Go to ScienceDirect® Home Skip Main Navigation Links
 Register or Login:  Password:    
 Athens/Institution Login 
HomeSearchBrowse JournalsBrowse Book Series, Handbooks and Reference WorksBrowse Abstract DatabasesMy ProfileAlertsHelp (Opens New Window)
 Quick Search:  within Quick Search searches abstracts, titles, keywords, and authors. Click here for more information. 
Clinical Psychology Review
Volume 25, Issue 4 , June 2005, Pages 459-486

This Document
SummaryPlus
Full Text + Links
   ·Full Size Images
PDF (359 K)
External Links
Abstract + References in Scopus
Actions
Cited By
Save as Citation Alert
E-mail Article
Export Citation

doi:10.1016/j.cpr.2005.03.001    How to Cite or Link Using DOI (Opens New Window)  
Copyright © 2005 Elsevier Ltd All rights reserved.

Empirical evaluation of the assumptions in identifying evidence based treatments in mental health

Sidney J. Blatta, Corresponding Author Contact Information, E-mail The Corresponding Author and David C. Zuroffb

aYale University, Suite 901, 300 George Street, New Haven, CT 06511, United States
bMcGill University, Department of Psychology, 1205 Dr. Penfield Ave., Montreal, Quebec, Canada H3X 285

Received 3 March 2005;  accepted 7 March 2005.  Available online 12 May 2005.


Abstract

Extensive analyses of data from the remarkably comprehensive data set established by the Treatment of Depression Collaborative Research Program (TDCRP), initiated and conducted by the National Institute of Mental Health (NIMH), enabled us to examine the contributions of three dimensions of the treatment process (type of treatment, aspects of the therapeutic relationship, and patients' pretreatment personality characteristics) to three assessments of therapeutic change (symptom reduction, reduction of vulnerability, and development of adaptive capacities) evaluated at termination and extended follow-up. The most consistent factors predicting therapeutic gain were the quality of the therapeutic relationship and patients' pretreatment personality dimensions. The implications of these findings for clinical practice, training, and research are discussed.


Article Outline

1. The NIMH-sponsored TDCRP
1.1. Factors that could contribute to therapeutic change
1.1.1. Type of treatment
1.1.2. The quality of the therapeutic relationship
1.1.3. Patient pretreatment characteristics
1.2. Assessment of therapeutic change
1.2.1. Assessment of symptom reduction
1.2.2. Assessment of vulnerability to depression
1.2.3. Assessment of resilience (Enhanced Adaptive Capacity—EAC)
2. Integration of further analyses of the TDCRP data set
2.1. Impact of type of treatment on therapeutic outcome as assessed by three measures of therapeutic outcome
2.1.1. Symptom reduction
2.1.2. Reduction of vulnerability
2.1.3. Development of enhanced adaptive capacities (EAC)
2.1.4. Summary of difference among treatment conditions
2.2. Impact of therapeutic relationship as assessed on three measures of therapeutic outcome
2.2.1. Impact of therapeutic relationship on symptom reduction
2.2.2. Impact of therapeutic relationship on the reduction of vulnerability
2.2.3. Impact of therapeutic relationship on development of enhanced adaptive capacities
2.2.4. Summary of the impact of therapeutic relationship on therapeutic outcome
2.3. Impact of patient pretreatment personality characteristics (level of DAS perfectionism) on three measures of therapeutic outcome
2.3.1. Symptom reduction
2.3.2. Development of the therapeutic alliance
2.3.3. The development of resilience
2.3.4. Summary of effects of patients' pretreatment characteristics on therapeutic outcome
3. Discussion
References


A remarkably diverse data set on the brief outpatient treatment of serious depression, the NIMH-sponsored Treatment of Depression Collaborative Research Program (TDCRP), became available to the scientific community in 1994.1 The TDCRP is probably the most extensive and comprehensive data set ever established in psychotherapy research. In addition to being a very well-designed and carefully conducted randomized clinical trial (RCT) comparing two forms of brief outpatient psychotherapy for depression (Cognitive-behavioral [CBT] and Interpersonal [IPT]) with antidepressant medication (Imipramine) and a double-blind placebo, the TDCRP data set contains extensive and diverse evaluations of the patients before, during, and subsequent to treatment. Because of our prior findings (e. g., Blatt, 1992, Blatt & Felsen, 1993 and Blatt & Ford, 1994; see a summary in Blatt & Shahar, 2004) demonstrating the importance of patient–treatment (P–T) and patient–outcome (P–O) interactions (e.g., Cronbach, 1953) in understanding aspects of therapeutic process and outcome in long-term, intensive treatment, we became interested in investigating P–T and P–O interactions in brief outpatient treatments in the extensive TDCRP data set. Over the past 10 years, we and our colleagues2 have explored various facets of the TDCRP data set and have published a series of papers that have provided considerable understanding about the complex interactions among aspects of the treatment and characteristics of the patients and the therapists and how these interactions impact on therapeutic process and outcome in the brief outpatient treatment of serious depression. In the years since we began to explore the diverse TDCRP data set, we have become increasingly aware that our findings have important implications for the contemporary emphasis on evidence-based treatment and the current attempts to identify empirically supported treatments within mental health. This paper attempts to integrate the diverse findings from our various analyses of the TDCRP data set and to consider their implications for establishing evidence based treatments and for identifying empirically supported treatments for psychological disorders.

Responsibility to our patients and to our disciplines demands that we systematically evaluate our interventions in mental health services. Most efforts to identify empirically supported treatments (ESTs) in mental health thus far have been based on two implicit fundamental assumptions: 1) that the type of treatment is the primary factor in determining therapeutic outcome and 2) that therapeutic outcome is most effectively assessed by a reduction of manifest symptoms. Because these two fundamental assumptions determine much of the contemporary effort to identify empirically supported treatments—that the type of treatment is the primary factor in facilitating therapeutic change and that this change is best assessed by the reduction of symptoms—it is essential that we evaluate these assumptions. Though various meta-analyses (e.g., Lambert & Barley, 2002) and extensive literature reviews (e.g. Westen, Novonty, & Thompson-Brenner, 2004) have addressed these issues, the TDCRP data allows us to test these assumptions within a single comprehensive data set. Thus, the purpose of this paper is to consider the implications of our various findings from the TDCRP data set to evaluate empirically, within a single comprehensive data set, these two fundamental assumptions underlying contemporary efforts to identify empirically supported treatments.

Much of treatment research has compared various forms of treatment with either waiting list controls or “treatment as usual” (TAU) as defined by contemporary clinical practice in the community. Both of these designs, however, frequently involve weak control groups. Contrast with waiting list controls indicates only that doing something is better than doing nothing; this design provides little information about the nature of the treatment process and the mechanisms of therapeutic change (Blatt, Shahar, & Zuroff, 2002). Likewise, TAU is also usually a weak control because most studies using a TAU control group provide relatively little information about the treatment offered by individual clinicians in the community. Additionally, a number of important collateral differences often exist between treatment provided within a research context by members of a research team and treatment provided by a clinician in solo practice in the community beyond the type of the treatment provided. Participation in a research team usually involves a degree of commitment, enthusiasm, a social support structure, and an organization that facilitates communication among the participating clinicians that are usually lacking for the clinician in solo practice in the community. Given these limitations of waiting lists and of TAU as possible controls in psychotherapy research, a more effective and possibly more powerful research design is to compare active treatments, but findings from comparative treatment trials are usually equivocal about the superiority of particular forms of treatment. Extensive meta-analyses of comparative treatment trials often indicate relatively few differences in therapeutic efficacy and effectiveness among various forms of active treatment (e.g., American Psychiatric Association, 1982, Frank, 1979, Shapiro & Shapiro, 1982, Smith et al., 1980 and Wampold, 2001). This lack of significant differences among active treatments is so frequent that it has been labeled, from Alice in Wonderland, the Dodo bird effect (Luborsky et al., 2002 and Luborsky et al., 1975) in which the declaration is made that “Everyone has won and all must have prizes.” Additionally, the recent identification of the impact of “allegiance effects” in influencing outcome (Luborsky et al., 1999), again possibly through the effects of enthusiasm and commitment, raises further complications about many of the findings in the comparison of active treatments. Thus, while a few types of treatment have been identified as effective in reducing some specific symptoms, most studies report a functional equivalence among various forms of treatment. This functional equivalence suggests that either our research methods are insensitive to differences among various forms of treatment (e.g., Kazdin, 1986, VandenBos & Pino, 1980 and Wortman, 1983) or that these various forms of treatment share common processes (e.g., Frank, 1982 and Strupp & Binder, 1984) that make them functionally equivalent (Lambert et al., 1986 and Stiles et al., 1986). These shared processes, including the therapeutic alliance, are often referred to as “non-specific effects.”

The frequent findings of fundamental equivalence among different types of active treatment, and the conclusion that the shared “non-specific” or common processes may have an important role in determining therapeutic outcome in various types of treatment suggests that a necessary, if not vital, prior step to the identification of empirically supported treatments is a fuller understanding of the factors and processes that contribute to therapeutic change (Blatt et al., 2002). In addition to examining various dimensions that might influence therapeutic outcome, it also seems important to evaluate dimensions of therapeutic change beyond symptom reduction—to develop broad evaluations of the efficacy and effectiveness of various forms of therapeutic intervention (Blatt et al., 2002).

Prior research indicates that several dimensions, in addition to the nature of the particular therapeutic method or technique, may make important contributions to the therapeutic process and facilitate therapeutic change including: 1) the quality of the therapeutic relationship, 2) patient pretreatment characteristics, and 3) characteristics of the therapist. Additionally, the evaluation of treatment efficacy and effectiveness should be based on assessments of dimensions of therapeutic change in addition to symptom reduction, including the reduction of vulnerability and the development of capacities for adaptation both at termination of treatment and in extended follow-up assessments. Thus, the goal of this paper is to integrate our various analyses of the data from within a single comprehensive data set, the TDCRP, in order to evaluate the relative contributions of different aspects of the treatment process (the type of treatment, the quality of the therapeutic relationship, and patient pretreatment characteristics) to therapeutic change as assessed on several measures of therapeutic progress (reduction of symptoms, reduction of vulnerability, and the development of resilience to stressful events).

1. The NIMH-sponsored TDCRP

In the early 1980s, under the leadership of Dr. Gerald Klerman, NIMH commissioned a major research program to evaluate definitively the brief outpatient treatment of serious depression. The NIMH-sponsored TDCRP was a very sophisticated, comprehensive, well-designed, and carefully conducted randomized clinical trial that assessed a multitude of dimensions of patient, therapist and the treatment process prior to, during, and subsequent to 16 weeks of two manual-directed forms of psychotherapy (CBT and IPT), an antidepressant medication ((Imipramine, the antidepressant of choice at that time3) plus clinical management (IMI-CM), and a double-blind placebo, also with clinical management4 (PLA-CM). Because the TDCRP research program was initiated by NIMH, the principal investigators (Irene Elkin, Morris Parloff, Tracie Shea, and John Docherty), all senior members of the NIMH extramural staff at the time of the investigation, were able to include a large and diverse range of assessments of the patients and of the treatment process at various points in the treatment and follow-up, much more so than might ordinarily be possible in the typical research grant. Also, because the TDCRP was an NIMH initiated research program, the extensive empirical data became available to the scientific community after the primary investigators (the NIMH staff and the research staff at the three participating clinical research sites) had sufficient time to explore these data. So in 1994, some ten years after the initiation of this very elaborate and very diverse research program, the TDCRP data set became available to the scientific community.

