<%BANNER%>

Development of a Measure of Family Accommodation for Pediatric Anxiety Disorders

Permanent Link: http://ufdc.ufl.edu/UFE0041382/00001

Material Information

Title: Development of a Measure of Family Accommodation for Pediatric Anxiety Disorders
Physical Description: 1 online resource (53 p.)
Language: english
Creator: Grabill, Kristen
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: Clinical and Health Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The present study developed a measure of family accommodation (Pediatric Accommodation Scale; PAS) for children with Generalized Anxiety Disorder (GAD), Social Phobia (SP), and Separation Anxiety Disorder (SAD). Participants were 59 caretakers and their children ages 7 to 17 with a principle diagnosis of GAD, SP, or SAD. The PAS, a clinician-administered measure, was administered at two sites along with other study measures. Results provided preliminary evidence for the psychometric properties of the PAS, including internal consistency, inter-rater reliability, and some evidence of convergent validity. Examination of the frequency of individual items on the PAS showed that up to 90% of parents endorsed at least minimal accommodation. There were significant site differences on the PAS, suggesting that results may have been influenced by the interviewer's level of training and experience with an anxiety disorders population. Based on findings in the Obsesive Compulsive Disorder (OCD) literature, a series of mediation models were tested that did not support family accommodation as a mediator in the relationship between symptom severity and functional impairment or in the relationship between parent psychopathology and symptom severity. It is possible that accommodation has a different role in these anxiety disorders compared with OCD, but is also possible that the current study was underpowered to detect mediation effects.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Kristen Grabill.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Geffken, Gary R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041382:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041382/00001

Material Information

Title: Development of a Measure of Family Accommodation for Pediatric Anxiety Disorders
Physical Description: 1 online resource (53 p.)
Language: english
Creator: Grabill, Kristen
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: Clinical and Health Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The present study developed a measure of family accommodation (Pediatric Accommodation Scale; PAS) for children with Generalized Anxiety Disorder (GAD), Social Phobia (SP), and Separation Anxiety Disorder (SAD). Participants were 59 caretakers and their children ages 7 to 17 with a principle diagnosis of GAD, SP, or SAD. The PAS, a clinician-administered measure, was administered at two sites along with other study measures. Results provided preliminary evidence for the psychometric properties of the PAS, including internal consistency, inter-rater reliability, and some evidence of convergent validity. Examination of the frequency of individual items on the PAS showed that up to 90% of parents endorsed at least minimal accommodation. There were significant site differences on the PAS, suggesting that results may have been influenced by the interviewer's level of training and experience with an anxiety disorders population. Based on findings in the Obsesive Compulsive Disorder (OCD) literature, a series of mediation models were tested that did not support family accommodation as a mediator in the relationship between symptom severity and functional impairment or in the relationship between parent psychopathology and symptom severity. It is possible that accommodation has a different role in these anxiety disorders compared with OCD, but is also possible that the current study was underpowered to detect mediation effects.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Kristen Grabill.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Geffken, Gary R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041382:00001


This item has the following downloads:


Full Text





DEVELOPMENT OF A MEASURE OF FAMILY ACCOMMODATION FOR PEDIATRIC
ANXIETY DISORDERS




















By

KRISTEN M. GRABILL


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2010

































2010 KRISTEN GRABILL
































To my Dad









ACKNOWLEDGMENTS

I would like to thank my parents and my mentors, all of whom have been
supportive throughout my graduate training.









TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ............................ ....... ....................... 4

LIST O F TA BLES ......... ................ ..................... ...... ............... 6

LIST O F FIG U R ES ....................................... .................... 7

A BSTRA CT ........... ....... ..... ..... ........ ....................... ......... 8

CHAPTER

1 INTRODUCTION .................. ...... ......... ......... 10

Family Variables related to Pediatric Anxiety................................... 10
Family Accommodation........................... ................................. 12
Family Accommodation and Pediatric Anxiety Treatment................... 15
Aims of the Current Study............................................. ............... 17

2 M E T H O D .............. ..... ............ ................. ............................................... 1 8

P a rtic ip a n ts .............. ...... ........... ............................................................... 1 9
Clinical Interview ers .................................................................... ....... 20
M e a s u re s ........................................................................................ 2 1
C linician A dm inistered .............................................................. 2 1
C hild R report .................................... ..................... .... 22
Parent Report .................................. ........................................ 23
P ro c e d u re s ..................................................................................................... 2 5

3 R E S U L T S ....................................................................................................... 2 7

Evaluation of Scale Item s ......................................................... .. ....... 27
R e lia b ility ......................................................... ....... .............................. 3 0
Convergent and Discrim inant Validity .............................................. 30
Exploratory Analysis of Family Accommodation Construct ............................... 34
Mediation Analyses ......... ..... ............................... 34

4 DISCUSSION ........................................................................ ............ .......... ........ 36

APPENDIX: PEDIATRIC ACCOMMODATION SCALE................. ........ ..... 43

LIST OF REFERENCES ................. ...................... ......... 47

BIOGRAPHICAL SKETCH ...... ........... .... ............. .................... 53









LIST OF TABLES


Table page

3-1 PAS item characteristics by site ................................. ............... 28

3-2 Pearson correlations among study variables............. ...... .................. 32









LIST OF FIGURES

Figure page

3-1 Models of conditional indirect effects .......... ............ ........... ........... 35









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

DEVELOPMENT OF A MEASURE OF FAMILY ACCOMMODATION FOR PEDIATRIC
ANXIETY DISORDERS

By

Kristen Grabill

August 2010

Chair: Gary R. Geffken
Major: Psychology

The present study developed a measure of family accommodation (Pediatric

Accommodation Scale; PAS) for children with Generalized Anxiety Disorder (GAD),

Social Phobia (SP), and Separation Anxiety Disorder (SAD). Participants were 59

caretakers and their children ages 7 to 17 with a principle diagnosis of GAD, SP, or

SAD. The PAS, a clinician-administered measure, was administered at two sites along

with other study measures. Results provided preliminary evidence for the psychometric

properties of the PAS, including internal consistency, inter-rater reliability, and some

evidence of convergent validity. Examination of the frequency of individual items on the

PAS showed that up to 90% of parents endorsed at least minimal accommodation.

There were significant site differences on the PAS, suggesting that results may have

been influenced by the interviewer's level of training and experience with an anxiety

disorders population. Based on findings in the Obsessive Compulsive Disorder (OCD)

literature, a series of mediation models were tested that did not support family

accommodation as a mediator in the relationship between symptom severity and

functional impairment or in the relationship between parent psychopathology and

symptom severity. It is possible that accommodation has a different role in these anxiety









disorders compared with its role in OCD, but is also possible that the current study was

underpowered to detect mediation effects.









CHAPTER 1
INTRODUCTION

Anxiety disorders represent the most common group of psychiatric disorders in

children, with prevalence rates of 12% to 20% (Achenbach, Howell, McConaughy, &

Stanger, 1995; Velting, Setzer, & Albano, 2004). Anxiety disorders in childhood cause

impairment across multiple domains, including academic and interpersonal functioning,

and are associated with lower levels of social supports, academic underachievement,

underemployment, substance use, and high comorbidity with other disorders (Bernstein

& Borchardt, 1991; Velting et al., 2004). Additionally, untreated pediatric anxiety

disorders are often unremitting into adulthood (Last, Phillips, & Statfield, 1987).

Although Cognitive Behavioral Therapy (CBT) has been identified as a probably

efficacious treatment for childhood anxiety, there remain children who are treatment

refractory (Cartwright-Hatton et al., 2004), and the majority of treatment studies have

failed to identify mechanisms of change in treatment or predictors of treatment

refractoriness (Cartwright-Hatton et al., 2004). Identification of mechanisms of change

and predictors of treatment outcome is important to improve treatments and to address

the needs of treatment refractory patients. Development of measures to assess these

variables and inclusion of such measures in treatment studies will be necessary to

identify ways to improve treatment.

Family Variables related to Pediatric Anxiety

Applied practice has frequently implicated parent and family variables in child

symptom expression and in treatment outcome (Gosch et al., 2006; Kingery et al.,

2006). Although a number of studies have explored the role of parent and family

variables, the range of variables investigated to date has been limited. More









specifically, parental factors associated with childhood anxiety disorders include

parenting stress (Mash & Johnston, 1990), and presence of parental psychopathology

(McClure, Brennan, Hammen, & le Broque, 2001). Parents of anxious children are also

less likely to encourage autonomy and more likely to support avoidant coping strategies

(Dadds, Barrett, Rapee, & Ryan, 1996). Family factors, such as low levels of family

adaptiveness and high levels of family cohesion (Poleg-Popko & Dar, 2001) have also

been associated with anxiety in children.

Similarly, those parental and family factors have been associated with treatment

outcome in anxious children. Children with high levels of family dysfunction, parental

frustration, and parenting stress may have poorer response to treatment (Crawford &

Manassis, 2001). This may be due to presence of family processes with the potential to

interfere with progress in treatment or to prevent families from fully engaging in

treatment. In contrast to the results of Poleg-Popko and Dar (2001), Victor, Bernat,

Bernstein, and Layne (2007) demonstrated that families high in cohesion had better

treatment outcome than those low in cohesion, following CBT with parent training. They

also reported no differences in treatment outcome related to family adaptability,

parenting stress, or parental psychopathology.

Despite a number of theoretical and empirical reports of the role of parenting in

development and maintenance of pediatric anxiety, meta-analysis of 47 investigations

revealed that only 4% of variance in child symptoms was accounted for by parenting

practices (McLeod, Wood, & Weisz, 2007). However, the vast majority of parenting

practices investigated were some form of parental rejection or parental control. Given

that studies inconsistently identify family variables associated with pediatric anxiety









disorders and outcomes of treatment, and the limited range of family variables that have

previously been studied, it is likely that research has not yet identified relevant family

variables.

Family Accommodation

One reason for the failure to identify family variables related to childhood anxiety

may be a lack of appropriate measures to assess them. Family accommodation, for

example, has been clinically reported as a barrier to pediatric anxiety treatment by virtue

of being counter to the principles of CBT (Storch et al., 2007a). Family accommodation

refers to ways in which family members accommodate patient symptoms, such as

providing reassurance, avoiding feared stimuli, taking over a patient's duties,

participation in rituals and modifying family routines. Theoretically, family

accommodations negatively reinforce the patient's symptoms, thereby affecting

symptom severity and adversely affecting treatment outcome (Leane, 1991). For

example, provision of reassurance is negatively reinforcing because it temporarily

reduces anxiety. Similarly, allowing the patient to avoid feared situations is negatively

reinforcing. Given that family accommodation has been related to symptom severity and

treatment outcome in children and adults with Obsessive-Compulsive Disorder (OCD;

Amir, Freshman, & Foa, 2000; Storch et al., 2007a; Storch et al., 2007b), it may be a

family variable that is important to investigate in other pediatric anxiety disorders.

Similar to the manner in which family accommodation is theorized to interfere with

treatment through provision of negative and positive reinforcement of OCD symptoms,

families of children with other anxiety disorders may also be reinforcing symptoms

through accommodation. Although children with other anxiety disorders do not have









rituals, parents may accommodate in many other ways (e.g., providing reassurance,

allowing avoidance of feared stimuli, taking over a patient's responsibilities).

Although there is not currently a measure of family accommodation for general

child anxiety, Calvocoressi et al. (1995; 1999) developed a 13-item measure of family

accommodation specific to symptoms of OCD, demonstrating good internal consistency

(a = .76 .82), interrater reliability (ICCs = .72 to 1.0), and construct validity through

higher correlations with measures of family functioning and family stress than with

measures of financial stress and stress associated with caring for those with physical

disabilities The FAS targets the following forms of accommodation: facilitation of

compulsions, providing reassurance, facilitating avoidance, modifying family routines,

and assuming responsibilities of the patient. These accommodating behaviors were

associated with reduced global functioning and increased symptom severity.

Following development of the FAS, researchers have been able to study family

accommodation in adults and children with OCD. Up to 90% of families of OCD adults

report at least minimal accommodation of symptoms (Allsop & Verduyn, 1990; Storch et

al., 2007b). Family accommodation has also been positively related to depression and

anxiety in family members of adult OCD patients (Amir, Freshman, & Foa, 2000). In

family members of patients with OCD, disengagement coping strategies such as family

accommodation were positively related to depression and inversely related to hope and

adaptive coping (Geffken, Storch, Duke, Monaco, and Goodman, 2006). In a study of

children with OCD, most families reported some degree of family accommodation

(Storch et al., 2007a). Family accommodation was also related to symptom severity,

functional impairment, and to comorbid internalizing and externalizing behavior









problems. Storch et al. (2007b) found that family accommodation mediates the

relationship between symptom severity and parent-rated functional impairment, further

demonstrating the potential importance of addressing family accommodation in

treatment.

In a study of adults with OCD, Amir, Freshman, and Foa (2000) found that family

accommodation and family distress were related to symptom severity post-treatment,

suggesting that family accommodation hinders progress in treatment. They argue that

family functioning, including family accommodation, should be addressed as part of

treatment. Similarly, Ferrao et al. (2006) reported that while most families of patients

with OCD entering treatment reported some level of accommodation, treatment

refractory patients demonstrated the highest levels. In that study, 52.4% of families of

refractory patients were classified as having extreme family accommodation, compared

with only 3.8% of treatment responder families. In a study of family-based CBT for

pediatric OCD, family accommodation was reduced following treatment, although this

study did not examine the mechanism of change (Storch et al., 2007a).

Despite its importance in treatment outcome of pediatric OCD, family

accommodation has not been studied in children with other pediatric anxiety disorders.

Similar to the way in which family accommodation reinforces OCD symptoms, it may

reinforce symptoms in other pediatric anxiety disorders. In Generalized Anxiety

Disorder (GAD), for example, parents may provide reassurance about many different

worries, or allow their children to avoid situations which might elicit anxiety (e.g.,

watching the news). In Social Phobia (SP), parents may allow their children to avoid

feared social situations, or rearrange family routines to accompany the child to social









situations which the child might otherwise fear. By nature of the disorder, parents of

children with Separation Anxiety Disorder (SAD) are involved in symptoms. Parents of

children with SAD likely spend extra time upon separating trying to reassure the patient,

or rearrange family schedules to allow children to avoid separation. Accommodating

behaviors such as these have the potential to interfere with treatment, as they may

undermine techniques that are central to completion of CBT.

Family Accommodation and Pediatric Anxiety Treatment

Similar to OCD, other anxiety disorders are hypothesized to be maintained

through negative reinforcement. Existence of contingency management as an integral

part of treatments for pediatric anxiety implies that many parents need to cease

providing negative reinforcement, through learning ways to reinforce appropriate

behaviors and not to reinforce inappropriate, anxiety-driven behaviors (Kendall, 1994;

Kendall, 2000; Spence et al., 2000). Studies have not investigated which behaviors

parents engage in that necessitate contingency management training. It may be that in

some cases parents engage in accommodation of symptoms that inadvertently reinforce

anxiety.

Kendall (1990) reported that when family members are not included in treatment

of pediatric anxiety disorders, one-third to one-half of patients continue to have clinically

significant anxiety problems after treatment. However, studies directly comparing

treatments with and without family involvement have yielded mixed results (Barrett et

al., 1996; Cobham et al., 1998; Barrett et al., 1998; Mendlowitz et al., 1999; Nauta et al.,

2003, Spence et al., 2000; Wood, Piacentini, Southam-Gerow, Chu, & Sigman, 2006).

Of seven studies to date, two demonstrated no difference between treatments with and

without parent involvement (Nauta et al., 2003; Spence et al., 2000), four found









increased benefits of parents involvement on only some outcome measures (Barrett et

al., 1996; Cobham et al., 1998; Barrett et al., 1998; Mendlowitz et al., 1999), and only

one demonstrated clear benefit of parent involvement on a majority of outcome

measures (Wood et al., 2006). Although lack of significant findings may indicate low

power for finding differences between active treatments, the majority of these studies do

not thoroughly describe the family intervention, and it appears that family components of

treatment may vary widely among treatment studies in general. Thus, it is likely that

some of the interventions tested have not targeted family variables that have an effect

on child anxiety.

In the only study to date demonstrating better outcome on a majority of measures

for parent involvement in treatment, Wood et al. (2006) investigated a family-based

intervention including modules designed to target "parent intrusiveness" and

discouragement of child autonomy. Parental intrusiveness has been described as

parent performance of tasks a child should be performing independently (Chorpita &

Barlow, 1998; Whaley et al., 1999; Wood et al., 2003), which is similar to some family

accommodations described by Calvocoressi et al. (1999). There was a greater decline

in anxiety severity for children assigned to the family-based intervention, and 78.9% of

those in the family-based intervention remitted diagnostic status, compared with 52.6%

in the child treatment condition. However, this study failed to investigate the

mechanism of change responsible for the difference between groups, so conclusions

about the role of these factors cannot be drawn. In summary, although it has been

shown that inclusion of families in treatment can enhance treatment outcome, few

studies have found a consistent relationship between specific family variables and









childhood symptom expression or treatment outcome. This discrepancy suggests that

there are family variables that influence treatment outcome which have not yet been

identified.

Aims of the Current Study

Currently, there is a need to identify family variables that have a role in pediatric

anxiety symptom presentation and treatment outcome. Although early research with

pediatric and adult OCD has indicated the potential importance of family

accommodation, studies have not investigated its role in other pediatric anxiety

disorders. Study of this construct is limited by lack of a measure of family

accommodation for general pediatric anxiety. The current study aims to develop and

validate such a measure, and to collect preliminary data regarding correlates of family

accommodation in pediatric anxiety. This study also aims to test for a mediation effect

of family accommodation in the relationship between symptom severity and functional

impairment, previously reported by Storch et al. (2007b) in a pediatric OCD sample. As

well, we will test for family accommodation as a possible mediator of the relationship

between parent psychopathology and child symptom severity.









