Veterinary Dermatology 2003, 14, 323 332

Evaluation of intravenous fluorescein in intradermal allergy testing

in psittacines
Blackwell Publishing Ltd.

CLAUDIA S. NETT*, GISELLE HOSGOOD*, J. JILL HEATLEY, CAROL S. FOIL* and

THOMAS N. TULLY JR*
*Department of Clinical Sciences, School of Veterinary Medicine, Louisiana State University,
Baton Rouge, LA 70803, USA
Department of Clinical Sciences, College of Veterinary Medicine, Auburn University, AL 36849, USA
(Received 2 January 2003; accepted 30 May 2003)

Abstract This study was designed to improve the clinical feasibility of intradermal skin testing of psittacine birds
using intravenous fluorescein stain. Twenty-five healthy, anaesthetized Hispaniolan Amazon parrots (Amazona
ventralis) were injected intravenously with 10 mg kg1 fluoresceinsodium 1% followed by intradermal injections
of 0.02 mL phosphate-buffered saline, histamine phosphate (1:100 000 w/v) and codeine phosphate (1:100 000 w/v)
at the sternal apteria. Wheal diameters of reaction sites were measured grossly and under illumination with a
Woods lamp after 5 and 10 min. Fluorescence-enhanced injection sites were scored between 0 and 2, with 0
equivalent to normal skin and 2 equivalent to a plucked feather follicle. The presence of a fluorescent halo around
intradermal injections was also recorded. Under Woods light illumination at 10 min, histamine and saline were
evaluated as positive and negative controls, respectively, based on a positive test having a halo and a score of 2.
Sensitivity and specificity were each 76% for halo, 84 and 42% for score and 64 and 77% for combination of score
and halo, respectively. Further, mean histamine reactions were significantly larger than codeine phosphate and
saline (8.8 0.4 mm; 7.2 0.3 mm; 5.9 0.6 mm); however, this finding was not consistent in individual birds.
Wheal size, halo presence and score were affected by site location independent from the injected compound. Intravenous fluorescein improved the readability of avian skin tests; however, the compounds tested raised inconsistent
reactions in wheal size, score or halo presence. The compound-independent site effect raises concern on the validity of avian skin testing and warrants investigation of other techniques such as in vitro allergy testing. Based on
our findings, intradermal allergy testing in psittacines with or without fluorescein is unreliable and cannot be recommended for practical clinical use.
Keywords: avian, codeine phosphate, histamine, intravenous fluorescein, skin testing, ultraviolet light, Woods
lamp.

INTRODUCTION
Pathological feather-picking is a frequently observed
dermatological syndrome in psittaciformes. Underlying aetiologies include malnutrition, internal organ
pathology and behavioural problems, as well as a
variety of infectious agents including parasites, bacteria,
fungi and viruses. Idiopathic cases, which are common
and considered pruritic are often assumed to be caused
by psychological disorders or cutaneous allergies.19
Allergies and behavioural disorders are believed to be
the cause of many cases in which no obvious aetiological agent or disease is identified. Further evidence for
an underlying allergic aetiology includes the clinical
observation that antihistamines, essential fatty acids,
steroids, and dietary and environmental changes have
been successful in the treatment of feather plucking
birds.7,9,10 In addition, Macwhirter et al. described

Correspondence: Claudia S. Nett, School of Veterinary Medicine,

Louisiana State University, Skip Bertman Drive, Baton Rouge,
LA 70803, USA. E-mail: cnett@vetmed.lsu.edu.
2003 European Society of Veterinary Dermatology

positive intradermal skin test reactions to common

environmental allergens in feather picking birds.7
In many mammalian species intradermal allergy
testing (IDT) is the gold standard to identify offending
allergens in allergic dermatitis. Two previous studies
investigating IDT in birds encountered difficulties
in providing gross evidence of positive reactions to
injected allergens.7,8 One protocol evaluated IDT in
healthy psittacines and based on their findings the
authors recommended codeine phosphate as the most
suitable positive control substance and the proventer
region as an appropriate skin test site. The optimal
injection volume was found to be 0.02 mL and the most
appropriate reading time was defined to be 5 min after
IDT.8 However, although statistically significant, the
differences between positive and negative reactions were
too small for routine use in a clinical setting, because
the most common scoring method is subjective and
relies upon erythema, induration and wheal size.
In an attempt to provide more clinically useful
results, we designed an IDT protocol that used intravenous fluorescein; a similar protocol is available for
cats. Cats are another species in which it is difficult to
323

324

C. S. Nett et al.

subjectively score injection sites because of poorly

developed wheal and flare reactions. Differentiation
of positive from negative reaction sites in cats was
improved by the accumulation at injection sites of previously injected IV fluorescein, which is readily visible
with the aid of a ultraviolet light source (Woods lamp).11
The purpose of the study was to investigate whether
intravenous injection of fluorescein in psittacine birds
undergoing IDT, and the subsequent use of a Woods
lamp illumination enhanced the readability of sites
allowing for more consistent differentiation of negative
and positive reaction sites.

