Swallowing Dysfunction in Patients
Receiving Prolonged Mechanical
Ventilation*
Kenneth Tolep, MD, FCCP; Cathy Leonard Getch, MA; and
Gerard J. Criner, MD, FCCP
Several studies have suggested that swallowing dys
function and pulmonary aspiration occur in patients
receiving prolonged ventilation. However, the inci
dence of swallowing dysfunction, its rate of resolution,
and the sensitivity of tests used to characterize swal
lowing abnormalities are not well defined. The goals of
our study were to evaluate swallowing function in this
group of patients by (1) defining the specific swallow
ing abnormalities that occur in this patient population,
(2) comparing the sensitivity of bedside evaluations to
modified barium swallow with videofluoroscopy (MBS/
VF), (3) performing endoscopic evaluation of the up
per airway to characterize glottic function during
swallowing, (4) evaluating the relationship between
swallowing dysfunction and neuromuscular disorders,
and (5) studying the temporal resolution of swallowing
abnormalities. Swallowing function was evaluated in
35 patients receiving prolonged ventilation (ie, >3
weeks) admitted to a specialized rehabilitation unit
dedicated to the care of patients requiring prolonged
ventilation. The average age of the 35 patients was
61 15 years. The total duration of intubation at the
time of the initial swallowing evaluation was 29 34
days via a cuffed tracheostomy tube and 15 9 days via
an endotracheal tube. Neuromuscular disorders were
present in 16 patients (45%). Thirty-four percent of the
patients had at least one swallowing abnormality
detected by bedside examination. Results of bedside
"D ecent studies have suggested that swallowing dys-
*-*
function and pulmonary aspiration occur in pa
tients receiving ventilatory support via translaryngeal
intubation followed by a cuffed tracheostomy tube.1,2
In patients who receive prolonged ventilation (ie, >21
days), multiple factors maybe present that increase the
prevalence of swallowing dysfunction. These include
the following: glottic injury due to preceding transla
ryngeal intubation; the effect of tracheostomy on
laryngeal movement; prolonged inactivity of the skel
etal muscles of the larynx; the use of anxiolytics/
*From the Division of Pulmonary and Critical Care Medicine,
Department of Medicine, Temple University School of Medicine,
Philadelphia.in
Supported part by a grant from the Health Care Financing Ad
ministration (HCFA) 29-P-99401/3/01.
Manuscript received April 12, 1995; revision accepted June 23.
Dr. Criner, Pulmonary and Critical Care Medi
Reprint requests:
cine, Parkinson Pavilion 925, Temple University School ofMedicine,
3401 N. Broad Street, Philadelphia, PA 19140
Downloaded From: http://journal.publications.chestnet.org/ on 04/15/2016
swallowing examination were abnormal in 31% of pa
tients with a neuromuscular disorder and 37% of pa
tients without a neuromuscular disorder. MBS/VF was
abnormal in 83% of patients (85% in patients with and
80% in patients without a neuromuscular disorder).
Results of early (<1 month) repeated MB SAT exami
nations usually remained unchanged; however, in a
small group of patients, later studies (>1 month)
revealed significant improvement. In 50% of patients
who underwent direct laryngoscopy, important abnor
malities were found that contributed to swallowing
dysfunction. Our data show that patients requiring
prolonged mechanical ventilation have a high inci
dence of swallowing abnormalities, regardless of the
presence or absence of neuromuscular disorders.
MB SAT and direct laryngoscopy can provide useful
information about laryngeal action and swallowing
dysfunction, and can facilitate the implementation of
corrective actions to prevent respiratory complica
tions. (CHEST 1996; 109:167-72)
MBSATsmodified barium swallow with videofluoros
copy
Key words: long-term ventilation; pulmonary aspiration;
swallowing; tracheostomy
neuromuscular blocking agents; and an underlying
neuromuscular illness. Any or all of these factors may
contribute to the swallowing abnormalities that have
been reported previously.12 Recognition of swallowing
dysfunction in these patients may identify patients at
high risk of pulmonary aspiration, and thereby help to
avoid respiratory complications such as recurrent
pneumonia and atelectasis.
