PRACTALL consensus report
Standardizing double-blind, placebo-controlled oral food
challenges: American Academy of Allergy, Asthma &
Immunology–European Academy of Allergy and Clinical
Immunology PRACTALL consensus report
Hugh A. Sampson, MD,aRoy Gerth van Wijk, MD,bCarsten Bindslev-Jensen, MD, PhD,cScott Sicherer, MD,a
Suzanne S. Teuber, MD,dA. Wesley Burks, MD,eAnthony E. J. Dubois, MD,fKirsten Beyer, MD,g
Philippe A. Eigenmann, MD,hJonathan M. Spergel, MD, PhD,iThomas Werfel, MD,jand Vernon M. Chinchilli, PhDk
New York, NY, Rotterdam and Groningen, The Netherlands, Odense, Denmark, Davis, Calif, Chapel Hill, NC, Berlin and Hannover, Germany,
Geneva, Switzerland, and Philadelphia and Hershey, Pa
Key words: Food allergy, oral food challenge, double-blind,
placebo-controlled food challenge
Discuss this article on the JACI Journal Club blog: www.jaci-
In an article reviewing the status of gastrointestinal allergy in
the New England Journal of Medicine in 1949, Ingelfinger et al1
decried the reliance on patients’ ‘‘incrimination’’ of specific
foods, outcome of trial diets, or association of abdominal com-
plaints with symptoms believed to be allergic in making the diag-
existedbutbelieved that the criteria used to diagnose food allergy
should begiven in capsules, bystomach tube, or in such a manner
that the patient is unaware of its nature; (2) reproducible symp-
toms should consistently follow administration of the disguised
food at a more or less constant interval; (3) other foods given to
the patient in the same manner should not produce similar
changes; and (4) suspected foods given in the same manner to
nonallergic healthy subjects should not cause the observed
AD: Atopic dermatitis
APT: Atopy patch test
DBPCFC: Double-blind, placebo-controlled food challenge
LOAEL: Lowest observable adverse event level
ML: Maximum likelihood
NO: Nitric oxide
NOAEL: No observable adverse event level
OFC: Oral food challenge
SPT: Skin prick test
FromaMount Sinai School of Medicine, New York;bErasmus Medical Center, Rotter-
dam;cOdense University Hospital, Odense;dthe University California-Davis School
Medicine, Davis;ethe University North Carolina School of Medicine, Chapel Hill;
fthe University of Groningen, University Medical Center Groningen, Department of
Pediatric Pulmonology and Pediatric Allergy, and the GRIAC Research Institute,
Groningen;gCharit? e University Hospital, Clinic for Pediatric Pneumology and Immu-
nology, Berlin;hChildren’s University Hospital of Geneva;iChildren’s Hospital of
Philadelphia and Perelman School of Medicine, University of Pennsylvania, Philadel-
Disclosure of potential conflict of interest: H. A. Sampson is on the Danone Scientific
Advisory Board; has received consultancy fees from Allertein Therapeutics and the
Food Allergy Initiative (FAI); is employed by Mount Sinai Medical School; has
received research support from the National Institute of Allergy and Infectious
Diseases (NIAID) and the National Institutes of Health (NIH); has received royalties
is 42.5% shareholder of Herbs Springs, LLC. R. Gerth van Wijk has received travel
fees from MSD, Crucell, and Haarlems Allergenen Lab; is employed by Erasmus
Medical Center; has received research support from Zon-MW, STW, and Biomay; has
received lecture fees from Allergopharma; and receives royalties from Handboek
Allergologie, Het Allergie Formulatius, and Het Allergie Boek. C. Bindslev-Jensen
has received lecture fees from MSD and Thermo-Fisher. S. Sicherer has received
consultancy fees from the FAI. S. S. Teuber has consulted for Sanofi-Aventis. A. W.
Burks has received grants and support for travel from the NIH/NIAID; is on boards for
the NIH HAI, US Food and Drug Administration (FDA), and American Academy of
Allergy, Asthma & Immunology (AAAAI); has consulted for Dannon Company,
Merck & Co, Exploramed Development Intelliject, McNeil Nutritionals, Novartis
jHannover Medical School; and
kPennsylvania State University College of
Pharma AG, and Schering-Plough; has received grants from the Food Allergy &
Anaphylaxis Network, the FAI, the NIH, and Wallace Research Foundation; has
received royalties from UptoDate; has received payment for educational presentations
from Current Views; and owns stock or stock options in Allertein and MastCell
Pharmaceuticals. A. Dubois has received travel and hotel costs from the EAACI, is on
the ALK-Abell? o Adrenaline Board, and has received research support from ALK-
Abell? o. K. Beyer has received consultancy fees from Danone, Novartis, and ALK-
Abell? o; has received research support from the German Research Foundation,
European Union, Danone, and Foundation of the Treatment of Peanut Allergy; and
has received lecture fees from Danone, Phadia, Infectopharm, CSL Behring, Novartis,
UCB, Meda Pharma, and Med Update. P. A. Eigenmann has received travel support
from the EACCI; has consulted for DBV Technologies, Danone, Nestl? e, and ALK-
Abell? o; has received grants from LETI and Nestl? e; has received payment for lectures
from Stallergenes, Sodilac, and Phadia; has received royalties from UpToDate and
advisory board for and has stock/stock options in DBV; has received consultancy fees
NIH, and Nutricia; has received lecture fees from Abbott and Nutricia; receives
royalties from UpToDate; and has received payment for development of educational
presentations from MEI. T. Werfel has received travel support from the EAACI. V. M.
Chinchilli has received travel support from the AAAAI.
Received for publication August 22, 2012; revised October 11, 2012; accepted for pub-
lication October 11, 2012.
Corresponding author: Hugh A. Sampson, MD, Department of Pediatrics, Box 1198,
Mount Sinai School of Medicine, One Gustave L. Levy Place, New York, NY
10029-6574. E-mail: firstname.lastname@example.org.
? 2012 American Academy of Allergy, Asthma & Immunology
symptoms. One year later, Mary Loveless, a New York allergist,
published 3 articles outlining the use of blind, placebo-
controlled oral food challenges to demonstrate that the practice
of the day for diagnosing food allergy (ie, elimination diets and
open feedings) inadequately accounted for confounding fac-
tors.2-4Dr Lovelesscalled for ‘‘the introduction ofcontrolled, ob-
jective methods to the study offood allergy,’’ but her appeal went
largely unnoticed for more than 10 years until A. S. Goldman and
his colleagues published a series of articles describing their eval-
uation of children with milk allergy. In this series of 89 children,
metthefollowingcriteria5-7:(1)thepatienthad toexperience res-
olution of symptoms when milk was eliminated from the diet and
milk was added back to the diet, preferably in a blinded fashion.
However, it was not until 1976, when Charles May published his
seminal article,8that the allergy community began to accept the
need for double-blind, placebo-controlled food challenges
(DBPCFCs), which became known as the gold standard, to accu-
ative challenge result had to be followed up with an open feeding
of the challenged food prepared in the usual way and given in a
normal proportion at a later time point before a patient could be
defined as nonallergic. This was to ensure that the use of
dehydrated powdered foods, as advocated for use in challenges
by May, did not alter the allergenicity of the food and that the
use of progressively increasing doses did not lead to the theoretic
‘‘rapid desensitization’’ of the patient, resulting in a ‘‘false-nega-
by Niggemann et al9when using lower initiating doses, semilog-
arithmic dose increases, and more prolonged periods between
doses, as advocated by some investigators. Overall, 13% of neg-
positive when given as a cumulative dose compared with 1% to
3% reported by American investigators.10
A number of expert panels from various organizations have
published consensus documents on the appropriate diagnosis of
food allergies. Recently, the National Institutes of Health–spon-
of food allergy,’’ reaffirmed the utility of the DBPCFC for
diagnosing food allergy after an extensive review of the current
literature.11However, the expert panel noted that open or single-
come is negative or when objective symptoms are elicited that
exactly recapitulate the history of the reaction. This is in agree-
ment with an algorithm proposed by Niggemann and Beyer,12
as shown in Fig 1. Although most authorities agree that the
DBPCFC is the gold standard, there is yet to be universal stan-
dardization of the challenge procedure and its interpretation.
Both the European Academy of Allergology and Clinical Immu-
nology13and the American Academy of Asthma, Allergy & Im-
of oral food challenges (OFCs).
In October 2008, a number of food allergy experts from the
European Academy of Allergology and Clinical Immunology and
the American Academy of Asthma, Allergy & Immunology met to
develop a consensus document on the conduct and interpretation of
the DBPCFC. Members of both academies realized the need to
develop an international standard that would advance the field by
history, and therapeutic trials in food allergy done around theworld.
In the sections that follow, the authors have summarized various
aspects of the OFC that were considered. Although this consensus
document was meant to provide a standard approach for researchers
in the field, most sections are equally applicable to the clinician in
expertise of thoseadministering the challenge. Considerable discus-
sion was directed at the appropriate prescreening and preparation of
the patient undergoing the DBPCFC; the prechallenge and intra-
challenge assessment, including how to deal with subjective symp-
positive challenge; and how outcomes should be reported.
With the increasing number of immunotherapeutic trials for
it is imperative that investigators use a standardized approach.
Regulatory agencies, such as the European Medical Agency and
the US Food and Drug Administration, are increasingly demand-
ing more stringent standards for the conduct of these studies, and
therefore it is critically important that the allergy/immunology
these standard operating procedures. It will also be important for
the allergy/immunology community to stress the importance of
children, who are more frequently affected by food allergies.
