Académique Documents
Professionnel Documents
Culture Documents
doi: 10.1111/j.1440-0960.2011.00800.x
ORIGINAL RESEARCH
259..263
Philip Singh,1 H. Peter Soyer,1,2 Jason Wu,1,2 Wolfgang Salmhofer3 and Stephen Gilmore1
1
Dermatology Research Centre, The University of Queensland, School of Medicine, and 2Department of
Dermatology, Princess Alexandra Hospital, Brisbane, Australia; and 3Department of Dermatology,
Medical University of Graz, Graz, Austria
ABSTRACT
Background: The implementation of remote Psoriasis Area and Severity Index (PASI) determinations
would greatly enhance the delivery of specialist dermatological care to patients with severe psoriasis
unable to attend face-to-face dermatological consultations. Here we investigate the feasibility of
the remote determination of PASI scores by comparing the results of face-to-face with digital image
assessment.
Methods: Twelve patients with confirmed psoriasis
were recruited for the study. Initially, two dermatologists scored the PASI at the patients usual scheduled
face-to-face visits, at which time standardized digital
images were obtained. PASI scoring based on digital
images was then performed on two separate occasions by three dermatologists with a time-interval
period between assessments, facilitating an assessment of score reproducibility. Linear weighted kappa
statistics were applied to the PASI scores to ascertain
agreement between sets of observations.
Results: While we found a moderate (k = 0.51)
agreement between the face-to-face scores, there was
very good (k = 0.83) agreement between the first
round of telescores and moderate (k = 0.60) agreement between the second round of telescores. Comparison between the face-to-face and telescores
revealed good (k = 0.67 and 0.63) agreement for the
scorers respectively.
Conclusion: We demonstrate that PASI scores can
be determined with moderate to good accuracy by
INTRODUCTION
Psoriasis is a common, chronic, papulosquamous inflammatory skin condition of varying severity, affecting up to 2%
of Western populations.1 Management demands individualized treatment approaches, and while most patients will
only ever require topical therapy, a significant number will
require systemic intervention.1,2 These patients undergoing
systemic therapy require ongoing clinical evaluation and
monitoring for response as well as any development of drug
toxicities.2 With the introduction of government-funded
potent immunomodulating agents, otherwise known as
biologics, the determination of Psoriasis Area and Severity
Index (PASI) scores has become a routine part of dermatologists workload.25 The PASI score is used as the gold
standard for assessing efficacy of systemic agents in the
majority of recently published data on the treatment of psoriasis. Indeed, for patients with severe psoriasis requiring
ongoing treatment with biological therapy, the determination of PASI scores is mandatory in order to objectively
document efficacy.6
Given the documented successes of teledermatology in
general,7 we are motivated to investigate the feasibility of
tele-PASI the remote determination of PASI scores in
patients with psoriasis. For incapacitated patients with
severe psoriasis, and for patients with severe psoriasis where
distances are prohibitive, the implementation of a tele-PASI
service would greatly augment their dermatological care by
permitting the remote introduction or continuation of biological therapy. In addition, such a service would lessen the
260
P Singh et al.
strain on government health budgets since the costs of transport and accommodation would be greatly reduced.8,9
Table 1
Patient
Age
Sex
Site
Psoriasis Type
Mean
PASI score
1
2
3
4
5
6
7
8
9
10
11
12
55
63
27
46
39
22
33
36
29
43
33
58
Male
Male
Male
Male
Female
Male
Male
Female
Female
Male
Male
Male
PAH
QID
QID
QID
PAH
QID
PAH
PAH
QID
QID
QID
QID
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
Chronic
12.5
27.1
5.2
11.1
5.2
13.9
36.8
21.4
20.2
24.9
11.9
15.5
Plaque-Type
Plaque-Type
Plaque-Type
Plaque-Type
Plaque-Type
Plaque-Type
Plaque-Type
Plaque-Type
Plaque-Type
Plaque-Type
Plaque-Type
Plaque-Type
Figure 1 Example of the digital images used in the calculation of Psoriasis Area and Severity Index (PASI) scores. For this patient, the PASI
scores are represented by the red disk in each panel of Figures 2 and 3, 4 and 5.
2011 The Authors
Australasian Journal of Dermatology 2011 The Australasian College of Dermatologists
261
(a)
(b)
(c)
(d)
(e)
(f)
Figure 3 Interobserver Psoriasis Area and Severity Index (PASI) scores for the first and second photo evaluations for clinicians 1 and 2 (a,
d), clinicians 1 and 3 (b, e), and clinicians 2 and 3 (c, f ). The red disk in each panel corresponds to the patient shown in Figure 1. Here the
respective agreements are given by k = 0.87 (a), k = 0.88 (b), k = 0.74 (c), k = 0.56 (d), k = 0.72 (e) and k = 0.51 (f ).
262
P Singh et al.
to ~6 as equivalent. Following standard protocol, we interpreted kappa agreement scores as follows:11,12 0.0 k < 0.2
(poor), 0.2 k < 0.4 (fair), 0.4 k < 0.6 (moderate),
0.6 k < 0.8 (good), 0.8 k < 1.0 (very good).
(a)
(b)
(c)
(d)
RESULTS
We first examined interobserver agreement using the faceto-face PASI scores (Fig. 2), followed by the first and second
(Fig. 3) digital image PASI scores. We found a moderate
agreement between clinician 1 and clinician 2 who were
involved in the face-to-face assessments (k = 0.51). While
the first round of digital PASI scores revealed an average
kappa value of 0.83, indicating a very good agreement, the
agreement on the second round was moderate (k = 0.60).
Averaging all statistics revealed good overall agreement of
PASI scores between all three dermatologists (k = 0.68).
We next examined the intra-observer agreement of PASI
scores, first by comparing the face-to-face scores with the
first and second (Fig. 4) rounds of digital image PASI scores,
and then by comparing the results of the first and second
rounds of digital image PASI scores (Fig. 5). We found moderate and good agreement between the face-to-face and the
first round of digital image PASI scores (k = 0.41 and 0.64 for
clinician 1 and clinician 2, respectively), and moderate and
good agreement between the face-to-face and the second
round of digital image PASI scores (k = 0.55 and 0.71 for
clinician 1 and clinician 2, respectively). Finally, we found
good, very good and good agreement between the first and
second rounds of digital image PASI scores (k = 0.72, 0.93
and 0.65) for clinicians 1, 2 and 3, respectively).
DISCUSSION
(a)
(b)
(c)
Figure 5 Intra-observer Psoriasis Area and Severity Index (PASI) scores for the first and second photo evaluations for clinician 1 (a),
clinician 2 (b), and clinician 3 (c). The red disk in each panel corresponds to the patient shown in Figure 1. Here the respective agreements
are given by k = 0.72 (a), k = 0.93 (b) and k = 0.65 (c).
2011 The Authors
Australasian Journal of Dermatology 2011 The Australasian College of Dermatologists
263
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.