Comparison of Two Different Approaches to Therapy - Report Example

Different Approaches To Therapy [Name Of Student] [Name Of Institution] INTRODUCTION The therapeutic alliance, defined as the attachment and collaboration between the client and therapist (Bordin, 1979), has been identified as a factor crucial to the change process across different forms of psychotherapy, including psychodynamic–interpersonal (e.g., Strupp & Binder, 2002) and cognitive–behavioral (e.g., Arnkoff, 2003) psychotherapies. Despite different theoretical assumptions about the nature and function of the therapeutic alliance (Gaston et al., 2003), a good alliance predicts client improvement in both forms of therapy (e.g., Bachelor, 2000; Castonguay, Goldfried, Wiser, Raue, & Hayes, 2004; Gaston, 2000; Horvath & Symonds, 2000; Raue & Goldfried, 2002; Windholz & Silberschatz, 1998). The concept of the alliance was originally developed by psychoanalytic theorists (e.g., Greenson, 1965; Sterba, 1934), who discussed the patient's ability to form a “working alliance” to work with the therapist's interpretations. Although ignored in early writings of behavioral techniques, cognitive–behavioral therapists have more recently recognized the value of the “therapeutic relationship” as a means to facilitate the process of therapy (e.g., Goldfried & Davison, 1976, 2002; Raue & Goldfried, 2002). Reflecting the early greater interest in the alliance by psychoanalytic and psychodynamic theorists, most attempts to measure the alliance have emerged from this background. The Working Alliance Inventory (WAI; Horvath, 2001), however, was specifically developed to apply to all forms of therapy (Horvath, 2001). This scale reflects Bordin's (1979) transtheoretical and dyadic conceptualization of the working alliance as consisting of agreement between the client and therapist on the tasks of therapy, agreement on the goals of therapy, and the development of a therapeutic bond. LITERATURE REVIEW Empirically, studies examining aspects of the alliance in these two therapies have provided mixed evidence. Sloane, Staples, Cristol, Yorkston, and Whipple (1975) found outside observers to rate behaviorally oriented therapists as displaying significantly higher levels of empathy, genuineness, and interpersonal contact than psychoanalytically oriented therapists, but they rated both sets of therapists equally on the display of warmth. Brunink and Schroeder (1979) found that observers rated behavior therapists as using more supportive communications such as reassurance, praise, and sympathy than psychoanalytic therapists did, but they rated both sets of therapists equally on empathy and rapport. Marmar, Gaston, Gallagher, and Thompson (1999) failed to demonstrate significant differences among behavioral, cognitive, and brief psychodynamic therapies in the level of the alliance as measured by the California Psychotherapy Alliance Scales (CALPAS; Marmar, Horowitz, Weiss, & Marziali, 2004) from the client, therapist, or observer perspective. By contrast, Raue, Castonguay, and Goldfried (2001) compared clinically significant psychodynamic–interpersonal and cognitive–behavioral therapy sessions and found that observers rated the cognitive–behavioral group significantly higher on the WAI. In addition, greater variability in alliance scores occurred in the psychodynamic–interpersonal group, indicating a greater spread of higher and lower scores. This study differs from the former three in that all sessions chosen for examination were identified by the therapists as being significant in terms of facilitating therapeutic change. METHODOLOGY PARTICIPANTS Five therapists, espousing an even-handed approach to both cognitive–behavioral and psychodynamic–interpersonal therapies, participated in the study. Three therapists were recent recipients of a master's-level degree in the United Kingdom (representing a level of experience between master's level and doctoral level in the United States), one had 6 years of additional clinical experience, and one had 16 years of additional clinical experience. Three therapists were male, and two were female. Before taking on cases for the study, all therapists satisfactorily completed a minimum of four training cases in each therapy (two in the 8-session condition and two in the 16-session condition). Fifty-seven clients who were diagnosed with major depressive disorder and whose problems in some way affected their occupational functioning participated in the study. Clients were recruited by health care providers, who received information in the mail about the research clinic offering psychological therapy to depressed and anxious professional, managerial, and white-collar workers. TREATMENTS Manuals were developed that define the objectives, strategies, and techniques that are appropriate to both treatments (see Coren & Firth, 2005, for more detailed descriptions). The cognitive–behavioral therapy manual emphasizes teaching anxiety management through relaxation exercises and exposure; self-management through self-monitoring, homework assignments, and self-reinforcement; cognitive restructuring; a focus on problems related to work through assertiveness training, time management, and appropriate delegation of responsibility. The