[acb-hsp] Interpersonal Predictors of therapeutic alliance ...

[J.Rayl]

Interpersonal predictors of early therapeutic alliance in a transdiagnostic cognitive-behavioral

treatment for adolescents with anxiety and depression.

Author:

Levin, Laura1; Henderson, Heather A.1; Ehrenreich-May, Jill11Department of Psychology,

University of Miami, Coral Gables, FL, US j.ehrenreich at miami.edu

Publication info:

Psychotherapy 49. 2 (Jun 2012): 218-230.

ProQuest document link

Abstract:

The importance of therapeutic alliance in predicting treatment success is well established,

but less is known about client characteristics that predict alliance. This study

examined alliance predictors in adolescents with anxiety and/or depressive disorders

(n = 31) who received a transdiagnostic cognitive-behavioral treatment, the Unified

Protocol for the Treatment of Emotional Disorders in Youth (Ehrenreich, Buzzella,

Trosper, Bennett, & Barlow, 2008) in the context of a larger randomized controlled

trial. Alliance was assessed at session three by therapists, clients, and independent

observers. Results indicated that alliance ratings across the three informant perspectives

were significantly associated with one another, but that pretreatment interpersonal

variables (e.g., social support, attachment security, and social functioning in current

family and peer relationships) were differentially associated with varying informant

perspectives. Adolescent and observer ratings of alliance were both predicted by

adolescent self-reports on measures reflecting how they perceive their interpersonal

relationships. In addition, adolescent-reported symptom severity at pretreatment

predicted observer ratings of alliance such that adolescents who indicated greater

anxiety and depressive symptoms were rated as having stronger early alliances by

independent observers. Therapists perceived having weaker early alliances with adolescents

evidencing clinically significant depression at intake as compared with adolescents

diagnosed with anxiety disorders alone. Future research is needed to examine whether

identification of relevant interpersonal factors at intake can help improve initial

therapeutic engagement and resulting outcomes for the psychosocial treatment of adolescents

with anxiety and depressive disorders. (PsycINFO Database Record (c) 2012 APA, all

rights reserved)(journal abstract)

Links:

Look for Full-Text

Full Text:

Contents

Abstract

Method

Participants

Treatment Context and Therapists

Measures

Clinical status

Anxiety Disorders Interview Schedule, Child Version, Parent and Child Reports

Symptom severity

Revised Child Anxiety and Depression Scale

Adolescent-rated pretreatment relational characteristics

Adolescent Attachment Questionnaire

Multidimensional Scale of Perceived Social Support

Child and Adolescent Social and Adaptive Functioning Scale

Alliance assessment

Working Alliance Inventory

Vanderbilt Therapeutic Alliance Scale

Procedure

Alliance Raters and Training

Alliance Observational Coding Procedure

VTAS: Interrater reliability and scale properties

Results

Preliminary Analyses

Descriptive statistics for independent (interpersonal) variables

Descriptive data on alliance

Cross-informant agreement on therapeutic alliance

Predictors of alliance

Demographic and clinical variables

Interpersonal variables

Discussion

Show less

Figures and Tables

Table 1

Table 2

Table 3

Table 4

Table 5

Show less

Abstract

The importance of therapeutic alliance in predicting treatment success is well established,

but less is known about client characteristics that predict alliance. This study

examined alliance predictors in adolescents with anxiety and/or depressive disorders

(

n

= 31) who received a transdiagnostic cognitive-behavioral treatment, the Unified

Protocol for the Treatment of Emotional Disorders in Youth (

Ehrenreich, Buzzella, Trosper, Bennett, & Barlow, 2008

) in the context of a larger randomized controlled trial. Alliance was assessed at

session three by therapists, clients, and independent observers. Results indicated

that alliance ratings across the three informant perspectives were significantly

associated with one another, but that pretreatment interpersonal variables (e.g.,

social support, attachment security, and social functioning in current family and

peer relationships) were differentially associated with varying informant perspectives.

Adolescent and observer ratings of alliance were both predicted by adolescent self-reports

on measures reflecting how they perceive their interpersonal relationships. In addition,

adolescent-reported symptom severity at pretreatment predicted observer ratings of

alliance such that adolescents who indicated greater anxiety and depressive symptoms

were rated as having stronger early alliances by independent observers. Therapists

perceived having weaker early alliances with adolescents evidencing clinically significant

depression at intake as compared with adolescents diagnosed with anxiety disorders

alone. Future research is needed to examine whether identification of relevant interpersonal

factors at intake can help improve initial therapeutic engagement and resulting outcomes

for the psychosocial treatment of adolescents with anxiety and depressive disorders.

The patient-therapist bond and degree of collaboration observed in the therapeutic

relationship, often referred to as therapeutic alliance (

Bordin, 1994

), predicts treatment progress and outcomes across various client populations and

treatment modalities (

Horvath, 2001

;

Horvath, Del Re, Fluckiger & Symonds, 2011

;

Martin, Gaske, & Davis, 2000

;

Shirk & Karver, 2003

;

Shirk, Karver, & Brown, 2011

). In the current study, we focus on the development of early alliance among adolescents

with anxiety and/or depressive disorders, many of whom had additional comorbid emotional

and behavioral difficulties, within the context of a larger randomized control trial

examining the efficacy of a novel, transdiagnostic cognitive-behavioral therapy (CBT)

for adolescents (

Ehrenreich et al., 2008

). Recent meta-analytic findings suggest that alliance may be particularly relevant

to the outcome of behavioral treatments (

Shirk, Karver & Brown, 2011

), increasing the importance of examining alliance development in a new cognitive-behavioral

approach such as this. In youth with internalizing disorders, a strong therapeutic

alliance has found to predict engagement in therapeutic tasks (

Karver et al., 2008

) and successful treatment outcomes (

Chu et al., 2004

). Given the importance of alliance-to-treatment process and outcome in this population,

knowledge of client characteristics that contribute to the development of a stronger

or weaker alliance is essential to increase treatment effectiveness.

The ways in which adolescents perceive their interpersonal relationships at pretreatment

have been thought to play an especially important role in treatment engagement and

outcome (e.g.,

Green, 2006

). However, few studies have examined linkages between client pretreatment interpersonal

characteristics and early alliance formation with adolescents, and none have examined

this specifically among anxious and depressed adolescent populations. Identifying

alliance predictors in adolescents seems especially critical, given that adolescents

are at heightened risk to drop-out of treatment prematurely (

Kazdin, 1996

;

Wierzbicki & Pekarik, 1993

)-and that adolescents who do not have strong alliances are even more likely to leave

treatment prematurely (

Garcia & Weisz, 2002

;

Robbins et al., 2006

). As such, the lack of research examining additional correlates of alliance development

represents a serious gap in the current adolescent literature (

Green, 2006

;

Shirk & Saiz, 1992

;

Zack, Castonguay, & Boswell, 2007

). To address this gap, the present study focuses on adolescent alliance and its

pretreatment correlates. Clinically, such research may be especially useful in helping

therapists anticipate sources of resistances that may interfere with alliance formation,

and may also contribute to the development of strategies for therapists to use when

working with challenging adolescent clients to mitigate the influence of preexisting

predictor variables (e.g.,

Diamond, Liddle, Hogue, & Dakof, 1999

;

DiGiuseppe, Linscott, & Jilton, 1996

;

Robbins, Turner, Alexander, & Perez, 2003

).

Although only a few studies with adolescents have directly investigated correlates

of alliance, interpersonal and social relationship variables have emerged as significant

predictors of this patient-therapist bond, whereas clinical variables have not generally

related to alliance formation (

Eltz, Shirk, & Sarlin, 1995

;

Garner, Godley, & Funk, 2008

).

Eltz et al. (1995)

examined interpersonal correlates of adolescent and therapist-rated alliance in

a sample of psychiatrically hospitalized adolescents with a history of maltreatment.

In this study, adolescents with higher levels of relationship problems (using the

Interpersonal Problems scale of the Child Behavior Checklist;

Achenbach, 1991

) had poorer alliance, according to both therapist and adolescent perspectives. In

addition, adolescents' ratings of alliance were predicted by their pretreatment ratings

of interpersonal expectancies about close relationships and willingness to seek out

and use social support resources. Symptom severity, on the other hand, did not relate

significantly to alliance in this study (Eltz et al., 2005). In adolescent substance

abuse populations, measures assessing levels of social support and relationships

with family and friends have been identified as predictors of therapeutic alliance

(

Garner et al., 2008

) and treatment engagement (

Broome, Joe, & Simpson, 2001

), whereas clinical variables (e.g., substance use severity) have been unrelated

to alliance formation. Given the findings from other treatment populations indicating

the importance of relational characteristics to alliance and outcome, a central question

in the current study is whether relational characteristics are associated with alliance

early in therapy among adolescents with anxiety and depressive disorders.

