See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/261581172

The Core Conflictual Relationship Theme (CCRT) and psychopathology in

patients selected for dynamic psychotherapy

Article  in  Psychotherapy Research · April 2010

DOI: 10.1093/ptr/10.1.100

CITATIONS READS

23 3,732

5 authors, including:

Alexander Wilczek Jacques P. Barber

Karolinska Institutet Adelphi University
19 PUBLICATIONS   235 CITATIONS    288 PUBLICATIONS   10,630 CITATIONS   

SEE PROFILE SEE PROFILE

Petter Gustavsson Marie Asberg

Karolinska Institutet Karolinska Institutet
194 PUBLICATIONS   6,059 CITATIONS    312 PUBLICATIONS   31,722 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Attempted suicide and shame View project

Panic Disorder STudy View project

All content following this page was uploaded by Jacques P. Barber on 19 September 2014.

The user has requested enhancement of the downloaded file.

Psychotherapy Research 10(1) 100–113, 2000
100 WILCZEK ET AL.
© 2000 Society for Psychotherapy Research

THE CORE CONFLICTUAL RELATIONSHIP THEME

(CCRT) AND PSYCHOPATHOLOGY IN PATIENTS
SELECTED FOR DYNAMIC PSYCHOTHERAPY
Alexander Wilczek
Robert M. Weinryb
Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden
Jacques P. Barber
Department of Psychiatry, University of Pennsylvania, Philadelphia, USA
J. Petter Gustavsson
Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden
Department of Health Sciences, University of Skövde, Sweden
Marie Åsberg
Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden

The present naturalistic study examined the relation between core inter-
personal patterns measured by the Core Conflictual Relationship Theme
(CCRT) method and psychopathology in 55 patients selected for long-term
dynamic psychotherapy. Psychopathology was assessed by the DSM-III-
R, symptom self-report, and the Karolinska Psychodynamic Profile (KAPP).
Patients with different DSM-III-R diagnoses did not differ in their CCRTs
abstracted from the Relationship Anecdote Paradigm interview. Further-
more, lack of flexibility (pervasiveness) in the use of different CCRT com-
ponents was not associated with psychiatric symptoms. Only one signifi-
cant correlation between character pathology and the CCRT was found:
problems concerning the patients’ experience of social significance were
associated with more negative responses of other. Limitations of the CCRT
method and the sample used are discussed.

Relating to others is a central aspect of human life, and the way people relate to
each other often forms characteristic patterns that differ between individuals. Psy-
choanalysts understand the peculiar characteristics of these relationship patterns as
expressions of underlying, often unconscious motives. According to psychoanalytic
theory, features of an individual’s object relations might be expressions of underlying

Ann Sofie Bárány and Lena Norman are gratefully acknowledged for their work as CCRT judges, and
Kerstin Holmér for her transcription of the RAP interviews. The staff at the Institute of Psychotherapy
is gratefully acknowledged for their generous support. Financial support was received from the LJ
Boethius Foundation, the Söderström-König Foundation, Praktikertjänst AB, the Salus Foundation, the
Magn Bergvall Foundation, the Thuring Foundation, the Swedish Medical Research Council (grants nos.
10592 and 5454), the Claes Groschinskys Foundation, and Karolinska Institutet. Additional support was
received from grants R01 DA-08237 from NIDA and RO-1 MH-49902 from NIMH.
Correspondence regarding this article should be addressed to Alexander Wilczek, MD, Bastugatan
19, S-118 25 Stockholm, Sweden. E-mail: alexander.wilczek@pi.ki.se.

100
CCRT AND PSYCHOPATHOLOGY 101

psychopathology. Strong dependent traits, for example, can underlie an anxious ad-
aptation to others as a means of avoiding separation; or obsessive-compulsive traits
may be expressed as patterns of hierarchical relationships, with domination or sub-
mission as a main theme.
Because of the theoretical link between psychopathology and object relations,
several instruments have been constructed to measure different aspects of the latter.
For a recent review see Barber and Crits-Christoph (1993). One of these instruments
is the Core Conflictual Relationship Theme method (CCRT), developed by Luborsky
and Crits-Christoph (1990). The CCRT consists of three basic components: (1) the
individual’s wish, need, or intention; (2) the experienced, anticipated, or fantasized
response from others; and (3) the experienced, anticipated, or fantasized response
of the self. The CCRT has been used in a large number of studies since its introduc-
tion 20 years ago. Several studies suggest that it is a reliable instrument for system-
atic description of central relationship patterns (Crits-Christoph, Luborsky, Dahl, Popp,
Mellon, & Mark, 1988; Luborsky & Diguer, 1995).
The validity of the CCRT has been examined in explorations of the psychothera-
peutic process itself as well as in outcome studies. For example, Crits-Christoph,
Luborsky, and Cooper (1988) found that positive outcome was related to accuracy of
interpretations, and that accurate interpretations early in treatment correlated with stron-
ger therapeutic alliance later in treatment (Crits-Christoph, Barber, & Kurcias, 1993).
Turning to outcome studies, Crits-Christoph and Luborsky (1990) found a sig-
nificant increase in positive responses of both other and self, a decrease in negative
responses, and no changes in wishes when comparing CCRTs early and late in psy-
chotherapy. This finding is in accordance with the clinical experience of many psy-
chotherapists, namely that patients do remain faithful to their wishes and desires,
but they improve their capacity to relate to themselves and to others. In a recent
study Staats, Strack, and Seinfelt (1997), however, found similar changes in responses
of other (ROs) and responses of self (RSs) in a sample of healthy volunteers who
were interviewed twice within a six-month interval, indicating that such changes can
occur even without treatment. It is possible that their finding could be the result of
an artifact of the procedure used to collect the data, namely the Relationship Anec-
dotes Paradigm (RAP) interview (Luborsky, 1990).
Crits-Christoph and Luborsky (1990) have also reported that pervasiveness in
the individual’s repertoire of wishes, responses from others, and responses of self
decreased during the course of psychotherapy. High pervasiveness was defined as
the stereotypical use of only a few different CCRT components, whereas low perva-
siveness implied a more flexible use of a larger number of CCRT components. Re-
cently, Cierpka et al. (1998) reported that pervasiveness of CCRT components was
associated with severity of psychopathology. Thus, there are indications that stereo-
typical relationship patterns can be regarded as an expression of psychopathology.
Although the CCRT has been used for a long time, only one published study
(Cierpka et al., 1998) has explored its capacity to discriminate between different
diagnostic groups. The general aim of the present study was to investigate whether
central relationship patterns as measured by the CCRT were associated with other
measures of psychopathology. We had four specific objectives: first, to compare the
CCRTs of patients in different DSM-III-R (American Psychiatric Association, 1987)
defined diagnostic subgroups. We expected to find different CCRTs in patients with
and without a DSM diagnosis. More specifically, it was expected, for example, that
patients with manifest depression would have more masochistic wishes (e.g., to be
hurt, to be controlled) as well as more depressed RSs.
102 WILCZEK ET AL.