MATERIALS AND METHODS

Birds
Twenty-five Hispaniolan Amazon parrots (Amazona
ventralis), originally obtained from the Department of
Natural and Environmental Resources, Areceibo, Puerto
Rico, were housed at the Louisiana State DLAM facility from 2000 to 2002. All birds were allowed food and
water ad libitum and cared for according the guidelines
set forth in the NIH guides for the care and use of
laboratory animals. All birds were healthy and evaluated
by veterinarians on a regular basis.
At the time of study no bird had gross evidence of
skin disease.
Intradermal skin testing procedure
For the skin testing procedure, birds were anaesthetized employing a mask and isoflurane as the inhalant
anaesthetic agent at 2.5% with an oxygen flow rate
of 1 L min1, as needed for maintenance. The skin on
both sides of the keel was prepared using a water swab
to moisten and part the feathers over the proventer area.
No feathers were plucked at the injection sites to avoid
skin trauma. Using a black permanent marking pen,
six dots were made on the unfeathered skin over the
keel area to serve as guides for the intradermal injections.
Intradermal injections of 0.02 mL of each test compound were applied on either side of the keel area
next to a dot in a staggered pattern using a 30-G needle
(Fig. 1). Compounds included phosphate-buffered
saline (as a negative control), histamine phosphate
(histamine phosphate 0.275 mg mL1, Center Laboratories, Port Washington, NY) at 1:100 000 w/v (as a
putative positive control) and codeine phosphate
(codeine phosphate injection, USP 30 mg mL1, Abbott
Laboratories, North Chicago, IL) at 1:100 000 w/v
(as a putative positive control).8 Each compound was
administered in duplicate in each bird. Skin test site,
injection volumes, reading times, and compounds and
their concentrations were selected based on the findings
of a previously published study.8 Although histamine
yielded poor results as a positive control in previous
studies, it was re-evaluated in combination with fluorescein in this study, as it is a readily accessible compound and because it is a widely used standard positive
control in mammalian pets.7,8 To exclude injection-

Figure 1. Proventer region; Hispaniolan parrot, case 9. The parrot

was placed in dorsal recumbency. Six dots were made on the skin
over the keel area to serve as guides for intradermal injections.
Injections were applied on either side of the keel area next to a dot
in a staggered pattern using a 30-G needle.

Figure 2. Right wing (ventral aspect); Hispaniolan parrot, case 7.

Fluorescein was injected into the superficial ulnar vein prior to
intradermal injection.

associated oedema and trauma of the skin and related

extravasation of fluorescein, skin beneath one dot on
the right side of the keel remained untouched and
served as negative control. To discern false-positive
reactions associated with the trauma of the pinprick,
skin beneath another dot on the left side of the keel
was only pricked with the needle but no compound was
injected. Also a small feather adjacent to the injection
sites was plucked immediately prior to fluorescein
injection to serve as a (traumatic) positive control.
Fluoresceinsodium 10% ophthalmic solution (Fluorescite 10%, Alcon Laboratories, Inc. Fort Worth,
TX) was diluted to 1% with sterile saline for injection
for use in each bird at a dose of 10 mg kg1. The dye was
injected into the superficial ulnar vein or basilic vein
immediately prior to intradermal allergy testing (Fig. 2).
The fluorescein dose was selected based on preliminary
data obtained in six psittacines, in which dosages of 10,
20, 30 and 40 mg kg1 fluorescein were injected. All six
birds showed fluorescence at feather plucked sites.
However, at 30 and 40 mg kg1 untraumatized skin

2003 European Society of Veterinary Dermatology, Veterinary Dermatology, 14, 323 332

Intravenous fluorescein and intradermal allergy testing in psittacines

325

also began to fluoresce. In this pilot study no adverse

effects to IV fluorescein injection were noted in any of
the birds immediately after IV injection, at the end of
the procedure or 24 h later.
Heart rate and respiratory rate were recorded prior
to IV fluorescein injection, immediately after the injection and at the end of the procedure to monitor for possible systemic side effects associated with fluorescein
injection. Birds were watched carefully for other side
effects including salivation and vomiting upon anaesthetic recovery.
Because the possibility existed that the order in
which the compounds were injected could have an
effect on the results, the injection protocols were
predetermined using a randomized block design that
accounted for all possible site agent combinations.
Twenty-five of these protocols were randomly selected
and used in this study. The principal investigator was
blinded to the injected compounds.
Scoring of sites
At 5 and 10 min post injection, skin test sites were
measured using digital calipers under normal light
conditions as well as under the Woods light illumination. Subjective grading was also performed under
Woods lamp illumination. The fluorescence at the
untouched skin site served as a negative control and
was graded 0, whereas the fluorescence at the plucked
feather follicle served as a positive control and was
graded 2. Fluorescence at the injection sites was
graded as 0, 1 or 2, depending on the intensity of the
dye at the injection sites compared with the negative
and the positive control. A score of 1 represented fluorescence intensity somewhat between the negative and
the positive control.
In a pilot study, a rim of increased fluorescence
around the bleb of the injected compound was perceived at various injection sites when using the Woods
lamp for skin test reading (Fig. 3). The presence or
absence of this halo effect was reported for each injection site.
Twenty-four hours after intradermal allergy testing,
birds were evaluated for local and systemic side effects
of intravenous fluorescein administration. Skin test
sites were again inspected under direct and under
Woods light for any skin reaction that may have indicated a possible late phase reaction.
Statistics
The subjective fluorescent grades of skin test sites were
considered as ordinal categorical variable and were
evaluated using CochranMantelHaenszel methods
for repeated categorical data. Each bird served as its
own control and the data were analysed blocking by
bird. A significant difference between the injected
compounds was considered at P < 0.05 based on the
Chi statistic.
Wheal size (mm) under direct light and a Woods
lamp was considered continuous data, found to
follow a normal distribution with rejection of the null

Figure 3. Proventer region; Hispaniolan parrot, case 20. Intradermal reaction sites were evaluated under the Woods lamp.
Blebs at injection sites were readily detectable under the Woods
lamp. Intensity of fluorescence was most significant at cranial
reaction sites. Sites IV show the presence of a halo.

