Discriminating Between Cognitive and Supportive Group Therapies for Chronic Mental Illness

Rhode Island School of Design, Providence, Rhode Island, United States
Journal of Nervous & Mental Disease (Impact Factor: 1.69). 09/2006; 194(8):603-9. DOI: 10.1097/01.nmd.0000230635.03400.2d
Source: PubMed
ABSTRACT
This descriptive and comparative study employed a Q-sort process to describe common factors of therapy in two group therapies for inpatients with chronic mental illness. While pharmacological treatments for chronic mental illness are prominent, there is growing evidence that cognitive therapy is also efficacious. Groups examined were part of a larger study comparing the added benefits of cognitive versus supportive group therapy to the treatment milieu. In general, items described the therapist's attitudes and behaviors, the participants' attitudes and behaviors, or the group interactions. Results present items that were most and least characteristic of each therapy and items that discriminate between the two modalities. Therapists in both groups demonstrated good therapy skills. However, the cognitive group was described as being more motivated and active than the supportive group, indicating that the groups differed in terms of common as well as specific factors of treatment.

Full-text

Available from: Somaia Mohamed, Jul 29, 2015
ORIGINAL ARTICLES
Discriminating Between Cognitive and Supportive Group
Therapies for Chronic Mental Illness
Sarah A. Hayes, MA,* Debra A. Hope, PhD,* Lori S. Terryberry-Spohr, PhD,†
William D. Spaulding, PhD,* Melanie VanDyke, PhD,‡ Dirk T. Elting, PhD,§ Jeffrey Poland, PhD,
Somaia Mohamed, PhD,¶ Calvin P. Garbin, PhD,* Dorie Reed, PhD,# and Mary Sullivan, MSW#
Abstract: This descriptive and comparative study employed a
Q-sort process to describe common factors of therapy in two group
therapies for inpatients with chronic mental illness. While pharma-
cological treatments for chronic mental illness are prominent, there
is growing evidence that cognitive therapy is also efficacious. Groups
examined were part of a larger study comparing the added benefits of
cognitive versus supportive group therapy to the treatment milieu. In
general, items described the therapist’s attitudes and behaviors, the
participants’ attitudes and behaviors, or the group interactions. Results
present items that were most and least characteristic of each therapy and
items that discriminate between the two modalities. Therapists in both
groups demonstrated good therapy skills. However, the cognitive group
was described as being more motivated and active than the supportive
group, indicating that the groups differed in terms of common as well
as specific factors of treatment.
Key Words: Severe mental illness, common factors, treatment
process, psychiatric rehabilitation, cognitive therapy.
(J Nerv Ment Dis 2006;194: 603– 609)
C
ognitive impairments have been seen as key characteris-
tics in schizophrenia and other chronic mental illnesses
since the time of Bleuler (1911/1950). More recently, cogni-
tive based treatments for schizophrenia and other chronic
mental illnesses have been developed. Much of this research
involves specific cognitive impairments that were detected in
the laboratory and then targeted for change, such as atten-
tional deficits and difficulties with conceptual flexibility (see
Corrigan and Storzbach, 1993; Reed et al., 1992; Spaulding et
al., 1986; Storzbach and Corrigan, 1996). This approach,
termed the cognitive process targeting approach (Twamley et
al., 2003), is a key component of integrated psychological
therapy (IPT; Brenner et al., 1992, 1994), a comprehensive
psychosocial treatment modality.
Integrated psychological therapy is a highly structured
group therapy approach consisting of five sections that ad-
dress various aspects of social behavioral functioning. The
first three sections make up the cognitive component of this
treatment: cognitive differentiation, social perception, and ver-
bal communication. In each of these sections, the patients par-
ticipate in a series of graduated group activities designed to
exercise specific cognitive abilities. The final two sections focus
on social skills and interpersonal problem solving. Overall, it
appears that IPT produces a benefit when compared with less
extensive psychosocial treatments (Brenner et al., 1992; Heim et
al., 1989; Kraemer et al., 1987; van der Gaag, 1992).
In a more recent study, Spaulding et al. (1999) specif-
ically examined a modified version of the cognitive compo-
nent of IPT (cognitive differentiation, social perception, and
verbal communication sections) in comparison to a support-
ive group focused on maintaining and enhancing personal and
social functioning as part of a standard rehabilitation regimen
for a population of patients with schizophrenia and other
chronic, severe, and disabling psychiatric disorders. This
study found that patients in both the supportive and the IPT
groups showed improvement. However, patients in the IPT
groups showed greater gains in social competency, psychotic
disorganization, and attentional processing. The supportive
group was designed as an active treatment condition that
contained the nonspecific elements of IPT to test the added
benefit of the cognitive treatment to the treatment milieu.
The specific elements of treatment differed between the
IPT group (i.e., focus on cognitive exercises) and the sup-
portive group (i.e., focus on social cooperation and under-
standing). While it is likely that these specific elements
contribute to the differences between the groups on the
outcome variables, the groups may also differ on common
factors. In 1936, Rosenzweig (1936) first noted that various
forms of psychotherapy may be equally efficacious because
of factors common to all techniques rather than the tech-
niques themselves. Today, common factors of treatment are
seen as essential components of the therapeutic process in all
types of psychotherapy (Wampold, 2001); however, there
*University of Nebraska–Lincoln, Lincoln, Nebraska; †Madonna Institute
for Rehabilitation, Science, and Engineering, Lincoln, Nebraska; ‡Saint
Louis Behavioral Medicine Institute, St. Louis, Missouri; §Hawaii State
Hospital, Kaneohe, Hawaii; Rhode Island School of Design, Providence,
Rhode Island; ¶University of Cincinnati, Cincinnati, Ohio; and #Lincoln
Regional Center, Lincoln, Nebraska.
