Veterinary Surgery

38:434–438, 2009

Comparison of Three Clinical Techniques for the Diagnosis of

Laryngeal Paralysis in Dogs

MARYANN G. RADLINSKY, DVM, MS, Diplomate ACVS, JAMIE WILLIAMS, MS, DVM,
PAUL M. FRANK, DVM, Diplomate ACVR, and TANYA C. COOPER, RVT

Objectives—To evaluate laryngeal function using 3 diagnostic techniques: echolaryngography (EL),

transnasal laryngoscopy (TNL), and laryngoscopy per os (LPO).
Study Design—Prospective clinical study.
Animals—Dogs with laryngeal paralysis (n ¼ 5) and control dogs (n ¼ 10); 5 age- and breed-
matched dogs and 5 young, breed-matched dogs.
Methods—Laryngeal function was evaluated in conscious dogs using EL. All examinations were
recorded and evaluated by separate, blinded observers upon completion of the study. The methods
were compared with a standard evaluation incorporating all clinical knowledge of the case (STD)
using sensitivity, specificity, positive, and negative predictive values.
Results—Three dogs with bilateral laryngeal paralysis requiring surgery were diagnosed as uni-
laterally affected or normal on EL. Three dogs had paradoxic motion on TNL and LPO, 2 of those
were considered normal on EL, and 1 had no motion on EL. Paralysis was diagnosed in 1 age-
matched and 3 young control dogs on EL. LPO and TNL falsely diagnosed lack of arytenoid
movement in 2 age-matched controls and 1 young control. Two age-matched and 1 young control
dog were misdiagnosed as paralyzed with TNL and LPO.
Discussion—Direct observation of the larynx allowed better evaluation of laryngeal function com-
pared with EL. TNL did not require induction of anesthesia, but did not improve the ability to
assess laryngeal function compared with LPO.
Conclusions—EL was not as effective as direct observation of the larynx. TNL did not improve the
evaluation of laryngeal function compared with LPO.
Clinical Relevance—We use LPO combined with knowledge of the clinical history and physical
examination to diagnose laryngeal paralysis in preference to EL and TNL.
r Copyright 2009 by The American College of Veterinary Surgeons

INTRODUCTION struction may occur in severely affected dogs. Differential

diagnoses include laryngeal or pharyngeal neoplasms,

I DIOPATHIC LARYNGEAL paralysis is a disease of

older, medium to giant breed dogs. Some studies re-
port an increased frequency of diagnosis in male dogs.1–6
The history may be that of a long, insidious change in
bark, decreased activity level, progressive inspiratory
stridor, and coughing or gagging. Signs may progress to
exercise intolerance. Life-threatening upper airway ob-
cervical or mediastinal inflammatory or neoplastic pro-
cesses, polyneuropathies, and generalized neuromuscular
disorders.1–6
The diagnostic plan typically includes a complete
physical examination, complete blood count, biochemical
profile, thyroid function testing, and urinalysis to iden-
tify concurrent conditions in older dogs, a complete

From the Department of Small Animal Medicine & Surgery and the Department of Anatomy & Radiology, College of Veterinary
Medicine, University of Georgia, Athens, GA.
Funded by the Veterinary Medical Experiment Station, University of Georgia. Presented at the Annual meeting of the American
College of Veterinary Surgeons, Washington, DC, October 2006.
Address reprint requests to MaryAnn Radlinsky, DVM, MS, Diplomate, ACVS, Department of Small Animal Medicine & Surgery,
College of Veterinary Medicine, University of Georgia, Athens, GA 30602. E-mail: radlinsk@vet.uga.edu.
Submitted October 2007; Accepted August 2008
r Copyright 2009 by The American College of Veterinary Surgeons

