Research

JAMA Dermatology | Original Investigation

Pediatric Contact Dermatitis Registry Data on Contact Allergy

in Children With Atopic Dermatitis
Sharon E. Jacob, MD; Maria McGowan, MD; Nanette B. Silverberg, MD; Janice L. Pelletier, MD; Luz Fonacier, MD;
Nico Mousdicas, MD; Doug Powell, MD; Andrew Scheman, MD; Alina Goldenberg, MD, MAS

IMPORTANCE Atopic dermatitis (AD) and allergic contact dermatitis (ACD) have a dynamic
relationship not yet fully understood. Investigation has been limited thus far by a paucity of
data on the overlap of these disorders in pediatric patients.

OBJECTIVE To use data from the Pediatric Contact Dermatitis Registry to elucidate the
associations and sensitizations among patients with concomitant AD and ACD.

DESIGN, SETTING, AND PARTICIPANTS This retrospective case review examined 1142 patch test
cases of children younger than 18 years, who were registered between January 1, 2015, and
December 31, 2015, by 84 health care providers (physicians, nurse practitioners, physician
assistants) from across the United States. Data were gathered electronically from
multidisciplinary providers within outpatient clinics throughout the United States on pediatric
patients (ages 0-18 years).

EXPOSURES All participants were patch-tested to assess sensitizations to various allergens;

history of AD was noted by the patch-testing providers.

MAIN OUTCOMES AND MEASURES Primary outcomes were sensitization rates to various
patch-tested allergens.

RESULTS A total of 1142 patients were evaluated: 189 boys (34.2%) and 363 girls (65.8%) in
the AD group and 198 boys (36.1%) and 350 girls (63.9%) in the non-AD group (data on
gender identification were missing for 17 patients). Compared with those without AD,
patch-tested patients with AD were 1.3 years younger (10.5 vs 11.8 years; P < .001) and had
longer history of dermatitis (3.5 vs 1.8 years; P < .001). Patch-tested patients designated as
Asian or African American were more likely to have concurrent AD (odds ratio [OR], 1.92; 95%
CI, 1.20-3.10; P = .008; and OR, 4.09; 95% CI, 2.70-6.20; P <.001, respectively). Patients with
AD with generalized distribution were the most likely to be patch tested (OR, 4.68; 95% CI,
3.50-6.30; P < .001). Patients with AD had different reaction profiles than those without AD,
with increased frequency of reactions to cocamidopropyl betaine, wool alcohol, lanolin,
tixocortol pivalate, and parthenolide. Patients with AD were also noted to have lower
frequency of reaction to methylisothiazolinone, cobalt, and potassium dichromate.

CONCLUSIONS AND RELEVANCE Children with AD showed significant reaction patterns to

allergens notable for their use in skin care preparations. This study adds to the current
understanding of AD in ACD, and the continued need to investigate the interplay between
these disease processes to optimize care for pediatric patients with these conditions.

Author Affiliations: Author

affiliations are listed at the end of this
article.
Corresponding Author: Alina
Goldenberg, MD, MAS, Resident,
Department of Dermatology,
University of California–San Diego,
8899 University Center Lane,
JAMA Dermatol. doi:10.1001/jamadermatol.2016.6136 Ste 350, Loma Linda, CA 92122
Published online February 22, 2017. (a1golden@ucsd.edu).

(Reprinted) E1

Copyright 2017 American Medical Association. All rights reserved.

Downloaded From: http://archderm.jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/0/ by a Fudan University User on 02/28/2017

