Content uploaded by Paolo Gisondi
Author content
All content in this area was uploaded by Paolo Gisondi on Dec 02, 2016
Content may be subject to copyright.
Original article
The early psoriatic arthritis screening questionnaire:
a simple and fast method for the identification of
arthritis in patients with psoriasis
Ilaria Tinazzi
1
, Silvano Adami
1
, Elisabetta Maria Zanolin
2
, Cristian Caimmi
1
,
Silvia Confente
1
, Giampiero Girolomoni
3
, Paolo Gisondi
3
, Domenico Biasi
1
and
Dennis McGonagle
4
Abstract
Objective. Dermatologists usually see patients with psoriasis before arthritis develops, making them well
placed to diagnose early PsA (ePsA). This study aimed to develop a rapid and robust screening ques-
tionnaire for predicting PsA in patients with psoriasis referred to a specialized joint dermatologyrheuma-
tology combined clinic.
Methods. In all, 228 psoriasis patients naı
¨ve to DMARD treatment were administered two screening
questionnaire: the new Early ARthritis for Psoriatic patients (EARP) questionnaire and the existing
Psoriatic Arthritis Screening and Evaluation (PASE) questionnaire. The diagnostic accuracy of the two
questionnaires for the diagnosis of ePsA was compared by receiving operating characteristics curves.
Results. After psychometric analysis, a simplified questionnaire of 10 items was found to have good
internal reliability (Cronbach’s a= 0.83) and was much faster and simpler to administer than the PASE.
Both the EARP and PASE questionnaires presented similar receiving operating characteristics curves
(specificity 91.6 and 67.2 and sensitivity 85.2 and 90.7, respectively) in identifying ePsA patients by
using the cut-off value of 3 for EARP-10 and the standard cut-off value of 44 for PASE. The CASPAR
criteria for PsA were present in 61 (26.7%) of the patients at clinical presentation and in 32.9% at 1-year
follow-up, and the EARP score of 53 correlated with clinically determined arthropathy by a
rheumatologist.
Conclusion. The EARP questionnaire is simple and fast to administer and proved robust for the identi-
fication of PsA in the dermatological setting. Dermatologists should consider the EARP for patients
attending clinics, as it correlates well with early PsA diagnosis.
Key words: screening questionnaire, ePsA, PASE, psoriasis.
Introduction
PsA is an inflammatory arthritis associated with psoriasis
that can have either an indolent or rapidly progressive
course. The exact proportion of patients with psoriasis
who will develop PsA is an area of significant controversy,
with studies demonstrating a range from as low as 6% to
as high as 42% [1, 2]. Typically psoriasis patients initially
seek treatment for their skin and then are referred to a
rheumatologist as arthritic symptoms develop. However,
concurrent PsA in psoriasis patients may be overlooked in
dermatology and general medicine practices, particularly
in the early stages of the disease, as enthesitis-related
manifestations at clinically inaccessible sites may be dif-
ficult to recognize, and inflammatory markers may remain
normal.
Several factors contribute to the delay in diagnosis of
PsA. These include the lack of awareness among patients
1
Unit of Rheumatology,
2
Department of Public Health and Medicine,
Unit of Epidemiology and Medical Statistics,
3
Department of Medicine,
Section of Dermatology and Venereology, University of Verona,
Verona, Italy and
4
Section of Musculoskeletal Diseases, NIHR Leeds
Biomedical Research Unit, Leeds Institute of Molecular Medicine,
University of Leeds, Leeds, UK.
Correspondence to: Ilaria Tinazzi, Unit of Rheumatology, University of
Verona, P. le Scuro, 1 37134 Verona, Italy. E-mail: ilariatinazzi@yahoo.it
Submitted 5 January 2012; revised version accepted 1 June 2012.
!The Author 2012. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com
RHEUMATOLOGY
Rheumatology 2012;51:20582063
doi:10.1093/rheumatology/kes187
Advance Access publication 9 August 2012
CLINICAL
SCIENCE
by guest on December 1, 2016http://rheumatology.oxfordjournals.org/Downloaded from
of the relationship between skin symptoms and joint
symptoms and the absence of a specific diagnostic
marker. Ideally every psoriasis patient with musculoskel-
etal pain should be evaluated by a rheumatologist, but this
is not practical. Accordingly, several groups developed
screening questionnaires for use in the dermatology or
general practice setting to identify individuals that might
be at high risk for the development of PsA [35]. These
include the Toronto Psoriatic Arthritis Screen (ToPAS)
screening questionnaire, the Psoriatic Arthritis Screening
and Evaluation (PASE) questionnaire, the Psoriasis and
Arthritis Questionnaire [35] and the Psoriasis
Epidemiology Screening Tool (PEST) questionnaire [6].