Portions of this paper were supported by NIMH grant R01 MH44756,
awarded to William D. Spaulding, PhD.
Send reprint requests to Debra A. Hope, PhD, Department of Psychology,
University of Nebraska-Lincoln, 238 Burnett Hall, Lincoln, NE 68588-
0308.
Copyright © 2006 by Lippincott Williams & Wilkins
ISSN: 0022-3018/06/19408-0603
DOI: 10.1097/01.nmd.0000230635.03400.2d
The Journal of Nervous and Mental Disease Volume 194, Number 8, August 2006 603
Page 1
have been numerous uses of the term common factors. For the
purpose of this study, we are defining common factors as
those aspects of treatment that are not specific to a given
intervention. For example, under our definition, the in-session
activity level of clients would be considered a common factor
since many treatments could be characterized as having
active client participation. However, having an active treat-
ment is separate from, although possibly related to, the
specific treatment components. Therefore, active participa-
tion would be considered a common treatment factor.
The difficulty in addressing questions about common
factors is that so few procedures have been developed to
measure common factors quantitatively across therapy mo-
dalities. One way of investigating the differences in these
common factor elements is to use a measure similar to Jones’
Psychotherapy Process Q-sort (PQS; Jones, 1985). Q-sort instru-
ments involve items being sorted into a forced distribution
allowing for a quantitative evaluation of characteristic and un-
characteristic items. The Q-sort procedure has the benefit of
utilizing direct measurement from audiotaped or videotaped
records of actual sessions. In the case of the PQS, the items
represent statements about the therapy process. The PQS has
been used to compare psychodynamic and cognitive-behavioral
therapies for depression (Jones and Pulos, 1993) as well as to
examine therapeutic factors for inpatients and outpatients receiv-
ing group treatment of schizophrenia (Gonza´lez de Cha´vez et al.,
2000). Since the current study is designed to be descriptive, the
use of an adapted PQS procedure allows for a quantitative
method to describe the relative similarities and differences be-
tween the IPT and the supportive groups.
To understand better the treatment effects seen in the
Spaulding et al. (1999) outcome study, it is essential to
understand all differences between the groups. Based on the
outcome study, we know that there were difference between
the cognitive and the supportive groups in terms of outcome.
Additionally, all therapists demonstrated fidelity to the re-
spective treatment manuals. However, we do not know the
extent to which these two treatments also differed on com-
mon factors. Before outcome can be attributed to the specific
treatment elements, other potential differences need to be
ruled out. The purpose of this study was to illustrate any
additional differences between these two established treat-
ment modalities. This descriptive study uses the Q-sort
method to describe how the IPT group and the supportive
group compared across common elements of the psychother-
apeutic process in a group treatment of patients with chronic
mental illness. We did not have specific expectations about
the various constructs in this exploratory study but were
instead interested in the common factor differences between
the two groups in this understudied population.
METHODS
Participants
The data presented here are from the first three (of eight)
6-month cohorts of a study of the effectiveness of cognitive
therapy for a chronic inpatient population (Spaulding et al.,
1999). Please see Spaulding et al. (1999) for the inclusion and
exclusion criteria for the larger study. Table 1 presents the
demographic and clinical characteristics of this sample.
Procedure
Treatments
This study was completed as part of a large-scale study of
treatment efficacy. Each participant, after giving his or her
informed consent, was randomly assigned to either a cognitive
TABLE 1. Demographic Information and Clinical Characteristics of the Sample
Cognitive (N 18) Supportive (N 19) Group t or
2
(p)
Age, mean (SD) 35.18 (10.62) 36.23 (12.27) 0.28 (NS)
Education, mean (SD) 11.83 (2.46) 12.10 (2.10) 0.36 (NS)
Gender (men/women) 9/9 9/10 0.03 (NS)
Ethnicity 1.34 (NS)
European-American 16 17
African-American 1 2
Hispanic 1 0
Diagnosis 9.36 (NS)
Schizophrenia, paranoid 8 6
Schizophrenia, undifferentiated 4 7
Schizophrenia, disorganized 2 1
Schizoaffective disorder 0 4
Other
a
31
Global Assessment Scale, mean (SD) 36.88 (9.26) 37.16 (8.33) 0.09 (NS)
PANSS positive symptoms, number rated
“severe,” mean (SD)
1.53 (1.33) 1.63 (1.46) 0.22 (NS)
PANSS negative symptoms, number rated
“severe,” mean (SD)
1.29 (1.21) 1.26 (1.28) 0.07 (NS)
Antipsychotic medication CPZ equivalent
(mg/d), mean (SD)
1920.18 (2271.78) 1996.71 (1932.46) 0.11 (NS)
a
Other psychosis NOS, organic personality disorder, personality disorder NOS (cognitive), major depression (supportive).
Hayes et al. The Journal of Nervous and Mental Disease Volume 194, Number 8, August 2006
© 2006 Lippincott Williams & Wilkins604
Page 2
therapy group or a supportive-attention control group. Treatment
lasted for 6 months and occurred three times per week. A
therapist trained in the cognitive therapy modalities led the
cognitive therapy cohorts while three therapists trained in gen-
eralized supportive modalities conducted the cohorts of support-
ive-control therapy. All therapists had received treatment man-
uals that described how their modality was to be conducted. To
assess treatment fidelity, pairs of graduate students rated 10 tapes
from each modality on a 5-point Likert scale with higher
numbers indicating better adherence to the cognitive treatment
manual. As expected, cognitive group sessions were highly
adherent to the cognitive manual (M 4.80), whereas the
supportive group sessions contained few elements from the
cognitive manual (M 0.21; Elting et al., 1992).
