Veterinary Anaesthesia and Analgesia, 2006, 33, 143–148
RESEARCH PAPER
Clinical evaluation of intranasal benzodiazepines,
a2-agonists and their antagonists in canaries
Nasser Vesal DVM, MVetSc & Payman Zare DVM
Department Of Veterinary Clinical Sciences, School of Veterinary Medicine, Shiraz University, Shiraz, Iran
Correspondence: Nasser Vesal, Department Of Veterinary Clinical Sciences, School of Veterinary Medicine, Shiraz University, Postal Code
71345-1731, Shiraz, Iran. E-mail: nv1340@shirazu.ac.ir
Abstract
Objective To evaluate the effects of
intranasal benzodiazepines (midazolam and diaze-
pam), a2-agonists (xylazine and detomidine) and
their antagonists (flumazenil and yohimbine) in
canaries.
Study design Prospective randomized study.
Animals Twenty-six healthy adult domesticated
canaries of both sexes, weighing 18.3 ± 1.0 g.
Methods In Study 1 an attempt was made to
determine the dose of each drug that allowed
treated canaries to be laid in dorsal recumbency
for at least 5 minutes, i.e. its effective dose. This
involved the evaluation of various doses, during
which equal volumes of the tested drug were
administered slowly into each nostril. In study 2
the onset of action, duration and quality of sedation
induced by each drug at its effective dose were
evaluated. The efficacy of flumazenil and yohimbine
in antagonizing the effects of the sedative drugs was
also studied.
Results In study 1 administration of 25 lL per
nostril diazepam (5 mg mL)1 solution) or midazo-
lam (5 mg mL)1 solution) to each bird caused
adequate sedation within 1–2 minutes; birds did
not move when placed in dorsal recumbency. After
administration of 12 lL per nostril of either xylazine
(20 mg mL)1) or detomidine (10 mg mL)1), birds
doi:10.1111/j.1467-2995.2005.00244.
seemed heavily sedated and assumed sternal
recumbency but could not be placed in dorsal
recumbency. Higher doses of xylazine (0.5 mg per
nostril) or detomidine (0.25 mg per nostril) pro-
longed sedation but did not produce dorsal recum-
bency. In study 2 in all treatment groups, onset of
action was rapid. Duration of dorsal recumbency
was significantly longer (p < 0.05) with diazepam
(38.4 ± 10.5 minutes) than midazolam (17.1 ±
2.2 minutes). Intranasal flumazenil (2.5 lg per
nostril) significantly reduced recumbency time.
Duration of sedation was longer with a2-agonists
compared with benzodiazepines. Detomidine had
the longest duration of effect (257.5 ± 1.5 minutes)
and midazolam the shortest (36.9 ± 2.4 minutes).
Nasally administered flumazenil significantly
reduced the duration of sedation with diazepam
and midazolam while yohimbine (120 lg per
nostril) effectively antagonized the effects of xylazine
and detomidine.
Conclusion Intranasal benzodiazepines produce
rapid and effective sedation in canaries. Intranasal
a2 agonists produce sedation but not sustained
recumbency. Specific antagonists are also effective
when used by this route.
Clinical relevance Intranasal sedative drug admin-
istration is an acceptable alternative method of drug
delivery in canaries.
Keywords a2-agonists, benzodiazepines, canary,
intranasal, sedation.
143
Clinical evaluation of intranasal benzodiazepines and a2-agonists N Vesal and P Zare
Introduction
Many varieties of small pet birds are delicate and
react adversely to prolonged handling which is re-
quired during procedures such as surgical sexing,
radiography, wound dressing, blood collection,
endoscopy and fracture splinting. Chemical restraint
is necessary to avoid stress, anxiety and struggling,
which may reduce post-operative survival (Harrison
1986; Coles 1997). Although inhalation anaesthesia
is generally safer than injectable anaesthesia in birds,
anaesthetic machines with precision vapourizers
may not always be available. Several sedative and
anaesthetic drugs and drug combinations have been
used for the induction of sedation and anaesthesia in
small birds (Boever & Wright 1975; Fedde 1978;
Samour et al. 1984; Lawton 1996; Coles 1997).
