BJD

Q U A L I TA T I V E A N D O U T CO M E S R E S E A R CH British Journal of Dermatology

Severity strata for Eczema Area and Severity Index (EASI),

modified EASI, Scoring Atopic Dermatitis (SCORAD), objective
SCORAD, Atopic Dermatitis Severity Index and body surface
area in adolescents and adults with atopic dermatitis
R. Chopra,1 P.P. Vakharia,1 R. Sacotte,1 N. Patel,1 S. Immaneni,1 T. White,2 R. Kantor,1 D.Y. Hsu1
and J.I. Silverberg iD 1,2
1
Department of Dermatology, Feinberg School of Medicine at Northwestern University, Chicago, IL 60611, U.S.A.
2
Northwestern Medicine Multidisciplinary Eczema Center, Chicago, IL, U.S.A.

Linked Comment: Drucker. Br J Dermatol 2017; 177:1158–1159.

Summary

Correspondence Background Scoring systems for assessing the signs of atopic dermatitis (AD) are
Jonathan I. Silverberg. complex and difficult to interpret. Severity strata are helpful to interpret these
E-mail: JonathanISilverberg@Gmail.com assessments properly.
Objectives To confirm previously reported strata for the Eczema Area and Severity Index
Accepted for publication
27 April 2017
(EASI), Scoring Atopic Dermatitis (SCORAD) and the objective component of SCORAD
(oSCORAD), and to develop strata for the modified EASI (mEASI), Atopic Dermatitis
Funding sources Severity Index (ADSI) and body surface area (BSA) for use in adults with AD.
This publication was made possible with support Methods Skin examination was performed in 673 adolescents and adults (age
from the Agency for Healthcare Research and ≥ 13 years) with diagnosed AD, in a dermatology practice setting. Strata were selected
Quality (AHRQ), grant number K12 HS023011,
using an anchoring approach based on a four-point Investigator’s Global Assessment
and the Dermatology Foundation.
of severity (clear of active skin lesions, mild, moderate or severe disease).
Conflicts of interest Results We determined potential severity strata for EASI (0 clear, 0
1–5
9 mild, 6
0–
None declared. 22
9 moderate, 23
0–72 severe; j = 0
69), mEASI (0–0
9 clear, 1–8
9 mild, 9
0–
29
9 moderate, 30
0–90 severe; j = 0
71), oSCORAD (0–7
9 clear, 8
0–23
9
DOI 10.1111/bjd.15641 mild, 24
0–37
9 moderate, 38
0–83 severe; j = 0
70), SCORAD (0–9
9 clear,
10
0–28
9 mild, 29
0–48
9 moderate, 49
0–103 severe; j = 0
68), ADSI (0–1
9
clear, 2–5
9 mild, 6
0–8
9 moderate, 9
0–15 severe; j = 0
55) and BSA (0 clear,
0
1–15
9 mild, 16
0–39
9 moderate, 40
0–100 severe; j = 0
66). oSCORAD val-
ues > 0 were found in clear skin due to the presence of xerosis, which is scored in
oSCORAD. Similarly, SCORAD values > 0 were found in clear skin due to the scor-
ing of xerosis, pruritus and sleeplessness. Similarly, mEASI and ADSI scores > 0
occurred in patients with clear skin due to scoring of pruritus.
Conclusions We recommend using these strata for interpretation of their respective
measures in clinical trials of AD. There are important differences between the five
assessments, which profoundly impact the interpretation of their scores.

What’s already known about this topic?

• Multiple assessments have been used for the signs of atopic dermatitis. However,
scores from these instruments can be difficult to interpret.

What does this study add?