CHAPTER 2
METHOD

Development of the PAS occurred in three phases.

Phase 1: Initial scale development. Items were initially generated from the FAS

(Calvocoressi et al., 1999; see Table 1), review of the extant literature, and clinical

experience of a writing committee. In addition to the first author, the writing committee

was comprised of 5 members representing psychologists and psychiatrists with

expertise and interest in pediatric anxiety disorders. Members were Dr. Eric Storch,

Ph.D. (University of South Florida), Gary Geffken, Ph.D. (University of Florida), Tanya

Murphy, M.D. (University of South Florida), Ayesha Lall, M.D. (University of Florida),

and Golda Ginsburg, Ph.D. (Johns Hopkins University).

Phase 2: Pilot testing. We initially examined the measure in a sample of 5

children and their caregivers. Based on feedback from these families, we added

several items to the PAS to reflect responses of children when accommodation is not

provided (i.e. added sad/down and angry/abusive) and altered the response choices to

provide more concrete frequency options (i.e. changed response range from "never" to

"always" to a new response range of "never" to "11 + times per day"). After making

these revisions, we piloted the PAS in a second group of 5 patients and their caregivers.

Those families did not have any suggestions for substantive changes to the measure.

Phase 3: Scale validation. The final version of the PAS was complete following

phases 1 and 2. The PAS, including all items, is included in the appendix. The PAS was

designed to be a clinician-rated scale incorporating information from the patient,

caregivers, and clinician judgment. It may be administered by a psychologist,

psychiatrist, or a trained clinical interviewer. The final PAS contains 16 items, each of









which includes a specific question followed by a series of examples that can be altered

by the administering clinician. For example, to assess whether parents facilitate

avoidance, the PAS asks "How often in the last week did you help your child avoid

things or situations that might make him/her more anxious, such as talking to others,

separating from you, or watching the news?" Clinicians are able to substitute known

symptoms for the provided examples, tailoring the administration for each family. In

addition, the semi-structured format permits follow-up questions as necessary to clarify

a response or to probe more carefully around a family routine. Responses are then

provided on a 0 (never) to 8 (11+ times per day) scale. Participants, measures, and

procedures for validation of the PAS are described below.

Participants

Participants were 59 children (n = 27 male, n = 32 female) and their caretakers.

Children ranged in age from 7 to 17 (M = 12.59, SD = 2.97). Children were primarily

Caucasian (91.4%; 3.4% Hispanic, 5.2% Other) and average household income was

$77,633.20 (SD = $43,618.64, Median = $69,000). All children had a principal

diagnosis of Generalized Anxiety Disorder (GAD; n = 30), Social Phobia (SP; n = 19), or

Separation Anxiety Disorder (SAD; n = 11) according to DSM-IV-TR criteria (American

Psychiatric Association, 2000). These disorders were chosen based on evidence that

anxiety symptoms are similar in all three disorders (Ginsburg, Riddle, & Davies, 2006)

and that these disorders are very frequently combined in treatment studies (Cartwright-

Hatton et al., 2004). There were no differences on any study variables based on

diagnosis (ps > .05).

Inclusion criteria were: (a) principal diagnosis of GAD, SAD, or SP, (b) child age

between 6 and 17 years, and (c) parent age above 21 years. Exclusion criteria were (a)









positive diagnosis of psychosis or autism, (b) principal diagnosis other than GAD, SAD,

or SP and (c) positive diagnosis in the caregiver of mental retardation, psychosis, or

other psychiatric disorders or conditions that would limit their ability to complete

measures. In order to maximize generalizability, children with comorbid diagnoses

(except psychosis or autism) were included. Comorbid diagnoses, when present,

included presence of another anxiety disorder (n = 26), Attention Deficit Hyperactivity

Disorder (ADHD; n = 10), Major Depressive Disorder (n = 8), OCD (n = 2), Oppositional

Defiant Disorder (ODD; n = 1), and Tourette's Disorder (n = 1).

All children and their caretakers were recruited through one of two outpatient

anxiety disorder specialty clinics: one clinic affiliated with the University of Florida (UF)

and one clinic affiliated with the University of South Florida (USF). Examination of all

study variables by site revealed no significant differences in demographic information,

child-report, or parent-report data (ps > .05). However, significant differences emerged

on the clinician-administered measures: the PAS, t (56) = -2.04, p < .05, and the

Pediatric Anxiety Rating Scale (PARS; see Measures section for a full description), t

(55) = -2.90, p < .05). It is possible that differences were due to the level of experience

of clinical interviewers at the two sites.

Clinical Interviewers

Interviews were conducted by trained clinical interviewers. At the UF site, clinical

interviewers included undergraduate level research assistants with prior training in child

psychopathology and additional experience with OCD patient populations. At the USF

site, clinical interviewers were post-undergraduate level research assistants with

experience with OCD patient populations; though these research assistants had more

assessment experience with this population. Although there were differences between









sites in the experience of clinicians with the population, use of both sites was necessary

for feasibility and recruitment of an adequate sample size. To ensure fidelity across sites,

all clinical interviewers were trained to a reliable standard on clinician rated measures

through a training workshop, joint interviews, and discussion. More specifically, to

complete training, clinical interviewers completed practice administrations using

standardized vignettes, observed two administrations of each measure by the first author

or another trained clinical interviewer, and completed two practice interviews under

supervision. In addition, clinical interviewers attended weekly assessment meetings to

promote quality assurance of study assessments.

Measures

Clinician Administered

Pediatric Anxiety Rating Scale. The Pediatric Anxiety Rating Scale (PARS;

Research Units on Pediatric Pharmacology Anxiety Study Group, 2002) was used to

rate the severity of anxiety symptoms, and contains separate symptom checklist and

severity rating scales. The symptom checklist contains 50 items assessing presence or

absence of anxiety symptoms related to social interactions, separation, generalized

anxiety, specific phobias, and physical symptoms. The severity rating scales contain

seven items rating severity of anxiety feelings, physical symptoms, avoidance,

interference, overall severity, and overall frequency. The PARS has demonstrated

inter-rater reliability (ICC = .97), convergent (with the HAM-A and SCARED-P) and

divergent validity (with CBCL-Externalizing and CDRS), and sensitivity to treatment

effects (Research Units on Pediatric Pharmacology Anxiety Study Group, 2002). In this

sample, reliability was adequate at both sites (Cronbach's alpha = .79 at UF and .74 at

USF) and inter-rater reliability was good (ICCs ranged from .89 to 1.00 at both sites).









Pediatric Accommodation Scale. The PAS, described in the present study, was

administered to examine family accommodation of child anxiety symptoms. The PAS

contains 16 items and the psychometric properties are detailed in the present

investigation.

Child Report

Child Anxiety Impact Scale-Child Version. The Child Anxiety Impact Scale-

Child version (CAIS-C; Langley, Bergman, McCracken, & Piacentini, 2004) was used to

examine functional impairment due to anxiety symptoms. The CAIS-C is a 34-item

scale measuring the amount of impairment across domains: school, social activities,

home/family activities, and globally. Responses are given on a four point scale from

"not at all" to "very much". The CAIS-C has demonstrated good psychometric

properties, including internal consistency (Cronbach's alpha = 0.73 to 0.87), convergent

validity (with CBCL-lnternalizing and Child Depression Inventory), and divergent validity

(CBCL-Externalizing; Langley et al., 2004). Reliability was excellent in this sample

(Cronbach's alpha = .92).

Children's Depression Inventory. The Children's Depression Inventory (CDI;

Kovacs, 1992) is a 27-item child-report measure that assesses the presence and

severity of cognitive, affective, or behavioral symptoms of depression during the

previous two weeks. Widely used, the CDI has adequate internal consistency (r = .71-

.89) and differentiates between depressed and non-depressed youth (Carlson &

Cantwell, 1979). In addition, the CDI has good test-retest reliability (r= .87) and

construct validity as determined by high correlations with similar depression measures

and through factor analysis (Kovacs, 1992; Craighead et al., 1998). Reliability was

excellent in this sample (Cronbach's alpha = .91).









Multidimensional Anxiety Scale for Children. The Multidimensional Anxiety

Scale for Children (MASC; March et al., 1997) is a 39-item child-report index that

assesses symptoms of general, social, and separation anxiety. Good psychometrics

have been reported, including construct validity through confirmatory factor analysis,

internal consistency (Cronbach's alpha = .87), convergent validity (with the STAI-C and

RCMAS), and divergent validity (with the BDI, CDRS, and HAM-D; Ryn et al., 2006;

March et al., 1997, 1999). Reliability was good in this sample (Cronbach's alpha = .89).

Parent Report

Beck Depression Inventory-Second Edition. Based on the original Beck

Depression Inventory (Beck et al., 1961), the Beck Depression Inventory-Second

Edition (BDI-II; Beck et al., 1996) is a 21-item self-report measure of depressive

symptoms experienced during the past week. Extensive reliability and validity data have

been reported in clinical and non-clinical samples, including evidence of construct

validity using confirmatory factor analysis, high internal consistency (Cronbach's alpha =

.90), convergent validity with other measures of depression, and divergent validity with

measures of anxiety (Beck et al., 1996; Whisman et al., 2000; Storch et al., 2004).

Reliability was good in this sample (Cronbach's alpha = .87).

Child Anxiety Impact Scale-Parent Version. The Child Anxiety Impact Scale-

Parent version (CAIS-P; Langley, Bergman, McCracken, & Piacentini, 2004) was used

to examine functional impairment due to anxiety symptoms. The CAIS-P is a 34-item

scale measuring the amount of impairment across domains: school, social activities,

home/family activities, and globally. Responses are given on a four point scale from

"not at all" to "very much". The CAIS-P has demonstrated good psychometric

properties, including internal consistency (Cronbach's alpha = 0.73 to 0.87), convergent









validity (with CBCL-lnternalizing and Child Depression Inventory), and divergent validity

(CBCL-Externalizing; Langley et al., 2004). Reliability was good in this sample

(Cronbach's alpha = .89).

Child Behavior Checklist. The Child Behavior Checklist (CBCL; Achenbach,

1991) is a parent-rated questionnaire assessing the frequency and intensity of

behavioral and emotional problems exhibited by the child in the past six months. The

CBCL consists of eight syndrome scales (withdrawn, somatic complaints,

anxious/depressed, social problems, thought problems, attention problems, delinquent

behavior, and aggressive behavior) and two composite scales (externalizing and

internalizing problems). Overall, the CBCL has excellent psychometric properties

including one-week test-retest reliability, adequate internal consistency (Cronbach's

alpha = .62-.92), and construct validity (e.g., strong associations with subscales of other

measures that assess similar constructs; Achenbach, 1991). Reliability was good in this

sample (Cronbach's alpha = .93 externalizing and .95 internalizing).

Family Adaptability and Cohesion Evaluation Scale fourth edition (FACES-

IV). This questionnaire is designed to assess family cohesion and family flexibility

(Olsen, Gorall, and Tiesel, 2007). Within cohesion, the FACES assesses both overall

Cohesion and unbalanced cohesion in the form of Enmeshment and Disengagement.

Within flexibility, the FACES assesses both overall Flexibility and unbalanced flexibility

in the form of Chaos and Rigidity. The FACES consists of 62 statements about family

characteristics, asking family members to indicate agreement on a 5-point scale from 1

(strongly disagree) to 5 (strongly agree). The FACES-IV has demonstrated good

psychometric properties, including reliability (Cronbach's alpha ranging from .77 to .87)









and convergent validity (with Self-Report Family Inventory, Family Satisfaction Scale,

and Family Assessment Device). Reliability for the subscales was good in this sample,

with Cronbach's alpha ranging from .78 (rigid) to .94 (cohesion).

State-Trait Anxiety Inventory Trait version (STAI-T). The State-Trait Anxiety

Inventory Trait version (STAI-T; Spielberger, 1983) is a 20-item scale rated on a 4-

point Likert scale that measures the tendency of people to experience general anxiety

and view stressful situations as threatening. The widely-used STAI-T has been found to

have good internal consistency (Cronbach's alpha = .89 to .91), convergent validity (with

the Penn State Worry Questionnaire and Worry Domains Questionnaire) and divergent

validity (with Lykken's Activity Questionnaire), and can discriminate between patients

with anxiety and non-clinical controls (Bieling et al., 1998; Spielberger, 1983, 1989).

Reliability was excellent in the present sample (Cronbach's alpha = .92).

Procedures

The University of Florida and University of South Florida institutional review

boards provided appropriate human subjects ethical approval of this project. A waiver of

informed consent was granted so that the first author or another study representative

could identify qualifying participants by diagnosis. Diagnoses were made by the

patient's clinician (a licensed child clinical psychologist or board-certified child

psychiatrist), through clinical interview, behavioral observation, parent-report, child-

report, and all available information. The first author verified diagnosis through review

of medical records and through information collected during clinician interview.

Additionally, the first author discussed evidence for diagnoses of participants with the

diagnosing clinician. The first author and clinician shared 100% diagnostic agreement

about all participants. Following verification of diagnosis, clinicians of qualifying









patients were then notified of patient eligibility. The clinician subsequently informed a

study representative when parents and patients expressed interest in participating.

Thereafter, the study representative fully described the study and obtained informed

parental consent and child assent. Participants then completed study measures in a

private setting. First, the clinical interviewer administered the PARS and the PAS,

following which children and their caregivers completed self-report measures. Children

completed the MASC, CDI, and CAIS-C while parents completed the CBCL, CAIS-P,

FACES-ll1, BDI-II, and STAI-T. Although some very young children were not able to

complete self-report measures alone, inclusion of young children in the current study is

important to determine the role of family accommodation for children of all ages. This is

particularly critical given a lack of any treatment research in very young children

(Verdeli, Mufson, Lee, & Keith, 2006). Therefore, the clinical interviewer verbally

administered child-report measures to children who were unable to complete them

alone (n = 3). Completion of all measures took between 60 and 90 minutes.









CHAPTER 3
RESULTS

Evaluation of Scale Items

Means, standard deviations, and item-total correlations for each of the 16 items on

the PAS are presented in Table 1. Examination of skewness and kurtosis values (see

Table 1) reveal that at the UF site, 7 items were positively kurtotic and 15 items were

positively skewed such that items indicating less accommodation were most frequently

endorsed. At the USF site, 8 items were positively kurtotic and 11 items were positively

skewed. Percentage of parents endorsing the item (i.e. indicating that it occurred at

least once in the preceding week) is also reported in Table 1. The most frequently

endorsed items include item 1/providing reassurance (UF = 70.1%, USF = 90.5%), item

2/allowing children to avoid anxiety provoking stimuli (UF = 63.2%, USF = 71.4%), and

item 12/parents becoming distressed when providing accommodation (UF = 55.3%,

USF = 75.2%). The least frequently endorsed items include item 5/parents changing

their work schedule (UF = 36.8%, USF = 19.0%), item 9/changing child social activities

(UF = 13.2%, USF = 38.1%), and item 11/changing the family schedule (UF = 34.2%,

USF = 23.8%). Examination of item-total correlations revealed scores ranging from .79

to .84 at the UF site, and from .09 to .67 at the USF site. Low item-total correlations (<

.3) may be accounted for by some items that were susceptible to influence of severity

rather than frequency. For example, allowing a child to sleep with parents nightly may

earn a rating of "8" (11+ times per day) because it is a level of accommodation that the

rater deemed not adequately captured by a rating of "4" (1 time per day). Note that this

was not a standardized instruction, and thus, ratings that might be influenced by

severity were made based on individual judgment and not provided uniformly.









Table 3-1. PAS item characteristics by site


Item

1. Reassurance

2. Avoidance

3. Parent social change

4. Parent routine change

5. Parent work change

6. Parent leisure change

7. Child responsibility
change
8. Child school change

9. Child social change

10. Child sleep change

11. Family schedule change

12. Parent distress

13. Child distress/anxiety

14. Child oppositional/defiant

15. Child angry/abusive


M (SD) Item-total (r)


2.41(2.11)
4.48(2.58)
1.76(1.88)
2.33(2.24)
0.24(0.60)
0.81(0.93)
1.54(1.88)
1.05(1.83)
0.81(1.35)
0.81(2.04)
0.81(2.00)
0.91(1.58)
1.27(1.66)
1.71(2.45)
0.43(0.83)
2.10(2.84)
0.16(0.44)
0.95(1.91)
1.41(2.01)
1.71(2.24)
0.73(1.12)
0.57(1.16)
1.43(2.08)
1.86(1.59)
1.54(1.77)
1.86(1.80)
1.27(1.68)
1.33(2.29)
0.89(1.51)
1.23(1.82)


.80
.43
.80
.48
.82
.61
.80
.67
.81
-.37
.80
.45
.81
.55
.81
.58
.81
.20
.84
.09
.80
.40
.79
.49
.80
.43
.79
.44
.79
.63


Skewness

0.28
-0.22
0.80*
0.83
2.41*
0.83
1.07*
1.66*
1.42*
2.78*
2.98*
1.66*
1.31*
1.20*
2.29*
1.31*
2.92*
2.93*
1.23*
1.25*
1.24*
2.01*
1.64*
0.67
0.94*
0.52
1.05*
1.76*
1.78*
1.12*


Kurtosis % Endorsing


-1.21
-.088
-0.43
0.21
4.65*
-0.29
0.39
1.60*
0.65
7.75*
8.51*
1.69*
0.87
0.36*
5.58*
0.50
8.59*
9.52*
1.09
1.32
-0.05
3.20*
2.03*
-0.33
-0.19
-1.15
-0.27
2.49*
2.59*
-0.24


71.1
90.5
63.2
71.4
15.8
52.4
50.0
33.3
36.8
19.0
23.7
33.3
50.0
42.9
31.6
52.4
13.2
38.1
42.1
47.6
34.2
23.8
55.3
76.2
56.8
66.7
50.0
33.3
39.5
42.9


Inter-rater
Reliability (ICC)a
.99
1.00
1.00
1.00
.78
1.00
.99
1.00
1.00
1.00
1.00
.33
.94
.36
.88
1.00
.93
.79
.99
1.00
1.00
1.00
.84
1.00
1.00
1.00
.95
1.00
.93
1.00









Table 3-1. Continued
Item M (SD) Item-total (r) Skewness Kurtosis % Endorsing Inter-rater
Reliability (ICC)a
16. Child sad/down 1.03(1.42) .80 1.76* 3.34* 47.4 .80
1.43(1.57) .36 1.27* 0.96 66.7 1.00
an = 21(UF), n = 7 (USF), *p <.05. Data are presented by site, with line 1 = UF (n = 38), line 2 = USF (n = 21). Responses to the PAS are provided
on a 0 (never) to 8 (11+ times per day) scale based on the frequency the behavior occurred in the preceding week.