The major goal of this study was to define swallow
ing function in patients receiving prolonged ventila
tion. We attempted to do this by the following: (1)
defining the specific swallowing abnormalities that
occur in this patient population; (2) comparing the
sensitivity of bedside evaluations, made byexperienced
speech pathologists, to modified barium swallow with
videofluoroscopy (MBSAT); (3) evaluating the rela
tionship between swallowing dysfunction and neuro
muscular disorders; and (4) studying the temporal
CHEST /109 /1 / JANUARY, 1996 167
 Table 1.Redside Evaluation in Patients With Abnormal could tolerate 30 to 60 min of spontaneous unassisted ventilation.
Results of Studies (n=12) During the MBS/VF, patients swallowed three different textures of
material (thin, thick liquids, and semisolid paste) and fluoroscopic
NMD* Present NMD Absent imaging was obtained that was recorded on videotape for detailed
(n=5) (n=7) analysis. All patients were seated upright during the study. Attempts
I-1 I-1 at swallowing were made with the tracheostomy cuff inflated and
Parameters Normal Abnormal Normal Abnormal deflated while subjects breathed spontaneously without ventilatory
Lingual control 35 2 2 support. Parameters that reflected the oral and pharyngeal phases
Palatal elevation 17 0 4 of swallowing {ie, lingual propulsion, swallowing reflex, laryngeal
Swallowing reflex 40
14 elevation) were judged to be normal, mildly, moderately, or severely
abnormal. If tracheal aspiration (penetration of barium-coated
Laryngeal elevation 2 3 4 3 material below the glottis) or abnormalities that increased the risk
Laryngeal control 26 1 3
for delayed, intermittent aspiration (ie, premature spillage, vallec-
Tinted tracheal secretions 4 5
12
ular and pyriform stasis, pharyngeal coating, cricopharyngeal spasm)
*NMD=neuromuscular disorders. were present, they were graded in a similar fashion (Table 2).
Early repeated MBS/VF studies (<1 month after the initial
resolution of swallowing abnormalities. In a subgroup MBS/VF) were performed in nine patients who had abnormal re
sults of initial studies. In a smaller subset of patients (n=5), delayed
of patients, endoscopic evaluation of the upper airway repeated MBS/VF studies (>1 month after the initial MBS/VF)
was performed to better characterize glottic function were performed to evaluate the response to a program of swallow

during swallowing. ing therapy.

Methods
All patients were admitted to a tertiary care, noninvasive respi
ratory care unit that is one of four national demonstration sites for
the Health Care Financing Administration Chronic Ventilator
Demonstration Project. Patients admitted to this unit receive
aggressive whole body and respiratory reconditioning. The admis
sion criteria and care plan are outlined in prior publications.3
Thirty-five consecutive ventilator-dependent patients underwent
a bedside swallowing evaluation performed by an experienced
7 days after they were accepted into the
speech pathologist within
unit. The evaluation included a review of the patient's medical his
tory and an examination of the oral anatomy. Patients were closely
observed while they swallowed materials of different textures. Six
parameters that reflected the oral and pharyngeal phases of swal
lowing were judged to be either normal or abnormal (Table 1). Food
to be swallowed and recovery
coloring was added to the materials
of tinted trachealsecretions was considered as evidence for aspira
tion.
An MBS/VF was performed on all patients in whom results of the
bedside evaluation were abnormal (n=12) and in 11 patients who
had normal findings from the bedside evaluation but who were
because of unexplained fevers and/or
suspected of aspiratingabnormalities.
recurrent radiographic The MBS/VF was performed
as soon as the patient's clinical condition was stable and the patient
Endoscopic evaluation of thein upper airway via flexible fiberop
tic endoscopy was performed ten patients with abnormal results
of initial MBS/VF examinations to further define glottic abnormal
ities. Topical anesthesia was applied only to the nasal mucosa so that
laryngeal sensation could be evaluated.
Data Analysis
Group data are shown as the means SDs. An unpaired t test was
used to compare the frequency of abnormalities seen on bedside
evaluation and MBS/VF between the groups with and without a
neuromuscular disorder. Patients with underlying neuromuscular
disorders were examined separately because of the previously
reported high association of swallowing abnormalities. An unpaired
t test was used to compare the number of days that a tracheostomy
was in place at the time of the MBS/VF between patients with
normal and abnormal results of studies. A p value less than 0.05 was
considered to be significant.
