Dysphagia 19:43–47 (2004)
DOI: 10.1007/s00455-003-0030-x
Evaluating Swallowing Dysfunction Using a 100-ml Water Swallowing
Test
Meng-Chun Wu, MD,1 Yeun-Chung Chang, MD,2 Tyng-Guey Wang, MD,1 and Li-Chan Lin3
1
Department of Physical Medicine and Rehabilitation, and 2Department of Medical Imaging, National Taiwan University Hospital, National
1 Taiwan University College of Medicine; and 3Institute of Clinical Nursing, National Yang-Ming University, Taipei, Taiwan
Abstract. This study used comparison with video-
fluoroscopic examination of swallowing (VFES) to
examine the validity of a 100-ml water swallowing
test (WST) in assessing swallowing dysfunction. Fif-
ty-nine consecutive outpatients (15 females, 44 males)
with clinically suspected dysphagia were enrolled in
this study. Each subject underwent a 100-ml WST
followed by VFES. Data was obtained on swallowing
speed and signs of choking (coughing and a wet-
hoarse voice). The analytical results revealed that 49
subjects had abnormal swallowing speeds (<10 ml/s)
in the 100-ml WST, and 47 of them were identified as
having dysphagia by VFES. Among the ten partici-
pants with normal swallowing speed (>10 ml/s), eight
were diagnosed with dysphagia by VFES. Notably,
14 participants choked in the 100-ml WST, 11 of
whom exhibited aspiration or penetration in VFES.
Among the 45 participants without choking in WST,
12 displayed aspiration or penetration in VFES. The
sensitivity of swallowing speed in detecting the
swallowing dysfunction was 85.5%, and the specificity
was 50%. Moreover, the sensitivity of using choking
or wet-horse voice in the 100-ml WST as the sole
factor for predicting the presence of aspiration was
47.8%, while the specificity was 91.7%. Therefore, this
study concluded that swallowing speed is a sensitive
indicator for identifying patients at risk for swal-
lowing dysfunction. Moreover, choking in the 100-ml
WST may be a potential specific indicator for fol-
lowup aspiration.
Correspondence to: Tyng-Guey Wang, M.D., Department of
Physical Medicine & Rehabilitation, National Taiwan University
Hospital, 7 Chung-Shan South Road, Taipei 100, Taiwan, ROC.
Telephone: 886-2-23123456, ext. 7588; E-mail: tgw@ha.mc.ntu.
edu.tw.
Key words: Water test — Videofluoroscopy —
Dysphagia — Aspiration — Deglutition —
Deglutition disorders.
Dysphagia, a symptom of swallowing dysfunction,
can be evoked by numerous disorders, including
both neurological and non-neurological ones. Dys-
phagia not only deprives people of the pleasure from
eating but also endangers patient health by creating
a risk of aspiration pneumonia and malnutrition [1].
Accurately assessing swallowing dysfunction, espe-
cially aspiration risk, thus is extremely important to
clinicians in providing crucial information to pre-
vent morbidity and mortality from swallowing dis-
orders.
Videofluoroscopic examination of swallowing
(VFES) has been widely recommended as a gold
standard in identifying swallowing dysfunction [2].
However, VFES has radiation exposure and is rela-
tively costly and time consuming; thus, its use is not
practical for every patient suspected of having dys-
phagia. Consequently, designing effective screening
methods to identify patients for VFES referral is the
most clinically feasible approach.
Several methods have been developed for
identifying subjects suitable for VFES referral. One
method with significant potential is observing water
swallowing. Ninety- to 150-ml water swallowing test
(WST) have been used to evaluate swallowing dis-
orders but the efficiency and accuracy of these
methods remains uncertain. For example, DePippo
et al. [2] reported that choking on a 3-oz. WST was
a sensitive indicator for identifying patients requir-
ing referral for VFES. Moreover, Horner and
Massey [3] reported that aspirating subjects differed
44
from nonaspirating subjects in having a low preva-
lence of subjective complaints of swallowing diffi-
culties and a higher prevalence of weak cough reflex
and dysphonia. Linden and Siebens [4] noted that
wet-hoarse voice quality and impaired pharyngeal
gag reflex were correlated with the presence of as-
piration on VFES examination in nine of a total
sample of 11 patients. Scales et al. [5] identified
seven clinical signs displayed by aspirating patients
and applied these clinical parameters to identify up
to 42% of subjects with aspiration on VFES ex-
amination.