 Research Original Investigation
A
topic dermatitis (AD) and allergic contact dermatitis
(ACD) are 2 of the most commonly recognized pediat-
ric inflammatory skin disorders, but they have a dy-
namic relationship that is not fully understood. For patients with
AD, concomitant undiagnosed ACD may result in treatment-
resistant disease1 leading to increased patient morbidity, eco-
nomic burden, and lost work and/or school days.2 Diagnosis of
ACD in patients with AD is associated with improvements in
cutaneous symptoms as patients become able to avoid de-
tected allergens that could potentially aggravate their under-
lying AD.1 Because AD is most commonly a disease of child-
hood, and trends show that the incidence of AD and atopy are
increasing,3 understanding the interplay between AD and ACD
is especially important in the pediatric population.
A limited number of studies have investigated the over-
lap of AD and ACD in children. Multiple studies found in-
creased sensitization rates to various allergens among chil-
dren with AD, including potassium dichromate, Compositae
mix, disperse blue, balsam of Peru, fragrance mix, and
lanolin.1,4,5 However, these study were limited by small sample
size, narrow age range, and restricted geographical distribu-
tion of test patients. One of the largest evaluations of chil-
dren with ACD was done by Zug et al6 from the North Ameri-
can Contact Dermatitis Group (NACDG), which evaluated
7 years of patch-test results from children 0 to 18 years of
age, with a total of 883 children within the sample. Zug et al6
reported that children had increased prevalence of atopic dis-
orders compared with adults. However, AD was not a pri-
mary focus of this study, and no demographic analysis was per-
formed for patients with concomitant AD and ACD.6
The Pediatric Contact Dermatitis Registry (PCDR) was de-
veloped to gather data on pediatric ACD, its associated demo-
graphics, and related conditions. Established in 2014, the PCDR
is a collaborative, multidisciplinary, nationwide registry of 1142
patch-tested children with data provided by 252 health care pro-
viders (physicians, nurse practitioners, physician assistants)
from all 50 states.7 Herein, using data from the PCDR, we fo-
cus specifically on the interplay between AD and ACD, with the
goal of elucidating the sensitizations among patients with AD.
Methods
This retrospective case review was approved by the Loma Linda
University internal review board. In 2015, data for this study
were gathered from the PCDR, a multidisciplinary, electronic
data registry of patch-test results from children 0 to 18 years
of age in the United States. Methodology for the PCDR devel-
opment and a full description of providers contributing to the
registry may be found elsewhere.7
The data collected for each case included a deidentified
database collecting demographic information (age, sex, race);
medical history, including diagnosis of AD; clinical informa-
tion on active dermatitis (anatomic location and duration); and
logistical information regarding the patch-testing technique
(type of patch test used, date of patch test, time intervals for
patch-test evaluation, and allergen results). Allergen results
were reported as either positive, relevant positive, or irritant;
E2 JAMA Dermatology Published online February 22, 2017 (Reprinted)
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: http://archderm.jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/0/ by a Fudan University User on 02/28/2017
Contact Allergy in Children With Atopic Dermatitis
Key Points
Question Does having atopic dermatitis (AD) influence
sensitization patterns and allergic contact dermatitis development
among children?
Findings In this retrospective case review, which included 1142
children younger than 18 years, those with AD had statistically
significant increased frequency of reactions to cocamidopropyl
betaine, wool alcohol, lanolin, tixocortol pivalate, and
parthenolide, and lower frequency of reaction to
methylisothiazolinone, cobalt, and potassium dichromate.
Meaning Children with AD showed significant reaction patterns
to allergens found in their skin care preparations..
grading of results was not included in the survey. Positive
patch-test (PPT) reactions included any reaction from macu-
lar erythema to a 3+ reaction as defined by the International
Contact Dermatitis Research Group. Relevance of a PPT was
interpreted and defined by the diagnosing providers and sub-
mitted within the survey, not verified by the primary investi-
gative team. The medical history and diagnostic findings were
not verified by independent review because medical records
were not accessible to the primary investigative team. Spe-
cifically, the diagnosis of AD was made by the case-entering
provider on a case-by-case basis.
Within this analysis, the percentage of PPT (PPT%) was de-
fined for each allergen as the sum of patients who had a posi-
tive reaction divided by the total number of patients patch-
tested for that specific allergen. The percentage of clinically
relevant positive patch tests (RPPT%) for each allergen was cal-
culated using the same approach. The percentage of having at
least 1 positive or relevant reaction was calculated by divid-
ing the sum of patients with at least 1 positive or relevant re-
action, respectively, by the total patients within the study.
Cases were arranged into 2 statistically independent groups
(patients with AD and those with non-AD medical history) based
on case-entering provider’s documentation. Data were ana-
lyzed with the independent t test for continuous variables, the
χ2 test for binary variables; odds ratios (ORs) and 95% CIs and
P values were confirmed through univariate logistic regres-
sion using SPSS statistical software (version 22.0; SPSS Inc).
Results
Overall, of 1142 pediatric patch test cases registered between
January 1 and December 31, 2015, by 84 unique providers, 552
patients (49%) had a history of AD (data on concurrent AD were
available for 98% of the entries).7 In 338 patients (30%), AD and
ACD were concurrent diagnoses reported by the data-entering
provider. The mean age of patients with AD was significantly
younger than the rest of the study populous—10.5 vs 11.8 years
old (P < .001). Almost two-thirds of both AD and non-AD groups
were female; sex was not significantly different between pa-
tients with AD and the rest of the study population. (Gender
identification was missing for 17 patients.) Most patients were
jamadermatology.com
 Contact Allergy in Children With Atopic Dermatitis Original Investigation Research