None of these questionnaires is focused on early diagno-
sis of PsA, and the reference population included patients
already on DMARDs. The ToPAS screening questionnaire
[5] was also designed for family medicine to help to rec-
ognize psoriasis, nail lesions and joint manifestations. The
other two questionnaires are focused more on the preva-
lence of joint disease among patients with psoriasis.
Peloso et al. [3] developed the Psoriasis and Arthritis
Questionnaire, a 12-item questionnaire in a cohort of
108 patients, and they found a sensitivity of 0.85 and spe-
cificity of 0.88 in predicting PsA for a score of 57. By
testing the same questionnaire in a larger cohort of psor-
iasis patients, Alenius et al. [3] could not confirm the same
diagnostic accuracy, and the cut-off was set to 4.
Ibrahim et al. [6] developed a five-question question-
naire—the PEST—starting from a community-based
survey of people with psoriasis and found a sensitivity of
0.92 and specificity of 0.78 in predicting PsA.
Recently Dominguez et al. [4] developed the PASE
questionnaire, which consists of symptom and function
subscales. This was predominantly designed to score
joint involvement for the early identification of PsA.
However, in our hands, the PASE was cumbersome for
patients and rather time consuming. In the present study
we evaluated a cohort of psoriasis cases using a screen-
ing questionnaire that was progressively simplified without
loosing diagnostic accuracy. We undertook this in a joint
dermatologyrheumatology early psoriasis clinic, and we
report the development of a new questionnaire—Early
Arthritis for Psoriatic Patients (EARP)—that is very user
friendly and easy to administer. We compared the EARP
with the PASE, as the latter existing questionnaire was
developed for screening potential early PsA (ePsA).
Herein we report that the EARP is a new rapid screening
method for the identification of PsA in the psoriasis clinic
of Verona University Hospital.
Patients and methods
Study design
The study was done in a randomized, investigator-blinded
design. Patients completed both the EARP (initial format
having 14 questions) and the Italian translation of the
PASE questionnaire, both of which were self-administered
before the patient was clinically evaluated by the
rheumatologist. The two questionnaires were adminis-
tered in a random sequence (1/1).
The PASE questionnaire [4] consists of two subscales:
symptoms and function. The back-translation from Italian
to English yielded no appreciable differences. The pos-
sible answers consist of five categories (15) related to
the level of agreement with the statement (strongly
agree to strongly disagree). A total score was calculated
by summing the score for each question, with a range of
1575.
The EARP questionnaire was developed through review
of the typical symptoms and signs seen among patients
with an established diagnosis of PsA. Fourteen questions
were jointly selected by a group of experts of the Verona
University. Patients were asked about possible recent epi-
sodes of joint swelling, enthesitis, dactylitis, and back and
hand morning stiffness occurring during the past 12
months. The questionnaire was composed of dichotom-
ous (yes/no) items; the total score was calculated by sum-
ming the score of each question.
The time taken by the patient to complete the PASE and
EARP questionnaires (14 items) was measured. The study
was approved by the Ethics Boards of the University of
Verona and written consent was obtained from all cases.
Patients
The patients were approached to enroll in the study while
in the waiting room of the psoriasis clinic of the University
of Verona. This clinic is jointly staffed by both dermatolo-
gists and rheumatologists. GPs in the Verona region typ-
ically refer the most severe cases and possibly those with
psoriasis and joint symptoms to this particular outpatient
clinic. Adults between the ages of 18 and 85 years, with an
established diagnosis of psoriasis that were naı
¨ve to sys-
temic treatment with a DMARD, were enrolled. We
excluded individuals known to be affected by long-stand-
ing PsA or with a concomitant diagnosis of RA, gout and
other rheumatic diseases.
Rheumatological assessment
After filling out the questionnaires, all patients underwent
blood tests (CRP, ESR, RF, ACPA) and a full clinical, radio-
graphic assessment independently of knowledge of the
questionnaire findings. The clinical examination was per-
formed by an experienced rheumatologist. This included
the 68-joint count for swelling and tender joints. The pres-
ence of possible enthesitis was assessed by the
Maastricht Ankylosing Spondylitis Enthesis Score
(MASES) index [7]. Patients were also questioned about
the presence of previous bouts of dactylitis, namely,
a history of sausage-like digital swelling in the previous
6 months. The Classification Criteria for Psoriatic
Arthritis (CASPAR) diagnostic criteria were applied [8].