>From a methodological point of view, a major challenge in studying alliance is in

dealing with similarities and differences among rater perspectives on alliance (

Horvath, 1994

). Past research comparing agreement between alliance rater perspectives indicates

low-to-moderate or inconsistent agreement between raters (

Fenton, Cecero, Nich, Frankforter, & Carroll, 2001

;

Hilliard, Henry, & Strupp, 2000

;

Tichenor & Hill, 1989

;

Tryon, Blackwell, & Felleman, 2007

), with clients reporting higher estimates of alliance than therapist or independent

observers. Studies comparing the predictive validity of alliance measurements taken

from each perspective demonstrate differential relationships between such ratings

and both treatment variables and client characteristics (e.g.,

Bachelor, Laverdiere, Gamache, & Bordeleau, 2007

;

Horvath & Greenberg, 1994

).

Bachelor et al. (2007)

observed that clients and therapists may have different perceptions of certain aspects

of alliance, such as collaboration. In studies of youth alliance, therapist reports

have shown stronger predictive relations to outcome variables than youth self-reports

(

Shirk & Karver, 2003

), which tend to overestimate the quality of alliance in comparison with therapist

or observer perspectives (

Kendall, 1994

;

Shelef, Diamond, Diamond, & Liddle, 2005

).

In prior studies of children and adolescents receiving CBT for anxiety disorders

(

Kendall, 1994

), nonsignificant associations have been reported between youth-rated alliance and

outcome, with limited variability in the youths' ratings (all high), whereas significant

associations between therapist-rated alliance and outcome have been documented (

Kendall et al., 1997

). In one of the few studies examining observer ratings of adolescent-therapist alliance,

Shelef et al. (2005)

found that observer ratings of alliance were both more normally distributed and

more predictive of outcome (i.e., substance abuse and dependence symptoms) than adolescent

self-report, suggesting the additive value of using observer informants to rate alliance

in adolescents; however, therapist alliance reports were not considered in this study.

To overcome methodological limitations of past research, and in accordance with recommendations

made by researchers in the field (

Faw, Hogue, Johnson, Diamond, & Liddle, 2005

;

Green, 2006

;

Shirk & Karver, 2003

;

Zack et al., 2007

), the current study examines the therapeutic alliance from all three basic perspectives.

Adolescents in the current study were participating in an ongoing randomized controlled

trial of the Unified Protocol for the Treatment of Emotional Disorders in Youth (UP-Y;

Ehrenreich, Buzzella, Trosper, Bennett, & Barlow, 2008

). The UP-Y is a transdiagnostic, cognitive-behavioral approach to treating adolescents

with emotional disorders (in the context of this study, this refers to all anxiety

and depressive disorders), which was adapted from its adult progenitor (

Barlow et al., 2010

), with revisions based on consideration of empirical research regarding normative

adolescent development and a lengthy treatment development and open trial evaluation

process during which modifications were systematically added and examined for their

utility with an adolescent sample (e.g.,

Ehrenreich, Goldstein, Wright & Barlow, 2009

;

Trosper, Buzzella, Bennett & Ehrenreich, 2009

). Similar to other cognitive-behavioral treatments, the overall goals for treatment

involve altering cognitive reappraisal, prevention of emotional avoidance, and reduction

of maladaptive behaviors (

Barlow, Allen, & Choate, 2004

). However, the UP-Y is unique from other cognitive-behavioral treatments in its

transdiganostic approach, which attempts to treat adolescents with a range of emotional

disorders, rather than a single disorder (

Barlow et al, 2004

;

Ehrenreich-May & Bilek, 2012

). Influenced by a model of emotion regulation described by

Gross and Thompson (2007)

, the treatment is designed to target maladaptive emotion regulation strategies hypothesized

to underlie the range of internalizing disorders. Preliminary analyses from the ongoing

randomized controlled trial indicate that for UP-Y study completers, this treatment

is superior to waitlist at a postcondition assessment regarding clinician-rated severity

of principal anxiety or depressive disorder diagnosis,

F (1, 46) = 15.51, p

< .001, ?2

= .26, and in terms of global impression of severity,

F (1, 46) = 8.82, p < .01, ?2

= .20, as assessed by an independent evaluator.

The central aim of this study was to identify factors predicting early alliance in

adolescents with anxiety and/or depressive disorders in the treatment arm of a large

randomized control study. Alliance was assessed from the three basic rater perspectives.

Potential predictor client variables were identified from the literature (e.g.,

Bachelor, Meunier, Laverdiere, & Gamache, 2010

) and included symptom severity (rated by adolescents on a dimensional scale), diagnostic

severity (rated by intake clinicians based on adolescent and parent reports of the

adolescent's symptoms during a semistructured interview), demographic characteristics,

and adolescent-rated interpersonal or social relationship aspects of functioning.

In this study, adolescent self-reports of social support, attachment security, and

social functioning in current family and peer relationships were obtained at pretreatment

to provide global indicators of interpersonal functioning. It was anticipated that

clients' ratings of security in their attachment relationships, high quality social

and family relationships, and high levels of social support would predict stronger

alliances from all three basic rater perspectives. Moreover, client, therapist, and

observer alliance ratings were expected to demonstrate moderate correlations.

Method

Participants

Potential participants were adolescents receiving therapy in the arm of a larger

randomized controlled trial for which data on early alliance were collected. This

study focused on a subset of UP-Y clients who completed at least three initial treatment

sessions during a portion of this trial. The final sample included 31 of 37 participants

who were eligible to take part in the study. The remaining six were not included

because they either did not complete three treatment sessions (

n

= 4) or they did not initiate treatment (

n

= 2) despite providing informed consent and assent to participate following an intake

assessment. Adolescents were 45% male, aged 12 to 17 years (

M = 15.9 years; standard deviation (SD

) = 1.7). Six (19.4%) were parent identified as Caucasian, 19 (61.3%) as Hispanic

American/Latino, 1 (3.2%) as African American, 1 (3.2%) as Asian, and 4 (12.9%) were

identified as "multiethnic" or "other" ethnicity. Adolescents' grade in school ranged

from sixth to twelfth. Parents of adolescents identified their relationship status

as married (

n

= 18; 60.0%), divorced (

n = 5; 16.7%), never married (n

= 1; 3.3%), or remarried (

n

= 5; 16.7%). Mean annual family income was $82,966 (range = $19,000-$300,000). Three

adolescents were using antidepressant medications (SSRI) at or before the intake

evaluation.

Table 1

presents details of the sample.

Click to view image

Enlarge this Image.

Description of Client Sample Completing Working Alliance Inventory

Of the 31 adolescents, 20 (64.5%) were diagnosed with a primary anxiety disorder,

and the remaining 10 (35.5%) were diagnosed with either a primary mood disorder or

with coprimary anxiety and mood disorders. The majority (93.5%) of the sample was

diagnosed with multiple Axis I disorders. Principal (or coprincipal) Axis I diagnoses

were generalized anxiety disorder (33.3%), social phobia (22.2%), obsessive compulsive

disorder (7.4%), panic disorder (3.7%), specific phobia (3.7%), anxiety disorder

not otherwise specified (3.7%), major depressive disorder (18.5%), dysthymic disorder

(3.7%), and depressive disorder not otherwise specified (3.7%). About two-third of

the sample (67.7%) was diagnosed with a comorbid mood disorder. Eight were diagnosed

with at least one comorbid disorder other than an anxiety or depressive disorder

(attention-deficit/hyperactivity disorder,

n = 2; oppositional-defiant disorder, n

= 1; selective mutism,

n = 1; enuresis, n

= 1; impulse control disorder,

n = 2; substance-related disorder,

n = 1; stuttering disorder, n = 1).