In each of three medical centers at the University of Oklahoma, University of Pittsburgh, and George Washington University, approximately 80 patients5 who met well-specified criteria for severe depression, were randomly assigned to one of 4 treatment protocols (CBT, IPT, IMI-CM, and PLA-CM). Twenty-eight well trained and experienced therapists in these three sites (18 psychiatrists and 10 Ph.D. clinical psychologists), with an average of over 11 years of clinical experience, participated in this study. These therapists were carefully selected and supervised throughout the treatment to assure adherence to the treatment protocols in which they were participating.6 Extensive clinical evaluations were conducted at each site by a Ph.D. level clinical evaluator (CE) prior to treatment, every 4 weeks during the treatment process, and again at 6, 12 and 18 months following the termination of treatment.7

Several facets of the therapeutic process were systematically assessed during treatment in the TDCRP that enable us to compare their relative contribution to the therapeutic change. Additionally, since extensive assessments were made of therapeutic progress during treatment, at termination, and at the three follow-up evaluations, we were able to compare therapeutic outcome across a number of different dimensions in addition to symptom reduction.

1.1. Factors that could contribute to therapeutic change

The design of the TDCRP enabled us to evaluate three different factors that might have contributed to therapeutic change:

1. Type of treatment

2. The quality of the therapeutic relationship

3. Patients' pretreatment personality characteristics.

1.1.1. Type of treatment

Patients were randomly assigned to 16 weeks of manual directed CBT8, IPT, IMI-CM, or PLA-CM. These treatments were designed and supervised by recognized authorities in each of the three treatment modalities. All treatment sessions were video recorded.9

1.1.2. The quality of the therapeutic relationship

The Barrett-Lennard Relationship Inventory (B-L RI; Barrett-Lennard, 1962) was included in the TDCRP protocol to assess aspects of the therapeutic relationship both at the beginning (after the second treatment session) and at the end of treatment (Elkin, 1994 and Elkin et al., 1989). The B-L RI is based on Carl Rogers, 1951, Rogers, 1957 and Rogers, 1959 concepts that the therapist's empathic understanding, unconditional positive regard, and congruence with the patient are the necessary and sufficient conditions for therapeutic change. On the basis of these formulations, Barrett-Lennard (1962) developed the four scales of the B-L RI—Empathic Understanding, Level of Regard, Unconditionality of Regard, and Congruence—to assess the patient's experience of the therapeutic relationship. The B-L RI presents 64 items, 16 for each of the four scales, with each item as a 6-point, anchored scale. These four scales were found to predict therapeutic change and to be related significantly to independent estimates of the therapist's competence (Barrett-Lennard, 1962). Several research reviews (e.g., Barrett-Lennard, 1986, Gurman, 1977a and Gurman, 1977b) indicate acceptable levels of reliability, internal consistency, and validity for the four scales of the B-L RI.

To identify possible dimensions within the four scales of the B-L RI, Blatt, Zuroff, Quinlan and Pilkonis (1996) performed a principal-components factor analysis on the scores for the four B-L RI four scales obtained in the TDCRP after the second therapy session. A single factor was extracted with an eigenvalue of 2.70, accounting for 67.39% of the common variance; all subsequent factors failed to exceed an eigenvalue of 1.0. Three of the scales (Empathy, Positive Regard, and Congruence) loaded at very high levels on this factor (.93, .87 and .92, respectively), and the fourth, Unconditionality of Regard, loaded only at a moderate level (.49). In our analyses, we assessed the B-L RI using a scaled score derived from summing the three component scales that loaded highly (> .80) on this single factor. Fig. 1 presents items illustrating these 3 scales that comprise the BL-RI.


Enlarge Image
(20K)

Fig. 1. Assessment of the therapeutic relationship.

1.1.3. Patient pretreatment characteristics

All patients had been give the Dysfunctional Attitudes Scale (DAS: Weissman & Beck, 1978) as part of the intake procedure and as part of the monthly (including termination) evaluations conducted during treatment. The DAS, a 40-item inventory consisting of items expressing typical depressogenic attitudes, was designed to measure cognitive vulnerability to depression (Oliver & Baumgart, 1985 and Weissman & Beck, 1978). The DAS has high internal and test–retest reliability (Dobson & Breiter, 1983 and Weissman & Beck, 1978), and is sensitive to severity of depression (Dobson & Shaw, 1986 and Weissman & Beck, 1978). It had been included in the TDCRP primarily as an outcome measure, especially to assess therapeutic change that might occur in CBT.

Imber et al. (1990) conducted a factor analysis on the initial, pretreatment, administration of the DAS in the TDCRP and identified two robust primary factors in a principal-components analysis with varimax rotation (done alternatively with three-, four-, and five-factor solutions). The two factors that emerged, labeled Need for Approval and Perfectionism, were highly congruent with prior factor analyses of the DAS in other samples (e.g., Cane et al., 1986 and Oliver & Baumgart, 1985). As in other samples, these two factors of the DAS in the TDCRP data had acceptable internal consistency (.91 and .82, respectively). Fig. 2 presents illustrative items that load highly on one of these two DAS factors.


Enlarge Image
(15K)

Fig. 2. Patients' pretreatment characteristics in the TDCRP.

The differentiation of these two primary factors, Need for Approval and Perfectionism, within the DAS is consistent with the differentiation of issues of relatedness and self-definition (or agency and communion) as two fundamental dimensions in personality theory (e.g., Angyal, 1941, Angyal, 1951, Bakan, 1966, Blatt, 1990, Blatt & Blass, 1990, Blatt, in press, Blatt & Blass, 1996, Freud, 1957, Loewald, 1962 and Wiggins, 1991) and in psychopathology (Blatt & Shichman, 1983). The dimensions of relatedness and self-definition, for example, have enabled investigators from different theoretical orientations (e.g., Arieti & Bemporad, 1978, Arieti & Bemporad, 1980, Beck, 1983, Blatt, 1974, Blatt, 1998, Blatt, 2004, Blatt et al., 1979, Blatt et al., 1976, Blatt et al., 1982, Bowlby, 1988a and Bowlby, 1988b) to identify two unique dimensions of depression (Blatt & Maroudas, 1992)—an anaclitic (dependent, dominant-other, or sociotropic) depression centered on feelings of loneliness, abandonment, and neglect and an introjective (self-critical, dominant-goal, or autonomous) depression focused on issues of self-worth and feelings of failure and guilt. Extensive empirical investigation (see summaries in Blatt, 2004, Blatt & Zuroff, 1992 and Luyten, 2002) indicates consistent differences in the life experiences (both current and early) of these two types of depressed individuals (Blatt & Homann, 1992), as well as major differences in their basic character style and the clinical expression of their depression (Blatt & Zuroff, 1992).

The differentiation between individuals preoccupied with issues of relatedness and with issues of self-definition has also enabled investigators to identify an empirically derived structure for integrating the diversity of personality disorders described in Axis II of DSM-IV. Systematic empirical investigation of outpatients (Morse et al., 2002 and Ouimette et al., 1994) and of inpatients (Levy et al., 1995) found that the various personality disorders can be organized into two primary configurations—one organized around issues of relatedness and the other around issues of self-definition. Ouimette et al. and Morse, Robins, and Gittes-Fox with outpatients, and Levy et al with inpatients, found that dependent, histrionic, and borderline personality disorders (anaclitic patients; Blatt & Shichman, 1983) had significantly greater preoccupation with issues of relatedness than with issues of self-definition. Conversely, individuals with paranoid, schizoid, schizotypic, antisocial, narcissistic, avoidant, obsessive–compulsive, and self-defeating personality disorders (introjective patients; Blatt & Shichman, 1983) had significantly greater preoccupation with issues of self-definition than with issues of relatedness (Blatt & Levy, 1998).10

Thus, the fundamental polarity of relatedness and self-definition has facilitated the differentiation of two primary configurations of psychopathology—anaclitic and introjective—based on differences between an excessive preoccupation with issues of relatedness and an excessive focus on issues of self-definition (Blatt, 1990, Blatt, 1995a and Blatt & Shichman, 1983). Prior research (e.g., Blatt, 1992 and Blatt & Ford, 1994, and Blatt and Shahar (2004) demonstrated the differential role of these two forms of psychopathology (anaclitic and introjective) in the therapeutic process and outcome in long-term, intensive treatment.

Because of these theoretical considerations and empirical findings, Blatt et al. (1995) used the differentiation of these two factors in the DAS (Need for Approval and Perfectionism), assessed prior to the beginning of treatment in the TDCRP, as the pre-dispositional personality variables in our analyses of the TDCRP data to evaluate the P–T and P–O interactions in the four therapy conditions in the TDCRP. It is important to note that extensive empirical evidence (see summaries in Blatt, 1995b, Blatt, 2004, Blatt & Zuroff, 1992 and Flett & Hewitt, 2002) indicates that perfectionism, or introjective personality dimensions, can be a particularly disruptive and pernicious personality characteristic.

1.2. Assessment of therapeutic change

The design of the TDCRP also allowed us to identify three different measures of therapeutic gain:

1. Symptom reduction

2. Reduction of vulnerability to depression

3. Development of resilience: Enhanced Adaptive Capacity (EAC).

1.2.1. Assessment of symptom reduction

The TDCRP utilized both interview and self-report assessments of depression and general clinical functioning throughout treatment and the follow-up evaluations. Interview assessments were conducted by a Ph.D. level clinical evaluator (CE) at each of the three treatment sites. Depression was assessed in a clinical interview utilizing the Hamilton Rating Scale for Depression (HRS-D) and by self-report using the Beck Depression Inventory (BDI). Level of general clinical functioning was assessed in a clinical interview utilizing the Global Assessment Scale (GAS) and by self-report using the Hopkins Symptom Checklist (HSCL-90). In addition, social adjustment was assessed in an interview using the Social Adjustment Scale (SAS; Paykel, Weissman, & Prusoff, 1978). Though the original TDCRP investigators used the HRS-D as their primary outcome measure (e.g., Watkins et al., 1993), we (Blatt, Zuroff, Quinlan, &, Pilkonis, 1995) found that the residualized gain scores at termination of these five outcome measures (HRS-D, BDI, GAS, HSCL-90, and SAS) were highly inter-correlated and each loaded substantially (> .70) on a common factor. Converting these measures to standard scores, we constructed a composite maladjustment score based on all five of these outcome measures. Thus, this composite maladjustment score reflects an overall level of clinical functioning with a particular emphasis on depressive symptoms.

1.2.2. Assessment of vulnerability to depression

Dysfunctional Attitude Scale (DAS)

a. Need for approval (NFA)

b. Perfectionism (PFT).

As discussed earlier, the DAS assesses two types of vulnerability to depression (Cane et al., 1986, Oliver & Baumgart, 1985 and Weissman & Beck, 1978). Extensive research (see summaries in Blatt, 1995b and Blatt, 2004, Blatt and Zuroff (1992), and Luyten (2002)) indicates that the PFT factor of the DAS assesses a particularly powerful vulnerability to depression.