Table 1. Demographics of 1142 Children Referred for Patch Testing

AD Group Non-AD Group
Variable (n = 552)a (n = 565) OR (95% CI) P Value
Age, mean (SD), range, y 10.5 (4.7) 11.8 (4.6) (0.91-0.96) <.001
Sexb
Male 189 (34.2) 198 (36.1)
.50
Female 363 (65.8) 350 (63.9) 0.90 (0.70-1.20)
Race/ethnicity
White, non-Hispanic 274 (49.6) 377 (66.7) 0.49 (0.38-0.63) <.001
Hispanic/Latino 98 (17.8) 81 (14.3) 1.29 (0.94-1.80) .12
Asian 52 (9.4) 29 (5.1) 1.92 (1.20-3.10) .008
African American 106 (19.2) 31 (5.5) 4.09 (2.70-6.20) <.001
Location of dermatitis
Generalized 220 (39.9) 70 (12.4) 4.68 (3.50-6.30) <.001
Head
All 191 (34.6) 236 (41.8) 0.74 (0.58-0.94) .01
Face 70 (12.7) 88 (15.6) 0.79 (0.56-1.10) .17
Ears 14 (2.5) 25 (4.4) 0.56 (0.29-1.10) .10
Eyes 20 (3.6) 20 (3.5) 0.94 (0.54-1.90) >.99
Oral 8 (1.4) 38 (6.7) 0.20 (0.94-0.44) <.001
Neck 75 (13.6) 52 (9.2) 1.50 (1.10-2.30) .02
Torso 122 (22.1) 107 (18.9) 1.20 (0.91-1.60) .20
Abbreviations: AD, atopic dermatitis;
UEs 198 (35.9) 167 (29.6) 1.30 (1.00-1.70) .03 LE, lower extremity; PPT, positive
LEs 148 (26.8) 125 (22.1) 1.30 (0.90-1.70) .07 patch test result; RPPT, relevant
Groin 14 (2.5) 25 (4.4) 0.56 (0.30-1.10) .10 positive patch test result; UE, upper
Buttocks 18 (3.3) 28 (5) 0.65 (0.40-1.20) .17 extremity.
a
Duration of dermatitis, mean (SD), y 3.5 (4) 1.8 (2.6) (1.10-1.20) <.001 Data were missing for 25 cases
regarding medical history of AD.
≥1 RPPT 233 (42.2) 308 (54.5) 0.61 (0.48-0.77) <.001
b
Data on gender identification were
≥1 PPT 337 (61.1) 499 (88) 0.64 (0.49-0.82) <.001
missing for 17 patients.

listed as white non-Hispanic, and this category was signifi-
cantly more likely to be within the non-AD study population
(P < .001). However, those patch-tested patients designated as
Asian (OR, 1.92; 95% CI, 1.2-3.1; P = .008) or African American
(OR, 4.09; 95% CI, 2.7-6.2; P < .001) were significantly more
likely to have AD (Table 1).
Location of dermatitis prompting patch testing was sig-
nificantly differed among patients with AD compared with
those without. Patients with AD had 4.68 times the odds of hav-
ing a generalized dermatitis (OR, 4.68; 95% CI, 3.5-6.3; P < .001)
and 1.3 times the odds of having dermatitis on the upper ex-
tremities (OR, 1.3; 95% CI, 1.0-1.7; P = .03) compared with pa-
tients without AD. Patients with AD had significantly longer
history of dermatitis than those without AD: 3.5 years vs 1.8
years (P < .001). Overall, 42.2% of patch-tested children with
AD had at least 1 RPPT, and 61.1% had at least 1 PPT (Table 1).
Furthermore, the top 21 prevalent allergens with PPT were
assessed among patients with AD, in comparison with the rest
of the study sample (Table 2). Specifically, in comparison with
the patients without AD, those with AD had 7.4 times the odds
of having an RPPT to cocamidopropyl betaine (CAPB)
(P = .008), 4.2 times the odds of having an RPPT to wool al-
cohol (P = .047), 4 times the odds of having an RPPT to lano-
lin (P = .053), 5.3 times the odds of having an RPPT to tixocor-
tol pivalate (P = .02), and 7.6 times the odds of having an RPPT
to parthenolide (P = .006). For MI, cobalt, and potassium
dichromate, patients with AD had statistically significantly
lower odds of having an RPPT compared with the non-AD study
population. No statistically significant difference was seen in
RPPT rates to nickel among patients with AD compared with
the rest of the study sample (OR, 2.39; P = .13) (Table 2).
Discussion
Demographics
Patients with AD who were referred for patch testing were, on
average, more than 1 year younger than their counterparts with-
out AD and had lived with some form of dermatitis for more
than twice as long (3.5 years vs 1.6 years, respectively). Al-
though sex was not significantly different between AD and
non-AD groups, girls were patch-tested more often than boys
at a 2:1 ratio, consistent with referral patterns noted in prior
studies.6,8,9 It remains to be determined as to why girls are re-
ferred more frequently for patch testing; however, this may in-
dicate that female children have a higher incidence of con-
tact allergy. If so, increased topical product exposure may be
a relevant sensitization route in girls.
White children who underwent patch testing were found
to be the least likely to have concomitant AD, whereas patch-
tested Asians and African Americans had significantly higher
odds of having AD. The predilection for those with various an-
cestries and ethnic backgrounds who are referred for patch test-
ing to have a greater likelihood to have ACD and concomitant
AD has not been fully elucidated in the literature, and, to our
knowledge, we are the first to report such an association. Prior

jamadermatology.com (Reprinted) JAMA Dermatology Published online February 22, 2017 E3

