Therapeutic Advances in Musculoskeletal Disease Original Research

Ther Adv Musculoskel Dis

Simplifying fibromyalgia assessment: (2011) 3(5) 215226
DOI: 10.1177/
the VASFIQ Brief Symptom Scale 1759720X11416863
! The Author(s), 2011.
Reprints and permissions:
Chad S. Boomershine, Birol Emir, Yi Wang and Gergana Zlateva http://www.sagepub.co.uk/
journalsPermissions.nav

Abstract:
Objectives: We tested the ability of the VASFIQ, a seven-item scale composed of Fibromyalgia
Impact Questionnaire (FIQ) visual analog scales (VASs), to quantify fibromyalgia global disease
severity and identify fibromyalgia patients with significant symptoms of fatigue, poor sleep,
depression or anxiety.
Methods: Spearman rank correlations were used to compare global VASFIQ, FIQ and Patient
Global Impression of Change (PGIC) scores and individual FIQ VAS scores with full-length,
validated questionnaire scores for fatigue (Multidimensional Assessment of FatigueGlobal
Fatigue Index [MAF-GFI]), poor sleep (Medical Outcomes Study Sleep Problems Index [SPI])
and depression and anxiety (Hospital Anxiety and Depression Scale [HADS]). Patient scores
used in the analyses were derived from 2229 patients enrolled in three pregabalin fibromyalgia
trials. Receiver operating characteristic analyses determined VASFIQ cutoff scores identifying
patients with clinically significant symptom levels using full-length, validated symptom ques-
tionnaires to define cases.
Results: Global VASFIQ and FIQ scores correlated highly at baseline and study endpoints
(r 0.94 and 0.97, respectively; both p < 0.0001). Change in global VASFIQ and FIQ scores
correlated similarly to PGIC scores at study endpoints (r 0.58 and 0.61, respectively; both
p < 0.0001). Individual FIQ VAS scores correlated with corresponding full-length symptom
questionnaire scores at baseline and study endpoints (VASfatigue with MAF-GFI, r 0.64 and
0.76; VASsleep with SPI, r 0.50 and 0.67; VASdepression with HADS-D, r 0.43 and 0.62;
VASanxiety with HADS-A, r 0.47 and 0.67, respectively; p < 0.0001 for all). Patients with
significant symptoms of fatigue were identified by VASfatigue >7.5, poor sleep by VASsleep
>7.9, depression by VASdepression >5.8 and anxiety by VASanxiety >6.0. VASFIQ global scores
31.4 and 45.0 identified patients with moderate and severe global fibromyalgia symptoms,
respectively.
Conclusions: The VASFIQ scale accurately quantifies global fibromyalgia severity and identifies
patients with significant symptoms of fatigue, poor sleep, depression or anxiety with brevity,
enabling rapid patient assessment and informing treatment decisions in busy clinics.

Keywords: anxiety, depression, fatigue, fibromyalgia, pain, poor sleep, screening, symptoms
Correspondence to:
Chad S. Boomershine,
Introduction symptoms is required for diagnosis under new MD, PhD
Department of Medicine,
Fibromyalgia (FM) is a complex disorder of preliminary diagnostic criteria for FM recom- Vanderbilt University,
chronic widespread pain and tenderness associ- mended by the American College of T3219 MCN, 1161 21st
Avenue South, Nashville,
ated with numerous other symptoms including Rheumatology (ACR) [Wolfe et al. 2010], and TN 37232, USA
fatigue, poor sleep quality, depression and anxi- management of all clinically significant symp- chad.boomershine@
vanderbilt.edu
ety [Wolfe, 1990]. The complexity of FM can toms is recommended for effective FM manage-
Birol Emir, PhD
make patients challenging to treat. While all ment [Carville et al. 2008]. However, assessment Gergana Zlateva, PhD
patients with FM experience pain, individual is often hampered by the inability of many Pfizer Inc, New York, NY,
USA
patients often vary widely in the severity of asso- patients to effectively articulate their symptom
Yi Wang, MS
ciated FM symptoms from which they suffer. burden and the inability of clinicians to interpret Columbia University,
Quantification of the severity of associated FM the patients complaints into a coherent New York, NY, USA