Before undertaking a DBPCFC, the patient or research subject
should be avoiding the food to be challenged and have attained a
stable baseline with regard to atopic disease. The length of time
required for chronic disease to improve on an elimination diet can
vary by disease and individual patient, but at least 2 weeks is
suggested. For safety considerations, the clinician should be aware
that prolonged elimination of a food to which IgE is detectable
might occasionally be associated with acute severe reactions on
reintroduction.15,16Challenges are preferably performed on an
empty stomach, although age and practical issues might require
compromises toallow liquidsor light nonfatty snacks. The follow-
ing additional prechallenge assessmentsand considerationsshould
be reviewed in deciding on selection of patients/research partici-
pants and for decisions regarding proceeding with administering
a DBPCFC to increase the safety and accuracy of the DBPCFC:
FIG 1. Toward a standardization of food challenges. Reprinted with per-
mission from Niggemann et al.12
J ALLERGY CLIN IMMUNOL
VOLUME 130, NUMBER 6
SAMPSON ET AL 1261
(AD), urticaria, or allergic rhinitis. Chronic atopic disease
symptoms (eg, rhinitis, AD, and asthma) should be stable
and defined as controlled/mild as possible before challenge.
Challenges should be deferred for acute infections that
might interfere with challenge interpretation.
2. Diseases and conditions that might affect safety. Patients
should not be challenged ifthey have chronic medical condi-
ditions include unstable angina pectoris, cardiac disease or
dysrhythmias, severe chronic lung disease, and pregnancy.
Pregnancy should be ruled out by testing or based on history
(eg, before menarche or after menopause), as appropriate.
3. Medications that might interfere with assessment or affect
safety. Antihistamines and medications with antihistaminic
properties can mask symptoms and should be avoided for a
period of 5 half-lives of the specific agent. Inhaled and
topical steroids or anti-inflammatory medications, such
as calcineurin inhibitors or leukotriene antagonists and
b-agonists, used at the lowest doses possible and on an
established schedule by using fixed doses of medication
to maintain a low and stable baseline of atopic disease
can usually be continued because their use is unlikely to
significantly influence challenge outcomes and their with-
drawal might result in exacerbations, affecting disease
management and challenge interpretation.
Patients should not be challenged proximate to treatment with
systemic steroids (eg, within 7-14 days) because disease rebound
might confound the interpretation of the food challenge result.
Prolonged high-dosesteroids,omalizumab, orpossibly othernew
drugs to control atopic disease are likely to modify challenge
outcomes and should be avoided. Aspirin/nonsteroidal anti-
inflammatory drugs, angiotensin-converting enzyme inhibitors,
alcohol, and antacids can act as eliciting factors that increase
reactivity in susceptible patients.17b-Blockers can pose safety
concerns if epinephrine is required for treatment.18
PRECHALLENGE ASSESSMENT: SURROGATE
Because food challenges are time-consuming and not always
without risk to the patient, much focus has been given to other
diagnostic procedures aimed at reducing the need for food
challenges. Among these are case history, skin prick testing,
measurement of specific IgE levels to the food or components in
the food in question, and atopy patch testing.
Much effort has been invested in trying to establish standards
for the size of skin prick test (SPT)responses, the levelof specific
IgE, or both to food proteins above which the probability of a
point). Decision points have been established for various foods
(milk, egg, peanut, and hazelnut) in various patient groups, but
thus far, no common values have been accepted.
Only in very clear-cut cases of an acute severe reaction to a
specific single food can the patient’s case history substitute for
challenge.19-21In most cases the incidence of self-reported
adversereactions to food exceeds the challenge-proved incidence
by up to a factor of 10.13,22-25Furthermore, in cases of delayed
reactions to foods, case history is rarely of clinical relevance.26
Skin prick testing
SPTs can be performed either with commercial extracts or with
fresh foods.27-30The sensitivity and specificity varies between
extracts, and direct comparisons between different reports are
therefore difficult. In the most positive reports31the specificity
of a positive SPT response compared with challenge outcome is
100%, but this is achieved at the expense of sensitivity because
in the studies by Sporik et al,31positive challenge outcomes were
seen in patients with negative SPT responses. Normally, the use
extracts, but the opposite has also been demonstrated.28,30There
are major differences between different foods, as demonstrated
by Verstege et al,29who established decision points for milk and
egg in a cohort of children but failed to do so with wheat and soy.
accuracy by using single allergenic proteins, such as casein in
milk32: the sensitivity of SPTs with fresh milk was superior to the
individual components, whereas the specificity of SPTs could be
improved by skin testing with purified casein in a cohort of chil-
dren with milk allergy. Another attempt to increase diagnostic
performance has recently been published by Tripodi et al30using
increased the diagnostic accuracy up to 99%.
Allergenic cross-reactions (caused by common allergenic
epitopespresentinvarious relatedorunrelatedfoodsand pollens)
might play a role in the evaluation of a positive SPT response
(or serum IgE measurement), as demonstrated in a cohort of
young patients with grass pollen allergy in which a number were
also found to have positive results to peanut without clinical
relevance.33On the other hand, panallergens, such as profilin,
were recently suggested to increase diagnostic accuracy by using
SPTs with the recombinant protein.34
Only inveryselected cases, such as the presence of IgE against
v-5 gliadin in patients with food-dependent, exercise-induced
anaphylaxis, does the presence of measurable levels of specific
IgE to a food establish a diagnosis of food allergy.35Quantifica-
tion of v-5 gliadin in patients with wheat allergy without food-
dependent, exercise-induced anaphylaxis has not been found to
be so specific.36
Decision points predicting a positive outcome of a food
challenge have been investigated in several centers worldwide
and provide a useful tool for predicting outcome, especially if the
levelis high (>20 kUA/L, as determined by using the CAP system
FEIA or UniCAP[Thermo
Sweden]).37-44Most data are available for egg and milk, followed
by peanut and tree nuts, whereas decision points for soy or wheat
of specific IgE can sometimes correlate with the severity of a
reaction, such as egg and peanut allergy,45,46no correlation
between the clinical threshold of the patient during challenge
and the specific IgE level has been demonstrated.39
One major point hampering the use of decision points as a
surrogate parameter for food allergy is the diversity of the
reported levels of specific IgE resulting in a greater than 95%
J ALLERGY CLIN IMMUNOL
1262 SAMPSON ET AL
probability for a positive challenge outcome. Regarding egg
allergy in children, published decision points vary between
1.5 kUA/L,397 kUA/L or greater,47and 10.2 kUA/L,44and the
decision points can even vary within the same clinical
center.41,43,44,47One reason for these discrepancies is the age of
the patients studied because Celik-Biligi et al41calculated the de-
that in infants greater than 1 year of age to be 13.2 kUA/L, giving
rise to the overall level of 12.6 kUA/L.
Decision points should thus be used with caution and adjusted
to the actual patient population in a center. Furthermore, most
data are established with the Phadia CAP System FEIA or
UniCAP system, and data generated with these assays are not
readily transferrable to other systems because differences
between different commercial
Another issue is the probability of serologic, clinically insignif-
icant cross-reactivity seen between different allergens, such as in
peanut33orinpatientswith a challenge-provedfoodallergytoone
at least partially resolved in the future, when component-resolved
single purified allergenic proteins in the food might improve the
diagnostic accuracy and eliminate insignificant cross-reactions to
pollen allergens. Antibodies toward numerous proteins have been
identified,52and data are now emerging on the relative allergenic
significance of various proteins in hazelnut,53,54soy,55egg,56,57
However, at the current time,
component-resolved diagnostics cannot serve as a substitute for
OFCs in determining either thresholds or severity.44
Atopy patch testing might improve the diagnostic accuracy in
ited benefit overmeasurement of specificIgE or SPTs in the daily
routine. Measurement of specific IgG levels or subclasses thereof
have not been demonstrated to predict outcomes of food chal-
lenges.21,63The use of basophil leukocyte activation tests as an
in vitro marker for clinical reactivity has not been thoroughly
investigated, and no data exist on the use of these methods as sur-
rogate markers for clinical food allergy.64
offood allergy, but certain levels of specific IgE might reduce the
need for challenges. However, the patient population in question,
is important to always keep in mind that the surrogate parameters
measure sensitization and not clinical disease in the patient.
PRECHALLENGE ASSESSMENT: CHALLENGE
Food challenges, whether DBPCFCs, single-blind challenges,
or open challenges, pose certain safety issues that must be
addressed before beginning any challenge. The purpose of the
challenge, the type of expected allergic symptoms, and the
patient’s previous clinical history of a reaction, among others,
are all important factors to consider.12-14,65,66
Purpose of challenge
The purpose of the challenge is important; it might simply be a
certain food. Challenges can be done for research purposes and
again might confirm or refute a patient is allergic to the food or
might show the minimal threshold of food needed to cause initial
suggestive of a reaction but without likelihoodof a reaction, and a
itive.13,67,69Other important factors to consider include whether
the suspected allergic reaction is likely IgE or non–IgE mediated,
the age of the patient, and the food to be used for challenge.
There are several options for a possible location of the food
challenge, including the intensive care unit, a regular hospital
room, a hospital clinic room, an outpatient clinic room, a private
practice office clinic, or the home of the patient.20Between 20%
dose and schedule of dose escalation is part of overall safety. For
example, Sicherer et al69reported on a series of 196 children who
underwent DBPCFCs in whom the results of 26 milk challenges
and 22 egg challenges were positive at the first dose of 250 mg
and 11% ofthe firstdose reactions weresevere. Additionally, cer-
tain foods seem to cause more severe reactions on challenge than
with the severity of reaction on food challenge.46,71,72The likeli-
hood of a late-phase or delayed reaction should be considered be-
subject; the presence of other concomitant diseases, such as AD,
asthma, or risk factors for cardiovascular disease; whether they
tiveand subjectivesymptoms); and whether the patient was on an
elimination diet before challenge.14,73-75
The personnel involved in a challenge procedure must be
and anaphylaxis practice drills should be conducted periodically
with the challenge team. Centers for performing food challenges
are desirable, especially for challenges conducted for research
purposes, so that personnel have adequate experience in conduct-
ing food challenges.76
All necessary medication should be readily available if needed
during the food challenge. Epinephrine, oxygen, antihistamines
(both H1and H2), b-agonists, corticosteroids, and intravenous
fluids are all needed for specific types of allergic reactions.18,76
All patients with a history of a previous severe reaction who are
being challenged should have intravenous access established.