Copyright 2017 American Medical Association. All rights reserved.

Downloaded From: http://archderm.jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/0/ by a Fudan University User on 02/28/2017

 Research Original Investigation Contact Allergy in Children With Atopic Dermatitis

Table 2. Top Allergens Among 1142 Children With AD Referred for Patch Testing

No. (%)a Abbreviations: AD, atopic dermatitis;

Patient No. Allergen RPPT PPT OR P Valueb MCI/MI, methylchloroisothiazolinone/
1 Nickel sulfate 63 (12.0) 121 (22.0) 2.38 .13 methylisothiazolinone; OR, odds ratio;
2 Fragrance mix 1 52 (9.6.0) 63 (12.0) 0.05 .84 PPT, positive patch test result;
RPPT, relevant positive patch test
3 Balsam of Peru (Myroxylon pereirae) 38 (7.0) 45 (8.3) 0.35 .63
result.
4 Bacitracin 29 (5.4) 41 (7.6) 0.9 .40 a
Percentages reflect the proportion of
5 Formaldehyde 28 (5.2) 34 (6.3) 1.2 .31 PPT out of the total number of
6 Cocamidopropyl betainec 27 (6.3) 36 (8.4) 7.4 .008 children with AD known to be
7 Propylene glycolc 25 (5.8) 33 (7.8) 2.8 .11 patch-tested with that allergen. The
denominator varied for each
8 Wool alcohol 25 (4.6) 29 (5.4) 4.2 .047
allergen—representing the total
9 Lanolinc 26 (6.0) 30 (7.0) 4.0 .053 number of children with AD known
10 Bronopol 19 (3.5) 23 (4.3) 0.005 >.99 to be patch-tested with that specific
11 Neomycin sulfate 18 (3.3) 33 (6.1) 3.77 .06 allergen. Unless otherwise marked,
12 Quaternium 15 18 (3.3) 27 (5.0) 1.91 .19 the denominator was 540 for
allergens; and for the allergens not
13 Colophony 17 (3.1) 17 (3.1) 3.53 .07
on the TRUE test (SmartPractice) the
14 Tixocortol-21-pivalate 15 (2.8) 17 (3.1) 5.33 .02 denominator was 429.
15 MCI/MI 13 (2.4) 18 (3.3) 2.17 .17 b
P value compares RPPT with each
16 Cobaltd 13 (2.4) 47 (8.7) 0.16 .02 allergen among patients with AD,
17 Fragrance mix 2c 13 (3.0) 19 (4.4) 3.64 .07 compared with everyone else.
c
18 MIc,d 12 (2.8) 14 (3.2) 0.16 .01 Represents allergen not on the
19 Potassium dichromate d
11 (2.0) 15 (2.8) 0.19 .03 TRUE test.
d
20 Compositae mixc 10 (2.3) 13 (3.0) 1.81 .20 For MI, cobalt, and potassium
dichromate, the ORs and P values
21 Parthenolide 10 (1.8) 11 (2.0) 7.65 .006
represent negative association.