Total scores on the PAS ranged from 0 to 49 (M = 17.73, SD = 13.22) at UF and

from 6 to 54 at USF (M = 25.19, SD = 15.37), with Shapiro-Wilk tests of normality

indicating significantly non-normal distribution of scores at both sites (ps < .05).

Examination of skewness and kurtosis values for the PAS indicated significant positive

skewness (Skewness = .749, SD = .389) and absence of kurtosis (Kurtosis = -.333, SD

= .778). This finding is similar to that of Calvocoressi et al. (1999), who reported

positively skewed total scores on the FAS for OCD.

Reliability

Reliability statistics were calculated separately for each site. Cronbach's alpha

was good at both sites (UF = .81 and USF = .78). Inter-rater reliability was assessed by

two live, independent raters who were blind to the ratings of the other. Inter-rater

reliability was available for a portion of the sample (n = 27 of 59), based on the

availability of raters during measure administration. Feasibility did not permit two raters

to be available for all administrations of the PAS. Intraclass correlation coefficients

were calculated for each item (see Table 1) and ranged from .78 to 1.0 at the UF site (n

= 20) and from .33 to 1.0 at the USF site (n = 7).

Convergent and Discriminant Validity

Given that the distribution of scores on the PAS was significantly positively

skewed, square root transformation was applied to total PAS scores. Square root

transformation is an approach that normalizes positively skewed data and is a

conservative transformation relative to other approaches (e.g., logarithmic

transformation, inverse transformation). Following square root transformation, Shapiro

Wilk tests of normality indicated that the assumption of an underlying normal distribution

could be maintained (p > .05). As well, examination of individual skewness scores









indicated that square root transformed PAS scores were not significantly skewed

(Skewness = -.23, SD = .31). In subsequent analyses, where PAS scores are reported

they reflect the square root transformation unless otherwise noted.

PAS total scores were examined by site for evidence of convergent and

discriminant validity using Pearson's correlations (see Table 2). To correct for the risk of

making a Type I error through multiple analyses, significance was set at p < .01. Given

the small sample size in the current study, particularly when examining site differences,

Bonferroni correction was not used to avoid the risk of Type II error. UF site. As

expected, the PAS correlated significantly with parent-rated impairment (CAIS-P; r=

.46, p < .01). It did not correlate significantly with child-rated symptoms of depression

(CDI; r = -.05, p > .01). Fisher's r to z comparison test shows that the PAS is more

strongly correlated with the CAIS-P than with the CDI, providing some evidence for

convergent and divergent validity (z = 2.22, p < .05). Contrary to expectations, the PAS

did not correlate significantly with other similar constructs such as child-rated anxiety

symptoms (MASC; r= .20, p > .01), overall anxiety symptoms (PARS; r= .25, p > .01),

child rated impairment (CAIS-C; r= .17, p > .01), or internalizing symptoms (CBCL-

Internalizing; r= .15, p > .01). USF site. As expected, the PAS correlated significantly

with overall anxiety symptoms (PARS; r= .59, p < .01), child-rated anxiety symptoms

(MASC; .58, p < .01), and internalizing symptoms (CBCL-Internalizing; r= .58, p < .01).

Contrary to expectations, the PAS did not correlate significantly with either child-rated or

parent-rated impairment (CAIS-C; r = .40, p > .01; CAIS-P; r = .44, p > .01). As well, the

PAS correlated significantly with child-rated depression, a theoretically distinct construct

(CDI; r= .68, p< .01).










Table 3-2. Pearson correlations among study variables
PASa PARS CAISP CAISC CDI MASC CBCL CBCL- BDI-II STAI- Cohes Flex Disen Enme Rigid Chao
-Int Ext T
PAS -- .25 .46* .17 -.05 .20 .15 .26 .37 .23 -.12 -.02 .00 .36 .28 .15
.59* .44 .40 .68** .58* .58* .52 .57* .67** -.53 -.56* .13 .49 .15 .48
PARS -- .37 .53** .39 .56** .29 .10 -.03 .24 .05 .09 -.15 -.01 .04 -.08
.28 .35 .49 .50 .33 .32 .31 .49 -.41 -.25 .23 .34 .04 .44
CAISP -- .38 .23 .40 .43* .26 .17 .26 -.18 -.23 .10 .39 .18 .09
.48 .38 .41 .44 .59* .31 .49 -.41 -.40 .16 .32 .27 .41
CAISC -- .47* .69** .53** .03 -.11 .12 -.11 -.03 .00 .25 -.03 .18
.71** .49 .64* .39 .26 .36 -.20 -.43 .16 .32 .17 .40
CDI -- .62** .50** .08 -.20 .01 -.25 -.03 .03 -.08 -.23 .09
.66* 81** .56 .35 .57 -.63* -.68** .28 .18 .05 .61*
MASC -- .43 -.08 -.10 .36 -.22 -.13 .10 .17 -.11 .27
.40 .31 .49 .56* -.60* -.76** .41 .35 -.01 .72**
CBCL- -- .46* .01 .09 -.27 .09 .15 .42 .03 .21
Int .52 .54 .68* -.38 -.43 .08 .24 .15 .52
CBCL- -- .16 .17 -.15 .04 .12 .11 .19 .12
Ext .29 .43 -.48 -.34 .07 .47 .14 .50
BDI-II -- .71** -.40 -.49* .51** .36 -.02 .44*
.87** -.43 -.37 .19 .46 -.22 .61**
STAI-T -- -.36 -.55** .47* .31 .01 .45*
-.50 -.45 .20 .52 -.16 .69**
Cohes -- .64** -.70** -.52** .20 -.59**
.76** -.65** -.54 .11 -.51
Flex -- -.54** -.33 .27 -.44*
-.45 -.38 .03 -.64**
Disen -- .41* -.16 .67**
.48 .01 .27
Enme -- .18 .39
.31 .50
Rigid -- -.26
-.03
Chao

*p<.01, **p<.003 (.05/16 = .003), apAS scores were square-root transformed. Data are presented by site, with line 1 = UF (n = 38), line 2 = USF (n
= 21). PAS = Pediatric Accommodation Scale, PARS = Pediatric Anxiety Rating Scale, CAISP = Child Anxiety Impact Scale (Parent-rated),










CAISC = Child Anxiety Impact Scale (Child-rated), CDI = Child Depression Inventory, MASC = Multidimensional Anxiety Scale for Children, CBCL-
Int = Child Behavior Checklist-Internalizing scale, CBCL-Ext = Child Behavior Checklist-Externalizing scale, BDI-II = Beck Depression Inventory
Second Edition, STAI-T = State-Trait Anxiety Inventory-Trait scale, Cohes = Family Adaptablilty and Cohesion Scale-Cohesion subscale, Flex =
Family Adaptablilty and Cohesion Scale-Flexibility subscale, Disen = Family Adaptablilty and Cohesion Scale-Disengaged subscale, Enme =
Family Adaptablilty and Cohesion Scale-Enmeshed subscale, Rigid = Family Adaptablilty and Cohesion Scale-Rigid subscale, Chao = Family
Adaptablilty and Cohesion Scale-Chaotic subscale.









The PAS was not significantly correlated with externalizing symptoms (CBCL-

Externalizing, r = .52, p > .01), although the size of the correlation was large. Fisher's r

to z comparison tests do not show that theoretically similar constructs show a

significantly stronger relationship to the PAS compared with those that are theoretically

dissimilar (ps > .05).

Exploratory Analysis of Family Accommodation Construct

Pearson's correlations were conducted to explore relationships of the PAS with

constructs of interest (see Table 2). UF Site. Contrary to expectations, data from the

UF site revealed that the PAS was not significantly related to any measures of parent

psychopathology or family functioning (see Table 2). USF Site. As expected, the PAS

was significantly positively related to parent symptoms of depression (BDI-II; .57, p <

.01) and anxiety (STAI-T; r= .66, p < .01). The PAS was negatively related to family

flexibility (FACES-Flexibility; r = -.56). Inconsistent with previous findings, the PAS was

not significantly related to family cohesion (FACES-Cohesion; r= -.53), although the

size of the correlation was large and approached significance. Again, different findings

between the two sites are likely due to the difference in skill of clinical interviewers

administering both the PAS and the PARS.

Mediation Analyses

This study aimed to test a mediation model in which family accommodation

mediates the relationship between anxiety symptom severity and functional impairment,

as reported by Storch et al. (2007b). As well, we proposed a second mediation model

in which family accommodation mediates the relationship between parent

psychopathology and anxiety symptom severity. However, given the site differences in

the PAS data, we tested a series of models in which site moderates the mediation effect









(see Figure 1). All models included site as a moderator and used bootstrapping (n =

5000), as other techniques (i.e. Sobel's test) may result in unstable estimates when

sample sizes are small. Models were analyzed in SPSS using the macro provided by

Preacher, Rucker, and Hayes (2007). Model 1. This model tested PAS as a mediator

of the relationship between symptom severity (PARS) and parent-rated functional

impairment (CAIS-P). Results indicated nonsignificant indirect effects at both UF (B =

.04, z = 1.31, p > .05) and USF (B = .11, z = 1.06, p > .05). Model 2. This model tested

PAS as a mediator of the relationship between symptom severity (PARS) and child-

rated functional impairment (CAIS-C). Results indicated nonsignificant indirect effects at

both UF (B = .01, z = .26, p > .05) and at USF (B = .15, z = .94, p > .05). Model 3. This

model tested PAS as a mediator of the relationship between parent depressive

symptoms (BDI-II) and symptom severity (PARS). Results indicated nonsignificant

indirect effects at both UF (B = .09, z = 1.07, p > .05) and USF (B = .13, z = .92, p >

.05). Model 4. This model tested PAS as a mediator of the relationship between parent

anxiety symptoms (STAI-T) and symptom severity (PARS). Results indicated

nonsignificant indirect effects at both UF (B = .03, z = .87, p > .05) and at USF (B = .12,

z = 1.06, p> .05).

Model 1

Site

-1.91 .20


PARSAS CAIS-P
.08 .63

Model 2


-2.35









CHAPTER 4
DISCUSSION

The current study investigated the reliability and validity of the PAS in a sample

of children with GAD, SAD, and SP and is the first investigation of the family

accommodation construct in this population. Data were collected from two anxiety-

disorder specialty clinics, and results showed a significant difference in PAS scores

based on site. Notably, the two sites showed significant differences on both clinician-

rated measures. These results may have been influenced by the interviewer's level of

training and experience with an anxiety disorders population. Although we made every

effort to standardize clinical interviewer training, those at the USF site had prior

specialty assessment experience with an OCD population that may have benefited

administration of the PAS. It may be that capturing family accommodation requires

careful probing around symptoms and family activities. For example, a parent who

initially denies changing his/her work schedule but previously reported changing the

child's school schedule might be asked about the details of the family routine when the

child is not in school (e.g., "Earlier you told me that you had to take Mary to school late

on Tuesday. Did you get to work on time that day?"). Therefore, results of the present

investigation may indicate the importance of clinician experience with an anxiety

population when administering this measure. However, we did not measure clinical

interviewer experience and therefore cannot be certain that this is the cause of the site

differences.

Overall, the PAS showed good internal consistency and inter-rater reliability

across two sites. Inter-rater reliability was low on two items at the USF site, but this was

likely due to the small number of participants (n = 7) for whom two raters were available









at that site. As well, verbal indication that some clinical interviewers incorporated

severity into ratings, while some did not, may have contributed to low inter-rater

reliability. Evidence for convergent validity was mixed, with the PAS showing a

significant, positive relationship with other measures of anxiety symptoms, anxiety

related impairment, and internalizing symptoms. Results at the UF site indicated that

the PAS did not correlate significantly with a measure of child-rated depression, which

supports divergent validity, but results from the USF site show a strong relationship with

this measure. Given that there were no significant differences in depressive symptoms

or comorbidity between sites, the difference may be related to the relationship between

accommodation and depression at each site or possibly to differential ratings of

interviewers at each site.

Examination of individual items on the PAS revealed that item-total correlations

ranged from .79 to .84 at the UF site and from .09 to .67 at the USF site. Items with low

item total correlations (< .3) at USF were item 9 (changing the child's social activities)

and item 10 (changing the child's sleep routine). One discrepancy among items that

may explain this is the susceptibility of some items to severity of accommodation. For

example, for a child that sleeps with parents nightly, some interviewers provided a rating

of "11 + times per day" because a rating of "1 time per day" does not adequately capture

the seriousness of the accommodation. In this case, although clinical interviewers were

not trained to incorporate severity in ratings of accommodation, severity of the behavior

interfered with the frequency rating. If some items partially reflect severity, which may

be a separate construct, those items may not be as closely related to the scale.

Additionally, it is possible that this interference occurred to a greater degree at the USF









site if clinical interviewers, who may have probed more closely than those at the UF site,

relied more heavily on clinical judgment rather than parent report. However, given that

we did not measure clinician judgment, we cannot draw this conclusion with certainty

and it is possible that future analysis of the PAS will identify these items as those that

should be removed from the scale. In future studies of the PAS, addition of a severity

scale may benefit this measure so that we may separately assess the severity

construct.

Further examination of individual items shows that accommodation of anxiety

symptoms in the previous week was endorsed by a high percentage of parents. Most

commonly, parents reported providing reassurance to their children (UF = 71.1%, USF

= 90.5%) and helping their children avoid anxiety-provoking stimuli (UF = 63.2%, USF =

71.4%). Additionally, a large number of parents reported feeling distressed when

accommodating their child's symptoms (UF = 55.3%, USF = 76.2%), which suggests

that many parents accommodate reluctantly. Over half of parents reported their child's

anxiety increasing when accommodation was not provided (UF = 56.8%, USF = 66.7%).

Some parents also endorsed oppositional/defiant behavior (UF = 50.0%, USF = 33.3%),

angry/abusive behavior (UF = 39.5%, USF = 42.9%), and sad/down behavior (UF =

47.4%, USF = 66.7%) when accommodation was not provided. Endorsement of these

items demonstrates that child emotional reaction or behavior is one factor that could

maintain parent accommodation of pediatric anxiety symptoms. Least commonly

endorsed forms of accommodation were changing parent work schedules (UF = 36.8%,

USF = 19.0%) and changing the child's social activities (UF = 13.2%, USF = 38.1%).

However, some infrequent forms of accommodation may have high potential for









negative consequences (e.g., parents missing work). Therefore, it is important to

continue assessing these behaviors in the context of accommodation. Overall, these

percentages suggest that a large number of parents accommodate anxiety symptoms in

this population of patients.

In addition to the high number of parents reporting accommodation, results

indicated that family accommodation is positively related to both parent-rated functional

impairment (UF site) and to clinician-, parent-, and child-rated measures of anxiety

symptoms (USF site). These results suggest that the role of accommodation in

pediatric anxiety may be similar to its role in an OCD population and that family

accommodation may have eventual implications for treatment. However, these results

should be interpreted as preliminary given that the present study is complicated by site

differences and does not provide information about causality in this relationship (i.e.

whether accommodation leads to increased severity/decreased functioning).

Exploratory analyses of the PAS with measures of parent psychopathology

revealed that family accommodation was strongly related to both parent symptoms of

depression and anxiety (USF site). While previous studies have shown inconsistent

findings regarding the relationship of parent psychopathology to child anxiety

symptoms, our results suggest that parents with higher levels of symptoms engage in

more accommodation of child symptoms. As well, exploratory analyses of the PAS with

measures of family functioning revealed an association with family flexibility (USF site),

indicating that higher frequency of accommodation is associated with lower levels of

flexibility. As well, the relationship of the PAS to family cohesion suggested that

frequent accommodation is related to low levels of family cohesion, but this relationship









only approached statistical significance (USF site). This finding is contrary to previous

reports suggesting high cohesion is linked to higher symptom severity (Poleg-Popko &

Dar, 2001) but supports treatment research suggesting that high cohesion is related to

improved outcomes (Victor et al., 2007). Positive relationships with other family

constructs in our sample (i.e. Enmeshment and Chaos) were large but did not achieve

statistical significance. Future investigations of the PAS should explore relationships

with family constructs in a larger sample. However, these results suggest that families

that are less flexible and cohesive may need additional support to target

accommodation that could interfere with CBT treatment. Additionally, families in which

parents are experiencing symptoms of anxiety and depression may also need targeted

intervention to facilitate reduced accommodation.