1.2.3. Assessment of resilience (Enhanced Adaptive Capacity—EAC)

In the follow-up assessments, conducted 6, 12, and 18 months after termination of treatment, patients were also asked to rate 8 questions on 7-point Likert scales:

The degree to which their treatment improved:

a. Their interpersonal skills

b. Their ability to recognize their symptoms of depression

c. Their interpersonal relationships

d. Their ability to control the impact of negative thoughts

e. Their ability to cope with their symptoms of depression

f. Their attitudes associated with depression

g. Their understanding of the effects of their rigid attitudes toward themselves on their depression

h. Their understanding of the impact of close relationships on their depression.

Blatt, Zuroff, Bondi, and Sanislow (2000) found that the responses to these 8 items were highly correlated and formed a single factor, which Zuroff, Blatt, Krupnick, and Sotsky (2003) called Enhanced Adaptive Capacity (EAC). In contrast to the other two outcome dimensions (symptoms and vulnerability), however, these 8 items had not been used in prior research, so no evaluations were available on the psychometric properties or the validity of this scale. To address these issues, Zuroff et al. (2003) examined the impact of EAC scores obtained at the 6-month follow-up assessment on patients' reactivity to the stress they reported encountering during the follow-up period (at the 6, 12 and 18 month assessments). In each of the follow-up assessments, patients had been asked to report the degree of stress they had experienced in the prior 6 months. Zuroff et al. (2003) examined the impact of the level of reported stress on the level of depression, as measured by HRS-D, at the 6, 12 and 18 month follow-up assessments for patients with lower or higher levels of EAC at 6 months. They found a significant (p < .01) effect indicating that patients with lower EAC at the 6-month assessment had greater reactivity to stress (i.e., higher levels of depression, as assessed by the HRS-D) throughout the remainder of the follow-up period than patients with higher EAC scores at the 6-month follow-up. Thus, the EAC scale appears to measure a capacity to be stress-resilient.

In summary, our overall strategy for integrating these various analyses of the data from the TDCRP was to assess the relative contributions of three dimensions of the treatment process in the TDCRP—the type of treatment, the quality of the therapeutic relationship early in the treatment process, and patients' pretreatment personality characteristics to therapeutic change as assessed in the TDCRP by three dimensions of clinical change—symptom reduction (the composite maladjustment score), reduction of vulnerability to depression (the NFA and PFT factors of the DAS), and the development of resilience (EAC).

2. Integration of further analyses of the TDCRP data set

The presentation of findings from our various analyses of the TDCRP data will examine, in turn, the impact of each of the three factors of the therapeutic process (type of treatment, quality of the therapeutic relationship, and patients' pretreatment characteristics) on the three measures of therapeutic gain (symptom reduction, reduction of vulnerability, and the development of resilience). Most of these analyses of the TDCRP data set were conducted on the 162 patients who were defined as “treatment completers” (having participated in at least 12 of the 16 treatment sessions). Occasionally, statistical techniques (e.g., Random Regression Modeling, RRM) enabled us to also include data from all available patients in addition to those who were defined as treatment completers. This occasional increase in sample size is noted in text.

2.1. Impact of type of treatment on therapeutic outcome as assessed by three measures of therapeutic outcome

2.1.1. Symptom reduction

As discussed earlier, five primary measures (HRS-D, BDI, GAS, HSCL-90, and SAS) had been used in the TDCRP to assess therapeutic gain. In our analyses, we (Blatt et al., 1996 and Blatt et al., 1996) found that the residualized gain scores of these five outcome variables at termination were highly inter-correlated and each loaded substantially on a common factor. As also noted earlier, we standardized these five variables and combined them into a composite maladjustment measure that we used in our various analyses of the TDCRP data. Because these measures had been obtained at intake, every four weeks thereafter until termination at the end of 16 weeks of manual-directed treatment, and at the three follow-up assessments, the decline of this composite measure of symptom severity could be tracked throughout the treatment process.

Fig. 3 presents this composite symptom severity measure for patients with complete data at the 5 assessment points through the treatment process for the 4 treatment conditions (CBT, IPT, IMI-CM, PLA-CM) in the TDCRP. Using Repeated Measures Anova, Zuroff and Blatt (unpublished data) found that IMI-CM resulted in a significantly (p < .01) more rapid decline in symptoms at mid-treatment than the two psychotherapy treatment conditions. These results are consistent with findings reported earlier by Elkin et al. (1996) based on their RRM analyses of therapeutic change with only the Hamilton Rating Scale for Depression (HRS-D). Over the second 8 weeks of treatment, the linear trend for the IPT group became significantly greater than for the IMI-CM group. In other words, as Fig. 3 suggests, the IPT group showed greater improvement over the second half of treatment, while the IMI-CM group remained essentially on a plateau. Thus, at termination, no significant differences were found in the reduction of symptoms among the three active treatment conditions (Blatt, Zuroff et al., 1998). Though IMI-CM and IPT, but not CBT, had a significantly (p < .05) greater reduction in symptoms than the placebo group (PLA-CM) at termination, no significant differences were found in degree of symptom reduction at termination among the three active treatments (these findings are consistent with the findings reported earlier by Watkins et al. (1993) on analyses based only on the HRS-D).11 Even further, no significant differences were found in symptom reduction among all 4 treatment groups (including placebo) at the 18-month follow-up evaluation (Blatt, Zuroff et al., 1998). Thus, in terms of symptom reduction, though medication had a more rapid effect in terms of symptom reduction, no significant differences in degree of symptom reduction were found among the three active treatment conditions at termination or at follow-up.12


Enlarge Image
(35K)

Fig. 3. Composite outcome during 16 weeks of treatment for four treatment conditions.

2.1.2. Reduction of vulnerability

No significant differences were found among the three active treatments in the reduction of the level of Perfectionism (PFT) at termination or at follow-up, though CBT was somewhat more effective (p < .05) than IPT in reducing Need for Approval (NFA) at termination (Imber et al., 1990).

2.1.3. Development of enhanced adaptive capacities (EAC)

Fig. 4 presents the level of EAC at the 18-month follow-up assessment and illustrates a significantly (p < .01) greater level of EAC in the two psychotherapy conditions in contrast to both IMI-CM and placebo, as well as the lack of a significant difference between these two medication conditions (Blatt et al., 2000). The same essential findings were obtained at the two earlier follow-up assessments at 6 and 12 months following termination (Zuroff et al., 2003).13


Enlarge Image
(27K)

Fig. 4. Enhanced adaptive capacities as a function of treatment condition.

2.1.4. Summary of difference among treatment conditions

Medication led to more rapid symptom reduction, but no significant differences were found among the three active treatment conditions at termination or at follow-up, findings consistent with the frequently cited Dodo bird effect in psychotherapy research. No significant differences were found among the three active treatment conditions in the reduction of Perfectionism (PFT), a major vulnerability factor, but CBT was significantly more effective than the other treatment conditions in reducing Need for Approval (NFA). The two psychotherapy conditions (CBT and IPT) were significantly more effective than medication in facilitating the development of Enhanced Adaptive Capacities in the follow-up assessments. In summary, while some significant treatment effects were identified, the results are mixed. Some findings are supportive of medication, some supportive of psychotherapy. Short-term effects appear to favor medication; long-term effects appear to favor psychotherapy.14

2.2. Impact of therapeutic relationship as assessed on three measures of therapeutic outcome15

Discussion in the literature (e.g., Beck et al., 1979 and Kazdin, 2002) has questioned the frequent findings about the importance of the therapeutic alliance in determining treatment outcome because of the possibility that the apparent contributions of the therapeutic relationship to treatment outcome may be confounded by the degree of symptomatic improvement. That is, the more effective treatment has been in reducing symptoms, the greater the likelihood that the patient will view more favorably the therapist and the therapeutic relationship. Thus, the evaluation of the role of the contribution of the therapeutic relationship to treatment outcome requires statistical control of the degree of concurrent symptomatic improvement (see Barber et al., 2000 and Klein et al., 2003). Using Random Regression (Multilevel) Modeling (RRM), we (Zuroff & Blatt, unpublished data) controlled statistically for the concurrent level of clinical improvement at the time of the assessment of the quality of the therapeutic relationship. When indicated, we also controlled for the concurrent level of clinical improvement in the assessment of other independent variables at the various assessment points. These are very conservative analyses.

2.2.1. Impact of therapeutic relationship on symptom reduction

As illustrated in Fig. 5, the patient's experience of the therapeutic relationship early in the treatment process (as assessed by the B-L RI at the end of the 2nd treatment session) was significantly (p < .01) related to the degree of symptom reduction across all four treatment conditions, even after controlling statistically for the degree of clinical improvement at the time of the assessment of the quality of the therapeutic relationship (Zuroff & Blatt, unpublished data).


Enlarge Image
(5K)

Fig. 5. Maladjustment as a function of quality of therapeutic relationship (B−L RI).

2.2.2. Impact of therapeutic relationship on the reduction of vulnerability

In assessing the impact of the quality of the therapeutic relationship at the end of the second treatment session on the reduction of factors that create a vulnerability to depression (DAS Need for Approval and Perfectionism), we controlled statistically for the level of symptom reduction at the time of the assessment of the quality of the therapeutic relationship as well as at each time of the assessment of vulnerability; that is, controlling for the concurrent level of symptom severity. As illustrated in Fig. 6 and Fig. 7, patients who reported experiencing a more constructive relationship with their therapist at the end of the second treatment session (higher scores on the B-L RI) had significantly greater decreases (p < .05) in Need for Approval as well as in level of Perfectionism (p < .01) during the treatment process than patients with a less constructive therapeutic relationship (Zuroff & Blatt, unpublished data).16


Enlarge Image
(5K)

Fig. 6. Need for approval as a function of quality of therapeutic relationship (B-L RI), controlling for effects of decreases in maladjustment.


Enlarge Image
(4K)

Fig. 7. Perfectionism as a function of quality of therapeutic relationship (B-L RI), controlling for effects of decreases in maladjustment.

2.2.3. Impact of therapeutic relationship on development of enhanced adaptive capacities

In evaluating the impact of the therapeutic relationship on EAC in the follow-up, we again controlled statistically for the level of symptomatic improvement at the time of the assessment of the therapeutic relationship, as well as the level of symptoms at the time of assessment of EAC. As illustrated in Fig. 8, patients who reported experiencing a more constructive relationship with their therapist at the end of the second treatment session, had significantly greater levels of EAC at all three follow-up assessments. Thus, the quality of the therapeutic relationship assessed early in the treatment process was significantly (p < .001) related to the development of Enhanced Adaptive Capacities (EAC) at the follow-up assessments.


Enlarge Image
(31K)

Fig. 8. Enhanced adaptive capacities (EAC) as a function of the quality of the therapeutic relationship (B-L RI).

2.2.4. Summary of the impact of therapeutic relationship on therapeutic outcome

The quality of the therapeutic relationship assessed at the end of the second treatment session was significantly related to all three outcome measures—to symptom reduction, reduction of vulnerability (Perfectionism), and the development of resilience (EAC). These findings are consistent with the recent emphasis on the importance of the therapeutic relationship in the treatment process (e.g., Norcross, 2002 and Wampold, 2001).