However, intravenous injection is difficult and so the
intramuscular (IM) or subcutaneous routes are pre-
ferred. The former is usually made into the pectoral
muscles, which have a small mass in caged birds
such as finches, canaries, parakeets and lovebirds.
Attempted pectoral muscle injection incurs the risk
of accidental intravascular or intracoelomic drug
administration (Harrison 1986). Injections into the
thigh muscles of small birds are not recommended
because of the potential for nerve damage. Pain
associated with IM injection may be considerable,
particularly with irritant agents. In order to avoid the
pain and anxiety related to IM injections in children,
the intranasal (IN) route has been evaluated for the
induction of sedation or analgesia; several studies
have demonstrated the efficacy and safety of mida-
zolam, sufentanil and diamorphine given IN in
young children (Wilton et al. 1988; Rey et al. 1991;
Karl et al. 1992; Kendall et al. 2001). The IN
administration of anaesthetics, alone or in combi-
nation, has been reported in adult rabbits (Robertson
& Eberhart 1994).
The objective of this study was to determine
whether IN benzodiazepines (midazolam and diaze-
pam), a2-agonists (xylazine and detomidine) and
their antagonists (flumazenil and yohimbine) were
effective in canaries.
Materials and methods
The project was approved by the Institutional Ani-
mal Care and Use Committee. Twenty-six healthy
adult domestic canaries (Serinus canarius) of both
sexes were studied. The birds’ body masses ranged
from 16 to 20 g [18.3 ± 1.0 g, mean (±SD)]. All
144
birds were housed in a temperature-controlled
environment (18–20
C) with groups of three to
four birds per cage. They were acclimatized for at
least 4 weeks before the study began. The birds had
free access to water and were fed a variety of seeds,
vegetables and fruits ad libitum. They were not fas-
ted before the administration of medication.
Study 1
The objective was to determine the dose of diazepam
(5 mg mL)1; Phoenix Pharma Ltd, Gloucester, UK),
midazolam (5 mg mL)1; Dormicum; Hoffman-La
Roche Ltd, Basel, Switzerland), xylazine
(20 mg mL)1; Alfasan; Woerden, Holland) and de-
tomidine (Dormosedan, 10 mg mL)1; Orion Cor-
poration, Farmos, Finland) that produced sedation to
the extent that treated birds could be laid in dorsal
recumbency for at least 5 minutes without strug-
gling, moving or the need for further physical re-
straint. Canaries were manually restrained in a dorsal
position and equal volumes of drug administered
slowly into each nares using a micropipette (Varipet
4810; Eppendorf, Hamburg, Germany) (Fig. 1). One
minute after drug administration, each bird was
placed in dorsal recumbency in separate cages for
observation. Noise, movement and other stimuli were
minimized after drug administration.
Study 2
The objective was to evaluate the onset of action,
duration and the quality of sedation produced by each
drug using the doses that proved effective in study 1.
The efficacy of flumazenil (Anexate 0.1 mg mL)1;
Hoffman-La Roche Ltd) and yohimbine (Reverzine
10 mg mL)1; Parnell Laboratories, Alexandria,
Figure 1 Intranasal drug administration in canary.
 2006 Association of Veterinary Anaesthetists, 33, 143–148
 Clinical evaluation of intranasal benzodiazepines and a2-agonists N Vesal and P Zare

Table 1 Treatment groups performed using one-way analysis of variance

(ANOVA) followed by Duncan’s test where appropri-
Benzodiazepines ate. All results are expressed as the mean and
Group 1: Midazolam standard deviation (±SD) and differences were
A: Midazolam considered significant at p < 0.05.