Results
The average age of the 35 patients was 61 15 years;
21 were men.The total duration of intubation at the
time of the initial swallowing evaluation was 29 34
days via a cuffed tracheostomy tube and 15 9 days via
an endotracheal tube. Most patients required pro-

Table 2.Findings on Initial MRS/VF in Patients With Abnormal Results of Studies (n=19)
NMD Present NMD Absent
(n=ll) (n=8)
Abnormal Abnormal
Parameters Normal Mild Moderate Severe Normal Mild Moderate Severe

Lingual propulsion 5 2
Premature spillage* 3 4
Swallowing reflex 4 5
Tracheal aspiration* 7 1
Vallecular stasis* 3 2
Pyriform sinus stasis* 3 2
Pharyngeal coating* 4 2
Laryngeal elevation 5 0 0
Cricopharyngeal spasm* 7
*"Normal" indicates that this finding was absent on the study.
168 Clinical Investigations in Critical Care

Downloaded From: http://journal.publications.chestnet.org/ on 04/15/2016

 Table 3.Changes Seen on Repeated MBS/VF in Patients Without Neuromuscular Disorders (n-7)
Initial
Parameters Normal Mild
Lingual propulsion
Premature spillagef
Swallowing reflex
Tracheal aspiration'
Vallecular stasis1
Pyriform sinus stasis*
Pharyngeal coating*
Laryngeal elevation
*
Average length of time between studies=19ll days.
^"Normal" indicates that this finding was absent on the study.
severe. In the one patient who had only mild abnor
malities noted on the initial MBSAT, there was no
improvement on the repeated MBSAT. Similarly, no
significant improvement was seen in the two patients
with a neuromuscular disorder on the early repeated
MBSAT. Delayed repeated MBSAT examinations
showed dramatic improvement in two of three patients
without a neuromuscular disorder. One patient with
neurofibromatosis ofthe spinal cord had no significant
improvement seen on the delayed repeated MBSAT;
however, in a patient with Parkinson's disease, signif
icant improvement was seen on the delayed study af
ter levodopa (L-dopa) was given on a regular basis.
Laryngoscopy Findings
Abnormalities were seen in five often patients who
underwent direct laryngoscopy. These included the
following:secretions
decreased sensation ofthe vocal cords (n=2);
pooled above the vocal cords (n=2); limited
vocal cord motion (n=2); and edema of arytenoids
(n=l). In the seven patients without a neuromuscular
disorder, four had abnormalities detected. The relative
frequency of the abnormalities was similar in patients
with and without a neuromuscular disorder.
Discussion
Our data show that patients with and without neu
romuscular disorders requiring prolonged intubation
(translaryngeal followed by tracheostomy) and positive
pressure ventilation have a high incidence of swallow
ing abnormalities. Moreover, the types of swallowing
abnormalities are complex and more than one type of
abnormality is commonly present. Finally,
and direct laryngoscopy can provide useful information
MBSAT
about swallowing dysfunction that may help to prevent
subsequent respiratory complications.
Earlier studies that used dysphagia as a marker of
swallowing dysfunction suggested that swallowing ab
normalities in patients with tracheostomies were as low
as 7%.4 When speech pathologists have evaluated pa
170
Downloaded From: http://journal.publications.chestnet.org/ on 04/15/2016
Repeated Study*
Abnormal Abnormal
Moderate Severe Normal Mild Moderate Severe
tients at the bedside, a greater incidence of abnormal
ities in the oral and pharyngeal phases of swallowing
has been observed compared with the complaint of
dysphagia.and
Ancillary methods such as the tinting of
feedings13 the measurement of thefurther
centration of tracheal secretions6
glucose con
may enhance
the sensitivity of bedside evaluations. MBS/VF appears
to be even more sensitive for the detection of swal
lowingand
dysfunction, and as shown in our study, can be
safely easily employed patients requiring me
in
chanical ventilation.
The 43% incidence of aspiration detected by
MBSAT in our study (60% in the patients without a
neuromuscular disorder), is quite similar to the 45%
and 50% incidences reported in two other recent
studies.1,2 We believe that although actual aspiration is
observed in only 50% of patients receiving ventilatory
support via a cuffed tracheostomy tube, intermittent
aspiration likely occurs due to pooling ofmaterial in the
vallecula and pyriform sinuses (Fig 2), especially when
coupled with an abnormal swallowing reflex. The high
incidence of swallowing abnormalities detected by
MBSAT in patients considered to have a normal
finding from a bedside examination pointsevenoutwhen the
problem of relying ontrained
bedside evaluations
performed by highly speech pathologists. We
believe that bedside evaluation should be considered
only a screening procedure to detect gross distur
bances of swallowing function.