Our second objective was to study the relationship between psychopathology

and pervasiveness of patients’ repertoire of wishes, ROs, and RSs. Psychopathology
was measured in three different ways: presence or absence of DSM-III-R diagnosis;
self-rating on a measure of psychiatric symptoms; and an interview-based instrument
for psychodynamic character assessment, the Karolinska Psychodynamic Profile
(KAPP) (Weinryb & Rössel, 1991). Our hypothesis was that high pervasiveness (i.e.,
the stereotypical use of the CCRT components) would prevail more in patients with
a DSM diagnosis than in patients without, and would correlate significantly with more
self-reported symptoms and to character pathology as measured by the KAPP.
The third objective was to examine the relationship between, on the one hand, the
ratio of positive and negative responses of other (PositiveRO/PositiveRO+NegativeRO)
and self (PositiveRS/PositiveRS+NegativeRS), and on the other hand, psychopathol-
ogy as measured by the DSM-III-R, symptom self-report and the KAPP. A low ratio
was expected to prevail in patients with a DSM diagnosis and to correlate signifi-
cantly with self-reported symptoms and character pathology.
Finally, we wanted to explore whether negative RSs represented a subjective
way to describe symptoms. This was done by examining the association between RS
and self-reported symptoms. The hypothesis was that the RS Depression and Disap-
pointment would be highly correlated with self-reported depressive symptoms.

METHODS AND PROCEDURE

SUBJECTS

The study was conducted at the Institute of Psychotherapy in Stockholm, which

provides long-term psychodynamic psychotherapy. Patients applied for psychotherapy
by telephone, and were interviewed briefly during the call. About one third of the
patients who applied were offered one to three exploratory interviews. One third of
the patients offered an interview were considered suitable for long-term psychody-
namic psychotherapy and were put on a waiting list for treatment at the Institute.
Fifty-eight consecutive patients from the waiting list were invited to participate
in the study. Fifty-five patients (95%) agreed to participate. Fourty-four patients (80%)
were women, mean age 33 years (range 21–54); 11 patients (20%) were men, mean
age 37 years (range 27–53). For a more detailed presentation of the patients see
Wilczek, Weinryb, Gustavsson, Barber, Schubert, and Åsberg (1998).

ASSESSMENTS

Assessment of central relationship patterns. The CCRT method was used to as-
sess relationship patterns (Luborsky & Crits-Christoph, 1990). The patients’ narratives
about their interaction with others provide the information needed to formulate the
CCRT. Three components are defined in the narratives: wishes, needs, or intentions
expressed by the subject (Ws); actual, fantasized, or anticipated ROs; and actual, fan-
tasized, or anticipated RSs. The most frequent of each of the three components is
used to formulate the CCRT.
The most common way to collect data for the CCRT is by assessing relationship
episodes (REs) spontaneously told by patients during psychotherapy sessions. The
sessions are transcribed and REs are identified and scored for each of the three com-
ponents. Barber, Luborsky, Crits-Christoph, and Diguer (1995) compared the CCRTs
CCRT AND PSYCHOPATHOLOGY 103

obtained from transcripts of early therapy sessions with the CCRTs obtained by a spe-
cial interview (RAP) (Luborsky, 1990) made before therapy, and found a relatively high
level of agreement between the two. In the present study, the RAP interview was used
to collect REs. During a RAP interview the patient is asked to describe 10 specific episodes
in which he or she interacted with another person. The patient is asked to describe
what happened, what was said, how the patient her- or himself reacted, and how the
interaction ended. These interviews were transcribed and the 10 episodes from the
RAP interview were scored in the standard CCRT fashion (Luborsky & Crits-Christoph,
1990). All 55 transcripts were scored by the same judge. In five cases we used only 9
of the 10 episodes because one of the episodes did not provide enough information,
and in one case the patient only told 7 episodes.
CCRTs can be scored either by using tailor-made formulations of Ws, ROs, and
RSs or by using standard categories. In the former case the Ws, ROs, and RSs are
formulated to fit the content of the specific narrative as closely as possible. Using the
latter method, raters decide which of the 35 standard Ws, 30 standard ROs, and 31
standard RSs most closely matches their own formulations. Some of these standard
categories have similar meanings (e.g., wish to be understood, to be respected, and
to be accepted), which complicates the study of reliability. Thus, looking at judges’
agreement on the most frequent standard categories may be too stringent a criteria
for calculating reliability. One way to avoid nil agreement between closely related
categories is to use Barber, Crits-Christoph, and Luborsky’s (1990) grouping of the
standard categories, the clustered standard categories. Each of the standard catego-
ries in the present study for Ws, ROs, and RSs were regrouped into one of eight
cluster standard categories.
For the reliability analysis 16 RAP interviews were rated by a second independent
judge. The two most frequent ratings for each CCRT component from each judge were
chosen. In order to correct for chance agreement, we followed Crits-Christoph et al.’s
(1988) use of weighted kappa (Cohen, 1968) for assessing interjudge reliability for each
of the three CCRT components. In contrast to regular kappa, weighted kappa allows
different weights for different levels of agreement. If the most frequent wish rated by
each judge matched, a weight of 1.00 was given; if the most frequent wish of one judge
matched the second most frequent of the other judge, a weight of .66 was assigned. If
only the two second most frequent categories matched, a weight of .33 was used. This
computation was performed separately for Ws, ROs and RSs. The weighted kappa be-
tween the two judges were for Ws, ROs, and RSs .83, .86 and .76, respectively.