Cognitive Therapy Groups
The cognitive therapy groups were designed to address
specific social information processing deficits. The cognitive
therapy was based on the cognitive subprograms (cognitive
differentiation, social perception, and verbal communication)
of IPT (Brenner et al., 1992, 1994). A cognitive therapy
manual (Spaulding and Reed, 1989) outlined 18 specific
exercises that formed the core of the cognitive rehabilitation
therapy. The therapist of the IPT group was instructed to
introduce the group activity, guide the participants, and eval-
uate participants’ responses as well as to facilitate social
interactions among the group. The cognitive differentiation
subprogram focused on concept manipulations. The social
perception subprogram focused on the processing of social
information. The verbal communication subprogram focused
on the cognitive substrates of verbal interactions, including
attention and short-term memory.
Supportive Therapy Groups
The supportive therapy groups were based on a generic
modality used to maintain or enhance personal and social func-
tioning among chronic schizophrenia patients. The supportive
manual (Spaulding, 1989), designed specifically for the larger
study, combined the nonspecific aspects of the IPT manual as
well as procedures adopted from accounts of supportive group
therapy work with patients with chronic schizophrenia. The
supportive therapy manual described the assumptions and goals
of supportive therapy (e.g., fostering and maintaining improve-
ments in social functioning), the general context and conditions
of therapy, and the role of the therapist in detail.
In the supportive therapy groups, therapists introduced
the group as an exercise in social cooperation and understand-
ing where patients were going to learn to help each other.
There was no set agenda for these groups; group members
were encouraged to bring in recent experiences, problems,
and concerns for group discussion. The therapists in the
supportive groups were social workers with extensive expe-
rience in supportive group therapy.
Treatment Outcomes
The larger treatment study (Spaulding et al., 1999)
compared the efficacy of cognitive therapy and supportive
therapy as components of a standard treatment regimen.
Overall, individuals in the cognitive group therapy showed
significantly greater improvement on the primary outcome
measure, the Assessment of Interpersonal Problem-Solving
Skills (Donahoe et al., 1990). Effect sizes for the supportive
group ranged from .34 to .46 and from .58 to .91 for the
cognitive group. Those in the cognitive group therapy dem-
onstrated greater improvement in terms of attentional pro-
cessing and the disorganization factor of the Brief Psychiatric
Rating Scale (Ventura et al., 1993). Overall, participants in
both groups showed improvement on measures of attention,
memory, and executive functioning.
Tapes Selected
For this study, each cohort was split into 2-week treatment
periods. Out of each 2-week period, one session of each type of
therapy was randomly chosen to rate with the Q-sort measure.
This resulted in a total of 66 tapes being rated, 34 of which were
tapes of the cognitive modality and 32 of which were tapes of
the supportive-control modality. With this sample size, an effect
size of .35 would yield a power of .80 and an
of .05 (Friedman,
1982). Effect sizes reported in this paper were .43 or greater,
indicating that the sample size was large enough to detect this
size of an effect.
Measures
For this study, the PQS (Jones, 1985) was adapted to
assess common factors occurring in the cognitive and sup-
portive group therapies by a team led by two doctoral level
psychologists (D. Hope and W. Spaulding) with experience in
treatment outcome research. A Q-sort instrument was chosen
since it allows for a descriptive, yet quantitative, method for
comparing relative similarities and differences across groups.
Using Jones’ instrument as a basis, the research team began
the process of adapting items, devising new items, and deleting
others to develop a suitable instrument for use in a group therapy
format. The items were classified as one of three types: those
which describe the therapist’s attitudes and behaviors, those
which describe the participant’s attitudes and behaviors, and
those which describe the interactions of the group. Initial pilot-
ing of the instrument and reliability analysis resulted in some
modifications and revisions, leaving 80 items that were deter-
mined to be suitable for the final Q-sort measure. A nine-
category Q-sort was used with ratings ranging from 1, repre-
senting most characteristic of the session, to 9, representing least
characteristic of the session. Research assistants were first in-
structed to sort the items into three piles: characteristic, unchar-
acteristic, and neutral. Then they sorted the three piles into the
nine categories. The item distribution agreed upon for the nine
categories was as follows: 3, 6, 10, 13, 16, 13, 10, 6, 3.
Reliability
For each tape rated, a minimum of two trained graduate
student research assistants were asked to complete separate
Q-sort ratings to determine interrater reliability. The average
intraclass correlation was 0.73 (McGraw and Wong, 1996).
The Journal of Nervous and Mental Disease Volume 194, Number 8, August 2006 Discriminating Between Therapies
© 2006 Lippincott Williams & Wilkins 605
Page 3
Coding of Items
Two graduate students (separate from those who com-
pleted the Q-sort; one the first author, the second blind to the
study’s purpose) classified the 80 Q-sort items into those that
represented common factors and those that represented spe-
cific factors. Common factors were considered those charac-
teristics that would be expected to be consistent across
treatment modalities. The intraclass correlation was .83
(McGraw and Wong, 1996). A total of 28 items were deter-
mined to represent specific factors of treatment (i.e., “The
therapist focuses on the causal antecedents of group mem-
bers’ disputes”) and were deleted from further analysis.