Statistical analysis
Statistical analyses were performed using the SPSS 17.0
(Chicago, IL, USA) and STATA10.1 software packages.
Continuous data with normal distribution were expressed
as mean (S.D.); data with normal distribution with the
www.rheumatology.oxfordjournals.org 2059
EARP questionnaire to detect early PsA
by guest on December 1, 2016http://rheumatology.oxfordjournals.org/Downloaded from
ShapiroWilk test were expressed as median (interquartile
range). P<0.05 was considered significant.
Principal component analysis
To identify the number of the underlying components of
the initial EARP-14 questionnaire, the Pearson’s correl-
ation matrix was explored by means of principal compo-
nent analysis (PCA) [9]; items weakly correlated with the
others (r<0.20) were excluded from further analysis. The
number of components was determined on the basis of
eigenvalues of the correlation matrix >1. Item-component
correlations, i.e component loading, >0.40, in absolute
value, were chosen to identify a simple component struc-
ture. At the end of this analysis, the component number
was forced to one to obtain a single scale.
Multiple correspondence analysis phase
The multiple correspondence analysis phase (MCA) or
homogeneity analysis (HOMALS) [10] is designed to test
item internal homogeneity and reliability for each dimen-
sion of the PCA. This method uses the yes/no answers as
nominal category responses, and enables optimal grading
for each category response of the questions (called opti-
mal weights); consequently, an optimal score for each
subject could be obtained. The optimal score of a subject
is the sum of the optimal weights of the item options
chosen. According to the criterion of internal consistency,
MCA computes a homogeneity index (called Guttman’s
Z), and the optimal reliability determined by Cronbach’s
acoefficient is a one-to-one transformation of Guttman’s
Z[11]. Good internal consistency has been suggested if
Cronbach’s a>0.70, and good homogeneity has been
suggested if Guttman’s Z>0.30 [11]. Correlations be-
tween each item and the total score of the questionnaire,
the total score excluding that specific item (itemrest) and
between the optimal item and the total score were
calculated.
Receiver operating characteristic curves
The receiver operating characteristic (ROC) curves
were constructed to investigate the diagnostic perform-
ance of PASE and of the newly developed EARP ques-
tionnaire. The area under the curve (AUC) provided a
measure of the overall discriminative ability of the predic-
tion rule. The ROC curves were used also to identify pos-
sible discriminatory cut-off points for the questionnaires
using the Youden test. The sensitivity and specificity were
determined for several cut-off values of the prediction
score.
Psoriasis patient follow-up
Patients with psoriasis but without a diagnosis of PsA at
baseline according to the CASPAR criteria were once
again evaluated after 1 year to ascertain whether clinical
PsA had evolved. The clinical examination was performed
by an experienced rheumatologist applying the CASPAR
criteria [8]. The baseline EARP scores were compared
with the total numbers of cases with PsA at the end of
this period.
Results
Patient assessment
The initial study population included 386 patients;
85 patients previously treated with DMARDs, 42 affected
by long-standing PsA and 31 with concomitant rheumatic
disease were excluded as per protocol (Fig. 1). The 228
remaining patients had never been previously investigated
by a rheumatologist, and all of them completed both
PASE and EARP questionnaires. The CASPAR criteria
were present in 61 (26.7%) of the patients at clinical pres-
entation. The clinical characteristics of these patients are
listed in Table 1.
All patients had a further rheumatological assessment
at the 3-month interval, and this allowed the identification
of 14 additional cases (6.1%) of ePsA by 12 months. The
clinical characteristics of this second group of patients are
also listed in Table 1.
Questionnaire completion and EARP refinement
The original items of the EARP questionnaire comprised
14 items (the EARP-14). The PASE questionnaire was
completed in 6 (2) [mean (SD)] min, whereas the
EARP-14 required 2 (1.5) min. The correlation matrix
among the questions of the EARP-14 revealed that four
questions (q2, q4, q12, q13) were weakly correlated
(r<0.20) with the others and were then excluded from
the PCA (Table 2). The PCA identified two dominant com-
ponents (eigenvalues >1), explaining 54% of the total
variance, and 75% of the pair-wise correlations of the
10 questions examined. As all items were binary, compo-
nent loading patterns are equal to optimal itemtotal cor-
relation (Table 2). This component includes the 10 items
composing the final 10-item EARP questionnaire
(EARP-10), henceforth referred to as the EARP (Table 3).