Mental disorder. The DSM-III-R was used to diagnose clinical syndromes (Axis
I), personality disorders (Axis II), and global assessment of function (Global Assess-
ment of Functioning, Axis V) (American Psychiatric Association, 1987). DSM-III-R
diagnoses were made by a senior board-certified psychiatrist and psychoanalyst (se-
nior author) with extensive experience conducting and teaching psychiatric inter-
views. No formal reliability test of the diagnoses was made. The diagnoses are pre-
sented in Table 1.

Symptom assessment. The Comprehensive Psychopathological Rating Scale-Self-

Affective (CPRS-S-A) (Svanborg & Åsberg, 1994), which was developed from the
Comprehensive Psychopathological Rating Scale (CPRS) (Åsberg, Montgomery, Perris,
Schalling, & Sedvall, 1978), is a self-report questionnaire containing 19 items that
cover core symptoms of depressive, anxiety, and obsessive-compulsive syndromes.
The complete CPRS consists of 40 reported and 25 observed items covering the full
104 WILCZEK ET AL.

Table 1. DSM-III-R Diagnoses in the Sample

DSM-III-R axis I n
none 26
affective disorders 17
anxiety disorders 6
somatoform disorder 2
sleeping disorder 1
eating disorder 1
sexual disorder 1
sexual dysfunction 1
DSM-III-R axis II
personality disorder NOS 4
borderline personality disorder 1
narcissistic personality disorder 1

GAF mean 70 (SD = 8.0), range 50–85

DSM-III-R = Diagnostic and Statistical Manual of Mental Disorders (3rd
edition, revised).
GAF = Global Assessment of Functioning.

range of psychopathology. Several subscales for different syndromes have been con-
structed from the CPRS, such as the Montgomery-Åsberg Depression Rating Scale
(MADRS) (Montgomery & Åsberg, 1979), the Brief Scale for Anxiety (BSA; Tyrer, Owen,
& Cicchetti, 1984) and the CPRS obsessive-compulsive disorder scale (Thorén, Åsberg,
Cronholm, Jörnestedt, & Träskman, 1980).
Each item contains a description of the symptom and four defined levels of se-
verity. Three additional intermediate levels may be used, resulting in a seven point
scale graded in half steps from 0 to 3, where level 0 represents “no symptoms” and
level 3 represents “extreme symptoms.” The patients are instructed to assess the
severity of the symptom during the last three days.
The CPRS-S-A has been shown to have satisfactory concordance between inter-
view-based ratings and self-ratings for patients with anxiety (r = 0.94 for the BSA
scores) and depressive syndromes (r = 0.80 for the MADRS scores) (Mattila-Evenden,
Svanborg, Gustavsson, & Åsberg, 1996).

Psychodynamic character assessment. The KAPP was used for psychodynamic

character assessment (Weinryb & Rössel, 1991; Weinryb, Rössel, Gustavsson, Åsberg,
& Barber, 1997). Based on psychoanalytic theory, the KAPP is designed to assess
relatively stable character traits and modes of mental functioning as they appear in
self-perception and interpersonal relationships. The instrument consists of 18 subscales.
Seventeen of the subscales measure specific character traits, and the last subscale
evaluates character organization. Each subscale contains an explanation of the subscale
and its three defined levels. Two additional intermediate levels may be used on each
subscale, resulting in a five-point scale (1, 1.5, 2, 2.5, and 3) where level 1 repre-
sents “most normal” and level 3 represents “least normal.” The KAPP subscales are
presented in Table 2. In the present study we used the nine KAPP subscales that
most directly assess aspects of object relations. The reliability of the KAPP scores
was tested by comparing the scores of the first author with the scores obtained from
an independent judge assessing 15 of the 55 audio taped interviews. The intraclass
correlations (Shrout & Fleiss, 1979) are indicated in parentheses below.
CCRT AND PSYCHOPATHOLOGY 105

Nine of the KAPP subscales were used in this study and were aggregated to cre-
ate four scores: (1) quality of object relations (0.60), which comprised the subscales
Intimacy and Reciprocity, and Dependency and Separation; (2) sense of one’s own
social significance (0.75) comprising the subscales Belonging, Feeling of Being
Needed, and Access to Advice and Help; (3) capacity to cope with conflicts between
one’s needs and wishes and the reality principle (0.85) comprising the subscales Frus-
tration Tolerance, Impulse Control, and Coping with Aggressive Affects; and (4) the
subscale Personality Organization (0.48) which does not describe a specific trait but
assesses overall character organization.

PROCEDURE

One interviewer conducted all interviews, in which the RAP interview was im-
mediately followed by the KAPP interview. In a few cases, due to the patient’s wish,
the order was reversed. The interview took approximately two hours and was
audiotaped. Information was also collected for DSM-III-R diagnoses. The CPRS-S-A
questionnaire was filled out at the end of the interview. The KAPP was scored im-
mediately after the interview without listening to the tape. The RAP interview was
transcribed and scored by an independent judge.

DATA ANALYSIS

One-way analysis of variance (ANOVA) was used to compare different groups. The
Pearson correlation coefficient was used in all the correlation analyses with p < 0.05 as
the level of significance.