• The present study developed novel severity strata for the Eczema Area and Severity
Index (EASI), modified EASI (mEASI), Scoring Atopic Dermatitis (SCORAD), the

1316 British Journal of Dermatology (2017) 177, pp1316–1321 © 2017 British Association of Dermatologists

objective component of SCORAD, Atopic Dermatitis Severity Index and body sur-
face area in a cohort of adults with atopic dermatitis.
What are the clinical implications of this work?
• The strata presented for these measures can be used to improve interpretation of
clinical and research data, and have important ramifications for future clinical trials
of atopic dermatitis.
Introduction
Atopic dermatitis (AD) can present with heterogeneous mani-
festations and varying degrees of severity. There is currently
no gold standard for evaluating the severity and extent of
AD.1 Severity assessment for AD is based on an assessment of
lesional severity and extent, and/or disease-related symptoms.
A recent systematic review by Harmonizing Outcome Mea-
sures in Eczema (HOME), an international consensus group
aimed at standardizing outcomes in clinical trials for AD, iden-
tified 56 different named and unnamed objective measure-
ments of disease severity.2 The recommended outcomes
assessments for the signs of AD were the Eczema Area and
Severity Index (EASI) followed by Scoring Atopic Dermatitis
(SCORAD), based on factors including demonstration of valid-
ity and good intra- and inter-rater reliability.1
However, these assessments have complicated scoring sys-
tems, with scores that can be difficult to interpret. EASI
assesses only active acute or chronic AD lesions, whereas
SCORAD also assesses xerosis, pruritus and sleeplessness. These
differences may result in differences of scores that can pro-
foundly impact the interpretation. Furthermore, modifications
of these scores have also been used in clinical trials. For exam-
ple, a modified version of EASI (mEASI) has been used, which
also includes patient report of pruritus. A modified version of
SCORAD including only the objective component (oSCORAD)
has been used in some trials. This may further confuse the
interpretation of these scores.
Severity strata have been used to improve the interpretation
of patient-reported outcomes in dermatology3–6 and have
recently been developed for EASI in AD.7 The proposed sever-
ity strata for oSCORAD (mild 0–14, moderate 15–39, severe
40–83) were developed by expert opinion without any formal
testing.8 The source for the commonly used SCORAD strata
(mild 0–25, moderate 26–50, severe 51–103) is unclear, but
appears to be a modification of the proposed oSCORAD strata.
Body surface area (BSA) of lesional involvement in itself is
an important aspect of AD severity. However, BSA as a mea-
sure of disease extent alone does not fully characterize AD, as
it does not account for lesional severity. Nevertheless, it has
been widely used as an inclusion criterion in AD trials with
variable definitions of severity, and also as a secondary efficacy
outcome measure in AD trials. Thus, proper interpretation of
BSA and its relation to AD severity is important. The Atopic
© 2017 British Association of Dermatologists
Severity strata for atopic dermatitis signs, R. Chopra et al. 1317
Dermatitis Severity Index (ADSI) has also been used to assess
the signs and symptoms of AD in clinical trials.9 Despite not
being a preferred assessment selected by the HOME consensus
group, ADSI was recently used in a randomized controlled
trial of a novel topical agent for AD.10,11 However, severity
strata have not been formally established for BSA and ADSI.
We sought to confirm previously proposed strata for EASI,
oSCORAD and SCORAD. Moreover, we sought to develop inter-
pretable severity strata for the mEASI, ADSI and BSA scales.
Materials and methods
Study design