RESULTS
Results are presented first as Q-sort items that were
most and least characteristic of cognitive therapy and of
supportive therapy. Then, items that discriminate between the
two treatment modalities are discussed. In the description that
follows, Q-sort item numbers correspond to the numbers in
Tables 2 to 4. Mean item ratings range from 2.07 to 7.87 for
the cognitive therapy group and from 2.91 to 7.66 for the
supportive therapy group.
Cognitive Therapy: Most and Least
Characteristic Q-Sort Items
Most and least characteristic items for the cognitive
therapy modality are presented in Table 2. In general, group
members in the cognitive group were attentive (Q-sort item 65;
Q 65), employed humor (Q 79), and were accepting of the
therapist’s comments (Q 76) without verbalizing negative feel-
ings toward the therapist (Q 62). The group members did not
appear withdrawn or aloof (Q 72), nor did they resist participa-
tion in group activities (Q 78). Group members did not appear to
have difficulty understanding or responding to questions (Q 63).
The group activities were not dominated by one or a few group
members (Q 52). In facilitating the cognitive groups, the thera-
pist was observed to employ positive feedback to facilitate group
members’ speech (Q 3) while conveying a sense of nonjudg-
mental acceptance (Q 56). In doing so, the therapist helped
clarify group members’ comments (Q 5); however, she did not
push group members beyond their emotional comfort level (Q
27). As the therapist facilitated the sessions, she addressed the
group members by their first names (Q 2) without being more
positively inclined to certain group members (Q 44). The ther-
TABLE 2. Most and Least Characteristic Q-Sort Items for
Cognitive Therapy
a
Item Description M (SD)
Ten most characteristic items
2 T uses first names 2.07 (0.49)
46 T is confident and self-assured 2.20 (0.65)
19 T communicates in a clear, coherent style 2.44 (0.50)
5 T helps to clarify GMs’ comments 2.65 (0.82)
7 T emphasizes the positive when giving
feedback to GMs
2.68 (0.55)
3 T employs positive feedback to facilitate
GMs’ speech
2.88 (0.97)
56 T conveys a sense of nonjudgmental
acceptance of GMs
2.90 (0.72)
65 GMs are attentive 3.07 (0.46)
76 GMs are accepting of the T’s comments,
observations, and guidance
3.79 (0.59)
79 GMs employ humor 3.93 (0.78)
Ten least characteristic items
25 T condescends to or patronizes the GMs 7.87 (0.75)
39 T appears aloof and avoids emotional
expressiveness
7.87 (0.88)
44 T calls on, or is more positively inclined
toward, certain GMs
7.68 (0.58)
52 One or a few GMs dominate group activity 7.57 (0.90)
21 T has a superficial enthusiasm that appears
disingenuous
7.25 (0.87)
27 T pushes GMs beyond their level of
emotional comfort
7.00 (0.85)
72 Three or more GMs are withdrawn or aloof 6.81 (1.03)
62 GMs verbalize negative feelings toward
therapist
6.62 (0.90)
78 GMs resist participation in group activities 6.50 (0.89)
63 Individual GMs have difficulty understanding
or responding as indicated by long
response latency or lack of response to a
direct question
6.50 (0.59)
a
Endpoints are most characteristic (1) and least characteristic (9). T therapist;
GM group members.
TABLE 3. Most and Least Characteristic Q-Sort Items for
Supportive Therapy
a
Item Description M (SD)
Ten most characteristic items
47 T uses self-disclosure of thoughts to assist GMs in
their coping efforts
2.91 (0.87)
32 T emphasizes commonalities of experience
among GMs
3.02 (1.20)
46 T is confident and self-assured 3.06 (0.59)
2 T uses first names 3.25 (1.39)
19 T communicates in a clear, coherent style 3.41 (0.78)
5 T helps to clarify GMs’ comments 3.76 (1.38)
52 One or a few GMs dominate group activity 3.80 (2.56)
73 One or two GMs are withdrawn or aloof 3.92 (1.26)
72 Three or more GMs are withdrawn or aloof 4.28 (2.02)
39 T appears aloof and avoids emotional expressiveness 4.66 (1.97)
Ten least characteristic items
18 T encourages lower functioning GMs to respond first 7.66 (1.19)
20 T avoids judgmental connotations in feedback 7.17 (0.91)
27 T pushes GMs beyond their level of emotional comfort 7.00 (1.61)
6 T provides nonjudgmental responses that are more
than simple rephrasing or clarifications to GMs’
verbal and nonverbal behavior
6.34 (1.61)
30 T comments on the “group process” 6.12 (0.83)
66 GMs are animated or excited 6.11 (1.54)
8 T selectively responds to positive contributions by GMs 5.98 (0.76)
44 T calls on, or is more positively inclined toward,
certain GMs
5.95 (1.70)
26 T clarifies and explains the purpose of the therapy 5.88 (2.07)
62 GMs verbalize negative feelings toward T 5.72 (1.38)
a
Endpoints are most characteristic (1) and least characteristic (9). T therapist;
GM group members.
Hayes et al. The Journal of Nervous and Mental Disease Volume 194, Number 8, August 2006
© 2006 Lippincott Williams & Wilkins606
Page 4
apist tended to communicate in a clear and coherent style (Q 19)
with an emphasis on the positive (Q 7). The therapist was
described as being confident and self-assured (Q 46) without
appearing superficially enthusiastic (Q 21), condescending or
patronizing (Q 25), or aloof so to avoid emotional expressive-
ness (Q 39).