RESULTS

The most common CCRT in the psychotherapy patients could be formulated in the
following way: I wish to be close to and accepted by others (W), but they are reject-
ing (RO), and that makes me feel depressed (RS) (see Table 3).

CCRT IN THE THREE DSM-III-R DEFINED SUBGROUPS

The two most common Ws, ROs, and RSs were computed for the entire sample
(n = 55), and for three subgroups: one comprising patients without a DSM-III-R diag-
nosis (n = 25); one comprising patients with DSM-III-R affective disorders (n = 17);
and one comprising patients with other DSM-III-R diagnoses (n = 13). The latter group
was very heterogeneous and comprised six different syndromes (see Table 1) which is
why these patients were not included in the subgroup with affective disorders. Fur-
thermore, the 17 affective patients also reported significantly more depressive symp-
toms, according to the CPRS-S-A, than the 13 patients with other disorders or the 25
patients without a diagnosis, F (2,52) = 9.15, p < 0.001. The most common W, RO, and
RS was the same for all subsamples as well as for the entire sample (Table 3).

PERVASIVENESS OF CCRT COMPONENTS AND PSYCHOPATHOLOGY

Following Crits-Christoph and Luborsky (1990), the pervasiveness variable for

Ws was calculated as follows: first, the most frequent W (as defined by the cluster
106 WILCZEK ET AL.

Table 2. The Karolinska Psychodynamic Profile (KAPP) Subscales

QUALITY OF INTERPERSONAL RELATIONS
1. Intimacy and reciprocity
Describes different ways of relating to others—from relations characterized by
intimacy, reciprocity and consideration, to unilateral relations based upon selfish
needs.
2. Dependency and separation
Describes different types of dependency—from relative independence, as a more
adult form of dependency, to infantile dependency.
3. Controlling personality traits
Describes different ways in which the need for power and control may be expressed—
ranging from mature and flexible attitudes, via covert and indirect bids for power or
control, to less mature and more compulsively rigid forms made manifest in relations to
both people and things.
SPECIFIC ASPECTS OF PERSONALITY FUNCTIONING
4. Frustration tolerance
Describes the capacity to endure the tension and displeasure arising from conflict
between wishes felt to be essential and the internal or external limitations involved. The
subscale describes different ways of responding to frustration—ranging from tolerance
and coming to terms with it, via “reactive” modes of functioning e.g., ego-restrictions, to
manifest difficulty in enduring the disagreeable feelings it engenders.
5. Impulse control
Describes different ways of containing urgent affects, wishes and needs of different
kinds, and the way these are expressed in action—ranging from a mature balance
between wishes and reality, via undue emphasis upon the dictates of reality at the
cost of wishes, to manifest difficulty in taking reality into consideration in the pursuit
of gratification.
6. Regression in the service of the ego
Describes the capacity to regress in the service of the ego—ranging from a satisfactory
capacity to relinquish the reality principle temporarily, both playfully, voluntarily and
under control, to manifest difficulty in doing so.
7. Coping with aggressive affects
The subscale ranges from adaptive and goal-directed attitudes, via non-adaptive
inhibition of aggression, to impulsive and destructive expression.
AFFECT DIFFERENTIATION, BOTH WITH REGARD TO EXPERIENCE AND EXPRESSION
8. Alexithymic traits
The subscale ranges from good ability to identify, experience and articulate variation
in feelings and emotional states in a subtle and differentiated manner, to great
difficulty in distinguishing between different feelings and sensations and in verbaliz-
ing them.
9. Normopathic traits
The subscale ranges from good ability to give active expression to personal and individu-
alized needs and wishes, to an incapacity for such personal fantasies and instead clinging
to social conventions or mores.
THE IMPORTANCE ATTACHED TO THE BODY AS A FACTOR
OF SELF-ESTEEM
10. Conceptions of bodily appearance and their significance for self-esteem
Assesses the individual’s more enduring conscious and unconscious conceptions of
the appearance of the body and its significance for self-esteem.
11. Conceptions of bodily function and their significance for self-esteem
Assesses the individual’s more enduring conscious and unconscious conceptions of
the function of the body and its significance for self-esteem.
CCRT AND PSYCHOPATHOLOGY 107

12. Current body image

Assesses the individual’s current conceptions, conscious and unconscious, of his
physical appearance and function, and their effect on his self-esteem.
SEXUALITY
13. Sexual functioning
Assesses the functional sexual capacity of the individual, with regard to sexual activity
with a partner.
14. Sexual satisfaction
Assesses sexual interest, desire, and satisfaction in relation to a partner. The subscale
is graded from an active attitude to sex toward greater inhibition and passivity.
THE INDIVIDUAL’S SENSE OF HIS OWN SOCIAL SIGNIFICANCE
15. Sense of belonging
16. Feeling of being needed
17. Access to advice and help
These three subscales assess the individual’s capacity to relate socially, though it is his
own experience of this and not “objective” fact that is assessed.
CHARACTER AS ORGANIZATION
18. Personality organization
Assesses the degree of differentiation and integration of internalized object relations,
and habitual defence strategies. The subscale is graded from neurotic to psychotic
personality organization.