Inclusion criteria for this randomized controlled trial included a diagnosis of an

anxiety or unipolar depressive disorder, based on specific guidelines outlined by

Silverman and colleagues (

Silverman & Nelles, 1988

), with a clinical severity rating (CSR) of four or higher, as rated by the clinician

administering a semistructured diagnostic interview at intake, the anxiety disorders

interview schedule for the Diagnostic and Statistical Manual of Mental Disorders,

4th ed

(DSM-IV)

, child version, parent and child reports (ADIS-IV-C/P;

Silverman & Albano, 1996

); no prior or ongoing experience receiving cognitive-behavioral treatment; the absence

of current suicidal or homicidal ideation; no evidence of schizophrenia, bipolar

disorder, pervasive developmental disorder, mental retardation, organic brain syndrome,

severe learning disorders, or any other cognitive impairment that may have prevented

basic comprehension of questionnaire or treatment materials. A medication stabilization

period (i.e., a consistent dose/type of medication for 3 months before the initial

diagnostic assessment; 1-month stabilization period for benzodiazepines) was required

for individuals taking psychotropic medications. Adolescents were also asked to refrain

from changes in medication regimen throughout the treatment program; however, adolescents

were not excluded for changes in medication usage, once they complied with initial

medication stabilization.

Mean pretreatment CSRs of the principal anxiety or depressive disorder was 5.62 (

SD

= .90), and the mean severity rating of the total number of diagnoses assigned at

intake was 5.12 (

SD

= .68). The mean current global assessment of functioning (

American Psychiatric Association, 2000

) at intake was 61.60 (range = 33-80). The mean T score for the Total Anxiety and

Depression scale of the Revised Child Anxiety and Depression Scale (RCADS;

Chorpita, Yim, Moffitt, Umemoto, & Francis, 2000

) was 56.23 (

SD = 15.32; range = 37-90).

Treatment Context and Therapists

Treatment was administered according to the protocol and guidelines of the UP-Y manualized

protocol (

Ehrenreich et al., 2008

) for a minimum of 8 and a maximum of 21 sessions, with treatment lasting a maximum

of 24 weeks. The protocol is broken up into eight modular sections, including five

required sections and three optional sections (

Trosper et al., 2009

). Clients were seen weekly, with each session lasting approximately 50 minutes.

All sessions were held at the Child and Adolescent Mood and Anxiety Treatment Program

at the University of Miami. Ongoing adherence ratings have been collected on a random

sample (approximately 20%) of treatment sessions to ensure compliance with procedures

as set forth in the UP-Y manual.

Therapists (n

= 12) were primarily doctoral-level graduate students in a clinical psychology program,

with approximately 1 to 7 years of experience. One therapist held a doctorate degree

in clinical psychology. They were mostly female (83.3% female). All therapists received

initial training from the primary treatment developer and attended weekly group supervision

led by the same developer. The average number of clients assigned to each therapist

in the current study sample during which alliance data were available, was 2.46,

although a wide range was observed from 1 to 10 clients each. No therapist was ever

seeing more than four UP-Y cases simultaneously during the trial.

Measures

Clinical status

Anxiety Disorders Interview Schedule, Child Version, Parent and Child Reports

The

Anxiety Disorders Interview Schedule, Child Version, Parent and Child Reports

(ADIS-IV-C/P;

Silverman & Albano, 1996

) is a semistructured clinical interview that assesses anxiety and related disorders

in children and adolescents on the basis of criteria set forth in the

DSM-IV (

American Psychiatric Association, 1994

). This interview was conducted at the initial intake by a trained graduate student

or postdoctoral fellow. Clinician-derived diagnoses and severity ratings, as made

by a trained examiner at intake (typically a graduate student clinician or postdoctoral

clinician), reflected a composite of parent- and adolescent-reported diagnoses and

severity ratings, summed using specific guidelines outlined by Silverman and colleagues

(

Silverman & Nelles, 1988

) and supervised weekly for adherence to these guidelines by an expert ADIS-IV-C/P

rater. Diagnoses assigned a CSR of four or above by the clinician on an 8-point scale

(e.g., 0 = absent; 8 = very severely interfering/disabling), were considered clinically

significant diagnoses, whereas those assigned a rating of less than four were considered

subclinical. A primary diagnosis of a specific anxiety disorder, depressive disorder,

or coprincipal anxiety and depression, was assigned by the trained examiner at intake

based on the diagnosis with the highest assigned CSR. Research demonstrates that

the ADIS-IV-C/P has good interrater (

r

= .98 for the ADIS-C;

r

= .93 for the ADIS-P) and test-retest reliability (

k = .76 for ADIS-C; k

= .67 for ADIS-P) (

Silverman & Nelles, 1988

;

Silverman & Eisen, 1992

). A kappa of .92 has been found for overall principal diagnoses using combined ADIS-IV-C/P

information (

Lyneham, Abbott, & Rapee, 2007

).

Symptom severity

Revised Child Anxiety and Depression Scale

The Revised Child Anxiety and Depression Scale (RCADS;

Chorpita et al., 2000

) is a 47-item self-report measure that was administered to adolescents at pretreatment

to assess symptom severity. Items ask about the frequency of symptoms and are rated

on a 4-point scale from 0 (never) to 3 (always). The RCADS was designed to assess

symptoms of DSM-IV anxiety disorders and depression, and demonstrates good internal

consistency, reliability, and validity (

Chorpita et al., 2000

). In the present study, the RCADS total anxiety and depression subscale is used

as a continuous measure of overall symptom severity. The mean coefficient alpha for

the total RCADS score was .96 in the current sample.

Adolescent-rated pretreatment relational characteristics

Adolescent Attachment Questionnaire

Adolescent Attachment Questionnaire (AAQ;

West, Rose, Spreng, Sheldon-Keller, & Adam, 1998

). The AAQ is a nine-item self-report questionnaire designed to measure adolescents'

perceptions of the relationship with their attachment figure. Items assess the adolescent's

confidence in the availability and responsiveness of the attachment figure (Availability),

the amount of anger in the adolescent-parent relationship (Angry Distress), and the

extent to which the adolescent considers and is empathetic to the needs and feelings

of the attachment figure (Goal-corrected partnership). Items are rated on a 5-point

Likert-type scale, from strongly disagree (1) to strongly agree (5). For ease of

interpretation, the AAQ score was reversed in this study, with higher scores indicating

fewer problems on the dimension being measured. For example, high scores on Availability

indicated higher perceived available responsiveness of the attachment figure. The

validity and reliability of the AAQ have been established with clinical and community-based

adolescent populations. The AAQ has also demonstrated strong convergent validity

with the Adult Attachment Interview (

George, Kaplan, & Main, 1985

), the most commonly used measure of attachment in adults (

West et al., 1998

). In the current study, the total score was used to index adolescents' perceptions

of the availability and responsiveness of their attachment figure. The AAQ total

score showed excellent internal consistency within this sample (coefficient a = .90).

Multidimensional Scale of Perceived Social Support

Multidimensional Scale of Perceived Social Support (MSPSS;

Zimet, Dahlem, Zimet, & Farley, 1988

). The MSPSS is a 12-item measure designed to assess perceptions of the adequacy

of social support from specific sources. Two adolescent-rated subscales (Friend and

Family subscales) from this measure were combined in this investigation to index

perceived relationship support. The Friend subscale consists of four items evaluating

the extent to which the individual perceives that he or she receives help and support

from friends. The Family subscale consists of four items assessing the individual's

perceptions of his or her family members' support. Items were rated on a 7-point

Likert-type scale, from not suitable at all (1) to very suitable (7). Both internal

consistency and test-retest reliability have been established, with a = .87 for the

Family subscale and a = .85 for the Friends subscale, and test-retest reliabilities

of .85 and .75, respectively. Construct validity has also been demonstrated, with

high levels of perceived social support associated with low levels of depression

and anxiety symptomology (

Zimet et al., 1988

). Total perceived support was calculated, based on the sum of items from the Friends

and Family subscales. Internal consistency for total perceived support in this study

was a = .85.

Child and Adolescent Social and Adaptive Functioning Scale

Child and Adolescent Social and Adaptive Functioning Scale (CASAFS;

Price, Spence, Sheffield, & Donovan, 2002

). Peer and Family Relationships subscales of the CASAFS were used to assess adolescents'

judgments about their competence in social relationships (i.e., friend and family

relationships). Peer Relationships subscale items assess the youth's judgments of

the extent to which he or she has friendships with the opposite sex, has close friendships,

has contact with friends, participates in social activities, spends spare time alone,

and has difficulty making friends. Family Relationships subscale items indicate how

well the youth gets along with individual family members (i.e., mother, father, siblings,

and relatives), the extent to which the youth fights with his or her parents, and

the availability of an adult with whom the youth can talk to about his or her problems.