To evaluate further how the quality of the therapeutic relationship facilitated therapeutic outcome, we (Blatt, Sanislow et al., 1996) differentiated more effective therapists from less effective therapists by aggregating therapeutic outcome, as measured by change on the composite maladaptive adjustment score at termination, of the patients each therapist had seen in active treatment (i.e., patients seen in the placebo condition were omitted from these analyses). The twenty-eight therapists participating in the TDCRP had completed two brief questionnaires describing their general clinical practice outside their participation in this study, particularly the percentage of time they usually devoted to providing psychotherapy alone, medication alone, and a combination of both medication and psychotherapy. More effective therapists, those therapists whose patients had greater reduction in the composite maladjustment scale at termination in the TDCRP, reported that they usually devoted significantly more of their practice to psychotherapy alone than did the less effective therapists. Also, more effective therapists placed a significantly greater emphasis on a psychological rather than a biological approach to the understanding and treatment of depression. More effective therapists, for example, to a significantly greater extent than less effective therapists thought that the treatment of depression required more time and they eschewed biological interventions (medication and ECT) in the treatment of depression (Blatt, Sanislow et al., 1996).17

2.3. Impact of patient pretreatment personality characteristics (level of DAS perfectionism) on three measures of therapeutic outcome18

2.3.1. Symptom reduction

Blatt, Zuroff et al., 1998 found that, as indicated in Fig. 9, pretreatment level of perfectionism significantly (p < .01) impeded therapeutic response in two-thirds of the patients in this study, beginning at mid-treatment (at the 8th treatment session). Patients with higher and moderate levels of pretreatment PFT, as compared to those with low levels of PFT, made no further significant therapeutic gain after the 8th treatment session. Only patients lower in pretreatment PFT continued to make significant (p < .001) therapeutic progress during the latter half of the treatment process (Blatt, Zuroff et al., 1998). Pretreatment levels of PFT were also significantly (p < .05) related to impeded therapeutic progress in the three follow-up assessments at 6, 12 and 18 months after termination (Blatt, 1999 and Blatt et al., 2000).19


Enlarge Image
(21K)

Fig. 9. Outcome in the TDCRP as a function of perfectionism. This figure appeared in Blatt, Zuroff et al., 1998.

The findings of significant impairment of therapeutic gain in the latter half of the treatment process of patients with elevated pretreatment PFT is consistent with theoretical formulations and empirical findings (e.g., Blatt, 1974, Blatt, 1998, Blatt, 2004, Blatt & Ford, 1994, Blatt & Shahar, 2004 and Blatt & Shichman, 1983) about introjective patients–individuals dominated by concerns about self-definition, self-control, and self-worth, and who tend to be over-ideational and preoccupied with issues of control. Because introjective patients (patients with elevated PFT) are very concerned about issues of control and self-worth, as they approach termination they may be more likely to have negative reactions to the imposition of an externally defined termination date and may be more likely to be dissatisfied with the progress they have made. As suggested by Seligman (1995) in his analyses of data gathered as part of a survey conducted by Consumers Report in 1995, many patients appear to respond more constructively when they feel that they have participated in the decision to terminate their treatment than when treatment is arbitrarily terminated by an external authority. One would expect this observation might be particularly true of introjective patients who are especially concerned about issues of control and who are very evaluative of their performance and accomplishments.

It is noteworthy that while introjective patients (patients with higher perfectionism) did significantly less well in brief outpatient treatment in the TDCRP, other investigations (Blatt, 1992, Blatt & Ford, 1994 and Blatt & Shahar, 2004) found that introjective patients are relatively more responsive to long-term, intensive, insight-oriented treatment (see also Fonagy et al., 1996 and Gabbard, 1995).20

2.3.2. Development of the therapeutic alliance

In an attempt to understand more fully the mechanisms through which pretreatment PFT impeded therapeutic progress in brief treatment in the TDCRP, we gained access to the ratings made by Janice Krupnick and colleagues (Krupnick et al., 1996) of the video recordings of the 3rd, 9th, and 15th treatment sessions of the TDCRP. Using a modification of the Vanderbilt Therapeutic Alliance Scale (VTAS; Hartley & Strupp, 1983), Krupnick et al. (1996) reliably rated the degree to which the patient and therapist contributed to the development of the therapeutic alliance.21 They found that the contributions of the patients, but not the therapists, to the therapeutic alliance were significantly related to the level of therapeutic outcome at termination. As illustrated in Fig. 10, we (Zuroff et al., 2000) found that pretreatment perfectionism significantly (p < .01) interfered with patients' participation in the therapeutic alliance in the latter half of the treatment process (beginning at the 9th treatment session). Consistent with the findings that perfectionism generally disrupts the quality of interpersonal relationships (e.g., Mongrain, 1998, Priel & Shahar, 2000, Shahar & Priel, 2003, Shahar et al., 2003 and Zuroff & Duncan, 1999), patients with elevated pretreatment levels of perfectionism (PFT) were less involved in the therapeutic process as they approached termination.


Enlarge Image
(15K)

Fig. 10. Patient alliance as a function of perfectionism. This figure appeared in Zuroff et al., 2000.

Path analyses indicated that while patients' increased engagement in the therapeutic alliance significantly (p < .01) mediated the impact of pretreatment perfectionism on therapeutic outcome, a substantial portion of the variance of the impact of pretreatment PFT on therapeutic outcome was still unaccounted for (Zuroff et al., 2000). Based on the psychoanalytic conception of transference, we (Shahar, Blatt, Zuroff, Krupnick, & Sotsky, 2004) explored whether a parallel existed between the disengagement from the treatment process of patients with elevated levels of PFT and their disengagement from their social network, external to the treatment process. Data had been collected as part of the TDCRP assessing the breadth and intensity of patients' available social support system at intake and following treatment (Elkin et al., 1985). Shahar et al. (2004), using this assessment of the breadth and intensity of patients' interpersonal relationships at intake and termination, found, as indicated in Fig. 11, that the impact of pretreatment PFT on treatment outcome was mediated equally by the degree of patients' disengagement in both of these social contexts—in both therapy and their social network. These two factors accounted for much of the variance of the negative impact of pretreatment perfectionism on therapeutic outcome.22 Also as indicated in Fig. 11, patients' constructive involvement in the therapeutic alliance early in treatment (as assessed by Krupnick et al. (1996) in the 3rd treatment session) was significantly related to the quality of patients' interpersonal relationships at termination. Thus, patients' capacity to engage in the therapeutic process early in the treatment process appears to reflect their general capacity to establish and maintain interpersonal bonds (see Andersen & Miranda, 2000, for a further discussion of transference in everyday life from a social-cognitive perspective [Shahar et al., 2004]).


Enlarge Image
(38K)

Fig. 11. Mediation of therapeutic outcome by patient alliance and perceived social support. This figure appeared in Shahar et al., 2004.

2.3.3. The development of resilience

As illustrated in Fig. 12, patients' pretreatment levels of perfectionism significantly interfered with the development of Enhanced Adaptive Capacity (EAC) at all three follow-up assessments.


Enlarge Image
(36K)

Fig. 12. Impact of pretreatment perfectionism on follow-up EAC.

The negative impact of PFT on the development of resilience was evident not only in patients' self-report of the impact of treatment on their development of EAC in the follow-up assessments, but it was also apparent as a moderating effect on stress reactivity in the follow-up assessments. The TDCRP investigators had assessed levels of stress during the follow-up period and, as illustrated in Fig. 13, patients with higher levels of PFT at termination were significantly (p < .01) more vulnerable to stressful life events than patients with lower levels of PFT at termination. Increasing levels of stress resulted in significant increases in level of depression, as measured by the HRS-D, for patients with higher levels of PFT at termination; while patients with lower levels of PFT at termination, in contrast, seemed relatively impervious to the occurrence of stressful life events (Zuroff & Blatt, 2002).


Enlarge Image
(24K)

Fig. 13. Impact of termination perfectionism on stress reactivity in follow-up assessment. This figure appeared in Zuroff and Blatt, 2002.

Thus, both pre-treatment and termination levels of perfectionism significantly interfered with the development of Enhanced Adaptive Capacities (EAC) and the ability to cope with stressful life events in the 18-month period following the termination of treatment (Zuroff & Blatt, 2002).

2.3.4. Summary of effects of patients' pretreatment characteristics on therapeutic outcome

Pretreatment perfectionism (PFT) significantly interfered with symptom reduction at termination and at follow-up. PFT also significantly interfered with the development of the therapeutic relationship in the latter half of the treatment process and with the ability to maintain and sustain a supportive social network. Disruption of interpersonal relationships in both treatment and in the broader social context significantly mediated the negative impact of pretreatment perfectionism on therapeutic progress. Pretreatment and termination levels of PFT also significantly impeded the development of resilience and the capacity to adapt to stressful life events in the 18 months following the termination of treatment.

In terms of the important role of the therapeutic relationship in treatment outcome as discussed earlier in Section 2.2, it is noteworthy that the negative impact of pretreatment perfectionism on therapeutic outcome is significantly moderated by the patient's perception of the quality of the therapeutic relationship at the end of the second treatment session as assessed by the B–L-RI (Blatt et al., 1996 and Blatt et al., 1996). As illustrated in Fig. 14, this buffering effect of the quality of the early therapeutic relationship on the negative impact of pretreatment perfectionism on treatment outcome was particularly apparent in the mid-range of perfectionism.


Enlarge Image
(35K)

Fig. 14. Impact of B-L RI at end of second treatment session on PFT and therapeutic outcome. This figure appeared in Blatt, Zuroff et al., 1996.

3. Discussion

The findings from our further exploration of data from the NIMH-sponsored TDCRP have important implications for the identification of empirically supported treatments as well as for clinical training and for clinical research.

Our detailed evaluation of data from the TDCRP does not support the two primary assumptions of contemporary efforts to identify empirically supported treatments—1) that the type of treatment is a primary factor in the therapeutic process and 2) that therapeutic gain is most effectively expressed in symptom reduction. Quite to the contrary, consistent with recent extensive literature reviews by Lambert and Barley (2002), and by Westen et al. (2004), our empirical findings indicate that primary among the mutative factors in the brief outpatient treatment for serious depression is the quality of the therapeutic relationship that the patient and therapist establish very early in the treatment process (at the second treatment session) and the pretreatment personality characteristics of the patients, especially their level of pretreatment perfectionism. Findings from our further analyses of data from the TDCRP also indicate that evaluation of therapeutic gain, in addition to symptom reduction, should include assessments of the reduction of the vulnerability to depression and the development of resilience as expressed in increased adaptive capacities in the ability to manage stressful life events. These findings raise serious questions about the validity of current efforts to identify empirically supported treatments by comparing different types of treatment in their relative efficacy and effectiveness in reducing symptoms. Our findings suggest that efforts to identify ESTs require a much more complex view of the treatment process, one that needs to include other dimensions of the treatment process, especially the quality of the therapeutic relationship and patients' pretreatment characteristics, and which investigates the impact of these factors of the therapeutic process across a range of outcome measures beyond symptom reduction. Consistent with formulations of Cronbach (e.g., 1953) about the importance of possible patient–treatment and patient–outcome interactions in psychotherapy research, Blatt et al. (2002) note that different types of patients may be responsive to different aspects of the treatment process, and different types of patients may respond to treatment in different ways) (see also Blatt & Shahar, 2004).

The findings that the type of treatment provided is not as important in influencing therapeutic change as the quality of the therapeutic relationship and the patients' pretreatment personality characteristics also have important implications for clinical training. These findings suggest that it is essential to focus training, not on learning particular manual-directed treatments for specific disorders, but on the development of fundamental clinical skills that will enable students to become competent clinicians capable of establishing effective therapeutic relationship with their patients as well as being able to identify personality characteristics of their patients that are likely to facilitate or impede their patients' response to various forms of treatment.