B: Midazolam–saline
C: Midazolam–flumazenil
Group 2: Diazepam Results
D: Diazepam
E: Diazepam–saline
F: Diazepam–flumazenil
Study 1
a2-agonists
Within 1–2 minutes of administering either 25 lL
Group 3: Xylazine
A: Xylazine
diazepam or midazolam (125 lg) into each nostril,
B: Xylazine–saline birds did not move when placed in dorsal recum-
C: Xylazine–yohimbine bency. The dose range was 12.5–15.6 mg kg)1
Group 4: Detomidine (mean ¼ 13.6 mg kg)1). The onset of action of xyl-
D: Detomidine
azine [12 lL (0.24 mg) per nostril] and detomidine
E: Detomidine–saline
F: Detomidine–yohimbine
[12 lL (0.12 mg) per nostril] was also rapid; the birds
became heavily sedated and assumed sternal
recumbency, although it was not possible to place
Australia) in reversing the effects of sedatives was them in a dorsal position. The dose range for xylazine
also studied. Saline (control) or antagonist drugs used was 24–30 mg kg)1 (mean ¼ 26.2 mg kg)1)
were administered IN 10 minutes (flumazenil) or 15 and 12–15 mg kg)1 (mean ¼ 13.1 mg kg)1) for
minutes (yohimbine) after the administration of detomidine. Higher doses of xylazine (0.5 mg per
benzodiazepine or a2-agonists, respectively. Canaries nostril) or detomidine (0.25 mg per nostril) failed to
were randomly assigned to one of two treatment induce dorsal recumbency, although the duration of
groups (Table 1) at weekly intervals in a crossover sedation was more prolonged. Consequently, the
design (five birds per group). Each bird was used two lower dose was used in study 2.
or three times. Following each drug trial, a radio-
paque dye, ioxitalamate meglumine (Telebrix; Labo-
Study 2
ratoire Guerbet, France) was instilled into the nares in
the same manner as described for drugs and radio- The onset and duration of each treatment are
graphy performed 5, 10 and 15 minutes thereafter. shown in Table 2. The onset of action was rapid in
Statistical analysis of data (onset and duration of all treatment groups, and was longest with detom-
sedation and duration of dorsal recumbency) was idine (2.9 ± 1.5 minutes) although there were no

Table 2 Time to onset and duration

of sedation and dorsal recumbency Onset time Duration of dorsal Duration of
(mean ± SD) with each drug treat- Drug(s) (minutes) recumbency (minutes) sedation (minutes)
ment
Diazepam 38.4 ± 23.5* 69.2 ± 24.4*
Diazepam–saline 1.4 ± 0.6 48.2 ± 22.9 73.6 ± 18.9
Diazepam–flumazenil 22.0 ± 9.9  25.4 ± 7.0 
Midazolam 17.1 ± 5.0 36.9 ± 5.3
Midazolam–saline 1.9 ± 1.0 21.7 ± 12.7 39.7 ± 13.4
Midazolam–flumazenil 9.3 ± 2.1  22.4 ± 3.6 
Xylazine – 115.2 ± 41.6à
Xylazine–saline 1.4 ± 0.3 – 172.3 ± 26.0
Xylazine–yohimbine – 19.9 ± 9.9 
Detomidine – 257.5 ± 16.8
Detomidine–saline 2.9 ± 1.5 – 271.8 ± 23.0
Detomidine–yohimbine – 168.6 ± 11.6 

*Significantly different from midazolam (p £ 0.05);  significantly different from control

value (p £ 0.05); àsignificantly different from detomidine (p £ 0.05).

 2006 Association of Veterinary Anaesthetists, 33, 143–148 145

Clinical evaluation of intranasal benzodiazepines and a2-agonists N Vesal and P Zare
significant differences between groups. The duration
of dorsal recumbency was significantly longer
(p < 0.05) with diazepam (38.4 ± 23.5 minutes)
compared with midazolam (17.1 ± 5.0 minutes).
Intranasal flumazenil significantly reduced
recumbency time (p < 0.05) in birds receiving
benzodiazepines. Xylazine and detomidine did not
allow birds to be placed in dorsal recumbency.
The duration of sedation was longer with
a2-agonists (257.5 ± 16.8 minutes; detomidine)
compared with benzodiazepines (36.9 ± 5.3 min-
utes; midazolam) (p < 0.05). Intranasal antagonist
[flumazenil, 25 lL (2.5 lg) per nostril, or yohim-
bine 12 lL (120 lg) per nostril] significantly
reduced the duration of sedation produced by the
benzodiazepines and a2-agonists respectively. The
dose range used was 0.25–0.31 mg kg)1 (mean ¼
0.27 mg kg)1) and 12–15 mg kg)1 (mean ¼
13.1 mg kg)1) for flumazenil and yohimbine
respectively. Saline administration tended to in-
crease the duration of recumbency (with ben-
zodiazepines) and the duration of sedation (with
benzodiazepines and a2-agonists) although the dif-
ference was not significant (p > 0.05).