0161-3499/09
doi:10.1111/j.1532-950X.2009.00506.x
434
 RADLINSKY ET AL
neurologic examination to rule out neuropathies or ne-
uromuscular disorders, and thoracic radiographs to eval-
uate the cardiopulmonary system and structures within
the mediastinum that could affect the recurrent laryngeal
nerve.1–6 Definitive diagnosis of laryngeal paralysis is
made with laryngoscopy per os (LPO).7 Anxiolytic, an-
algesic, and anesthetic agents used to facilitate oral ma-
nipulation may inhibit laryngeal motion and lead to false
positive diagnoses. Many sedative and anesthetic agents
have been suggested for LPO, including opioids and an-
xiolytics; however, recent studies have provided informa-
tion to guide selection of anesthetic agents in dogs.7–10
Clinicians agree that dogs should be lightly anesthetized
so laryngeal reflexes are maintained.
Ideally, the diagnosis of laryngeal paralysis would be
made in the conscious dog, thus avoiding confounding
variables of differing anesthetic regimens. Echolaryngo-
graphy (EL) has been described in dogs and cats for
evaluation of laryngeal structure and function and can be
performed on animals that are completely awake, which
is of great benefit in evaluating laryngeal function.11,12
Laryngeal mass lesions, laryngeal cysts, laryngeal thick-
ening, and laryngeal collapse have been diagnosed with
EL.11 Patient motion, however, may impede accurate
evaluation of laryngeal movement.
Transnasal laryngoscopy (TNL) has been used to di-
agnose laryngeal paralysis in large breed dogs.13 The
procedure was performed on sedated dogs with topical
anesthesia of the nasal passage before induction of an-
esthesia, with the goal of decreasing the amount of se-
dation required compared with LPO. Dogs were sedated
with a tranquilizer and an opioid, eliminating the need
for general anesthesia. A 2 mm flexible endoscope passed
into the oropharynx via the left nostril allowed direct
observation of laryngeal structure and function. TNL
was successfully used to diagnose laryngeal paralysis in 3
dogs and to evaluate laryngeal function in 2 normal dogs.
Topical laryngeal stimulation was used to further eval-
uate laryngeal function in overly sedated dogs.13
Because EL, TNL, and LPO require incrementally in-
creasing levels of manual or chemical restraint in non-
emergent patients, this study was designed to compare
the ability of EL, TNL, and LPO to evaluate laryngeal
function in dogs with and without laryngeal paralysis.
MATERIALS AND METHODS
Three breed-matched groups each of 5 dogs were enrolled
in the study: dogs with laryngeal paralysis, age-matched con-
trol dogs, and young control dogs. Dogs with laryngeal pa-
ralysis were admitted for surgical intervention. Control dogs
were selected from patients requiring general anesthesia for
reasons unrelated to respiratory disease, and were enrolled
after a dog with laryngeal paralysis was admitted to match
435
dogs for breed. Young control dogs were  4 years of age,
and age-matched controls were matched as closely as possible
to age of the dogs with laryngeal paralysis. Different age
groups were included in an attempt to ensure inclusion of dogs
with normal and potentially abnormal laryngeal function.
Both control groups were free of respiratory disease based on
history, physical examination, and thoracic radiographs; how-
ever, if laryngeal dysfunction was identified in any control
dog, its diagnosis would have been changed to paralyzed as
necessary.
Each dog had laryngeal function evaluated by 3 methods:
EL, TNL, and LPO. EL was performed on each dog by a
board-certified veterinary radiologist (P.F.) blinded to group
before sedation, and the entire examination was recorded.
Each dog was then sedated with acepromazine (0.02–0.04 mg/
kg) and butorphanol (0.2–0.4 mg/kg) intravenously (IV). After
10 minutes, topical anesthetic (2% lidocaine, 0.5–1.0 mg/kg)
was administered to the left nasal passage using a male cat
urinary catheter. TNL was performed by a board-certified
surgeon (M.G.R.) using a flexible endoscope (1.9 or 2.5 mm
OD); a minimum of 30 seconds or at least 5 breathing cycles
were recorded. Topical stimulation of the arytenoids was at-
tempted in any dog that lacked arytenoid motion during ex-
amination. General anesthesia was induced with IV propofol
to effect. Jaw tone, palpebral reflex, and withdrawal upon toe
pinch were maintained as completely as possible while still
allowing oral insertion of the same flexible endoscope used for
TNL. LPO was recorded for at least 30 seconds or 5 breathing
cycles. Controls then had their planned procedure and dogs
with laryngeal paralysis had left arytenoid lateralization.
A board-certified radiologist blinded to group (J.W.) re-
viewed the recorded segments of EL, and a separate observer
(T.C.), trained to evaluate laryngeal function by a boarded
surgeon (M.G.R.) to eliminate recollection bias, reviewed
TNL and LPO studies. Both reviewers were blinded to group,
and T.C. was also blinded to method of examination (TNL
versus LPO). Evaluations from each reviewer were compared
with the final diagnosis, standard (STD), which was derived
by the surgeon who combined history, physical examination,
and live visual examinations by both TNL and LPO. Partial
or unilateral function was not considered as paralyzed, and
paradoxical motion was classified as paralyzed for this study.
Group ages and weights were compared using a Kolmogorov–
Smirnov test for normality and a 1-way ANOVA with
Tukey’s test for post hoc comparison. TNL, EL, and LPO
were compared with STD for calculation of sensitivity, spec-
ificity, negative predictive value (NPV), and positive predictive
value (PPV) for diagnosis of laryngeal function using software
(SAS V 9.1, SAS, Cary, NC).
RESULTS
Each of the 3 groups was composed of 2 Labrador
retrievers and 1 each: Great Dane, Australian shepherd,
and Rottweiler cross. Mean  SD age of dogs with la-
ryngeal paralysis was 10  1.6 years which was not sig-
nificantly different from age-matched controls (9  1.6
436 DIAGNOSIS OF LARYNGEAL PARALYSIS IN DOGS