However, bedside swallowing tests have long
been criticized for lack of objectiveness in identifying
aspirating patients. Nathadwarawala et al. [6] first
used objective swallowing speed to assess swallowing
function and found that swallowing speed was sig-
nificantly reduced in subjects with swallowing prob-
lems. They further proposed that reduced swallowing
speed in chronic neurological diseases is one of the
compensatory or adaptive mechanisms used by many
patients before an overt clinical problem develops.
However, none of the above investigations used
VFES to present further evidence of swallowing dys-
function.
To address these issues, this study designed a
100-ml WST. The primary purpose of this study was
to determine the role of a 100-ml WST in evaluating
swallowing dysfunction. By observing both
swallowing speed and signs of choking, as well as
correlating these parameters with findings on VFES,
this work attempted to estimate the sensitivity and
specificity of a 100-ml WST in detecting swallowing
dysfunction.
Methods
Participants
Fifty-nine individuals, consecutively referred to videofluoroscopic
examination for clinical suspected swallowing disorders, were
recruited in this study. All participants preserved the ability to
eat orally at the time of referral and were fully cooperative. The
mean age of the sample group was 71.7 ± 11.4 years (range-43–
97 years); the group consisted of 44 males and 15 females.
Stroke was the main reason for dysphagia in the subjects;
51 subjects had a previous history of stroke, while the remaining
8 individuals had no definite neurological disorders. Nineteen
subjects had right-hemisphere strokes, 18 patients had left-hemi-
sphere strokes, 10 patients had bilateral hemisphere strokes, and
4 had brain stem and other strokes. All stroke patients had
suffered their attacks at least six months previously and had
stable neurological status on referral. The study procedures were
fully explained, and consent forms were obtained from all
participants.
M.-C. Wu et al.: Water Test in Swallow
100-ml Water Swallowing Test
Twenty-four hours before VFES, a 100-ml WST was performed for
each participant as follows. First, the participants were seated
upright and asked to place a glass of 100 ml of distilled water to
their lips. On receiving a ‘‘go’’ signal, they drank the water in the
glass as quickly as possible. Hand dexterity of the subjects was
confirmed to be acceptable by neurological examination. A stop-
watch with a readability of 1 ms was used to measure the swal-
lowing time, measured from the ‘‘go’’ signal to the end of WST.
Notably, for those who successfully finished the glass of the water,
the end of WST was defined as the completion of the last swallow,
indicated visually by the return of the thyroid cartilage to its resting
position. Signs of choking, defined as coughing from the beginning
of WST to 1 min after the end WST, or a wet-hoarse voice after
testing, were recorded. Participants who choked during swallowing
were asked to stop drinking immediately regardless of whether they
had finished the water. In such cases, the stopwatch was stopped as
soon as choking occurred. The amount of water drunk was de-
termined by subtracting the residual water from 100 ml, using a
minimum scale of 1 ml. Swallowing speed (ml/s), defined as the
amount of drunken water divided by elapsed time on the stop-
watch, was also calculated. Notably, an abnormal swallowing
speed was defined as swallowing speed below 10 ml/s in accordance
with the definition used in the literature [7].
Videofluoroscopic Examination of Swallowing
Standardized VFES was conducted at our institution using a
fluoroscope with a remote control (KOX-850, Toshiba Corp.,
Tokyo, Japan; RSZ-2000, Shimadzu Corp., Kyoto, Japan) and a
high-resolution Super-VHS recorder (BR 1200, JVC, Japan). The
videotape recorder offered a frame rate of 33 frames per second and
could display real-time dynamic images and frame-by-frame static
images. Participants were instructed to sit on a specially designed
chair (VESS chair, Vess Chairs Inc., Milwaukee, WI, USA) for
both lateral and frontal anterior–posterior views. Each participant
swallowed three standardized formulas (5 ml each of thin, thick,
and paste medium) of barium sulfate (E-Z-HD, E-Z-EM, Inc.,
Westbury, NY, USA) sequentially [8,9]. Notably, thin barium
sulfate (suspension of 340 g E-Z-HD and 65 ml of water) is a
standard for routine gastrointestinal examinations. Thick barium
was prepared by adding an extra 7.5 ml of E-Z-HD powder in a 15-
ml standard thin barium preparation, while paste barium was
prepared with adding extra 12 ml of E-Z-HD powder in a 15-ml
standard thin barium preparation. By viewing the VEFS video-
tapes, an experienced radiologist timed the oral transit time, pha-
ryngeal transit time, and trigger time for each formula. Oral transit
time was defined as the time elapsed from placing the formula at
the anterior dorsal tongue to the formula completely leaving the
valleculae. Pharyngeal transit time was determined by the time
from the formula leaving the valleculae to completely leaving the
pharynx. Swallowing trigger time was defined as the interval be-
tween the formula head arriving at the valleculae and the con-
traction of the upper pharyngeal constrictor muscle. The presence
of oral stasis, pharyngeal stasis, pyriform stasis, vallecular stasis,
and penetration as well as aspiration was also recorded. Swallow-
ing dysfunction was defined as the presence of at least one of the
following five determinants in VFES: (1) oral transit time exceeding
2 s, (2) pharyngeal transit time exceeding 1 s, (3) trigger time ex-
ceeding 1 s, (4) laryngeal penetration, (5) aspiration, and (6) stasis
at more than two sites of valleculae and pyriform sinuses.