Finally, the present investigation aimed to test several mediation models that

were moderated by site. We hypothesized that family accommodation mediates the

relationship between parent psychopathology and child anxiety symptoms. However,

results did not support mediation models based on either parent anxiety or depressive

symptoms. We also hypothesized that family accommodation mediates the relationship

between symptom severity and functional impairment, similar to the role of family

accommodation in OCD (Storch et al., 2007b). However, results did not support the

mediation model for either parent or child reported functional impairment. It may be that

family accommodation does not have the same relationship to symptom severity and

functional impairment in this anxiety population compared with the mediation effect

found for patients with OCD. One possible explanation is that patients with OCD may

have very visible symptoms (i.e. compulsions) that both contribute to impairment and









draw accommodation. However, it is also likely that the present investigation was

underpowered to detect mediation effects. It will be important to test family

accommodation as a possible mediator in these relationships in future investigations

with a larger sample.

Although this study offers promising preliminary data for the development of the

PAS, its limitations should be noted. As noted previously, the sample size limits our

ability to draw conclusions from the data. As well, the site differences related to clinical

interviewer ratings limited our ability to generalize results across settings and further

reduced sample size given the need to conduct many analyses separately. However,

given that our site differences may have been related to clinical interviewer experience,

direct examination of interviewer experience as a variable that influences the validity of

this and other clinician-administered scales may be warranted. Finally, several items on

the PAS showed low item-total correlations, perhaps due to some clinical interviewers

incorporating severity into ratings. The PAS may benefit from inclusion of a severity

scale for this reason, and a severity scale may also contribute to overall understanding

of the family accommodation construct and its relationship to other symptom and

treatment variables.

Overall, the present study is the first to investigate family accommodation in a

sample of children with GAD, SP, and SAD. Results show that a large percentage of

parents provide accommodation, particularly when providing reassurance and allowing

children to avoid anxiety-provoking stimuli. The PAS shows good preliminary

psychometric properties, including internal consistency, inter-rater reliability, and some

evidence of convergent validity. However, differences related to clinical interviewer









ratings and inability to rate severity of accommodation suggest that the PAS should be

revised and re-administered in a second sample. Beyond re-administration of the PAS

in a second sample, future studies should examine the role of family accommodation in

CBT, including its ability to predict treatment outcome and whether addressing family

accommodation in treatment is a mechanism of change. The current study may also

have implications for clinicians and researchers wishing to identify families who are

likely to accommodate prior to beginning treatment, as low family cohesion and flexibility

are related to high levels of accommodation. Studies should continue to examine family

variables related to accommodation in order to aid in identification of these families.









APPENDIX A
PEDIATRIC ACCOMMODATION SCALE

Pediatric Accommodation Scale (PAS)



Note: This measure is designed to be administered by a clinician or trained interviewer.

Responses should be chosen based on clinician judgment of responses. Each item

contains examples that may be substituted with a child's specific symptoms. After each

item, the clinician may ask follow-up questions to clarify or to probe more closely for

accommodation.



Clinician States: This questionnaire asks about different ways that you have dealt with

your child's anxiety, such as reassuring your child when he/she is anxious, rearranging

your child's schedule, and allowing your child to avoid certain things. These questions

are asking about things you have done during the past week, so please answer keeping

only the past week in mind.



In the past week, how often...

1. Did you reassure your child about his/her fears, for example, fears that something bad

would happen him/her, something bad would happen to you, or fear that others may be

thinking bad things about him/her?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day


2. Did you help your child to avoid things or situations that might make him/her more









anxious such as talking to others, separating from you, or watching the news?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

3. Have you avoided doing things such as going out to a social event or being with

people because your child might get nervous or scared?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

4. Have you modified your routine in any way to reduce your child's anxiety, for example,

spending less time with other people or getting up early to send your child to school

because of your child's symptoms?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

5. Have you modified your work schedule because of your child's symptoms, such as

going to work late because your child refuses to go to school or taking days off because

your child is frequently feeling sick?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

6. Have you modified your social leisure activities or hobbies because of your child's

symptoms, for example, staying home when he/she is distressed or spending less time

with friends because he/she is worried?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

7. Have you modified your child's responsibilities because of his/her symptoms, such as









completing chores for him/her or reducing difficulty of his/her schoolwork?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

8. Have you modified your child's school schedule because of his/her symptoms, such as

allowing him/her to miss school, go to school late, or leave school early because he/she is

worried or frequently ill?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

9. Have you changed your child's social or leisure activities as a result of his/her

symptoms, such as picking your child up early from peer gatherings or withdrawing

him/her from sports teams of other social organizations?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day
10. Have you changed your child's sleep or bedtime routine because of his/her

symptoms, such as checking under beds or in closets or allowing your child to sleep in a

room that is not his/her bedroom?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

11. Have you modified the schedule of other family members, such as siblings, because

of your child's symptoms, such as decreasing family member's activities or spending less

time with other family members?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day









12. Did you become upset, stressed, or distressed when you help your child cope with

anxiety in unhealthy ways (e.g., allowing him/her to miss school, rearranging your

schedule so your child can be with you, or constantly providing reassurance)?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

13. Has your child become more distressed or anxious when you have not provided

assistance during times of anxiety, such as providing reassurance or allowing him/her to

avoid social activities or school?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

14. Has your child become oppositional or defiant when you have not provided
assistance, such as refusing to obey or throwing a tantrum?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

15. Has your child become angry or abusive when you have not provided assistance,

such as saying hurtful things, being destructive, or physically hurting someone?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

16. Has your child become sad or down when you have not provided assistance, such as

becoming tearful or withdrawing from you or others?

0 = Never 3 = 4-6 times per week 6 = 4-6 times per day
1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day
2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day









LIST OF REFERENCES


Achenbach, T. M. (1991). Manual for the Child Behavior Checklist/4-18 and 1991
profile. Burlington, VT: University of Vermont, Department of Psychiatry.

Achenbach, T. M., Howell, C. T., McConaughy, S. H., & Stanger, C. (1995). Six-year
predictors of problems in a national sample of children and youth: I. Cross-
Informant syndromes. Journal of the American Academy of Child and
Adolescent Psychiatry, 34, 336-347.;

Achenbach, T. M., Howell, C. T., McConaughy, S. H., & Stanger, C. (1995). Six-year
predictors of problems in a national sample of children and youth: I. Cross -
informant syndromes. Journal of the American Academy of Child and
Adolescent Psychiatry, 34, 336-347.

American Psychiatric Association (2000). Diagnostic and statistic manual of mental
disorders (4th ed. text revision). Washington, DC: Author.

Amir, N., Freshman, M., & Foa, E. (2000). Family distress and involvement in relatives
of obsessive-compulsive disorder patients. Journal of Anxiety Disorders, 14,
209-217.

Barrett, P. M. (1998), Group therapy for childhood anxiety disorders. Journal of Clinical
Child Psychology, 27, 1-468.

Barrett, P. M., Dadds, M. R., Rapee, R. M. (1996). Family treatment of childhood
anxiety: a controlled trial. Journal of Consulting and Clinical Psychology, 64,
333-342.

Beck, A. T., Steer, R. A., Brown, G.K. (1996). Beck Depression Inventory- Second
Edition manual. San Antonio, TX: The Psychological Corporation.

Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., Erbaugh, J. (1961). An inventory for
measuring depression. Archives of General Psychiatry 4, 561 571.

Bernstein, D. C., & Borchardt, C. M. (1991). Anxiety disorders of childhood and
adolescence: A critical review. Journal of the American Academy of Child and
Adolescent Psychiatry, 30, 519-532.

Bieling, P. J., Antony, M. M., Swinson, R. P. (1998). The State-Trait Anxiety Inventory,
Trait version: structure and content re-examined. Behavioral Research and
Therapy, 36, 777-788.

Bolton D., Collins S., Steinberg, D. (1983). The treatment of obsessive compulsive
disorder in adolescence: A report of fifteen cases. British Journal of Psychiatry
142, 456-64.









Calvocoressi, L., Lewis, B., Harris, M., Goodman, W. K., Trufan, S.J, McDougle, C. J.,
Price, L. H. (1995). Family accommodation in obsessive compulsive disorder.
American Journal of Psychiatry, 152, 441-443.

Calvocoressi, L., Mazure, C. M., Kasl, S., Skolnick, J., Fisk, D., Vegso, S. J., et al.
(1999). Family accommodation of obsessive-compulsive symptoms: Instrument
development and assessment of family behavior. Journal of Nervous and Mental
Disease, 187, 636-642.

Cartwright-Hatton, S., Roberts, C., Chitsabesan, P., Fothergill, C., and Harrington, R.
(2004). Systematic review of the efficacy of cognitive behavior therapies for
childhood and adolescent anxiety disorders. British Journal of Clinical
Psychology, 43, 421-436.

Cartwright-Hatton, S., Roberts, C., Chitsabesan, P., Fothergill, C., and Harrington, R.
(2004). Systematic review of the efficacy of cognitive behavior therapies for
childhood and adolescent anxiety disorders. British Journal of Clinical
Psychology, 43, 421-436.

Chorpita, B. F., Barlow, D. H. (1998). The development of anxiety: the role of control in
the early environment. Psychological Bulletin, 124, 3-21.

Cobham, V. E., Dadds, M. R., Spence, S. H. (1998). The role of parental anxiety in the
treatment of childhood anxiety. Journal of Consulting and Clinical Psychology,
66, 893-905.

Craighead, W. E., Smucker, M. R., Craighead, L. W., & Ilardi, S. S. (1998). Factor
analysis of the Children's Depression Inventory in a community sample.
Psychological Assessment, 10, 156-165.

Crawford, M. & Manassis, K. (2001). Familial predictors of treatment outcome in
childhood anxiety disorders, Journal of the American Academy of Child &
Adolescent Psychiatry 40, 1182-1189.

Dadds, M. R., Barrett, P. M., Rapee, R. M., & Ryan, S. (1996). Family process and
child anxiety and aggression: an observational analysis, Journal of Abnormal
Child Psychology 24, 715-734.

Faul, F., & Erdfelder, E. (2007). GPOWER: A Priori, post-hoc, and compromise power
analyses for MS-DOS (Computer Program). Bonn, FRG: Bpnn University,
Department of Psychology.

Ferrao, Y. A., Shavitt, R. G., Bedin, N. R., de Mathis, M. E., Lopes, A. C., Fontenelle, L.
F., et al. (2006). Clinical features associated to refractory obsessive-compulsive
disorder. Journal of Affective Disorders, 94, 199-209.









Geffken, G. R., Storch, E. A., Duke, D. C., Monaco, L., Lewin, A. B., & Goodman, W. K.
(2006). Hope and coping in family members of patients with obsessive-
compulsive disorder. Journal of Anxiety Disorders, 20, 614-629.

Ginsburg, G. S., Riddle, M. A., & Davies, M. (2006). Somatic symptoms in children and
adolescents with anxiety disorders. Journal of the American Academy of Child
and Adolescent Psychiatry, 45, 1179-1187.

Gosch, E. A., Flannery-Schroeder, E., Mauro, C. F., & Compton, S. C. (2006).
Principles of cognitive-behavioral therapy for anxiety disorders in children.
Journal of Cognitive Psychotherapy: An International Quarterly, 20, 247-262.

Hudson, J. L., & Rapee, R. M. (2001), Parent-child interactions and anxiety disorders:
an observational study. Behavioral Research Therapy, 39, 1411-1427.

Kendall, P. C. (1994). Treating anxiety disorders in children: Results of a randomized
clinical trial. Journal of Consulting and Clinical Psychology, 62, 100-110.

Kendall, P. C. (2000). Coping Cat Workbook. Ardmore, PA: Workbook Publishing.

Kendall, P. C., Kane, M., Howard, B., & Siqueland, L. (1990). Cognitive-Behavioral
Treatment of Anxious Children: Treatment Manual. Ardmore, PA: Workbook
Publishing.

Kingery, J. N., Roblek, T. L., Suveg, C., Grover, R. L., Sherrill, J. T., & Bergman, R. L.
(2006). They're not just "little adults": Developmental considerations for
implementing cognitive-behavioral therapy with anxious youth. Journal of
Cognitive Psychotherapy: An International Quarterly, 20, 263-273.

Kovacs, M. (1992). The Children's Depression Inventory. Manual. Toronto, Ontario,
Canada: Multi-Health Systems, Inc.

Langley, A., Bergman, L., McCracken, J., & Piacentini, J. (2004). Impairment in
childhood anxiety disorders: Preliminary examination of the Child Anxiety Impact
Scale-Parent Version. Journal of Child and Adolescent Psychopharmacology,
14, 105-114.

Leane, M. (1991). Family therapy for children with obsessive-compulsive disorder. In
MT Fato, J Zohak (Eds), Current treatments of obsessive-compulsive disorder.
Washington, DC: American Psychiatric Press.

March, J. S. (1995). Cognitive-behavioral psychotherapy for children and adolescents
with OCD: A review and recommendations for treatment. Journal of the American
Academy of Child and Adolescent Psychiatry, 34, 7-18.

March, J. S., Sullivan, K., Parker, J. D. (1999). Test-retest reliability of the
Multidimensional Anxiety Scale for Children. Journal of Anxiety Disorders, 13,
349-358.









Mash, E. J., & Johnston, C. (1990). Determinants of parenting stress: illustrations from
families of hyperactive children and families of physically abused children,
Journal of Clinical Child Psychology 19, 313-328.

McClure, E. B., Brennan, P.A., Hammen, C. & Le Brocque, B. C. (2001). Parental
anxiety disorders, child anxiety disorders, and the perceived parent-child
relationship in an Australian high-risk sample, Journal of Abnormal Child
Psychology 29, 1-10.

McLeod, B. D., Wood, J. J., & Weisz, J. R. (2007). Examining the association between
parenting and child anxiety: A meta-analysis. Clinical Psychology Review, 27,
155-172.

Mendlowitz, S. L., Manassis, K., Bradley, S., Scapillato, D., Mietzitis, S., & Shaw, B.
(1999). Cognitive-behavioral group treatments in childhood anxiety disorders: the
role of parental involvement. Journal of the American Academy of Child and
Adolescent, Psychiatry 38, 1223-1229.

National Institute of Mental Health (1985). Rating scales and assessment instruments
for use in pediatric psychopharmacology research. Psychopharmacological
Bulletin, 21, 839-843.

Nauta, M. H., Scholing, A., Emmelkamp, & P. M. G., Minderaa, R. B. (2003). Cognitive-
behavioral therapy for children with anxiety disorders in a clinical setting: no
additional effect of a cognitive parent-training. Journal of the American Academy
of Child and Adolescent Psychiatry, 42, 1270-1278.

Olson, D. H., Portner, J. & Bell, R. Q. (1982). FACES II: Family Adaptability and
Cohesion Evaluation Scales. Minnesota: Family Social Science, University of
Minnesota.

Peleg-Popko, O. & Dar, R. (2001). Marital quality, family patterns, and children's fears
and social anxiety, Contemporary Family Therapy 23, 465-487.

Place, M., Hulsmeier, J., Brownrigg, A., & Soulsby, A. (2005). The Family Adaptability
and Cohesion Evaluation Scale (FACES): An instrument worthy of rehabilitation?
Psychiatric Bulletin, 29, 215-218.

Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation
hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral
Research, 42, 185-227.

Research Units on Pediatric Pharmacology Anxiety Study Group, (2002). The Pediatric
Anxiety Rating Scale (PARS): Development and psychometric properties.
Journal of the American Academy of Child and Adolescent Psychiatry, 41, 1061
1069.









Ryn, M. A., Barber, J. P., Sarosh, K., Siqueland, L., Dembiski, M., McCarthy, K. S., et
al. (2006). The psychometric properties of the MASC in a pediatric sample.
Anxiety Disorders, 20, 139-157.

Scahill, L., Riddle, M. A., McSwiggin-Hardin, M., Ort, S. I., King, R. A., Goodman, W. K.,
Cicchetti, D., Leckman, J. F. (1997). Children's Yale-Brown Obsessive
Compulsive Scale: Reliability and validity. Journal of the American Academy of
Child and Adolescent Psychiatry, 36, 844 852

Shafran, R., Ralph, J., & Tallis, F. (1995). Obsessive-compulsive symptoms and the
family. Bulletin of the Menninger Clinic, 59, 472-479.

Spence, S., Donovan, C., Brechman-Toussaint, M. (2000). The treatment of childhood
social phobia: The effectiveness of a social skills training-based, cognitive-
behavioral intervention, with and without parental involvement. Journal of Child
Psychology and Psychiatry and Allied Disciplines, 41, 713-726.

Spielberger, C. D. (1983). Manual for the State-Trait Anxiety Inventory STAI (Form Y).
Palo Alto, CA: Mind Garden.

Spielberger, C. D. (1989). State-Trait Anxiety Inventory: a comprehensive bibliography.
Palo Alto, CA: Consulting Psychologists Press.

Storch, E. A., Geffken, G. R. Merlo, L., Jacob, M., Murphy, T. K., Goodman, W. K., et al.
(2007b). Family accommodation in obsessive-compulsive disorder. Journal of
Clinical Child and Adolescent Psychology, 36, 207-216.

Storch, E. A., Geffken, G. R., Merlo, L. J., Mann, G., Duke, D., Munson, M., et al.
(2007a). Family-based cognitive-behavioral therapy for pediatric obsessive-
compulsive disorder: Comparison of intensive and weekly approaches. Journal
of the American Academy of Child Psychiatry, 46, 469-478.

Velting, O. N., Setzer, N. J., & Albano, A. M. (2004). Update on and advances in
assessment and cognitive-behavioral treatment of anxiety disorders in children
and adolescents. Professional Psychology: Research and Practice, 35, 42-54.

Verdeli, H., Mufson, L., Lee, L., & Keith, J. A. (2006). Review of evidence-based
psychotherapies for pediatric mood and anxiety disorders. Current Psychiatry
Reviews, 2, 395-421.

Victor, A. M., Bernat, D. H., Bernstein, G. A., & Layne, A. E. (2007). Effects of parent
and family characteristics on treatment outcome of anxious children. Journal of
Anxiety Disorders, 21, 835-848.