Canaries accepted IN administration of drugs
without resistance. Increased appetite was observed
in birds receiving benzodiazepines after sedation had
waned. Muscle tremors and movement in response
to noise were frequently observed in canaries
receiving xylazine or detomidine. An increased
frequency of defecation was noted following a2-
agonist administration. Radiography revealed the
presence of radio-opaque material in the nasal
cavity, but not in the crop, stomach, trachea or
lung. No adverse reactions or complications were
encountered after IN drug administration in this
study.
Discussion
This study showed that IN drug administration
could provide fast and reliable sedation in canaries.
Antagonist agents were also effective when admin-
stered by this route. In our study, IN drug delivery
was well tolerated and no complications were
observed. However physiologic data were not recor-
ded. All canaries had a smooth recovery and did not
show signs of respiratory distress during sedation.
The results of this study indicate that IN mida-
zolam or diazepam can provide adequate sedation
for diagnostic and minor therapeutic procedures.
Midazolam has the advantage of water solubility
146
and a shorter duration of action. Midazolam is a
short-acting, water-soluble benzodiazepine which is
slightly more potent than diazepam and has min-
imal cardiopulmonary effects (Ludders & Matthews
1996). Intramuscular injection caused no signifi-
cant changes in cardiopulmonary function in
Canada geese, pigeons and quail (Valverde et al.
1990; Smith & Muir 1992; Day & Roge 1996).
Intranasal midazolam has been successfully used for
pre-anaesthetic sedation in children, as it avoids the
discomfort associated with IV or IM injection. Mild
side effects such as transient burning sensation and
lacrimation have been reported, which may result
from its low pH (approximately 3.0–3.5). Olfactory
activity was not affected in children (Rosario et al.
1990; Lugo et al. 1993). Intranasal midazolam
decreased the respiratory rate in rabbits but had no
effect on haemoglobin saturation or on the values of
venous blood gas variables. Changes in arterial
oxygen saturation have not been observed following
IN midazolam administered in children (Wilton
et al. 1988; Karl et al. 1992). Endotracheal diaze-
pam caused deleterious histological changes in the
lungs of cats (Rusli et al. 1987) in contrast to
endotracheal midazolam, which produced no path-
ologic changes in the lungs of lambs (Jaimovich
et al. 1992). This difference in pulmonary irritancy
has been attributed to the presence of propylene
glycol in diazepam solutions.
The precise site of drug absorption was not
determined in this study, although the rapid onset
of sedation indicated that the drugs were probably
absorbed across the nasal, oral and/or pharyngeal
mucosae. In the present study, oral administration
of midazolam or diazepam did not produce desirable
sedation in two canaries. Radiography indicated
that instilled fluids remained in the nasopharyngeal
region, although the viscosity of contrast media
used in this study may not have been the same as
that of the drugs used. The nasal mucosa represents
a highly permeable route of drug absorption, and a
number of drugs are absorbed systemically at this
site. Intranasal administration of epinephrine during
cardiac arrest and cardiopulmonary resuscitation
has been attempted in dogs (Bleske et al. 1992)
while IN sedatives and analgesics have been used in
human paediatric patients (Wilton et al. 1988; Rey
et al. 1991; Karl et al. 1992; Kendall et al. 2001).
The attainment of significant plasma midazolam
concentrations occurs as early as 10 minutes after
IN administration in human beings (Walbergh et al.
1991). The mean onset and duration of sedation
 2006 Association of Veterinary Anaesthetists, 33, 143–148
 Clinical evaluation of intranasal benzodiazepines and a2-agonists N Vesal and P Zare
following IN midazolam in rabbits was 3.0 and
24.6 minutes respectively (Robertson & Eberhart
1994). The bioavailability of IN midazolam (55–
57%) is higher than that resulting from the use of
the oral and rectal routes; this is attributed to the
absence of first-pass metabolism (Rey et al. 1991;
Burstein et al. 1997).