The addition of direct laryngoscopy to the bedside
examination allows one to more fully assess hypopha-
ryngeal sensation, detect vallecular and pyriform sinus
pooling, observe hypopharyngeal muscle contraction,
and observe the integrity of epiglottic and vocal cord
function. We have found examination of the unanes-
thetized upper airway via direct laryngoscopy to pro
vide important information while simultaneouslybeing
well tolerated by the patient. Following visual inspec
tion of the upper airway for edema and pooling of se
cretions in the valleculae or pyriform sinuses, the oral
Clinical Investigations in Critical Care
 Table 3.Changes Seen on Repeated MBS/VF in Patients Without Neuromuscular Disorders (n-7)
Initial
Parameters Normal Mild
Lingual propulsion
Premature spillagef
Swallowing reflex
Tracheal aspiration'
Vallecular stasis1
Pyriform sinus stasis*
Pharyngeal coating*
Laryngeal elevation
*
Average length of time between studies=19ll days.
^"Normal" indicates that this finding was absent on the study.
severe. In the one patient who had only mild abnor
malities noted on the initial MBSAT, there was no
improvement on the repeated MBSAT. Similarly, no
significant improvement was seen in the two patients
with a neuromuscular disorder on the early repeated
MBSAT. Delayed repeated MBSAT examinations
showed dramatic improvement in two of three patients
without a neuromuscular disorder. One patient with
neurofibromatosis ofthe spinal cord had no significant
improvement seen on the delayed repeated MBSAT;
however, in a patient with Parkinson's disease, signif
icant improvement was seen on the delayed study af
ter levodopa (L-dopa) was given on a regular basis.
Laryngoscopy Findings
Abnormalities were seen in five often patients who
underwent direct laryngoscopy. These included the
following:secretions
decreased sensation ofthe vocal cords (n=2);
pooled above the vocal cords (n=2); limited
vocal cord motion (n=2); and edema of arytenoids
(n=l). In the seven patients without a neuromuscular
disorder, four had abnormalities detected. The relative
frequency of the abnormalities was similar in patients
with and without a neuromuscular disorder.
Discussion
Our data show that patients with and without neu
romuscular disorders requiring prolonged intubation
(translaryngeal followed by tracheostomy) and positive
pressure ventilation have a high incidence of swallow
ing abnormalities. Moreover, the types of swallowing
abnormalities are complex and more than one type of
abnormality is commonly present. Finally,
and direct laryngoscopy can provide useful information
MBSAT
about swallowing dysfunction that may help to prevent
subsequent respiratory complications.
Earlier studies that used dysphagia as a marker of
swallowing dysfunction suggested that swallowing ab
normalities in patients with tracheostomies were as low
as 7%.4 When speech pathologists have evaluated pa
170
Downloaded From: http://journal.publications.chestnet.org/ on 04/15/2016
Repeated Study*
Abnormal Abnormal
Moderate Severe Normal Mild Moderate Severe
tients at the bedside, a greater incidence of abnormal
ities in the oral and pharyngeal phases of swallowing
has been observed compared with the complaint of
dysphagia.and
Ancillary methods such as the tinting of
feedings13 the measurement of thefurther
centration of tracheal secretions6
glucose con
may enhance
the sensitivity of bedside evaluations. MBS/VF appears
to be even more sensitive for the detection of swal
lowingand
dysfunction, and as shown in our study, can be
safely easily employed patients requiring me
in
chanical ventilation.
The 43% incidence of aspiration detected by
MBSAT in our study (60% in the patients without a
neuromuscular disorder), is quite similar to the 45%
and 50% incidences reported in two other recent
studies.1,2 We believe that although actual aspiration is
observed in only 50% of patients receiving ventilatory
support via a cuffed tracheostomy tube, intermittent
aspiration likely occurs due to pooling ofmaterial in the
vallecula and pyriform sinuses (Fig 2), especially when
coupled with an abnormal swallowing reflex. The high
incidence of swallowing abnormalities detected by
MBSAT in patients considered to have a normal
finding from a bedside examination pointsevenoutwhen the
problem of relying ontrained
bedside evaluations
performed by highly speech pathologists. We
believe that bedside evaluation should be considered
only a screening procedure to detect gross distur
bances of swallowing function.