We performed a prospective dermatology practice-based study of
adolescents and adults (age ≥ 13 years) with AD as defined by the
Hanifin and Rajka diagnostic criteria.12 Patients underwent full-
body skin examination by a dermatologist (J.I.S.) and were
assessed with Investigator’s Global Assessment (IGA) (a gestalt
physician’s assessment of clear of active acute or chronic AD
lesions, or mild, moderate or severe overall disease), EASI (four
signs – erythema, excoriation, swelling and lichenification – on
four body sites; range 0–72),13 mEASI (four signs – erythema,
excoriation, swelling and lichenification – on four body sites, and
pruritus; range 0–90),14 SCORAD (six signs – erythema, excoria-
tion, swelling, oozing/crusting, lichenification and dryness – on
eight body sites, and pruritus and sleeplessness; range 0–103),15
oSCORAD (six signs – erythema, excoriation, swelling, oozing/
crusting, lichenification and dryness – on eight body sites, and no
symptoms; range 0–83),15 ADSI (four signs – erythema, excoria-
tion, exudation and lichenification – and pruritus; range 0–15)9
and BSA (0–100%) affected with active lesions.
Patients were enrolled at different stages of disease activity,
and disease severity ranged from clear to severe. Patients
found to be clear upon assessment were previously confirmed
to have AD based on history and physical exam. The patients
were evaluated between January 2014 and June 2016. The
study was approved by the institutional review boards of
Northwestern University and informed consent was waived.
Data processing and statistical methods
An anchor-based approach was used, where the severity
assessments are compared, or anchored, to a global assessment
British Journal of Dermatology (2017) 177, pp1316–1321
1318 Severity strata for atopic dermatitis signs, R. Chopra et al.
performed at the same time.3 The score is then stratified using
thresholds into severity strata comparable with the global
score. EASI, mEASI, oSCORAD, SCORAD, ADSI and BSA were
stratified against the IGA (three point and four point), which
was used as the severity anchor. IGA was used as a gestalt
assessment of overall AD severity at that encounter, irrespec-
tively of prior disease history.
All data analyses and statistical processes were performed
using SAS version 9.4 (SAS Institute Inc., Cary, NC, U.S.A.).
Spearman rank correlation was performed to compare IGA
with EASI, mEASI, oSCORAD, SCORAD, ADSI and BSA scores.
The frequency, mean, median and mode of clear skin and
mild, moderate and severe disease were determined for each
integer value of the EASI, mEASI, oSCORAD, SCORAD, ADSI
and BSA scales. Potential strata were selected based on these
summary statistics. Different values for each assessment that
were associated with a one-grade increase in mean, median
and/or mode IGA were considered possible threshold values
for severity strata. Potential strata were then constructed using
combinations of these different threshold values. Weighted
j-coefficients were determined for concordance between the
potential severity strata and IGA. In addition, we tested any
strata that were previously reported for the above-mentioned
scales. Subgroup analyses were performed to determine how
many patients had IGA scores that were outside the final
severity strata for each measure.
Results
Patient characteristics
Overall, 673 adolescents and adults (aged 13–93 years) were
included in the study, including 401 female (59
6%) and 403
white patients (59
9%). The mean age at enrolment was
43
4  19
6 years, and the mean age of self-reported AD
onset was 35
2  35
9 years. In total, 111 patients (16.5%)
were clear of active acute or chronic lesions on skin exam.
Xerosis was observed in 36 (38%) and pruritus was reported
in 69 (72%) of those who were otherwise clear. Furthermore,
320 (49