years). Laryngeal paralysis dogs and age-matched con-
trols were significantly older than the young controls
(2  1.4 years; Po.001). Mean weight was not signifi-
cantly different between groups: laryngeal paralysis,
42  12 kg; age-matched controls, 32  10 kg; and young
controls, 34.4  10.3 kg (Table 1).
Interpretation of EL was hindered by movement in 3
dogs and the diagnosis was difficult to make in 1 dog with
laryngeal paralysis and in 1 normal dog. Three dogs with
bilateral laryngeal paralysis requiring surgery were diag-
nosed as unilaterally affected or normal on EL. Three
dogs had paradoxic motion on TNL and LPO, 2 of those
were considered normal on EL, and 1 had no motion on
EL. Paralysis was diagnosed in 1 age-matched and 3
young control dogs on EL. LPO and TNL falsely diag-
nosed lack of arytenoid movement in 2 age-matched
controls and 1 young control.
Sensitivity and specificity of EL were 0.6 and for TNL
were 1.0 and 0.7, respectively. PPV and NPV of EL were
0.43 and 0.75, respectively, whereas for TNL they were
0.62 and 1.0, respectively.
One dog developed marked pyoderma requiring top-
ical cleaning and antimicrobial therapy after EL. Mild
epistaxis that resolved with no treatment occurred in 2
dogs after TNL.
DISCUSSION
We were successful in our attempt to evaluate multiple
dogs with normal and abnormal laryngeal function with
EL, TNL, and LPO. The sensitivity and specificity of EL
were both less than that of TNL, suggesting that direct
observation of the larynx (TNL) is better than indirect
(EL). The same is true of the PPV and NPV. Although
neither EL nor TNL had a high PPV, TNL had higher
PPV and specificity than EL. The suggested prevalence of
laryngeal paralysis in dogs undergoing anesthesia was
 25% in 1 study; however, the prevalence of dogs with
substantial signs of upper airway obstruction associated
with laryngeal paralysis diagnosed upon induction of an-
esthesia was  5%.14 The true prevalence of the condi-
tion may not be known, which can affect the usefulness of
PPV and NPV. However, the need for proper evaluation
of laryngeal functional is usually reserved for dogs with
clinical signs of upper airway obstruction. The difficulty
in diagnosing laryngeal paralysis by direct observation
may be related to the effects of sedatives and anesthetics
on laryngeal function.7,8,10
Despite the higher values of sensitivity, specificity,
PPV, and NPV, clinicians must also consider the com-
plications of applying a diagnostic test. The side effects of
EL and TNL were not severe; however, there were no
complications with LPO. TNL requires sedation, which
also carries an inherent risk in every patient. The risk of
damage to expensive endoscopic equipment is high when
used on conscious animals. Although ability to evaluate
laryngeal function improved with TNL over EL, our re-
sults suggest that the risks associated with performing
endoscopy on an awake yet sedated dog outweigh the
benefits of TNL. The diagnosis and observation achieved
with LPO is equal to or superior to that with TNL and
requires only a laryngoscope.
The main problem with EL was the presence of false-
negative results in 3 dogs diagnosed with paradoxical ar-
ytenoid motion on TNL and LPO. Dogs with paradox-
ical motion or ‘‘up and down’’ motion of the arytenoids
have been diagnosed with laryngeal paralysis on EL,12

Table 1. Results of Laryngeal Evaluation Using Echolaryngography (EL), Transnasal Laryngoscopy (TNL), Laryngoscopy Per Os (LPO), and the
Standard Diagnosis (STD)
Dog Group Age (years) Breed Weight (kg) EL TNL LPO STD Other