hypothesis of normality at P 0.05 using the ShapiroWilk

test. The effect of compound and time on the size of the
wheal under direct light and a Woods lamp was evaluated using a mixed effect linear model that accounted
for the random variance of bird and injection site, and
the repeated measurements on each bird. Where there
was significant interaction of compound type and time,
least squares means comparisons were made between
and within compounds to determine where significant effects occurred. Type I error was maintained at
0.05 for all between and within compound comparisons.
The data were summarized as mean SEM.
Because gross observation indicated that there might
be a difference in the size of the wheal based solely on
the site of injection, the analysis was also run with the
effect of site as a fixed effect. Where there was significant interaction of site and compound at P 0.05,
predetermined comparisons were made using least
squares means maintaining type I error at 0.05.
The association between compound type and time
with the presence of a halo and with wheal score was
evaluated using logistic regression. The model also
accounted for injection site. Significant association was
considered when the 95% confidence interval of the
odds ratio excluded 1.0.
, , and
were used for the analyses ( v 8.2, SAS
Institute, Cary, NC).

2003 European Society of Veterinary Dermatology, Veterinary Dermatology, 14, 323332

326

C. S. Nett et al.

RESULTS
Twenty-five birds underwent general anaesthesia and
injection of 10 mg kg1 intravenous fluorescein prior
to intradermal allergy testing with three compounds.
No complications were noted during the study or
24 h later.
All data are summarized and reported below. For
the purpose of statistical analysis, data from the
untouched skin site, feather site and needle site were
excluded as these were used for the purpose of scoring
but were not of statistical relevance
Summary of evaluation of skin test sites
Direct light. At 5 and 10 min post injection test sites
lacked erythema and induration under normal light
conditions. Gross observation did not allow distinction
of negative and positive reaction sites. Within compound type statistical analysis, no difference in wheal
size between the 5- and 10-min reading point was seen.
Codeine phosphate and histamine sites were significantly larger than saline sites at 5 and 10 min post
injection, but there was no difference between these
positive controls (Fig. 1). However, when individual
birds were examined, 17/25 (68%) of birds showed their
largest saline reaction to be bigger than their smallest
histamine or codeine phosphate reaction, independent
of the reading time. At 5 min post injection, 21/25
(84%) birds showed their largest saline reaction to be
bigger than their smallest histamine or codeine phosphate reaction, whereas at 10 min post injection, saline
was larger than codeine phosphate in 21/25 (84%) and
larger than histamine in 18/25 (72%) birds.
Scores and halo effect. When examining injection sites
under the Woods lamp illumination, blebs at most
injection sites could readily be detected because of
various intensities of fluorescence (Table 1, Fig. 3). At
10 min post injection, 97.7% (169/173) of the reactions
were scored 1 or 2 (Table 1). The untouched skin and
needle prick sites did not fluoresce in any bird, whereas
the feather plucked site consistently showed marked
fluorescence at the follicular opening.

5 min
Saline
Codeine phosphate
Histamine
Total
10 min
Saline
Codeine phosphate
Histamine
Total

There was significant association between compound and time with respect to score. Compared to
saline, an increased score (i.e. 01 or 12) was 3.3 times
more likely with codeine phosphate [95% confidence
interval (CI) 1.57.0] and 10.4 times more likely with
histamine (95% CI 4.523.8). Compared with codeine
phosphate, an increase in score with histamine was
3.2 times more likely (95% CI 1.66.1). Compared with
5 min, an increase in score was 3.2 times more likely
at 10 min (95% CI 1.95.3). When scores of saline and
histamine sites were used as a test to determine positive
and negative reaction sites (negative reaction = score 0
or 1, positive = score 2), sensitivity at 5 min was 62%
(31/50) and specificity was 58% (29/50). At 10 min,
sensitivity was 84% (42/50) and specificity was 42% (21/
50), whereas the positive predictive value was 69.2%
(42/71) and the negative predictive value was 72.4%
(21/29) (Table 1).
There was significant association between time and
compound type with the presence of a halo (Table 2).
Compared with saline, obtaining a halo with histamine
was 7.7 times more likely (95% CI 3.915.4). The odds
of having a halo with codeine phosphate were not different from saline. Compared with 5 min, obtaining a
halo at 10 min was 1.8 times more likely (95% CI 1.1
2.9). If the presence or absence of a halo was used to
distinguish positive (histamine) from negative (saline)
sites, the sensitivity at 5 min was 54% (27/50) and the
specificity was 82% (41/50). At 10 min sensitivity was
76% (38/50), specificity was 76% (38/50), whereas the
positive predictive value was 76% (38/50), and the negative predictive value was 76% (38/50) (Table 2).
When the presence of halo and a score of 2 were
required to account for a positive reaction using histamine as a positive and saline as a negative control at
10 min, the sensitivity was 64% (32/50) and the specificity was 78% (39/50). (Data not shown, but available
upon request.)
Ultraviolet light. For each compound, wheals appeared
larger under the Woods lamp illumination than
with direct light (Table 3, Fig. 4). Within compound
type histamine wheals were significantly larger at
10 min post injection compared with the 5-min reading

0 Reactions
No. (%)

+1 Reactions
No. (%)

+2 Reactions
No. (%)

5 (10.0)
4 (8.0)
7 (14.0)
16 (10.7)

24 (48.0)
23 (46.0)
12 (24.0)
59 (39.3)

21 (42.0)
23 (46.0)
31 (62.0)
75 (50.0)

1 (2.0)
2 (4.0)
1 (2.0)
4 (2.7)