0%) had mild, 215 (32
9%) moderate and 118
(18
1%) severe disease.
The distribution, mean, median and mode of AD severity
for each scale are presented in Tables S1–6 (see Supporting
Information). There were strong correlations of EASI (Spear-
man correlation, q = 0
85), mEASI (q = 0
85), oSCORAD
(q = 0
84), SCORAD (q = 0
83), ADSI (q = 0
72) and BSA
(q = 0
83) with both the three- and four-point IGA
(P < 0
001 for all).
Eczema Area and Severity Index
To illustrate the anchor-based approach, selection of potential
thresholds for moderate and severe AD will be elaborated. A
median and mode IGA score of 1 (mild) first occurred at an
EASI score of 1, whereas a median and mode IGA score of 2
(moderate) first occurred at an EASI score of 6, with a mean
British Journal of Dermatology (2017) 177, pp1316–1321
IGA of 1
8. EASI scores ≥ 7 all showed median and mode IGA
scores of 2 and mean IGA ≥ 1
6. Therefore, an EASI score of
6 was selected as the proposed threshold for moderate AD. In
addition, we tested an EASI score of 7 as a threshold for mod-
erate AD based on previously proposed severity strata.7 A
median and mode IGA score of 3 (severe) first occurred at an
EASI score of 17, with a mean IGA of 2
6. Therefore, an EASI
score of 17 was considered as a threshold for severe AD.
However, EASI scores of 18–22 were associated with median
and mode IGA scores of 2 and mean IGA scores < 2
5. An
EASI of 23 was the next score to be associated with a median
and mode IGA of 3, and a mean IGA of 2
8. Therefore, an
EASI of 23 was also considered as a potential threshold for
severe AD.
Four different strata were tested based on the combinations
of thresholds identified for moderate and severe disease. The
set of strata with the highest j-coefficient compared with a
four-point IGA (clear, mild, moderate, severe) was EASI of 0
clear, 1–5
9 mild, 6
0–22
9 moderate and 23
0–72 severe
(j = 0
69) (Table 1). A collapsed version of these strata (0–
5
9 clear–mild, 6
0–22
9 moderate, 23
0–72 severe) also had
the highest j-coefficient when compared against a three-point
IGA (clear-mild, moderate, severe) (j = 0
76). Previously
reported strata of 0–7
0, 7
1–21
0 and 21
1–727 had the low-
est j-value of all strata tested in this cohort of adults with AD
(j = 0
73).
Only seven patients (1
1%) had an IGA score > 1 point out-
side of that predicted by the final EASI severity strata. There
were 82 patients (12
6%) whose IGA score was 1 point
higher and 27 (4
1%) whose IGA score was 1 point lower
than the final EASI strata predicted. Patients falling outside the
possible strata had a similar distribution of sex, age and race/
ethnicity to the overall study cohort.
Modified Eczema Area and Severity Index
Four different strata were selected for testing. The set of strata
with the highest j-coefficient when compared against a four-
point IGA was mEASI of 0 clear, 1–8
9 mild, 9
0–29
9 mod-
erate and 30
0–90 severe (j = 0
71). Similarly, a collapsed
version of these strata (0–8
9 clear–mild, 9
0–29
9 moderate,
30
0–90 severe) had the highest j-value compared against a
three-point IGA (j = 0
76).
Only seven patients (1
1%) had an IGA score > 1 point out-
side of that predicted by the final mEASI severity strata. There
were 101 patients (15
6%) whose IGA score was 1 point
higher and 62 (9
6%) whose IGA score was 1 point lower
than the final mEASI strata predicted. Patients falling outside
the possible strata had a similar distribution of sex, age and
race/ethnicity to the overall study cohort.
Objective Scoring Atopic Dermatitis
Five different strata were selected for testing. The set with the
highest j-coefficient compared against a four-point IGA was
oSCORAD of 0–7
9 clear, 8
0–23
9 mild, 24
0–37
9 moderate
© 2017 British Association of Dermatologists
 Severity strata for atopic dermatitis signs, R. Chopra et al. 1319