http://tab.sagepub.com 215
Therapeutic Advances in Musculoskeletal Disease 3 (5)
intellectual framework from which to make
diagnostic and treatment decisions. While self-
report questionnaires to quantify symptom
severity exist, including the Multidimensional
Assessment of Fatigue (MAF) for fatigue
[Belza, 1995], the Medical Outcomes
StudySleep Scale (MOSSleep) for sleep qual-
ity [Hays and Stewart, 1992] and the Hospital
Anxiety and Depression Scale (HADS) for anxi-
ety and depression [Zigmond and Snaith, 1983],
their length and complexity typically preclude use
in busy clinical practices.
The Fibromyalgia Impact Questionnaire (FIQ) is
the standard patient self-report assessment for
quantifying the global severity of FM symptoms
in research studies [Burckhardt et al. 1991]. The
FIQ is composed of 20 items that assess disease
severity over the past week with 11 questions to
assess physical functioning, two day-of-the-
week items to quantify the number of days
patients felt good or missed work, and seven
visual analog scales (VASs) to assess symptoms of
fatigue, sleep quality, depression, anxiety, stiff-
ness, pain and work disability. However, clinical
utility of the FIQ is limited by its length, the scor-
ing complexity of physical functioning and day-
of-the-week items, and the lack of functionality
to identify patients with clinically significant
symptom levels that require treatment. While
clinical utility of the entire FIQ is limited, the
FIQ VASs have properties that make them ideal
for clinical use. VASs are simple to score and have
been shown to perform as well as longer scales in
respect to sensitivity to change and correlation
with clinical variables [Wolfe, 2004]. In addition,
the FIQ VASs evaluate domains that have been
identified as important in assessing patients with
FM [Mease et al. 2009]. Finally, prior research in
a small FM cohort showed the FIQ VASs per-
formed as well as the full FIQ in assessing
global disease severity, and cutoff scores on indi-
vidual FIQ VASs could be established to identify
patients with significant symptoms of fatigue,
poor sleep and depression [Boomershine et al.
2008].
In this work, we challenged the hypothesis that a
scale composed of the seven FIQ VASs (the
VASFIQ, Figure 1) could provide a brief, clini-
cally useful FM assessment measure, with VAS
scores summed to provide a measure of global
disease severity and cutoff scores developed for
individual VASs that identify patients with
216
clinically significant symptoms of fatigue, poor
sleep quality, depression and anxiety.
Patients and methods
Combined data from 2229 patients enrolled in
three pregabalin (Lyrica ) FM treatment trials
(A0081056 [ClinicalTrials.gov identifier:
NCT00645398], A0081077 [ClinicalTrials.gov
identifier: NCT00230776] and A0081100
[ClinicalTrials.gov identifier: NCT00333866])
[Arnold et al. 2008; Mease et al. 2008b; Pauer
et al. 2008] were analyzed. All trials had similar
inclusion and exclusion criteria. Patients were
included if they were  18 years of age, had a
baseline pain score of at least 40 mm on a 100-
mm pain VAS and fulfilled ACR classification
criteria for FM (i.e. widespread pain present for
3 months and pain on palpation at 11 of 18
specific tender point sites) [Wolfe et al., 1990].
Exclusion criteria included evidence of inflam-
matory rheumatic disease, severe painful disor-
ders other than FM, clinically significant or
unstable medical or psychiatric disorders, history
of illicit drug or alcohol abuse within the past 2
years, previous pregabalin treatment and receiv-
ing or applying for disability benefits. Patients
were also required to discontinue use of other
concomitant medications taken for FM as well
as agents used to treat pain and insomnia prior
to entering the trials. Each trial complied with
Declaration of Helsinki and International
Conference on Harmonization Good Clinical
Practice Guidelines, was approved by local
research ethics committees, and written informed
consent was obtained from all patients.
Questionnaire data analyzed from the trials
included baseline and endpoint scores for the
FIQ, HADS, MOS-Sleep, MAF and Patient
Global Impression of Change (PGIC).
Endpoints scores were determined using last
observation carried forward. The MAF measures
fatigue and the effect of fatigue on activities using
16 items resulting in a Global Fatigue Index
(MAF-GFI) that can range from 1 to 50, with
higher scores indicating greater fatigue severity
[Belza, 1995]. The MAF-GFI has been shown
a valid fatigue measure in a variety of patient pop-
ulations with chronic conditions [Neuberger,
2003]. Since there are no set cut-off scores on
the MAF to define clinically significant fatigue,
a MAF-GFI score of 30 was used to define clin-
ically significant fatigue since this score is one
standard deviation (SD) above the mean seen in
healthy controls [Belza, 1995]. The MOS-Sleep
http://tab.sagepub.com
 CS Boomershine, B Emir et al.

(for internal
1. VASwork: When you worked how much did pain or other symptoms of your fibromyalgia use only)
interfere with your ability to do your work, including housework?
No problem Great difficulty
with work with work
Score
2. VASpain: How bad has your pain been?
Very severe
No pain
pain
Score
3. VASfatigue: How tired have you been?

No tiredness Very tired

4. VASsleep: How have you felt when you get up in the morning? Score

Awoke well Awoke very

rested tired
Score
5. VASstiff: How bad has your stiffness been?

No stiffness Very stiff

6. VASanxiety: How nervous or anxious have you felt? Score

Not anxious Very anxious

Score
7. VASdepression: How depressed or blue have you felt?

Not depressed Very depressed

Score

Figure 1. The VASFIQ Questionnaire. FIQ, Fibromyalgia Impact Questionnaire; VAS, Visual Analogue Scale. FIQ VASs reproduced with
permission from R.M. Bennett and figure reproduced with permission from The Journal of Rheumatology! 1991 [Burckhardt et al.
1991].

comprises 12 items that measure sleep parame-
ters across six domains yielding a composite
Sleep Problems Index II (SPI) score that can
range from 0 to 100, with higher scores indicating
worse sleep problems [Hays and Stewart, 1992].
The MOS-Sleep has proven relevance for the
impact of FM on sleep quality in FM patients
[Martin et al. 2009] and is considered the best
questionnaire for assessing sleep disturbance in
patients with pain [Cole et al. 2007]. The
MOS-Sleep has traditionally utilized a recall
period of 4 weeks, as utilized in trial A0081056
[Arnold et al. 2008]. However, the US Food and
Drug Administration (FDA) has recommended
against use of patient-reported outcomes with
long recall periods [US Department of Health
and Human Services, 2009], and a 1-week
recall period has been shown reliable for use of
the MOS-Sleep in studies evaluating sleep distur-
bance in patients with FM [Sadosky et al. 2009],
so a 1-week recall period was used in studies
A0081077 and A0081100 [Mease et al. 2008b;
Pauer et al. 2008]. The MOS-Sleep has been val-
idated for use in patients with FM and has good
reliability over time using a testretest model
[Cappelleri et al. 2009; Sadosky et al. 2009].
Since there are no set cutoff scores on the
MOS-sleep to define clinically significant sleep
problems, an SPI score of 50 was used to
define clinically significant sleep problems since
this score is one SD above the mean seen in
healthy controls [Hays and Stewart, 1992] and
identifies patients with moderately severe sleep
interference [Viala-Danten et al. 2008]. The
HADS is one of the most frequently used scales
to assess anxiety and depression symptoms in
somatically ill patients since its items exclude
somatic symptoms of psychologic distress
[Zigmond and Snaith, 1983]. The HADS has
been validated in numerous patient populations
including patients with musculoskeletal disorders
and found to be internally consistent for assessing
the severity of anxiety and depression symptoms
[Pallant and Bailey, 2005]. The HADS consists
of two, seven-item subscales to assess anxiety
(HADS-A) and depression (HADS-D) symp-
toms with scores for each that can range from 0
to 21, with higher scores indicating higher levels
of each symptom. HADS-A and HADS-D scores
11 were used to define clinically significant anx-
iety and depression levels, respectively, since
these are the most widely accepted cutoff scores