20 (40.0)
19 (38.0)
7 (14.0)
46 (30.7)

29 (58.0)
29 (58.0)
42 (84.0)
100 (66.7)

Twenty-five birds were tested in duplicate therefore there were 50 injection sites scored for each
test compound. Intradermal skin test sites were scored as 0, +1, or +2.
2003 European Society of Veterinary Dermatology, Veterinary Dermatology, 14, 323 332

Table 1. Scores for reaction sites at 5 and

10 min under the Woods light

Intravenous fluorescein and intradermal allergy testing in psittacines

Table 2. Observation of halo at 5 and 10 min
Halo absent
No. (%)

Halo present
No. (%)

5 min
Saline
CP
Histamine
Total

41 (82.0)
39 (78.0)
23 (46.0)
103 (68.7)

9 (18.0)
11 (22.0)
27 (54.0)
47 (31.3)

10 min
Saline
CP
Histamine
Total

38 (76.0)
36 (72.0)
12 (24.0)
86 (57.3)

12 (24.0)
14 (28.0)
38 (76.0)
64 (42.7)

Twenty-five birds were tested in duplicate therefore there were 50

injection sites scored for each test compound.

327

point; whereas codeine phosphate and saline did not

differ within compound either at the 5 or 10 min
reading.
Between compounds, at 5 min, histamine wheals
were significantly larger than with codeine phosphate
and saline, but saline and codeine phosphate were not
different. At 10 min, histamine wheals were significantly larger than codeine phosphate (1.55 mm/21.5%)
and saline (2.87 mm/49%), and codeine phosphate
was significantly larger than saline (1.32 mm/22%)
(Table 3).
However, when individual birds were examined and
the largest saline wheal was compared with the smallest
codeine phosphate or histamine wheal, 12/25 (48%)
birds showed their saline was the largest reaction at

Table 3. Mean (SEM) wheal diameter under direct light and Woods light
Wheal diameter (SEM)
Time

Saline

CP

Histamine

Comparisons of significance at times

(5/10 = 5 min/10 min)

Direct light at 5 min

Woods lamp at 5 min
Direct light at 10 min
Woods lamp at 10 min

5.28 (0.24)
6.204 (0.29)
5.101 (0.28)
5.89 (0.32)

5.83 (0.08)
6.77 (0.14)
6.01 (0.09)
7.21 (0.17)

5.99 (0.11)
8.06 (0.21)
5.78 (0.11)
8.76 (0.2)

CP, H > S
H > CP, S H5 > H10
CP, H > S
H > CP, S CP > S

Mean (SEM) wheal diameter using 0.02 mL injections in 25 birds compared under direct light and Woods light at 5- and 10-min reading time.
Under direct light: codeine phosphate and histamine are significantly different from saline at 5 and 10 min. Under Woods light: histamine at
5 min is significantly different from histamine at 10 min and from codeine phosphate and saline at 5 min. At 10 min: histamine is significantly
different from codeine phosphate and saline, and codeine phosphate is significantly different from saline. Within compound: saline under Woods
light is significantly different from saline under direct light at 5 and 10 min. All other reactions were not statistically different from each other.

Figure 4. Proventer region; Hispaniolan parrot, case 24. Comparison of the same skin test evaluated first under direct light (a) and then under
ultraviolet light (b). Wheals at all injection sites appear larger under the Woods lamp.
2003 European Society of Veterinary Dermatology, Veterinary Dermatology, 14, 323332

328

C. S. Nett et al.

5 min and 5/25 (20%) at 10 min, respectively. In 13/25

(52%) birds and in 12/25 (48%) birds, saline was larger
than histamine at 5 min and at 10 min, respectively.
Saline wheals were larger in 84% (21/25) of birds at
5 min and 76% (19/25) of birds at 10 min (data not
shown but available on request).
Evaluation of site of injection
Direct light. Site of injection was evaluated within and
between compounds. Under direct light, saline reactions at site I were significantly larger than saline at
site II and site III at 5 and at 10 min. Codeine phosphate and histamine did not reveal a site effect within
compound.
In between sites, saline reactions were not different
from codeine phosphate or histamine reactions at any
site at 5 and 10 min (Fig. 5a,b)
When wheal sizes of the three compounds were
compared within sites, saline reactions were not different from codeine phosphate or histamine at any site at
either the 5- or 10-min readings. However, histamine
was significantly larger than codeine phosphate at site
I at 5 min (Fig. 5a) and codeine phosphate was significantly larger than histamine at site IV at 10 min
(Fig. 5b).
Ultraviolet light. Under ultraviolet light, saline reactions at site I were larger than at site III at 10 min but
not at 5 min. Histamine reactions were larger at sites I
and II than sites IV, V and VI at 5 min but not at
10 min. No difference in site effect was observed for
codeine phosphate at either 5 or 10 min.

When wheal size of all tested compounds was

compared within sites, saline was not different from
codeine phosphate at any site and histamine was
larger than saline at sites II and III only at the 5-min
reading (Fig. 5c).
When the largest saline reaction was compared with
the smallest histamine or codeine phosphate reaction,
saline at site I was not different from histamine or
codeine phosphate at site VI at 5 and 10 min.
When compounds were evaluated within sites, at
10 min, codeine phosphate reactions were larger than
saline at site III only. Histamine reactions were significantly larger than saline at sites I, II, III but not at sites
IV, V or VI (Fig. 5d).