large-scale assessments of patch-tested patients by the NACDG
may have had a type II error in assessing the component of race
in the association of AD and ACD because it was noted that blacks
were underrepresented in NACDG studies.10,11 Possible theo-
ries for our findings among African American and Asian pa-
tients with AD include a possibility that they are more likely to
be referred for patch testing because of increased rates of
treatment-resistant AD as associated with filaggrin mutations,12
underlying genetic predilection for AD,12 reduced minority ac-
cess to epicutaneous allergy testing, or more frequent misdi-
agnosis of AD in these racial groups. Conversely, Asian and Afri-
can American patients without AD may have been less likely to
be referred for patch testing for various reasons and, thus, were
not captured by our data registry. Future larger studies are nec-
essary to further elucidate the aforementioned associations be-
tween AD and ACD across diverse populations.
Dermatitis Type
Patients with generalized dermatitis were significantly more likely
to be patch-tested than those with other distributions. This find-
ing is consistent with those of other studies, such as the study by
Isaksson et al,1 which found that scattered generalized dermati-
tis was the most common distribution in patch-tested children.
Patch-tested patients with AD were less likely to have a PPT
and RPPT than those in the non-AD group. According to one
pathophysiologic theory, this finding may be due to an under-
lying misbalanced TH1/TH2 ratio in severe atopy.13-16 It has been
noted that patients with severe AD have a conspicuously lower
rate of ACD compared with the general population, and com-
pared with those with mild to moderate AD.17-19 TH1 response
is limited in severe atopy, with an underlying increase in TH2;
however, TH1 is critical in development of ACD; thus, these
patients with severe AD are in a sense protected from highly
sensitizing agents.16,19 With our current data set, it is not pos-
sible to separate severe AD from mild to moderate AD; there-
fore, further research is necessary to elucidate this point.
Allergens
Certain trends in sensitization have been noted in our study. Many
of the most common allergens, such as fragrance mix, balsam of
Peru, bacitracin, formaldehyde, or propylene glycol, were found
in similar ratios across both groups (patients with AD and non-
AD). Nickel, for example, is the most common positive allergen,
but there was no significant difference in rates of PPT from pa-
tients with AD compared with those without AD, consistent with
results of previous studies.4 Specifically, sensitization to 5 aller-
gens was noted more frequently in patients with AD than those
without AD to a statistically significant degree: CAPB, wool
alcohol, lanolin, tixocortol pivalate, and parthenolide (a compo-
nent of Compositae). These sensitizers are frequently associated
with topical medicaments and skin care products.
Patients with AD face inherent amplified risk of sensitiza-
tion because they are primarily treated with topical regimens
leading to increased exposure to product-based chemicals, they
pathophysiologically have an impaired epidermal barrier, and
these chemicals have been reported to have potentially en-
hancing absorption through the skin.15,20-23
Cocamidopropyl betaine is a surfactant composed of a co-
conut oil derivative combined with dimethylaminopropyl-
amine and monochloroacetic acid. It is frequently found in
shampoos, cleansers, contact lens solutions, and antiseptics.24
In 2004, it was named “Allergen of the Year” by the American
Contact Dermatitis Society (ACDS) and has been noted as an
emerging allergen in pediatrics, especially in younger age
groups.4,25 Our findings correlated with those of other studies,
such as Shaughnessy et al,20 who noted that atopic patients were

E4 JAMA Dermatology Published online February 22, 2017 (Reprinted) jamadermatology.com

Copyright 2017 American Medical Association. All rights reserved.

Downloaded From: http://archderm.jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/0/ by a Fudan University User on 02/28/2017