Supportive Therapy: Most and Least
Characteristic Q-Sort Items
Most and least characteristic items for the supportive
therapy modality are presented in Table 3. In the supportive
groups, group members were animated or excited (Q 66) and
did not verbalize negative feelings toward the therapist (Q
62). In these groups, the activities tended to be dominated by
one or a few group members (Q 52) while one or more group
members were withdrawn or aloof (Q 72; Q 73). In facilitat-
ing the supportive groups, the therapists used self-disclosure
to assist group members’ coping efforts (Q 47), emphasized
commonalities of experiences among group members (Q 32),
and helped to clarify group members’ comments (Q 5),
without pushing group members beyond their level of emo-
tional comfort (Q 27) or clarifying the purpose of therapy
(Q 26). The supportive group therapists did not comment on the
“group process” (Q 30). The therapists appeared to be confident
and self-assured (Q 46), yet aloof and avoidant of emotional
expressiveness (Q 39). They did not encourage lower function-
ing group members to respond first (Q 18), nor did they call on
or were they inclined toward certain group members (Q 44). The
therapists did not selectively respond to group members’ con-
tributions (Q 8). Therapists used group members’ first names
(Q 2) and communicated in a clear and coherent style (Q 19).
They were unlikely to provide nonjudgmental responses that
were more than simple rephrasing of group members’ behavior
(Q 6); however, it was more characteristic for them to use
judgmental connotations during feedback (Q 20).
TABLE 4. Differences in the Therapy Process Across Cognitive and Supportive Therapies
a
Item Description Cognitive M (SD) Supportive M (SD)
More characteristic of cognitive therapy
7 T emphasizes the positive when giving feedback to GMs 2.68 (0.55) 5.42 (1.68) .75
65 GMs are attentive 3.07 (0.46) 4.83 (1.32) .67
76 GMs are accepting of the T’s comments, observations, and guidance 3.79 (0.59) 5.09 (0.88) .66
3 T employs positive feedback to facilitate GMs’ speech 2.88 (0.97) 5.17 (1.62) .66
56 T conveys a sense of nonjudgmental acceptance of GMs 2.90 (0.72) 5.31 (1.88) .66
68 GMs are motivated to participate in group activities 4.10 (0.66) 5.31 (0.84) .63
8 T selectively responds to positive contributions by GMs 4.56 (1.07) 5.98 (0.76) .61
19 T communicates in a clear, coherent style 2.44 (0.50) 3.41 (0.78) .60
46 T is confident and self-assured 2.20 (0.65) 3.06 (0.59) .57
66 GMs are animated or excited 4.50 (1.00) 6.11 (1.54) .53
2 T uses first names 2.07 (0.49) 3.25 (1.39) .50
18 T encourages lower functioning members to respond first 6.13 (1.50) 7.66 (1.19) .49
71 GMs interact in a cooperative manner 4.03 (0.77) 4.95 (1.06) .45
5 T helps to clarify GMs’ comments 2.65 (0.82) 3.76 (1.38) .45
45 T employs humor in the therapy sessions 4.01 (0.62) 5.09 (1.43) .45
30 T comments on the “group process” 5.38 (0.66) 6.12 (0.83) .45
More characteristic of supportive therapy
39 T appears aloof and avoids emotional expressiveness 7.87 (0.88) 4.66 (1.97) .73
32 T emphasizes commonalities of experience among GMs 5.04 (0.79) 3.02 (1.20) .71
52 One or a few GMs dominate group activity 7.57 (0.90) 3.80 (2.56) .71
25 T condescends to or patronizes the GMs 7.87 (0.75) 5.22 (2.00) .67
47 T uses self-disclosure of thoughts to assist GMs in their coping efforts 4.81 (1.26) 2.91 (0.87) .66
21 T has a superficial enthusiasm that appears disingenuous 7.25 (0.87) 5.59 (1.06) .66
72 Three or more GMs are withdrawn or aloof 6.81 (1.03) 4.28 (2.02) .63
60 GMs have difficulty getting started in group activities 6.40 (0.91) 4.83 (1.16) .61
44 T calls on, or is more inclined toward, certain GMs 7.68 (0.58) 5.95 (1.70) .57
78 GMs resist participation in group activities 6.50 (0.89) 5.34 (1.06) .52
73 One or two GMs are withdrawn or aloof 5.44 (1.51) 3.92 (1.26) .48
63 GMs have difficulty understanding or responding as shown by long response
latency or lack of response to a direct question
6.50 (0.59) 5.58 (1.06) .48
80 GMs express angry or aggressive feelings 6.24 (1.12) 4.91 (1.46) .46
67 GMs are provocative: they test the limits of group membership versus compliant
with normal expectations of group behavior
6.26 (1.06) 5.33 (0.90) .43
a
Endpoints are most characteristic (1) and least characteristic (9). T therapist; GM group members;
effect size. Significant differences between Q-sort item means were
obtained by one-way analyses of variance; df 1, 64, p 0.001.
The Journal of Nervous and Mental Disease Volume 194, Number 8, August 2006 Discriminating Between Therapies
© 2006 Lippincott Williams & Wilkins 607
Page 5
Differences in the Therapy Process Across the
Two Treatment Modalities
Each of the 52 common factor items of the Q-sort was
submitted to a one-way analysis of variance test to compare
the mean of each item for cognitive versus supportive ther-
apy. Differences between the two treatment modalities are
presented in Table 4. Using a Bonferroni correction for
inflation (.05/52 .001), a p value less than 0.001 was
considered significant. Overall, there were 30 items with
significant differences between the groups. Of these items, 16
were more characteristic of cognitive therapy and 14 were
more characteristic of supportive therapy. Below is a sum-
mary of the relative differences between the two groups for
items that have large effect sizes (
.50).