DISCUSSION
Intradermal allergy testing is increasingly being used in
an attempt to diagnose allergic dermatitis in feather
picking psittacines when other underlying factors have
been excluded or symptomatic treatment was to no
avail. A previously published protocol suggested a suitable injection site, injection volume, reading time and
positive and negative control compounds.8 Although
this study achieved statistical significance with codeine
phosphate as the best positive control and a suggested
reading time of 5 min, the protocol lacked clinical
feasibility as it failed to allow subjective grading based
on injection site diameter and wheal and flare reaction,
parameters used practically to interpret intradermal
allergy tests in companion animals. In cats, which
similarly exhibit faint positive reactions, reading of

Figure 5. (ad): Site dependent wheal sizes (mm) under direct light and ultraviolet light at 5 and 10 min. Error bars indicate the mean standard
deviation. Within each site, compounds with the same letter are not significantly different (P 0.05).
2003 European Society of Veterinary Dermatology, Veterinary Dermatology, 14, 323 332

Intravenous fluorescein and intradermal allergy testing in psittacines

intradermal allergy testing can be facilitated by injecting
intravenous fluorescein prior to skin testing and subsequent reading with a Woods lamp.11
Fluoresceinsodium is a highly fluorescent chemical
compound synthesized from petroleum derivatives.
After intravenous injection, free fluorescein reaches
capillaries and invades interstitial tissue through vascular fenestrations and rapidly diffuses out of capillaries into the extravascular tissue.12 The uptake and
distribution pattern of fluorescein in the skin reflects
both local blood flow and capillary permeability.13 The
dye molecules fluoresce when excited with a blue (ultraviolet) light source such as the Woods lamp, which
produces light at the appropriate wavelength to excite
fluorescein. This lamp is readily available because it is
commonly used as a screening tool for dermatophytosis in domestic animals and man.14,15 Fluorescein has
been widely used as a contrast dye for investigation of
the chorioretinal microcirculation.14,16,17 Fluorescein
is metabolized by the kidneys and eliminated through
the urine 24 36 h after administration. Minimal side
effects have been reported, in animals and humans.
Adverse gastrointestinal reactions are the most
common.18,19
Fluorescent angiography in raptors evaluated filling
patterns in retinal capillaries, fluorescein reached the
ophthalmic tissue within seconds and persisted for
several hours.12,20 The observation of side effects in birds
were dose dependent and consisted of salivation, vomiting, head shaking and somnolence.12,20,21
Intravenous fluorescein was evaluated in this study
as an aid to improve intradermal skin testing protocols
in birds, because it is a small molecule that rapidly diffuses out of capillaries if increased vascular permeability exists such as in positive reaction sites. Chickens
showed marked increase of vascular permeability
when intradermally injected with histamine and several
other known inflammatory compounds. This finding
led to the hypothesis that fluorescein may increase
visualization of positive reactions when skin testing
birds.22
In this study, gross observation of the injection sites
under direct light did not allow grading of the reactions
owing to a lack of erythema and induration. Furthermore, subjective reading of the skin test was difficult
and unreliable as reported previously.7,8 Under direct
light in this study, codeine phosphate and histamine
were suitable positive controls compared to saline at
5 and 10 min post injection. These results vary from
those obtained by Colombini et al. in which codeine
phosphate produced the largest reactions at 5 min and
histamine was significantly smaller than saline at the
10 min.8 However, if compounds were compared
within sites, saline was not different from histamine
or codeine phosphate at any site regardless of time. If
individual birds were considered, saline sometimes
caused larger reactions than histamine or codeine
phosphate rendering direct light unsuitable for skin
test evaluation in birds. This difference compared with
the Colombini et al. study8 may be due to a failure to

329

randomize the injection sites, therefore not accounting

for possible site-dependent difference in wheal size,
which was observed in our study. In addition, observations in our study were blinded to compound in contrast to the prior study. The similarly small standard
mean errors (SEM) indicate a consistent injection and
reading technique in both studies.
When injection sites were read using the Woods
lamp, most reactions were graded a score 1 or 2 at
10 min and were readily detectable under ultraviolet
light. It was likely that some degree of fluorescence was
due to the injected volume associated with injection
trauma instead of a positive reaction to the injected
compound. Hence, scores of 1 and 0 were considered as
negative reactions. When skin test evaluation at 10 min
was based on score only (with saline as the negative
and histamine as the positive control), sensitivity and
specificity (false positives) were 84 and 42%, which is
unacceptable for a positive and negative control. The
low specificity further highlights the possibility that
trauma rather than compound attributes to the intensity of fluorescence. Based on the finding of this study,
score alone can not be recommended for subjective
grading of intradermal allergy tests in psittacines.
Halo presence was commonly observed at 10 min for
histamine and to a lesser extent for codeine phosphate
compared with saline. This halo effect was interpreted
as a sign of positive reaction. If a type-1 hypersensitivity reaction is truly present in avian species, mast cell
degranulation would cause vascular dilatation and
increased vascular permeability at positive injection
sites with subsequent fluorescein extravasation producing the observed halo. Halo presence at 10 min was a
poor indicator of positive reactions when saline sites
were considered negative and histamine sites positive
with a sensitivity and specificity of 76% each. These
findings suggest that halo presence occurs not only
because of increased permeability due to a positive
allergic reaction, but also may be associated with
trauma due to the injected volume. The presence or
absence of halo is unsuitable as a diagnostic criterion
for subjective grading of avian intradermal allergy
testing.
The combination of a score of 2 and the presence of
halo to define positive reactions increased specificity,
but resulted in a marked decrease in sensitivity rendering this combination unsuitable to interpret avian skin
tests.
There was a tendency to observe larger wheals under
the Woods light than under direct light which suggests
that intravenous fluorescein may improve visualization
of avian skin test reaction sites.
Histamine reactions consistently showed the largest
increase in wheal size compared with saline and
codeine phosphate at both 5 and 10 min. Saline and
codeine phosphate wheals did not increase significantly over time in contrast to histamine. This timedependent increase in histamine wheal size supports
the existence of type-1 hypersensitivity in birds and
suggests that histamine is a better positive control than