Adolescents answer questions on a 4-point scale from never (1) to always (4). Family

relationship items include a fifth scoring category stating "does not apply to me,"

which was included for scenarios in which the question was inapplicable (e.g., adolescents

without siblings or one of their parents). The CASAFS has adequate internal consistency,

with coefficients alphas of .67 for Peer Relationships and .74 for Family Relationships

(

Price et al., 2002

). The CASAFS also has excellent stability with 12-month test-retest correlations

of .59 for Peer Relationship subscale and .54 for Family Relationship subscale. The

CASAFS has good concurrent validity with significant correlations between the subscales

and the Beck Depression Inventory. The Peer Relationships subscale and Family Relationships

subscale were combined in this study to form a measure of social relationship functioning;

higher scores indicate higher level of social functioning. Internal consistency of

the combined subscales was a = .69.

Alliance assessment

Working Alliance Inventory

Working Alliance Inventory (WAI;

Horvath & Greenberg, 1989

;

Tracey & Kokotovic, 1989

). Adolescents and therapists completed the short form of the WAI after the third

therapy session (or as soon as possible thereafter). The short version of the WAI

includes 12-items rated on a 7-point Likert-type scale from never (1) to always (7),

with items reflecting the three components of alliance proposed by

Bordin (1979)

: agreement on tasks (e.g., "My therapist and I agree about the things I will need

to do in therapy to help improve my situation"), agreement on goals (e.g., "My therapist

and I are working toward goals that we both agree on"), and bond (e.g., "My therapist

and I trust one another"). Client and therapist versions of the scale are identical

in content and format. The items on each measure are summed to provide a total score;

higher scores reflect a greater quality of therapeutic alliance.

This measure was selected because of its common use in research with adolescent populations

(e.g.,

Hintikka, Laukkanen, Marttunen, & Lehtonen, 2006

), its ease of application to cognitive-behavioral treatment samples (

Raue & Goldfried, 1994

), and its excellent reliability and validity (

Horvath & Greenberg, 1994

;

Tichenor & Hill, 1989

). Extensive research has attested to diverse forms of reliability and validity of

this inventory in adult psychotherapy (

Horvath & Bedi, 2002

). The short form has been found to have comparably good reliability and validity

as compared with the WAI and has been used commonly in youth studies (e.g.,

Shelef et al., 2005

;

Tetzlaff et al., 2005

;

Wintersteen, Mensinger, & Diamond, 2005

). The short form has excelled reliability with Cronbach's a = .98 for the patient

version and a = .95 for the therapist version (

Tracey & Kokotovic, 1989

;

Tryon & Kane, 1993

). Only the total score was used in the current study, given the evidence of a general

alliance factor in adolescents (e.g.,

DiGiuseppe et al., 1996

). The WAI showed excellent internal consistency within this study sample, as indicated

by a = .89 for the adolescent report and a = .91 for the therapist report.

Vanderbilt Therapeutic Alliance Scale

Vanderbilt Therapeutic Alliance Scale-Revised (VTAS-R;

Diamond, Liddle, Dakof, & Hogue, 1996

). Observer ratings of therapeutic alliance were made using a revised version of

the Vanderbilt Therapeutic Alliance Scale (VTAS,

Hartley & Strupp, 1983

), which defines the therapeutic alliance as a collaborative and task-oriented relationship

determined by client behaviors and therapist-client relationship characteristics.

The revised version includes 24 items, split into Patient Contribution (e.g., "To

what extent did the patient acknowledge that he had a problem, which the therapist

could help him with?") and Patient-Therapist Contribution (e.g., "To what extent

did the therapist and patient share a common viewpoint about the definition, possible

causes, and potential alleviation of the patient's problems?") subscales; these items

reflect the three aspects of the therapeutic alliance (i.e., the bond, task, and

goal components) proposed by

Bordin (1979)

. Each item is rated on a Likert-type scale ranging from not at all (0) to a great

deal (5). The revised VTAS has been implemented in a number of studies with adolescents

(e.g.,

Diamond et al., 1999

;

Hogue, Dauber, Stambaugh, Cecero, & Liddle, 2006

;

Robbins et al., 2003

;

Shelef et al., 2005

) and has demonstrated adequate reliability and construct and predictive validity.

Analyses of the revised version indicate strong interrater agreement, as assessed

by intraclass correlation (ICC) of .80 and a coefficient alpha estimate of .95 (

Diamond et al., 1999

). Based on prior research with this measure suggesting one general alliance dimension

in adolescents, only the total alliance score was used in this study.

Procedure

Institutional Review Board approval was obtained for this investigation before the

initiation of data collection. After an initial intake, adolescents were randomized

to either immediate UP-Y treatment or an 8-week waitlist/attentional-control condition.

Those randomized to the waitlist condition began treatment approximately 8 weeks

after randomization, immediately before which they were administered a brief version

of the ADIS-IV-C/P. Written informed consent from parents and assent from adolescents

were obtained at the onset of the first interview and before beginning treatment.

Data on client pretreatment characteristics were obtained at the intake assessment,

during which the adolescent and caregiver completed a semistructured interview (the

ADIS-IV-C/P), as well as various questionnaires. Data on client, therapist, and observer

ratings of alliance were obtained during or immediately after the third therapy session

with the adolescent (or as soon as possible thereafter). Session 3 was chosen because

prior research (

O'Malley, Suh, & Strupp, 1983

) indicates that alliance assessed early in treatment is more predictive of treatment

outcomes than alliance assessed later in treatment (

Hersoug, Monsen, Havik, & Hoglend, 2002

;

Horvath, Del Re, Fluckiger & Symonds, 2011

;

Martin et al., 2000

), and because alliance assessed early in treatment reduces the potential confound

between alliance scores and symptom improvement over the course of therapy (

Feeley, DeRubeis, & Gelfand, 1999

). Adolescents and therapists independently completed the therapeutic alliance measure

(the WAI) immediately after the third treatment session. Adolescents were informed

that their therapist would not see or have access to the survey and placed the completed

survey in a sealed envelope. Session 3 was recorded on a DVD and subsequently viewed

by trained raters assessing alliance.

Alliance Raters and Training

Raters were one graduate student in a Clinical Psychology PhD program and two external

raters (one undergraduate and one postbaccalaureate rater; both with majors in Psychology)

who completed rater-training procedures; all three raters were women, aged 19 to

28 years, and were of European American, Hispanic American, and South Asian/American

descents, respectively. The two nongraduate student raters were selected to be naive

to the study design and hypotheses. The graduate rater received training to use the

VTAS-R for research purposes, conducted primarily by one of the developers of the

revised version of this measure for use with adolescents. After training, the two

nongraduate student raters were trained by the graduate rater. Training sessions

were 2 hours in duration and were conducted twice a week over the course of 2 months.

Coder training included studying the manual, viewing and independently rating videotapes,

and ongoing discussions to clarify scoring dilemmas. Training tapes were not drawn

from participants included in the present study sample. After this training, raters

were given five practice sessions to rate. Analyses indicated that for these practice

sessions, raters achieved excellent interrater reliability (intraclass correlation

coefficient [ICC = .94]) (

Shrout & Fleiss, 1979

), and they were therefore deemed competent to begin coding actual study tapes (

Diamond et al., 1999

;

Robbins et al., 2003

).

Alliance Observational Coding Procedure

The total number of observer-rated alliance sessions coded was 24. The number of

observer alliances collected was slightly lower than the number of self- and therapist-rated

alliances because of some technical problems recording or viewing select sessions

and because two participants did not consent videotaping of sessions. In 20 of the

24 cases, the observer ratings of the alliance were based on the same session as

was available for therapist and client reports. In the remaining four cases, observer

ratings were obtained within one session after self-reports were collected.

Session alliance was coded by one to two raters. Of the 24 alliance sessions that

were coded, 15 (62.5%) were coded by a second observer to provide an estimate of

observer agreement. Raters were assigned sessions in rotating, random order. Ratings

were made independently after viewing the entire session, with one to two raters

separately coding each alliance. Without consulting one another, a pair of raters

watched session videotapes together. Final alliance scores were generated after discussion

between the raters. Rater disagreements were discussed using their notes to substantiate

their score. The ratings were averaged if an agreement could not be reached or if

the ratings differed by only one point. Raters were not allowed to review the session

DVD in the case of disagreements, to prevent any rater bias that could confound reliability.