Recovery from IN midazolam or diazepam in the
current study was completed in all birds within
45 minutes, while full recovery from IN xylazine
and detomidine took hours. Furthermore, the plu-
mage appeared ruffled in birds receiving a2-agon-
ists, which remained sluggish for prolonged periods.
In contrast, birds recovering from midazolam or
diazepam were alert and appeared to be hungry.
There were no complications associated with the
prolonged recovery following a2-agonist adminis-
tration, but these were nevertheless felt to be
undesirable: by preventing eating, hypothermia
and hypoglycaemia resulting from the bird’s high
metabolic rate is likely to be aggravated (Harrison
1986; Coles 1997). With a2-agonists the degree of
sedation may be insufficient to perform a procedure.
In addition, a2-agonists cause bradycardia and
partial atrioventricular heart block, respiratory
depression and muscle tremors in most species
(Ludders & Matthews 1996).
Instilling saline after agonist drugs tended to
increase the duration of sedation and recumbency.
This may have resulted from a flushing effect and
drug dispersion over a wider area. Although the
surface area and volume of the nasal cavity were
not investigated in this study, the optimum volume
for IN administration in canaries appears to be
approximately 0.02–0.03 mL per nostril. Antagon-
ist drugs were effective when given IN. Full recovery
occurred within 10–15 minutes after flumazenil
administration in midazolam or diazepam groups.
Flumazenil has been used successfully to reverse
midazolam sedation in pigeons (Smith & Muir
1992) and quail (Day & Roge 1996). Administra-
tion of yohimbine resulted in rapid arousal in
canaries given xylazine, but it was less effective in
the detomidine group. A more specific a2-antagon-
ist, e.g. atipamezole, may have been more effective
in antagonizing the sedative effects of detomidine.
Yohimbine is an effective reversal agent for xyla-
zine/ketamine anaesthesia in raptors and budgeri-
gars (Degernes et al. 1988; Heaton & Brauth 1992).
Tolazoline, another a2-antagonist, has been used to
successfully reverse xylazine/ketamine anaesthesia
in turkey vultures (Allen & Oosterhuis 1986). The
 2006 Association of Veterinary Anaesthetists, 33, 143–148
use of these reversal agents may render the use of
xylazine safer and more practical for birds. The
increased frequency of defecation following xylazine
or detomidine administration observed in the
current study may be attributed to the diuretic
effect of a2-agonists.
Conclusion
Intranasal drug administration appears to be an
acceptable noninvasive alternative method of drug
delivery in canaries. The method requires no special
technical skills and only a short period of restraint is
needed for intranasal drug installation. This route of
drug administration may supercede intramuscular
or subcutaneous routes in avian medicine. Intra-
nasal midazolam or diazepam can be used effectively
to provide adequate sedation and muscle relaxation
for diagnostic procedures. Further studies are nee-
ded to determine whether these and other drugs are
safe and effective for use in clinical cases.
Acknowledgements
This research was financially supported by grant no.
79-VE-1340-C111 of the Research Council of Shiraz
University.
References
Allen JL, Oosterhuis JE (1986) Effects of tolazoline on
xylazine-ketamine-induced anesthesia in turky vultures.
J Am Vet Med Assoc 189, 1011–1012.
Bleske BE, Warren EW, Rice TL et al. (1992) Comparison
of intravenous and intranasal administration of epi-
nephrine during CPR in canine model. Ann Emerg Med
219, 1125–1130.
Boever WJ, Wright W (1975) Use of ketamine for restraint &
anesthesia of birds. Vet Med/Small Anim Clin 70, 86–88.
Burstein AH, Modica R, Hatton M et al. (1997) Pharma-
cokinetics and pharmacodynamics of midazolam after
intranasal administration. J Clin Pharmacol 37, 711–
718.
Coles BH (1997) Avian Medicine and Surgery (2nd edn).
Blackwell Science, Oxford, UK, pp. 125–147.