2003 European Society of Veterinary Dermatology, Veterinary Dermatology, 14, 323332

330

C. S. Nett et al.

codeine phosphate, with a suggested reading time at

10 min.
Wheal size appeared to be strongly dependent on the
site of injection. Saline reactions at site I were significantly larger than at more caudal injection sites independent of time or light condition. Under Woods light
at 5 min, histamine reactions at cranial sites were larger
than at caudal sites. This may be due to a difference in
skin pliability at various areas of the proventer region
and subsequent different distribution of the injected
compound. The increased wheal sizes, at least under
ultraviolet light, may also be associated with the fact
that sites in the cranial region of the proventer area
(site I) directly overlie the largest muscle bodies of the
pectoralis muscle. Compared with more caudal sites
where the muscle becomes smaller, this area may
have a higher blood supply which may contribute to
increased fluorescein distribution to those sites and
subsequently larger wheal diameters of positive and
negative compounds. At 10 min, the site effect for histamine faded and histamine reactions were no longer
site dependent. This may be explained by equilibration
of the previous physical effect by a potential histamine
receptor-mediated reaction. The site dependency on
wheal size has not been evaluated in previous studies
and may have contributed to the differences in previous
results from ours.7,8
Skin testing with saline as a negative and histamine
as a positive control with the use of fluorescein and
evaluation at 10 min under ultraviolet light yielded the
most significant differences in wheal size. However,
saline reactions at sites I and VI were still not different
from histamine at site VI and codeine phosphate at site
IV. Further, when compounds were compared within
the same site, histamine was significantly larger than
saline only at sites IIII; no difference was observed in
the more caudal sites. The gradient observed for histamine reactions may be explained by increased fluorescein distribution to the more cranial sites because of
higher local blood flow which augmented vascular
permeability as a reaction to histamine injection.
Codeine phosphate was only larger than saline at one
site rendering this compound unsuitable as a positive
control.
In individual birds, saline was often larger than
histamine and codeine phosphate, independent of
light condition or time. This finding lessens the usefulness of the employed compounds as valid positive and
negative controls in intradermal allergy testing in
psittacines.
We found that intravenous fluorescein prior to intradermal injections facilitates the reading of intradermal
allergy testing in psittacines with the aid of a Woods
lamp, the optimal reading time was at 10 min. However, subjective assessment based on score and the
presence or absence of a halo produced poor specificity
and sensitivity making these criteria, both alone and in
combination, unsuitable for clinical skin test evaluation.
None of the compounds tested raised consistent
reactions in wheal size, score or presence of halo. These

findings combined with compound-independent sitedependent wheal sizes, compromise the validity of this
protocol including the choice of negative and positive
controls. Based on our findings, intradermal allergy
testing in psittacines with or without fluorescein is
unreliable and cannot be recommended for practical
clinical use.
If future protocols for allergy testing are to be evaluated they should concentrate on compound injection
into or around feather follicles or the use of atopy
patch testing at follicular sites. Feather follicles are the
primary target of the great majority of plucking birds.
They contain a well-differentiated capillary bed, higher
cellularity of the dermis and high numbers of nerve
endings compared with interfollicular skin, which may
lead to a substantial wheal and flare reaction.
However, intradermal allergy testing in birds may
never become a feasible clinical tool and future investigations may be directed at developing in vitro allergy
testing based on immunoglobulin Y, the major circulating immunoglobulin, which is theorized to possess
reaginic function in birds.23

ACKNOWLEDGEMENTS
The study was funded by a grant from the Department
of Veterinary Clinical Sciences, Organized Research
Fund of Louisiana State University and supported by
the Kaytee Avian Foundation. The authors would like
to thank Ms Gaye Gomilla for her invaluable technical
assistance. This study was approved by the Animal
Care and Use Committee of the Louisiana State
University.

REFERENCES
1. Rosskopf, W., Woerpel, R. Feather-picking and therapy
of skin and feather disorders. In: Diseases of Cage and
Aviary Birds. Baltimore: Williams & Wilkins, 1996: 397
405.
2. Lung, N., Romagnano, A. Current approaches to feather
picking. In: Bonagura, J., Kirk, R., eds. Kirks Current
Veterinary Therapy XII. Small Animal Practice. Philadelphia: W.B. Saunders, 1995: 13037.
3. Rosenthal, K. Differential diagnosis of feather-picking in
pet birds. Proceedings of the Annual Conference of the
Association of Avian Veterinarians. Nashville, TN: Association of Avian Veterinarians, 1993: 10812.
4. Hillyer, E., Quesenberry, K., Baer, K. Basic avian dermatology. Proceedings of the Annual Conference of the
Association of Avian Veterinarians. Phoenix, AZ: Association of Avian Veterinarians, 1989: 10121.
5. Harrison, G. Disorders of the integument. In: Harrison,
G., Harrison, L., eds. Clinical Avian Medicine and Surgery. Philadelphia: W.B. Saunders, 1986: 50924.
6. Johnson-Delaney, C. Feather picking: diagnosis and
treatment. Journal of the Association of Avian Veterinarians 1992; 6: 823.
7. Macwhirter, P., Mueller, R. Comparison of immediate
skin test reactions in clinically normal and self-mutilating

2003 European Society of Veterinary Dermatology, Veterinary Dermatology, 14, 323 332

Intravenous fluorescein and intradermal allergy testing in psittacines

8.