Whaley, S. E., Pinto, A., Sigman, M. (1999). Characterizing interactions between
anxious mothers and their children. Journal of Consulting and Clinical
Psychology, 67, 826-836.









Whisman, M. A., Perez, J. E., and Ramel, W. (2000). Factor structure of the Beck
Depression Inventory-Second edition (BDI-II) in a student sample. Journal of
Clinical Psychology, 56, 545-551.

Wood, J. J., McLeod, B. D., Sigman, M., Hwang, W., Chu, B. C. (2003). Parenting
behavior and childhood anxiety: theory, empirical findings, and future directions.
Journal of Child Psychology and Psychiatry, 44, 134-151

Wood, J. J., Piacentini, J. C., Southam-Gerow, M., Chu, B., & Sigman, M. (2006).
Family cognitive behavioral therapy for child anxiety disorders. Journal of the
American Academy of Child Adolescent Psychiatry, 45, 314-321.









BIOGRAPHICAL SKETCH

Kristen Grabill graduated magna cum laude with a Bachelor of Arts in psychology

from Saint Mary's College of Maryland in 2003. Following that, she completed an

intramural post-baccalaureate fellowship at the National Institutes of Health in

Bethesda, MD. She completed her Predoctoral Internship at Brown University in

Providence, RI and graduated from the University of Florida in August 2010.





PAGE 1

1 DEVELOPMENT OF A MEASURE OF FAMI LY ACCOMMODATION FOR PEDIATRIC ANXIETY DISORDERS By KRISTEN M. GRABILL A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORID A IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 KRISTEN GRABILL

PAGE 3

3 To my Dad

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my parents and my mentors, all of whom have been supportive throughout my graduate training.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDG MENTS .................................................................................................. 4LIST OF TABLES ............................................................................................................ 6LIST OF FIGURES .......................................................................................................... 7ABSTRACT ..................................................................................................................... 8 CHA PTER 1 INTRODUC TION .................................................................................................... 10Family Variables related to Pediatri c Anxiety .......................................................... 10Family Accommodation ........................................................................................... 12Family Accommodation and Pedi atric Anxiety Treatment ....................................... 15Aims of the Cu rrent St udy ....................................................................................... 172 METHOD ................................................................................................................ 18Participants ............................................................................................................. 19Clinical Interviewers ................................................................................................ 20Measur es ................................................................................................................ 21Clinician Admi nistered ...................................................................................... 21Child R eport ..................................................................................................... 22Parent Re port ................................................................................................... 23Procedur es ............................................................................................................. 253 RESULTS ............................................................................................................... 27Evaluation of Scale Ite ms ....................................................................................... 27Reliabi lity ................................................................................................................ 30Convergent and Discrim inant Vali dity ..................................................................... 30Exploratory Analysis of Fam ily Accommodati on Constr uct ..................................... 34Mediation Analyses ................................................................................................. 344 DISCUSSI ON ......................................................................................................... 36APPENDIX: PEDIATRIC ACCOMMO DATION SCALE ................................................. 43LIST OF RE FERENCES ............................................................................................... 47BIOGRAPHICAL SKETCH ............................................................................................ 53

PAGE 6

6 LIST OF TABLES Table page 3-1 PAS item characte ristics by site ......................................................................... 283-2 Pearson correlations among study va riables ...................................................... 32

PAGE 7

7 LIST OF FIGURES Figure page 3-1 Models of conditional indirect e ffe cts. ................................................................. 35

PAGE 8

8 Abstract of Dissertation Pr esented to the Graduate School of the University of Florida in Partial Fulf illment of the Requirements for t he Degree of Doctor of Philosophy DEVELOPMENT OF A MEASURE OF FAMI LY ACCOMMODATION FOR PEDIATRIC ANXIETY DISORDERS By Kristen Grabill August 2010 Chair: Gary R. Geffken Major: Psychology The present study developed a measure of family accommodation (Pediatric Accommodation Scale; PAS) for children with Generalized Anxiety Disorder (GAD), Social Phobia (SP), and Separation Anxiet y Disorder (SAD). Participants were 59 caretakers and their children ages 7 to 17 with a principle diagnosis of GAD, SP, or SAD. The PAS, a clinician-administered meas ure, was administered at two sites along with other study measures. Results provided preliminary evidence for the psychometric properties of the PAS, including internal co nsistency, inter-rater reliability, and some evidence of convergent validity. Examination of the frequency of indivi dual items on the PAS showed that up to 90% of parents endor sed at least minimal accommodation. There were significant site differences on the PAS, suggesting that results may have been influenced by the interviewers level of training and experience with an anxiety disorders population. Based on findings in the Obsessive Compulsive Disorder (OCD) literature, a series of m ediation models were tested th at did not support family accommodation as a mediator in the re lationship between symptom severity and functional impairment or in the rela tionship between parent psychopathology and symptom severity. It is possible that accomm odation has a different role in these anxiety

PAGE 9

9 disorders compared with its role in OCD, but is also possible that the current study was underpowered to detect mediat ion effects.

PAGE 10

10 CHAPTER 1 INTRODUCTION Anxiety dis orders represent the most common group of psychiatric disorders in children, with prevalence rates of 12% to 20% (Achenbach, Howell, McConaughy, & Stanger, 1995; Velting, Setzer, & Albano, 2004). Anxiety disorders in childhood cause impairment across multiple domains, including academic and interpersonal functioning, and are associated with lower levels of so cial supports, academic underachievement, underemployment, substance use, and high como rbidity with other disorders (Bernstein & Borchardt, 1991; Velting et al., 2004). Additionally, untreated pediatric anxiety disorders are often unremitting into adult hood (Last, Phillips, & Statfield, 1987). Although Cognitive Behavioral Therapy (CBT) has been identified as a probably efficacious treatment for childhood anxiety, there remain children who are treatment refractory (Cartwright-Hatton et al., 2004), and the majority of treatment studies have failed to identify mechanisms of change in treatment or predictors of treatment refractoriness (Cartwright-Hatton et al., 2004). Identificat ion of mechanisms of change and predictors of treatment out come is important to impr ove treatments and to address the needs of treatment refracto ry patients. Development of measures to assess these variables and inclusion of such measures in treatment studies will be necessary to identify ways to im prove treatment. Family Variables related to Pediatric Anxiety Applied practice has frequently im plicated parent and family variables in child symptom expression and in tr eatment outcome (Gosch et al., 2006; Kingery et al., 2006). Although a number of studies have explored the role of parent and family variables, the range of variables invest igated to date has been limited. More

PAGE 11

11 specifically, parental factors associat ed with childhood anxiety disorders include parenting stress (Mash & Johns ton, 1990), and presence of parental psychopathology (McClure, Brennan, Hammen, & le Broque, 2001). Parents of anxious children are also less likely to encourage autonomy and more likely to support avoidant coping strategies (Dadds, Barrett, Rapee, & Ryan, 1996). Family fa ctors, such as low levels of family adaptiveness and high levels of family cohes ion (Poleg-Popko & Dar, 2001) have also been associated with anxiety in children. Similarly, those parental and family factors have been associated with treatment outcome in anxious children. Children with high levels of family dysfunction, parental frustration, and parent ing stress may have poorer respons e to treatment (Crawford & Manassis, 2001). This may be due to presence of family processes with the potential to interfere with progress in tr eatment or to prevent fam ilies from fully engaging in treatment. In contrast to the results of Poleg-Popko and Dar (2001), Victor, Bernat, Bernstein, and Layne (2007) dem onstrated that families high in cohesion had better treatment outcome than those low in cohesion, following CBT with parent training. They also reported no differences in treatment outcome related to family adaptability, parenting stress, or pa rental psychopathology. Despite a number of theoretical and empirical reports of the role of parenting in development and maintenance of pediatric anxiety, meta-analysis of 47 investigations revealed that only 4% of variance in child symptoms was accounted for by parenting practices (McLeod, Wood, & We isz, 2007). However, the va st majority of parenting practices investigated were some form of par ental rejection or parent al control. Given that studies inconsistently identify family variables associated with pediatric anxiety

PAGE 12

12 disorders and outcomes of treatment, and the lim ited range of family variables that have previously been studied, it is lik ely that research has not ye t identified relevant family variables. Family Accommodation One reason for the failure to identify fam ily variables related to childhood an xiety may be a lack of appropriate measures to assess them. Family accommodation, for example, has been clinically reported as a barrier to pediatric anxiety treatment by virtue of being counter to the princi ples of CBT (Storch et al., 2007a). Family accommodation refers to ways in which family member s accommodate patient symptoms, such as providing reassurance, avoi ding feared stimuli, taking over a patients duties, participation in rituals and modifying family routines. Theoretically, family accommodations negatively reinforce the patients symptoms, thereby affecting symptom severity and adversely affecti ng treatment outcome (Leane, 1991). For example, provision of reassurance is negat ively reinforcing because it temporarily reduces anxiety. Similarly, allowing the patient to avoid f eared situations is negatively reinforcing. Given that family accommodation has been related to symptom severity and treatment outcome in children and adults with Obsessive-Compulsive Disorder (OCD; Amir, Freshman, & Foa, 2000; St orch et al., 2007a; Storch et al., 2007b), it may be a family variable that is import ant to investigate in other pediatric anxiety disorders. Similar to the manner in which family a ccommodation is theorized to interfere with treatment through provision of negative and positive reinforcement of OCD symptoms, families of children with other anxiety di sorders may also be reinforcing symptoms through accommodation. Although children wit h other anxiety disorders do not have

PAGE 13

13 rituals, parents may accommodat e in many other ways (e.g., providing reassurance, allowing avoidance of feared stimuli, taking over a patients responsibilities). Although there is not curr ently a measure of family accommodation for general child anxiety, Calvocoressi et al. (1995; 1999) developed a 13-item measure of family accommodation specific to sym ptoms of OCD, demonstrating good internal consistency ( = .76 .82), interrater reli ability (ICCs = .72 to 1.0), and construct validity through higher correlations with measures of fam ily functioning and family stress than with measures of financial stress and stress asso ciated with caring for those with physical disabilities The FAS targets the followi ng forms of accommodation: facilitation of compulsions, providing reassurance, facilitat ing avoidance, modifying family routines, and assuming responsibilities of the patient. These accommodating behaviors were associated with reduced global functioning and increased symptom severity. Following development of the FAS, resear chers have been able to study family accommodation in adults and children with OCD. Up to 90% of families of OCD adults report at least minimal accommodation of sym ptoms (Allsop & Verduy n, 1990; Storch et al., 2007b). Family accommodation has also been positively related to depression and anxiety in family members of adult OCD pat ients (Amir, Freshm an, & Foa, 2000). In family members of patients with OCD, diseng agement coping strategies such as family accommodation were positively related to depression and inversely related to hope and adaptive coping (Geffken, Storch, Duke, M onaco, and Goodman, 2006). In a study of children with OCD, most families repor ted some degree of family accommodation (Storch et al., 2007a). Family accommodation was also related to symptom severity, functional impairment, and to comorbid internalizing and externalizing behavior

PAGE 14

14 problems. Storch et al. (2007b) found t hat family accommodation mediates the relationship between symptom severity and parent-rated functional impairment, further demonstrating the potential importance of addressing family accommodation in treatment. In a study of adults with OCD, Amir, Fr eshman, and Foa (2000) found that family accommodation and family distress were related to symptom severity post-treatment, suggesting that family accommodation hinders progress in treatment. They argue that family functioning, including family acco mmodation, should be addressed as part of treatment. Similarly, Ferrao et al. (2006) reported that whil e most families of patients with OCD entering treat ment reported some level of accommodation, treatment refractory patients demonstrated the highest leve ls. In that study, 52.4% of families of refractory patients were classified as hav ing extreme family accommodation, compared with only 3.8% of treatment re sponder families. In a study of family-based CBT for pediatric OCD, family accommodation was r educed following treatment, although this study did not examine the mechanism of change (Storch et al., 2007a). Despite its importance in treatment outcome of pediatric OCD, family accommodation has not been studied in children with other pediat ric anxiety disorders. Similar to the way in which family accommodation reinforces OCD symptoms, it may reinforce symptoms in other pediatric anxiety disorders. In Generalized Anxiety Disorder (GAD), for example, parents may provide reassurance about many different worries, or allow their children to avoid situations which might elicit anxiety (e.g., watching the news). In Social Phobia (SP) parents may allow their children to avoid feared social situations, or r earrange family routines to a ccompany the child to social

PAGE 15

15 situations which the child might otherwise f ear. By nature of t he disorder, parents of children with Separation Anxiety Disorder ( SAD) are involved in sym ptoms. Parents of children with SAD likely spend extra time upon separating trying to reassure the patient, or rearrange family schedules to allow children to avoi d separation. Accommodating behaviors such as these have the potential to interfere with treatment, as they may undermine techniques that are cent ral to completion of CBT. Family Accommodation and Pediatric Anxiet y Treatment Similar to OCD, other anx iety disorders are hypothesized to be maintained through negative reinforcement. Existence of contingency management as an integral part of treatments for pediatric anxiety implies that many parents need to cease providing negative reinforcement, through l earning ways to reinforce appropriate behaviors and not to reinforce inappropriate, anxiety-driven behaviors (Kendall, 1994; Kendall, 2000; Spence et al., 2000). Studies have not investigated which behaviors parents engage in that necessitate contingency management training. It may be that in some cases parents engage in accommodation of symptoms that inadvertently reinforce anxiety. Kendall (1990) reported that when family members are not included in treatment of pediatric anxiety disorders, one-third to one-half of patients continue to have clinically significant anxiety problems after treatment However, studies directly comparing treatments with and wit hout family involvement have yielded mixed results (Barrett et al., 1996; Cobham et al., 1998; Barrett et al., 1998; Mendlowitz et al., 1999; Nauta et al., 2003, Spence et al., 2000; W ood, Piacentini, Southam-Gerow, Chu, & Sigman, 2006). Of seven studies to date, two demonstrated no difference between treatments with and without parent involvement (Nauta et al., 2003; Spence et al., 2000), four found

PAGE 16

16 increased benefits of par ents involvement on only some outcome measures (Barrett et al., 1996; Cobham et al., 1998; Barrett et al., 1998; Mendlowitz et al., 1999), and only one demonstrated clear benefit of parent involvement on a majority of outcome measures (Wood et al., 2006). Although lack of significant findings may indicate low power for finding differences between active treat ments, the majority of these studies do not thoroughly describe the family intervention, and it appears that family components of treatment may vary widely amon g treatment studies in general. Thus, it is likely that some of the interventions tested have not ta rgeted family variables that have an effect on child anxiety. In the only study to date demonstrating be tter outcome on a majority of measures for parent involvement in treatment, Wood et al. (2006) investigated a family-based intervention including modul es designed to target parent intrusiveness and discouragement of child autonomy. Par ental intrusiveness has been described as parent performance of tasks a child should be performing independently (Chorpita & Barlow, 1998; Whaley et al., 1999; Wood et al., 2003), which is similar to some family accommodations described by Calvocoressi et al. (1999). There was a greater decline in anxiety severity for children assigned to the family-based intervention, and 78.9% of those in the family-based intervention remi tted diagnostic status compared with 52.6% in the child treatment condition. Howeve r, this study failed to investigate the mechanism of change responsible for the di fference between groups, so conclusions about the role of these factors cannot be drawn. In summary, although it has been shown that inclusion of families in tr eatment can enhance treatment outcome, few studies have found a consistent relationshi p between specific family variables and

PAGE 17

17 childhood symptom expression or treatment outcome. This discrepancy suggests that there are family variables that influence treatment outcome which have not yet been identified. Aims of the Current Study Currently, there is a need to identify family variables that have a role in pediatric anxiety symptom presentati on and treatment outcome. Although early research with pediatric and adult OCD has indicated the potential im portance of family accommodation, studies have not investigated its role in other pediatric anxiety disorders. Study of this construct is limited by lack of a measure of family accommodation for general pediatric anxiety. The current study aims to develop and validate such a measure, and to collect preliminary data rega rding correlates of family accommodation in pediatric anxiety. This study also aims to te st for a mediation effect of family accommodation in the relationship between symptom severity and functional impairment, previously reported by Storch et al. (2007b) in a pediatric OCD sample. As well, we will test for family accommodation as a possible mediator of the relationship between parent psychopathology and child symptom severity.

