Transcript View Poster - pantherFILE - University of Wisconsin–Milwaukee

Confirmatory Factor Analysis of the SCARED
in a Pediatric Chronic Pain Sample:
Implications for Use
[email protected]
Jenny R. Evans, M.A.,1 Kristen Jastrowski Mano, Ph.D.,1 Susan T. Tran, M.S.,2 Kimberly Anderson Khan, Psy.D.,3,4 Steven J. Weisman, M.D.,3,4 and Keri R. Hainsworth, Ph.D.3
1Alliant International University, San Diego, CA, 2University of Wisconsin-Milwaukee, 3Medical College of Wisconsin, 4Children’s Hospital of Wisconsin
Abstract
Purpose
The psychometric properties of the Screen for Child Anxiety
Related Emotional Disorders (SCARED; Birmaher et al., 1997,
1999) were examined in a pediatric pain sample. Higher
SCARED scores were associated with increased pain
catastrophizing and decreased quality of life. The 5 factor
model proposed by Birmaher et al. (1999) evidenced
acceptable fit. Questions remain regarding the adequacy of
the SCARED in a pediatric pain setting, particularly in the
assessment of school anxiety.
The present study examined the psychometric properties of
the SCARED in order to begin establishing an evidence base
for using the SCARED in pediatric chronic pain.
PAN
.93
(.96)
.97
(.88)
Participants
.89
(.99)
1.03
(1.0)
SAD
.96
(1.01)
SOC
Although anxiety symptoms are often reported by children
with chronic pain, anxiety is not routinely assessed. To date,
no pediatric anxiety measures have been specifically
developed or normed for use in pediatric chronic pain. The
SCARED is a commonly used measure and has been
validated for use with children presenting for anxiety
treatment in an outpatient clinical setting (Birmaher 1999;
Hale 2011). The degree to which the SCARED adequately
assesses anxiety in pediatric pain remains unknown. Of
particular relevance to this population is its effectiveness in
assessing school anxiety, which is often reported by youth
with chronic pain (Ladwig & Khan, 2007).
Methods
.98
(.99)
GAD
Introduction
.93
(.80)
.87
(.75)
.94
(.78)
.91
(.82)
SCH
Figure 1. Confirmatory factor analyses for the SCARED
model proposed by Birmaher et al. (1999) in a pediatric pain
sample with factor loadings for mother-report indicated
parenthetically. Subscale names are: PAN (Panic Disorder),
GAD (Generalized Anxiety Disorder), SAD (Separation
Anxiety Disorder), SOC (Social Anxiety), SCH (School Phobia).
Participants were 349 treatment seeking youth presenting
to an outpatient multidisciplinary pain program and their
parents (311 mothers and 162 fathers). Youth (69% female;
79% Caucasian) varied in age (range = 8-18 years, M = 14.21,
SD = 2.54) and medical diagnosis (37% headaches, 16%
abdominal pain, 15% back pain, 15% lower extremity pain).
Measures
Participants and their parents completed the SCARED,
measures of quality of life (PedsQL) and pain catastrophizing
(PCS-C). The authors of the SCARED propose a 5-factor
model: Panic/Somatic Symptoms, Generalized Anxiety,
Separation Anxiety, Social Anxiety, and School Phobia.
Procedure
Questionnaire packets were mailed to families as part of the
standard clinic intake procedure. Parents and youth
completed the questionnaires individually and returned
them prior to their pain clinic evaluation.
Results
As predicted, SCARED scores from both youth and parents
were positively related to Pain Catastrophizing and inversely
related to Quality of Life scores (p-values < .01; see Table 1
for SCARED descriptives).