psychodynamic–interpersonal therapy manual emphasizes the exploration and modification of interpersonal problems within the therapeutic relationship; a focus on in-session experience of emotion; making links to similar experiences in other times and settings; and interpreting underlying reasons for experiences. Adherence to these manuals was rated by means of the Sheffield Psychotherapy Rating Schedule (Coren & Startup, 1999) on a selected number of sessions. INTAKE MEASURES Clinical interviews were administered to assess symptomatology at intake using the Present State Examination (PSE; Wing, Cooper, & Sartorius, 1974). To make the diagnoses of major depressive disorder, panic disorder, and generalized anxiety disorder, we used relevant questions from the Diagnostic Interview Schedule (DIS; Robins, Helzer, Croughan, Williams, & Spitzer, 2000). The BDI was also administered at intake. Additional measures were administered (see Coren et al., 2002 for details), but these were not used in the present study. SESSION SELECTION Two sessions were selected from each therapist–client pair, one having the highest impact on the client and the other having the lowest impact. The initial and ending three therapy sessions were excluded from selection to avoid initial assessment, agenda setting, and termination issues. Impact was defined primarily in terms of therapist ratings of session helpfulness taken at the end of each session. Therapists made their ratings on a 9-point scale on which 9 = greatly helpful and 1 = greatly hindering. Using these ratings, we selected two sessions having the largest spread between high and low, of two points or more. Therapist ratings of helpfulness were chosen over client ratings on the basis of their relatively greater ability to discriminate between sessions. However, if client ratings of high and low helpfulness directly reversed those of the therapist, an alternate high or low therapist-rated session that maintained a maximum gap between therapist high and low ratings was selected. If there were ties between session ratings for high or low session, or when therapist ratings failed to differentiate between sessions by at least two points, other ratings of impact were used in combination with therapist ratings of helpfulness. Specifically, the raw scores of the valuable-worthless, special-ordinary, and full-empty scales of the therapist-rated SEQ (Stiles, Trowe, 2000; Stiles, Trowe, Coren, & Firth-Cozens, 1998) were added, and the two sessions with the highest and lowest scores were selected. PROCESS MEASURES After every session, both the client and the therapist independently evaluated the session using the SEQ. The SEQ is a measure of the immediate impact of a particular session. The first part consists of session evaluations on twelve 7-point scales of bipolar adjectives such as valuable-worthless, shallow-deep, and rough-smooth. Factor analyses have revealed two major factors: depth-value and smoothness-comfort (Stiles, Trowe & Snow, 2002). The second part of the SEQ consists of ratings of affective state on twelve 7-point scales such as happy-sad, angry-pleased, and confident-afraid. Factor analyses on this section have revealed two major factors: positivity and arousal. The internal consistency of the SEQ has been demonstrated to be high for both client and therapist versions, and consensus between client and therapist ratings of the same session has been shown to be low to moderate, suggesting that there are differences in how each perspective evaluates the session (Stiles, Trowe, Coren, & Firth-Cozens, 1998). CODING Six advanced doctoral-level students, coming from a clinical psychology program emphasizing cognitive–behavioral approaches to treatment, served as raters. In light of findings from Raue, Putterman, Goldfried, and Wolitzky (2003) that rater orientation may, at times, influence absolute ratings of alliance, an effort was made to select raters who were more acquainted with psychodynamic therapy. Four of the raters participated in a seminar on psychodynamic theory and therapy, which involved academic work and supervised therapy with 2 clients a week for 6 months. The fifth rater had experience doing dynamic therapy at an inpatient hospital, and the sixth had limited direct experience but had done extensive reading on psychodynamic theory and therapy. Before coding, the raters were trained for approximately 1 month to adequate reliability among themselves (intraclass correlation [ICC] > .60) on a mix of cognitive–behavioral and psychodynamic–interpersonal therapy sessions. The order of sessions coded was randomized to prevent possible biasing effects related to order of receipt. Rotating pairs of raters independently coded the 114 sessions by listening to the audiotape and reading the transcript. Each rater coded 38 sessions, completing approximately 2 per week. No rater coded the same client twice. To maintain calibration, raters met regularly to discuss any potential discrepancies. RESULTS RATER RELIABILITY Consistent with past research on the WAI (Horvath & Greenberg, 1999), all subscales were highly intercorrelated (r = .83 for Bond and Task subscales, .77 for Bond and Goal subscales, and .92 for Task and Goal subscales; all p values Read More