J ALLERGY CLIN IMMUNOL
VOLUME 130, NUMBER 6
SAMPSON ET AL 1263
are food challenges for peanuts, tree nuts, fish, and shellfish,
depending on the situation and the patient.14,20,69In small chil-
dren intravenous access is necessary only in selected patients; if
there is any doubt as to the outcome (severity) of the challenge,
intravenous access might be needed.13,66Also, patients with
severe asthma, even without a history of an anaphylactic reaction
to food, and older children and adults with difficult intravenous
access should also be considered for prior line insertion.14
Patients can generally be discharged after an observation period
of at least 2 hours, provided no reaction occurred. If the clinical
history indicates the allergic symptoms on the initial reaction
occurred later, then a longer period of observation might be
necessary. If the patient had significant clinical symptoms, then a
period of up to 4 hours of observation might be needed before
requiring significant treatment, then he or she should be kept under
Conclusions: future directions
challenge and where the challenge should be done. A food
challenge might be done for various reasons, generally to identify
canputintohis orherdiet. ItisimportantthatOFCsbeconducted
in a location that is well equipped to deal with anaphylactic
reactions by physicians and staff who are well trained in dealing
with such reactions. This does not exclude practicing allergists’
offices because most allergists administering immunotherapy are
fully equipped to deal with anaphylactic reactions. Food chal-
lenges for research purposes will continue to allow us to better
understand the mechanism of food allergy, as well as to begin to
develop new therapies.
CHALLENGE PARAMETERS: SCHEDULES AND
Numerous dosage schedules using diverse materials are cur-
rently in use for performing DBPCFCs for research purposes.
These schedules differ in starting dose, incremental scale, time
between doses, and top dose (see Fig 2).13,14,41,65,68,69,71,78
A distinction must be made between dosage schedules designed
for studies determining lowest observable adverse event levels
(LOAELs) and noobservable
(NOAELs)13,68,79and those used for other purposes. Similarly,
schedules will be different if late-phase reactions, such as those
seen in patients with AD, are the object of the study.
DBPCFCs with cow’s milk, in which the challenge dose is ex-
pressed as protein content of the particular dose. Although
schedules for other foods are similar, calculation from protein
content to weight of food in either the dried or native form might
vary, especially when protein contents of such foods are high
(eg, fish) or low (eg, celery).13For these high and low protein-
containing foods, other amounts might be necessary to deliver
reasonable quantities of food to the patient. Dosages have
been derived from either absolute quantities13,41,68,78,79or per-
centages of either maximum single dose65or cumulative
doses.65,70As can be seen from the figure below, starting doses
vary widely from the low microgram level for studies aimed at
NOAEL/LOAEL determinations to doses from the low milli-
gram level to several hundred milligrams in schedules used for
other purposes. Incremental scales vary from 10-fold increases
through 5-fold,19,68semilogarithmic (ie, 1, 3, 10, 30, 100, and
300 mg and so on),13,41doubling-dose,79or even smaller65,69,71
increases, with the latter being associated with schedules aiming
to deliver cumulative doses. These last-mentioned schedules
also generally have higher starting doses and shorter time inter-
vals between doses (15 minutes) than the schedules aimed at de-
livery of discrete doses, where intervals between doses are
typically 20 to 30 minutes. One of the pitfalls of using lower
starting doses and longer intervals is the increased likelihood
of partial desensitization and false-negative results.9
With schedules developed for NOAEL/LOAEL determination
to achieve meaningful top doses with acceptable increments in an
acceptable period of time. A combination of logarithmic and
semilogarithmic increases, as done in EuroPrevall, can be
Double-blind challenge schedules
13 (doubling dose)
13 (semi log)
71 (age >5)
FIG 2. Incremental doses of various schedules used for DBPCFCs. Individual doses are expressed as
micrograms of protein in the food administered during the active food challenge. Schedules shown are
either generic or derived from those devised for challenges with cow’s milk. The numbers shown with
symbols refer to the source publications in the reference list.
J ALLERGY CLIN IMMUNOL
1264 SAMPSON ET AL
resulting from studies using different schedules, robust evidence
demonstrating the relative superiority of any of these schedules is
limited. Comparing schedules is further hampered by the inability
to assess the contribution of a given schedule parameter to a
purported difference in outcome. Many parameters have been
established on the basis of practical considerations, such as the
these limitations, the following recommendations are proposed.
The preferred method for DBPCFCs is the administrations of
active and placebo challenges on separate days. If the active and
placebo test challenges are administered on the same day, these
challenges should be separated by at least 3 hours. However, such
schedules will not be capable of diagnosing reactions more than
3 hours after the challenge. Administration of active and placebo
doses in an interspersed fashion (ie, administering both placebo
and active doses in the same challenge) is not recommended.
with more severe reactions during the challenge. Starting doses at
the low milligram level are generally safe and seem to result in
fewer severe reactions than higher doses.70,81Experience with
10-fold dose increases as being relatively safe over the entire
dose range is limited. Many schedules using semilogarithmic
increases are associated with good safety. Smaller increments (eg,
Longer time intervals between doses will also lengthen the chal-
lenge procedure but will probably decrease the chance of accumu-
lating high doses, which might result in more severe reactions. An
The top dose required to avoid false-negative DBPCFC results
is uncertain but seems to be at least 2 g. Sicherer et al69reported
approximately 5% false-negative test results with a schedule
with a top dose of 875 mg (cumulative dose, 3500 mg) of protein.
Schedules with top doses of approximately 1750 mg (cumulative
dose, 2190 mg) report a false-negative rate of approximately 9%
and 12.7%,82asgauged byrecurrence ofsymptoms duringsubse-
quent introduction of the food. The original recommendation of
May8of 8 to 10 g of dried food as being an adequate top dose is
show that top doses might need to be higher for certain foods (eg,
those of which the usual portion contains relatively high amounts
of protein, such as fish), patients, or situations. Follow-up of neg-
ative DBPCFC results with open food challenges with the native
food in amounts usually consumed thus remains recommended
if clinical tolerance is to be unambiguously demonstrated.
On the basis of the above recommendations, a general
challenge schedule consisting of 3, 10, 30, 100, 300, 1000, and
3000 mg offood protein at intervals of at least 20 minutes (Fig 2)
would be suitable. Such a schedule is likely to be appropriate for
most foods, patients, clinical situations, and settings. Appropriate
challenge doses for common foods are shown in Table I and can
also be found in a recent working group report.14On the basis of
ing false-negative test outcomes.
Materials used in DBPCFCs should be well documented. The
form and source of the food to be used should preserve maximum
allergenicity. Matrices should be capable of blinding the sensory
qualities of the food in a volume that is as small as possible and
compatible with use in the target patient group (eg, children).
Blinding should be verified by using appropriate testing in a
sensory laboratory.78Current evidence suggests that a high fat
content of matrices slows absorption and masks early warning
symptoms, leading to more severe reactions.83A large variety
of matrices and matrix food combinations is important for palat-
ability, especially in young children.
PARAMETERS TO FOLLOW DURING ORAL
The assessment of a food challenge is mostly clinical and will
always largely depend on the investigator. However, decisions on
thepositivityofa foodchallengeresultneedto bestandardized as
much as possible and should, if possible, include tools for
Before each dose, the patient should have a clinical examina-
tion, including inspection of the skin, lung auscultation, abdom-
inal auscultation, and blood pressure, heart rate, and oxygen
saturation measurement. In addition, the patient should be
questioned for pruritus (mouth mucosa, skin, or both), laryngeal
symptoms, abdominal pain, dizziness, and any other new com-
plaints.Although thephysicalexamination will provide objective
symptoms, most complaints arising from the patient without
observable changes need to be classified as subjective and, if
isolated, might not account for a positive challenge result (see the
section on interpretation).
Assessment of respiratory parameters
Respiratory reactions are often difficult to assess early on, and
an objective tool that could identify pulmonary changes early is
desirable because it might predict the potential severity of the
reaction. Respiratory reactions were reported in 59% of patients
with a positive food challenge result, mostly as nasal symptoms
(63%), laryngeal reactions (43%), or involvement of the lower
airways (64%).84Interestingly, significant spirometric changes
recognition of pulmonary changes secondary to the food chal-
lenge. The same authors hypothesized that patients undergoing
food challenges might also present with more subtle reactions,
implicating only a change in bronchial reactivity.85They investi-
gated a group of 26 asthmatic children undergoing DBPCFCs by
performing methacholine challenges both before and 4 to 6 hours
after a DBPCFC. Chest symptoms (cough, wheeze, or both)
developed during 12 of 38 DBPCFCs. Of these, 7 subjects had
a significant increase in bronchial hyperreactivity in addition to
1 patient who had no apparent chest symptoms. Overall, metha-
assessment of clinical reactivity in research protocols or espe-
cially clinical practice involving DBPCFCs, although they might
provide an objective measurement of a respiratory reaction, even
without a significant change in pulmonary function testing. How-
ever,use ofmethacholine inhalationchallenges will notprovide a
useful early indicator of ongoing pulmonary changes.
The measurement of nitric oxide (NO) has been shown to
provide an objective and reliable tool to quantify airway
J ALLERGY CLIN IMMUNOL
VOLUME 130, NUMBER 6
SAMPSON ET AL 1265
been measured in 44 infants undergoing open milk challenges
before and immediately after the termination of the challenges.86
Awide range of exhaled NO values were observed, and no signif-
icant changes were observed in children with positive or negative
milk challenge results. A more recent study measured NO levels
up to 2 hours after the last challenge dose and reported a signifi-
cant NO level decrease, peaking at 90 minutes after positivechal-
lenge results.87Although these procedures might provide an
objective measurement for respiratory reactions, the delay in
assessment of the positivity of a DBPCFC result.