Table 1. SCARED Descriptive Statistics
Subscale
Total SCARED
Panic Disorder
Generalized Anxiety Disorder
Separation Anxiety
Social Anxiety
School Phobia
Male
(n = 110)
M (SD)
16.99 (12.42)*
3.18 (3.68)*
4.33 (3.92)*
2.47 (3.24)
4.12 (3.63)
2.89 (1.67)
Female
(n = 239)
M (SD)
21.38 (14.27)*
4.69 (4.67)*
5.95 (4.56)*
3.26 (3.39)
4.41 (3.74)
3.07 (1.96)
Child Total
(N = 349)
M (SD)
19.99 (13.85)
4.21 (4.43)
5.44 (4.43)
3.01 (3.36)
4.32 (3.71)
3.01 (1.87)
Mother Total
(n = 313)
M (SD)
15.47 (12.42)
2.42 (3.53)
4.92 (4.36)
2.07 (2.92)
3.33 (3.44)
2.75 (1.92)
Father Total
(n = 163)
M (SD)
13.96 (10.60)
2.15 (3.12)
4.85 (3.77)
1.69 (2.33)
2.95 (3.37)
2.28 (1.78)
Note. SCARED = Screen for Child Anxiety Related Emotional Disorders; SD = standard deviation.
*Significant differences between male and female youth on Panic Disorder, Generalized Anxiety Disorder, and Separation Anxiety Disorder
subscales, p < .01. All other differences were non-significant (p > .05).
Internal consistency (Cronbach’s α) of SCARED Total scores
was excellent (.92 or higher). All subscales except for School
Avoidance (α = .59 - .62) exhibited acceptable internal
consistency (α = .77 or higher; see Table 2 for complete
total and subscale values).
A CFA of the 5-factor SCARED measurement model was
conducted using a maximum-likelihood estimation
procedure in LISREL 8.80 (see Figure 1). Youth- and motherreport were analyzed separately. Due to insufficient sample
size, the model was not examined using father-report data.
For youth, the model provided acceptable fit to the
observed data, χ2 = 1148.07, df = 769 (χ2/df = 1.49), p < .05,
CFI = 0.992, RMSEA = 0.042; GFI = 0.836; NNFI = 0.992; AIC
= 1332.068; ECVI = 4.674. Likewise, for mother-report the
model demonstrated acceptable fit, χ2 = 1296.95, df = 769
(χ2 /df = 1.68), p < .05, CFI = 0.982, RMSEA = 0.047; GFI =
0.829; NNFI = 0.981; AIC = 1480.955; ECVI = 4.808.
Discussion
Conclusion
The results of this study provide mixed evidence regarding
the appropriateness of the SCARED for assessing anxiety in
youth with pediatric chronic pain. The model evidenced
acceptable fit, however of particular concern is the
measurement of school avoidance symptoms, calling into
question the adequacy of the SCARED in pediatric pain. Our
findings are consistent with previous research
demonstrating the presence of anxiety in pediatric chronic
pain, as well as the associations among anxiety, pain
catastrophizing and HRQOL.
Table 2. Internal Consistency of SCARED Subscales
Child
Mother
Father
(N = 349) (n = 313) (n = 163)
Subscale
α
α
α
Total SCARED
0.93
0.93
0.92
Panic Disorder
0.86
0.86
0.85
Generalized Anxiety
0.87
0.86
0.85
Separation Anxiety
0.85
0.84
0.77
Social Anxiety
0.87
0.89
0.90
School Phobia
0.59
0.59
0.62
Note. α = Cronbach’s alpha internal consistency coefficient.
Future Directions
Given the dearth of previous research in this area, further
investigation is warranted to confirm these findings and
better understand the presentation of anxiety symptoms in
pediatric chronic pain. Considering the prevalence of school
anxiety and absenteeism among youth with chronic pain, it
is important to expand existing measures or develop new
measures that accurately and reliably assess school anxiety
in this population.
It may also be important to reconsider the current DSMbased conceptualization of anxiety and examine ways in
which the construct of anxiety may be uniquely experienced
by youth with chronic pain. Perhaps physiological measures,
in conjunction with self-report may provide additional
insight on the impact of anxiety on patient functioning and
treatment outcomes.