The cognitive therapy group was more likely to have a
therapist who communicated in a clear, coherent style (Q 19)
and was confident and self-assured (Q 46). When interacting
with group members, the therapist in the cognitive group was
more likely to emphasize the positive when giving feedback
(Q 7), to employ positive feedback to facilitate group members’
speech (Q 3), to convey a sense of nonjudgmental acceptance of
group members (Q 56), to respond selectively to positive con-
tributions by group members (Q 8), and to use first names (Q 2).
In the cognitive therapy group, the group members were more
likely to be attentive (Q 65); animated or excited (Q 66);
motivated to participate in group activities (Q 68); and accepting
of the therapist’s comments, observations, and guidance (Q 76).
On the other hand, the supportive therapy group was
more likely to have therapists described as aloof and avoiding
emotional expressiveness (Q 39), as condescending or pa-
tronizing to group members (Q 25), and as having superficial
enthusiasm that appeared disingenuous (Q 21). When inter-
acting with group members, the therapists in the supportive
group were more likely to emphasize commonalities of ex-
perience among group members (Q 32) and to use self-
disclosure of thoughts to assist group members in their coping
efforts (Q 47). In terms of group dynamics, it was more
characteristic for the therapists to call on certain group
members (Q 44) and for group activities to be more likely to
be dominated by one or a few group members (Q 52) while
three or more group members appeared withdrawn or aloof
(Q 72). Group members were more likely to have difficulty
getting started in group activities (Q 60) and were more likely
to resist participation in group activities (Q 78).
DISCUSSION
This study sought to examine the psychotherapeutic
process in treatment of inpatients diagnosed with chronic
mental illness by employing a modified version of the PQS to
compare cognitive group therapy to supportive group ther-
apy. In particular, this study sought to examine the common
rather than the specific factors in each therapy modality.
Groups examined for this study were part of a larger study
(Spaulding et al., 1999) comparing the added benefits of
cognitive versus supportive group therapy to the treatment
milieu. The larger study found patients in both groups bene-
fited from the treatments, while patients in the cognitive
group showed more improvement. In the larger study, the
supportive group was designed specifically to provide all of
the common elements and none of the specific elements of the
cognitive therapy group. Thus, the two groups were designed
to differ only in specific factors. The strong treatment fidelity
observed in the groups indicates that the groups were con-
ducted as expected and that the specific treatment factors
differed between the groups. As would be expected, thera-
pists in both groups used good therapeutic skills such as
referring to group members by their first names and keeping
group members within their level of emotional comfort. It is
worth noting that the majority of items listed as least char-
acteristic for both groups were items that countered good
therapy practice. Since these items were least characteristic of
the groups, this may indicate that all therapists used good
therapy skills.
When examining the 10 most and least characteristic
items for the cognitive and supportive therapy groups there
were 10 items in common: four on both most characteristic
lists, three on both least characteristic lists, and three on the
most characteristic list for the supportive group and the least
characteristic list for the cognitive group. Most items that
were consistent across groups, either characteristic in both or
uncharacteristic in both, were items describing basic thera-
peutic skills such as communicating in a clear way and
clarifying group members’ comments without being more
inclined to certain group members. Also, group members
were not verbalizing negative feelings toward the therapists
in either group.
Interestingly, three items were most characteristic of
the supportive group and least characteristic of the cognitive
group. In the supportive group, one or a few group members’
dominated the group activity, while three or more appeared
withdrawn or aloof. Also, the therapists in the supportive
groups appeared aloof and avoidant of emotional expressive-
ness. Taken together, these items indicate that both therapists
and group members appeared to take a more active role in the
cognitive groups. This distinction is also apparent when
looking at the significant differences between the groups.
While this may relate to the structure of the groups (i.e., the
cognitive group had a strict agenda to follow, while the
supportive group did not), it could also be characterized as a
common factor under our definition. Client participation and
involvement in therapy is viewed as an important common
factor variable (Tallman and Bohart, 1999). Both Garfield
(1994) and Orlinsky et al. (1994) reviewed several studies
that show the importance of client involvement for positive
outcome. Therefore, if group members are more actively
involved in the session, then they may be more likely to
benefit from the group than if they are less involved regard-
less of the specific treatment elements. Therefore, it could be
that involvement in the group, regardless of the type of group,
benefits therapeutic outcome.
Overall, it seems that items more characteristic of
cognitive therapy were more positive than items more char-
acteristic of supportive therapy. However, just because an
item was rated more characteristic of the cognitive group
does not indicate that it occurred frequently; rather, it indi-
cates that it was more characteristic in the cognitive group.