http://tab.sagepub.com 217
Therapeutic Advances in Musculoskeletal Disease 3 (5)
with a sensitivity and specificity for identifying
patients of approximately 80% [Bjelland et al.
2002].
The FIQ is the accepted standard for quantifying
the global severity of FM symptoms [Bennett
et al. 2009]. The FIQ is composed of 20 items
that assess disease severity over the past week
with 11 questions to assess physical functioning,
two day-of-the-week items to quantify the
number of days patients felt good or missed
work, and seven VASs to assess symptoms of
fatigue, sleep quality, depression, anxiety, stiff-
ness, pain and work disability. The 11 FIQ phys-
ical functioning items evaluate the ability of
patients to perform large muscle activities and
each were scored on a 0-to-3 Likert scale, with
0 indicating always and 3 indicating never
[Bennett et al. 2009]. A composite physical func-
tioning score was derived by summing the phys-
ical functioning items, dividing by the number of
questions answered, and multiplying by 3.33 to
yield a 010 score. The felt good day-of-the-
week score was derived by reverse scoring (to
obtain the number of days patients felt bad)
and multiplying the result by 1.43 to yield a
010 score. The missed work day-of-the-
week score was derived by multiplying the
number of days by 1.43 to yield a 010 score.
VAS scores were derived by measuring the dis-
tance from the origin to the patient mark in cen-
timeters to yield a 010 score for each symptom.
FIQ global scores were derived by summing the
composite physical functioning, day-of-the-
week and VAS scores, with higher scores indicat-
ing more severe FM. To maintain a maximum
possible score of 100, an equalization calcula-
tion was used if patients did not answer all 10
sections by multiplying the global score by 10 and
dividing by the number of sections answered.
VASFIQ global scores were derived by summing
scores from the seven FIQ VASs to yield a 070
score, with higher scores indicating more severe
FM. Owing to its brevity, VASFIQ global scores
were only calculated for patients who answered
all VASs.
Statistical analysis
All analyses were conducted on an intention-to-
treat population (i.e. randomized patients that
took a dose and contributed efficacy data).
Summary statistics were described for the demo-
graphics data, such as sex, race and age. Baseline
efficacy characteristics were descriptively sum-
marized for the FIQ global, VASFIQ global,
218
HADS-D, HADS-A, SPI and MAF-GFI.
Spearman rank correlations compared FIQ
global scores with VASFIQ global scores at base-
line and study endpoints, change in FIQ and
VASFIQ global scores from baseline to study
endpoints, and individual FIQ VAS scores with
corresponding full-length symptom question-
naires at baseline and study endpoints.
Spearman rank correlations also were used to
compare PGIC scores at study endpoints with
change in global FIQ and VASFIQ scores.
Percentage of patients with minimal (floor) and
maximal (ceiling) values also were determined for
individual FIQ VASs and the corresponding full-
length symptom scales. Receiver operating char-
acteristic (ROC) [Harris, 2010; Zweig and
Campbell, 1993; Hanley and McNeil, 1982;
Metz, 1978] analyses of baseline patient data
identified cutoff scores on individual FIQ VASs
that classified patients with clinically significant
symptom levels using corresponding validated
questionnaires to define cases (clinically signifi-
cant cases were defined for anxiety if HADS-A
scores were 11, depression if HADS-D scores
were 11, fatigue if MAF-GFI scores were 30,
and poor sleep if SPI scores were 50). We also
calculated the sensitivity and specificity of the
tests using the identified cutoffs. The intersection
of the sensitivity and specificity curves estab-
lished the cutoff scores on individual FIQ VASs,
according to methods previously described
[Harris, 2010]. Scatterplot and regression analy-
ses were used to determine corresponding values
between global FIQ and VASFIQ scores.
Results
While questionnaire data were derived primarily
from white, middle-aged females (Table 1), the
cohort included a number of patients from other
groups including Hispanics, Blacks, and men.
Owing to its brevity, VASFIQ global scores
were only calculated for patients who answered
all VASs, resulting in the exclusion of only 18 of
2225 patients with FIQ global scores (Table
2).The patients typically had moderate-to-
severe FM as indicated by an average FIQ
global score >60 (Table 2) [Burckhardt et al.
1991]. Patients also typically had clinically signif-
icant symptoms of fatigue and poor sleep quality
as evidenced by average MAF-GFI scores >30
and average SPI scores >50 (Table 2). Median
scores in the HADS-A (score 9) and HADS-D
(score 7) indicate that our FM cohort had
levels of anxiety and depression consistent with
those previously recorded in patients with
http://tab.sagepub.com
 CS Boomershine, B Emir et al.

Table 1. Summary of FM patient characteristics.

Characteristic All patients (n 2229)
Mean age (SD), years 49.12 (11.18)
Range, years 1882
Race, n (%)
White 1912 (85.8)
Black 69 (3.1)
Hispanic 149 (6.7)
Other 99 (4.4)
Gender, n (%)
Women 2083 (93.4)
Men 146 (6.6)
Mean FM duration, months (SD) 111.28 (95.42)
Median 85
Range 1656

FM, fibromyalgia; SD, standard deviation.

Table 2. Summary of fibromyalgia patient baseline questionnaire scores.

Questionnaire N Mean (SD) Median Range
FIQ global 2225 61.72 (14.73) 61.9 8.899.1
VASFIQ global 2207 46.90 (10.63) 47 5.270
VASfatigue 2216 7.86 (1.85) 8.3 0.110
VASsleep 2215 7.81 (1.94) 8.3 0.210
VASanxiety 2217 4.99 (2.94) 5.2 010
VASdepression 2214 4.59 (2.98) 4.7 010
HADS-A 2215 9.02 (4.42) 9 021
HADS-D 2215 7.63 (4.20) 7 021
SPI 2197 60.54 (17.75) 61.4 4.5100
MAF-GFI 2193 36.39 (7.95) 37.8 7.550

FIQ, Fibromyalgia Impact Questionnaire; FM, fibromyalgia; HADS-A, Hospital Anxiety and Depression ScaleAnxiety sub-
scale; HADS-D, Hospital Anxiety and Depression ScaleDepression subscale; MAF-GFI, Multidimensional Assessment of
Fatigue Global Fatigue Index; SD, standard deviation; SPI, Medical Outcomes Study Sleep Problems Index II; VAS, visual
analog scale.