standard categories) was identified for each individual. Then the number of times
this W occurred was divided by the number of relationship episodes. The mean
pervasiveness score for Ws was 0.49 (range 0.29–0.80), and this figure was used as
the pervasiveness measure. The same procedure was used to calculate pervasive-
ness for ROs (0.51, range 0.30–1.00) and RSs (0.54, range 0.30–0.90).
In order to explore whether the pervasiveness of Ws, ROs, and RSs could be
regarded as an expression of psychopathology, the three DSM-III-R defined subgroups
were compared. The ANOVAs yielded no significant differences in pervasiveness
between the three groups on any of the three CCRT components. The F(2,52)-value
for Ws, ROs, and RSs was 1.71, 2.90, and 0.49 respectively, all nonsignificant. Nor
were there any significant correlations between the GAF and pervasiveness of Ws
(r = –0.10, n.s.), ROs (r = –0.03, n.s.), or RSs (r = –0.07, n.s.).
The expected associations between pervasiveness of RSs and self-reported de-
pressive (r = 0.02, n.s.) and obsessive-compulsive symptoms (r = 0.04, n.s.) were
not found.
In order to test the association between pervasiveness and character pathology,
the former was correlated with the four different aspects of KAPP character pathol-
ogy. No significant correlations were found between the pervasiveness of Ws, ROs,
and RSs and any of the four aspects of character pathology (correlations ranged from
r = –0.09 to 0.14).
Recently another measure of pervasiveness, dispersion, has been proposed by
Cierpka et al. (1998). This measure considers not only the most common CCRT com-
ponent for each patient but the whole spectrum of Ws, ROs, and RSs given in each
relationship episode. Thus, for each interview dispersion measures the flexibility of
all themes. Low dispersion scores indicate more stereotypical patterns and high scores
more flexible interpersonal patterns.
108 WILCZEK ET AL.

In the present study the mean dispersion score for Ws was 0.77 (range 0.54–
0.94), for ROs it was 0.69 (range 0.17–0.88), and for RSs it was 0.73 (range 0.54–
0.89).
The same statistical analyses as mentioned above were conducted with the dis-
persions scores and yielded the following nonsignificant results: (1) the ANOVA re-
garding the three DSM defined groups, F(2,52)-values for Ws, ROs, and RSs were 1.12,
1.91, and 0.15 respectively; (2) the correlations between the GAF and Ws (r = –0.21),
ROs (r = 0.25), and RSs (r = 0.04) were nonsignificant; (3) the correlations between the
dispersion of RSs and self-report depression (r = 0.09, n.s.) and obsessive/compulsive
symptoms (r = 0.01, n.s.) were nonsignificant; (4) the same was true for the relation-
ship between dispersion of the three CCRT components and the KAPP measures of
character pathology (correlations ranged from r = –0.13 to 0.19).

RATIO OF POSITIVE TO (POSITIVE+NEGATIVE) RESPONSES FROM OTHER

AND RESPONSES OF SELF, AND PSYCHOPATHOLOGY

In order to examine the relationship between the prevalence of positive and

negative responses on the one hand, and the measures of psychopathology on the
other, we computed a ratio of positive and negative responses separately for ROs
and RSs for each patient (positive to [positive+negative]). We used the cluster stan-
dard categories (Barber, Crits-Christoph, & Luborsky, 1990) and the following clusters
were defined as positive responses: RO cluster 1 (strong), 6 (helpful), 7 (likes me),
and 8 (understanding); and RS cluster 1 (helpful), 3 (respected and accepted), and
5 (self-controlled and self-confident). All the other clusters were considered nega-
tive responses. Our definition of positive and negative responses differs from the
definition suggested by Luborsky (1990), i.e., that the judges decide whether a cer-
tain RO or RS is positive or negative for a given patient. No significant differences
were found on the RO and the RS ratios when comparing patients with affective
disorders (n = 17), no diagnosis (n = 25), and other syndromes (n = 13) (Table 4).
Neither did we find any significant correlations between the RO and RS ratios and
self-reported anxious, depressive, obsessive/compulsive symptoms or the GAF
(Table 4).

Table 3. Most and (Second Most) Frequent Wish (W), Response From
Other (RO), and Response of Self (RS) for the Total Sample and Three
Diagnostic Subgroups
CCRT
standard category cluster

n W RO RS

Total sample 55 be close (0.36) rejecting (0.45) depressed (0.44)

(to be hurt, controlled) (0,23) (understanding) (0.22) (helpless) (0.26)
No DSM 25 be close (0.40) rejecting (0.39) depressed (0.39)
diagnosis (to achieve & help) (0.21) (understanding) (0.24) (helpless) (0.23)
Affective 17 be close (0.32) rejecting (0.51) depressed (0.51)
disorder (to be hurt, controlled) (0.29) (understanding) (0.21) (helpless) (0.29)
Other DSM 13 be close (0.33) rejecting (0.50) depressed (0.44)
diagnosis (be loved) (0.22) (understanding) (0.19) (anxious & ashamed)(0.28)

Note. The cluster standard categories are used. Frequencies in parentheses.

CCRT AND PSYCHOPATHOLOGY 109

Table 4. Summary of Analyses Computed With the Positive to (Positive +

Negative) Ratio for ROs (PRO/PRO + NRO) and RSs (PRS/PRS + NRS)
PRO/PRO + NRO PRS/PRS + NRS
Comparison of 3 DSM-III-R defined groups. F (2,52) = 0.47, ns F (2,52) = 0.71, ns
CPRS-S-A—anxiety r = –0.02, ns r = –0.02, ns
CPRS-S-A—depression r = –0.06, ns r = –0.11, ns
CPRS-S-A—obsessive/compulsive r = –0.02, ns r = –0.12, ns
GAF r = –0.13, ns r = –0.04, ns
KAPP subscales 1, 2 (object relations) r = –0.10, ns r = –0.15, ns
KAPP subscales 4, 5, 7 (handling conflicts) r = –0.23, ns r = –0.22, ns
KAPP subscales 15, 16, 17 (social significance) r = –0.33, p < 0.05 r = –0.24, ns
KAPP subscale 18 (personality organization) r = –0.19, ns r = –0.23, ns
CPRS-S-A = The Comprehensive Psychopathological Rating Scale-Self-Affective; GAF = The Global As-
sessment of Function; KAPP = The Karolinska Psychodynamic Profile (the subscales are presented in
Table 2); PRO = positive response from other; NRO = negative response from other; PRS = positive
response of self; NRS = negative response of self.