Table 1 Concordance of proposed severity strata for EASI, mEASI, oSCORAD, SCORAD and BSA with IGA

IGA (four point) IGA (three point)

Clear Mild Moderate Severe j-coefficient Clear–mild Moderate Severe j-coefficient
Possible EASI strata
0 0
1–5
9 6
0–16
9 17–72 0
69 0–5
9 6
0–16
9 17
0–72 0
75
0 0
1–5
9 6
0–22
9 23–72 0
69 0–5
9 6
0–22
9 23
0–72 0
76
0 0
1–6
9 7
0–16
9 17–72 0
68 0–6
9 7
0–16
9 17–72 0
73
0 0
1–6
9 7
0–22
9 23–72 0
68 0–6
9 7
0–22
9 23–72 0
74
0–7
07 7
1–21
0 21
1–72 0
73
Possible mEASI strata
0–0
9 1–8
9 9
0–29
9 30
0–90 0
71 0–8
9 9
0–29
9 30
0–90 0
76
0–0
9 1–8
9 9
0–30
9 31
0–90 0
71 0–8
9 9
0–30
9 31
0–90 0
76
0–0
9 1–9
9 10
0–29
9 30
0–90 0
70 0–9
9 10
0–29
9 30
0–90 0
75
0–0
9 1–9
9 10
0–30
9 31
0–90 0
70 0–9
9 10
0–30
9 31
0–90 0
74
Possible oSCORAD strata
0–3
9 4–23
9 24
0–37
9 38
0–83 0
69
0–3
9 4–23
9 24
0–39
9 40
0–83 0
68
0–7
9 8–23
9 24
0–37
9 38
0–83 0
70 0–23
9 24
0–37
9 38
0–83 0
75
0–7
9 8–23
9 24
0–39
9 40
0–83 0
69 0–23
9 24
0–39
9 40
0–83 0
75
0–14
98 15
0–40
9 41
0–83 0
57
Possible SCORAD strata
0–1
9 2–28
9 29
0–45
9 46
0–103 0
64
0–1
9 2–28
9 29
0–48
9 49
0–103 0
64
0–5
9 6–28
9 29
0–45
9 46
0–103 0
66
0–5
9 6–28
9 29
0–48
9 49
0–103 0
67
0–9
9 10–28
9 29
0–45
9 46
0–103 0
68 0–28
9 29
0–45
9 46
0–103 0
72
0–9
9 10–28
9 29
0–48
9 49
0–103 0
68 0–28
9 29
0–48
9 49
0–103 0
73
0–25
916 26
0–50
9 51
0–103 0
70
Possible ADSI strata
0–1 2–5 6–8 9–15 0
55 0–5 6–8 9–15 0
58
0–1 2–6 7–8 9–15 0
51 0–6 7–8 9–15 0
52
0–1 2–5 6–9 10–15 0
54 0–5 6–9 10–15 0
57
0–1 2–6 7–9 10–15 0
49 0–6 7–9 10–15 0
50
0–2 3–5 6–8 9–15 0
54
0–2 3–6 7–8 9–15 0
50
0–2 3–5 6–9 10–15 0
53
0–2 3–6 7–9 10–15 0
49
Possible BSA strata
0 0
1–14
9 15
0–39
9 40
0–100 0
65 0–14
9 15
0–39
9 40
0–100 0
70
0 0
1–14
9 15
0–41
9 42
0–100 0
65 0–14
9 15
0–41
9 42
0–100 0
70
0 0
1–14
9 15
0–45
9 46
0–100 0
65 0–14
9 15
0–45
9 46
0–100 0
70
0 0
1–15
9 16
0–39
9 40
0–100 0
66 0–15
9 16
0–39
9 40
0–100 0
71
0 0
1–15
9 16
0–41
9 42
0–100 0
66 0–15
9 16
0–41
9 42
0–100 0
71
0 0
1–15
9 16
0–45
9 46
0–100 0
66 0–15
9 16
0–45
9 46
0–100 0
71
0 0
1–16
9 17
0–39
9 40
0–100 0
66 0–16
9 17
0–39
9 40
0–100 0
71
0 0
1–16
9 17
0–41
9 42
0–100 0
65 0–16
9 17
0–41
9 42
0–100 0
71
0 0
1–16
9 17
0–45
9 46
0–100 0
66 0–16
9 17
0–45
9 46
0–100 0
71

EASI, Eczema Area and Severity Index; mEASI, modified EASI; SCORAD, Scoring Atopic Dermatitis; oSCORAD, objective SCORAD; ADSI, Ato-
pic Dermatitis Severity Index; BSA, body surface area; IGA, Investigator’s Global Assessment. The strata with the highest j-coefficients are
highlighted in green. Selected literature values are given for comparison.

and 38
0–83 severe (j = 0
70). Similarly, a collapsed version
of these strata (0–23
9 clear–mild, 24
0–37
9 moderate,
38
0–83 severe) had the highest j-value compared against a
three-point IGA (j = 0
75). The previously reported strata for
oSCORAD8 of 0–14
9 mild, 15
0–40
9 moderate and 41–83
severe had a considerably lower j-coefficient than the above-
mentioned strata when compared against a three-point IGA
(j = 0
57).
Only seven patients (1
1%) had an IGA score > 1 point
outside of that predicted by the final oSCORAD severity
strata. There were 70 patients (10
7%) whose IGA score
was 1 point higher and 42 (6
4%) whose IGA score was 1
point lower than the final oSCORAD strata predicted.
Patients falling outside the possible strata had a similar dis-
tribution of sex, age and race/ethnicity to the overall study
cohort.