1 Paralyzed 9 Labrador retriever 38  þ þ þ

2 Paralyzed 12 Labrador retriever 28.6  þ þ þ Paradoxic motion
3 Paralyzed 8 Great Dane 60  þ þ þ Paradoxic motion
4 Paralyzed 11 Australian shepherd 36 þ þ þ þ Paradoxic motion
5 Paralyzed 10 Rottweiler cross 47.7 þ þ þ þ
6 Age-matched 9 Labrador retriever 34.6  þ þ 
7 Age-matched 11 Labrador retriever 27.3    
8 Age-matched 7 Great Dane 47.7  þ þ 
9 Age-matched 10 Australian shepherd 21.8    
10 Age-matched 8 Rottweiler cross 28.6 þ   
11 Young 1 Labrador retriever 35.6  þ þ 
12 Young 0.8 Labrador retriever 33.5    
13 Young 4 Great Dane 51 þ   
14 Young 3 Australian shepherd 26 þ   
15 Young 1 Rottweiler cross 25.9 þ   

The results for each examination were either positive ( þ ) or negative () for laryngeal paralysis.
 RADLINSKY ET AL
but the diagnosis of paradoxical motion seems difficult to
obtain based on our results. It is possible that EL is
associated with a steep learning curve, and improvement
in the sensitivity, specificity, PPV, and NPV would in-
crease with experience. It is also possible that evaluation
of the recorded sessions, separation of the observer from
the patient, lack of real-time examination, and the op-
portunity to optimize the image adversely affected our
results. Motion hindered diagnosis in some dogs during
EL. These dogs were not in the young control group: 1
was in the age-matched controls and 1 had laryngeal pa-
ralysis. Unfortunately reexamining the dog at a later time
or providing sedation represents the only means of de-
creasing. The addition of sedation would eliminate the
primary benefit of EL, i.e., examination of an awake dog.
The method of diagnosing laryngeal paralysis was
combining clinical signs with evaluation by TNL and
LPO (the STD method) in this study and was chosen
because it was similar to the standard method of com-
bining clinical signs with LPO. Results of LPO and TNL
were not different if the 2 dogs with diminished arytenoid
motion were considered false positives on LPO. TNL and
LPO require sedation and general anesthesia, respec-
tively, which may diminish arytenoid function.7,8,10 Se-
dation was required in dogs undergoing TNL to decrease
stress as well as risk of equipment damage.13 Ace-
promazine does not produce a significant effect in horses
and cattle undergoing TNL, and its use for anxiolysis for
passage of the endoscope during TNL in this study was
based on those results and previous use for TNL.13–15
Opioids can produce profound effects on the respira-
tory system in all species.16–18 However, in people sedated
with hydromorphone, there was neither significant
change in pharyngeal resistance nor change in incidence
of sleep-disordered breathing.16 This suggests hydromor-
phone does not selectively depress upper airway muscular
function. Butorphanol was therefore selected for our
study to aid in endoscope passage during TNL, and was
used as a standard premedicant for laryngeal examina-
tion at the teaching hospital. Our clinical impression with
this combination of sedatives in LPO and during TNL
bears this out.13
Research on the effects of anxiolytics, analgesics, and
anesthetic agents on laryngeal function in the dog have
provided some guidelines for selection of drugs for la-
ryngeal examination. One study suggested that ace-
promazine enhanced the negative effects of other
opioids and induction agents on laryngeal function and
suggested that thiopental resulted in less depression of
arytenoid motion than propofol.8 We used lower doses of
acepromazine and butorphanol than previously reported
for endoscope passage for TNL.8 The protocol had been
used in the development of TNL and seemed adequate if
topical stimulation of the larynx was used to supplement
437
the examination in dogs with diminished or absent ar-
ytenoid motion.13 Doxapram may augment ventilatory
efforts and arytenoid motion during laryngeal examina-
tion.9,10 Its inclusion may have significantly decreased the
false-positive diagnoses during both TNL and LPO in
our study.
Lidocaine was used to desensitize the nasal mucosa
before insertion of the flexible endoscope. We presume
that lidocaine diminished dog stress and discomfort as-