Patients' pretreatment personality factors, especially pretreatment levels of perfectionism, also have important implications for psychotherapy research. The findings from our further detailed analyses of the TDCRP data set indicate that psychotherapy research must go beyond the simple comparisons of the efficacy and effectiveness of particular forms of treatment in the reduction of specific symptoms. Rather, treatment research needs to address more complex questions like what kinds of treatments are effective, for what kinds of patients, in what kinds of ways, and through what kinds of mechanisms or processes of therapeutic change (Blatt & Shahar, 2004 and Blatt et al., 2002).

These findings from the TDCRP data also raise a number of important considerations about the role of randomized clinical trials (RCTs) in clinical research. Much of contemporary efforts to identify empirically supported treatments rely on RCTs as the gold standard for psychotherapy research. Randomization of patients to treatment groups clearly controls for a number of important factors that could influence treatment outcome. But our findings suggest that different types of patients may respond to participation in RCTs in different ways and that some of the observed differences among patients may be a function of the differential response of different types of patients to these experimental conditions rather than the consequence of different types of treatment. Patients with elevated perfectionism, patients who in other research contexts we (Blatt, 1992, Blatt & Ford, 1994, Blatt & Shahar, 2004, Blatt et al., 2002 and Blatt & Zuroff, 1992) have described as “introjective,” patients who are very concerned about issues of self-control and self-worth, appear to encounter greater difficulty toward the end of treatment. Two-thirds of the patients in the TDCRP, those with higher pretreatment scores on perfectionism, failed to make any substantial further gain after mid-treatment, as they began to approach termination. And this impaired therapeutic progress appears to be a function of their withdrawal from engagement in the treatment process. This disengagement from the treatment process in the latter half of the treatment process suggests that they may be responding negatively to feeling that they are dissatisfied with their therapeutic progress and with the imposition of an externally imposed, arbitrary, termination of the treatment process. These possible explanations of the disengagement of a large segment of the sample in the latter half of the treatment process need to be investigated systematically.

These issues about the possible unique reactions to the conditions of the treatment process by introjective patients, those patients with elevated pretreatment scores on perfectionism or self-criticism, also raise the possibility that the overall lack of progress of these perfectionistic patients in brief treatment for depression in the TDCRP may possibly be a function of their untoward reactions to being part of a research program in which the patients' progress was frequently evaluated.23 This hypothesis about the specific impact of the evaluative process in treatment research on introjective patients should also be studied systematically.

Though RCTs may provide the basis for establishing good internal validity, they may create problems in establishing external or ecological validity. Though random assignment to treatment groups can control for a number of potential experimental artifacts, it may also induce other experimental artifacts as a consequence of the meaning to the patients of participating in a design in which they feel controlled, observed, and evaluated beyond what would occur in a more natural treatment relationship. Randomized assignment to treatment groups which have fixed, predefined termination dates and involve continual evaluations conducted external to the therapeutic relationship may create conditions that could compromise the therapeutic progress of one particular group of patients. Even further, if one of the goals of treatment, in addition to the reduction of symptoms, is to enable individuals to assume responsibility for their lives—to develop a sense of agency—this sense of agency may be seriously compromised in a therapeutic experience in which the patients are essentially passive in the selection of the type of treatment they receive and in the decision when to terminate the treatment process.24 Thus, while RCTs may be considered the gold standard of experimental design, we must begin to evaluate the differential impact that these experimental conditions can have on different types of patients and come to appreciate more fully the importance of more naturalistic studies (Blatt, Berman et al., 1998) that attempt to evaluate the impact of treatment in ways that are more consistent with the natural occurrence of treatment in which patients seek assistance in therapeutic settings they select and in which they actively participate in determining who to see and when to terminate treatment—in settings in which patients can experience a sense of agency in the initiation as well as during and at the termination of the treatment process (Seligman, 1995).

Additionally, the focus on symptom reduction in many attempts to identify empirically supported treatments has frequently resulted in the selection of “purified” samples of patients with focal symptoms without co-morbid complications. As has been noted by Westen and colleagues (e.g., Westen et al., 2004), this type of research design may result in the selection and study of an atypical group of patients, patients who would be most unusual in a traditional range of patients applying for treatment at most clinical facilities. Even further, patients with a focal symptom without co-morbid complications may represent a sample of more treatment accessible and responsive patients, thereby inflating the estimates of the efficacy and effectiveness of the treatment, which might be much more limited when the particular therapeutic technique is used with patients from a more general clinical population. It is interesting, in this regard, to note that while the patients in the TDCRP were seriously depressed, patients with suicidal potential and with psychotic and manic tendencies were excluded from this investigation. Even with this somewhat restricted range of severely depressed patients, only approximately 20% of the patients in this study achieved full recovery (Elkin et al., 1989 and Shea et al., 1992). These findings eventually led the original TDCRP investigators (e.g. Elkin et al., 1989) to question the value of short-term treatment for depression.

In summary, the findings from the TDCRP suggest that we must use a variety of research designs, RCTs as well as more naturalistic designs, to assess the efficacy and effectiveness (Seligman, 1995) of different types of treatments in our efforts to identify empirically supported treatments. As in most efforts to study psychological processes, it is important to use systematic as well as representative research designs (Brunswick, 1956)—designs that provide considerable experimental control as well as designs that try to study the phenomena as closely as possible to ways they occur in more natural settings (Blatt, Berman et al., 1998).

In closing, we want to express our gratitude to the members of the TDRCP research team and to acknowledge our indebtedness for their establishing a remarkably comprehensive data set that went well beyond simply noting the type and degree of symptom reduction in different treatment conditions, but instead gathered extensive evaluations of the psychological characteristics of the patient as well as important measures of aspects of the therapeutic process. It was these more extended aspects of the data set that enabled us to conduct the analyses reported in this paper. We also want to acknowledge our indebtedness to NIMH for making available this extensive data set to the scientific community that allowed us, and other investigators, from different theoretical orientations and with different assumptions about the nature of the therapeutic process, to explore these different perspectives in a remarkably rich data set.


References

American Psychiatric Association, 1982 American Psychiatric Association Commission on Psychotherapies, Psychotherapy research: Methodological and efficacy issues, American Psychiatric Institute, Washington, DC (1982).

Andersen & Miranda, 2000 S.M. Andersen and R. Miranda, Transference: How past relationships emerge in the present, Psychologist 13 (2000), pp. 608–609. Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Angyal, 1941 A Angyal, Foundations for a science of personality, Viking Press, New York (1941).

Angyal, 1951 A. Angyal In: E. Hanfmann and R.M. Jones, Editors, Neurosis and treatment: A holistic theory, Wiley, New York (1951).

Arieti & Bemporad, 1978 S. Arieti and J.R. Bemporad, Severe and mild depression: The therapeutic approach, Basic Books, New York (1978).

Arieti & Bemporad, 1980 S. Arieti and J.R. Bemporad, The psychological organization of depression, American Journal of Psychiatry 137 (1980), pp. 1360–1365. Abstract-EMBASE | Abstract-MEDLINE   | Abstract + References in Scopus | Cited By in Scopus

Bakan, 1966 D. Bakan, The duality of human existence: An essay on psychology and religion, Rand McNally, Chicago (1966).

Barber et al., 2000 J.P. Barber, M.B. Connolly, P. Crits-Cristoph, L. Gladis and L. Siqueland, Alliance predicts patients' outcome beyond in-treatment change in symptoms, Journal of Consulting and Clinical Psychology 68 (2000), pp. 1027–1032. Abstract | Abstract + References in Scopus | Cited By in Scopus

Barbui & Hotopf, 2001 C. Barbui and M. Hotopf, Amitriptyline v. the rest: Still the leading antidepressant after 40 years of randomized controlled trials, British Journal of Psychiatry 178 (2001), pp. 129–144. Abstract-EMBASE | Abstract-Elsevier BIOBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus

Barrett-Lennard, 1962 G.T. Barrett-Lennard, Dimensions of therapist responses as causal factors in therapeutic change, Psychological Monographs 76 (1962) (43, Whole No. 562).

Barrett-Lennard, 1986 G.T. Barrett-Lennard, The relationship inventory now: Issues and advances in theory, method, and use. In: L.S. Greenberg and W.M. Pinsof, Editors, The psychotherapeutic process: A research handbook, Guilford Press, New York (1986), pp. 439–476.

Beck, 1983 A.T. Beck, Cognitive therapy of depression: New perspectives. In: P.J. Clayton and J.E. Barrett, Editors, Treatment of depression: Old controversies and new approaches, Raven, New York (1983), pp. 265–290.

Beck et al., 1979 A.T. Beck, A.J. Rush, B.F. Shaw and G.L. Emery, Cognitive theory of depression, Guilford Press, New York (1979).

Blatt, 1974 S.J. Blatt, Levels of object representation in anaclitic and introjective depression, Psychoanalytic Study of the Child 29 (1974), pp. 107–157. Abstract-MEDLINE   | Abstract + References in Scopus | Cited By in Scopus

Blatt, 1990 S.J. Blatt, Interpersonal relatedness and self-definition: Two personality configurations and their implications for psychopathology and psychotherapy. In: J.L. Singer, Editor, Repression and dissociation: Implications for personality theory, psychopathology & health, University of Chicago Press, Chicago (1990), pp. 299–335.

Blatt, 1992 S.J. Blatt, The differential effect of psychotherapy and psychoanalysis on anaclitic and introjective patients: The Menninger Psychotherapy Research Project revisited, Journal of the American Psychoanalytic Association 40 (1992), pp. 691–724. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Blatt, 1995a S.J. Blatt, Representational structures in psychopathology. In: D. Cicchetti and S. Toth, Editors, Rochester symposium on developmental psychopathology, Emotion, Cognition, and Representation vol. 6, University of Rochester Press, Rochester, NY (1995), pp. 1–33. Abstract-PsycINFO  

Blatt, 1995b S.J. Blatt, The destructiveness of perfectionism: Implications for the treatment of depression, American Psychologist 50 (1995), pp. 1003–1020. Abstract | Abstract + References in Scopus | Cited By in Scopus

Blatt, 1998 S.J. Blatt, Contributions of psychoanalysis to the understanding and treatment of depression, Journal of the American Psychoanalytic Association 46 (1998), pp. 723–752. Abstract-EMBASE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Blatt, 1999 S.J. Blatt, Personality factors in the brief treatment of depression: Further analyses of the NIMH sponsored treatment for depression collaborative research program. In: D.S. Janowsky, Editor, Psychotherapy indications and outcomes, American Psychiatric Press, Washington, DC (1999), pp. 23–45. Abstract-PsycINFO  

Blatt, 2004 S.J. Blatt, Experiences of depression: Theoretical, clinical and research perspectives, American Psychological Association, Washington, DC (2004).

Blatt, in press Blatt, S. J. (in press). A fundamental polarity in psychoanalysis: Implications for personality development, psychopathology, and the therapeutic process. Psychoanalytic Inquiry.