M.-C. Wu et al.: Water Test in Swallow
Table 1. Videofluoroscopic characteristics of patients
Findings on VFES Number of patients
Normal 2
Oral transit time more than 2 s 15
Pharyngeal transit time more than 1 s 21
Trigger time more than 1 s 17
Occurrence of laryngeal penetration 12
Aspiration 11
Stasis more than 2 sites stated above 25
Table 2. Relationship between water swallowing speed and VFES
VFES
Swallowing speed Abnormal Normal
Abnormal (<10 ml/s) 47 2 Yes
Normal (‡10 ml/s) 8 2
Aspiration was defined as the penetration of food or liquid below
the vocal folds. Laryngeal penetration was defined as the entry of
food or liquid into the laryngeal vestibule and above the vocal
folds. Stasis was defined as the barium accumulation.
Statistics
The data were analyzed by comparing the results of the 100-ml
WST with those of the VFES using the chi-squared statistic
method. The sensitivity and specificity of the 100-ml WST as an
indicator of choking or penetration on VFES were then deter-
mined.
Results
Of the 59 subjects, VFES identified 55 as having
swallowing dysfunction by VFES. Specifically, 23
subjects were denoted as having aspiration or pene-
tration in VFES, while only two of them had con-
comitant coughing response. Table 1 summarizes the
VFES results.
Table 2 presents the relationship between ab-
normal swallowing speed and swallowing dysfunc-
tion. Forty-nine subjects had abnormal swallowing
speeds in the 100-ml WST. Among these, 47 subjects
simultaneously exhibited swallowing dysfunction in
VFES. In addition, 8 of the remaining 10 subjects
with normal swallowing speed were identified as
having swallowing dysfunction by VFES. Accord-
ingly, the sensitivity and specificity of detecting
swallowing dysfunction by employing swallowing
speed in the 100-ml WST was 85.5% and 50%, re-
spectively.
45
Table 3. Relationship between choking on 100-ml WST and as-
piration or penetration on VFES
Aspiration or penetration on VFES
Choking on 100-ml WST Yes No
Yes 11 3
No 12 33
Table 4. Relationship between choking on 100-ml WST and as-
piration on VFES
Aspiration on VFES
Choking on 100-ml WST Yes No
4 10
No 7 33
Table 3 lists the correlation among choking
and both aspiration and penetration on VFES.
Choking was noted in 14 participants during the 100-
ml WST. Notably, 11 of these subjects had concomi-
tant penetration or aspiration in VFES. In contrast,
among the 45 subjects who did not choke during the
100-ml WST, only 12 had coinstantaneous aspiration
or penetration in VFES. The VFES findings for these
12 participants included 5 cases of silent aspiration, 5
cases of penetration alone, and 2 cases of obvious
coughing during aspiration. Consequently, the sensi-
tivity and specificity of using the presence of choking
during the 100-ml WST to predict penetration or as-
piration in VFES was 47.8% and 91.7%, respectively.
Since aspiration in VFES is more clinically
serious than penetration, this study further correlated
aspiration with choking during the 100-ml WST. As
listed in Table 4, among the 14 participants who
choked during the 100-ml WST, 4 were found to as-
pirate on VFES. In comparison, 7 of the 45 partici-
pants who did not choke during the 100-ml WST
aspirated during VFES. Moreover, 5 of the 7 par-
ticipants displayed silent aspiration and 2 exhibited
coughing during aspiration. Consequently, the sen-
sitivity and specificity of using choking during the
100-ml WST to detect aspiration in VFES was 36.4%
and 20.8%, respectively.