Raters were instructed to fast-forward any segments of the session during which parent(s)

of the adolescent were present (e.g., parent "check-ins"). This decision was made

to minimize the possibility of confounding effects of a parents' presence on ratings

of adolescent-therapist alliance scores. Weekly recalibration meetings were held

during the study to prevent rater drift. Interrater reliability was computed regularly

throughout the study.

VTAS: Interrater reliability and scale properties

An analysis of interrater reliability using intraclass correlation coefficients (

Shrout & Fleiss, 1979

) was conducted with data from 15 sessions that were rated by multiple raters. Consistent

with prior research on the VTAS (

Shelef et al., 2005

), raters were able to achieve a high degree of rater reliability in this investigation.

The ICCs for the items ranged from good (.65) to excellent (.91) (

Cicchetti, 1994

), except for three items: "Refer back to experiences they have been through together

(.23)," "Accept their different roles and responsibilities as part of their relationship

(.33)," and "Make an effort to carry out therapeutic procedures suggested by the

therapist (.59)." Because the three items could not be coded reliably, they were

eliminated from all subsequent analyses. Raters achieved a mean ICC of .88, when

the three items were removed for the total scale. An internal consistency analysis

performed on the 21 VTAS items produced a Cronbach's coefficient alpha of .97. Similar

results were reported by

Shelef et al. (2005)

, suggesting that the VTAS is a reliable measure of therapist-adolescent alliance

for this population.

Results

Preliminary Analyses

Descriptive statistics for independent (interpersonal) variables

Table 2

presents the means (or percentages) and SDs for pretreatment measures obtained for

the present study sample. The sample size for each measure varies slightly because

of variations in the availability of data. Visual inspection of histograms for the

total scales of the MSPSS, the CASAFS, and the AAQ suggest near-normal distributions.

Adolescents' scores on the CASAFS were truncated, with most items rated at a 3 or

a 4 on a 4-point Likert scale. Total scores across the three interpersonal measures

(i.e., the CASAFS, MSPSS, and AAQ) were moderately to strongly and significantly

correlated (see

Table 3

).

Click to view image

Enlarge this Image.

Means and Standard Deviations of Study Variables

Click to view image

Enlarge this Image.

Correlations of Independent Variables

Descriptive data on alliance

The means and SDs for alliance scores are also presented in

Table 2

. The total number of WAI ratings made by adolescents and therapist were 30 and 29,

respectively. Neither therapist nor adolescent WAI data from one participant in the

full sample (

n

= 31) were collected because of a clerical error. In addition, therapist ratings

from one participant treated by the first author of this investigation were eliminated

to minimize study bias. Adolescent self-reported alliance scores (

N

= 30) ranged from 42.00 to 84.00, with a mean total score of 69.57 (

SD

= 11.29). Therapist-reported alliance scores (

N

= 29) ranged from 40.00 to 77.00, with a mean total score of 64.44 (

SD

= 8.41). The mean ratings (from 1 to 7) for adolescent and therapist alliances were

5.71 (

SD

= .94) and 5.37 (SD

= .70), respectively. Independent observer alliance scores on the VTAS-R (

N

= 24) ranged from 10 to 93, with a mean total score of 67 (

SD

= 22.5). The difference in magnitudes between therapist and adolescent-reported alliance

fell short of significance,

t = 1.86,

p

< .07, but indicates a trend for adolescents to rate alliance more positively than

therapists. Visual inspection of histograms and frequency distributions indicate

that adolescent self-report scores were truncated, with 47% of the sample scoring

an average of 6 or above on the scale. In contrast, only 14% of the therapist-rated

alliance scores fell above 6. This difference was significant,

T = 2.83, p < .01.

Cross-informant agreement on therapeutic alliance

Correlational analyses using two-tailed Pearson r

's were conducted to examine the intercorrelations among the three alliance rater

perspectives. Results indicate that observer ratings of alliance were strongly correlated

with both therapist and adolescent ratings,

r = .55, p < .01, and r = .65, p

< .01, respectively. Adolescent and therapist alliance ratings were moderately correlated,

r = .39, p

< .05. This moderate correlation between youth and therapist reports is consistent

with findings reported in past research (

Hawley & Garland, 2008

;

Shirk & Karver, 2003

). These correlations demonstrate that although the three perspectives are related,

there is still significant and unique variance captured by each informant's report.

Predictors of alliance

Demographic and clinical variables

Initial analyses were conducted to determine whether client demographic and clinical

variables were associated with therapeutic alliance (see

Table 4

). Examination of the client demographic characteristics failed to reveal any significant

associations. There was a trend for observers to rate alliance more positively for

female participants, although this was not significant

T = -1.87, p

= .07. The presence of a comorbid depressive disorder was significantly associated

with therapist-rated alliance scores,

T

= 1.95,

p

< .05, indicating that therapists perceived a poorer alliance with participants exhibiting

clinically significant depressive symptoms. Inspection of mean differences indicated

that therapists rated alliance lower in participants with a comorbid depressive disorder

(

n = 20, M = 62.50, SD

= 9.21) than they did patients without a comorbid depressive disorder (

n = 9, M = 68.78, SD

= 3.96). Neither adolescent nor observer rater perspectives were found to differ

significantly based on the presence of a depressive disorder. In addition, adolescents'

self-reported ratings of symptoms based on the RCADS total scale score were significantly

associated with observer ratings of alliance,

r = .41, p

< .05, indicating that adolescents who reported a higher level of overall anxiety

and depressive symptomatology were observed to establish stronger alliances with

their therapists than adolescents who reported fewer symptoms. Adolescent symptom

severity ratings were not associated with therapist or adolescent ratings of alliance.

Click to view image

Enlarge this Image.

Relationship Between Pretreatment Variables and Alliance Ratings

Interpersonal variables

Stepwise regression analyses (

Table 5

) were used to identify the most salient interpersonal predictors of alliance in

this sample. The three measures reflecting adolescents' self-reported perceptions

of their interpersonal relations (i.e., CASAFS, MSPSS, AAQ), as well as adolescent

self-reported symptomology (based on RCADS Total score), were entered as predictors

into three separate regression models. The rationale for including RCADS into the

regression was based on the recognition that adolescents present with varying levels

of symptom severity that may have a confounding effect on their self-report (

Muran, Segal, & Samstag, 1994

). The dependent variable was the alliance score, with separate regressions for each

alliance rater perspective.

Click to view image

Enlarge this Image.

Stepwise Regression Statistics For Equations Predicting Alliance From Interpersonal

Pretreatment Measures

Stepwise analyses identified the AAQ as a significant predictor of adolescents' alliance

ratings, ß = .43,

T

= 2.42, p < .05, accounting for 18% of the variance,

R 2 = .18, p

< .05. This indicates that adolescents' perceptions of security in their relationships

with their caregivers predicted their perceptions of alliance with their therapist.

In addition, the MSPSS significantly predicted observer ratings of alliance, ß =

.51,

T

= 2.69,

p < .05, accounting for 26% of the variance,

R 2 = .26, p

< .05. This finding indicates that adolescents who reported higher levels of support

from their relationships with family and peers were observed by independent raters

to have stronger alliances. None of the three interpersonal measures were related

to therapists' ratings of alliance.

Discussion

The present study addresses the need for additional research on therapeutic alliance

with adolescent populations and is among the first to examine client pretreatment

variables predictive of early alliance among adolescents receiving cognitive-behavioral

treatment for anxiety and depressive disorders. While much is known about the essential

role of therapeutic alliance in treatment outcome, less is known about what predicts

this critical relationship. This study sought to expand this knowledge base by examining

predictors of early alliance among a sample of adolescents being treated for anxiety

and depressive disorders.

One goal of the study was to examine the degree of associations between three distinct

alliance rater perspectives (adolescent, therapist, observer). The association between

therapist- and client-alliance ratings was moderate, but each informant's report

was strongly correlated with observer ratings. Such patterns are consistent with

findings from previous research on adolescent alliance (

Garner, Godley, & Funk, 2008

;

Hawley & Garlan, 2008

;

Shelef & Diamond, 2008

;

Shirk & Karver, 2003

), and suggest that adolescent and therapist perceptions of alliance may capture

similar yet distinct processes. Consistent with past research, adolescents rated

the alliance more positively than therapists, with nearly 50% of adolescents rating

alliance at an average of 6 or above (on a 7-point scale), whereas less than 15%

of therapists provided an average alliance score of 6 or above. This elevated level

in client ratings of alliance is consistent with previous studies of both adolescent

and adult alliance (e.g.,

Shelef et al., 2005

;

Tryon, Blackwell, & Felleman, 2008

).