© 2017 British Association of Dermatologists British Journal of Dermatology (2017) 177, pp1316–1321
1320 Severity strata for atopic dermatitis signs, R. Chopra et al.
Scoring Atopic Dermatitis
Six different strata were selected for testing. The set of strata
with the highest j-coefficient when compared against a four-
point IGA was SCORAD of 0–9
9 clear, 10
0–28
9 mild,
29
0–48
9 moderate and 49
0–103 severe (j = 0
68). Simi-
larly, a collapsed version of these strata (0–28
9 clear–mild,
29
0–48
9 moderate, 49
0–103 severe) had the highest
j-value compared against a four-point IGA (j = 0
73). The
previously reported strata for SCORAD16 of 0–25
9 mild,
26
0–50
9 moderate and 51
0–103 severe had a lower j-coef-
ficient than the above-mentioned strata when compared
against a three-point IGA (j = 0
70).
Only two patients (0
3%) had an IGA score > 1 point out-
side of that predicted by the final SCORAD severity strata.
There were 56 patients (8
4%) whose IGA score was 1 point
higher and 79 (11
1%) whose IGA score was 1 point lower
than the final SCORAD strata predicted. Patients falling outside
the possible strata had a similar distribution of sex, age and
race/ethnicity to the overall study cohort.
Atopic Dermatitis Severity Index
Eight different ADSI strata were selected for testing. The set of
strata with the highest j-coefficient when compared against a
four-point IGA was ADSI of 0–1
9 clear, 2
0–5
9 mild, 6
0–
8
9 moderate and 9
0–15 severe (j = 0
55). Similarly, a
collapsed version of these strata (0–5
9 clear–mild, 6
0–8
9
moderate, 9
0–15 severe) had the highest j-value compared
against a three-point IGA (j = 0
58).
Body surface area
Nine different strata were selected for testing. The set of strata
with the highest j-coefficient compared against a four-point
IGA was BSA of 0 clear, 0
1–15
9 mild, 16
0–39
9 moderate
and 40
0–100 severe (j = 0
66). Similarly, a collapsed ver-
sion of these strata (0–15
9 clear–mild, 16
0–39
9 moderate,
40
0–83 severe) had the highest j-value compared against a
three-point IGA (j = 0
71).
Only eight patients (1
2%) had an IGA score > 1 point out-
side of that predicted by the final BSA severity strata. There
were 58 patients (8
9%) whose IGA score was 1 point higher
and 87 (13
3%) whose IGA score was 1 point lower than the
final BSA strata predicted. Patients falling outside the possible
strata had a similar distribution of sex, age and race/ethnicity
to the overall study cohort.
Discussion
The present study determined potential severity strata for EASI,
mEASI, oSCORAD, SCORAD, ADSI and BSA in AD. The severity
strata for EASI and SCORAD were similar to previously
reported strata overall.7,16 However, we found that moderate
disease was associated with even lower EASI scores and a
broader range of values (6
0–22
0 vs. 7
1–21
0). BSA ≤ 15
British Journal of Dermatology (2017) 177, pp1316–1321
and ≥ 40
0 were potential thresholds for defining mild and
severe disease, respectively. However, moderate disease was
associated with a higher threshold for oSCORAD (24
0–37
9
vs. 15–40) and SCORAD (16–39 vs. 26–50), with narrower
ranges of values. Interestingly, the potential severity band for
clear skin of active acute or chronic AD lesions on oSCORAD
was 0–7
9. This was due to the presence of xerosis, which is
included in oSCORAD and can be present even in the absence
of active lesions. Similarly, the potential severity band for clear
skin on SCORAD was 0–9
9, secondary to the presence of
xerosis, pruritus and sleeplessness.
Furthermore, the potential severity band for clear skin on
mEASI was 0–0
9 and for ADSI 0–1, secondary to the presence
of pruritus. While severity strata were not previously estab-
lished for ADSI, the original study of ADSI9 used 6 as a cut-
off for moderate disease, which is consistent with the range
of moderate disease observed in this study (range 6–8). ADSI
was not a preferred outcome measure for AD and its use was
not encouraged by the HOME consensus group.1 Nevertheless,
it has been recently used in a randomized controlled trial for
AD,10,11 and potential severity strata would be useful for
interpretation of those results. These are important for the
interpretation of oSCORAD, SCORAD, mEASI and ADSI scores,
because there may be clinical clearance of active AD lesions
without scores of zero for these scales.
The potential severity strata observed in this study have
important ramifications for study inclusion criteria. We found
that moderate disease includes up to 40% BSA. Yet some trials
of mild-to-moderate AD have used BSA < 20% as an inclusion
or exclusion criterion, which may bias the study population
towards only mild disease. Furthermore, some trials of moder-
ate-to-severe AD have used EASI scores > 12 or > 16 as an
inclusion or exclusion criterion, which likely biases the study
population towards more severe disease. There are also ramifi-
cations on interpretation of studies that examine the absolute
or percentage change of oSCORAD, SCORAD, mEASI and ADSI
as efficacy end points. These measures may hit a relative floor,
owing to the persistence of xerosis and symptoms. As EASI
scores only active acute or chronic lesions, it may be a more
responsive end point, particularly when trying to demonstrate
skin clearance.
There are a number of advantages of applying these strata
for the interpretation of their respective instruments. They can
be used to assess more accurately clinically meaningful
improvement in clinical trials. This is particularly important in
very large trials that are overpowered to detect differences of
continuous mean scores. A better approach in that scenario
might be to assess the proportion of persons who improve
from severe at baseline to mild or moderate at follow-up.
Another approach that has been developed is to determine the
proportion of patients achieving a prespecified percentage
improvement from baseline. For example, EASI 50 and EASI
75 represent the proportion of patients achieving ≥ 50% or
≥ 75% improvement from baseline EASI scores, respectively.
There were strong correlations between all of the examined
outcome measures and global AD severity. EASI had the
© 2017 British Association of Dermatologists