Assessment of blood parameters
Early studies have linked positive DBPCFC results with
increased histamine levels in peripheral blood samples,88leading
several investigators to explore changes in various parameters
measurable in the blood for the assessment of a food challenge.
Niggemman et al89have shown that peripheral blood eosinophil
numbers decreased immediately after positive DBPCFC results
ilarly, other mediators suggestive of mast cell and basophil de-
granulation have been investigated. Some investigators have
reported significant increases in urinary 1-methylhistamine levels
1 hour after a positiveOFC result, as well as a significant increase
in serum tryptase levels, but the sensitivity of these parameters is
low.90Similar results were observed by Ohtsuka et al91when
measuring plasma tryptase and histamine levels up to 4 hours af-
the blood or saliva.92
Other objective measures of clinical reactivity
For immediate availability of objectives measurements, Clark
et al93measured the temperatures of 3 areas of the face using fa-
perature measured on the nose during a positive food challenge
result. The same method has been recently used with success to
evaluate peanut challenges on the nasal mucosa.94
The clinical judgment of an experienced investigator will
always be the most important factor in deciding when to call a
DBPCFC result positive or negative. However, an objective, easy
to perform,andimmediately availableobjective assessment,such
as facial thermography or exhaled NO measurement, would
provide a decisive tool that could help to further standardize
SCORING AND STOPPING OFCs
To allow comparison of outcomes of DBPCFCs, standards
must be followed to report results, includingwhich symptoms are
classified as subjective or objective and what outcomes constitute
a positive challenge result. There are currently no agreed upon
and circumstances might vary by patient or study characteristics.
according to requirements of specific study protocols, the symp-
toms observed, safety issues, and patients’ characteristics. In
response to safety concerns, dosing can be ceased before eliciting
clear objective reactions, which in turn might reduce the diag-
nostic accuracy of the procedure. It is beyond the scope of this
report to develop a comprehensiveapproach to conducting OFCs,
but guidelines toward improved reporting are presented.
Challenge results are typically considered positive, and dosing
is stopped when objective symptoms occur. However, in some
situations mild objective symptoms might be considered insuffi-
cient to discontinue dosing or to consider a challenge result
positive (eg, 1 or 2 transient perioral urticarial lesions from
contact with the food or 1 episode of vomiting in a child with
anxiety and a distaste for the challenge substance).12In some cir-
cumstances subjective symptoms can indicate a positive chal-
lenge response and present a good reason to cease dosing, such
as by having repetitive symptoms or multiple subjective symp-
toms in several organ systems. However, stopping a challenge
for subjective symptoms increases the risk of a false-positive
dicate a positive test result. Subjective or initially mild objective
symptoms can be subtle in a young child, who can become sud-
denly quiet or begin to show behaviors such as food refusal or
scratching in the ears or at the tongue and neck.
At a minimum, the parameters for stopping and declaring a
challenge result positive or negative should be prespecified in
challenge protocols and the details should be reported in publi-
cations. Symptoms typically considered subjective are listed in
Table II. Regarding the options for discontinuing dosing and con-
sidering a challenge result positive, options include ‘‘worsening’’
subjectivesymptoms,repeated elicitation (eg,on3 doses), orper-
sistence (eg, 40 minutes).12-14These parameters for stopping a
challenge and declaring a challenge result positive have not
been evaluated with regard to the effect on the sensitivity and
TABLE I. Equivalent amount of food in milligrams (and milliliter) for the major foods in various forms often used in challenges
Pasteurized cow’s milk
with 3.3% protein content
Skim milk powder with
36% protein content
Pasteurized whisked hen’s egg
with 12.8% protein content
Hen’s egg powder with
47% protein content
3 mg 91 mg z 0.1 mL 8.3 mg23.4 mg 6.4 mg
303 mg z 0.3 mL
909 mg z 0.9 mL
3,030 mg z 3.0 mL
9,091 mg z 9.1 mL
30,303 mg z 30.3 mL
90,909 mg z 90.9 mL
J ALLERGY CLIN IMMUNOL
1266 SAMPSON ET AL
specificity of the test. When mild objective or subjective symp-
ger for the next dose, or repeating a dose.12,75Judgments about
proceeding must balance safety against the certainty of the chal-
lenge outcome. Although more time-consuming, undertaking ad-
ditional challenges (eg, 5 challenges per food with 3 challenges
containing placebo and 2 challenges containing the food allergen
or vice versa randomly) can increase the accuracy of conclusions
when symptoms are subjective.13
A scoring system appropriate for acute allergic responses is
shown in Fig 3. This scoring system has not been validated but
a dose, delaying a dose, and consideration for stopping) or are
clear enough to warrant stopping a challenge and declaring the
result positive. When DBPCFCs are used to evaluate disease out-
comes in which symptoms are delayed (eg, isolated gastrointesti-
nal reactions and AD) or controversial (migraine, behavior, and
arthritis), different approaches, such as symptom diaries and
SCORAD scores, might be needed. Similarly, evaluation of
isolated oral symptoms from pollen-food–related syndrome re-
quires modification of outcome measures. Challenge protocols
might require modifications relevant to specific ages, diseases,
and challenge circumstances; it is suggested to report in detail
how symptoms were assessed with regard to stopping dosing
results to allow comparison of outcomes.
LATE REACTIONS TO FOOD IN PATIENTS WITH AD
Clinical reaction patterns to food in patients
can be readily suspected in many patients with a detailed history,
skin test responses, and demonstration of specific IgE toward the
suspected food, the identification of triggers for delayed reac-
tions, such as eczematous reactions in patients with AD, is often
more difficult.26Provocation tests can lead to 3 different reaction
patterns in patients with AD95:
1. noneczematous reactions, which are usually IgE mediated
and present on the skin as pruritus, urticaria, or flush reac-
tions, as well as with other immediate-type reactions of the
gastrointestinal or respiratory tracts, or as anaphylaxis;
2. isolated eczematous delayed reactions (ie, usually flare of
eczema after hours or 1-2 days); or
3. a combination of a noneczematous early reaction and an
eczematous delayed reaction.
Analyses of food challenges in children show that isolated late
reactions (ie, those after 2 hours) occur in about 12% of children
with AD.41,96Adolescents and adults can also react to foods, but
reactions to ‘‘classical’’ food allergens, such as hen’s eggs and
cow’s milk, are not as common as in childhood. Some patients
with AD react to pollen-associated foods with late reactions.97,98
in adolescent and adult patients.
Practical aspects of the diagnostic algorithm in
suspected eczematous reactions to food allergens
Diagnosing food allergy in patients with AD does not differ in
principle from the diagnostic workup in other allergic diseases.67
However, serum food-specific IgE levels, SPT responses, and
clinical history often do not correlate well with clinical findings.
In particular, thevalue of history has been shown to be less useful
in suspected late reactions compared with immediate reac-
tions.96,99,100When food allergy is suspected, appropriate
in vivo tests (eg, skin tests), in vitro tests (eg, serum specific
IgE), or both should performed. For routine skin testing, only
SPTs are recommended for the diagnosis of food allergy in pa-
tients with AD.
The atopy patch test (APT) with cow’s milk, hen’s egg,
cereals, and peanut might be of some value, identifying specific
food allergies in patients with AD in the following cases61: (1)
suspicion of food allergy without predictive specific IgE levels
or positive SPT responses; (2) severe AD, persistent AD, or
both with unknown trigger factors; and (3) multiple IgE sensiti-
zations without proved clinical relevance in patients with AD.
The APT can therefore be considered an additional diagnostic
tool that can be used in specialized institutions. However, stan-
dardized reagents and methods for performing APTs and inter-
preting their results have not been established, and therefore
APTs are not recommended for routine diagnoses of food-
induced eczema. Of note, it was concluded from an evaluation
of a large number of children with AD that the APT does not
lead to a significant reduction in the need for OFCs when food-
induced eczema is suspected.101
In cases of suspected food allergy (by history, specific sensi-
tization, or both), a diagnostic elimination diet of suspected food
items over a period of up to 4 to 6 weeks is recommended.
However, OFCs should be performed, even if the skin improved
Peanut butter with 24%
Peanut flour with 50%
Gluten powder with 80%
Soy drink with 3.3%
Soy powder with 50%
12.5 mg 6.0 mg3.8 mg 91 mg z 0.1 mL6.0 mg
303 mg z 0.3 mL
909 mg z 0.9 mL
3,030 mg z 3.0 mL
9,091 mg z 9.1 mL
30,303 mg z 30.3 mL
90,909 mg z 90.9 mL
TABLE I. (Continued)
J ALLERGY CLIN IMMUNOL
VOLUME 130, NUMBER 6
SAMPSON ET AL 1267
generally not sufficient to accept the result of an elimination diet
as diagnostic because a variety of factors can influence the
outcome during the elimination period.
Oral provocation tests in suspected eczematous
In patients with AD, oral provocation should be performed
after a period of suspected food allergen elimination when
symptoms are minimal. If a stable situation cannot be attained
by avoidance alone, topical therapy must be intensified before the
provocation and continued in the same fashion throughout the
OFCs. Accompanying therapy and conditions should be changed
as little as possible. If a corticosteroid is absolutely necessary,
then the application of a weak preparation (eg, 1% hydrocorti-
sone) once daily is possible. Clinically relevant reactions will not
be prevented with this minimal treatment, although on the other
hand, natural fluctuations of the eczema are minimized. Steroid
therapy should be continued over the entire provocation; no
antihistamine or UV therapy should be given.