 Contact Allergy in Children With Atopic Dermatitis
more likely to be allergic to CAPB, and Zug et al,6 who found that
there was an increased prevalence of PPT to amidoamine, a pre-
cursor to CAPB. The frequency of sensitization in patients with
mild to moderate AD may be related to increased absorption,
due to both the penetrating nature of CAPB and the inherent sus-
ceptibility for absorption in AD skin.20-22
Parthenolide is a sesquiterpene lactone found in Compositae
plants, such as ragweed, feverfew, dandelions, sunflowers, and
daisies.24 Exposure to parthenolide may occur in nature, and is
associated with summer-exacerbated dermatitis and airborne-
pattern dermatitis, or as a component of topical preparations such
as moisturizers and cleansers.24,26 While our study showed that
patients with AD had significantly higher rates of allergy to par-
thenolide, the same was not true for Compositae, suggesting that
other elements of Compositae may more frequently be respon-
sible for sensitization in patients without AD.
Lanolin is a compound of esters, polyesters, alcohols,
and acids extracted from natural wool by using solvents or
detergents.4,27 It is a popular component of many skin prepara-
tions, especially commonly used products in AD, because of its
properties as an emollient, moisturizer, emulsifier, adhesive, and
plasticizer.24 Because lanolin is a natural compound, it varies in
composition based on its source and processing methods, ex-
plaining why some patients may tolerate certain lanolin prepa-
rations but not others.6,24,28 Wool alcohols may be removed from
lanolin, and this refined lanolin has been noted to elicit fewer re-
sponses than lanolin with retained wool alcohol.24
Tixocortol is a class A corticosteroid found in nasal sprays,
inhalers, topical hemorrhoid preparations, and buccal and
throat medications. Its use in nasal sprays is noteworthy be-
cause many patients with AD may also be using these to treat
another significant atopic syndrome, allergic rhinitis. Al-
though tixocortol is not typically an ingredient in topical
medicaments,24 other class A corticosteroids, including hy-
drocortisone, are frequently available over the counter.
Sensitization to 3 specific allergens was found to be sta-
tistically more common in patients without AD. These were
methylisothiazolinone (MI), cobalt, and potassium dichro-
mate. MI is a preservative that is commonly used as a cos-
metic preservative for its activity against bacteria and fungi.
MI has been well established as a potent allergen, similar to
poison ivy. Following with the theory of TH1/TH2 imbalance,3
patients with AD would be less likely to be sensitized to strong
sensitizers than their counterparts without AD,29 which is con-
sistent with our findings on MI.
Cobalt is a metal that is commonly combined with other
metals, such as nickel.24 These cobalt alloys are then used in
commonplace items (eg, jewelry, dental and orthopedic im-
plants, and belt buckles).24 Cobalt may also be used as pig-
ment for porcelain and glass, watercolor paints, crayons, tat-
toos, and hair dye. Historically, cobalt allergy was seen to be
inexorably linked to concomitant exposure to nickel. How-
ever, it is increasingly recognized as an independent sensitizer.30
Two studies31,32 from the 1980s and 1990s noted cobalt PPT in-
dependent of nickel PPT at rates ranging from 21% to 32%, and
the most recent NACDG data showed 40% of patients with PPT
to cobalt did not have PPT to nickel.30 Cobalt was named
“Allergen of the Year” in 2016 by the ACDS.30
jamadermatology.com
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: http://archderm.jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/0/ by a Fudan University User on 02/28/2017
Original Investigation Research
Potassium dichromate, an inorganic salt, used in manufac-
turing and construction. Historically, those at risk for ACD to chro-
mium were textile workers, leather tanners, smelters, and con-
struction workers who work with cement.24 Similar to cobalt, it
is used as a pigment in yellow and green tattoo inks, green felt
fabric, cosmetics, radiator coolants, and dental and orthopedic
implants.24 Presumably, in children, sensitization to potassium
dichromate is more likely due to pigments rather than occupa-
tional metal work. Our findings that potassium dichromate was
skewed toward patients without AD are not fully in concordance
with other studies. For example, Belloni-Fortina et al4 found that
Italian children with AD had greater prevalence in PPT to potas-
sium dichromate. Future larger studies are necessary to further
elucidate the relationship of potassium dichromate sensitivity
in patients with AD with specific focus on relevant sources of
exposure in children worldwide.
Overall, our study provides vital information on the relation-
ship between AD and ACD in pediatric patch-tested patients in
the United States. This data set demonstrated that of children
evaluated with patch testing, those who had an underlying his-
tory of AD tended to be significantly younger, a higher percent-
age were of Asian or African American descent, and they were
more likely to have a generalized-distribution and/or forearm dis-
ease and prolonged duration dermatitis compared with those
without AD. The allergens found to have significantly higher rates