The four questions removed were as follows: Do you feel
that your joints are limiting your movement? Is your back
stiff for >15 min when you wake up? Do you have plantar
burning and pain when you walk? and Do your knees
swell? The internal criterion index (Guttman’s Z) of the
MCA/HOMALS scaling was equal to 0.394, and the
Cronbach’s acoefficient was 0.83, indicating a good scal-
ing [11]. The raw itemtotal score correlations varied from
0.79 (q1) to 0.48 (q10) (Table 2); the same trend was found
for the raw itemrest correlation from 0.71 (q1) to 0.36
(q10) and for the optimal itemtotal score correlations
from 0.81 (q1) to 0.46 (q10). This suggests that the sum
row scores or the sum optimal scores were good [10]
values of the true scores of the continuous uni-
dimensional factor underlying the 10-item set and are
useful to detect ePsA.
The ROC curves were constructed to investigate the
diagnostic performance of PASE and EARP (Fig. 2). For
EARP, a cut-off point of 3 was found to be associated with
the best diagnostic performances, whereas for PASE, the
cut-off of 44 defined by the authors of the questionnaire
was used. The measured AUC was 0.895 [standard error
(S.E.) 0.030; 95% CI 0.836, 0.954] for PASE, whereas the
2060 www.rheumatology.oxfordjournals.org
Ilaria Tinazzi et al.
by guest on December 1, 2016http://rheumatology.oxfordjournals.org/Downloaded from
AUC of EARP was slightly higher at 0.906 (S.E. 0.025; 95%
CI 0.857, 0.955) (Fig. 2).
The EARP false-positive rate (i.e. with a score 53 but
without PsA) was 22.3% (19 of 85 patients), and the
false-negative rate was 3.5% (5 of 143 patients); i.e.
ePsA patients with a score <3 on the EARP but with
PsA. This yields a sensitivity of 85.2% and a specificity
of 91.6%. The sensitivity and specificity of the PASE was
90.7% (95% CI 54.3, 78.4) and 67.2% (95% CI 79.7,
96.9), respectively. Sensitivity and specificity rose to
87.5% (95% CI 76.8, 94.4) and to 81.5% (95% CI 68.5,
90.7), respectively, with a cut-off equal to 36 rather than
44. In the 14 patients who developed PsA within the
following year, the baseline EARP score was >3 in 10,
whereas a PASE score >44 was present in only 5 of the
patients.
Discussion
In this study we have shown that a large proportion of
patients with psoriasis exhibit previously unrecognized
signs of ePsA, and that this can be reliably collected
using the new EARP screening questionnaire. Moreover,
the EARP was extremely quick to administer in
FIG.1Study population flow chart.
A total of 386 patients affected by psoriasis (PsO) were included; 85 patients previously treated with DMARDs, 42
affected by long-standing PsA and 31 with concomitant rheumatic disease were excluded.
TABLE 1Baseline characteristics of patients
PsA (n = 71) PsO (n = 147) Incident PsA (n = 14) P
Age at inclusion, mean (S.D.), years 50.1 (11.4) 49.2 (12.9) 45.8 ± 8.4 0.31
Sex, women, % 50.7 47.9 35.5 0.01
BMI, median (range) 26 (1831) 25 (1936) 27 (2034) 0.58
Clinical presentation, n (%)
Polyarthritis (53) 14 (19.4)
Oligoarthritis (<3) 22 (30.5)
Dactylitis 11 (15.3)
Spinal involvement 2 (2.8)
MASES, median (IQR) 2 (04) 0 (01) 1 (03) 0.01
TJ 68, median (IQR) 4 (210) 0 (02) 1 (02) 0.01
SJ 68, median (IQR) 2 (05) 0 (01) 0 (01) 0.05
BASDAI, median (IQR) 3.3 (2.19.7) 1.4 (0.73.2) 1.6 (0.84) 0.04
PASI, mean (S.D.) 6.8 (3.18) 5.8 (2.88) 6.2 (2.54) 0.18
Onicopathy, n (%) 41 (56.9) 65 (45.7) 8 (57.1) 0.22
Baseline characteristics of the 228 patients included in the study. All these patients completed EARP and PASE questionnaires
and were followed over 1 year. PsO: psoriasis, IQR: interquartile range, TJ: tender joints, SJ: swollen joints; PASI: Psoriasis
Area and Severity Index.
www.rheumatology.oxfordjournals.org 2061
EARP questionnaire to detect early PsA
by guest on December 1, 2016http://rheumatology.oxfordjournals.org/Downloaded from
the psoriasis population and was useful in a real-world
setting of psoriasis patients presenting in the dermatology
clinical setting for the first time. Moreover, patients with
symptoms without clinical PsA often progressed to PsA
within the following year. The relatively large number of
cases recruited and the longitudinal follow-up for 1 year
in a dermatological setting suggest that the EARP could
be a useful screening tool for ePsA. This needs validation
in a new cohort of psoriasis cases.