Even when no immediate-type reactions are reported, provo-
cation should only be performed by clinicians with experience in
who have been maintaining elimination diets for a prolonged
period of time,verysevere tolife-threatening reactions havebeen
reported after reintroduction of a food allergen.15,16
expected immediate reactions, as discussed above. The dose
should be increased every 20 minutes up to the highest dose or
until a clinical reaction occurs. The total dose should approxi-
mately correspond to the average serving at a meal (eg, 1 egg or
150 mL of milk). Because of the expected delayed reaction, the
skin must be inspected by a medical doctor who is familiar in
If possible, repetitive provocation with the same food for another
1 to 2 days is recommend in those patients who did not react after
the first challenge day. Clinical evaluation with a standardized
eczema scoring system (eg, SCORAD or Eczema Area and
Severity Index score) must be uniform throughout the period.
Fig 4 summarizes the diagnostic algorithm for the identification
of late reactions in patients with AD.
DEALING WITH SUBJECTIVE SYMPTOMS IN
The DBPCFC is a rigorous tool that has become popular for
evaluating adverse reactions to foods. The standard use of the
DBPCFC has been to document food allergies for individual
patients, but it recently has been gaining acceptance as a
procedure for investigating the effectiveness of therapies to
prevent/minimize food-induced anaphylaxis. The DBPCFC is
administered in increasing (fixed) doses to each participant, and
the challenge is discontinued when the participant exhibits
objective symptoms (eg, vomiting, diarrhea, and urticaria) at a
specific dose or when the top dose is consumed without evidence
of reactivity. This presentation addresses statistical design and
analysis issues for the DBPCFC when subjects exhibit subjective
symptoms (eg, abdominal pain and throat tightness).
Assume that the DBPCFC consists of K increasing doses of the
food item, denoted as follows:
d1< d2< ::: < dk:
A subject undergoing the active component of the DBPCFC
progresses through the increasing doses of the food item. The
DBPCFC result can be discontinued for a subject because of any
of the following events:
A. objective symptoms occurring at one of the doses;
B. no objective symptoms occurring through dose dK; and
C. withdrawal before objective symptoms occur.
IfBor Coccurs,thenthat subjectissaid tobe ‘‘rightcensored’’
at the final dose that was administered. Censoring is said to occur
when results of an observation are only partially known. The
observation is considered right censored when the final (true)
result is greater than the observed result, but it is not known by
how much. In the case of B and C, the investigator would know
that the subject has tolerated a certain dosewithout symptoms but
how much more the subject could tolerate and remain symptom
free is unknown.
For the ith subject in the study (i 5 1, 2, . . . , n),2 items, Diand
li, are rerecorded, where Diis defined as the final dose adminis-
and liis defined as 0 if objective symptoms were not observed at
As an example, supposethere are 5 dose levels of thefood item
designated as follows:
d155mg; d2510mg; d3520mg; d4540mg; d5580mg;
and 4 subjects undergo the active component of the DBPCFC
symptoms? What should be done? Hourihane et al102continued
the challenge after the subjective symptoms were completely
resolved. Wensing et al103discontinued the challenge if the sub-
jective symptoms endured for at least 60 minutes. Flinterman
et al16discontinued the challenge if the subjective symptoms
occurred on 3 consecutive doses. Peeters et al104discontinued
the challenge if the subjective symptoms lasted for at least
An approach that minimizes the potential bias caused by the
occurrence of subjective symptoms is as follows:
TABLE II. Subjective symptoms
General nonspecific pruritus
Dyspnea (without objective signs)
Complaints of weakness, dizziness, not feeling well, etc
J ALLERGY CLIN IMMUNOL
1268 SAMPSON ET AL
A. Erythematous Rash- % area involved______
0 = Absent
1 = Mild, occasional scratching
2 = Moderate -scratching continuously for > 2 minutes at a time
3 = Severe hard continuous scratching excoriations
0 = Absent
1 = Mild < 3 hives, or mild lip edema
2 = Moderate - < 10 hives but >3, or significant lip or face edema
3 = Severe generalized involvement
0 = Absent
1 = Mild few areas of faint erythema
2 = Moderate areas of erythema
3 = Severe generalized marked erythema (>50%)
II. UPPER RESPIRATORY
0 = Absent
1 = Mild rare bursts, occasional sniffing
2 = Moderate bursts < 10, intermittent rubbing of nose, and/or eyes or frequent
3 = Severe continuous rubbing of nose and/or eyes, periocular swelling and/or long
bursts of sneezing, persistent rhinorrhea
III. LOWER RESPIRATORY
1 = Mild expiratory wheezing to auscultation
2 = Moderate inspiratory and expiratory wheezing
3 = Severe use of accessory muscles, audible wheezing
1 = Mild >3 discrete episodes of throat clearing or cough, or persistent throat
2 = Moderate hoarseness, frequent dry cough
3 = Severe stridor
A. Subjective Complaints
0 = Absent
1 = Mild complaints of nausea or abdominal pain, itchy mouth/throat
2 = Moderate frequent c/o nausea or pain with normal activity
3 = Severe - notably distressed due to GI symptoms with decreased activity
B. Objective Complaints
0 = Absent
1 = Mild 1 episode of emesis or diarrhea
2 = Moderate 2-3 episodes of emesis or diarrhea or 1 of each
3 = Severe >3 episodes of emesis or diarrhea or 2 of each
0 = normal heart rate or BP for age/baseline
1 = mild-subjective response (weak, dizzy), or tachycardia
2 = moderate-drop in blood pressure and/or >20% from baseline, or significant
change in mental status.
3 = severe-cardiovascular collapse, signs of impaired circulation (unconscious)
- Not usually an indication to alter dosing.
- Not generally sufficient to consider a challenge positive.
Orange (scores increasing to orange):
- Caution, dosing could proceed, be delayed, have a dose repeated rather than escalated.
- If clinically indicated, dosing is stopped.
- Symptoms that recur on 3 doses, or persist (e.g., 40 minutes) are more likely indicative
of a reaction than when such symptoms are transient and not reproducible.
- 3 or more scoring areas in orange more likely represent a true response.
- Objective symptoms likely to indicate a true reaction
- Usually an indication to stop dosing.
FIG 3. Scoring the challenge outcome (modified from Bock et al65and Nowak-Wegrzyn et al14). The scoring
system proposed here can be used for IgE-mediated reactions to determine the degree of response in var-
ious target organs and changes from baseline. Challenges should usually not commence if there are base-
line symptoms exceeding descriptions in green (an exception might be AD that remains moderate despite
maximal therapy). See the text for additional comments. (Please note that I,C,1 was made orange because it
is similar to mild objective symptoms in other areas, is not a stopping indication, and, depending on clinical
judgment, might or might not represent contact urticaria).
J ALLERGY CLIN IMMUNOL
VOLUME 130, NUMBER 6
SAMPSON ET AL 1269
1. continue the challenge, in spite of the subjective symp-
toms, until the subject exhibits objective symptoms, or
2. if the subject is unable or unwilling to continue the chal-
lenge because of the subjective symptoms, then discon-
tinue the challenge and consider the final administered
dose to be right censored.
less precise (more variable) because of the increased amount of
right censoring. Less precision results in a larger sample size
The most appealing approach for the between-group compar-
ison is a discrete-time survival analysis. This analysis is based on
a model for the conditional probability of a subject responding to
to previously administered doses.
This conditional probability is called the hazard function, and
an appropriate model to describe this process is the discrete-time
hazard function (in this situation the discrete-dose hazard func-
tion) as follows:
where x1, x2, . . ., xrdenote a set of r regressors, such as treat-
ment group assignment, baseline score, sex, and age.
This particular form for the discrete-time hazard function, as
proposed by Prentice and Gloeckler105and Allison,106is called
the extremevalue hazard function, and a statistical analysis based
plementary log-log link function. The extremevalue hazard func-
tion is the discrete-time analog of the Cox proportional hazards
function. Researchers might not be familiar with the extreme
value hazard function because they are more familiar with the ex-
The parameters a1, a2, . . . , aKand the parameters b1, b2, . . . ,
brare estimated through maximum likelihood (ML) estimation.
Statistical inference, such as hypothesis testing for comparing
the hazard rates of the experimental therapy and control groups,
is based on the ML estimates of b1, b2, . . . , br. SAS PROC LO-
GISTIC (SAS, Institute, Cary, NC) is available for the logistic re-
gression analysis with the complementary log-log link function.
An example is taken from Chinchilli et al.107A dataset for a
clinical trial comparing an experimental therapy with a control
therapy in patients with peanut allergy was simulated as follows:
Example: Six doses of peanut flour were administered during a
baseline DBPCFC and study’s end DBPCFC:
d15 1 mg
d25 4 mg
d35 16 mg
d4tt 5 64 mg
d55 256 mg
d65 1024 mg
log4(d1) 5 0
log4(d2) 5 1
log4(d3) 5 2
log4(d4) 5 3
log4(d5) 5 4
log4(d6) 5 5.
Fifty patients were randomized to each of the experimental
therapy and control groups. The final dose administered was
coded as 0, 1, 2, 3, 4, or 5 (log4of the dose). The results of the
analysis indicate that the baseline score and group assignment
are statistically significant (Table IV).
The negative sign for the ML estimate of the experimental
therapy group (21.157) effect indicates that the hazard (condi-
tional probability of objective symptoms) for the experimental
therapy group is less than the hazard for the control group; that is,
the experimental therapy is significantly more effective than the
Estimating sample size
Because the discrete-time survival analysis is the discrete
analog of the Cox proportional hazards regression model, sample
size for the former can be approximated by using a sample size
formula based on the latter. One such formula is provided by
Schoenfeld.108This sample size formula requires that the effect
size be expressed in terms of the hazard ratio, which is the ratio
of the hazard functions for the experimental and control groups.