of PPT and RPPT in patients with AD compared with those with-
out AD were ones commonly found in over-the-counter skin care
products associated with either treatment of AD or gentle skin
care, including CAPB, wool alcohol, lanolin, tixocortol pivalate,
and parthenolide. Patients with AD had notably lower rates of PPT
or RPPT to MI.
Limitations
Our study had several limitations. Our results may have been af-
fected by misclassification bias because our data were gathered
from a variety of providers, which could lead to possible variabil-
ity in classifying patients as having AD, a relevant positive re-
sponse, or a particular distribution of dermatitis. The variability
in protocol and antigen panels from patch tests performed by the
enrolled providers heightens this potential bias. In addition, be-
cause the rates of positive and relevant positive results depend
on the relative frequency of allergens tested, the overall results
may have been affected by differences in the types of patch tests
used by individual providers, possibly leading to certain allergens
being overrepresented or underrepresented. Because the primary
study investigators did not have access to patient medical records,
no confirmation of AD status or any additional medical history
was available as a quality-control measure. Finally, our results may
also be privy to selection bias for the providers who registered into
the study, those who entered cases into the registry, and ulti-
mately, the patients evaluated by these providers—because they
may not be representative of the general population.
Conclusions
Future larger studies are necessary to further elucidate the results
presented in this discussion. To date, the PCDR continues to pro-
vide the largest comprehensive collection of US-only pediatric
(Reprinted) JAMA Dermatology Published online February 22, 2017 E5
 Research Original Investigation
patch-test cases from which ongoing research findings and ques-
tions may continue to stem. Continued collaboration between pa-
tients, health care providers, manufacturers, and policy makers
ARTICLE INFORMATION
Accepted for Publication: December 22, 2016.
Published Online: February 22, 2017.
doi:10.1001/jamadermatol.2016.6136
Author Affiliations: Department of Dermatology,
Loma Linda University, Loma Linda, California
(Jacob); Department of Internal Medicine, Loma
Linda University, Loma Linda, California
(McGowan); Departments of Dermatology and
Pediatrics, Icahn School of Medicine at Mount Sinai,
New York, New York (Silverberg); Department of
Pediatric Dermatology, Eastern Maine Medical
Center, Bangor (Pelletier); University of Vermont
School of Medicine, Burlington (Pelletier);
Department of Allery Immunology, State University
of New York at Stony Brook, Stony Brook, New York
(Fonacier); Department of Allery Immunology,
Winthrop University Hospital, Mineola (Fonacier);
Department of Dermatology, Indiana University
Health, Indianapolis (Mousdicas); Department of
Dermatology, University of Utah, Salt Lake City
(Powell); Clinical Dermatology, Northwestern
University, Chicago, Illinois (Scheman); Department
of Dermatology, University of California–San Diego,
San Diego (Goldenberg).
Author Contributions: Drs Jacob and Goldenberg,
had full access to all of the data in the study and
take responsibility for the integrity of the data and
the accuracy of the data analysis.
Study concept and design: Jacob, Goldenberg.
Acquisition, analysis, or interpretation of data: All
authors.
Drafting of the manuscript: Jacob, McGowan,
Scheman, Goldenberg.
Critical revision of the manuscript for important
intellectual content: All authors.
Statistical analysis: Jacob, McGowan, Goldenberg.
Obtained funding: Jacob, Goldenberg.
Administrative, technical, or material support:
Jacob, Silverberg, Fonacier, Goldenberg.
Supervision: Jacob.
Conflict of Interest Disclosures: Dr Jacob received
an American Contact Dermatitis Society Mid-Career
Development award to support an education
endeavor in information technology acquisition in
association with this project. She was the
coordinating principal investigator for the PREA-1
and PREA-2 Trials. No other disclosures are
reported.
Funding/Support: This study was supported in
part by Society for Pediatric Dermatology pilot
project grant.
Role of the Funder/Sponsor: The funding source
had no role in the design and conduct of the study;
collection, management, analysis, and
interpretation of the data; preparation, review, or
approval of the manuscript; and decision to submit
the manuscript for publication.
Additional Contributions: We thank Susan
Nedorost, MD, Director, Graduate Medical
Education, UH Cleveland Medical Center, and
Professor, Dermatology, CWRU School of Medicine,
for insightful review and mentorship. She was not
compensated for her assistance.
E6 JAMA Dermatology Published online February 22, 2017 (Reprinted)
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: http://archderm.jamanetwork.com/pdfaccess.ashx?url=/data/journals/derm/0/ by a Fudan University User on 02/28/2017
is needed to shed light on the global and local epidemic of ACD,
and to prevent future morbidity from commercial allergens to our
children.
REFERENCES
1. Isaksson M, Olhardt S, Rådehed J, Svensson Å.
Children with atopic dermatitis should always be
patch-tested if they have hand or foot dermatitis.
Acta Derm Venereol. 2015;95(5):583-586.