chronic musculoskeletal pain (i.e. higher than in
patients with other chronic medical conditions
but lower than in outpatients with psychiatric
conditions) [Pallant and Bailey, 2005; Zigmond
and Snaith, 1983]. Consistent with the full-
length questionnaire scores, VAS scores for fati-
gue and sleep quality tended to be in the severe
range (averaging near 8 for both; Table 2), while
anxiety and depression VAS scores indicated
more moderate symptoms (averaging near 5 for
both; Table 2).
To compare global scores on the brief question-
naire VASFIQ with its full-length counterpart
FIQ, Spearman correlations for baseline, end-
point and change scores were performed (Table
3). VASFIQ global scores were highly correlated
with FIQ global scores at baseline (r 0.94,
p < 0.0001) and at study endpoints (r 0.97,
p < 0.0001). Change in VASFIQ and FIQ
global scores at study endpoints were also
highly correlated with one another (r 0.97,
p < 0.0001) and moderately correlated with
PGIC scores (r 0.61, p < 0.0001 for FIQ
global and r 0.58, p < 0.0001 for VASFIQ
global). Individual FIQ VASs scores correlated
well with corresponding full-length questionnaire
scores at baseline (VASfatigue with MAF-GFI,
r 0.67; VASanxiety with HADS-A, r 0.64;
VASdepression with HADS-D, r 0.57;
VASsleep with SPI, r 0.50; p < 0.0001 for all)
and at study endpoints (VASfatigue with MAF-
GFI, r 0.76; VASanxiety with HADS-A,
r 0.67; VASdepression with HADS-D,

http://tab.sagepub.com 219
Therapeutic Advances in Musculoskeletal Disease 3 (5)
Table 3. Summary of correlations between questionnaire scores.
Questionnaire comparison
FIQ global versus VASFIQ, r
PGIC versus FIQ global, r
PGIC versus VASFIQ, r
MAF-GFI versus VASfatigue, r
HADS-A versus VASanxiety, r
HADS-D versus VASdepression, r
SPI versus VASsleep, r
FIQ, Fibromyalgia Impact Questionnaire; HADS-A, Hospital Anxiety and Depression Scale Anxiety subscale; HADS-D,
Hospital Anxiety and Depression Scale Depression subscale; MAF-GFI, Multidimensional Assessment of Fatigue
Global Fatigue Index; n/a, not applicable; PGIC, Patient Global Impression of Change; SPI, Medical Outcomes Study
Sleep Problems Index II; VAS, visual analog scale.
r 0.62; VASsleep with SPI, r 0.67;
p < 0.0001 for all). Similarly, change in individual
FIQ VAS scores were also correlated with change
in the corresponding full-length questionnaire
scores at study endpoints (VASfatigue with
MAF-GFI, r 0.64; VASanxiety with HADS-
A, r 0.47; VASdepression with HADS-D,
r 0.43; VASsleep with SPI, r 0.57;
p < 0.0001 for all; Table 3).
Overlapping scoring density plots of baseline
scores were used to compare the distribution of
VAS symptom scores to their full-length counter-
parts (Figure 2). Depiction of scores in the den-
sity plots were made comparable by subtracting
the mean from each score and dividing by the
SD. There was considerable overlap of the scor-
ing distributions between the VASs and the full-
length questionnaires. Distributions of full-length
questionnaires were normally distributed, with
HADS-A and HADS-D scores shifted towards
lower scores (Figure 2(a) and (b), respectively),
whereas the MAF-GFI and SPI were shifted
towards higher scores (Figure 2(c) and (d),
respectively). The distribution of VASanxiety
and VASdepression scores showed a bimodal pat-
tern centered on the mean (Figure 2(a) and (b),
respectively). In contrast, the VASfatigue and
VASsleep scoring distributions were unimodal
(Figure 2(c) and (d), respectively). VASFIQ
global, FIQ global, individual FIQ VAS and
full-length symptom questionnaire baseline
scores at various patient percentile levels were
calculated to compare floor and ceiling effects.
No floor effects were seen for either the global
VASFIQ or FIQ scales, and ceiling effects were
low with 0% for FIQ global and 0.3% for
VASFIQ global scores. Floor and ceiling effects
220
Baseline Endpoint Change
0.94 0.97 0.97
n/a n/a 0.61
n/a n/a 0.58
0.67 0.76 0.64
0.64 0.67 0.47
0.57 0.62 0.43
0.50 0.67 0.57
were also low for all full-length symptom ques-
tionnaires at 1% or less. Floor effects for the
VASanxiety and VASdepression scores were
2.2% and 2.7%, respectively, and floor effects
for the VASfatigue and VASsleep scores were
zero. Ceiling effects for the VASfatigue and
VASsleep scores were 6.8% and 7.1%, respec-
tively, while VASanxiety and VASdepression ceil-
ing effects were lower at 1.6% and 1.4%,
respectively.
ROC analyses were performed to determine VAS
cutoff scores that identified FM patients with
clinically significant symptom levels with maxi-
mum sensitivity and specificity using the corre-
sponding full-length questionnaires as reference
standards. A VASfatigue score of >7.5 was 76.0%
sensitive and 82.2% specific for clinically signifi-
cant fatigue (ROC area under the curve
[AUC] 0.87, 95% confidence interval
[CI] 0.850.90; Figure 3(a)). A VASsleep
score of >7.9 was 68.7% sensitive and 71.2%
specific for significant sleep disturbance (ROC
AUC 0.76, 95% CI 0.740.79; Figure
3(b)). A VASdepression score of >5.8 was
71.3% sensitive and 72.6% specific for significant
depressive symptoms (ROC AUC 0.80, 95%
CI 0.770.82; Figure 3(c)), and a VASanxiety
score of >6.0 was 71.2% sensitive and 77.7%
specific for clinically significant anxiety (ROC
AUC 0.83, 95% CI 0.810.84; Figure 3(d)).
Correspondence between the global VASFIQ and
FIQ baseline scores were demonstrated in a scat-
terplot (Figure 4). FIQ global scores that define
moderate (39) and severe (59) symptoms
[Bennett et al. 2010] corresponded to VASFIQ
http://tab.sagepub.com
 CS Boomershine, B Emir et al.