Typical values that researchers use for effect sizes in sample
size calculations range from a hazard ratio of 1.5 (small effect
size) to 3.0 (moderate effect size).
the anticipated censoring rate (averaged across the 2 treatment
arms) at the study’s end food challenge. For example, if a
researcher anticipates that the experimental therapy group will
experience approximately 20% censoring and the control group
will experience approximately 0% censoring, then this is an
anticipated, then there will be a higher level of right censoring,
and hence an increased sample size will be necessary. Sample
sizes (per treatment arm) for various hazard ratios (1.5, 2.0. 2.5,
Persistent moderate to severe AD
History of possible food allergy, specific IgE, SPT (APT)
Diagnostic elimination diet over a period of some weeks (e.g. 4-6 weeks)
First step of OFC in stable phase of disease, evaluation of eczema score before OFC
First titrated oral food challenge
Evaluation of immediate symptoms during titration and the following two hours
Evaluation of eczema score for at least 16-24 hours after OFC
In cases of a negative reaction:
Repeat challenge with the average daily intake of food for another 1 – 2 days
Evaluation of eczema score on the following day
At least one challenge free day
Next step of OFC
OFC: oral food challenge (ideally DBPCFC should be performed); APT: Atopy Patch
Test, SPT: Skin prick test, AD: atopic dermatitis
FIG 4. Diagnostic algorithm for the identification of food allergy in patients
TABLE III. Subject responses to doses in DBPCFC
d15 5 mg
d45 40 mg
d55 80 mg
d35 20 mg
Objective symptoms at d1
Objective symptoms at d4
Right censored at d5
Right censored at d3
J ALLERGY CLIN IMMUNOL
1270 SAMPSON ET AL
and 3.0) and censoring rates (0%, 10%, 20%, and 30%) for a
2-sided, .05 significance level test with 90% statistical power are
presented in the Table V. Notice that the sample size increases as
(1) thehazardratio decreasesand(2) thecensoringrateincreases.
the presence of objective symptoms is a discrete-time survival
analysis. Statistical software packages, such as SAS, have avail-
able routines for applying such an analysis. When planning a
DBPCFC, it is important to calculate sample size based on the
assumed hazard ratio and the level of right censoring (the
percentage of participants who will not experience objective
symptoms). The formula presented by Schoenfeld108can be in-
voked to approximate the sample size. The statistical approach
described above is a very conservative approach with respect to
subjective symptoms in that a participant who discontinues be-
cause of subjective symptoms is regarded as right censored.
A more sensitive statistical approach needs to be developed that
accounts for subjective symptoms.
OFC: REPORTING RESULTS
Not only should the reasons for stopping a challenge be
reported (ie, objective symptoms vs repeated subjective, wors-
ening subjective, or persistent subjective symptoms) but also
studies should tabulate the numbers of patients who fall into the
different categories. The number of patients with severe systemic
reactions and their characteristics should be reported. The use of
epinephrine to treat a reaction is not an adequate measure of
threshold to administer epinephrine might be lower in an outpa-
tient setting), and the physician’s judgment will affect the
treatment of a reaction. Investigators are encouraged to make
individual data available when possible in online repositories. In
summary, Tables VI and VII list the recommended data that
should optimally be included in research reports.
Outcomes of DBPCFCs for research purposes should be
published, with the results of the placebo challenges clearly
described. The interpretation of equivocal challenge results
should be determined in advance, as shown in the example in
Fig 3.19Placebo responses are important to publish because these
rates should be used in the statistical handling of the results to ac-
commodate and estimate false-positive rates; such information is
also important in study design.109
Eliciting versus cumulative dose
Both the eliciting and cumulative dose in milligrams of protein
should be reported because this provides the maximum informa-
ple, it might seem that the eliciting dose is the sole dose
responsible for a reaction that begins in the mouth with pruritus
and angioedema, but to date, it cannot be excluded that previous
delayed absorption, and therefore both should be reported. Stud-
ies are currently lacking that compare a possible discrete dose ef-
fect versus the cumulative dose contribution.
Reporting of randomized, controlled therapeutic
As food allergy research continues to move beyond character-
ization to randomized therapeutic trials, the importance of
adopting consistent parameters for reporting results is evident
because this will allow a critical comparison of parameters
TABLE IV. Outcome of peanut challenges in 100 patients
randomized to experimental therapy or control group
b (ML estimate)SEP value
TABLE V. Sample size varies dependent upon hazard ratio and
TABLE VI. Data to report in studies using OFCs
1. Demographics (can be summarized in larger studies but helpful to have in
an online repository for individual data)
Age, sex, age at diagnosis, reaction symptoms, SPT responses, in vitro
specific IgE levels, other atopic disorders
2. Were patients with a history of anaphylaxis included?
Subanalysis might be indicated
3. Food matrix used (recipe), including percentage fat content
4. Method to validate blinding between placebo and verum
5. Exact description of challenge food’s physical state (eg, raw, cooked,
dehydrated, or defatted)
6. Dose escalation schedule
Should be expressed as mg of protein
7. Setting for challenges
8. Scoring system with predetermined stopping/delaying/repeating/
progressing decision points
9. Percentage of patients stopped for objective responses and subjective
10. Eliciting dose in mg of protein
11. Cumulative dose in mg of protein
12. Percentage positive, negative, or inconclusive, with criteria defined in
TABLE VII. Example of a decision matrix
Active food challengePlacebo challengeInterpretation
Positive (clearly more
Negative (or positive, but
positive than placebo)
Reprinted with permission from Vlieg-Boerstra et al.19
J ALLERGY CLIN IMMUNOL
VOLUME 130, NUMBER 6
SAMPSON ET AL 1271
between studies and facilitate the application of meta-analyses.
A framework developed in the 1990s (as the use of meta-analysis
an international level. The Consolidated Standards of Reporting
and execution that are essential components of a high-quality
ing of trials include full disclosure of each step of enrollment (eg,
eligibility and recruitment), allocation to intervention, follow-up,
and analysis. A flow chart facilitates visualization of the numbers
of participants at each phase of the study and whether analysis
study is extended when combined with other trials, it is important
for food allergy intervention trials to adhere to CONSORT to en-
sure high quality, even if the proposed trial is small or of inade-
quate power for generalization of results.
1. Ingelfinger FJ, Lowell FC, Franklin W. Gastrointestinal allergy. N Engl J Med
2. Loveless MH. Milk allergy: a survey of its incidence; experiments with a masked
ingestion test. J Allergy 1950;21:489-99.
3. Loveless MH. Allergy for corn and its derivatives: experiments with a masked
ingestion test for its diagnosis. J Allergy 1950;21:500-9.
4. Loveless MH. Cottonseed protein vs. cottonseed oil sensitivity; an objective ap-
proach to the diagnosis of food-allergy as applied to cottonseed atopy. Ann Al-
5. Goldman AS, Anderson DW, Sellers WA, Saperstein A, Kniker WT, Halpern SR.
Milk allergy. I. Oral challenge with milk and isolated milk proteins in allergic
children. Pediatrics 1963;32:425-43.
6. Goldman AS, Sellers WA, Halpern SR, Anderson DW, Furlow TE, Johnson CH.
Milk allergy. II. Skin testing of allergic and normal children with purified milk
proteins. Pediatrics 1963;32:572-9.
7. Saperstein S, Anderson DW Jr, Goldman AS, Kniker WT. Milk allergy. III. Im-
munological studies with sera from allergic and normal children. Pediatrics 1963;
8. May CD. Objective clinical and laboratory studies of immediate hypersensitivity
reactions to food in asthmatic children. J Allergy Clin Immunol 1976;58:500-15.
9. Niggemann B, Lange L, Finger A, Ziegert M, Muller V, Beyer K. Accurate oral
food challenge requires a cumulative dose on a subsequent day. J Allergy Clin
10. Sicherer SH, Sampson HA. Food allergy. J Allergy Clin Immunol 2010;
11. Boyce JA, Assa’ad A, Burks AW, Jones SM, Sampson HA, Wood RA, et al.
Guidelines for the diagnosis and management of food allergy in the United States:
report of the NIAID-sponsored expert panel. J Allergy Clin Immunol 2010;
12. Niggemann B, Beyer K. Diagnosis of food allergy in children: toward a standard-
ization of food challenge. J Pediatr Gastroenterol Nutr 2007;45:399-404.
13. Bindslev-Jensen C, Ballmer-Weber BK, Bengtsson U, Blanco C, Ebner C, Hour-
ihane J, et al. Standardization of food challenges in patients with immediate re-
actions to foods—position paper from the European Academy of Allergology
and Clinical Immunology. Allergy 2004;59:690-7.
14. Nowak-Wegrzyn A, Assa’ad AH, Bahna SL, Bock SA, Sicherer SH, Teuber SS.
Work Group report: oral food challenge testing. J Allergy Clin Immunol 2009;
15. David TJ. Anaphylactic shock during elimination diets for severe atopic derma-
titis. Arch Dis Child 1984;59:983-6.
16. Flinterman AE, Knulst AC, Meijer Y, Bruijnzeel-Koomen CA, Pasmans SG.
Acute allergic reactions in children with AEDS after prolonged cow’s milk elim-
ination diets. Allergy 2006;61:370-4.
17. Untersmayr E, Jensen-Jarolim E. The role of protein digestibility and antacids on
food allergy outcomes. J Allergy Clin Immunol 2008;121:1301-8.
18. Lieberman P, Kemp SF, Oppenheimer JJ, Land I, Bernstein IL, Nicklas RA, et al.
The diagnosis and management of anaphylaxis: an updated practice parameter.
J Allergy Clin Immunol 2005;115(suppl):S485-523.
19. Vlieg-Boerstra BJ, van der Heide S, Bijleveld CM, Kukler J, Duiverman EJ, Du-
bois AE. Placebo reactions in double-blind, placebo-controlled food challenges in
children. Allergy 2007;62:905-12.
20. Sampson HA. Food allergy—accurately identifying clinical reactivity. Allergy
21. Burks AW, Jones SM, Boyce JA, Sicherer SH, Wood RA, Assa’ad A, et al.
NIAID-sponsored 2010 guidelines for managing food allergy: applications in
the pediatric population. Pediatrics 2011;128:955-65.