PAGE 18

18 CHAPTER 2 METHOD Developme nt of the PAS o ccurred in three phases. Phase 1: Initial scale development. Items were initiall y generated from the FAS (Calvocoressi et al., 1999; see Table 1), review of the extant literature, and clinical experience of a writing committe e. In addition to the firs t author, the writing committee was comprised of 5 members representi ng psychologists and psychiatrists with expertise and interest in pediatric anxiety di sorders. Members were Dr. Eric Storch, Ph.D. (University of South Flor ida), Gary Geffken, Ph.D. (U niversity of Florida), Tanya Murphy, M.D. (University of South Florida), Ayesha Lall, M.D. (University of Florida), and Golda Ginsburg, Ph.D. (Johns Hopkins University). Phase 2: Pilot testing. We initially examined the measure in a sample of 5 children and their caregivers. Based on feedback from these families, we added several items to the PAS to reflect res ponses of children when accommodation is not provided (i.e. added sad/down and angry/abusive) and altered the response choices to provide more concrete frequency options (i .e. changed response range from never to always to a new response range of never to + times per da y). After making these revisions, we piloted the PAS in a se cond group of 5 patients and their caregivers. Those families did not have any suggestions for substantive changes to the measure. Phase 3: Scale validation. The final version of the PAS was complete following phases 1 and 2. The PAS, including all items, is included in the appendix. The PAS was designed to be a clinician-rated scale incorporating information from the patient, caregivers, and clinician judgment. It may be administered by a psychologist, psychiatrist, or a trained clinical interviewer. The final PAS contai ns 16 items, each of

PAGE 19

19 which includes a specific question followed by a series of examples that can be altered by the administering clinician. For exam ple, to assess whether parents facilitate avoidance, the PAS asks How often in the last week did you help your child avoid things or situations that might make him/her more anxious, such as talking to others, separating from you, or watc hing the news? Clinicians are able to substitute known symptoms for the provided examples, tailoring the administration for each family. In addition, the semi-structured fo rmat permits follow-up questions as necessary to clarify a response or to probe more carefully around a family routine. Responses are then provided on a 0 (never) to 8 (11+ times per day) scale. Participants, measures, and procedures for validation of the PAS are described below. Participants Participant s were 59 children ( n = 27 male, n = 32 female) and their caretakers. Children ranged in age from 7 to 17 ( M = 12.59, SD = 2.97). Children were primarily Caucasian (91.4%; 3.4% Hispanic, 5.2% Other) and average household income was $77,633.20 ( SD = $43,618.64, Median = $69,000). All children had a principal diagnosis of Generalized Anxiety Disorder (GAD; n = 30), Social Phobia (SP; n = 19), or Separation Anxiet y Disorder (SAD; n = 11) according to DSM-IV-TR criteria (American Psychiatric Association, 2000). These di sorders were chosen based on evidence that anxiety symptoms are similar in all three di sorders (Ginsburg, Riddle, & Davies, 2006) and that these disorders are very frequently co mbined in treatment studies (CartwrightHatton et al., 2004). There were no differ ences on any study variables based on diagnosis ( p s > .05). Inclusion criteria were: (a) principal diag nosis of GAD, SAD, or SP, (b) child age between 6 and 17 years, and (c) parent age above 21 years. Exclusion criteria were (a)

PAGE 20

20 positive diagnosis of psychosis or autism, (b) principal diagnosis other than GAD, SAD, or SP and (c) positive diagnosis in the caregi ver of mental retardation, psychosis, or other psychiatric disorders or conditions that would limit their ability to complete measures. In order to maxi mize generalizability, children with comorbid diagnoses (except psychosis or autism) were included. Comorbid diagnoses, when present, included presence of another anxiety disorder ( n = 26), Attention Deficit Hyperactivity Disorder (ADHD; n = 10), Major Depressive Disorder ( n = 8), OCD (n = 2), Oppositional Defiant Disorder (ODD; n = 1), and Tourettes Disorder ( n = 1). All children and their caretakers were recruited through one of two outpatient anxiety disorder specialty clinics: one clinic affiliated with the University of Florida (UF) and one clinic affiliated with the University of South Florida (USF). Examination of all study variables by site revealed no signifi cant differences in demographic information, child-report, or parent-report data ( p s > .05). However, signi ficant differences emerged on the clinician-administe red measures: the PAS, t (56) = -2.04, p < .05, and the Pediatric Anxiety Rating Scal e (PARS; see Measures section for a full description), t (55) = -2.90, p < .05). It is possible that differences were due to the level of experience of clinical interviewe rs at the two sites. Clinical Interviewers Interviews were conducted by trained clinic al interviewers. At the UF site, clinical interviewers included undergraduate level resear ch assistants with prior training in child psychopathology and additional experience wit h OCD patient populati ons. At the USF site, clinical interviewers were postundergraduate level research assistants with experience with OCD patient populations; t hough these research assistants had more assessment experience with this population. Although there were differences between

PAGE 21

21 sites in the experience of clinicians with t he population, use of both sites was necessary for feasibility and recruitment of an adequate sample size. To ensure fidelity across sites, all clinical interviewers were trained to a reliable standard on clinician rated measures through a training workshop, joint interviews and discussion. More specifically, to complete training, clinical interviewers completed practice administrations using standardized vignettes, observed two administrati ons of each measure by the first author or another trained clinical interviewer, and completed two practice interviews under supervision. In addition, clinical intervie wers attended weekly assessment meetings to promote quality assurance of study assessments. Measures Clinician Administered Pediatric Anxiet y Rating Scale. The Pediatric Anxiety Rating Scale (PARS; Research Units on Pediatric Pharmacology Anxiety Study Group, 2002) was used to rate the severity of anxiety symptoms, and contains separate symptom checklist and severity rating scales. The symptom checkl ist contains 50 items assessing presence or absence of anxiety symptoms related to so cial interactions, separation, generalized anxiety, specific phobias, and physical symptom s. The severity rating scales contain seven items rating severity of anxiety feelings, physical sym ptoms, avoidance, interference, overall severity, and overall frequency. The PARS has demonstrated inter-rater reliability (ICC = .97), conv ergent (with the HAM-A and SCARED-P) and divergent validity (with CBCL-Externalizing and CDRS), and sensitivit y to treatment effects (Research Units on Pediatric Pharmaco logy Anxiety Study Gr oup, 2002). In this sample, reliability was adequate at both sites (Cronbachs alpha = .79 at UF and .74 at USF) and inter-rater reliability was good (ICCs ranged from .89 to 1.00 at both sites).

PAGE 22

22 Pediatric Accommodation Scale. The PAS, described in the present study, was administered to examine family accommodation of child anxiety symptoms. The PAS contains 16 items and the psychometric properties are detailed in the present investigation. Child Report Child Anxiet y Impact Scale-Child Version. The Child Anxiety Impact ScaleChild version (CAIS-C; Langley, Bergman, Mc Cracken, & Piacentini, 2004) was used to examine functional impairment due to anxiety symptoms. The CAIS-C is a 34-item scale measuring the amount of impairment across domains: school, social activities, home/family activities, and globally. Respons es are given on a four point scale from not at all to very much. T he CAIS-C has demonstrated good psychometric properties, including internal consistency (C ronbachs alpha = 0.73 to 0.87), convergent validity (with CBCL-Int ernalizing and Child Depression In ventory), and divergent validity (CBCL-Externalizing; Langley et al., 2004). Reliability was excellent in this sample (Cronbachs alpha = .92). Childrens Depression Inventory. The Childrens Depression Inventory (CDI; Kovacs, 1992) is a 27-item child-report m easure that assesses the presence and severity of cognitive, affective, or behavioral symptoms of depression during the previous two weeks. Widely used, the CDI has adequate internal consistency ( r = .71.89) and differentiates bet ween depressed and non-dep ressed youth (Carlson & Cantwell, 1979). In addition, the CDI has good test-retest reliability ( r = .87) and construct validity as determined by high corre lations with similar depression measures and through factor analysis (Kovacs, 1992; Craighead et al., 1998). Reliability was excellent in this sample (Cronbachs alpha = .91).

PAGE 23

23 Multidimensional Anxiety Scale for Children. The Multidimensional Anxiety Scale for Children (MASC; March et al., 1997) is a 39-item child-report index that assesses symptoms of general, social, and separation anxiety. Good psychometrics have been reported, including construct validit y through confirmatory factor analysis, internal consistency (Cronbachs alpha = .87) convergent validity (with the STAI-C and RCMAS), and divergent validity (with the BDI, CDRS, and HAM-D; Ryn et al., 2006; March et al., 1997, 1999). Reliability was good in this sample (Cronbachs alpha = .89). Parent Report Beck Depr ession Inventory-Second Edition Based on the original Beck Depression Inventory (Beck et al., 1961) the Beck Depression Inventory-Second Edition (BDI-II; Beck et al., 1996) is a 21-item self-report measure of depressive symptoms experienced during the past week. Ex tensive reliability and validity data have been reported in clinical and non-clinical sa mples, including evidence of construct validity using confirmatory factor analysis, high internal consistency (Cronbachs alpha = .90), convergent validity with other measures of depression, and divergent validity with measures of anxiety (Beck et al., 1996; Whisman et al., 2000; Storch et al., 2004). Reliability was good in this sample (Cronbachs alpha = .87). Child Anxiety Impact Scale-Parent Version. The Child Anxiety Impact ScaleParent version (CAIS-P; Langley, Bergman, McCracken, & Piacentini, 2004) was used to examine functional impairm ent due to anxiety symptoms. The CAIS-P is a 34-item scale measuring the amount of impairment across domains: school, social activities, home/family activities, and globally. Respons es are given on a four point scale from not at all to very much. The CAIS-P has demonstrated good psychometric properties, including internal consistency (C ronbachs alpha = 0.73 to 0.87), convergent

PAGE 24

24 validity (with CBCL-Int ernalizing and Child Depression In ventory), and divergent validity (CBCL-Externalizing; Langley et al., 2004) Reliability was good in this sample (Cronbachs alpha = .89). Child Behavior Checklist. The Child Behavior Che cklist (CBCL; Achenbach, 1991) is a parent-rated questionnaire asse ssing the frequency and intensity of behavioral and emotional problems exhibited by the child in the past six months. The CBCL consists of eight syndrome sca les (withdrawn, somatic complaints, anxious/depressed, social problems, thought problems, attention problems, delinquent behavior, and aggressive behavior) and two co mposite scales (externalizing and internalizing problems). Over all, the CBCL has excellent psychometric properties including one-week test-rete st reliability, adequat e internal consistency (Cronbachs alpha = .62-.92), and construct validity (e.g., strong associations with subscales of other measures that assess similar constructs; Achenbac h, 1991). Reliability was good in this sample (Cronbachs alpha = .93 externalizing and .95 internalizing). Family Adaptability and Cohesion Eval uation Scale fourth edition (FACESIV). This questionnaire is designed to assess family cohesion and family flexibility (Olsen, Gorall, and Tiesel, 2007). Within cohesion, the FACES assesses both overall Cohesion and unbalanced cohes ion in the form of Enmeshment and Disengagement. Within flexibility, the FACES assesses both overall Flexibility and unbalanced flexibility in the form of Chaos and Rigidity. The FA CES consists of 62 statements about family characteristics, asking family members to in dicate agreement on a 5-point scale from 1 (strongly disagree) to 5 (strongly agree) The FACES-IV has demonstrated good psychometric properties, including reliability (Cronbachs alpha ranging from .77 to .87)

PAGE 25

25 and convergent validity (with Se lf-Report Family Inventory, Family Satisfaction Scale, and Family Assessment Device). Reliability for the subscales was good in this sample, with Cronbachs alpha ranging from .78 (rigid) to .94 (cohesion). State-Trait Anxiety Inventory Trait version (STAI-T). The State-Trait Anxiety Inventory Trait version (STAI-T; Spielberger, 1983) is a 20-item scale rated on a 4point Likert scale that measures the t endency of people to experience general anxiety and view stressful situations as threatening. The widely-used STAI-T has been found to have good internal consistency (Cronbachs alpha = .89 to .91), convergent validity (with the Penn State Worry Questi onnaire and Worry Domains Questionnaire) and divergent validity (with Lykkens Activity Questionna ire), and can discriminate between patients with anxiety and non-clinical controls (Bie ling et al., 1998; Sp ielberger, 1983, 1989). Reliability was excellent in the present sample (Cronbachs alpha = .92). Procedures The Univer sity of Florida and University of South Florida institutional review boards provided appropriate human subjects ethi cal approval of this project. A waiver of informed consent was granted so that the first author or another study representative could identify qualifying participants by diagnosis. Diagnoses were made by the patients clinician (a licensed child clinic al psychologist or board-certified child psychiatrist), through clinical interview, behavioral observation, parent-report, childreport, and all available information. The fi rst author verified diagnosis through review of medical records and through information collected during clinician interview. Additionally, the first author discussed evidence for diagnoses of participants with the diagnosing clinician. The first author and clinician shared 100% diagnostic agreement about all participants. Following verification of diagnosis, clinicians of qualifying

PAGE 26

26 patients were then notified of pat ient eligibility. The clinician subsequently informed a study representative when parents and patients expressed intere st in participating. Thereafter, the study representative fu lly described the study and obtained informed parental consent and child assent. Participan ts then completed study measures in a private setting. First, the clinical in terviewer administered the PARS and the PAS, following which children and their caregivers completed self-report measures. Children completed the MASC, CDI, and CAIS-C whil e parents completed the CBCL, CAIS-P, FACES-III, BDI-II, and STAI-T Although some very young children were not able to complete self-report measures alone, inclusi on of young children in the current study is important to determine the role of family ac commodation for children of all ages. This is particularly critical given a lack of any treatment research in very young children (Verdeli, Mufson, Lee, & Keith, 2006). Theref ore, the clinical interviewer verbally administered child-report measures to child ren who were unable to complete them alone ( n = 3). Completion of all measures took between 60 and 90 minutes.

PAGE 27

27 CHAPTER 3 RESULTS Evaluation of Scale Items Means, standard deviations, and item-total co rrelations for each of the 16 items on the PAS are presented in Table 1. Examin ation of skewness and kurtosis values (see Table 1) reveal that at the UF site, 7 item s were positively kurtotic and 15 items were positively skewed such that items indicati ng less accommodation were most frequently endorsed. At the USF site, 8 items were pos itively kurtotic and 11 items were positively skewed. Percentage of parents endorsing the item (i.e. indica ting that it occurred at least once in the preceding week) is also reported in Table 1. The most frequently endorsed items include item 1/providing reassurance (UF = 70.1%, USF = 90.5%), item 2/allowing children to avoid anxiety provoking stimuli (U F = 63.2%, USF = 71.4%), and item 12/parents becoming distressed when providing accommodation (UF = 55.3%, USF = 75.2%). The least frequently endors ed items include item 5/parents changing their work schedule (UF = 36.8%, USF = 19.0%), item 9/changing child social activities (UF = 13.2%, USF = 38.1%), and item 11/ changing the family schedule (UF = 34.2%, USF = 23.8%). Examination of item-total correlations revealed scores ranging from .79 to .84 at the UF site, and from .09 to .67 at the USF site. Low item-total correlations (< .3) may be accounted for by some items that were susceptible to influence of severity rather than frequency. For exam ple, allowing a child to sleep with parents nightly may earn a rating of (11+ times per day) becaus e it is a level of accommodation that the rater deemed not adequately captured by a rating of (1 time per day) Note that this was not a standardized instruction, and thus, ratings that might be influenced by severity were made based on individual j udgment and not provi ded uniformly.

PAGE 28

28 Table 3-1. PAS item characteristics by site Item M (SD) Item-total ( r )Skewness Kurtosis % EndorsingInter-rater Reliability (ICC)a 1. Reassurance 2.41(2.11) 4.48(2.58) .80 .43 0.28 -0.22 -1.21 -.088 71.1 90.5 .99 1.00 2. Avoidance 1.76(1.88) 2.33(2.24) .80 .48 0.80* 0.83 -0.43 0.21 63.2 71.4 1.00 1.00 3. Parent social change 0.24(0.60) 0.81(0.93) .82 .61 2.41* 0.83 4.65* -0.29 15.8 52.4 .78 1.00 4. Parent routine change 1.54(1.88) 1.05(1.83) .80 .67 1.07* 1.66* 0.39 1.60* 50.0 33.3 .99 1.00 5. Parent work change 0.81(1.35) 0.81(2.04) .81 -.37 1.42* 2.78* 0.65 7.75* 36.8 19.0 1.00 1.00 6. Parent leisur e change 0.81(2.00) 0.91(1.58) .80 .45 2.98* 1.66* 8.51* 1.69* 23.7 33.3 1.00 .33 7. Child responsibility change 1.27(1.66) 1.71(2.45) .81 .55 1.31* 1.20* 0.87 0.36* 50.0 42.9 .94 .36 8. Child school change 0.43(0.83) 2.10(2.84) .81 .58 2.29* 1.31* 5.58* 0.50 31.6 52.4 .88 1.00 9. Child social change 0.16(0.44) 0.95(1.91) .81 .20 2.92* 2.93* 8.59* 9.52* 13.2 38.1 .93 .79 10. Child sleep change 1.41(2.01) 1.71(2.24) .84 .09 1.23* 1.25* 1.09 1.32 42.1 47.6 .99 1.00 11. Family schedule change 0.73(1.12) 0.57(1.16) .80 .40 1.24* 2.01* -0.05 3.20* 34.2 23.8 1.00 1.00 12. Parent distress 1.43(2.08) 1.86(1.59) .79 .49 1.64* 0.67 2.03* -0.33 55.3 76.2 .84 1.00 13. Child distress/anxiety 1.54(1.77) 1.86(1.80) .80 .43 0.94* 0.52 -0.19 -1.15 56.8 66.7 1.00 1.00 14. Child oppositional/ defiant 1.27(1.68) 1.33(2.29) .79 .44 1.05* 1.76* -0.27 2.49* 50.0 33.3 .95 1.00 15. Child angry/abusive 0.89(1.51) 1.23(1.82) .79 .63 1.78* 1.12* 2.59* -0.24 39.5 42.9 .93 1.00

PAGE 29

29 Table 3-1. Continued Item M (SD) Item-total ( r )Skewness Kurtosis % EndorsingInter-rater Reliability (ICC)a 16. Child sad/down 1.03(1.42) 1.43(1.57) .80 .36 1.76* 1.27* 3.34* 0.96 47.4 66.7 .80 1.00 an = 21(UF), n = 7 (USF), *p <.05. Data are presented by site, wi th line 1 = UF (n = 38), line 2 = USF (n = 21). Responses to th e PAS are provided on a 0 (never) to 8 (11+ times per day) scale based on t he frequency the behavior occurred in the preceding week.