Twenty-six per cent of patients at baseline fulfilled the
CASPAR criteria; such high prevelance rates of PsA
in psoriasis have been previously reported by others
[2, 12]. However, we cannot rule out a kind of referral
bias, given that ours is a joint dermatologyrheumatology
psoriasis clinic that might attract a larger proportion of
patients with joint problems. For the primary purpose of
this study, the development of a PsA screening question-
naire, such a referral pattern would be useful. The clinical
signs of the PsA appeared within a year in an additional
6.1% of patients. This is much more than that reported in
epidemiological surveys, where only 10% of cases de-
velop PsA after a decade of follow-up [1]. However, in
previous studies the incidence of PsA was reported,
rather than assessed, and this inevitably might consider-
ably underestimate the true disease incidence.
Our results underline the importance of periodic
rheumatology assessment, or at least rheumatic symptom
evaluation, among psoriasis patients with tools such as
EARP.
A delayed diagnosis of PsA may result in joint destruc-
tion and permanent disabilities, whereas early diagnosis
and prompt therapy could prevent this irreversible joint
damage [13]. The EARP questionnaire we developed
was able to identify most of the PsA patients. The EARP
questionnaire revealed an internal homogeneity, as
Guttman’s Zwas equal to 0.394, and >0.30 is considered
the cut-off for high homogeneity. Cronbach’s awas 0.83,
denoting very good reliability [11] and indicating a low
degree of error in measuring the characteristic of interest.
The question with the best itemtotal and optimal
itemtotal correlation was q1, followed by q7, which are
the most generic questions about general joint pain or
hand joint involvement, respectively.
The sensitivity and specificity were 91% and 85%,
respectively, for the cut-off value of 3.
In this study, the diagnostic accuracy of EARP was
somewhat superior to that achieved by PASE with the
cut-off score of 44 identified by the authors [4], and it
increased slightly when the cut-off was optimized to 36
(data not shown). However, the main problem for PASE
remains its complexity, which makes it unsuitable for a
quick survey in outpatient dermatology clinics. The EARP
can be administered or even self-administered within a
couple of minutes, which favourably compares with
PASE requiring >5 min with a trained specialist.
The authors of the PEST questionnaire [6] used 5 items
from their original 16 questions. This questionnaire also
included questions about familiarity, patient’s thoughts
and nail assessment to identify PsA in general practice,
and the authors did not exclude known PsA cases [6]. The
EARP questions focused specifically on patients’ symp-
toms to identify new diagnoses of PsA in a dermatological
clinical setting.
An obvious limitation of this study is the lack of valid-
ation in a different cohort of patients. This will be per-
formed in future EARP validation studies. An additional
limitation is the language of the questionnaire, which will
need validation in other languages. However, the ques-
tions are extremely simple and intuitive and problems in
transferring the results into different languages appear
unlikely. In conclusion, the EARP questionnaire we de-
veloped for identifying psoriasis patients with ePsA is ex-
tremely simple and can be self-administered. Its
sensitivity and specificity in identifying psoriasis
patients who should be referred for a rheumatological
TABLE 3The EARP questionnaire
Question Yes No
Do your joints hurt? 1 0
Have you taken anti-inflammatory more than
twice a week for joint pain in the last 3 months?
10
Do you wake up at night because of low back
pain?
10
Do you feel stiffness in your hands for more than
30 minutes in the morning?
10
Do your wrists and fingers hurt? 1 0
Do your wrists and fingers swell? 1 0
Does one finger hurt and swell for more than
3 days?
10
Does your Achilles tendon swell? 1 0
Do your feet or ankles hurt? 1 0
Do your elbow or hips hurt? 1 0
The final 10 items of the EARP questionnaire were chosen
after the psychometric analysis. Each positive answer is
scored as one, and the final score is calculated by summing
the positive answers.
TABLE 2Raw itemtotal, raw itemrest and optimal
itemtotal correlations of EARP
Question no.