Anxiety Depression
(a) 0.4 (b) 0.4
HADS-A HADS-D
VASanxiety VASdepression
0.3 0.3

Density

Density
0.2 0.2

0.1 0.1

0.0 0.0
2 1 0 1 2 3 4 2 1 0 1 2 3 4
Fatigue Sleep
(c) 0.6 (d) 0.6
MAF-GFI SPI
VASfatigue VASrested

0.4 0.4
Density

Density
0.2 0.2

0.0 0.0
2 1 0 1 2 3 4 2 1 0 1 2 3 4

Figure 2. Overlapping scoring density plots of baseline VAS and full-length questionnaire scores. HADS-A,
Hospital Anxiety and Depression Scale Anxiety subscale; HADS-D, Hospital Anxiety and Depression Scale
Depression subscale; MAF-GFI, Multidimensional Assessment of Fatigue Global Fatigue Index; SPI, Medical
Outcomes Study Sleep Problems Index II; VAS, visual analog scale.

global scores of 31.4 and 45.0, respectively
(Figure 4).
Discussion
The complexity and heterogeneity of symptoms
experienced by patients with FM can make them
very challenging to manage. To improve manage-
ment, evidence-based FM guidelines recommend
development of an individualized treatment plan
tailored according to pain intensity, function and
associated features such as depression, fatigue
and sleep disturbance; this should be based on
a comprehensive assessment of pain, function
and psychosocial context [Carville et al. 2008].
In addition, new preliminary diagnostic criteria
recommended by the ACR for FM require phy-
sicians to quantify symptom severity [Wolfe et al.
2010]. However, comprehensive assessment of
FM symptoms can be difficult to perform
within the time available in a busy clinical prac-
tice. While the FIQ VASs have previously been
recommended as a rapid FM assessment
[Boomershine and Crofford, 2009], this advice
was based on research conducted in a single
rheumatology practice on a small number of
patients [Boomershine et al. 2008].
The work outlined herein provides support for
use of the VASFIQ in FM symptom assessment.
Baseline, endpoint and change global scores on
the VASFIQ correlated with corresponding FIQ
global scores (Table 3 and Figure 4) [Bennett
et al. 2010]. This, taken together with the finding
that global VASFIQ and FIQ change scores had
similar correlations with PGIC endpoint scores
(Table 3) and similarly low floor and ceiling
effects, suggests that the seven FIQ VAS items
can be summed to provide global FM disease
severity information. In addition to providing
global disease severity information in aggregate,
correlations between baseline, followup and
change scores of individual VASs and corre-
sponding full-length symptom questionnaires
(Table 3) indicate VASFIQ items can quantify
the severity of symptoms of fatigue, anxiety,
depression and sleep quality. ROC analyses
(Figure 3) also demonstrate that cutoff scores
on the FIQ VASs can be identified that charac-
terize FM patients with clinically significant
symptoms of fatigue (>7.5), poor sleep (>7.9),
anxiety (>6.0) and depression (>5.8).
The symptoms measured by the VASFIQ are
widely regarded as clinically relevant to patients
with FM. A Delphi exercise studying the impact
of FM symptoms on patients lives demonstrated
that pain, fatigue and sleep disturbance are
among the most detrimental [Mease et al.

http://tab.sagepub.com 221
 (a) Operating characteristics for MAF-GFI 30 (b) Operating characteristics for SPI 50

222
AUC: 100% sensitivity: Intersection: 100% specificity: 95% CI for AUC: AUC: 100% sensitivity: Intersection: 100% specificity: 95% CI for AUC:
0.8724 2 7.5000 (0.85340.8914) 0.7629 0.4000 7.9000 (0.73960.7862)
Fatigue No fatigue Sleep disturbance No sleep disturbance
Yes 1362 (62%) 70 (3.2%) Yes 1127 (52%) 157 (7.2%)
No 429 (20%) 323 (15%) No 514 (24%) 389 (18%)
1.0 Specificity 1.0 Specificity
Sensitivity Sensitivity
0.8 0.8

0.6 0.6

0.4 0.4

Proportion

Proportion
0.2 0.2

0.0 0.0
0 2 4 6 8 10 0 2 4 6 8 10
FIQ VASfatigue at baseline FIQ VASsleep at baseline

Yes Yes

No No

Disease

Disease
0 2 4 6 8 10 0 2 4 6 8 10

FIQ VASfatigue at baseline FIQ VASsleep at baseline

Confusion matrix is for the FIQ VASfatigue at baseline at the cutoff = 7.5 Confusion matrix is for the FIQ VASsleep at baseline at the cutoff = 7.9

(c) Operating characteristics for HADS-D 11 (d) Operating characteristics for HADS-A 11
AUC: 100% sensitivity: Intersection: 100% specificity: 95% CI for AUC: AUC: 100% sensitivity: Intersection: 100% specificity: 95% CI for AUC:
0.7951 0 5.8000 (0.77340.8167) 0.8257 0 6 (0.80840.8430)
Depression No depression Anxiety No anxiety
Therapeutic Advances in Musculoskeletal Disease 3 (5)

Yes 382 (17%) 457 (21%) Yes 560 (25%) 317 (14%)
No 154 (7%) 1211 (55%) No 227 (10%) 1103 (50%)
1.0 Specificity 1.0 Specificity
Sensitivity Sensitivity
0.8 0.8

0.6 0.6

0.4 0.4

Proportion
Proportion
0.2 0.2

0.0 0.0
0 2 4 6 8 10 0 2 4 6 8 10
FIQ VASdepression at baseline FIQ VASanxiety at baseline

Yes Yes

No No

Disease
Disease

0 2 4 6 8 10 0 2 4 6 8 10
FIQ VASdepression at baseline FIQ VASanxiety at baseline
Confusion matrix is for the FIQ VASdepression at baseline at the cutoff = 5.8 Confusion matrix is for the FIQ VASanxiety at baseline at the cutoff = 6.0

Figure 3. Receiver operating characteristic analyses for (a) VASfatigue, (b) VASsleep, (c) VASdepression and (d) VASanxiety baseline scores. AUC, area under the
curve; CI, confidence interval; FIQ, Fibromyalgia Impact Questionnaire; HADS-A, Hospital Anxiety and Depression Scale Anxiety subscale; HADS-D, Hospital Anxiety
and Depression Scale Depression subscale; MAF-GFI, Multidimensional Assessment of Fatigue Global Fatigue Index; ROC, receiver operating characteristic; SPI,
Medical Outcomes Study Sleep Problems Index II; VAS, visual analog scale.