Discussion
The study successfully assessed the significance of
detecting swallowing dysfunction using the 100-ml
46
WST. Correlated with abnormal findings presented in
VFES, swallowing speed and signs of choking in
WST had sensitivity and specificity of up to 85.5%
and 91.7%, respectively. Thus this study concludes
that for dysphagic individuals capable of eating
orally, estimating swallowing speed for 100 ml of
water provides an effective tool for screening for
VFES referral, while measuring choking in the 100-
ml WST can be used to monitor the progress of
swallowing dysfunction.
Although VFES has been widely accepted as
an accurate method of assessing swallowing dys-
function, it is not practical to utilize it for all subjects
suspected of swallowing dysfunction. Hence, numer-
ous clinical bedside approaches, such as neurological
examination [10,11], WST [2,12], swallowing provo-
cation test [13], and cough reflex [14], have been used
to identify or predict swallowing dysfunction. Cor-
related with VFES, the sensitivities of these ap-
proaches for predicting swallowing dysfunction
varied from 42% to 80%. This work employed the
WST because of its simplicity, high efficiency, and
objectivity in assessing swallowing dysfunction. Spe-
cifically, swallowing speed and evidence of choking
were selected as indicators for VFES referral.
Swallowing speed has been demonstrated to
slow markedly in individuals that display abnormal
symptoms or signs during swallowing [6]. Further-
more, a swallowing speed of below 10 ml/s was pro-
posed as the cutoff point for defining swallowing
dysfunction [7]. This phenomenon probably results
from the compensated or adapted mechanics for
disordered swallowing [6]. Additionally, individuals
with swallowing dysfunction may reduce the size of
the swallowed bolus to reduce the risk of aspiration,
thus slowing their swallowing speed [15]. This study
successfully validated the accuracy of swallowing
speed for detecting swallowing dysfunction, revealing
a sensitivity of 85.5%, which exceeded that reported
in the previous literature [2].
In our study, selection bias primarily ac-
counted for the relatively low specificity by using
swallowing speed to determine swallowing dysfunc-
tion. Since all study participants had perceived
symptoms or signs of swallowing disorder, a lower
specificity was inevitable, and it is acceptable that only
four of the subjects displayed normal VFES results.
Besides neurological impairment, factors re-
lated to gender and age may also contribute to altered
swallowing speed. For instance, the study defined a
swallowing speed below 10 ml/s as abnormal, fol-
lowing the definition used by Nathadwarawala et al.
[6,7], whose study included subjects below 70 years
old. However, the mean age of the subjects in our
M.-C. Wu et al.: Water Test in Swallow
study was 71.7 years old (range-43–97 years). Since
aging is normally associated with decreasing swal-
lowing speed [16], using a swallowing speed of 10 ml/s
as a reference is likely to underestimate the swal-
lowing abilities of our participants, biasing the find-
ings. Furthermore, swallowing speed is normally
slower in females than in males [11]. In the present
study, females comprised 25.4% of the samples, and
their swallowing abilities were probably also under-
estimated by using the same speed of 10 ml/s as a
reference. Although the statistical results in the pre-
sent study were not corrected for age and gender,
they were consistent with the reports in the literature
[2,6,7,12]. Future works will focus more carefully on
the above issues.
Another central feature of this study was to
determine the sensitivity of using choking, including
coughing and a wet-hoarse voice, to detect penetration
and aspiration. The sensitivity found by this study was
47.8%, relatively low compared with other studies
[2,5,11,17–19]. This finding thus implies that choking is
a poor indicator of aspiration and penetration and is
probably attributed partly to the presence of silent
aspiration in choking-free cases. Similarly, in contrast
with the statement of DePippo et al. [2], recent studies
also recommend against using choking as an indicator
of aspiration or as an alternative to VFES [12,13].
Moreover, the sensitivity and specificity of applying
choking to predict aspiration in VFES was further
decreased in this study, implying that choking is related
not only to aspiration but also to penetration.