strongest correlation, but mEASI, oSCORAD and SCORAD had
nearly identical correlation coefficients. These data provide
further support for the HOME consensus group’s selection of
EASI as the preferred outcome measure for the clinical signs
of AD,1 although SCORAD was close behind. ADSI had the
weakest correlation with AD severity, suggesting that lesional
severity alone is inadequate to describe AD severity fully.
In contrast, BSA had virtually the same strength of corre-
lation with global AD severity as EASI and SCORAD, sug-
gesting that lesional extent may be a more important
construct. That is, outcome measures that include lesional
extent appear to correlate better with overall AD severity.
Furthermore, BSA by itself may be a reasonable AD out-
come measure in this respect. Nevertheless, it may be that
interpretation of AD severity using BSA alone is inadequate.
Importantly, none of the outcome measures correlated per-
fectly with overall AD severity, suggesting that no one mea-
sure is perfect at characterizing the severity of AD signs.
EASI is recommended by the HOME group as the assess-
ment of AD signs to be included in AD trials. Even so,
additional inclusion of other validated assessments of AD
signs might add value in describing AD severity in clinical
trials and research.
This study has several strengths, including the large sample
size and good representation across sex, race/ethnicity and AD
severity. However, there are some limitations. The study
cohort included adolescents and adults recruited from a single
academic centre, which may limit generalizability. In particu-
lar, an assessor’s notion of disease severity may reflect their
experience and may not match others’ experience and inter-
pretations. However, we do not believe this to be a major
issue given that the potential strata for EASI and SCORAD are
similar to those in previous reports. This suggests that the
results from our cohort are generalizable to the broader popu-
lation of adult patients with AD. Nevertheless, the proposed
potential severity strata are not the final word and may not be
optimal for all settings. Future, even larger multicentre studies
are needed to confirm these findings. Additional studies are
also needed to confirm the potential severity strata to apply to
younger children.
In conclusion, the present study identified potential severity
strata for EASI, mEASI, oSCORAD, SCORAD, ADSI and BSA in
AD. These strata can be used to improve interpretation of clin-
ical and research data using these instruments, and have
important ramifications for future clinical trials of AD.
References
1 Schmitt J, Langan S, Williams HC et al. What are the best outcome
measurements for atopic eczema? A systematic review. J Allergy Clin
Immunol 2007; 120:1389–98.
2 Charman C, Chambers C, Williams H. Measuring atopic dermatitis
severity in randomized controlled clinical trials: what exactly are
we measuring? J Invest Dermatol 2003; 120:932–41.
3 Charman CR, Venn AJ, Ravenscroft JC et al. Translating patient-
oriented eczema measure (POEM) scores into clinical practice by
© 2017 British Association of Dermatologists
Severity strata for atopic dermatitis signs, R. Chopra et al. 1321
suggesting severity strata derived using anchor-based methods. Br J