9.

10.

11.

12.

13.

14.
15.

psittaciformes. Proceedings of the International Virtual

Conference in Veterinary Medicine: Diseases of Psittacine
Birds. http://www.vet.uga.edu/ivcvm/1998/macwhirter/
macwhirter.htm.
Colombini, S., Foil, C., Hosgood, G. et al. Intradermal
skin testing in Hispaniolan parrots. Veterinary Dermatology 2000; 11: 271 6.
Cooper, J., Harrison, G. Dermatology. In: Ritchie, B.,
Harrison, G., Harrison, I., eds. Avian Medicine: Principles and Application. Lake Worth, FL: Wingers, 1994:
607 39.
Johnson-Delaney, C., Nye, R. Allergies in pet birds. In:
Rosskopf, W., Woerpel, R., eds. Diseases of Cage and
Aviary Birds, 3rd edn. Baltimore: Williams & Wilkins, 1996:
397 405.
Schenkel, M., Bigler, B., Jungi, T. The use of fluorescein
for intradermal skin testing in cats (Abstract). Veterinary
Dermatology 2000; 11 (Suppl. 1): FC6.
Korbel, R., Redig, P., Jones, R. Investigations on fluorescein angiography in raptor eyes. Tieraerztliche Praxis
(Kleintiere) 1999; 27: 396 407.
Oh, D., Jones, R., Marshall, W. et al. Fluorescein dermofluorometry for the assessment of diabetic microvascular disease. Skin Research and Technology 2001; 7:
105 11.
Jacobs, J. Fluorescein sodium What is it? Journal of
Ophthalmic Photography 1992; 14: 62.
Scott, D., Miller, W., Griffin, C. Small Animal Dermatology, 6th edn. Philadelphia: W. B. Saunders, 2001: 3479.

331

16. Sargento, L., Sobral do Rosario, H., Saldanha, C. et al.

Hemorheological effects of sodium fluorescein in rats.
Clinical Hemorheology and Microcirculation 2001; 24:
17581.
17. Gelatt, K. Fluorescein angiography and fundus photography. In: Gelatt, K., ed. Veterinary Ophthalmology, 3rd
edn. Philadelphia: Lippincott, Williams & Williams,
1999: 4448.
18. Lopez-Saez, M., Tornero, P., Baeza, A. et al. Fluoresceininduced allergic reaction. Annals of Allergy, Asthma and
Immunology 1998; 81: 42830.
19. Yannuzi, L., Rohrer, K., Tindel, L. et al. Fluorescein
angiography complication survey. Ophthalmology 1986;
93: 61117.
20. Korbel, R., Nell, B., Redig, P. et al. Video fluorescein
angiography in the eyes of various raptors and mammals.
Proceedings of the Annual Conference of the Association
of Avian Veterinarians. Portland, OR: Association of
Avian Veterinarians, 2000: 8995.
21. Pettigrew, J., Wallman, J., Wildsoet, C. Saccadic oscillations facilitate ocular perfusion from the avian pecten.
Nature 1990; 343: 3623.
22. Awadhiya, R., Vegad, J., Kolte, G. A topographical study
of increased vascular permeability in acute and inflammatory reaction in the chicken skin. Research in Veterinary Science 1980; 29: 20310.
23. Warr, G., Magor, K., Higgins, D. IgY: clues to the origins
of modern antibodies. Immunology Today 1995; 16:
3928.

Rsum Cette tude a t ralise pour amliorer la faisabilit des tests intradermiques chez les perroquets en
utilisant la fluorescine par voie intraveineuse. Vingt cinq perroquets (Amazona ventralis) sains, anesthsis, ont
reu une injection intraveineuse de 10 mg kg1 d'une solution de fluorescine 1%, suivie par l'injection intradermique de 0.02 ml de solut sal, de phosphate d'histamine (1:100,000 w/v) et de phosphate de codine
1:100,000 w/v) sur le sternum. Le diamtre des plaques orties a t mesur directement et l'aide d'une lampe
de Wood aprs 5 et 10 minutes. Les sites d'injection matrialiss par une fluorescence ont t cots entre 0 et 2,
0 tant quivalent la peau normale, et 2 correspondant un follicule plumeux arrach. La prsence d'un halo
fluorescent autour des sites d'injection a galement t note. Avec illumination par lampe de Wood, 10 minutes,
l'histamine et le solut sal ont t valus comme tmoins positifs et ngatifs, en se basant sur un halo et un
score de 2 comme raction positive. La sensibilit et la spcificit taient de 76% pour le halo, 84% et 42% pour
le score et 64% et 77% pour une combinaison des deux. En outre, les ractions moyennes l'histamine taient
significativement plus importantes que les ractions la codine ou au solut sal (8.8 mm 0.4; 7.2 mm 0.3;
5.9 mm 0.6); cependant cette dcouverte n'tait pas valable pour tous les oiseaux. La taille de la plaque ortie,
la prsence d'un halo et le score taient affects par la localisation du test, indpendamment du compos inject.
L'injection intraveineuse de fluorescine amliore la lisibilit des tests cutans chez les oiseaux; cependant les composs tests ont donn des ractions inconsistentes pour la taille, le score ou la prsence d'un halo. L'existence
d'un effet dpendant de la localisation du test soulve des questions quant la validfit des tests cutans chez les
oiseaux et d'autres techniques, notamment in vitro, devraient donc tre essayes. Nos donnes permettent de ne
pas recommander l'utilisation des tests cutans, avec ou sans injection de fluorescine, chez les psittacids en
pratique clinique.
Resumen Este estudio fue diseado para mejorar la viabilidad clnica de pruebas cutneas intradrmicas en aves
psitcidas utilizando una tincin de fluorescena intradrmica. Se inyectaron veinticinco loros sanos anestesiados
de la especie Amazona Hispaniola (Amazona ventralis) por va intravenosa con 10 mg kg1 de sodio de
fluorescena 1%, seguido de inoculaciones intradrmicas de 0.02 ml de solucin salina tamponada con fosfato,
fosfato de histamina (1:100,000 p/v) y fosfato de codena (1:100,000 p/v) en la arteria esternal. Se evalu el
dimetro de los habones en los puntos de reaccin macroscpicamente y mediante lmpara de Wood a los 5 y
10 minutos. Se puntuaron los puntos de inoculacin destacados por fluorescencia entre 0 y 2 siendo 0 la piel
normal y 2 un tamao equivalente a un folculo de una pluma arrancada. Se registr tambin la presencia de un
halo fluorescente alrededor de las inoculaciones intradrmicas. Con lmpara de Wood durante 10 minutos, se
evaluaron las reacciones a histamina y suero salino como controles positivo y negativo respectivamente, basado
en el positivo mostrando un halo con puntuacin 2. La sensibilidad y especificidad fueron del 76% para el halo,
84% y 42% para la puntuacin y 64% y 77% para la combinacin de la puntuacin y halo, respectivamente.
2003 European Society of Veterinary Dermatology, Veterinary Dermatology, 14, 323332