Day TK, Roge CK (1996) Evaluation of sedation in quail
induced by use of midazolam and reversed by use of
flumazenil. J Am Vet Med Assoc 209, 969–971.
Degernes LA, Kreeger TJ, Mandsager R et al. (1988)
Ketamine-xylazine anesthesia in red-tailed hawks with
antagonism by yohimbine. J Wildl Dis 24, 322–326.
Fedde MR (1978) Drugs used for avian anesthesia: a re-
view. Poultry Sci 57, 1376–1399.
147
Clinical evaluation of intranasal benzodiazepines and a2-agonists N Vesal and P Zare
Harrison GJ (1986) Anesthesiology. In: Clinical Avian
Medicine and Surgery. Harrison GJ, Harrison LR (eds.),
WB Saunders, Philadelphia, PA,, pp. 549–559.
Heaton J, Brauth SE (1992) Effects of yohimbine as a
reversing agent for ketamine-xylazine anesthesia in
budgerigars. Lab Anim Sci 42, 54–56.
Jaimovich DG, Osborne JS, Shabino CL (1992) Compar-
ison of intravenous and endotracheal administration of
midazolam and the effects on pulmonary function
and histology in the lamb. Ann Emerg Med 21,
480–485.
Karl HW, Keifer AT, Rosenberger JL et al. (1992) Com-
parison of the safety and efficacy of intranasal midazo-
lam or sufentanil for preinduction of anesthesia in
pediatric patients. Anesthesiology 76, 209–215.
Kendall JM, Reeves BC, Latter VS (2001) Multicenter
randomised controlled trial of nasal diamorphine for
analgesia in children and teenagers with clinical frac-
tures. Br Med J 322, 261–265.
Lawton MPC (1996) Anaesthesia. In: Manual of Psittacine
Birds. Forbes NA, Lawton MPC (eds). British Small
Animal Veterinary Association, Gloucestershire, pp. 49–
59.
Ludders JW, Matthews N (1996) Anesthesia and
immobilization of birds. In: Lumb and Jones’ Veterin-
ary Anesthesia (3rd edn). Thumon JC, Tranquilli WJ,
Benson GJ (eds). William & Wilkins, Baltimore, MD,
pp. 645–669.
Lugo RA, Fishbein M, Nahata MC et al. (1993) Compli-
cations of intranasal midazolam. Pediatrics 92, 638
(letter).
148
Rey E, Delaunay L, Pons G et al. (1991) Pharmacokinetics
of midazolam in children: comparative study of intra-
nasal and intravenous administration. Clin Pharmacol
41, 355–357.
Robertson SA, Eberhart S (1994) Efficacy of intranasal
route for administration of anesthetic agents to addult
rabbits. Lab Anim Sci 44, 159–165.
Rosario M, Alves I, Luis AS (1990) Intranasal midazolam
for sedation in upper gastrointestinal endoscopy. Gas-
troenterology 98, A10 (abstract).
Rusli M, Spivey WH, Bonner H et al. (1987) Endotracheal
diazepam: absorption and pulmonary pathologic effects.
Ann Emerg Med 16, 314–318.
Samour JH, Jones DM, Knight JA et al. (1984) Compar-
ative studies of the use of some injectable anaesthetic
agents in birds. Vet Rec 115, 6–11.
Smith J, Muir WW (1992) Cardiopulmonary effects of
midazolam and flumazenil in racing pigeons. Vet Surg
21, 499 (abstract).
Valverde A, Honeyman VL, Smith DA et al. (1990)
Determination of a sedative dose and influence of mid-
azolam on cardiopulmonary function in Canada geese.
Am J Vet Res 51, 1071–1074.
Walbergh EJ, Wills RJ, Eckhert J (1991) Plasma concen-
trations of midazolam in children following intranasal
administration. Anesthesiology 74, 233–235.
Wilton NCT, Leigh J, Rosen DR et al. (1988) Preanesthetic
sedation of preschool children using intranasal midazo-
lam. Anesthesiology 69, 972–975.
Received 1 August 2004; accepted 24 March 2005.
 2006 Association of Veterinary Anaesthetists, 33, 143–148