22. Sampson HA. Food Allergy. Part 2: diagnosis and management. J Allergy Clin
23. Host A, Husby S, Osterballe O. A prospective study of cow’s milk allergy in ex-
clusively breast-fed infants. Incidence, pathogenetic role of early inadvertent ex-
posure to cow’s milk formula, and characterization of bovine milk protein in
human milk. Acta Paediatr Scand 1988;77:663-70.
24. Osterballe M, Hansen TK, Mortz CG, Bindslev-Jensen C. The clinical relevance
of sensitization to pollen-related fruits and vegetables in unselected pollen-
sensitized adults. Allergy 2005;60:218-25.
25. Osterballe M, Mortz CG, Hansen TK, Andersen KE, Bindslev-Jensen C. The
prevalence of food hypersensitivity in young adults. Pediatr Allergy Immunol
26. Werfel T, Ballmer-Weber B, Eigenmann PA, Niggemann B, Rance F, Turjanmaa
K, et al. Eczematous reactions to food in atopic eczema: position paper of the
EAACI and GA2LEN. Allergy 2007;62:723-8.
27. Bock S, Buckley J, Holst A, May C. Proper use of skin tests with food extracts in
diagnosis of food hypersensitivity. Clin Allergy 1978;8:559-64.
28. Rosen J, Selcow J, Mendelson L, Grodofsky M, Factor J, Sampson H. Skin testing
with natural foods in patients suspected of having food allergies.is it necessary?
J Allergy Clin Immunol 1994;93:1068-70.
29. Verstege A, Mehl A, Rolinck-Werninghaus C, Staden U, Nocon M, Beyer K, et al.
The predictive value of the skin prick test weal size for the outcome of oral food
challenges. Clin Exp Allergy 2005;35:1220-6.
30. Tripodi S, Businco AD, Alessandri C, Panetta V, Restani P, Matricardi PM. Pre-
dicting the outcome of oral food challenges with hen’s egg through skin test end-
point titration. Clin Exp Allergy 2009;39:1225-33.
31. Sporik R, Hill DJ, Hosking CS. Specificity of allergen skin testing in predicting
positive open food challenges to milk, egg and peanut in children. Clin Exp Al-
32. Calvani M, Alessandri C, Frediani T, Lucarelli S, Miceli SS, Panetta V, et al.
Correlation between skin prick test using commercial extract of cow’s milk pro-
tein and fresh milk and food challenges. Pediatr Allergy Immunol 2007;18:
33. Mortz CG, Andersen KE, Bindslev-Jensen C. The prevalence of peanut
sensitization and the association to pollen sensitization in a cohort of un-
selected adolescents—the Odense Adolescence Cohort Study on Atopic
Diseases and Dermatitis (TOACS). Pediatr Allergy Immunol 2005;16:
34. Asero R, Monsalve R, Barber D. Profilin sensitization detected in the office by
skin prick test: a study of prevalence and clinical relevance of profilin as a plant
food allergen. Clin Exp Allergy 2008;38:1033-7.
35. Inomata N. Wheat allergy. Curr Opin Allergy Clin Immunol 2009;9:238-43.
36. Beyer K, Chung D, Schulz G, Mishoe M, Niggemann B, Wahn U, et al. The role
of wheat omega-5 gliadin IgE antibodies as a diagnostic tool for wheat allergy in
childhood. J Allergy Clin Immunol 2008;122:419-21.
37. Sampson H, Ho D. Relationship between food-specific IgE concentration and the
risk of positive food challenges in children and adolescents. JAllergy Clin Immu-
38. Sampson HA. Utility of food-specific IgE concentrations in predicting sympto-
matic food allergy. J Allergy Clin Immunol 2001;107:891-6.
39. Osterballe M, Bindslev-Jensen C. Threshold levels in food challenge and specific
IgE in patients with egg allergy: is there a relationship? J Allergy Clin Immunol
40. Mehl A, Verstege A, Staden U, Kulig M, Nocon M, Beyer K, et al. Utility of the
ratio of food-specific IgE/total IgE in predicting symptomatic food allergy in chil-
dren. Allergy 2005;60:1034-9.
41. Celik-Bilgili S, Mehl A, Verstege A, Staden U, Nocon M, Beyer K, et al. The pre-
dictive value of specific immunoglobulin E levels in serum for the outcome of
oral food challenges. Clin Exp Allergy 2005;35:268-73.
42. Ridout S, Matthews S, Gant C, Twiselton R, Dean T, Arshad SH. The diagnosis of
Brazil nut allergy using history, skin prick tests, serum-specific immunoglobulin
E and food challenges. Clin Exp Allergy 2006;36:226-32.
43. Maloney JM, Rudengren M, Ahlstedt S, Bock SA, Sampson HA. The use of
serum-specific IgE measurements for the diagnosis of peanut, tree nut, and
seed allergy. J Allergy Clin Immunol 2008;122:145-51.
44. Ott H, Baron JM, Heise R, Ocklenburg C, Stanzel S, Merk HF, et al. Clinical use-
fulness of microarray-based IgE detection in children with suspected food allergy.
J ALLERGY CLIN IMMUNOL
1272 SAMPSON ET AL
45. Benhamou AH, Zamora SA, Eigenmann PA. Correlation between specific immu-
noglobulin E levels and the severity of reactions in egg allergic patients. Pediatr
Allergy Immunol 2008;19:173-9.
46. Hourihane JO, Grimshaw KEC, Lewis SA, Briggs RA, Trewin JB, King RM,
et al. Does severity of low-dose, double-blind, placebo-controlled food challenges
reflect severity of allergic reactions to peanut in the community? Clin Exp Allergy
47. Sampson HA. Food Allergy. J Allergy Clin Immunol 2003;111:S540-7.
48. Wang J, Godbold JH, Sampson HA. Correlation of serum allergy (IgE) tests
performed by different assay systems. J Allergy Clin Immunol 2008;121:
49. Eller E, Kjaer HF, Host A, Andersen KE, Bindslev-Jensen C. Food allergy and
food sensitization in early childhood: results from the DARC cohort. Allergy
50. Sicherer SH. Clinical implications of cross-reactive food allergens. J Allergy Clin
51. Lin J, Bardina L, Shreffler WG, Andreae DA, Ge Y, Wang J, et al. Development
of a novel peptide microarray for large-scale epitope mapping of food allergens.
J Allergy Clin Immunol 2009;124:315-22.
52. Sastre J. Molecular diagnosis in allergy. Clin Exp Allergy 2010;40:1442-60.
53. Hansen KS, Ballmer-Weber BK, Sastre J, Lidholm J, Andersson K, Oberhofer H,
et al. Component-resolved in vitro diagnosis of hazelnut allergy in Europe.
J Allergy Clin Immunol 2009;123:1134-41.
54. Flinterman AE, Akkerdaas JH, Knulst AC, van Ree R, Pasmans SG. Hazelnut al-
lergy: from pollen-associated mild allergy to severe anaphylactic reactions. Curr
Opin Allergy Clin Immunol 2008;8:261-5.
55. Holzhauser T, Wackermann O, Ballmer-Weber BK, Bindslev-Jensen C, Scibilia J,
Perono-Garoffo L, et al. Soybean (Glycine max) allergy in Europe: Gly m 5 (beta-
conglycinin) and Gly m 6 (glycinin) are potential diagnostic markers for severe
allergic reactions to soy. J Allergy Clin Immunol 2009;123:452-8.
56. Ando H, Moverare R, Kondo Y, Tsuge I, Tanaka A, Borres MP, et al. Utility of
ovomucoid-specific IgE concentrations in predicting symptomatic egg allergy.
J Allergy Clin Immunol 2008;122:583-8.
57. D’Urbano LE, Pellegrino K, Artesani MC, Donnanno S, Luciano R, Riccardi C,
et al. Performance of a component-based allergen-microarray in the diagnosis of
cow’s milk and hen’s egg allergy. Clin Exp Allergy 2010;40:1561-70.
58. Flinterman AE, Pasmans SG, Hoekstra MO, Meijer Y, Van Hoffen E, Knol EF,
et al. Determination of no-observed-adverse-effect levels and eliciting doses in
a representative group of peanut-sensitized children. J Allergy Clin Immunol
59. Nicolaou N, Poorafshar M, Murray C, Simpson A, Winell H, Kerry G, et al. Al-
lergy or tolerance in children sensitized to peanut: prevalence and differentiation
using component-resolved diagnostics. J Allergy Clin Immunol 2010;125:191-7.
60. Codreanu F, Collignon O, Roitel O, Thouvenot B, Sauvage C, Vilain AC, et al. A
novel immunoassay using recombinant allergens simplifies peanut allergy diagno-
sis. Int Arch Allergy Immunol 2011;154:216-26.
61. Turjanmaa K, Darsow U, Niggemann B, Rance F, Vanto T, Werfel T. EAACI/
GA2LEN position paper: present status of the atopy patch test. Allergy 2006;
62. Spergel JM, Brown-Whitehorn T. The use of patch testing in the diagnosis of food
allergy. Curr Allergy Asthma Rep 2005;5:86-90.
63. Stapel SO, Asero R, Ballmer-Weber BK, Knol EF, Strobel S, Vieths S, et al. Test-
ing for IgG4 against foods is not recommended as a diagnostic tool: EAACI Task
Force Report. Allergy 2008;63:793-6.
64. Kleine-Tebbe J, Erdmann S, Knol EF, MacGlashan DW Jr, Poulsen LK, Gibbs
BF. Diagnostic tests based on human basophils: potentials, pitfalls and perspec-
tives. Int Arch Allergy Immunol 2006;141:79-90.
65. Bock SA, Sampson HA, Atkins FM, Zeiger RS, Lehrer S, Sachs M, et al. Double-
blind, placebo-controlled food challenge (DBPCFC) as an office procedure: a
manual. J Allergy Clin Immunol 1988;82:986-97.