332

C. S. Nett et al.
Adems, las reacciones medias a la histamina fueron significativamente mayores que las de fosfato de codena y
suero salino (8.8 mm 0.4; 7.2 mm 0.3; 5.9 mm 0.6); sin embargo, este hallazgo no fue constante en animales
individuales. El tamao del habn, la presencia de halo y la puntuacin se vean afectados por la localizacin
independientemente del producto inoculado. La fluoresceina intravenosa mejoraba la lectura de las pruebas
cutneas aviares; sin embargo, los productos probados inducan reacciones inesperadas en el tamao del habn,
puntuacin y presencia de halo. El factor localizacin, independiente del compuesto, genera dudas sobre la
validez de las pruebas cutneas aviares y hace recomendable investigar otras tcnicas como pruebas alrgicas
in-vitro. Basndonos en nuestros resultados, las pruebas alrgicas intradrmicas en psitcidas con o sin fluorescena
no son fiables y no pueden ser recomendadas para uso clnico prctico.
Zusammenfassung Diese Studie soll die Durchfhrbarkeit von Intrakutantests bei Psittaziden durch die intravense Verabreichung eines Fluoreszeinfarbstoffes verbessern. 25 gesunden, ansthesierten Papageien
(Hispaniolan Amazon parrots, Amazona ventralis) wurde 10 mg/kg KGW Natrium-Fluoreszein i.v. verabreicht,
gefolgt von 0.02 ml phosphat-gepufferter Kochsalzlsung, Histamin-Phosphat (1:100,000 w/v) und CodeinPhosphat (1:100,000 w/v) intracutan im Bereich der sternalen Apteria. Nach 5 und 10 Minuten wurden die
Durchmesser der gebildeten Quaddeln makroskopisch und unter Verwendung einer Woodschen Lampe
beurteilt. Die durch Fluoreszenz hervorgehobenen Injektionsstellen wurden mit Werten zwischen 0 und 2
bewertet, wobei 0 normaler Haut und 2 einem Federbalg entsprach, dessen Feder ausgerupft wurde. Es wurde
auch ber einen fluoreszierenden Halo um die intradermalen Injektionsstellen herum berichtet. Histamin- und
Kochsalzreaktionen wurden nach 10 Minuten unter Verwendung einer Woodschen Lampe als Positiv- und
Negativkontrollen ausgewertet, bzw. aufgrund der Tatsache, dass ein positives Testergebnis einem Halo und eine
Bewertung von 2 hat. Sensitivitt und Spezifitt waren jeweils76% fr den Halo, 84 und 42% fr den Bewertungsmasstab und 64% und 77% fr die Kombination von Bewertungsmasstab und Halo. Darberhinaus waren die
durchschnittlichen Histaminreaktionen signifikant grsser als die von Codeinphosphat und Kochsalz (8.8 mm
0.4; 7.2 mm 0.3; 5.9 mm 0.6); diese Befunde waren jedoch nicht konsistent bei jedem einzelnen Vogel.
Quaddelgrsse, Vorhandensein eines Halos und Bewertung wurden durch die Lokalisation unabhngig von der
Art der verabreichten Substanz beeinflusst. Die intravense Gabe von Fluoreszein erleichtert die Lesbarkeit von
Hauttesten bei Vgeln; die getesteten Substanzen riefen jedoch widersprchliche Reaktionen hinsichtlich
Quaddelgrsse, Halo und Bewertung hervor. Die von der getesteten Substanz unabhngige, auf die Lokalisation
zurckzufhrende Reaktion lsst Bedenken hinsichtlich der Validitt von Hauttesten bei Vgeln aufkommen
und rechtfertigt die Erforschung anderer Techniken, wie z.B. von in-vitro Allergiediagnostik. Aufgrund unserer
Ergebnisse ist die intradermale Allergiediagnostik bei Psittaziden mit oder ohne Fluoreszein unzuverlssig und
kann fr den praktischen klinischen Gebrauch nicht empfohlen werden.

2003 European Society of Veterinary Dermatology, Veterinary Dermatology, 14, 323 332