Blatt & Auerbach, 1988 S.J. Blatt and J.S. Auerbach, Differential cognitive disturbances in three types of “borderline” patients, Journal of Personality Disorders 2 (1988), pp. 198–211. Abstract-EMBASE   | Abstract + References in Scopus | Cited By in Scopus

Blatt et al., 1998 S.J. Blatt, W.H. Berman, B.P. Cook and R.Q. Ford, Effectiveness of long-term, intensive, inpatient treatment for seriously disturbed young adults: A reply to Bein, Psychotherapy Research 8 (1998), pp. 42–53. Abstract-PsycINFO   | Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus

Blatt & Blass, 1990 S.J. Blatt and R.B. Blass, Attachment and separateness: A dialectic model of the products and processes of psychological development, Psychoanalytic Study of the Child 45 (1990), pp. 107–127. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Blatt & Blass, 1996 S.J. Blatt and R. Blass, Relatedness and self definition: A dialectic model of personality development. In: G.G. Noam and W.H. Fischer, Editors, Development and vulnerabilities in close relationships, Lawrence Erlbaum Associates, Hillsdale, NJ (1996), pp. 309–338. Abstract-PsycINFO  

Blatt et al., 1979 Blatt, S. J., D'Afflitti, J., & Quinlan, D. M. (1979). Depressive Experiences Questionnaire (DEQ). Unpublished research manual, Yale University.

Blatt et al., 1976 S.J. Blatt, J.P. D'Afflitti and D.M. Quinlan, Experiences of depression in normal young adults, Journal of Abnormal Psychology 85 (1976), pp. 383–389. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Blatt & Felsen, 1993 S.J. Blatt and I. Felsen, Different kinds of folks may need different kinds of strokes: The effect of patients' characteristics on therapeutic process and outcome, Psychotherapy Research 3 (1993), pp. 245–259. Abstract-PsycINFO  

Blatt & Ford, 1994 S.J. Blatt and R.Q. Ford, Therapeutic change: An object relations perspective, Plenum, New York (1994).

Blatt & Homann, 1992 S.J. Blatt and E. Homann, Parent–child interaction in the etiology of dependent and self-critical depression, Clinical Psychology Review 12 (1992), pp. 47–91. Abstract | Abstract + References in Scopus | Cited By in Scopus

Blatt & Levy, 1998 S.J. Blatt and K.N. Levy, A psychodynamic approach to the diagnosis of psychopathology. In: J.W. Barron, Editor, Making diagnosis meaningful, American Psychological Association Press, Washington, DC (1998), pp. 73–109. Full Text via CrossRef

Blatt & Maroudas, 1992 S.J. Blatt and C. Maroudas, Convergence of psychoanalytic and cognitive behavioral theories of depression, Psychoanalytic Psychology 9 (1992), pp. 157–190. Abstract-EMBASE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Blatt et al., 1982 S.J. Blatt, D.M. Quinlan, E.S. Chevron, C. McDonald and D. Zuroff, Dependency and self-criticism: Psychological dimensions of depression, Journal of Consulting and Clinical Psychology 50 (1982), pp. 113–124. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Blatt et al., 1995 S.J. Blatt, D.M. Quinlan, P.A. Pilkonis and T. Shea, Impact of perfectionism and need for approval on the brief treatment of depression: The National Institute of Mental Health Collaborative Research Program revisited, Journal of Consulting and Clinical Psychology 63 (1995), pp. 125–132. Abstract | Abstract + References in Scopus | Cited By in Scopus

Blatt et al., 1996 S.J. Blatt, C.A. Sanislow, D.C. Zuroff and P.A. Pilkonis, Characteristics of effective therapists: Further analyses of data from the NIMH TDCRP, Journal of Consulting and Clinical Psychology 64 (1996), pp. 1276–1284. Abstract | Abstract + References in Scopus | Cited By in Scopus

Blatt & Shahar, 2004 S.J. Blatt and G. Shahar, Das dialogische selbst: Adaptive und maladaptive dimesionen, The dialogic self: Adaptive and maladaptive dimensions. In: P. Giampieri-Deutsch, Editor, Psychoanlyse im dialog der wissenschaften. Band 2. Anglo-Amerikanishce perspektiven, Psychoanalysis as an Empirical, Interdisciplinary Science, An Anglo-American Perspective vol. 2, Kohlhammer, Stuttgart (2004), pp. 285–309.

Blatt et al., 2002 S.J. Blatt, G. Shahar and D.C. Zuroff, Anaclitic (sociotropic) and introjective (autonomous) dimensions. In: John C. Norcross, Editor, Psychotherapy relationships that work: Therapist contributions and responsiveness to patients, Oxford University Press, New York (2002), pp. 306–324.

Blatt & Shichman, 1983 S.J. Blatt and S. Shichman, Two primary configurations of psychopathology, Psychoanalysis and Contemporary Thought 6 (1983), pp. 187–254. Abstract-EMBASE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Blatt & Zuroff, 1992 S.J. Blatt and D.C. Zuroff, Interpersonal relatedness and self-definition: Two prototypes for depression, Clinical Psychology Review 12 (1992), pp. 527–562. Abstract | Abstract + References in Scopus | Cited By in Scopus

Blatt et al., 1998 S.J. Blatt, D.C. Zuroff, C.M. Bondi, C. Sanislow and P.A. Pilkonis, When and how perfectionism impedes the brief treatment of depression: Further analyses of the NIMH TDCRP, Journal of Consulting and Clinical Psychology 66 (1998), pp. 423–428. Abstract | Abstract + References in Scopus | Cited By in Scopus

Blatt et al., 2000 S.J. Blatt, D.C. Zuroff, C.M. Bondi and C.A. Sanislow, Short and long-term effects of medication and psychotherapy in the brief treatment of depression: Further analyses of data from the NIMH TDCRP, Psychotherapy Research 10 (2000), pp. 215–234. Abstract-PsycINFO   | Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus

Blatt et al., 1996 S.J. Blatt, D.C. Zuroff, D.M. Quinlan and P.A. Pilkonis, Interpersonal factors in brief treatment of depression: Further analyses of the National Institute of Mental Health Treatment of Depression Collaborative Research Program, Journal of Consulting and Clinical Psychology 64 (1996), pp. 162–171. Abstract | Abstract + References in Scopus | Cited By in Scopus

Bowlby, 1988a J. Bowlby, Developmental psychology comes of age, American Journal of Psychiatry 145 (1988), pp. 1–10. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Bowlby, 1988b J. Bowlby, A secure base: Clinical applications of attachment theory, Routledge & Kegan Paul, London (1988).

Brunswick, 1956 E. Brunswick, Perception and the representative design of psychological experiments, University of California Press, Berkeley, CA (1956).

Cane et al., 1986 D.B. Cane, L.J. Olinger, I.H. Gotlib and N.A. Kuiper, Factor structure of the dysfunctional attitude scale in a student population, Journal of Clinical Psychology 42 (1986), pp. 307–309. Abstract-EMBASE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Consumer Reports, 2004 Consumer Reports, Drug versus talk therapy, Consumer Reports (2004, October), pp. 22–29.

Crits-Christoph, 2004 Crits-Christoph, P. (2004). Personal communication.

Cronbach, 1953 L.J. Cronbach, Correlation between persons as a research tool. In: O.H. Mowrer, Editor, Psychotherapy: Theory and research, Ronald, New York (1953), pp. 376–389.

Dobson & Breiter, 1983 K.S. Dobson and H.J. Breiter, Cognitive assessment of depression: Reliability and validity of three measures, Journal of Abnormal Psychology 92 (1983), pp. 107–109. Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Dobson & Shaw, 1986 K.S. Dobson and B.F. Shaw, Cognitive assessment with major depressive disorders, Cognitive Therapy and Research 10 (1986), pp. 13–29. Abstract-PsycINFO   | Full Text via CrossRef

Elkin, 1994 I. Elkin, The NIMH treatment of depression collaborative research program: Where we began and where we are now. In: A.E. Bergin and S.L. Garfield, Editors, Handbook of psychotherapy and behavior change (4th ed.), Wiley, New York (1994), pp. 114–135.

Elkin et al., 1996 I. Elkin, R.D. Gibbons, M.T. Shea and B.F. Shaw, Science is not a trial (but it can be a tribulation), Journal of Consulting and Clinical Psychology 64 (1996), pp. 92–103. Abstract | Abstract + References in Scopus | Cited By in Scopus

Elkin et al., 1995 I. Elkin, R.D. Gibbons, M.T. Shea, S.M. Sotsky, J.T. Watkins, P.A. Pilkonis and D. Hedeker, Initial severity and differential treatment outcome in The National Institute of Mental Health Treatment of Depression Collaborative Research Program, Journal of Consulting and Clinical Psychology 63 (1995), pp. 841–848.

Elkin et al., 1985 I. Elkin, M.B. Parloff, S.W. Hadley and J.H. Autry, NIMH treatment of depression collaborative research program: Background and research plan, Archives of General Psychiatry 42 (1985), pp. 305–316. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Elkin et al., 1989 I. Elkin, M.T. Shea, J.T. Watkins, S.D. Imber, S.M. Sotsky and J.F. Collins et al., NIMH treatment of depression collaborative research program: General effectiveness of treatments, Archives of General Psychiatry 46 (1989), pp. 971–983. Abstract-MEDLINE  

Flett & Hewitt, 2002 G.L. Flett and P.L. Hewitt, Perfectionism and maladjustment: An overview of theoretical, definitional, and treatment issues. In: G.L. Flett and P.L. Hewitt, Editors, Perfectionism: Theory, research, and treatment, American Psychological Association, Washington, DC (2002), pp. 5–31. Full Text via CrossRef

Fonagy et al., 1996 P. Fonagy, T. Leigh, M. Steele, H. Steele, R. Kennedy and G. Matton et al., The relation of attachment status, psychiatric classification, and response to psychotherapy, Journal of Consulting and Clinical Psychology 64 (1996), pp. 22–31. Abstract | Abstract + References in Scopus | Cited By in Scopus

Frank, 1979 J.D. Frank, The present status of outcome studies, Journal of Consulting and Clinical Psychology 47 (1979), pp. 310–316. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Frank, 1982 J.D. Frank, Therapeutic components shared by all psychotherapies. In: H. Harvey and M.M. Parks, Editors, Psychotherapy Research and Behavior Change vol. 1, American Psychological Association, Washington, DC (1982), pp. 5–37.

Freud, 1957 S. Freud, Civilization and its discontents. In: J. Strachey, Editor, The Standard Edition of the Complete Psychological Works of Sigmund Freud vol. 1, Hogarth, London (1957) Original work published in 1930.

Gabbard, 1995 G.O. Gabbard, General considerations in psychiatric treatment. In: G.O. Gabbard, Editor (2nd ed.), Treatments of Psychiatric Disorders vol. 1, American Psychiatric Press, Washington, DC (1995), pp. 1–87. Abstract-PsycINFO  

Gurman, 1977a A.S. Gurman, The patient's perception of the therapeutic relationship. In: A.S. Gurman and A.M. Razin, Editors, Psychotherapy: A handbook of research, Pergamon, New York (1977), pp. 503–543.

Gurman, 1977b A.S. Gurman, Therapist and patient factors influencing the patient's perception of facilitative therapeutic conditions, Psychiatry 40 (1977), pp. 218–231. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Hartley & Strupp, 1983 D. Hartley and H. Strupp, The therapeutic alliance: Its relationship to outcome in brief psychotherapy. In: J. Masling, Editor, Empirical studies of psychoanalytic theories, Erlbaum, Hillsdale, NJ (1983), pp. 1–27.