Hayes et al. The Journal of Nervous and Mental Disease Volume 194, Number 8, August 2006
© 2006 Lippincott Williams & Wilkins608
Page 6
None of the items more characteristic of cognitive therapy
were negative; however, several of the items more character-
istic of supportive therapy were negative. For example, sup-
portive group members appeared more aloof or resisted
participation, and the therapist was more likely to be de-
scribed as condescending or disingenuous. In the larger treat-
ment study, members of the cognitive group benefited more
than members of the supportive group. While this may be
attributed to specific factors, there also appear to be some
common factor differences. Mainly, the characteristics of the
cognitive group appear to be more positive than the charac-
teristics of the supportive group. Therefore, it may be that the
more positive attitude and the higher levels of motivation
observed in the cognitive group may contribute to the greater
benefit of this group. At the same time, the specific cognitive
group activities are likely to contribute to the group differ-
ences. It is easier to be actively involved in a group consisting
of structured activities. While not examined in this study,
there may be differences among individual therapists that
could affect the differences between the groups; however,
there were no differences among therapists in terms of treat-
ment fidelity. Clients from both groups showed added benefit
from the groups. Therefore, being part of a group based on
common factors of therapy also led to improvement over
pretreatment status. The addition of a no group condition
would allow for the comparison of the common factors
supportive group to no treatment to determine the benefit
from the supportive group.
This study was an initial attempt to describe additional
characteristics of treatment of individuals with chronic men-
tal illness. Although we have described the group process, we
cannot make causal statements regarding which elements
were responsible for the superior outcome of the cognitive
modality. It appears that therapists in both groups exhibited
good therapeutic skills; however, cognitive group members
appeared more motivated and active in group activities than
those in the supportive group. Additionally, therapist factors
such as therapist responsiveness to clients were not specifi-
cally studied here. Future research should explore the rela-
tionship between treatment process and outcome for various
cognitive training approaches. For example, it may be that a
treatment based on specific factors invokes superiority on
common factors as well; therefore, it is unclear whether the
superiority of one treatment over another is driven by com-
mon factors, specific factors, or their combination. Addition-
ally, this study illustrates the feasibility of examining process
questions within standard treatments for individuals with
chronic mental illness.
REFERENCES
Bleuler E (1911/1950) Dementia Praecox or the Group of Schizophrenias
(J Zinkin, Trans). New York: International Universities Press.
Brenner HD, Hodel B, Roder B, Corrigan P (1992) Treatment of cognitive
rehabilitation of chronic psychiatric inpatients. Comput Human Behav.
10:359 –368.
Brenner HD, Roder B, Hodel B, Kienzle N, Reed D, Liberman R (1994)
Integrated Psychological Therapy for Schizophrenic Patients. Toronto:
Hogrefe & Huber.
Corrigan P, Storzbach D (1993) The ecological validity of cognitive reha-
bilitation for schizophrenia. J Cogn Rehabil. 11:14 –21.
Donahoe CP, Carter MJ, Bloem WD (1990) Assessment of interpersonal
problem-solving skills. Psychiatry J Study Interpers Proc. 53:329 –339.
Elting DT, Terryberry-Spohr LS, Quartarolo D, Mohamed S, Hope DA
(1992) Examining Treatment Fidelity in Outcome Studies: An Example
Involving Cognitive Rehabilitation Therapy for Chronic Inpatients. Pre-
sented at the annual meeting of the Association for the Advancement of
Behavior Therapy, Boston, MA.
Friedman H (1982) Simplified determinations of statistical power, magnitude
of effect and research sample sizes. Educ Psychol Meas. 42:521–526.
Garfield SL (1994) Research on client variables in psychotherapy. In AE
Bergin, SL Garfield (Eds), Handbook of Psychotherapy and Behavior
Change (4th ed, pp 190 –228). New York: Wiley.
Gonza´lez de Cha´vez MG, Gutierrez M, Ducaju M, Fraile JC (2000) Com-
parative study of the therapeutic factors of group therapy in schizophrenic
inpatients and outpatients. Group Anal. 33:251–264.
Heim M, Wolf S, Gothe U, Kretschmer J (1989) Kognitive training bei
schizophrenen erkrankungen. Psychiatr Neurol Med Psychol. 41:367–375.
Jones EE (1985) Psychotherapy Process Q-Sort Coding Manual. Unpub-
lished. University of California, Berkeley.
Jones EE, Pulos SM (1993) Comparing the process in psychodynamic and
cognitive-behavioral therapies. J Consult Clin Psychol. 61:306 –316.
Kraemer S, Sulz K, Schmid R, Lassle R (1987) Kognitive therapie bei
standardversorgten schizophrenen patienten. Nervenarzt. 58:84 –90.
McGraw KO, Wong SP (1996) Forming inferences about some intraclass
correlation coefficients. Psychol Methods. 1:30 46.
Orlinsky DE, Grawe K, Parks BK (1994) Process and outcome in psycho-
therapy–Noch einmal. In AE Bergin, SL Garfield (Eds), Handbook of
Psychotherapy and Behavior Change (4th ed, pp 270–378). New York:
Wiley.
Reed D, Sullivan M, Penn D, Stuve P, Spaulding W (1992) Assessment and
treatment of cognitive impairments. In RP Liberman (Ed), Effective
Psychiatric Rehabilitation (pp 7–20). San Francisco (CA): Jossey-Bass.
Rosenzweig S (1936) Some implicit common factors in diverse methods of
psychotherapy. Am J Orthopsychiatry. 6:412– 415.
Spaulding WD (1989) Procedural Manual for Supportive Group Therapy.
Unpublished.
Spaulding WD, Reed D (1989) Procedural Manual for Cognitive Group
Therapy: UN-L/LRC Version. Unpublished.
Spaulding WD, Reed D, Sullivan M, Richardson C, Weiler M (1999) Effects
of cognitive treatment in psychiatric rehabilitation. Schizophr Bull. 25:
657– 676.
Spaulding WD, Storms L, Goodrich V, Sullivan M (1986) Applications of
experimental psychopathology in psychiatric rehabilitation. Schizophr
Bull. 12:560 –577.