PAGE 30

30 Total scores on the PAS ranged from 0 to 49 ( M = 17.73, SD = 13.22) at UF and from 6 to 54 at USF (M = 25.19, SD = 15.37), with Shapiro-Wilk tests of normality indicating significantly non-normal dist ribution of scores at both sites ( p s < .05). Examination of skewness and kurtosis values for the PAS indicated significant positive skewness (Skewness = .749, SD = .389) and absence of kurt osis (Kurtosis = -.333, SD = .778). This finding is similar to that of Calvocoressi et al. (1999), who reported positively skewed total scores on the FAS for OCD. Reliability Reliability statistics were calculated s eparately for each site. Cronbachs alpha was good at both sites (UF = .81 and USF = .78) Inter-rater reliability was assessed by two live, independent raters who were blind to the ratings of the other. Inter-rater reliab ility was available for a portion of the sample (n = 27 of 59), based on the availability of raters during m easure administration. Feasibi lity did not permit two raters to be available for all administrations of the PAS. Intraclass correlation coefficients were calculated for each item (see Table 1) and ranged from .78 to 1. 0 at the UF site ( n = 20) and from .33 to 1. 0 at the USF site ( n = 7). Convergent and Discriminant Validity Given that the distribution of scores on the PAS was sig nificantly positively skewed, square root transformation was app lied to total PAS scores. Square root transformation is an approach that norma lizes positively skewed data and is a conservative transformation relative to other approaches (e.g., logarithmic transformation, inverse transformation). Fo llowing square root transformation, Shapiro Wilk tests of normality indicated that the a ssumption of an underlying normal distribution could be maintained ( p > .05). As well, examinati on of individual skewness scores

PAGE 31

31 indicated that square root transformed PAS scores were not significantly skewed (Skewness = -.23, SD = .31). In subsequent analyses, where PAS scores are reported they reflect the square root trans formation unless otherwise noted. PAS total scores were examined by site for evidence of convergent and discriminant validity using Pear sons correlations (see Table 2) To correct for the risk of making a Type I error through multiple analyses, significance was set at p < .01. Given the small sample size in the current study, particularly when examining site differences, Bonferroni correction was not used to avoi d the risk of Type II error. UF site. As expected, the PAS correlated significantly with parent-rated impairment (CAIS-P; r = .46, p < .01). It did not correla te significantly with child-ra ted symptoms of depression (CDI; r = -.05, p > .01). Fishers r to z comparis on test shows that the PAS is more strongly correlated with the CAIS-P than wit h the CDI, providing some evidence for convergent and divergent validity ( z = 2.22, p < .05). Contrary to expectations, the PAS did not correlate significantly with other similar constructs such as child-rated anxiety symptoms (MASC; r = .20, p > .01), overall anxie ty symptoms (PARS; r = .25, p > .01), child rated impairment (CAIS-C; r = .17, p > .01), or inte rnalizing symptoms (CBCLInternalizing; r = .15, p > .01). USF site. As expected, the PAS correlated significantly with overall anxiety symptoms (PARS; r = .59, p < .01), child-rated anxiety symptoms (MASC; .58, p < .01), and internalizing sym ptoms (CBCL-Internalizing; r = .58, p < .01). Contrary to expectations, the PAS did not correlate significantly with either child-rated or parent-rated impairment (CAIS-C; r = .40, p > .01; CAIS-P; r = .44, p > .01). As well, the PAS correlated significantly with child-rated depre ssion, a theoretically distinct construct (CDI; r = .68, p < .01).

PAGE 32

32 Table 3-2. Pearson correlations among study variables PASa PARS CAISP CAISC CDI MASC CBCL -Int CBCLExt BDI-II STAIT Cohes Flex Disen Enme Rigid Chao PAS -.25 .59* .46* .44 .17 .40 -.05 .68** .20 .58* .15 .58* .26 .52 .37 .57* .23 .67** -.12 -.53 -.02 -.56* .00 .13 .36 .49 .28 .15 .15 .48 PARS -.37 .28 .53** .35 .39 .49 .56** .50 .29 .33 .10 .32 -.03 .31 .24 .49 .05 -.41 .09 -.25 -.15 .23 -.01 .34 .04 .04 -.08 .44 CAISP -.38 .48 .23 .38 .40 .41 .43* .44 .26 .59* .17 .31 .26 .49 -.18 -.41 -.23 -.40 .10 .16 .39 .32 .18 .27 .09 .41 CAISC -.47* .71** .69** .49 .53** .64* .03 .39 -.11 .26 .12 .36 -.11 -.20 -.03 -.43 .00 .16 .25 .32 -.03 .17 .18 .40 CDI -.62** .66* .50** 81** .08 .56 -.20 .35 .01 .57 -.25 -.63* -.03 -.68** .03 .28 -.08 .18 -.23 .05 .09 .61* MASC -.43 .40 -.08 .31 -.10 .49 .36 .56* -.22 -.60* -.13 -.76** .10 .41 .17 .35 -.11 -.01 .27 .72** CBCLInt -.46* .52 .01 .54 .09 .68* -.27 -.38 .09 -.43 .15 .08 .42 .24 .03 .15 .21 .52 CBCLExt -.16 .29 .17 .43 -.15 -.48 .04 -.34 .12 .07 .11 .47 .19 .14 .12 .50 BDI-II -.71** .87** -.40 -.43 -.49* -.37 .51** .19 .36 .46 -.02 -.22 .44* .61** STAI-T --.36 -.50 -.55** -.45 .47* .20 .31 .52 .01 -.16 .45* .69** Cohes -.64** .76** -.70** -.65** -.52** -.54 .20 .11 -.59** -.51 Flex --.54** -.45 -.33 -.38 .27 .03 -.44* -.64** Disen -.41* .48 -.16 .01 .67** .27 Enme -.18 .31 .39 .50 Rigid --.26 -.03 Chao -*p<.01, **p<.003 (.05/16 = .003), aPAS scores were square-root transformed. Data are presented by site, with line 1 = UF (n = 38), line 2 = USF (n = 21). PAS = Pediatric Accommodation Scale, PARS = Pediatric Anxi ety Rating Scale, CAISP = Child Anxiety Impact Scale (Parentrated),

PAGE 33

33 CAISC = Child Anxiety Impact Scale (Child-rated), CDI = Child Depr ession Inventory, MASC = Multidimensional Anxiety Scale for C hildren, CBCLInt = Child Behavior Checklist-Internalizing scale, CBCL-Ext = Child Behavior Checklist-Externalizing scale, BDI-II = Beck Depression Inventory Second Edition, STAI-T = State-Trait Anxi ety Inventory-Trait scale, Cohes = Family Adaptablilty and Cohesion Scale-Cohesion sub scale, Flex = Family Adaptablilty and Cohesion Scale-Flexibility subscale, Disen = Family Adaptablilty and Cohe sion Scale-Disengaged subscale Enme = Family Adaptablilty and Cohesion Scale-Enmes hed subscale, Rigid = Family Adaptablilty and Cohesion Scale-Rigid subscale, Chao = Family Adaptablilty and Cohesion Scale-Chaotic subscale.

PAGE 34

34 The PAS was not significantly correlat ed with externalizing symptoms (CBCLExternalizing, r = .52, p > .01), although the size of the co rrelation was large. Fishers r to z comparison tests do not show that t heoretically similar constructs show a significantly stronger relationship to the PAS compared with those that are theoretically dissimilar ( p s > .05). Exploratory Analysis of Fa mily Accommodation Construct Pearsons correlations were conducted to explore relationships of the PAS with constructs of interest (see Table 2). UF Site. Contrary to expectations, data from the UF site revealed that the PAS was not signifi cantly related to any measures of parent psychopathology or family functioning (see Table 2). USF Site. As expected, the PAS was significantly positively related to parent symptoms of depre ssion (BDI-II; .57, p < .01) and anxiety (STAI-T; r = .66, p < .01). The PAS was negatively related to family flexibility (FACES-Flexibility; r = -.56). Inconsistent wit h previous findings, the PAS was not significantly related to fa mily cohesion (FACES-Cohesion; r = -.53), although the size of the correlation was large and approached significance. Again, different findings between the two sites are likely due to the di fference in skill of clinical interviewers administering both t he PAS and the PARS. Mediation Analyses This study aimed to test a mediati on model in which family accommodation mediates the relationship between anxiety symptom severity and functional impairment, as reported by Storch et al (2007b). As well, we proposed a second mediation model in which family accommodation mediat es the relationship between parent psychopathology and anxiety symptom severity. However, given the site differences in the PAS data, we tested a series of models in which site moderates the mediation effect

PAGE 35

35 (see Figure 1). All models included site as a moderator and us ed bootstrapping ( n = 5000), as other techniques (i.e. Sobels test ) may result in unstable estimates when sample sizes are small. Models were anal yzed in SPSS using the macro provided by Preacher, Rucker, and Hayes (2007). Model 1. This model tested PAS as a mediator of the relationship between symptom seve rity (PARS) and parent-rated functional impairment (CAIS-P). Results indicated nonsignificant indirect effects at both UF ( B = .04, z = 1.31, p > .05) and USF ( B = .11, z = 1.06, p > .05). Model 2. This model tested PAS as a mediator of the relationship bet ween symptom severity (PARS) and childrated functional impairment (CAI S-C). Results indicated nonsignificant indirect effects at both UF (B = .01, z = .26, p > .05) and at USF ( B = .15, z = .94, p > .05). Model 3. This model tested PAS as a mediat or of the relationship between parent depressive symptoms (BDI-II) and symptom severity (PARS). Results indicated nonsignificant indirect effects at both UF ( B = .09, z = 1.07, p > .05) and USF ( B = .13, z = .92, p > .05). Model 4. This model tested PAS as a mediat or of the relationship between parent anxiety symptoms (STAI-T) and symptom severity (PARS). Results indicated nonsignificant indirect effects at both UF ( B = .03, z = .87, p > .05) and at USF ( B = .12, z = 1.06, p > .05). PAS CAIS-P .08 .63 Site Model 2 -2.35 .94 PARS Site Model 1 -1.91 .20

PAGE 36

36 CHAPTER 4 DISCUSSION The current study investigated the reliabilit y and validity of the PAS in a sample of children with GAD, SAD, and SP and is t he first investigation of the family accommodation construct in this population. Data were collected from two anxietydisorder specialty clinics, and results s howed a significant difference in PAS scores based on site. Notably, the two sites show ed significant differences on both clinicianrated measures. These results may have been influenced by the interviewers level of training and experience with an anxiety diso rders population. Although we made every effort to standardize clinical interviewer training, those at the USF site had prior specialty assessment experience with an OCD population that may have benefited administration of the PAS. It may be that capturing family accommodation requires careful probing around symptoms and family acti vities. For ex ample, a parent who initially denies changing his/ her work schedule but previ ously reported changing the childs school schedule might be asked about the details of the family routine when the child is not in school (e.g., Earlier you told me that you had to take Mary to school late on Tuesday. Did you get to wo rk on time that day? ). Therefore, resu lts of the present investigation may indicate the importanc e of clinician exper ience with an anxiety population when administering this measure. However, we did not measure clinical interviewer experience and therefore cannot be cert ain that this is the cause of the site differences. Overall, the PAS showed good internal c onsistency and inter-rater reliability across two sites. Inter-rater reliability was lo w on two items at the US F site, but this was likely due to the small number of participants ( n = 7) for whom two raters were available

PAGE 37

37 at that site. As well, verbal indication t hat some clinical interviewers incorporated severity into ratings, while some did not, may have contributed to low inter-rater reliability. Evidence for convergent va lidity was mixed, with the PAS showing a significant, positive relationship with other measures of anxiety symptoms, anxiety related impairment, and internalizing symptoms. Re sults at the UF site indicated that the PAS did not correlate significantly with a measure of child-rated depression, which supports divergent validity, but results from t he USF site show a strong relationship with this measure. Given that there were no si gnificant differences in depressive symptoms or comorbidity between sites, the difference may be related to the relationship between accommodation and depression at each site or possibly to differential ratings of interviewers at each site. Examination of individual items on the PAS revealed that item-total correlations ranged from .79 to .84 at the UF site and from .09 to .67 at the USF site. Items with low item total correlations (< .3) at USF were it em 9 (changing the childs social activities) and item 10 (changing the childs sleep r outine). One discrepanc y among items that may explain this is the susceptibility of some items to severity of accommodation. For example, for a child that sleeps with parents nightly, some interviewers provided a rating of + times per day because a rating of time per da y does not adequately capture the seriousness of the accommodation. In this case, although clinical interviewers were not trained to incorporate severity in ratings of accommodation, severity of the behavior interfered with the frequency rati ng. If some items partially reflect severity, which may be a separate construct, those items may no t be as closely related to the scale. Additionally, it is possible that this interf erence occurred to a great er degree at the USF

PAGE 38

38 site if clinical interviewers, who may have probed more closely than those at the UF site, relied more heavily on clinical judgment rath er than parent report. However, given that we did not measure clinician judgment, we cannot draw this conclusion with certainty and it is possible that future analysis of the PAS will identify these items as those that should be removed from the scale. In future studies of the PAS, addition of a severity scale may benefit this measure so that we may separately assess the severity construct. Further examination of individual item s shows that accommodation of anxiety symptoms in the previous week was endorsed by a high percentage of parents. Most commonly, parents report ed providing reassurance to their children (UF = 71.1%, USF = 90.5%) and helping their children avoid anxi ety-provoking stimuli (UF = 63.2%, USF = 71.4%). Additionally, a la rge number of parents repor ted feeling distressed when accommodating their childs symptoms (UF = 55.3%, USF = 76.2%), which suggests that many parents accommodate reluctantly. Over half of parents reported their childs anxiety increasing when accommodation was not provided (UF = 56.8%, USF = 66.7%). Some parents also endorsed opp ositional/defiant behavior (U F = 50.0%, USF = 33.3%), angry/abusive behavior (UF = 39.5%, USF = 42.9%), and sad/down behavior (UF = 47.4%, USF = 66.7%) when accommodation was not provided. Endor sement of these items demonstrates that child emotional reac tion or behavior is one factor that could maintain parent accommodation of pediat ric anxiety symptoms. Least commonly endorsed forms of accommodation were changi ng parent work schedules (UF = 36.8%, USF = 19.0%) and changing the childs social activities (UF = 13.2%, USF = 38.1%). However, some infrequent forms of acco mmodation may have high potential for

PAGE 39

39 negative consequences (e.g., parents missing work). Therefore, it is important to continue assessing these behaviors in the cont ext of accommodation. Overall, these percentages suggest that a large number of parents accommodate anx iety symptoms in this population of patients. In addition to the high number of par ents reporting accommodation, results indicated that family accommodation is posit ively related to both parent-rated functional impairment (UF site) and to clinician-, par ent-, and child-rated measures of anxiety symptoms (USF site). These results s uggest that the role of accommodation in pediatric anxiety may be similar to its ro le in an OCD populatio n and that family accommodation may have eventual implications for treatment. However, these results should be interpreted as prelimin ary given that the present study is complicated by site differences and does not provide information about causality in this relationship (i.e. whether accommodation leads to increas ed severity/decreased functioning). Exploratory analyses of the PAS with measures of parent psychopathology revealed that family accommodation was strongly related to both parent symptoms of depression and anxiety (USF site). While previous studies have shown inconsistent findings regarding the relationship of par ent psychopathology to child anxiety symptoms, our results suggest that parents with higher leve ls of symptoms engage in more accommodation of child symptoms. As well, exploratory analyses of the PAS with measures of family functioning revealed an a ssociation with family flexibility (USF site), indicating that higher frequency of accommodati on is associated with lower levels of flexibility. As well, the relationship of the PAS to family cohesion suggested that frequent accommodation is related to low levels of family cohesion, but this relationship

PAGE 40

40 only approached statistical significance (USF site ). This finding is co ntrary to previous reports suggesting high cohesion is linked to higher symptom severity (Poleg-Popko & Dar, 2001) but supports treatment research suggesting that high cohesion is related to improved outcomes (Victor et al., 2007). Po sitive relationships with other family constructs in our sample (i.e. Enmeshment and Chaos) were large but did not achieve statistical significance. Future investigat ions of the PAS should explore relationships with family constructs in a larger sample. However, these results suggest that families that are less flexible and cohesiv e may need additional support to target accommodation that could interfere with CBT treatment. Additionally, families in which parents are experiencing symptoms of anxiety and depression may also need targeted intervention to facilitate reduced accommodation. Finally, the present investigation aimed to test several mediation models that were moderated by site. We hypothesized that family accommodation mediates the relationship between parent psychopathology and child anxiety symptoms. However, results did not support mediation models bas ed on either parent anxiety or depressive symptoms. We also hypothesized that fam ily accommodation mediates the relationship between symptom severity and functional impai rment, similar to the role of family accommodation in OCD (Storch et al., 2007b). However, results did not support the mediation model for either parent or child reported functional impairment. It may be that family accommodation does not have the same relationship to symptom severity and functional impairment in this anxiety popul ation compared with t he mediation effect found for patients with OCD. One possible explanation is that patients with OCD may have very visible symptoms (i .e. compulsions) that both contribute to impairment and

PAGE 41

41 draw accommodation. However, it is also likely that the present investigation was underpowered to detect mediation effects. It will be important to test family accommodation as a possible mediator in thes e relationships in future investigations with a larger sample. Although this study offers promising pre liminary data for the development of the PAS, its limitations should be noted. As noted previously, the sample size limits our ability to draw conclusions from the data. As well, the site differences related to clinical interviewer ratings limited our ability to generalize results across settings and further reduced sample size given the need to conduc t many analyses separately. However, given that our site differences may have been re lated to clinical interviewer experience, direct examination of interviewer experience as a variable that influences the validity of this and other clinician-administered scales may be warranted. Finally, several items on the PAS showed low item-total correlations, per haps due to some clinical interviewers incorporating severity into ratings. The PAS may benefit from inclusion of a severity scale for this reason, and a severity scale may also contribute to overall understanding of the family accommodation construct and its relationship to other symptom and treatment variables. Overall, the present study is the first to investigate family accommodation in a sample of children with GAD, SP, and SAD. Results show that a large percentage of parents provide accommodation, particularly when providing reassurance and allowing children to avoid anxiety-provoking st imuli. The PAS shows good preliminary psychometric properties, including internal c onsistency, inter-rater reliability, and some evidence of convergent validity. However, differences related to clinical interviewer

PAGE 42

42 ratings and inability to rate severity of accommodation suggest that the PAS should be revised and re-administered in a second sa mple. Beyond re-administration of the PAS in a second sample, future studies should ex amine the role of family accommodation in CBT, including its ability to predict treat ment outcome and whether addressing family accommodation in treatment is a mechanism of change. The current study may also have implications for clinicians and resear chers wishing to identify families who are likely to accommodate prior to beginning treatment, as low family cohesion and flexibility are related to high levels of accommodation. Studies should contin ue to examine family variables related to accommodation in order to aid in identificatio n of these families.