http://tab.sagepub.com

70
60
50
VAS FIQ global
45.0
40
31.4
30
Global FIQ categories:
20
10
39.0 59.0
0
0 20 40 60
FIQ global scores
Figure 4. Correspondence between baseline VASFIQ
and FIQ global scores. FIQ, Fibromyalgia Impact
Questionnaire; VAS, visual analog scale.
2008a]; this led researchers to classify these
symptoms among a core set of domains that are
essential to measure in all clinical FM trials
[Mease et al. 2009]. Depression was classified
among domains to be assessed at some point in
a clinical development program, whereas anxiety
and stiffness were considered domains of
research interest. In addition, a cluster analysis
performed to identify clinical features of FM
that patients would most like to see improved
found that clusters of pain and fatigue symptoms
were selected most frequently at 90% and 89%,
respectively [Bennett et al. 2010]. Disturbed
sleep was included as a clinical feature in the
pain cluster and was identified as the second
most important individual symptom behind
pain or discomfort. The affective cluster,
which included feelings of anxiety or depression,
was selected fifth overall at 21%.
VASs have been used since the 1920s to measure
subjective phenomena owing to their brevity, uni-
versality, ease of use and scoring [Wewers and
Lowe, 1990]. These properties, along with
research showing the performance of VASs rival
longer scales with respect to correlation with clin-
ical variables and sensitivity to change [Wolfe,
2004], have led to widespread use of VASs in
both clinical medicine and research. VASs have
shown great utility in measuring numerous symp-
toms relevant to FM including mood, sleep qual-
ity, functional ability and pain [Wewers and
Lowe, 1990]. VASs measuring fatigue also have
http://tab.sagepub.com
CS Boomershine, B Emir et al.
been shown to be highly reproducible and sensi-
tive for quantifying symptom severity and treat-
ment response in multiple patient populations
[Wolfe, 2004; Grant et al. 1999]. Similar to the
VASFIQ, VAS questionnaire sets have been used
previously in other populations with complex
symptom burdens such as cancer patients being
treated with chemotherapy [Padilla et al. 1983;
0<39 mild Presant et al. 1981]. As with the VASFIQ, these
39<59 moderate
scales used individual VAS scores to quantify spe-
59100 severe
cific symptoms and combined VAS scores to
measure overall quality of life.
The present work has several limitations.
80 100
Numerous limitations were introduced by the
types of patients included in the treatment trials
from which the data were collected [Arnold et al.
2008; Mease et al. 2008b; Pauer et al. 2008].
Study participants consisted primarily of white,
middle-aged females (Table 1), which may limit
applicability of these data to other populations.
All patients included in the trials were required to
meet ACR classification criteria for FM and have
baseline pain scores in the moderate-to-severe
range. These inclusion criteria are likely to have
skewed the study population toward patients with
more severe symptoms than the average patient
with FM typically seen in clinical practice, as
patients who fulfill the ACR criteria are known
to have more severe symptom levels [Katz et al.
2006]. Conversely, patients were excluded from
the trials if they had severe depression or unstable
psychiatric disorders, and this may have falsely
reduced scores on psychiatric assessments com-
pared with those seen in routine clinical practice.
However, the proportions of patients with signif-
icant depressive or anxiety symptoms as deter-
mined by the HADS (approximately 25% for
depression and 40% for anxiety) are similar to
those previously reported for patients in the com-
munity with FM [Herrmann, 1997].
The study also was limited by the full-length
questionnaires available for use as reference stan-
dards to characterize patients with clinically sig-
nificant symptom levels. The present analyses
were performed on data from pre-existing prega-
balin FM clinical trials [Arnold et al. 2008;
Mease et al. 2008b; Pauer et al. 2008], and so
we were dependent on the questionnaires that
were utilized in these trials. The HADS is the
best available self-report questionnaire for assess-
ing the severity of anxiety and depression symp-
toms in patients with FM since it was specifically
designed to exclude the effects of complicating
223
Therapeutic Advances in Musculoskeletal Disease 3 (5)
physical and emotional symptoms of medical ill-
ness [Zigmond and Snaith, 1983] and its struc-
ture has proven validity in numerous populations
including patients with chronic musculoskeletal
pain [Pallant and Bailey, 2005]. While the
HADS has a sensitivity and specificity for identi-
fying patients with clinically significant depres-
sion and anxiety symptoms of approximately
80% [Bjelland et al. 2002], it is not a diagnostic
instrument. Accurate diagnosis of depressive and
anxiety disorders requires a structured diagnostic
interview; an interview such as the Mini-
International Neuropsychiatric Interview
(MINI) would be preferred to categorize patients
with clinically significant symptoms [Sheehan
et al. 1998]. Similarly, while the MOS-Sleep has
demonstrated validity and reliability in quantify-
ing the severity of sleep symptoms in FM patients
[Cappelleri et al. 2009; Sadosky et al. 2009], it is
not a diagnostic instrument. A structured clinical
interview or a questionnaire such as the Athens
Insomnia Scale based on International
Classification of Diseases, Tenth Revision criteria
for insomnia diagnosis with predetermined cut-
off scores would have been preferable for identi-
fying patients with clinically significant sleep dis-
turbance [Soldatos et al. 2000]. Future studies
using gold-standard assessments to diagnose
FM patients with mood disorders and insomnia
are needed to assess the accuracy of FIQ VAS cut
points identified in the current study.
The VASFIQ has some inherent limitations that
could impact its clinical utility [Wewers and
Lowe, 1990]. Patients sometimes have difficulty
understanding the VAS concept, and initial train-
ing is required to ensure accurate completion.
VASFIQ completion requires self-administration
of a paper questionnaire and VAS scoring
requires distance measurement with a ruler,
both of which can be cumbersome to perform
in the clinic setting. Oral administration of the
VASFIQ using a 010 numeric rating scale has
been proposed as a way to circumvent these
issues [Boomershine, 2010; Boomershine and
Crofford, 2009]. However, since oral administra-
tion would require changes to the VASFIQ,
including addition of a small introductory expla-
nation and third-person wording of the items,
validation of an interviewer-administered version
is needed before it can be recommended. While
the VASFIQ assesses many symptoms of clinical
importance in FM [Mease et al. 2009], it fails to
evaluate the severity of cognitive dysfunction
224
necessary for ACR FM diagnostic criteria
[Wolfe et al. 2010].
Two revisions of the FIQ VASs have sought to
address these limitations [Boomershine, 2010;
Bennett et al. 2009]. A revision of the FIQ
(FIQR) has been developed recently that pro-
vides a rapid alternative to the full-length FIQ
[Bennett et al. 2009]. This patient questionnaire
is longer than the VASFIQ, uses a different
format for answering questions, and includes a
subscale for patient function. Although the
FIQR shows promise as a screening tool, it has
only been tested in two studies: an online and
focus group study [Bennett et al. 2009], and an
interventional pilot study [Carson et al. 2010].
Further validation in the clinical setting is neces-
sary before it can be recommended. While VASs
perform well in quantifying baseline symptom
severity and gauging symptom improvement at
follow up, they often perform poorly in assessing
symptom worsening over time [Farrar et al.
2001]. VASs perform poorly at follow up when
baseline scores are at or near the upper limit, as
seen with the VASfatigue and VASsleep items,
because ceiling effects do not allow patients to
indicate symptom worsening. This is an impor-
tant limitation of using VASs in FM manage-
ment, because medications that improve some
FM symptoms often worsen other symptoms
and many patients with FM have high baseline
VAS scores. A revision of the VASFIQ that uti-
lizes more extreme wording for the upper anchors
to decrease ceiling effects and a patient impres-
sion of change format at follow up has been pro-
posed to improve assessment of symptom
worsening in patients with FM [Boomershine,
2010], but this revision needs to be validated.
Finally, we want to stress that the VASFIQ
should not be used in isolation to make treat-
ment decisions. The VASFIQ is a rapid screen
that can quantify symptom severity, but it is not
intended for use as a diagnostic instrument. A
careful clinical evaluation is required to identify
the underlying cause of symptoms and determine
the most appropriate course of therapy. The
VASFIQ can improve FM management by pro-
viding a mechanism for patients to effectively
articulate their symptom burden and enhance
the ability of clinicians to interpret the patients
complaints into a coherent intellectual frame-
work from which to make diagnostic and treat-
ment decisions.
http://tab.sagepub.com