Raw
itemtotal
correlation
Raw
itemrest
correlation
Optimal
itemtotal
correlation
1 0.791 0.711 0.811
2 0.687 0.582 0.682
3 0.511 0.394 0.486
4 0.616 0.503 0.615
5 0.733 0.635 0.744
6 0.666 0.554 0.674
7 0.492 0.382 0.493
8 0.477 0.364 0.456
9 0.662 0.546 0.647
10 0.590 0.464 0.568
2062 www.rheumatology.oxfordjournals.org
Ilaria Tinazzi et al.
by guest on December 1, 2016http://rheumatology.oxfordjournals.org/Downloaded from
work-up appears to be as accurate as more complex
tools.
Rheumatology key messages
.ePsA identification is a difficult and controversial
topic.
.Many screening questionnaires to detect PsA are
not used by dermatologists in everyday practice
because of their complexity.
.The EARP is fast and easy for dermatologists to
use to identify early symptoms of PsA.
Disclosure statement: The authors have declared no
conflicts of interest.
References
1 Wilson FC, Icen M, Crowson CS et al. Time trends in
epidemiology and characteristics of psoriatic arthritis over
3 decades: a population-based study. J Rheumatol 2009;
36:3617.
2 Gisondi P, Girolomoni G, Sampogna F, Tabolli S, Abeni D.
Prevalence of psoriatic arthritis and joint complaints in a
large population of Italian patients hospitalized for psor-
iasis. Eur J Dermatol 2005;15:27983.
3 Alenius GM, Stenberg B, Stenlund H, Lundblad M,
Dahlqvist SR. Inflammatory joint manifestations are
prevalent in psoriasis: prevalence study of joint and axial
involvement in psoriatic patients, and evaluation of psori-
atic and arthtitic questionnaire. J Rheumatol 2002;29:
257782.
4 Dominguez PL, Husni ME, Holt EW, Tyler S, Qureshi AA.
Validity, reliability, and sensitivity-to-change propertiea of
the psoriatic arthritis screening and evalutation
questionnaire. Arch Dermatol Res 2009;301:57379.
5 Gladman DD, Schentag CT, Tom BD et al. Developement
and initial validation of a screening questionnaire for
psoriatic arthritis: the Toronto Psoriatic Arhritis
Screen (ToPAS). Ann Rheum Dis 2009;68:497501.
6 Ibrahim GH, Buch MH, Lawson C, Waxman R,
Helliwell PS. Evaluation of an existing screening tool for
psoriatic arthritis in people with psoriasis and the devel-
opment of a new instrument: the Psoriasis Epidemiology
Screening Tool (PEST) questionnaire. Clin Exp Rheumatol
2009;27:46974.
7 Kaya A, Ozgocmen S, Kamanli A, Aydogan R, Yildirim A,
Ardicoglu O. Evaluation of a quantitative scoring of
enthesitis in ankylosing spondylitis. J Clin Reumatol 2007;
13:3036.
8 Taylor W, Gladman D, Helliwell P, Marchesoni A, Mease P,
Mielants H, and the CASPAR Study Group. Classification
criteria for psoriatic arthritis: development of new criteria
from a large international study. Arthritis Rheum 2006;54:
266573.
9 Gifi A, ed. Nonlinear multivariate analysis. Chichester:
Wiley & Sons, 1990.
10 Gorsuch RL. Factor Analysis, 2nd edn. Hillsdale, NJ:
Lawrence Erlbaum, 1983.
11 Fayers PM, Machin D. Quality of life: the assessment,
analysis and interpretation of patient-reported outcomes,
vol. 123. Chichester: Wiley & Sons, 2007.
12 Zachariae H. Prevalence of joint disease in patients with
psoriasis: implications for therapy. Am J Clin Dermatol
2003;4:4417.
13 Kane D, Stafford L, Bresnihan B, FitzGerald O. A pro-
spective, clinical and radiological study of early psoriatic
arthritis: an early synovitis clinic experience.
Rheumatology 2003;42:14608.
FIG.2ROC curves of EARP items (A) and the PASE questionnaire (B).
Area under the curve of PASE is 0.884 (S.E. 0.031; 95% CI 0.823, 0.9). Area under the curve of EARP-10 is 0.903
(S.E. 0.025; 95% CI 0.854, 0.952).
www.rheumatology.oxfordjournals.org 2063
EARP questionnaire to detect early PsA
by guest on December 1, 2016http://rheumatology.oxfordjournals.org/Downloaded from