66. Rance F, Deschildre A, Villard-Truc F, Gomez SA, Paty E, Santos C, et al. Oral
food challenge in children: an expert review. Eur Ann Allergy Clin Immunol
67. Sicherer SH, Sampson HA. Food allergy. J Allergy Clin Immunol 2006;
68. Taylor SL, Hefle SL, Bindslev-Jensen C, Atkins FM, Andre C, Bruijnzeel-
Koomen C, et al. A consensus protocol for the determination of the threshold
doses for allergenic foods: how much is too much? Clin Exp Allergy 2004;34:
69. Sicherer SH, Morrow EH, Sampson HA. Dose-response in double-blind, placebo-
controlled oral food challenges in children with atopic dermatitis. J Allergy Clin
70. Perry TT, Matsui EC, Connover-Walker MK, Wood RA. Risk of oral food chal-
lenges. J Allergy Clin Immunol 2004;114:1164-8.
71. Perry TT, Matsui EC, Connover-Walker MK, Wood RA. The relationship of
allergen-specific IgE levels and oral food challenge outcome. J Allergy Clin Im-
72. Mankad VS, Williams LW, Lee LA, LaBelle GS, Anstrom KJ, Burks AW. Safety
of open food challenges in the office setting. Ann Allergy Asthma Immunol 2008;
73. Gellerstedt M, Bengtsson U, Niggemann B. Methodological issues in the diagnos-
tic work-up of food allergy: a real challenge. J Investig Allergol Clin Immunol
74. Niggemann B, Wahn U, Sampson HA. Proposals for standardization of oral food
challenge tests in infants and children. Pediatr Allergy Immunol 1994;5:11-3.
75. Niggemann B, Beyer K. Pitfalls in double-blind, placebo-controlled oral food
challenges. Allergy 2007;62:729-32.
76. Cox L, Nelson H, Lockey R, Calabria C, Chacko T, Finegold I, et al. Allergen
immunotherapy: a practice parameter third update. J Allergy Clin Immunol
77. Sampson HA, Munoz-Furlong A, Campbell RL, Adkinson NF Jr, Bock SA, Bra-
num A, et al. Second symposium on the definition and management of anaphy-
laxis: summary report—Second National Institute of Allergy and Infectious
Disease/Food Allergy and Anaphylaxis Network symposium. J Allergy Clin Im-
78. Vlieg-Boerstra BJ, Bijleveld MA, van der Heide S, Beusekamp BJ, Wolt-
Plompen SAA, Kukler J, et al. Development and validation of challenge materials
for double-blind, placebo-controlled food challenges in children. J Allergy Clin
79. Crevel RW, Ballmer-Weber BK, Holzhauser T, Hourihane JO, Knulst AC, Mackie
AR, et al. Thresholds for food allergens and their value to different stakeholders.
80. Cochrane SA, Salt LJ, Wantling E, Rogers A, Coutts J, Ballmer-Weber BK, et al.
Development of a standardized low-dose double-blind placebo-controlled chal-
lenge vehicle for the EuroPrevall project. Allergy 2012;67:107-13.
81. Vlieg-Boerstra BJ, Duiverman EJ, van der Heide S, Bijleveld CM, Kukler J, Du-
bois AE. Should children with a history of anaphylaxis to foods undergo chal-
lenge testing? Clin Exp Allergy 2008;38:1935-42.
82. Eigenmann PA, Caubet JC, Zamora SA. Continuing food-avoidance diets after
negative food challenges. Pediatr Allergy Immunol 2006;17:601-5.
83. Grimshaw KEC, King RM, Nordlee JA, Hefle SL, Warner JO, Hourihane JOB.
Presentation of allergen in different food preparations affects the nature of the
allergic reaction—a case series. Clin Exp Allergy 2003;33:1581-5.
84. James JM, Bernhisel-Broadbent J, Sampson HA. Respiratory reactions provoked
by double-blind food challenges in children. Am J Respir Crit Care Med 1994;
85. James JM, Eigenmann PA, Eggleston PA, Sampson HA. Airway reactivity
changes in food-allergic, asthmatic children undergoing double-blind placebo-
controlled food challenges. Am J Respir Crit Care Med 1996;153:597-603.
86. Gabriele C, Hol J, Kerkhof E, Elink Schuurman BE, Samsom JN, Hop W, et al.
Fractional exhaled nitric oxide in infants during cow’s milk food challenge.
Pediatr Allergy Immunol 2008;19:420-5.
87. Benhamou AH, Koehli A, Rochat I, Inci D, Moeller A, Taramarcaz P, et al.
Exhaled nitric oxide decreases after positive food-allergen challenge. Clin Transl
88. Sampson HA, Jolie PL. Increased plasma histamine concentrations after food
challenges in children with atopic dermatitis. N Engl J Med 1984;311:372-6.
89. Niggemann B, Beyer K, Wahn U. The role of eosinophils and eosinophil cationic
protein in monitoring oral challenge tests in children with food-sensitive atopic
dermatitis. J Allergy Clin Immunol 1994;94:963-71.
90. Beyer K, Niggemann B, Schulze S, Wahn U. Serum tryptase and urinary 1-meth-
ylhistamine as parameters for monitoring oral food challenges in children. Int
Arch Allergy Immunol 1994;104:348-51.
91. Ohtsuka T, Matsumaru S, Uchida K, Onobori M, Matsumoto T, Kuwahata K,
et al. Time course of plasma histamine and tryptase following food challenges
in children with suspected food allergy. Ann Allergy 1993;71:139-46.
92. Vila L, Sanz ML, Sanchez-Lopez G, Garcia-Aviles C, Dieguez I. Variations of
serum eosinophil cationic protein and tryptase, measured in serum and saliva,
during the course of immediate allergic reactions to foods. Allergy 2001;56:
93. Clark AT, Mangat JS, Tay SS, King Y, Monk CJ, White PA, et al. Facial thermog-
raphy is a sensitive and specific method for assessing food challenge outcome.
94. Clark A, Mangat J, King Y, Islam S, Anagnostou K, Foley L, et al. Thermo-
graphic imaging during nasal peanut challenge may be useful in the diagnosis
of peanut allergy. Allergy 2012;67:574-6.
95. Werfel T, Breuer K. Role of food allergy in atopic dermatitis. Curr Opin Allergy
Clin Immunol 2004;4:379-85.
J ALLERGY CLIN IMMUNOL
VOLUME 130, NUMBER 6
SAMPSON ET AL 1273
96. Breuer K, Heratizadeh A, Wulf A, Baumann U, Constien A, Tetau D, et al. Late Download full-text
eczematous reactions to food in children with atopic dermatitis. Clin Exp Allergy
97. Reekers R, Busch M, Whittmann M, Kapp A, Werfel T. Birch pollen-related
foods trigger atopic dermatitis in patients with specific cutaneous T-cell responses
to birch pollen antigens. J Allergy Clin Immunol 1999;104:466-72.
98. Breuer K, Wulf A, Constien A, Tetau D, Kapp A, Werfel T. Birch pollen-related
food as a provocation factor of allergic symptoms in children with atopic eczema/
dermatitis syndrome. Allergy 2004;59:988-94.
99. Atherton DJ, Soothill JF, Sewell M, Wells RS, Chilvers CED. A double-blind
controlled crossover trial of an antigen-avoidance diet in atopic eczema. Lancet
100. Niggemann B, Sielaff B, Beyer K, Binder C, Wahn U. Outcome of double-blind,
placebo-controlled food challenge tests in 107 children with atopic dermatitis.
Clin Exp Allergy 1999;29:91-6.
101. Mehl A, Rolinck-Werninghaus C, Staden U, Verstege A, Wahn U, Beyer K, et al.
The atopy patch test in the diagnostic workup of suspected food-related symptoms
in children. J Allergy Clin Immunol 2006;118:923-9.
102. Hourihane JO’B, Kilburn SA, Nordlee JA, Hefle SL, Taylor SL, Warner JO. An
evaluation of the sensitivity of subjects with peanut allergy to very low doses
of peanut protein: a randomized, double-blind, placebo-controlled food challenge
study. J Allergy Clin Immunol 1997;100:596-600.
103. Wensing M, Penninks AH, Hefle SL, Koppelman SJ, Bruijnzeel-Koomen CA,
Knulst AC. The distribution of individual threshold doses eliciting allergic
reactions in a population with peanut allergy. J Allergy Clin Immunol 2002;
104. Peeters KA, Nordlee JA, Penninks AH, Chen L, Goodman RE, Bruijnzeel-Koo-
men CA, et al. Lupine allergy: not simply cross-reactivity with peanut or soy.
J Allergy Clin Immunol 2007;120:647-53.
105. Prentice RL, Gloeckler LA. Regression analysis of grouped survival data with ap-
plication to breast cancer data. Biometrics 1978;34:57-67.
106. Allison PD. Discrete-time methods for the analysis of event histories. Sociologi-
cal Methodology Vol 13; 1982; American Sociological Assoc, Washington, DC;
107. Chinchilli VM, Fisher L, Craig TJ. Statistical issues in clinical trials that involve
the double-blind, placebo-controlled food challenge. J Allergy Clin Immunol
108. Schoenfeld DA. Sample-size formula for the proportional-hazards regression
model. Biometrics 1983;39:499-503.
109. Briggs D, Aspinall L, Dickens A, Bindslev-Jensen C. Statistical model for assess-
ing the proportion of subjects with subjective sensitisations in adverse reactions to
foods. Allergy 2001;56(suppl 67):83-5.
110. Moher D, Schulz KF, Altman DG. The CONSORT statement: revised recommen-
dations for improving the quality of reports of parallel-group randomized trials.
J Am Podiatr Med Assoc 2001;91:437-42.
111. Altman DG, Schulz KF, Moher D, Egger M, Davidoff F, Elbourne D, et al. The
revised CONSORT statement for reporting randomized trials: explanation and
elaboration. Ann Intern Med 2001;134:663-94.
J ALLERGY CLIN IMMUNOL
1274 SAMPSON ET AL