Acknowledgements
Editorial support was provided by Patricia
McChesney, PhD, of UBC Scientific Solutions
and was funded by Pfizer Inc.
Funding
This study was funded by Pfizer Inc. Chad S.
Boomershine is supported by the NIH (award
number K08DK080219).
Conflict of interest statement
Chad S. Boomershine has previously received
research support from Pfizer Inc. for developing
the VASFIQ tool. Dr. Boomershine also serves as
a consultant for Pfizer Inc., Eli Lilly and
Company and Forest Pharmaceuticals Inc. Birol
Emir and Gergana Zlateva are full-time employ-
ees of Pfizer Inc. Yi Wang performed a paid
internship at Pfizer Inc.
References
Arnold, L.M., Russell, I.J., Diri, E.W., Duan, W.R.,
Young Jr, J.P., Sharma, U. et al. (2008) A 14-week,
randomized, double-blinded, placebo-controlled
monotherapy trial of pregabalin in patients with
fibromyalgia. J Pain 9: 792805.
Belza, B.L. (1995) Comparison of self-reported fatigue
in rheumatoid arthritis and controls. J Rheumatol
22: 639643.
Bennett, R.M., Friend, R., Jones, K.D., Ward, R.,
Han, B.K. and Ross, R.L. (2009) The Revised
Fibromyalgia Impact Questionnaire (FIQR): valida-
tion and psychometric properties. Arthritis Res Ther
11: R120.
Bennett, R.M., Russell, J., Cappelleri, J.C.,
Bushmakin, A.G., Zlateva, G. and Sadosky, A. (2010)
Identification of symptom and functional domains that
fibromyalgia patients would like to see improved: a
cluster analysis. BMC Musculoskelet Disord 11: 134.
Bjelland, I., Dahl, A.A., Haug, T.T. and Neckelmann,
D. (2002) The validity of the Hospital Anxiety and
Depression Scale. An updated literature review. J
Psychosom Res 52: 6977.
Boomershine, C.S. (2010) The FIBRO System: a
rapid strategy for assessment and management of
fibromyalgia syndrome. Therapeut Adv Musculoskel Dis
2: 187200.
Boomershine, C.S. and Crofford, L.J. (2009) A
symptom-based approach to pharmacologic manage-
ment of fibromyalgia. Nat Rev Rheumatol 5: 191199.
Boomershine, C.S., Titova, D., Chung, C.P., Oeser,
A. and Stein, M. (2008) Five visual analogue scales
quantify global disease severity and identify clinically
significant symptoms in fibromyalgia syndrome.
Arthritis Rheum 58: S686S687.
http://tab.sagepub.com
CS Boomershine, B Emir et al.
Burckhardt, C.S., Clark, S.R. and Bennett, R.M.
(1991) The fibromyalgia impact questionnaire: devel-
opment and validation. J Rheumatol 18: 728733.
Cappelleri, J.C., Bushmakin, A.G., McDermott,
A.M., Dukes, E., Sadosky, A., Petrie, C.D. et al.
(2009) Measurement properties of the Medical
Outcomes Study Sleep Scale in patients with fibro-
myalgia. Sleep Med 10: 766770.
Carson, J.W., Carson, K.M., Jones, K.D., Bennett,
R.M., Wright, C.L. and Mist, S.D. (2010) A pilot
randomized controlled trial of the Yoga of Awareness
program in the management of fibromyalgia. Pain
151: 530539.
Carville, S.F., Arendt-Nielsen, S., Bliddal, H.,
Blotman, F., Branco, J.C., Buskila, D. et al. (2008)
EULAR evidence-based recommendations for the
management of fibromyalgia syndrome. Ann Rheum
Dis 67: 536541.
Cole, J.C., Dubois, D. and Kosinski, M. (2007) Use
of patient-reported sleep measures in clinical trials of
pain treatment: a literature review and synthesis
of current sleep measures and a conceptual model of
sleep disturbance in pain. Clin Ther 29(Suppl):
25802588.
Farrar, J.T., Young Jr, J.P., LaMoreaux, L., Werth, J.L.
and Poole, R.M. (2001) Clinical importance of
changes in chronic pain intensity measured on an 11-
point numerical pain rating scale. Pain 94: 149158.
Grant, S., Aitchison, T., Henderson, E., Christie, J.,
Zare, S., McMurray, J. et al. (1999) A comparison of
the reproducibility and the sensitivity to change of
visual analogue scales, Borg scales, and Likert scales in
normal subjects during submaximal exercise. Chest
116: 12081217.
Hanley, J.A. and McNeil, B.J. (1982) The meaning
and use of the area under a receiver operating charac-
teristic (ROC) curve. Radiology 143: 2936.
Harris, K. (2010) ROC hard? No ROC made easy! In:
SAS Global Forum 2010, Seattle, WA, available at:
http://support.sas.com/resources/papers/proceed-
ings10/2222010.pdf.
Hays, R.D. and Stewart, A.L. (1992) Sleep measures,
In: Stewart, A.L. and Ware, J.E. (eds). Measuring
Functioning and Well-being: The Medical Outcomes
Study Approach, Duke University Press: Durham,
NC, pp. 235259.
Herrmann, C. (1997) International experiences with
the Hospital Anxiety and Depression Scalea review of
validation data and clinical results. J Psychosom Res
42: 1741.
Katz, R.S., Wolfe, F. and Michaud, K. (2006)
Fibromyalgia diagnosis: a comparison of clinical,