Fenton et al. (2001)

suggested that elevated-alliance scores in adult clients with substance abuse disorders

might reflect a fear of expressing negative feelings about the therapist. This may

also be true for adolescents with anxiety and/or depressive disorders. Another factor

that could account for the extremely positive client ratings may be the lack of comparison

with other standards (

Fenton et al., 2001

). In this study, adolescents likely had fewer points of reference from which to

judge the quality of alliance, in contrast to therapist and observer ratings, which

were both conducted on multiple occasions and often across several adolescent participants.

Given the substantial evidence for a link between alliance and treatment outcome

indicating the role of a strong therapeutic alliance on successful treatment outcome

(

Horvath, Del Re, Fluckiger & Symonds, 2011

;

Martin, Gaske, & Davis, 2000

), the second goal of the study was to examine client factors contributing to the

development of a strong alliance. Adolescents who perceived their relationships with

their caregivers as more secure also reported more positive alliances with their

therapist. Not surprisingly, those adolescents who characterized their relationship

with their caregiver as higher in availability and responsiveness to their own needs,

higher in empathy, and lower in anger directed toward their caregivers were more

likely to perceive the therapeutic relationship as trusting, warm, and amenable.

This finding is consistent with the notion in attachment theory that children use

their relationships with caregivers to create internal working models, which carry

over to other relationships, including the therapeutic relationship (

Bowlby, 1973

). These results are all the more noteworthy in light of the limited variability

in adolescent-rated alliance scores, which would be expected to attenuate such associations.

In addition, observer-rated alliance scores were predicted by adolescents' self-reports

of social support from family and peers. That is, adolescents who reported having

people to talk to and rely on, or having people that could help them figure out how

to cope with problems, were more likely to receive positive ratings of alliance from

blind observers. This finding is consistent with previous research that has found

evidence for the importance of social support in predicting therapeutic alliance

(

Garner et al., 2008

;

Meier, Donmall, Barrowclough, McElduff, & Heller, 2005

). From an attachment-based perspective, higher self-reported ratings of social support

and attachment security may reflect a positive working model and trust in the benevolence

of other people. With their tendency to seek contact and help from the others, more

secure adolescents may find it easier to create a trusting and emotionally close

relationship with their therapists.

Adolescent-reported measures of attachment security and social support accounted

for 18% and 26% of the variance in adolescent-rated and observer-rated alliance,

respectively, whereas symptom severity at intake did not account for a significant

proportion of variability in either rating of alliance. The important contributions

of interpersonal factors to the formation of therapeutic alliances support the potential

usefulness of incorporating interpersonal relationship factors into assessment and

intervention protocols using this treatment modality. For example, more thoroughly

assessing pretreatment interpersonal factors in patients might improve early identification

of patients who might require additional focus on alliance building factors early

in treatment. Therapists could, in turn, use information both on patients' interpersonal

problems and strengths to consider ways of engaging and motivating patients and family

members more effectively. Therapists may also need to consider modifying their approach

for clients with pretreatment indications of interpersonal problems and their families

in order to engage them in a more productive relationship.

The UP-Y, as a transdiagnostic CBT modality, is potentially robust to these issues,

given therapist flexibility in the ability to use interpersonal issues as potential

examples of emotional experiences under consideration (see

Ehrenreich et al., 2008

) and opportunities to reinforce skills related to the protocol's main foci: cognitive

reappraisal, acceptance and experiencing of intense emotions, and behavioral engagement.

However, additional research examining the role of such interpersonal variables in

UP-Y outcomes would be required to better understand the protocol's ability to appropriately

handle such concerns thoroughly.

Therapist-rated alliance was not predicted by any of the three adolescent self-reported

interpersonal measures. The majority of therapists in this study were at relatively

early stages of training, and alliance ratings of more junior therapists may differ

in important ways from experienced therapists, particularly regarding the task and

goal dimensions of alliance (

Mallinckrodt, 1991

). It is unclear whether such experiential differences might help explain nonsignificant

findings observed for therapist reports in this investigation. An alternative or

additional explanation for the lack of prediction of interpersonal factors on therapist-rated

alliance may be that therapists were unaware of issues related to attachment security

or social support in their clients because assessment and discussion of these factors

were not explicitly part of the treatment manual or the clinical assessment data

they received before onset of treatment.

In contrast, therapist-reported alliance was related to the presence of a comorbid

depressive disorder. Therapists' perceptions of lower alliances in depressed patients

may be, in part, explained by the symptom constellation in depressed patients (i.e.,

decreased motivation and loss of interest), which may create reluctance on both sides

to engage in the therapeutic relationship. Given the research demonstrating a tendency

for depressed adolescents to withdraw from family and friends (

Puig-Antich, Kaufman, Ryan, & Williamson, 1993

), it would not be surprising if depressed adolescents were also more likely to refuse

to participate in aspects of treatment (e.g., homework assignments or in-session

activities). This may, in turn, increase negative emotions (e.g., hostility) in therapists.

This finding points to the potential importance of helping therapists learn to work

with and tolerate their negative emotional reactions and possible feelings of hostility

and/or withdrawal from depressed patients, who represent themselves as less motivated

and whose symptoms render them at odds for developing a productive therapeutic relationship.

Although symptomology did not yield significant predictions in the regression models,

correlational analyses revealed that adolescents who reported higher symptom levels

on the RCADS were rated as having stronger alliances by blind observers. The positive

association between alliance and symptom severity has been found in other research

studies (

Garner et al., 2008

). This makes sense intuitively because greater acknowledgment of one's problems

may be associated with an increased number and/or intensity of symptoms experienced.

Such client acknowledgment of problems is likely to increase motivation and engagement

in the therapeutic tasks to alleviate symptoms. For example, adolescents who acknowledge

more problems may be more likely to want and accept help, support, and feedback from

a therapist, creating the circumstances for a positive working alliance. Moreover,

therapists may be more comfortable with adolescents who report or acknowledge higher

levels or intensity of symptoms-therapists may both feel gratified by the client's

desire for their help and may also feel more confident in their own work because

of the broader range of issues to work with.

Although this study provides important information about potential interpersonal

predictors of alliance in youth CBT, certain study limitations are acknowledged.

First, given the small sample size, the power to detect significant associations

was reduced; therefore, future investigations should replicate these findings in

a larger sample size. Second, indices used to assess interpersonal factors at pretreatment

were only obtained from adolescents' own reports. This poses two distinct problems.

The first is the issue of shared method variance associated with using the same informant

to report on both independent predictors and the dependent variable (alliance). In

addition, self-reported measures of attachment and interpersonal functioning are

limited by the fact that some aspects of the interpersonal factors measured are implicit

and may occur outside patients' awareness. Future studies, therefore, need to assess

interpersonal qualities using multiple informant perspectives. A third limitation

in the current study is that only participant variables were assessed as alliance

predictors. The findings suggesting that some therapists consistently respond to

their patients in ways that systematically influences the alliance indicate that

future research should include therapist characteristics (

Dinger, Sachsse, & Schauenburg, 2009

).

It is also worth noting that results from this study are based on adolescent clients

for which sufficient data were collected (i.e., the completion of three treatment

sessions), whereas adolescents who dropped out before the third session were excluded

from the study. Although the number of adolescents that terminated early in this

investigation was minimal, it is possible that adolescents who dropped out before

session three had certain interpersonal characteristics that may have predisposed

them to problems in the alliance. In turn, ratings of both interpersonal pretreatment

characteristics and alliance may reflect overly positive scores with respect to characteristics

that potentially pose a threat to the generalizability of these findings. Future

studies would therefore do well by examining predictors of drop-out in conjunction

with alliance. A final factor worth considering is that therapists were aware of

which sessions were coded for alliance, which may have impacted the interactions

during the third session.