The correlations between the response ratios and the four different KAPP scores
were also computed. With the exception of the correlation between problems of social
significance and the RO ratio, none of the other correlations were significant (Table 4).
Thus, the more problems patients had with their senses of social significance, the more
negative responses from others they reported.

RESPONSES OF SELF AND SELF-REPORTED SYMPTOMS

In order to study the relationship between the eight cluster standard category RSs
and self-reported symptoms (CPRS-S-A), Pearson correlation coefficients were com-
puted. Significant correlations were found between the RS cluster Unreceptive (patients’
experience of lack of understanding, not being open, and disliking others) and self-
reported anxiety (r = 0.30, p < 0.05), depression (r = 0.33, p < 0.05), and obsessive-
compulsive symptoms (r = 0.48, p < 0.0001). Unreceptive was also inversely associated
with the GAF (r = –0.40, p < 0.01). The RS cluster Self-controlled and self-confident was
inversely correlated to self-reported obsessive compulsive symptoms (r = –0.33,
p < 0.05), and positively correlated to the GAF (r = 0.29, p < 0.05). However, the ex-
pected association between the cluster standard category Depression and Disappoint-
ment and self-reported depressive symptoms was not found (r = –0.10, n.s.).

DISCUSSION

In the present naturalistic study we examined whether CCRT-defined relationship

patterns in patients selected for long-term dynamic psychotherapy were associated
with measures of psychopathology. Three different measures of psychopathology
were used: presence of a syndrome diagnosis according to the DSM-III-R; symptom-
atic self-report according to the CPRS-S-A; and character pathology as defined by an
interview-based instrument, the KAPP. Only a few weak associations were found
between psychopathology as defined by these instruments and relationship patterns
110 WILCZEK ET AL.

as defined by the CCRT. This unexpected lack of difference in relationship patterns

between patients with different kinds and severity of psychopathology could be due
to characteristics of the sample, to the instruments used, or to the lack of validity of
our hypotheses.
The present sample consisted of patients who in socio-economic terms resemble
samples in other psychotherapy studies, such as the Penn study (Luborsky, Crits-
Christoph, Mintz, & Auerbach, 1988). In terms of psychopathology, 55% of the present
sample fulfilled criteria for a DSM-III-R axis I or II diagnosis, with depression as the
most common axis I diagnosis. There was a limited range of dynamically defined
character pathology. The most common kind of problem was inhibition, mainly due
to impaired frustration tolerance where anxiety, depression, or other psychological
and physiological symptoms were experienced when frustration could not be avoided
defensively (Wilczek et al., 1998). It could be argued that patients exposing a wider
range of character pathology would also reveal more significant differences in rela-
tionship patterns.
It is perplexing that only one study (Cierpka et al., 1998) investigating the dis-
criminant validity of the CCRT has been published. A possible explanation could be
that negative results are rarely published. Findings similar to ours were presented in
the 1997 international Society for Psychotherapy Research conference by Lefebvre,
Diguer, Morissette, and Rousseau (1997), who found very similar CCRTs across pa-
tients with psychotic, borderline, and neurotic personality organization.
We had expected to find a relationship between the prevailing RS, i.e., Depressed
and Disappointed, and the depression subscale of the symptom questionnaire (CPRS-
S-A). There was no such association, and moreover, we found hardly any associa-
tions between CCRT and self-reported symptoms. There were a few unexpected weak
but significant associations, mainly between the RS cluster Unreceptive and self-
reported anxiety, depression, and obsessive/compulsive symptoms. Naturally, this
might be due to low power for detecting such associations. Power, however, did not
seem to be the issue since a one-tail test with a 5% level of significance resulted in
a power of 0.99 for detecting a large correlation (r = 0.50), and a power of 0.75 for
detecting a moderate correlation (r = 0.30; Cohen, 1988). That is, we had power to
detect moderate to strong associations. Since the CCRTs in the three DSM-defined
small to moderate size subgroups were almost identical regarding both the most and
second most frequent component, it is very unlikely that this lack of difference in
relationship patterns was due to the present sample size. In this context, it is inter-
esting that very similar CCRTs were found by Eckert, Luborsky, Barber, and Crits-
Christoph (1990) in patients with major depression.
Finally, turning to the issue of the relationship between CCRT and character
pathology as assessed by the KAPP, one weak association was found, namely, be-
tween negative ROs and problems concerning social significance. This association
was expected, but more intriguing is the fact that this was the only association we
found. From a clinical point of view one would expect patients with impaired qual-
ity of object relations (as expressed in part-object relations and immature depen-
dency) to experience others as not trustworthy and rejecting. In the CCRT this could
be expressed by more negative ROs and RSs. The present results, however, indicate
no association between quality of object relations and negative ROs and RSs.
Moreover, our findings indicate that patients with an inhibition regarding inner
impulses and outer challenges leading to difficulties in dealing with conflicts, as
defined by the KAPP, did not reveal the expected association with negative responses.
Patients with problems in establishing meaningful and close personal relations often
CCRT AND PSYCHOPATHOLOGY 111