Although less reliable in predicting the occur-
rence of aspiration because of its low sensitivity,
choking in the 100-ml WST had a specificity of up to
91.7% for predicting aspiration. Accordingly, choking
in the 100-ml WST is an excellent monitor for follow-
ing up individuals who were already identified as
having aspiration. For these cases, frequent estimates
of the progress of aspiration are essential to improved
treatment strategy. Despite gold standard status in
assessing aspiration, repeated VFES within a short
interval is not acceptable because of the danger of ex-
cessive radiation exposure. Instead, choking in the 100-
ml WST can be used as an alternative approach to
followup on the aspiration status. This study recom-
mends applying choking in the 100-ml WST to clinical
followup rather than initial screening of aspiration.
In conclusion, this study explicitly measured
the validity of using a 100-ml WST for screening
VFES referral. Based on its high sensitivity and
simplicity, estimating swallowing speed in the 100-ml
WST appears to a useful bedside tool for the early
detection of swallowing dysfunction. Furthermore,
the presence of choking in the 100-ml WST may be a
 M.-C. Wu et al.: Water Test in Swallow
feasible indicator for clinically monitoring the pro-
gression of aspiration.
References
1. Noll SF, Bender CE, Nelson MC: Rehabilitation of patients
with swallowing disorders. In: Braddom RL (ed.): Physical
Medicine and Rehabilitation, 1st ed. Philadelphia: W.B.
Saunders, 1996, pp 533–554
2. DePippo K, Holas M, Reding M: Validation of the 3-Oz
WST for aspiration following stroke. Arch Neurol 49:1259–
1611, 1992
3. Horner J, Massey EW: Silent aspiration following stroke.
Neurology 38:317–319, 1998
4. Linden P, Siebens A: Dysphagia: predicting laryngeal pene-
tration. Arch Phys Med Rehabil 64:281–283, 1983
5. Scales K, Patterson C, Linden P: Correlation between clini-
cal and motion fluoroscopy observation of swallowing.
American Speech-Language-Hearing Association Annual
2 Meeting. Boston, November 21, 1998
6. Nathadwarawala KM, McGroary A, Wiles CM: Swallowing
in neurological outpatients: use of a timed test. Dysphagia
9:120–129, 1994
7. Nathadwarawala KM, Nicklin J, Wiles CM: A timed test of
swallowing capacity for neurological patients. J Neurol
Neurosurg Psychiatry 55:822–825, 1992
8. Logemann JA: Treatment for aspiration related to dyspha-
gia: an overview. Dysphagia 1:34–38, 1986
9. Palmer JB, Kuhlemeier KV, Tippett DC: A protocol for the
videofluorographic swallowing study. Dysphagia 8:209, 1993
47
10. Horner J, Massey EW, Riski JE, Lathrop DL, Chase KNL:
Aspiration following stroke: clinical correlates and outcome.
Neurology 38:1359–1362, 1998
11. Lin YN, Wang TG, Chang YC, Hsieh FM, Lien IN: Vali-
dation of clinical swallowing evaluation in stroke patients.
J Rehab Med Assoc ROC 26:175–180, 1998
12. Garon BR, Engle M, Ormiston C: Reliability of the 3-Oz
WST utilizing cough reflex as sole indicator of aspiration. J
Neurol Rehabil 9:139–143, 1995
13. Teramoto S, Fukuchi Y: Detection of aspiration and swal-
lowing disorder in older stroke patients: simple swallowing
provocation test versus WST. Arch Phys Med Rehabil
81:1517–1519, 2000
14. Addington WR, Stephens RE, Gilliland K, Rodriguez M:
Assessing the laryngeal cough reflex and the risk of devel-
oping pneumonia after stroke. Arch Phys Med Rehabil
80:150–154, 1999
15. Buchholz DW, Bosma JF, Donner MW: Adaptation, com-
pensation, and decompensation of the pharyngeal swallow.
Gastrointest Radiol 10:235–239, 1985
16. Hughes TAT, Wiles CM: Clinical measurement of swallow-
ing in health and in neurogenic dysphagia. Q J Med 89:109–
116, 1996
17. Linden P, Kuhlemeier KV, Patterson C: The probability of
correctly predicting subglottic penetration from clinical ob-
servation. Dysphagia 8:170–179, 1993
18. Logemann JA: Evaluation and treatment of swallowing dis-
orders. San Diego, CA: College Hill Press, 1983
19. Splaingard ML, Hutchins B, Sultan L, Chanudhuri G: As-
piration in rehabilitation patients: videofluoroscopy vs.
bedside clinical assessment. Arch Phys Med Rehabil 69:637–
640, 1988