Dermatol 2013; 169:1326–32.
4 Reich A, Heisig M, Phan NQ et al. Visual analogue scale: evaluation
of the instrument for the assessment of pruritus. Acta Derm Venereol
2012; 92:497–501.
5 Rogers A, DeLong LK, Chen SC. Clinical meaning in skin-specific
quality of life instruments: a comparison of the Dermatology Life
Quality Index and Skindex banding systems. Dermatol Clin 2012;
30:333–42.
6 Love EM, Marrazzo GA, Kini S et al. ItchyQoL bands: pilot clinical
interpretation of scores. Acta Derm Venereol 2015; 95:114–15.
7 Leshem YA, Hajar T, Hanifin JM et al. What the Eczema Area and
Severity Index score tells us about the severity of atopic dermatitis:
an interpretability study. Br J Dermatol 2015; 172:1353–7.
8 Kunz B, Oranje AP, Labreze L et al. Clinical validation and guideli-
nes for the SCORAD index: consensus report of the European Task
Force on Atopic Dermatitis. Dermatology 1997; 195:10–19.
9 Van Leent EJ, Graber M, Thurston M et al. Effectiveness of the
ascomycin macrolactam SDZ ASM 981 in the topical treatment of
atopic dermatitis. Arch Dermatol 1998; 134:805–9.
10 Stein Gold LF, Spelman L, Spellman MC et al. A phase 2, random-
ized, controlled, dose-ranging study evaluating crisaborole topical
ointment, 0.5% and 2% in adolescents with mild to moderate ato-
pic dermatitis. J Drugs Dermatol 2015; 14:1394–9.
11 Murrell DF, Gebauer K, Spelman L et al. Crisaborole topical
ointment, 2% in adults with atopic dermatitis: a phase 2a, vehicle-
controlled, proof-of-concept study. J Drugs Dermatol 2015; 14:
1108–12.
12 Hanifin J, Rajka G. Diagnostic features of atopic eczema. Acta Derm
Venereol 1980; 92:44–7.
13 Hanifin JM, Thurston M, Omoto M et al. The Eczema Area and
Severity Index (EASI): assessment of reliability in atopic dermatitis.
EASI Evaluator Group. Exp Dermatol 2001; 10:11–18.
14 Remitz A, Harper J, Rustin M et al. Long-term safety and efficacy
of tacrolimus ointment for the treatment of atopic dermatitis in
children. Acta Derm Venereol 2007; 87:54–61.
15 Stalder JF, Atherton DJ, Bieber T et al. Severity scoring of atopic
dermatitis: the SCORAD index. Consensus report of the European
Task Force on Atopic Dermatitis. Dermatology 1993; 186:23–31.
16 Wollenberg A, Oranje A, Deleuran M et al. ETFAD/EADV Eczema
Task Force 2015 position paper on diagnosis and treatment of ato-
pic dermatitis in adult and paediatric patients. J Eur Acad Dermatol
Venereol 2016; 30:729–47.
Supporting Information
Additional Supporting Information may be found in the online
version of this article at the publisher’s website:
Table S1 Distribution of atopic dermatitis severity across
different Eczema Area and Severity Index scores.
Table S2 Distribution of atopic dermatitis severity across
different modified Eczema Area and Severity Index scores.
Table S3 Distribution of atopic dermatitis severity across
different objective Scoring Atopic Dermatitis scores.
Table S4 Distribution of atopic dermatitis severity across
different Scoring Atopic Dermatitis scores.
Table S5 Distribution of atopic dermatitis severity across
different Atopic Dermatitis Severity Index scores.
Table S6 Distribution of atopic dermatitis severity across
different body surface areas.
British Journal of Dermatology (2017) 177, pp1316–1321