2. Bickers DR, Lim HW, Margolis D, et al; American
Academy of Dermatology Association; Society for
Investigative Dermatology. The burden of skin
diseases: 2004 a joint project of the American
Academy of Dermatology Association and the
Society for Investigative Dermatology. J Am Acad
Dermatol. 2006;55(3):490-500.
3. Kumar V, Abbas AK, Fausto N, Robbins SL,
Cotran RS. Robbins and Cotran Pathologic Basis of
Disease. 7th ed. Philadelphia, PA: Saunders; 2005.
4. Belloni Fortina A, Fontana E, Peserico A. Contact
sensitization in children: a retrospective study of
2,614 children from a single center. Pediatr Dermatol.
2016;33(4):399-404.
5. Herro EM, Matiz C, Sullivan K, Hamann C, Jacob
SE. Frequency of contact allergens in pediatric
patients with atopic dermatitis. J Clin Aesthet
Dermatol. 2011;4(11):39-41.
6. Zug KA, Pham AK, Belsito DV, et al. Patch testing
in children from 2005 to 2012: results from the
North American Contact Dermatitis Group.
Dermatitis. 2014;25(6):345-355.
7. Goldenberg A, Jacob SE. Demographics of US
Pediatric Contact Dermatitis Registry providers.
Dermatitis. 2015;26(4):184-188.
8. Zug KA, McGinley-Smith D, Warshaw EM, et al.
Contact allergy in children referred for patch
testing: North American Contact Dermatitis Group
data, 2001-2004. Arch Dermatol. 2008;144(10):
1329-1336.
9. Hammonds LM, Hall VC, Yiannias JA. Allergic
contact dermatitis in 136 children patch tested
between 2000 and 2006. Int J Dermatol. 2009;48
(3):271-274.
10. Deleo VA, Alexis A, Warshaw EM, et al. The
association of race/ethnicity and patch test results:
North American Contact Dermatitis Group,
1998-2006. Dermatitis. 2016;27(5):288-292.
11. Deleo VA, Taylor SC, Belsito DV, et al. The effect
of race and ethnicity on patch test results. J Am
Acad Dermatol. 2002;46(2)(Suppl Understanding):
S107-S112.
12. Margolis DJ, Gupta J, Apter AJ, et al. Filaggrin-2
variation is associated with more persistent atopic
dermatitis in African American subjects. J Allergy
Clin Immunol. 2014;133(3):784-789.
13. Guillet G, Guillet MH, Dagregorio G. Allergic
contact dermatitis from natural rubber latex in
atopic dermatitis and the risk of later type I allergy.
Contact Dermatitis. 2005;53(1):46-51.
14. Boone M, Lespagnard L, Renard N, Song M,
Rihoux JP. Adhesion molecule profiles in atopic
dermatitis vs. allergic contact dermatitis:
pharmacological modulation by cetirizine. J Eur
Acad Dermatol Venereol. 2000;14(4):263-266.
15. Halling-Overgaard AS, Kezic S, Jakasa I,
Engebretsen KA, Maibach H, Thyssen JP. Skin
absorption through atopic dermatitis skin:
Contact Allergy in Children With Atopic Dermatitis
a systematic review [published online September
17, 2016]. Br J Dermatol. doi:10.1111/bjd.15065
16. Kohli N, Nedorost S. Inflamed skin predisposes
to sensitization to less potent allergens. J Am Acad
Dermatol. 2016;75(2):312-317.e1.
17. Thyssen JP, Johansen JD, Linneberg A, Menné
T, Engkilde K. The association between contact
sensitization and atopic disease by linkage of a
clinical database and a nationwide patient registry.
Allergy. 2012;67(9):1157-1164.
18. Rystedt I. Contact sensitivity in adults with
atopic dermatitis in childhood. Contact Dermatitis.
1985;13(1):1-8.
19. Uehara M, Sawai T. A longitudinal study of
contact sensitivity in patients with atopic
dermatitis. Arch Dermatol. 1989;125(3):366-368.
20. Shaughnessy CN, Malajian D, Belsito DV.
Cutaneous delayed-type hypersensitivity in patients
with atopic dermatitis: reactivity to surfactants. J Am
Acad Dermatol. 2014;70(4):704-708.
21. Steele JC, Bruce AJ, Davis MD, Torgerson RR,
Drage LA, Rogers RS III. Clinically relevant patch test
results in patients with burning mouth syndrome.
Dermatitis. 2012;23(2):61-70.
22. Mertens S, Gilissen L, Goossens A. Allergic
contact dermatitis caused by cocamide
diethanolamine. Contact Dermatitis. 2016;75(1):
20-24.
23. Jakasa I, de Jongh CM, Verberk MM, Bos JD,
Kezić S. Percutaneous penetration of sodium lauryl
sulphate is increased in uninvolved skin of patients
with atopic dermatitis compared with control
subjects. Br J Dermatol. 2006;155(1):104-109.
24. Marks JG, Elsner P, DeLeo VA. Contact
Occupational Dermatology. 3rd ed. St Louis, MO:
Mosby; 2002.
25. Jacob SE, Amini S. Cocamidopropyl betaine.
Dermatitis. 2008;19(3):157-160.
26. Belloni Fortina A, Romano I, Peserico A.
Contact sensitization to Compositae mix in children.
J Am Acad Dermatol. 2005;53(5):877-880.
27. Jovanović M, Poljacki M, Duran V, Vujanović L,
Sente R, Stojanović S. Contact allergy to
Compositae plants in patients with atopic
dermatitis. Med Pregl. 2004;57(5-6):209-218.
28. Giménez-Arnau AM; Scientific Committee Of
Consumer Safety-SCCS. Opinion of the Scientific
Committee on Consumer safety (SCCS): opinion on
the safety of the use of methylisothiazolinone (MI)
(P94), in cosmetic products (sensitisation only).
Regul Toxicol Pharmacol. 2016;76:211-212.
29. Rees J, Friedmann PS, Matthews JN. Contact
sensitivity to dinitrochlorobenzene is impaired in
atopic subjects: controversy revisited. Arch Dermatol.
1990;126(9):1173-1175.
30. Fowler JF Jr. Cobalt. Dermatitis. 2016;27(1):3-8.
31. Rystedt I, Fischer T. Relationship between nickel
and cobalt sensitization in hard metal workers.
Contact Dermatitis. 1983;9(3):195-200.
32. Kranke B, Aberer W. Multiple sensitivities to
metals. Contact Dermatitis. 1996;34(3):225.
jamadermatology.com