Imber et al., 1990 S.D. Imber, P.A. Pilkonis, S.M. Sotsky, I. Elkin, J.T. Watkins and J.F. Collins et al., Mode-specific effects among three treatments for depression, Journal of Consulting and Clinical Psychology 58 (1990), pp. 352–359. Abstract | Abstract + References in Scopus | Cited By in Scopus

Jacobson & Hollon, 1996 N.S. Jacobson and S.D. Hollon, Cognitive-behavior therapy versus pharmacotherapy: Now that the jury's returned their verdict it's time to present the rest of the evidence, Journal of Consulting and Clinical Psychology 64 (1996), pp. 74–80. Abstract | Abstract + References in Scopus | Cited By in Scopus

Jones & Pulos, 1993 E.E. Jones and S.M. Pulos, Comparing the process in psychodynamic and cognitive-behavioral therapies, Journal of Consulting and Clinical Psychology 61 (1993), pp. 306–316. Abstract | Abstract + References in Scopus | Cited By in Scopus

Kazdin, 1986 A.E. Kazdin, Comparative outcome studies of psychotherapy: Methodological issues and strategies, Journal of Consulting and Clinical Psychology 54 (1986), pp. 95–105. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Kazdin, 2002 Kazdin, A.E. (2002, October). Personal communication.

Klein et al., 2003 D.N. Klein, J.E. Schwartz, N.J. Santiago, D. Vivian, C Vocisano and L.G. Castonguay et al., Therapeutic alliance in depression treatment: Controlling for prior change and patient characteristics, Journal of Consulting and Clinical Psychology 71 (2003), pp. 997–1006. Abstract | Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus

Krupnick et al., 1996 J.L. Krupnick, S.M. Sotsky, S. Simmens, J. Moyer, I. Elkin and J. Watkins et al., The role of the therapeutic alliance in psychotherapy and pharmacotherapy outcome: Findings in the NIMH treatment of depression collaborative research program, Journal of Consulting and Clinical Psychology 64 (1996), pp. 532–539. Abstract | Abstract + References in Scopus | Cited By in Scopus

Lambert & Barley, 2002 M.J. Lambert and D.E Barley, Research summary on the therapeutic relationship and psychotherapy outcome. In: J.C. Norcross, Editor, Psychotherapy relationships that work: Therapist contributions and responsiveness to patients, Oxford University Press, London (2002), pp. 17–32.

Lambert et al., 1986 M.J. Lambert, D.A. Shapiro and A.E. Bergin, The effectiveness of psychotherapy. In: S.L. Garfield and A.E. Bergin, Editors, Handbook of psychotherapy and behavior change (3rd ed.), Wiley, New York (1986).

Levy et al., 1995 Levy, K. N., Edell, W. S., Blatt, S. J., Becker, D. F., Quinlan, D. M., Kolligan, J., & McGlashan, T. H. (1995). Two configurations of psychopathology: The relationship of dependency, anaclitic neediness, and self-criticism to personality pathology. Unpublished manuscript.

Loewald, 1962 H.W. Loewald, Internalization, separation, mourning, and the superego, Psychoanalytic Quarterly 31 (1962), pp. 483–504. Abstract-MEDLINE | Abstract-PsycINFO  

Luborksy et al., 1999 L. Luborksy, L.A. Diguer, D.A. Seligman, R. Rosenthal, E.D. Kruse and S. Johnson et al., The researcher's own therapy allegiances: A “wildcard” in comparisons of treatment efficacy, Clinical Psychology: Science and Practice 6 (1999), pp. 95–106.

Luborsky et al., 2002 L. Luborsky, R. Rosenthal, L. Diguer, T.P. Andrusyna, J.S. Berman and J.T. Levitt et al., The dodo bird verdict is alive and well—mostly, Clinical Psychology 9 (2002), pp. 2–12. Abstract-PsycINFO   | Full Text via CrossRef

Luborsky et al., 1975 L. Luborsky, B. Singer and E. Luborsky, Comparation studies of psychotherapies: Is it true that “everyone has won and all must have prizes?”, Archives of General Psychiatry 32 (1975), pp. 995–1008. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Luyten, 2002 Luyten, P. (2002). Normbesef en depressie: Aanzet tot een integratief theoretisch kader en een empirisch onderzoek aan de hand van de depressietheorie van S. J. Blatt. (Personal standards and depression: An integrative psychodynamic framework, and an empirical investigation of S. J. Blatt's theory of depression). Doctoral dissertation, University of Leuven, Leuven, Belgium.

Mongrain, 1998 M. Mongrain, Parental representations and support-seeking behavior related to dependency and self-criticism, Journal of Personality 66 (1998), pp. 151–173. Abstract-MEDLINE  

Morse et al., 2002 J.A. Morse, C.J. Robins and M. Gittes-Fox, Sociotropy, autonomy, and personality disorder criteria in psychiatric patients, Journal of Personality Disorder 16 (2002), pp. 549–560. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus

Norcross, 2002 J.C. Norcross, Psychotherapy relationships that work: Therapist contributions and responsiveness to patients, Oxford University Press, New York (2002).

Oliver & Baumgart, 1985 J.M. Oliver and B.P. Baumgart, The dysfunctional attitude scale: Psychometric properties in an unselected adult population, Cognitive Theory and Research 9 (1985), pp. 161–169.

Ouimette et al., 1994 P.C. Ouimette, D.N. Klein, R. Anderson, L.P. Liso and H. Lizardi, Relationship of sociotropy/autonomy and dependency/self-criticism to DSM-III-R personality disorders, Journal of Abnormal Psychology 103 (1994), pp. 743–749. Abstract | Abstract + References in Scopus | Cited By in Scopus

Paykel et al., 1978 E.S. Paykel, M.M. Weissman and B.A. Prusoff, Social maladjustment and severity of depression, Comprehensive Psychiatry 19 (1978), pp. 121–128. Abstract | Abstract + References in Scopus | Cited By in Scopus

Priel & Shahar, 2000 B. Priel and G. Shahar, Dependency, self-criticism, social context and distress: Comparing moderating and mediating models, Personality and Individual Differences 28 (2000), pp. 515–525. Abstract | Abstract + References in Scopus | Cited By in Scopus

Rogers, 1951 C.R. Rogers, Client-centered therapy, Houghton Mifflin, Boston (1951).

Rogers, 1957 C.R. Rogers, The necessary and sufficient conditions of therapeutic personality change, Journal of Consulting Psychology 21 (1957), pp. 95–103. Abstract-MEDLINE | Abstract-PsycINFO  

Rogers, 1959 C.R. Rogers, A theory of therapy, personality, and interpersonal relationships as developed in the client-centered framework in psychology: A study of science. In: S. Koch, Editor, Formulations of the person and the social context, McGraw-Hill, New York (1959), pp. 184–256.

Seligman, 1995 M.E.P. Seligman, The effectiveness of psychotherapy: The consumer reports study, American Psychologist 50 (1995), pp. 965–974. Abstract | Abstract + References in Scopus | Cited By in Scopus

Shahar et al., 2004 G. Shahar, S.J. Blatt, D.C. Zuroff, J. Krupnick and S.M. Sotsky, Perfectionism impedes social relations and response to brief treatment of depression, Journal of Social and Clinical Psychology 23 (2004), pp. 140–154. Abstract-PsycINFO   | Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus

Shahar & Priel, 2003 G. Shahar and B. Priel, Active vulnerability, adolescent distress, and the mediating/suppressing role of life events, Personality and Individual Differences 35 (2003), pp. 199–218. Abstract | Abstract + References in Scopus | Cited By in Scopus

Shahar et al., 2003 G. Shahar, S.J. Blatt, D.C. Zuroff and P.A. Pilkonis, Role of perfectionism and personality disorder features in response to brief treatment for depression, Journal of Consulting and Clinical Psychology 71 (2003), pp. 629–633. Abstract | Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus

Shapiro & Shapiro, 1982 D.A. Shapiro and D. Shapiro, Meta-analysis of comparative therapy outcome studies: A replication and refinement, Psychological Bulletin 92 (1982), pp. 581–604. Abstract-MEDLINE | Abstract-PsycINFO   | Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus

Shea et al., 1992 T. Shea, I. Elkin, S.D. Imber, S.M. Sotsky, J.T. Watkins and J.F. Collins et al., Course of depressive symptoms over follow-up: Findings from the National Institute of Mental Health treatment of depression collaborative research program, Archives of General Psychiatry 49 (1992), pp. 782–787.

Smith et al., 1980 M.L. Smith, G.V. Glass and T.I. Miller, The benefits of psychotherapy, Johns Hopkins University Press, Baltimore, MD (1980).

Sotsky et al., 1991 S.M. Sotsky, D.R. Glass, M.T. Shea, P.A. Pilkonis, J.F. Collins and I. Elkin et al., Patient predictor of response to psychotherapy and pharmacotherapy: Findings in the NIMH treatment of depression collaborative research program, American Journal of Psychiatry 148 (1991), pp. 997–1008. Abstract-EMBASE | Abstract-MEDLINE | Abstract-PsycINFO   | Abstract + References in Scopus | Cited By in Scopus

Stiles et al., 1986 W.B. Stiles, D.A. Shapiro and R. Elliott, Are all psychotherapies equivalent?, American Psychologist 41 (1986), pp. 161–180.

Strupp & Binder, 1984 H.H. Strupp and J.L. Binder, Psychotherapy in a new key: A guide to time limited dynamic psychotherapy, Basic, New York (1984).

VandenBos & Pino, 1980 G.R. VandenBos and C.D. Pino, Research on the outcome of psychotherapy. In: G.R. Vandenbos, Editor, Psychotherapy: Practice, research, policy, Sage, Beverly Hills, CA (1980), pp. 23–69.

Wampold, 2001 B.E. Wampold, The great psychotherapy debate: Models, methods, and findings, Lawrence Erlbaum Associates, Mahwah, NJ (2001).

Watkins et al., 1993 J.T. Watkins, W.R. Leber, S.D. Imber, J.R. Collins, I. Elkin and P.A. Pilkonis et al., NIMH treatment of depression collaborative research program: Temporal course of change of depression, Journal of Consulting and Clinical Psychology 61 (1993), pp. 858–864. Abstract | Abstract + References in Scopus | Cited By in Scopus

Weissman & Beck, 1978 A.N. Weissman and A.T. Beck, Development and validation of the dysfunctional attitudes scale: A preliminary investigation Paper presented at the 86th annual convention of the American Psychological Association Toronto (1978, August–September).

Weissman & Paykel, 1974 M.M. Weissman and E.S. Paykel, The depressed woman: Study of social relationships, University of Chicago Press, Chicago (1974).

Westen et al., 2004 D. Westen, C.M. Novotny and H. Thompson-Brenner, The empirical status of empirically supported psychotherapies: Assumptions, findings, and reporting in controlled clinical trials, Psychological Bulletin 130 (2004), pp. 631–663. Abstract | Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus

Wiggins, 1991 J.S. Wiggins, Agency and communion as conceptual coordinates for the understanding and measurement of interpersonal behavior. In: W.W. Grove and D. Cicchetti, Editors, Thinking clearly about psychology, Personality and Psychotherapy vol. 2, University of Minnesota Press, Minneapolis (1991), pp. 89–113. Abstract-PsycINFO  

Wortman, 1983 P.M. Wortman, Meta-analysis: A validity perspective, Annual Review of Psychology 34 (1983), pp. 223–260. Abstract-PsycINFO  

Zuroff & Blatt, 2002 D.C. Zuroff and S.J. Blatt, Vicissitudes of life after the short-term treatment of depression: Roles of stress, social support, and personality, Journal of Social and Clinical Psychology 21 (2002), pp. 473–496. Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus

Zuroff & Blatt, unpublished data Zuroff, D. C., & Blatt, S. J. (2004). Brief treatment of depression: Contributions of the therapeutic relationship to clinical improvement, reduced vulnerability, and enhanced resilience. (unpublished data).

