CHAT Psychometrics

BACKGROUND
Independence of the Three Severity Subscales
 Hallucinations are a prominent symptom of schizophrenia and other psychotic disorders, but the
neuropathology underlying these aberrant sensory experiences remains poorly understood, particularly in
sensory modalities other than auditory.
 The lack of connection between hallucinations and underlying brain dysfunction could be due to low sensitivity
in hallucination measurement, which is typically one item from the PANSS, or frequency-focused items in the
SAPS.
 Although not commonly used, a “gold standard” instrument for measuring auditory hallucination severity is the
11-item PSYRATS-AH1. It reported 3 severity factors – physical characteristics, cognitive, and emotional.
 We developed a semi-structured interview, the Chicago Hallucination Assessment Tool (CHAT), to improve
hallucination assessment sensitivity for use in conjunction with neurobiological studies of hallucinations.
AUDITORY
Physical
Cognitive
Past
Current
Past
Current
Cognitive
.41
.29
*
*
Emotional
.46
.61
.38
.42
VISUAL
Physical
Inter-Rater Agreement
Cognitive
Past
Current
Past
Current
Cognitive
.35
.58
*
*
Emotional
.28
.53
.71
.46
Test-Retest Reliability
The Chicago Hallucination Assessment Tool
Total Past/Worst Visual
Hallucination Score
Time 1 and Time 2
r = 0.78
3-point
discrepancy
CHAT AH Physical Severity
SAPS Auditory (n = 30)
and…
r = 0.66
CHAT AH Cognitive Severity r = 0.03
CHAT AH Emotional Severity r = 0.32
CHAT VH Physical Severity r = 0.9
SAPS Visual (n = 21) and…
Auditory
• CHAT expands the PSYRATS-AH
SAPS Tactile (n = 9) and…
Visual
• All sensory modalities are assessed
• Scores for both Current and Worst/Past time
PSYCHOMETRIC STUDY PROCEDURES:
CHAT VH Cognitive Severity
r = -0.013
CHAT TH Cognitive Severity
r = 0.37
Olfactory
CHAT OH Physical Severity r = 0.5
psychiatry clinic and from community
advertisement to obtain a range of severity
SAPS Olfactory (n = 8)
and…
•3 new visual hallucinations that raters found “indeterminate”
•1 new olfactory hallucination
•2 new tactile hallucinations
CURRENT
SCORES
PAST/
WORST
SCORES
Total Sample
Test-Retest Reliability
Sample*
30 schizophrenia
20 schizoaffective
15 mood disorder w/
psychosis
11 schizophrenia
7 schizoaffective
2 mood disorder w/ psychosis
Mean (SD) Age
41 (+/- 11 years)
36 (+/- 11 years)
Mean (SD) Years
of Education
12.92 (+/- 2.6)
13.1 (+/- 1.7)
Gender
31 male
32 female
11 male
9 female
Neither interval length between test and retest nor general
illness severity predicted differences between test and retest
scores.
Types of Hallucinations Reported
CHAT OH Cognitive Severity
r = -0.41
CHAT OH Emotional Severity
r = -0.47
PANSS Total (n=65) and…
CHAT Current Total Severity r = .28
PANSS Positive and…
CHAT Current Total Severity r = .43
PANSS Negative and…
CHAT Current Total Severity r = .04
PANSS General and…
CHAT Current Total Severity
r = .25
CONCLUSIONS
• The Chicago Hallucination Assessment Tool (CHAT)
shows good psychometric properties:




High inter-rater agreement
High test-retest reliability
Wide range of scores (sensitivity)
High correlation with existing measures where expected (and low
correlation where expected)
• A larger sample is needed for factor analytic studies to confirm or
suggest alternative severity dimensional structure.
• Future work will include additional psychometric studies and use of
the scales alongside EEG and fMRI data as correlates of severity of
brain function abnormalities….
*Test-retest interval ranged from 32-206 days, mean = 90 (+/- 50 days)
1. Haddock et al. (1999), Psych Med 29, 879-889
r = 0.68
CHAT TH Emotional Severity r = 0.65
3 cases of imperfect reporting of sensory modalities ever experienced between
time 1 and time 2 interviews were all newly reported modalities at time 2
• Administered CHAT along with PSYRATS, SCID, PANSS, and SAPS
hallucination items
• Participants returned 4 weeks or more for test-retest assessment
CHAT TH Physical Severity
Tactile
Gustatory
• Outpatients recruited from medical center
REFERENCE
2-point
discrepancy
CHAT VH Emotional Severity r = 0.13
• A 30-60 minute semi-structured interview
Diagnosis
Total Past/Worst Auditory
Hallucination Score
Time 1 and Time 2
r = 0.73
CHAT Scores
What is the CHAT?
Perfect
Rating
Agreement
92%
Validity: Correlations with
Other Hallucination and
Psychopathology Measures
87% (20/23) of subjects reported the same sensory modalities for past/worst
Total Past/Worst Score
Time 1 and Time 2 r = 0.96
1-point
discrepancy
Visual: N= 35 past ; n = 11 current
N= 51 past; n = 29 current
hallucinations for both time 1 and 2
METHODS
Item Score Agreement on 355 CHAT
Items Across 7 Participants
SUPPORT
For such use of the CHAT, see talk Monday at 3:30 pm (Keedy, presenter):
“Abnormal Evoked Responses to Task-Irrelevant Stimuli in Psychosis Patients Correlates with
Hallucination Severity”
This research was supported by NIH grant K23MH092702 (Keedy)