Despite these limitations, this study is the first to assess predictors of therapeutic

alliance from the three basic rater perspectives within a research-based, cognitive-behavioral

intervention for adolescents with anxiety and depressive disorders. This is especially

important in light of research with similar treatment populations in which youth

have rated the therapeutic alliance as the most important aspect of treatment (Southam-Gerow,

1996). Additionally, the findings of this study have important implications for the

understanding factors related to the prediction of therapeutic alliance, which has

been documented as being important in therapeutic outcome. Indeed, interpersonal

factors may be useful to consider in identifying youth who may be more difficult

to engage across psychosocial treatment approaches for anxiety and depression.

This study increases our understanding of factors that may be of import in alliance

and ultimately treatment outcome in adolescents receiving CBT. If pretreatment interpersonal

factors may predispose therapeutic alliance problems, it seems important to identify

such characteristics as early as possible so that therapists can better recognize

potential pitfalls with patients and adjust accordingly. Such information seems essential,

as researchers move beyond understanding treatment outcome studies alone to understanding

how to remedy poor alliance and thereby improve treatment efficacy (

Muran et al., 1994

). Early therapists, including those explicitly focusing on training in evidence-based

interventions and/or behavioral approaches, might also benefit from explicit training

in factors pertaining to evaluating and developing an alliance, including how to

develop an alliance with difficult patients, how to recognize when the alliance is

fragile, and how to work to develop and repair weak alliances.

References

1

. Achenbach, T. M. (1991). Manual for the child behavior checklist/4-18 and 1991

profile. Burlington, VT: University of Vermont, Department of Psychiatry.

2

. (1994). Diagnostic and statistical manual of mental disorders. Washington, DC:

American Psychiatric Association.

3

. (2000). Diagnostic and statistical manual of mental disorders. Washington, DC:

American Psychiatric Association.

4

. Bachelor, A., Laverdiere, O., Gamache, D., & Bordeleau, V. (2007). Psychotherapy:

Theory, Research, Practice, Training.

5

. Bachelor, A., Meunier, G., Laverdiere, O., & Gamache, D. (2010). Psychotherapy:

Theory, Research, Practice, Training.

6

. Barlow, D. H., Allen, L. B., & Choate, M. L. (2004). Behavior Therapy.

7

. Barlow, D. H., Ellard, K. K., Fairholme, C. P., Farchione, T. J., Boisseau,

C. L., Allen, L. B., & Ehrenreich-May, J. (2010). The unified protocol for transdiagnostic

treatment of emotional disorders, therapist guide. New York, NY: Oxford University

Press.

8

. Bordin, E. S. (1979). Psychotherapy: Theory, Research, and Practice.

9

. Bordin, E. S., Horvath, A. O., & Greenberg, L. S. (1994). The working alliance:

Theory, research, and practice. New York, NY: Wiley.

10

. Bowlby, J. (1973). Separation: Anxiety and anger. New York: Basic Books.

11

. Bretherton, I., Munholland, K. A., Cassidy, J., & Shaver, P. R. (1999). Handbook

of attachment: Theory, research, and clinical applications. New York, NY: Guilford

Press.

12

. Broome, K. M., Joe, G. W., & Simpson, D. D. (2001). Journal of Adolescent Research.

13

. Chorpita, B. F., Yim, L., Moffitt, C., Umemoto, L. A., & Francis, S. E. (2000).

Behaviour Research and Therapy.

14

. Chu, B. C., Choudury, M. S., Shortt, A. L., Pincus, D. B., Creed, T. A., &

Kendall, P. C. (2004). Cognitive and Behavioral Practice.

15

. Cicchetti, D. V. (1994). Psychological Assessment.

16

. Diamond, G. M., Liddle, H. A., Dakof, G., & Hogue, A. (). Revised version of

the Vanderbilt Therapeutic Alliance Scale.

17

. Diamond, G. M., Liddle, H. A., Hogue, A., & Dakof, G. A. (1999). Psychotherapy:

Theory, Research, Practice, Training.

18

. DiGiuseppe, R., Linscott, J., & Jilton, R. (1996). Applied and Preventive Psychology.

19

. Dinger, U., Strack, M., Sachsse, T., & Schauenburg, H. (2009). Psychotherapy:

Theory, Research, Practice, Training.

20

. Ehrenreich-May, J. T., & Bilek, E. L. (2012). Cognitive and Behavioral Practice.

21

. Ehrenreich, J. T., Buzzella, B. A., Trosper, S. E., Bennett, S. M., & Barlow,

D. H. (). The Unified Protocol for treatment of emotional disorders in adolescents.

22

. Ehrenreich, J. T., Goldstein, C. R., Wright, L. R., & Barlow, D. H. (2009).

Child & Family Behavior Therapy.

23

. Eltz, M. J., Shirk, S. R., & Sarlin, N. (1995). Child Abuse & Neglect.

24

. Faw, L., Hogue, A., Johnson, S., Diamond, G. M., & Liddle, H. A. (2005). Psychotherapy

Research.

25

. Feeley, M., DeRubeis, R. J., & Gelfand, L. A. (1999). Journal of Consulting

and Clinical Psychology.

26

. Feeney, B. C., Cassidy, J., & Marcuse, F. (2008). Journal of Personality and

Social Psychology.

27

. Fenton, L. R., Cecero, J. J., Nich, C., Frankforter, T. L., & Carroll, K. M.

(2001). Journal of Psychotherapy Practice and Research.

28

. Garcia, J. A., & Weisz, J. R. (2002). Journal of Consulting and Clinical Psychology.

29

. Garner, B. R., Godley, S. H., & Funk, R. R. (2008). Journal of Psychoactive

Drugs.

30

. George, C., Kaplan, N., & Main, M. (). Attachment interview for adults.

31

. Green, J. (2006). Journal of Child Psychology and Psychiatry.

32

. Gross, J. J., Thompson, R. A., & Gross, J. J. (2007). Emotion regulation: Conceptual

foundations. New York, NY: The Guilford Press.

33

. Hartley, D. E., Strupp, H. H., & Masling, J. (1983). Empirical studies of psychoanalytic

theory. Hillsdale, NJ: Erlbaum.

34

. Hawley, K., & Garland, A. (2008). Child & Youth Care Forum.

35

. Hersoug, A. G., Monsen, J. T., Havik, O. E., & Hoglend, P. (2002). Psychotherapy

and Psychosomatics.

36

. Hilliard, R. B., Henry, W. P., & Strupp, H. H. (2000). Journal of Consulting

and Clinical Psychology.

37

. Hintikka, U., Laukkanen, E., Marttunen, M., & Lehtonen, J. (2006). Bulletin

of the Menninger Clinic.

38

. Hogue, A., Dauber, S., Stambaugh, L. F., Cecero, J. J., & Liddle, H. A. (2006).

Journal of Consulting and Clinical Psychology.

39

. Horvath, A. O., Bedi, R. P., & Norcross, J. C. (2002). Psychotherapy relationships

that work: Therapist contributions and responsiveness to patients. New York, NY:

Oxford University Press.

40

. Horvath, A. O., & Greenberg, L. S. (1989). Journal of Counseling Psychology.

41

. Horvath, A. O., & Greenberg, L. S. (1994). The working alliance: Theory, research,

and practice. New York, NY: John Wiley & Sons.

42

. Horvath, A. O., & Luborsky, L. (1993). Journal of Consulting and Clinical Psychology.

43

. Horvath, A. O., Del Re, A. C., Fluckiger, C., & Symonds, D. (2011). Psychotherapy.

44

. Horvath, A. O., Horvath, A. O., & Greenberg, L. S. (1994). The working alliance:

Theory, research, and practice. New York, NY: Wiley.

45

. Horvath, A. O. (2001). Psychotherapy: Theory, Research, Practice, Training.

46

. Karver, M. S., Shirk, S. R., Handelsman, J. B., Fields, S., Crisp, H., Gudmundsen,

G., & McMakin, D. (2008). Journal of Emotional and Behavioral Disorders.

47

. Kazdin, A. E. (1996). Clinical Child Psychology and Psychiatry.

48

. Kendall, P. C., Comer, J. S., Marker, C. D., Creed, T. A., Puliafico, A. C.,

Hughes, A. A., Martin, E. D., Suveg, C., & Hudson, J. (2009). Journal of Consulting

and Clinical Psychology.

49

. Kendall, P. C., Flannery-Schroeder, E., Panichelli-Mindel, S. M., Southam-Gerow,

M., Henin, A., & Warman, M. (1997). Journal of Consulting and Clinical Psychology.

50

. Kendall, P. C. (1994). Journal of Consulting and Clinical Psychology.