The addition of direct laryngoscopy to the bedside
examination allows one to more fully assess hypopha-
ryngeal sensation, detect vallecular and pyriform sinus
pooling, observe hypopharyngeal muscle contraction,
and observe the integrity of epiglottic and vocal cord
function. We have found examination of the unanes-
thetized upper airway via direct laryngoscopy to pro
vide important information while simultaneouslybeing
well tolerated by the patient. Following visual inspec
tion of the upper airway for edema and pooling of se
cretions in the valleculae or pyriform sinuses, the oral
Clinical Investigations in Critical Care
 ^k,
Figure 2. Severe stasis shown in the vallecula (arrowhead) and
pyriform sinuses (arrow) in a patient without a neuromuscular dis
order.
administration of color-tinted water-ice allows one to
visually assess the response of the pharyngeal and la
ryngeal structures to thermal stimulation. The reduc
tions in vocal cord sensitivity and vocal cord movement,
detected with this technique, increased the risk of as
piration in patients with these deficits.
Etiology of Swallowing Dysfunction in Ventilated
Patients
Many factors could contribute to swallowing dys
function in patients requiring prolonged ventilation.
These factors include the following: acute medical or
surgical illness; medications used to treat these condi
tions; prolonged inactivity of swallowing muscles;
injury due to translaryngeal intubation; the tracheo
stomy tube itself; and underlying neuromuscular dis
orders. Stroke,7 poliomyelitis,8 and Parkinson's dis
ease9 all are associated with an increased incidence of
swallowing dysfunction. It is possible that in some of
our patients with an underlying neuromuscular dis
ease, swallowing abnormalities were primarily related
to the neurologic condition, rather than intubation,
tracheostomy, or positiveseenpressure ventilation. The
dramatic improvement in one patient who was
treated for parkinsonism illustrates the close relation
Downloaded From: http://journal.publications.chestnet.org/ on 04/15/2016
ship between swallowing function and overall neuro
logic incidence
status in some patients. However, the equally
high of swallowing abnormalities, specifi
cally aspiration, in our patients without a neuromus
cular disorder supports the notion that prolonged
translaryngeal intubation and tracheostomy predispose
to these complications.
Nearly all patients receiving prolonged ventilation
via a tracheostomy have had prior translaryngeal intu
bation. Both translaryngeal intubation and tracheo
stomy affect glottic anatomy and swallowing function.
Fiberoptic laryngoscopy has shown that anatomic ab
normalities such as ulceration of the vocal cords and
laryngeal edema are extremely common in patients
and require weeks
receiving translaryngeal intubation
to resolve.10"12 These anatomic changes maybe severe
enough to compromise the ability of the upper airway
to protect itself during the act of swallowing.11 Laryn
geal spasm has also been reported in these patients,
even in the absence of laryngeal ulcerations or ede
ma.11 Two of our patients were found to have de
creased sensation at the level of the vocal cords. This
has been described in patients and animals who have
had their upper airways bypassed. It is not clear, how
ever, whether the absence of airflow through the vo
cal cords directly results in the reduction of sensation.
A tracheostomy has also been described to limit
swallowing function by either compressing the esoph
agus decreasing elevation and anterior displacement
or
of the larynx.4 The exact incidence and magnitude of
swallowing dysfunction mediated by the physical pres
ence of a tracheostomy is unknown. However, these
negative effects can be somewhat modified by ensur
ing the proper tracheostomy tube size and avoiding
tracheostomy cuff overinflation. The persistence of
swallowing dysfunction patients after decannulation
in
and significant improvement in two patients with tra
cheostomy tubes remaining in place illustrate that
mechanical factors related to the presence of a cuffed
tracheostomy tube do not fully explain the pathogen
esis of swallowing dysfunction.
However, swallowing dysfunction in patients receiv
ing prolonged ventilation may be more related to se
vere illness resulting in skeletal muscle catabolism or
agents used to treat critically ill patients (ie, high-dose
corticosteroids, neuromuscular blocking agents, ami-
noglycosides, and general sedatives). In selected pa
tients, the requirements for prolonged ventilation via
airway cannulation may be more a marker of severe
illness rather than a causative factor in the develop
ment of swallowing dysfunction. In any case, it appears
that the requirement for prolonged ventilation via a
tracheostomy identifies a group of patients at high risk
for swallowing dysfunction.
