric Bypass Patients
Interviewed 36 (47%) – 13 years post
33 (42%) – 14 years post
9 (12%) – 15 years post
Follow-up Gastric Bypass Patients
BMI
Pre-Surgery = 43.8 (32.1 - 57.2)
Minimum Post-surgery = 25.9 (18.4 - 38.0)
Follow-up = 32.8 (22.7 - 49.5)
Follow-up Gastric Bypass Patients
Involuntary Vomiting 52 69%
“Plugging” 33 43%
“Heartburn” 33 43%
Diarrhea 24 32%
Largest Change from Pre-surgery Baseline
Current Change from Pre-surgery Baseline
0
-2
Change in BMI
-4
-6
-8
-10
-12
-14
-16
-18
Never binged Binged pre-surgery only Binged pre- and post-
(n=36) (n=25) surgery (n=9)
Pre-Surgical BMI Lowest BMI (post-surgery) Current BMI
50
45
40
Body Mass Index
35
30
25
20
15
10
5
0
Never binged Binged Pre-Surgery Binged Pre- and Post-
(n=36) Only (n=25) Surgery (n=9)
International Eating Disorders
Standardized Data Base
• EDQ (8.7 Eating Disorder Questionnaire)
• SF-36 (Social Finding-36)
• EDE-Q4 (Eating Disorders Examination)
• EDI-II (Eating Disorders Inventory)
• QEWP-R (Questionnaire on Eating and
Weight Patterns-Revised)
Costs
Scanner $6,000
Computer $2,000
Printer $900 (duplex)
Teleform Program $3,070
IEDSDB Program N.C.
IEDSDB
Eating Disorder Institute 412
U. of South Florida 41
Med. College Ohio 276
Methodology Successful?
Paper/Pencil Testing ++
Direct Entry ++
Scramble Form Entry ++
Entry via Palm Top ++
Computers
Computer-Generated ++
Evaluation
Computer-Generated +/-
Reports
Internet Data Entry +/-
Other Technology Development
• Palm Pilots - Monitoring (EMA:
Ecological Momentary Assessment)
• Palm Pilots - Therapy Extenders
• Palm Pilots - Therapy Self-Help
• Telemedicine - Therapy (CBT for BN)
EMA
• Interval Contingent Recording (e.g., end
of day)
• Signal Contingent Recording (e.g.,
randomly)
• Event Contingent Recording (e.g., prior
to binge eating)
New Models of Service Delivery
• Self-help
• Supervised self-help
• Internet-based
• Phone-based
• Computer adjunct
• Computer-based
• Telemedicine
Overall Conclusions
• CBT best treatment BN
• Secondary treatments/sequencing
generally not helpful
• CBT associated with better long-term
outcomes
• Gastric Bypass associated with good
long-term outcome
• New methods of monitoring/therapy
delivery developing
Doing OK Not Doing OK/Goals
Efficacy Naturalistic F/U
Effectiveness Computer Reports
New Data Entry Internet Entry
Methods
Long Term F/U Therapist Feedback
Studies
Computer Evals