Storzbach D, Corrigan P (1996) Cognitive rehabilitation for schizophrenia. In P
Corrigan, S Yudofsky (Eds), Cognitive Rehabilitation for Neruopsychiatric
Disorders (pp 299 –328). Washington DC: American Psychiatric Press.
Tallman K, Bohart AC (1999) The client as a common factor: Clients as
self-healers. In MA Hubble, BL Duncan, SD Miller (Eds), The Heart and
Soul of Change: What Works in Therapy. Washington DC: American
Psychological Association.
Twamley EW, Jeste DV, Bellack AS (2003) A review of cognitive training in
schizophrenia. Schizophr Bull. 29:359 –382.
van der Gaag M (1992) The Results of Cognitive Training in Schizophrenic
Patients. Delft, The Netherlands: Eburon.
Ventura J, Green MF, Shaner A, Liberman RP (1993) Training and quality
assurance with the Brief Psychiatric Rating Scale: “The drift busters.” Int
J Methods Psychiatr Res. 3:221–244.
Wampold BE (2001) The Great Psychotherapy Debate: Models, Methods
and Findings. Mahwah (NJ): Lawrence Erlbaum Associates.
The Journal of Nervous and Mental Disease Volume 194, Number 8, August 2006 Discriminating Between Therapies
© 2006 Lippincott Williams & Wilkins 609
Page 7
  • Source
    [Show abstract] [Hide abstract] ABSTRACT: Introduction: The purpose of this study is the identification of the differences and similarities in the processes relative to Systemic Relational Therapy (SRT), Cognitive Constructivist Therapy (CCT), and Brief Strategic Therapy (BST). Methods: The research was undertaken following an observational systematic approach: 62 therapy sessions (15 SRT, 4 CCT, 42 BST) were used, recorded on audio or audio-video and transcribed exactly, relative to 21 adolescent and adult participants (8 males and 13 females), and a preliminary Italian adaptation of the Psychotherapy Process Q-sort (PQS; Jones, 1985) was used. In order to investigate the processes which seem to best characterize the three psychotherapeutic approaches, descriptive analyses were conducted. Results: The results of this study have substantially an indicative value, but they nevertheless allow us to grasp some indications and trace some lines of development In this sector of research, Conclusions: Specifically, the Psychotherapy Process Q-sort seem to be an important pan-theoretical instrument, capable of offering a further strategy for the analysis of processes in psychotherapy.
    Full-text · Article · Sep 2008
  • [Show abstract] [Hide abstract] ABSTRACT: The purpose of this study is to generate a substantive theory of hospital-based home care for people with severe mental illness in Taiwan. Despite the documented advantages of hospital-based home care services, there is a lack of information and understanding regarding the practices, functions and limitations. Currently, there is no model for how those services are to be provided and what factors will affect these services. The grounded theory method of Strauss and Corbin (Basics of Qualitative Research: Grounded Theory Procedures and Techniques, SAGE Publications, 1990) was used to develop a substantive theory through a paradigm model, including causal conditions, context, intervening conditions, action/interaction strategies and consequences. This study was conducted in six different hospital areas in Central Taiwan in 2007-2008. Data were collected using semi-structured face-to-face interviews. Constant comparative analysis continued during the open, axial and selective coding process until data saturation occurred. Participants were selected by theoretical sampling. When theoretical saturation was achieved, 21 clients with mental illness, 19 carers and 25 professionals were interviewed. A substantive theory of hospital-based home care for people with severe mental illness in Taiwan was developed. The core category was the process of hospital-based home care in helping people with severe mental illness, with 15 categories and 33 sub-categories of the substantive theory. The substantive theory is the first to emerge from hospital-based home care services in Taiwan. Results showed those services had several effective functions for helping people with severe mental illness and their families. The recommendations based on the findings of this research can be used as a guide to improve the delivery of hospital-based home care services to community-dwelling people with severe mental illness and their carers.
    No preview · Article · Oct 2009 · Journal of Clinical Nursing
  • [Show abstract] [Hide abstract] ABSTRACT: The purposes of this study were to understand the functions of hospital-based home care for people with severe mental illness in Taiwan, and the factors that affect functions of professionals who provide hospital-based home care. Hospital-based home care is a service which provides those people with serious mental illnesses who are in crisis and who are candidates for admission to hospital. Home care has been shown to have several advantages over inpatient treatment. However, there is a lack of knowledge about the functions of hospital-based home care for people with severe mental illness in Taiwan. This qualitative study was based on the grounded theory method of Strauss and Corbin. The study was conducted in six different hospital areas in central Taiwan in 2007-2008. Data were collected using semi-structured face-to-face interviews. Constant comparative analysis continued during the open, axial and selective coding processes until data saturation occurred. Participants were selected by theoretical sampling. When theoretical saturation was achieved, 21 clients with mental illness, 19 carers and 25 professionals were interviewed. Several functions were found when these professionals provided hospital-based home care services for people with severe mental illness in Taiwan, including stabilising the clients illness, supplying emergency care services, improving life-coping abilities, employment and welfare assistance, emotional support for both clients and carers, assistance with future and long-term arrangements and assistance with communication between carers and clients. Hospital-based home care provides several important services for helping clients and their families to live in the community. The recommendations based on the findings of this study can be used as a guide to improve the delivery of hospital-based home care services to community-dwelling clients with severe mental illness and their carers.
    No preview · Article · Feb 2010 · Journal of Clinical Nursing