Once identified, the resolution of swallowing dys
function in mechanically ventilated patients is uncer-
CHEST /109 / 1 / JANUARY, 1996 171
 tain. Devita and Spierer-Rundback1 suggested that
swallowing abnormalities are transient, but the time
interval between the initial and follow-up MBS/VF
studies was not clearly stated. Our study shows that
even in the absence of a neuromuscular disorder,
swallowing abnormalities do not quickly resolve. Re
peated MBSAT at intervals of several months seem to
be warranted, since some patients eventually demon
stratesignificant improvement.
Treatment ofSwallowing Dysfunction in the
Ventilated Patient
When swallowing abnormalities are detected, steps
need to be taken to prevent aspiration. The possibility
of neuromuscular disorders should be considered since
these may be reversible. Tracheostomy tube (and cuff)
size, location, and position should be closely evaluated.
Compensatorythat measures such as selecting a consis
results in the fewest abnormalities
tency of food
during MBSAT13 and emphasizing specific head pos
tures14 have been reported in the literature to prevent
reflex and swallowing se
aspiration. A coached cough These
quence may also be helpful. steps were utilized
in our patients who had mild or moderate abnormal
ities on the MBSAT and resulted in improved swal
lowing function with reinstitution of successful oral
feeding in most patients. abnormalities and
In conclusion, swallowing aspira
tion are common in patients receiving prolonged me
chanical ventilation via a cuffed tracheostomy tube
of the presence or absence of a neuromus
regardless
cular disorder. Bedside evaluations to exclude swal
lowing dysfunction are insensitive and newer methods
such as MBSAT and laryngoscopy should be used
frequentiy in these patients. The pathogenesis of
in this group has not been
swallowing dysfunction
clearly defined but is likely to be multifactorial. Routes
172
Downloaded From: http://journal.publications.chestnet.org/ on 04/15/2016
of feeding that avoid the risk of pulmonary aspiration
may be required in some patients, but most patients
can resume oral intake following identification of the
swallowing abnormality and implementation of cor
rective measures.
References
l Devita MA, Spierer-Rundback MS. Swallowing disorders in pa
tients with prolonged intubation or tracheostomy tubes. Crit Care
Med 1990; 18:1328-32
2 Elpern EH, Scott MG, Petro L, et al. Pulmonary aspiration in
mechanically ventilated patients with tracheostomies. Chest 1994;
105:563-66
3 Criner GJ, Kreimer DT, Pidlaoan L. Patient outcome following
prolonged mechanical ventilation via tracheostomy. Am Rev
Respir Dis 1993; 147:A874
4 Bonanno PC. Swallowing dysfunction after tracheostomy. Ann
Surg 1971; 174:29-33
5 Elpern EH, Jacobs ER, Bone RC. Incidence of aspiration in tra-
cheally intubated adults. Heart Lung 1987; 16:527-31
6 Kinsey GC, Murray MJ, Swensen SJ, et al. Glucose content of
tracheal aspirates: implication for the detection of tube feeding
aspiration. Crit Care Med 1994; 22:1557-62
7 Horner J, Massey MD, Riski JE, et al. Aspiration following stroke:
clinical correlates and outcome. Neurology 1988; 38:1359-62
8 Sonies BC, Dalakas MC. Dysphagia in patients with the post
polio syndrome. N Engl J Med 1991; 324:1162-7
9 Walton J, ed. The parkinsonian syndrome. In: Walton J, ed.
Brain's diseases of the nervous system. London: Oxford Univer
sity Press, 1985; 325-28
10 Whited RE. A prospective study of laryngotracheal sequelae in
long-term intubation. Laryngoscope 1984; 94:367-77
11 Colice GL, Stukel TA, Dain B. Laryngeal complications of pro
longed intubation. Chest 1989; 96:877-84
12 Colice GL. Resolution of laryngeal injury followingtranslaryngeal
intubation. Am Rev Respir Dis 1992; 145:361-64
13 Coster ST, Schwartz WF. Rheology and the swallow-safe bolus.
Dysphagia 1987; 1:113-18
14 Bucholz DW, Bosma JF, Donner MW. Adaptation, compensa
tion and decompensation of the pharyngeal swallow. Gastrointes
tinal Radio! 1985; 10:235-39
Clinical Investigations in Critical Care