51

. Liber, J. M., McLeod, B. D., Van Widenfelt, B. M., Goedhart, A. W., van, der

Leeden, A. J., Utens, E. M., & Treffers, P. D. (2010). Behavior Therapy.

52

. Liddle, H. A. (1998). Clinical Child Psychology and Psychiatry.

53

. Lyneham, H. J., Abbott, M. J., & Rapee, R. M. (2007). Journal of the American

Academy of Child & Adolescent Psychiatry.

54

. Mallinckrodt, B. (1991). Journal of Counseling Psychology.

55

. Mallinckrodt, B. (2000). Psychotherapy Research.

56

. Martin, D. J., Gaske, J. P., & Davis, M. K. (2000). Journal of Consulting and

Clinical Psychology.

57

. Meier, P. S., Donmall, M. C., Barrowclough, C., McElduff, P., & Heller, R.

F. (2005). Addiction.

58

. Muran, J. C., Segal, Z. V., Samstag, L. W., & Crawford, C. E. (1994). Journal

of Consulting and Clinical Psychology.

59

. Muran, J. C., Segal, Z. V., & Samstag, L. W. (1994). Behavior Therapy.

60

. O'Malley, S. S., Suh, C. S., & Strupp, H. H. (1983). Journal of Consulting and

Clinical Psychology.

61

. Price, C. S., Spence, S. H., Sheffield, J., & Donovan, C. (2002). Journal of

Clinical Child and Adolescent Psychology.

62

. Puig-Antich, J., Kaufman, J., Ryan, N. D., & Williamson, D. E. (1993). Journal

of the American Academy of Child and Adolescent Psychiatry.

63

. Raue, P. J., Goldfried, M. R., Horvath, A. O., & Greenberg, L. S. (1994). The

working alliance: Theory, research, and practice. New York, NY: Wiley & Sons.

64

. Robbins, M. S., Liddle, H. A., Turner, C. W., Dakof, G. A., Alexander, J. F.,

& Kogan, S. M. (2006). Journal of Family Psychology.

65

. Robbins, M. S., Turner, C. W., Alexander, J. F., & Perez, G. A. (2003). Journal

of Family Psychology.

66

. Shelef, K., Diamond, G. M., Diamond, G. S., & Liddle, H. A. (2005). Journal

of Consulting and Clinical Psychology.

67

. Shelef, K., & Diamond, G. M. (2008). Psychotherapy Research.

68

. Shirk, S. R., Gudmundsen, G., Kaplinski, H. C., & McMakin, D. L. (2008). Journal

of Clinical Child and Adolescent Psychology.

69

. Shirk, S. R., Karver, M. S., & Brown, R. (2011). Psychotherapy (Chicago, Ill).

70

. Shirk, S. R., & Karver, M. S. (2003). Journal of Consulting and Clinical Psychology.

71

. Shirk, S. R., & Saiz, C. (1992). Development and Psychopathology.

72

. Shrout, P. E., & Fleiss, J. L. (1979). Psychological Bulletin.

73

. Silverman, W. K., & Albano, A. M. (1996). Clinical Manual for the Anxiety Disorders

Interview Schedule for DSM-IV-Child Version. San Antonio, TX: The Psychological Corporation.

74

. Silverman, W. K., & Eisen, A. R. (1992). Journal of the American Academy of Child

and Adolescent Psychiatry.

75

. Silverman, W. K., & Nelles, W. B. (1988). Journal of the American Academy of

Child and Adolescent Psychiatry.

76

. Tetzlaff, B. T., Kahn, J. H., Godley, S. H., Godley, M. D., Diamond, G. S.,

& Funk, R. R. (2005). Psychology of Addictive Behaviors.

77

. Tichenor, V., & Hill, C. E. (1989). Psychotherapy: Theory, Research, Practice,

Training.

78

. Tracey, T. J., & Kokotovic, A. M. (1989). Psychological Assessment: A Journal

of Consulting and Clinical Psychology.

79

. Trosper, S. E., Buzzella, B. A., Bennett, S. M., & Ehrenreich, J. T. (2009).

Clinical Child and Family Psychology Review.

80

. Tryon, G. S., Blackwell, S. C., & Felleman, H. E. (2007). Psychotherapy Research.

81

. Tryon, G. S., Blackwell, S. C., & Felleman, H. E. (2008). Psychotherapy: Theory,

Research, Practice, Training.

82

. Tryon, G. S., & Kane, A. S. (1993). Journal of Counseling Psychology.

83

. Vaux, A. C., Burda, P. C., & Stewart, D. (1986). Journal of Community Psychology.

84

. West, M., Rose, M. S., Spreng, S., Sheldon-Keller, A., & Adam, K. (1998).

Journal of Youth and Adolescence.

85

. Wierzbicki, M., & Pekarik, G. (1993). Professional Psychology: Research and Practice.

86

. Wintersteen, M. B., Mensinger, J. L., & Diamond, G. S. (2005). Professional

Psychology: Research and Practice.

87

. Young, J. F., Mufson, L., & Davies, M. (2006). Journal of the American Academy

of Child and Adolescent Psychiatry.

88

. Zack, S. E., Castonguay, L. G., & Boswell, J. F. (2007). Harvard Review of Psychiatry.

89

. Zimet, G. D., Dahlem, N. W., Zimet, S. G., & Farley, G. K. (1988). Journal

of Personality Assessment.

Show less

Address for Correspondence:

Jill Ehrenreich-May, Department of Psychology, University of Miami, 5665 Ponce de

Leon, Coral Gables, FL 33146

Email:j.ehrenreich at miami.edu

© 2012 American Psychological Association

Subject:

Anxiety (major); Cognitive Behavior Therapy (major); Major Depression (major); Therapeutic

Alliance (major); Interpersonal Relationships (major); Adolescent Psychotherapy;

Diagnosis

Classification: 3311: Cognitive Therapy

Age:

Childhood (birth-12 yrs), School Age (6-12 yrs), Adolescence (13-17 yrs), Adulthood

(18 yrs&older)

Population: Human, Male, Female

Identifier (keyword):

adolescent, cognitive-behavioral, interpersonal, therapeutic alliance, transdiagnostic,

anxiety, depression

Test and measure:

Semistructured diagnostic interview, Anxiety disorders interview schedule for the

DSM-IV, child version, Anxiety disorders interview schedule for the DSM-IV, parent

version, Adolescent; Attachment Questionnaire, Vanderbilt Therapeutic Alliance Scale,

Working Alliance Inventory, Clinical Severity Rating, Child and Adolescent Social

and Adaptive Functioning Scale, Multidimensional Scaleof Perceived Social Support,

Revised Child Anxiety and Depression Scale

Methodology: Empirical Study, Quantitative Study

Author e-mail address: j.ehrenreich at miami.edu

Contact individual:

Ehrenreich-May, Jill,; Department of Psychology, University of Miami, 5665 Ponce

de Leon, Coral Gables, 33146, US,j.ehrenreich at miami.edu

Publication title: Psychotherapy

Grant/sponsorship:

National Institute of Mental Health; Grant K23MH073946; Ehrenreich-May, Jill

Volume: 49

Issue: 2

Pages: 218-230

Publication date: Jun 2012

Format covered: Electronic

Section: Cognitive-Behavioral Psychotherapy.

Publisher: Educational Publishing Foundation

Country of publication: United States

ISSN: 0033-3204

eISSN: 1939-1536

Peer reviewed: Yes

Document type:

Journal, Journal Article, Peer Reviewed Journal

Number of references: 89

Publication history :

Accepted date: 13 Mar 2012

Revised date: 12 Mar 2012

First submitted date: 06 Mar 2012

DOI:

http://dx.doi.org/10.1037/a0028265

Release date:

28 May 2012 (PsycINFO); ; 28 May 2012 (PsycARTICLES);

Accession number: 2012-13805-011

ProQuest document ID: 1017752891

Document URL:

https://login.libproxy.edmc.edu/login?url=http://search.proquest.com/docview/1017752891?accountid=34899

Copyright:

©American Psychological Association 2012

Database: PsycARTICLES

Contact ProQuest

Copyright © 2012 ProQuest LLC. All rights reserved. -

Terms and Conditions

Jessie Rayl
thedogmom63 at frontier.com
www.facebook.com/Eaglewings10
www.pathtogrowth.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.acb.org/pipermail/acb-hsp/attachments/20121119/37482890/attachment-0001.html>