sociated with passing the endoscope down the nasal pas-
sages, allowing sedation alone to be used for TNL. In
people, sedation with an opioid and benzodiazepine was
considered inadequate to obtund pharyngeal and laryn-
geal reflexes, laryngospasm was anticipated in response to
airway endoscopy, and topical local anesthetic was ad-
vised.19 Interestingly, topical sensitivity was subjectively
maintained, as touching the arytenoids with the endo-
scope was useful to increase arytenoid motion if it ini-
tially seemed diminished in some dogs in this and
previous research.13
In conclusion, although TNL was associated with a
higher sensitivity, specificity, PPV, and NPV over EL, the
need for expensive equipment and lack of significantly
different results from LPO do not justify its use for the
diagnosis of laryngeal paralysis in dogs. Minimal side
effects were noted with TNL and EL in all dogs. EL may
be associated with a steep learning curve, and care should
be taken to recognize paradoxical arytenoid motion in
dogs with marked clinical signs associated with laryngeal
paralysis. EL may still be useful for ruling out mass le-
sions in the pharyngeal and laryngeal regions before in-
duction of general anesthesia for the diagnosis of
laryngeal paralysis with LPO.
REFERENCES
1. MacPhail CM, Monnet E: Outcome of and postoperative
complications in dogs undergoing surgical treatment of la-
ryngeal paralysis: 140 cases (1985–1998). J Am Vet Med
Assoc 218:1949–1956, 2001
2. LaHue TR: Treatment of laryngeal paralysis in dogs by uni-
lateral cricoarytenoid laryngoplasty. J Am Anim Hosp As-
soc 25:317–324, 1989
3. Payne JT, Martin RA, Rigg DL: Abductor muscle prosthesis
for correction of laryngeal paralysis in 10 dogs and one cat.
J Am Anim Hosp Assoc 26:599–604, 1990
4. White RAS: Unilateral arytenoid lateralization; an assess-
ment of technique and long term results in 62 dogs with
laryngeal paralysis. J Small Anim Pract 30:543–549, 1989
5. Rosen E, Greenwood K: Bilateral arytenoid cartilage later-
alization for laryngeal paralysis in the dog. Am J Vet Res
180:515–518, 1982
6. Gaber CE, Amis TC, LeCouteur RA: Laryngeal paralysis in
dogs; a review of 23 cases. J Am Vet Med Assoc 186:377–
380, 1985
438 DIAGNOSIS OF LARYNGEAL PARALYSIS IN DOGS
7. Smith MM: Diagnosing laryngeal paralysis. J Am Anim
Hosp Assoc 36:383–384, 2000
8. Jackson AM, Tobias K, Long C, et al: Effects of various
anesthetic agents on laryngeal motion during laryngoscopy
in normal dogs. Vet Surg 33:102–106, 2004
9. Miller CJ, McKiernan BC, Pace J, et al: The effects of do-
xapram hydrochloride (Dopram-V) on laryngeal function
in healthy dogs. J Vet Intern Med 16:524–528, 2002
10. Tobias KM, Jackson AM, Harvey RC: Effects of doxapram
HCl on laryngeal function of normal dogs and dogs with
naturally occurring laryngeal paralysis. Vet Anesth Analg
31:258–263, 2004
11. Rudorf H, Barr F: Echolaryngography in cats. Vet Radiol
Ultrasound 43:353–357, 2002
12. Rudorf H, Barr FJ, Lane JG: The role of ultrasound in the
assessment of laryngeal paralysis in the dog. Vet Radiol
Ultrasound 42:338–343, 2001
13. Radlinsky MG, Mason DE, Hodgson D: Transnasal la-
ryngoscopy for the diagnosis of laryngeal paralysis in dogs.
J Am Anim Hosp Assoc 40:211–215, 2004
14. Broome C, Burbidge HM, Pfeiffer DU: Prevalence of laryn-
geal paresis in dogs undergoing general anesthesia. Aust
Vet J 78:769–772, 2000
15. Anderson DE, Gaughan EM, DeBowes RM, et al: Effects of
chemical restraint on the endoscopic appearance of laryn-
geal and pharyngeal anatomy and sensation in adult cattle.
Am J Vet Res 55:1196–1200, 1994
16. Robinson RW, Zwillich CS, Bixler EO, et al: Effects of oral
narcotics on sleep-disordered breathing in healthy adults.
Chest 91:197–203, 1987
17. Bolser DC, Hey HA, Chapman RW: Influence of central
antitussive drugs on the cough motor pattern. J Appl
Physiol 86:1017–1024, 1999
18. Wanke T, Lahrmann H, Formanek D, et al: The effect of
opioids on inspiratory muscle fatigue during inspiratory
resistive loading. Clin Physiol 13:349–360, 1993
19. McGuire G, Eg-Beheiry H: Complete upper airway obstruc-
tion during awake fiberoptic intubation in patients with
unstable cervical spinal fractures. Can J Anaesth 46:176–
178, 1999