survey, and American College of Rheumatology crite-
ria. Arthritis Rheum 54: 169176.
Martin, S., Chandran, A., Zografos, L. and Zlateva, G.
(2009) Evaluation of the impact of fibromyalgia on
225
Therapeutic Advances in Musculoskeletal Disease 3 (5)
patients sleep and the content validity of two sleep
scales. Health Qual Life Outcomes 7: 64.
Mease, P.J., Arnold, L.M., Choy, E.H., Clauw, D.J.,
Crofford, L.J., Glass, J.M. et al. (2009) Fibromyalgia
syndrome module at OMERACT 9: domain con-
struct. J Rheumatol 36: 23182329.
Mease, P.J., Arnold, L.M., Crofford, L.J., Williams,
D.A., Russell, I.J., Humphrey, L. et al. (2008a)
Identifying the clinical domains of fibromyalgia: con-
tributions from clinician and patient Delphi exercises.
Arthritis Rheum 59: 952960.
Mease, P.J., Russell, I.J., Arnold, L.M., Florian, H.,
Young Jr, J.P., Martin, S.A. et al. (2008b) A random-
ized, double-blind, placebo-controlled, phase III trial
of pregabalin in the treatment of patients with fibro-
myalgia. J Rheumatol 35: 502514.
Metz, C.E. (1978) Basic principles of ROC analysis.
Semin Nucl Med 8: 283298.
Neuberger, G.B. (2003) Measures of Fatigue, The
Fatigue Questionnaire, Fatigue Severity Scale,
Multidimensional Assessment of Fatigue Scale, and
Short Form-36 Vitality (Energy/Fatigue) Subscale of
the Short Form Health Survey. Arthritis Rheum
49: S175S183.
Padilla, G.V., Presant, C., Grant, M.M., Metter, G.,
Lipsett, J. and Heide, F. (1983) Quality of life index for
patients with cancer. Res Nurs Health 6: 117126.
Pallant, J.F. and Bailey, C.M. (2005) Assessment of
the structure of the Hospital Anxiety and Depression
Scale in musculoskeletal patients. Health Qual Life
Outcomes 3: 82.
Pauer, L., Danneskiold-Samsoe, B., Jespersen, A.,
Atkinson, G., Whelan, L., Leon, T. et al. (2008)
Pregabalin for management of fibromyalgia (FM): a
14-week, randomized, double-blind, placebo-con-
trolled, monotherapy trial (study A0081100). Ann
Rheum Dis 67(Suppl. II): 256.
Presant, C.A., Klahr, C. and Hogan, L. (1981)
Evaluating quality-of-life in oncology patients: pilot
observations. Oncol Nurs Forum 8: 2630.
Sadosky, A., Dukes, E. and Evans, C. (2009)
Visit SAGE journals online Reliability of a 1-week recall period for the Medical
http://tab.sagepub.com
Outcomes Study Sleep Scale (MOS-SS) in patients
with fibromyalgia. Health Qual Life Outcomes 7: 12.
Sheehan, D.V., Lecrubier, Y., Sheehan, K.H.,
Amorim, P., Janavs, J., Weiller, E. et al. (1998) The
Mini-International Neuropsychiatric Interview
226
(M.I.N.I.): the development and validation of a
structured diagnostic psychiatric interview for DSM-
IV and ICD-10. J Clin Psychiatry 59(Suppl. 20):
2233, , quiz 3457.
Soldatos, C.R., Dikeos, D.G. and Paparrigopoulos,
T.J. (2000) Athens Insomnia Scale: validation of an
instrument based on ICD-10 criteria. J Psychosom Res
48: 555560.
US Department of Health and Human Services
(2009) Guidance for Industry. Patient-Reported
Outcome Measures: Use in Medical Product
Development to Support Labeling Claims, available at:
http://www.fda.gov/downloads/Drugs/
GuideanceComplianceRegulatoryInformation/
Guidances/UCM193282.pdf.
Viala-Danten, M., Martin, S., Guillemin, I. and Hays,
R.D. (2008) Evaluation of the reliability and validity of
the Medical Outcomes Study sleep scale in patients
with painful diabetic peripheral neuropathy during an
international clinical trial. Health Qual Life Outcomes
6: 113.
Wewers, M.E. and Lowe, N.K. (1990) A critical
review of visual analogue scales in the measurement of
clinical phenomena. Res Nurs Health 13: 227236.
Wolfe, F. (1990) Fibromyalgia. Rheum Dis Clin North
Am 16: 681698.
Wolfe, F. (2004) Fatigue assessments in rheumatoid
arthritis: comparative performance of visual analog
scales and longer fatigue questionnaires in 7760
patients. J Rheumatol 31: 18961902.
Wolfe, F., Clauw, D.J., Fitzcharles, M.A., Goldenberg,
D.L., Katz, R.S., Mease, P. et al. (2010) The American
College of Rheumatology preliminary diagnostic cri-
teria for fibromyalgia and measurement of symptom
severity. Arthritis Care Res (Hoboken) 62: 600610.
Wolfe, F., Smythe, H.A., Yunus, B., Bennett, R.M.,
Bombardier, C., Goldenberg, D.C., et al. (1990) The
American College of Rheumatology 1990 Criteria for
the Classification of Fibromyalgia. Arthritis Rheum 33:
160172.
Zigmond, A.S. and Snaith, R.P. (1983) The hospital
anxiety and depression scale. Acta Psychiatr Scand
67: 361370.
Zweig, M.H. and Campbell, G. (1993) Receiver-
operating characteristic (ROC) plots: a fundamental
evaluation tool in clinical medicine. Clin Chem
39: 561577.
http://tab.sagepub.com