show a passive, inhibited attitude in social situations. Thus, the previously mentioned
association between negative ROs and social significance was expected, but the lack
of association with more negative RSs was surprising.
Pervasiveness, i.e. lack of flexibility in the use of the CCRT components, was
not associated with a disturbed personality organization. Nor was dispersion, a more
comprehensive measure of pervasiveness, associated with any of the psychopathol-
ogy measures. The latter finding is in contrast to Cierpka et al.’s (1998) recent find-
ings that normal adults had more flexible relationship patterns than psychiatric patients.
Although it cannot be ruled out that some of our lack of results could be due to
small sample size and problems with the instruments used to validate the CCRT, the
surprising lack of significant results in the present study raises concern about the
usefulness of the CCRT as a discriminating instrument. Since it is a common clinical
impression that patients with different kinds of psychopathology have different kinds
of relationship patterns, the question remains why the CCRT does not reveal them?
One problem with the CCRT might be the standard categories, which are formu-
lated on a level corresponding to conscious and ego syntonic processes in relatively
well-functioning patients. As a consequence, the CCRT might reveal more or less the
same intentions and responses in all patients. For example, the wish to be close (i.e.,
to be included, not alone, and to be friends) (Luborsky & Crits-Christoph, 1990) could
for a neurotic person mean a wish for a mature and mutual close relationship, whereas
a patient with a psychotic personality organization might wish to merge with the ob-
ject. The standard categories will not unveil this difference and will not capture the
possible development of this wish during the course of psychotherapy. If it were pos-
sible to reliably score also the maturity level of each standard category, this problem
might be overcome. The same argument could be raised for other rating systems, and
we are not aware of an existing solution to resolve this issue. A similar argument has
been expressed by Henry, Strupp, Schacht, & Gaston (1994) regarding the CCRT’s
capacity to distinguish between different kinds of depression. The CCRT clustered stan-
dard categories may also be too broad to capture patients’ idiosyncratic interpersonal
themes and thus could have contributed to the lack of differences between the diag-
nostic groups. Another problem with the CCRT could be the method used to collect
data for a CCRT formulation, that is, session narratives vs. the RAP interview. The find-
ings of Staats et al. (1997) offers, of course, reason to reconsider the use of the RAP
interview procedure. Although we cannot answer the question whether our results might
have differed had we used material from therapy sessions, it seems unlikely at this
point. Barber et al. (1995) showed that CCRTs obtained from RAP interviews share a
moderately high level of similarity with CCRT obtained from session material.
To conclude, theoretically anticipated associations between core interpersonal
patterns and psychopathology were not found in this study. This raises our concern
regarding the discriminative capacity of the CCRT, and also, considering the number
of CCRT studies that have been reported, we find that more studies regarding the
CCRT’s discriminant validity are warranted.

REFERENCES

American Psychiatric Association (1987). Diagnos- D., & Sedvall, G. (1978). The Comprehensive
tic and statistical manual of mental disorders Psychopathological Rating Scale (CPRS). Acta
(3rd., Revised ed.). Washington, DC: APA. Psychiatrica Scandinavica, Supplement 271, 5–
Åsberg, M., Montgomery, S., Perris, C., Schalling, 27.
112 WILCZEK ET AL.

Barber, J. P., & Crits-Christoph, P. (1993). Ad- chotherapy Research annual meeting, Amelia
vances in measures of dynamic formulations. Island, FL..
Journal of Consulting and Clinical Psychology, Luborsky, L. (1990). The Relationship Anecdotes
61, 574–585. Paradigm (RAP). In L. Luborsky & P. Crits-
Barber, J. P., Crits-Christoph, P., & Luborsky, L. Christoph (Eds.), Understanding Transference
(1990). A guide to the standard categories and (pp. 102–113). New York: Basic Books.
their classification. In L. Luborsky & P. Crits- Luborsky, L., & Crits-Christoph, P. (1990). Under-
Christoph (Eds.), Understanding transference standing Transference. New York: Basic Books.
(pp. 37–50). New York: Basic Books. Luborsky, L., Crits-Christoph, P., Mintz, J., &
Barber, J. P., Luborsky, L., Crits-Christoph, P., & Auerbach, A. (1988). Who will benefit from psy-
Diguer, L. (1995). A comparison of Core Con- chotherapy. New York: Basic Books.
flictual Relationship Themes before psycho- Luborsky, L., & Diguer, L. (1995). A novel reliabil-
therapy and during early sessions. Journal of ity study: Reply to Zander et al. Psychotherapy
Consulting and Clinical Psychology, 63, 145–148. Research, 5, 237–242.
Cierpka, M., Strack, M., Benninghoven, D., Staats, Mattila-Evenden, M., Svanborg, P., Gustavsson, P.,
H., Dahlbender, R., Pokorny, D., Frevert, G., & Åsberg, M. (1996). Determinants of self-rat-
Blaser, G., Kächele, H., Geyer, M., Körner, A., ing and expert rating concordance in psychiat-
& Albani, C. (1998). Stereotypical relationship ric out-patients, using the affective subscales of
patterns and psychopathology. Psychotherapy the CPRS. Acta Psychiatrica Scandinavica, 94,
and Psychosomatics, 67, 241–248. 386–396.
Cohen, J. (1968). Weighted kappa: Nominal scale Montgomery, S., & Åsberg, M. (1979). A new de-
agreement with provision for scaled disagree- pression scale designed to be sensitive to change.
ment on partial credit. Psychological Bulletin, British Journal of Psychiatry, 134, 382–389.
70, 213–220. Shrout, P. E., & Fleiss, J. L. (1979). Intraclass cor-
Cohen, J. (1988). Statistical power analysis for the relations: Uses in assessing rater reliability. Psy-
behavioral sciences. Hillsdale, NJ: Erlbaum. chological Bulletin,86, 420–428.
Crits-Christoph, P., Barber, J. P., & Kurcias, J. S. Staats, H., Strack, M., & Seinfeld, B. (1997). Veränder-
(1993). The accuracy of therapists’ interpreta- ungen des zentralen Beziehungskonfliktthemas
tions and the development of the therapeutic bei Probanden, die nicht in Psychotherapie sind.
alliance. Psychotherapy Research, 3, 25–35. Zeitschrift für Psychosomatiche Medizin und Psy-
Crits-Christoph, P., & Luborsky, L. (1990). Changes chotherapie, 2, 166–178.
in CCRT pervasiveness during psychotherapy. Svanborg, P., & Åsberg, M. (1994). A new self-
In L. Luborsky & P. Crits-Christoph (Eds.), Un- rating scale for depression and anxiety states on
derstanding Transference (pp. 133–146). New the Comprehensive Psychopathological Rating
York: Basic Books. Scale. Acta Psychiatrica Scandinavica, 89, 21–28.
Crits-Christoph, P., Cooper, A., Luborsky, L. Thorén, P., Åsberg, M., Cronholm, B., Jörnestedt,
(1988). The accuracy of therapist interpretation L., & Träskman, L. (1980). Clomipramine treat-
and the outcome of dynamic psychotherapy. ment of obsessive-compulsive disorder. A con-
Journal of Consulting and Clinical Psychology, trolled clinical trial. Archives of General Psychia-
56, 490–495. try, 37, 1281–1285.
Crits-Christoph, P., Luborsky, L., Dahl, L., Popp, Tyrer, P., Owen, R., & Cicchetti, D. (1984). The brief
C., Mellon, J., & Mark, D. (1988). Clinicians can scale for anxiety; a subdivision of the compre-
agree in assessing relationship patterns in psy- hensive psychopathological rating scale. Journal
chotherapy: The Core Conflictual Relationship of Neurology, Neurosurgery and Psychiatry, 47,
Theme method. Archives of General Psychiatry, 970–975.
45, 1001–1004. Weinryb, R. M., & Rössel, R. J. (1991). Karolinska
Eckert, R., Luborsky, L., Barber, J., & Crits-Christoph, Psychodynamic Profile—KAPP. Acta Psychia-
P. (1990). The narratives and CCRTs of patients trica Scandinavica, 83 (suppl 363), 1–23.
with major depression. In L. Luborsky & P. Crits- Weinryb, R. M., Rössel, R. J., Gustavsson, J. P.,
Christoph (Eds.), Understanding Transference Åsberg, M., & Barber, J. P. (1997). The Karolinska
(pp. 222–234). New York: Basic Books. Psychodynamic Profile (KAPP): Studies of char-
Henry, W. P., Strupp, H. H., Schacht, T. E., & acter and well-being. Psychoanalytic Psychology,
Gaston, L. (1994). Psychodynamic approaches. 14, 495–515.
In A. E. Bergin & S. L. Garfield (Eds.), Hand- Wilczek, A., Weinryb, R. M., Gustavsson, P. J.,
book of Psychotherapy and Behavior Change Barber, J. P., Schubert, J., & Åsberg, M. (1998).
(pp. 467–508). New York: Wiley & Sons. Symptoms and character traits in patients se-
Lefebvre, R., Diguer, L., Morissette, E., & Rousseau, lected for long-term psychodynamic psycho-
J. P. (1997). The CCRT: Preliminary results on therapy. Journal for Psychotherapy Practice and
borderline personality disorder. Society for Psy- Research, 7, 23–34.
 CCRT AND PSYCHOPATHOLOGY 113