PAGE 43

43 APPENDIX A PEDIATRIC ACCOMMODATION SCALE Pediatric Accommo dation Scale (PAS) Note: This measure is designed to be administered by a clinician or trained interviewer. Responses should be chosen based on clinici an judgment of responses. Each item contains examples that may be substituted wit h a childs specific symptoms. After each item, the clinician may ask follow-up questions to clarify or to probe more closely for accommodation. Clinician States: This questionnaire asks about diffe rent ways that you have dealt with your childs anxiety, such as reassuring y our child when he/she is anxious, rearranging your childs schedule, and allowing your child to avoid certain things. These questions are asking about things you have done during the past week so please answer keeping only the past week in mind. In the past week, how often 1. Did you reassure your child about his/her fears, for example, f ears that something bad would happen him/her, something bad would happen to you, or fear that others may be thinking bad things about him/her? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 2. Did you help your child to avoid things or situations that mi ght make him/her more

PAGE 44

44 anxious such as talking to others, separ ating from you, or watching the news? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 3. Have you avoided doing things such as going out to a social event or being with people because your child might get nervous or scared? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 4. Have you modified your routine in any way to reduce your childs anxiety, for example, spending less time with other people or getting up early to send your child to school because of your childs symptoms? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 5. Have you modified your work schedule because of your childs symptoms, such as going to work late because your child refuse s to go to school or taking days off because your child is frequently feeling sick? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 6. Have you modified your social leisure ac tivities or hobbies because of your childs symptoms, for example, staying home when he/she is distressed or spending less time with friends because he/she is worried? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 7. Have you modified your childs responsibili ties because of his/her symptoms, such as

PAGE 45

45 completing chores for him/her or reduc ing difficulty of his/her schoolwork? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 8. Have you modified your childs school sc hedule because of his/her symptoms, such as allowing him/her to miss school, go to school late, or leave school early because he/she is worried or frequently ill? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 9. Have you changed your childs social or leisure activities as a result of his/her symptoms, such as picking your child up early from peer gatherings or withdrawing him/her from sports teams of other social organizations? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 10. Have you changed your childs sleep or bedtime routine because of his/her symptoms, such as checking under beds or in cl osets or allowing your child to sleep in a room that is not his/her bedroom? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 11. Have you modified the schedule of other family members, such as siblings, because of your childs symptoms, such as decreasing fa mily members activities or spending less time with other family members? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

PAGE 46

46 12. Did you become upset, stressed, or di stressed when you help your child cope with anxiety in unhealthy ways (e.g., allowing hi m/her to miss school, rearranging your schedule so your child can be with you, or constantly providing reassurance)? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 13. Has your child become more distress ed or anxious when you have not provided assistance during times of anxiety, such as pr oviding reassurance or allowing him/her to avoid social activities or school? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 14. Has your child become oppositional or defiant when you have not provided assistance, such as refusing to obey or throwing a tantrum? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 15. Has your child become angry or abusiv e when you have not provided assistance, such as saying hurtful things, being destruc tive, or physically hurting someone? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day 16. Has your child become sad or down when you have not provided assistance, such as becoming tearful or withdrawi ng from you or others? 0 = Never 3 = 4-6 times per w eek 6 = 4-6 times per day 1 = 1 time per week 4 = 1 time per day 7 = 7-10 times per day 2 = 2-3 times per week 5 = 2-3 times per day 8 = 11+ times per day

PAGE 47

47 LIST OF REFERENCES Achenbach, T. M. (1991). Manual for the Child Behavio r Checklist / 4 and 1991 profile. Burlington, VT: University of Ve rmont, Departme nt of Psychiatry. Achenbach, T. M., Howell, C. T., McConaughy, S. H., & St anger, C. (1995). Six-year predictors of problems in a national sa mple of children and youth: I. CrossInformant syndromes. Journal of the American Academy of Child and Adolescent Psychiatry, 34 336 347.; Achenbach, T. M., Howell, C. T., McConaughy, S. H., & St anger, C. (1995). Six-year predictors of problems in a national sa mple of children and youth: I. Cross informant syndromes. Journal of the American Academy of Child and Adolescent Psychiatry, 34 336 347. American Psychiatric Association (2000). Diagnostic and statistic manual of mental disorders (4th ed. text revision). Washington, DC: Author. Amir, N., Freshman, M., & Foa, E. (2000). Family distress and involvement in relatives of obsessive-compulsive disorder patients. Journal of Anxiety Disorders, 14 209 217. Barrett, P. M. (1998), Group therapy for childhood anxiety disorders. Journal of Clinical Child Psychology, 27 1 468. Barrett, P. M., Dadds, M. R., Rapee, R. M. (1996). Family treatment of childhood anxiety: a controlled trial. Journal of Consulting and Clinical Psychology, 64 333 342. Beck, A. T., Steer, R. A., Brown, G.K. (1996). Beck Depression Inventory Second Edition manual. San Antonio, TX: The P sychological Corporation. Beck, A. T., Ward, C. H., M endelson, M., Mock, J., Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry 4, 561571. Bernstein, D. C., & Borchar dt, C. M. (1991). Anxiet y disorders of childhood and adolescence: A critical review. Journal of the American Academy of Child and Adolescent Psychiatry, 30 519 532. Bieling, P. J., Antony, M. M. Swinson, R. P. (1998). The State-Trait Anxiety Inventory, Trait version: structure and content re-examined. Behavioral Research and Therapy, 36, 777 788. Bolton D., Collins S., Steinberg, D. (1983). The treatment of obsessive compulsive disorder in adolescence: A report of fifteen cases. British Journal of Psychiatry 142 456 64.

PAGE 48

48 Calvocoressi, L., Lewis, B., Harris, M., Goodm an, W. K., Trufan, S. J, McDougle, C. J., Price, L. H. (1995). Family accommodation in obsessive compulsive disorder. American Journal of Psychiatry, 152, 441 443. Calvocoressi, L., Mazure, C. M. Kasl, S., Skolnick, J., Fisk, D., Vegso, S. J., et al. (1999). Family accommodation of obsessive-compulsive symptoms: Instrument development and assessment of family behavior. Journal of Nervous and Mental Disease, 187 636 642. Cartwright-Hatton, S., Roberts C., Chitsabesan, P., Fothergi ll, C., and Harrington, R. (2004). Systematic review of the effi cacy of cognitive behavior therapies for childhood and adolescent anxiety disorders. British Journal of Clinical Psychology, 43 421 436. Cartwright-Hatton, S., Roberts C., Chitsabesan, P., Fothergi ll, C., and Harrington, R. (2004). Systematic review of the effi cacy of cognitive behavior therapies for childhood and adolescent anxiety disorders. British Journal of Clinical Psychology, 43 421 436. Chorpita, B. F., Barlow, D. H. (1998). The dev elopment of anxiety: the role of control in the early environment. Psychological Bulletin, 124 3 21. Cobham, V. E., Dadds, M. R., Spence, S. H. (1998). The role of parental anxiety in the treatment of childhood anxiety. Journal of Consulting and Clinical Psychology, 66, 893 905. Craighead, W. E., Smucker, M. R., Craighead, L. W., & Ilardi, S. S. (1998). Factor analysis of the Childrens Depression Inventory in a community sample. Psychological Assessment, 10, 156165. Crawford, M. & Manassis, K. (2001). Familial predictor s of treatment outcome in childhood anxiety disorders, Journal of the American Academy of Child & Adolescent Psychiatry 40 1182. Dadds, M. R., Barrett, P. M., Rapee, R. M. & Ryan, S. (1996). Family process and child anxiety and aggression: an observational analysis, Journal of Abnormal Child Psychology 24, 715. Faul, F., & Erdfelder, E. (2007). GPOWER: A Priori, post-hoc, and compromise power analyses for MS-DOS (Computer Program). Bonn, FRG: Bpnn University, Department of Psychology. Ferrao, Y. A., Shavitt, R. G., Bedin, N. R., de Mathis, M. E., Lopes, A. C., Fontenelle, L. F., et al. (2006). Clinical features asso ciated to refractory obsessive-compulsive disorder. Journal of Affective Disorders, 94 199 209.

PAGE 49

49 Geffken, G. R., Storch, E. A., Duke, D. C., Monaco, L., Lewin, A. B ., & Goodman, W. K. (2006). Hope and coping in family me mbers of patients with obsessivecompulsive disorder. Journal of Anxiety Disorders, 20 614 629. Ginsburg, G. S., Riddle, M. A., & Davies, M. (2006). Somatic symptoms in children and adolescents with anxiety disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 45 1179 1187. Gosch, E. A., Flannery-Schroede r, E., Mauro, C. F., & Compton, S. C. (2006). Principles of cognitive-behavioral therapy for anxiety disorders in children. Journal of Cognitive Psychotherapy: An International Quarterly, 20 247 262. Hudson, J. L., & Rapee, R. M. (2001), Parent-child interact ions and anxiety disorders: an observational study. Behavioral Research Therapy, 39 1411 1427. Kendall, P. C. (1994). Treating anxiety diso rders in children: Results of a randomized clinical trial. Journal of Consulting and Clinical Psychology, 62 100 110. Kendall, P. C. (2000). Coping Cat Workbook. Ardmore, PA: Workbook Publishing. Kendall, P. C., Kane, M., Howard, B., & Siqueland, L. (1990). Cognitive-Behavioral Treatment of Anxious Ch ildren: Treatment Manual Ardmore, PA: Workbook Publishing. Kingery, J. N., Roblek, T. L., Suveg, C., Grov er, R. L., Sherrill, J. T., & Bergman, R. L. (2006). Theyre not just little adults : Developmental considerations for implementing cognitive-behavioral therapy with anxious youth. Journal of Cognitive Psychotherapy: An In ternational Quarterly, 20 263 273. Kovacs, M. (1992). The Childrens D epression Inventory Manual. Toronto, Ontario, Canada: Multi-Health Systems, Inc. Langley, A., Bergman, L., McCracken, J., & Piacentini, J. (2004). Impairment in childhood anxiety disorders: Pr eliminary examination of the Child Anxiety Impact Scale-Parent Version. Journal of Child and Adole scent Psychopharmacology, 14, 105 114. Leane, M. (1991). Family therapy for childr en with obsessive-compulsive disorder. In MT Fato, J Zohak (Eds), Current treatments of obsessive-compulsive disorder. Washington, DC: American Psychiatric Press. March, J. S. (1995). Cogn itive-behavioral psychotherapy for children and adolescents with OCD: A review and recommendations for treatment. Journal of the American Academy of Child and Adolescent Psychiatry, 34 7 18. March, J. S., Sullivan, K., Parker, J. D. (1999). Test-retest reliability of the Multidimen sional Anxiety Scale for Children. Journal of Anxiety Disorders, 13, 349 358.

PAGE 50

50 Mash, E. J., & Johnston, C. (1990). Dete rminants of parenting stress: illustrations from families of hyperactive children and families of physically abused children, Journal of Clinical Child Psychology 19, 313. McClure, E. B., Brennan, P.A., Hammen, C. & Le Brocque, B. C. (2001). Parental anxiety disorders, child anxiety diso rders, and the perceived parentchild relationship in an Australian high-risk sample, Journal of Abnormal Child Psychology 29, 1. McLeod, B. D., Wood, J. J., & Weisz, J. R. (2007). Examining the association between parenting and child anxiet y: A meta-analysis. Clinical Psychology Review, 27 155 172. Mendlowitz, S. L., Manassis, K., Bradley, S., Scapillato, D., Mietzitis, S., & Shaw, B. (1999). Cognitive-behavioral group treatments in childhood anxiety disorders: the role of parental involvement. Journal of the American Academy of Child and Adolescent, Psychiatry 38, 1223 1229. National Institute of Mental Health (1985). Rating scales and assessment instruments for use in pediatric p sychopharmacology research. Psychopharmacological Bulletin, 21, 839. Nauta, M. H., Scholing, A., Emmelkamp, & P. M. G., Minder aa, R. B. (2003). Cognitivebehavioral therapy for children with anxiety disorders in a clinical setting: no additional effect of a c ognitive parent-training. Journal of the American Academy of Child and Adolescent Psychiatry, 42 1270 1278. Olson, D. H., Portner, J. & Bell, R. Q. (1982). FACES II: Family Adaptability and Cohesion Evaluation Scales Minnesota: Family Social Science, University of Minnesota. Peleg-Popko, O. & Dar, R. (2001). Marital quality, family patterns, and children's fears and social anxiety, Contemporary Family Therapy 23, 465. Place, M., Hulsmeier, J., Brow nrigg, A., & Soulsby, A. (2005). The Family Adaptability and Cohesion Evaluation Scale (FACES): An instrument worthy of rehabilitation? Psychiatric Bulletin, 29, 215 218. Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressi ng moderated mediation hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research, 42 185 227. Research Units on Pediatric Pharmacology An xiety Study Group, ( 2002). The Pediatric Anxiety Rating Scale (PARS): Development and psychometric properties. Journal of the American Academy of Child and Adolescent Psychiatry, 41 10611069.

PAGE 51

51 Ryn, M. A., Barber, J. P., Sarosh, K., Siquel and, L., Dembiski, M., McCarthy, K. S., et al. (2006). The psychometric properties of the MASC in a pediatric sample. Anxiety Disorders, 20, 139 157. Scahill, L., Riddle, M. A., McS wiggin-Hardin, M., Ort, S. I., King, R. A ., Goodman, W. K., Cicchetti, D., Leckman, J. F. (1997). Childrens Yale-Brown Obsessive Compulsive Scale: Reliability and validity. Journal of the Amer ican Academy of Child and Adolescent Psychiatry, 36, 844852 Shafran, R., Ralph, J., & Tallis, F. (1995). Obsessive-c ompulsive symptoms and the family. Bulletin of the Menninger Clinic, 59 472 479. Spence, S., Donovan, C., Brechman-Toussain t, M. (2000). The tr eatment of childhood social phobia: The effectiveness of a social skills training-based, cognitivebehavioral intervention, with and without parental involvement. Journal of Child Psychology and Psychiatry and Allied Disciplines, 41 713 726. Spielberger, C. D. (1983). Manual for the State-Trait Anxiet y Inventory STAI (Form Y). Palo Alto, CA: Mind Garden. Spielberger, C. D. (1989). State-Trait Anxiety Inventory: a comprehensive bibliography. Palo Alto, CA: Consulting Psychologists Press. Storch, E. A., Geffken, G. R. Merlo, L., Jacob, M., Murp hy, T. K., Goodman, W. K., et al. (2007b). Family accommodation in obs essive-compulsive disorder. Journal of Clinical Child and Adolescent Psychology, 36 207 216. Storch, E. A., Geffken, G. R. Merlo, L. J., Mann, G., Du ke, D., Munson, M., et al. (2007a). Family-based cognitive-behavio ral therapy for pediatric obsessivecompulsive disorder: Comparison of intensive and weekly approaches. Journal of the American Academy of Child Psychiatry, 46, 469 478. Velting, O. N., Setzer, N. J., & Albano, A. M. (2004). Update on and advances in assessment and cognitive-behavioral treatment of anxiety disorders in children and adolescents. Professional Psychology: Research and Practice, 35 42 54. Verdeli, H., Mufson, L., Lee, L., & Keith, J. A. (2006) Review of evidence-based psychotherapies for pediatric m ood and anxiety disorders. Current Psychiatry Reviews, 2 395 421. Victor, A. M., Bernat, D. H., Bernstein, G. A., & Layne, A. E. (2007). Effects of parent and family characteristics on treatment outcome of anxious children. Journal of Anxiety Disorders, 21, 835 848. Whaley, S. E., Pinto, A., Sigman, M. ( 1999). Characterizing interactions between anxious mothers and their children. Journal of Consulting and Clinical Psychology, 67 826 836.

PAGE 52

52 Whisman, M. A., Perez, J. E., and Ramel, W. (2000). Factor st ructure of the Beck Depression InventorySecond edition (B DI-II) in a student sample. Journal of Clinical Psychology, 56 545 551. Wood, J. J., McLeod, B. D., Sigman, M., Hwang, W., Chu, B. C. (2003). Parenting behavior and childhood anxiety: theory, empiri cal findings, and future directions. Journal of Child Psychology and Psychiatry, 44 134 151 Wood, J. J., Piacentini, J. C., Southam-Gerow, M., Chu, B., & Sigman, M. (2006). Family cognitive behavioral therapy for child anxiety disorders. Journal of the American Academy of Child Adolescent Psychiatry, 45 314 321.

PAGE 53

53 BIOGRAPHICAL SKETCH Kristen Gra bill graduated magna cum laude with a Bachelor of Arts in psychology from Saint Marys College of Maryland in 2003. Following that, she completed an intramural post-baccalaureate fellowship at the National Institutes of Health in Bethesda, MD. She completed her Predoctoral Internship at Brown University in Providence, RI and graduated from the University of Florida in August 2010.