By contrast, Raue, Castonguay, and Goldfried (2001) compared clinically significant psychodynamic–interpersonal and cognitive–behavioral therapy sessions and found that observers rated the cognitive–behavioral group significantly higher on the WAI. In addition, greater variability in alliance scores occurred in the psychodynamic–interpersonal group, indicating a greater spread of higher and lower scores. This study differs from the former three in that all sessions chosen for examination were identified by the therapists as being significant in terms of facilitating therapeutic change.

METHODOLOGY PARTICIPANTS Five therapists, espousing an even-handed approach to both cognitive–behavioral and psychodynamic–interpersonal therapies, participated in the study. Three therapists were recent recipients of a master's-level degree in the United Kingdom (representing a level of experience between master's level and doctoral level in the United States), one had 6 years of additional clinical experience, and one had 16 years of additional clinical experience. Three therapists were male, and two were female.

Before taking on cases for the study, all therapists satisfactorily completed a minimum of four training cases in each therapy (two in the 8-session condition and two in the 16-session condition). Fifty-seven clients who were diagnosed with major depressive disorder and whose problems in some way affected their occupational functioning participated in the study. Clients were recruited by health care providers, who received information in the mail about the research clinic offering psychological therapy to depressed and anxious professional, managerial, and white-collar workers.

TREATMENTS Manuals were developed that define the objectives, strategies, and techniques that are appropriate to both treatments (see Coren & Firth, 2005, for more detailed descriptions). The cognitive–behavioral therapy manual emphasizes teaching anxiety management through relaxation exercises and exposure; self-management through self-monitoring, homework assignments, and self-reinforcement; cognitive restructuring; a focus on problems related to work through assertiveness training, time management, and appropriate delegation of responsibility.

The psychodynamic–interpersonal therapy manual emphasizes the exploration and modification of interpersonal problems within the therapeutic relationship; a focus on in-session experience of emotion; making links to similar experiences in other times and settings; and interpreting underlying reasons for experiences. Adherence to these manuals was rated by means of the Sheffield Psychotherapy Rating Schedule (Coren & Startup, 1999) on a selected number of sessions. INTAKE MEASURES Clinical interviews were administered to assess symptomatology at intake using the Present State Examination (PSE; Wing, Cooper, & Sartorius, 1974).

To make the diagnoses of major depressive disorder, panic disorder, and generalized anxiety disorder, we used relevant questions from the Diagnostic Interview Schedule (DIS; Robins, Helzer, Croughan, Williams, & Spitzer, 2000). The BDI was also administered at intake. Additional measures were administered (see Coren et al., 2002 for details), but these were not used in the present study. SESSION SELECTION Two sessions were selected from each therapist–client pair, one having the highest impact on the client and the other having the lowest impact.

The initial and ending three therapy sessions were excluded from selection to avoid initial assessment, agenda setting, and termination issues. Impact was defined primarily in terms of therapist ratings of session helpfulness taken at the end of each session. Therapists made their ratings on a 9-point scale on which 9 = greatly helpful and 1 = greatly hindering. Using these ratings, we selected two sessions having the largest spread between high and low, of two points or more. Therapist ratings of helpfulness were chosen over client ratings on the basis of their relatively greater ability to discriminate between sessions.

Read More