Zusammenfassung
Die vorliegende naturalistische Studie prüfte den Zusammenhang zwischen interpersonalen Mustern –
gemessen mit der ZBKT-Methode – und der Psychopathologie bei 55 Patienten, die für eine psychodynamische
Langzeittherapie indiziert waren. Psychopathologie wurde mit dem DSM-III-R, Selbstbeschreibungen und
dem Karolinska Psychodynamic Profile (KAPP) untersucht. Patienten mit verschiedenen DSM-III-R Diagnosen
unterschieden sich nicht im Hinblick auf ihre ZBKT in einem Beziehungsinterview. Mangel an Flexibilität
bezüglich unterschiedlicher ZBKT-Komponenten stand nicht in Beziehung zu psychiatrischen Symptomen.
Nur eine signifikante Korrelation zwischen Charakterpathologie und ZBKT ließ sich nachweisen: Probleme
der Patienten mit ihrer sozialen bedeuutng standen in Beziehung zu negativeren Reaktionen von Objekten.
Begrenzungen der ZBKT-Methode und der Stichprobe werden diskutiert.

Résumé
Cette étude naturaliste examine la relation entre des patterns interpersonnels de base mesurés par la
méthode du Thème Relationnel Conflictuel Central (CCRT) et la psychopathologie chez 55 patients
sélectionnés pour une psychothérapie psychodynamique de longue durée. La psychopathologie a été
évaluée selon DSM-III-R, par un auto-questionnaire de symptômes et par le Karolinska Profile Psycho-
dynamique (KAPP). Les patients avec des diagnostics DSM-III-R différents n’ont pas montré de différence
dans leur CCRT obtenu sur la base d’interviews RAP (Paradigme d’Anecdotes Relationnelles). Par ailleurs,
le manque de flexibilité (« pervasiveness ») dans l’emploi des différentes composantes du CCRT n’a pas
été associé à des symptômes psychiatriques. Une seule corrélation significative entre pathologie de
caractère et CCRT a été trouvée; une expérience de position sociale problématique a été associée à
des réponses de l’autre plus négatives. Nous discutons des limitations de la méthode du CCRT et de
l’échantillon.

Resumen
El presente estudio naturalístico examina la relación entre el núcleo de pautas interpersonales medida
por el método del Tema Nuclear Relacional Conflictivo (CCRT) y la psicopatología en cincuenta y cinco
pacientes seleccionados para psicoterapia dinámica a largo plazo. La psicopatología se evaluó según el
DSM-III-R, el informe auto-administrado de síntomas y el Perfil Psicodinámico de Karolinska (KAPP).
Pacientes con diferentes diagnósticos según el DSM-III-R no difirieron en el CCRT abstraído de la
entrevista Paradigma Anecdótico Relacional (Relationship Anecdote Paradigm). Más aun, la falta de
flexibilidad [‘pregnancia’ (‘pervasiveness’)] en el uso de diferentes componentes del CCRT no estuvo
asociada con síntomas psiquiátricos. Solo se encontró una correlación significativa entre la caracteropatía
y el CCRT; los problemas relacionados con la experiencia de significatividad social de los pacientes
estuvo asociada con respuestas más negativas del otro. Se discuten las limitaciones del método del
CCRT y de la muestra usada.

Received June 6, 1998

Revision Received February 2, 1999
Accepted March 17, 1999

View publication stats