Zuroff et al., 2003 D.C. Zuroff, S.J. Blatt, J.L. Krupnick and S.M. Sotsky, When brief treatment is over: Enhanced adaptive capacities and stress reactivity after termination, Psychotherapy Research 13 (2003), pp. 99–115. Abstract-PsycINFO   | Full Text via CrossRef

Zuroff et al., 2000 D.C. Zuroff, S.J. Blatt, S.M. Sotsky, J.L. Krupnick, D.J. Martin and C.A. Sanislow et al., Relation of therapeutic alliance and perfectionism to outcome in brief outpatient treatment of depression, Journal of Consulting and Clinical Psychology 68 (2000), pp. 114–124. Abstract | Abstract + References in Scopus | Cited By in Scopus

Zuroff & Duncan, 1999 D.C. Zuroff and N. Duncan, Self-criticism and conflict resolution in romantic couples, Canadian Journal of Behavioral Science 31 (1999), pp. 137–149. Abstract | Abstract + References in Scopus | Cited By in Scopus



Corresponding Author Contact InformationCorresponding author. Tel.: +1 203 785 2090; fax: +1 203 785 7357.


1 The NIMH-TDCRP was a multisite program initiated and sponsored by the Psychosocial Treatments Research Branch, Division of Extramural Research Programs (now part of the Mood, Anxiety, and Personality Disorders Research Branch, Division Clinical Research), NIMH. The program was funded by cooperative agreements to six participating sites: George Washington University, NH 33762; University of Pittsburgh, MH 33752; University of Oklahoma, MH 33760; Yale University, MH 33827; Clarke Institute of Psychiatry, MH 38231; and Rush Presbyterian-St. Luke's Medical Center, MH 35017. The principal NIMH collaborators were Irene Elkin, Coordinator; Tracie Shea, Associate Coordinator (formerly or George Washington University); John P. Docherty (now at New York Hospital-Cornell Medical Center); and Morris B. Parloff (now at American University). The principal investigators and project coordinators at three participating research sites were Stuart M. Sotsky and David Glass, George Washington University; Stanley D. Imber and Paul A. Pilkonis, University of Pittsburgh; and John T. Watkins and William Leber, University of Oklahoma. The principal investigators and project coordinators at the three research sites responsible for training therapists were Myrna Weissman (now at Columbia University), Eve Chevron, and Bruce J. Rounsaville, Yale University; Brian F. Shaw and T. Michael Vallis, Clarke Institute of Psychiatry; and Jan A. Fawcett and Phillip Epstein, Rush Presbyterian-St. Luke's Medical Center. Collaborators in the data management and data analysis aspects of the program were C. James Klett, Joseph F. Collins, and Roderic Gillis of the Veterans Administration Studies Program, Perry Point, Maryland.
2 We are indebted to Colin A. Bondi, Paul A. Pilkonis, Donald M. Quinlan, Charles M. Sanislow III, and Golan Shahar for their contributions to our further analyses of data from the NIMH sponsored TDCRP.
3 Though other antidepressant medications have been developed since the initiation of the TDCRP, they differ primarily, not in their effectiveness in reducing symptoms of depression, but in their likelihood of inducing undesirable side effects (Barbui & Hotopf, 2001).
4 The clinical management component of the IMI-CM and PLA-CM conditions was designed to manage the medication and to “provide a generally supportive atmosphere and to enable the psychiatrist to assess the patient's clinical status.” “The manual and training…include guidelines for providing support and encouragement to the patient and giving direct advice when necessary. This CM component thus approximates a ‘minimal supportive therapy’ condition” (Elkin, Parloff, Hadley, & Autry, 1985, p.311).
5 A total of 250 were randomly assigned across the three treatment sites and 239 patients started treatment (i.e., had at least one treatment session).
6 Therapists were selected to participate in a particular treatment modality because they had prior experience in and commitment to that treatment modality. Thus, distortions due to allegiance effects (Luborksy et al., 1999) were probably relatively minimal in the TDCRP.
7 Though some criticism has been leveled at the TDCRP for possible site difference in the effectiveness of the administration of CBT at one site (Elkin et al., 1996 and Jacobson & Hollon, 1996), our extensive analyses of the TDCRP yielded surprisingly few significant site differences. We believe, therefore, that the marginally significant difference of one site having somewhat less effective results with CBT than the other two sites is a statistical artifact of extensive multiple comparisons of 4 different treatments across three treatment sites.
8 Patients in CBT were seen, per Beck's recommendation, twice weekly for the first four weeks.
9 Because of the extensive space required to store these videotapes, they were recently (2004) destroyed, though audio recordings of all treatment sessions were retained (Crits-Christoph, 2004).
10 Ouimette et al. (1994) found that BPD patients had elevated concerns on issues of both relatedness and self-definition. Blatt and Auerbach (1988), in an earlier clinical theoretical contribution, differentiated between highly dependent borderline patients who conform to the BPD diagnosis as described in DSM, and a more over-ideational, introjective type of borderline patient with obsessive–complusive and paranoid features. Blatt and Auerbach suggest that the more dependent borderline patient, who is vulnerable to profound feelings of abandonment, would have greater concerns about issues of relatedness, while the more over-ideational obsessive–paranoid borderline patient would have greater concerns about issues of self-definition.
11 Elkin et al., 1989 and Elkin et al., 1995 reported that IMI-CM and IPT appeared to be somewhat more effective than CBT, but only with more seriously depressed patients.
12 Imber and colleagues (Imber et al., 1990) addressed the possibility of specific effects of each of the treatments on different outcome measures that had been included in the TDCRP because of their “presumed sensitivity to different treatments.” Although the TDCRP investigators used carefully derived outcome measures based on the rationales and procedures of three types of brief treatment for depression evaluated in the TDCRP (CBT, IPT, and IMI-CM), the results of Imber et al. (1990) provided little support for their basic hypothesis: “None of the therapies produced consistent effects on measures related to its theoretical origins” (p. 352).
13 These findings are consistent with the findings of Shea et al. (1992) that suggest a possibly greater relapse rate for patients in the medication as compared to the psychotherapy conditions.
14 These findings are consistent with the results of a recent survey, conducted by Consumer Reports (2004) on over three thousand respondents, that indicate that though “medication relieved symptoms faster than” psychotherapy, “psychotherapy of at least 13 sessions had better outcomes than those whose therapy was mostly medication” (p. 22).
15 In each of the analyses reported in this section and in Section 3 below, when appropriate, product terms representing the interaction of predictor variables with treatment conditions were entered as predictors to assess if the findings were specific to one or more of the treatment conditions. No significant treatment effects were found, so the findings reported in Sections 2 and 3 occurred equally in each of the treatment conditions.
16 B–L-RI scores at the end of the second treatment session, correlated − .09 with the pretreatment level of Perfectionism as measured on the DAS. When controlling for level of symptom reduction at the end of the 4th treatment session, the group with lower B–L RI had significantly (p < .05) higher pretreatment PFT scores. Despite this initial relationship between pretreatment PFT and the B–L RI, patients with higher BL-RI scores at the end of the second treatment session had a significantly greater reduction (p < .01) in PFT over the course of treatment than patients with lower scores on the BL-RI. This discussion of the quality of the therapeutic relationship as perceived and reported by the patient has to be distinguished from our later discussion (p 17 ff) of the patient's participation in the therapeutic alliance as rated by Krupnick et al. (1996) based on ratings of patient's behavior in the treatment sessions as observed in video recordings of the treatment sessions.
17 These findings are consistent with the results of a recent survey conducted by Consumer Reports (2004) that therapy with “psychologists and clinical social workers was perceived as effective as that given by a psychiatrist” (p. 22).
18 In the analyses reported in this section, when appropriate, product terms representing the interaction of predictor variables with treatment conditions were again added as predictors to assess if the findings were specific to one or more of the treatment conditions. Generally, we found no significant treatment effects, so the findings reported below occurred equally in each of the treatment conditions.
19 Shahar, Blatt, Zuroff, and Pilkonis (2003) found that while patients' pretreatment levels of perfectionism and their personality disorder features both predicted therapeutic outcome in the TDCRP, only the level of perfectionism predicted patients' contributions to the therapeutic alliance and the level of their satisfaction with their social relationships.
20 Sotsky et al. (1991) identified several patient characteristics that appeared to be predictive of outcome in particular treatment conditions in the TDCRP. They found that better social functioning assessed at intake on the social and leisure activities subscale of the Social Adjustment Scale (SAS; M. Weissman & Paykel, 1974) predicted generally favorable outcome, especially in IPT. Lower total score on the DAS (A. Weissman & Beck, 1978) predicted favorable outcome especially in IMI-CM and CBT. Patients with both impaired social and work functioning responded best to IMI-CM. Sotsky et al. (1991) concluded that “each psychotherapy relies on specific and different learning techniques to alleviate depression, and thus each may depend on an adequate capacity in the corresponding sphere of patient function to produce recovery with the use of that approach” (1991, p. 1006). The differences in therapeutic outcome between the two forms of psychotherapy in the TDCRP (CBT and IPT), however, are minimal and, even further, Jones and Pulos (1993) in their prototype analysis of these two psychotherapies in the TDCRP found considerable overlap between these two forms of manualized psychotherapy.
21 We are grateful to Janice Krupnick and Stuart Sotsky for their collaboration in this exploration of the impact of PFT on the therapeutic process. An example of the 20 items assessing the patient's contribution to the alliance was the item “Patient talks freely, openly and honestly about himself.” An example of the 11 items assessing the therapist contributions to the therapeutic alliance was the item, “Therapist commits himself and his skills to helping patient.”
22 Because of a concern that the inclusion of the assessment of social adjustment (SAS) as part of the composite outcome measure might overlap with the measurement of patients' social network, results with the Social Adjustment Scale (SAS) were excluded from the composite maladjustment measure for these analyses (Shahar et al., 2004).
23 These formulations are consistent with a recent report from a survey on differences between medication and psychotherapy conducted by Consumer Reports (2004) that indicate that “consumers who did their own research and monitored their own care reported better results” (p.22).
24 We are indebted to Rachel Sills-Shahar for her observations on these issues.


This Document
SummaryPlus
Full Text + Links
   ·Full Size Images
PDF (359 K)
External Links
Abstract + References in Scopus
Actions
Cited By
Save as Citation Alert
E-mail Article
Export Citation
Clinical Psychology Review
Volume 25, Issue 4 , June 2005, Pages 459-486


 
HomeSearchBrowse JournalsBrowse Book Series, Handbooks and Reference WorksBrowse Abstract DatabasesMy ProfileAlertsHelp (Opens New Window)

Contact Us  |  Terms & Conditions  |  Privacy Policy

Copyright © 2005 Elsevier B.V. All rights reserved. ScienceDirect® is a registered trademark of Elsevier B.V.