European Journal of Disorders of Communication, 28,213-226 (1993) 213

0 The College of Speech and Language Therapists, London

No place for motor speech acts in the

assessment of dysphagia?
Intelligibility and swallowing difficulties in stroke
and Parkinson’s disease patients

G. Kennedy
Speech Therapy Department, Whittington Hospital, London, UK.

T. Pring and R. Fawcus

Department of Clinical Communication Studies, The Ciiy University, London, UK

ABSTRACTS
Speech and language therapists are increasingly being asked to treat dysphagic patients. Concern
has been expressed and surveys have confirmed that radiological assessment procedures are rarely
available. Consequently, patients must often be assessed and their treatment planned on the basis
of bedside examinations. Despite evidence that swallowing disorders need not be related to problems
of articulation, recommendations on the procedure of such examinations frequently include an
evaluation of speech and non-speech articulatory movements. A study is reported of patients who
exhibit both dysphagia and dysarthria as a result of either stroke or Parkinson’s disease.
Assessments of the intelligibility of their speech and of their swallowing problems were found to
be unrelated. The implications of this finding for the assessment of dysphagia are discussed.
I1 est de plus en plus frkquent qu’on demande aux orthophonistes de traiter des sujets souffrant
de dysphagie. Ceci n’a pas manquk de soulever des inquidtudes, et les enquttes ont prouvd que
le diagnostic radiologique est rarement disponible. Le rdsultat est que c’est sur la base d’examens
pratiquks au pied du lit qu’il faut faire les diagnostics et ktablir le cours des traitements. Bien qu’il
se confirme que les troubles de la dkglutition ne sont pas ndcessairement liks aux problbmes
d‘articulation, les recommandations sur la marche a suivre lors de ces examens comprennent
souvent une kvaluation des mouvements articulatoires, qu’ils soient linguistiques ou non. L’on
rend compte ici d’une Ctude de sujets souffrant a la fois de dysphagie et de dysarthrie, par suite
soit d’une attaque soit de la maladie de Parkinson. L’examen de l‘intelligibilitk de leur langage et
de leurs difficultks de dkglutition montre que ces deux phknombnes sont sans rapport. Suit une
discussion des conskquences de ce rksultat du point de vue de l‘kvaluation de la dysphagie.
Sprachtherapeuten werden zunehmend gebeten, dysphagische Patienten z u behandeln. Die Sorge
kommt haufig zum Ausdruck, und es wird haufig in Berichten bestiitigt, dab radiologische
Untersuchungsverfahren selten zur Ve$igung stehen. Folglich miissen oft aufgrund von Kranken-
zimmeruntersuchungen Patienten eingestufr und ihre Behandlun festeglegt werden. Trotz der
Hinweise, daJ3 Schluckbeschwerden nicht notwendigenveise mit Arti&dationsproblemen einhergehen,
enthalten die Empfehlungen f i r solche Untersuchungsverfahren hiiufig eine Bewertun von sowohl
sprachlichen als auch nichtsprachlichen Artikulationsbewegungen. Ergebnisse einer 8ntersuchun
an Schlaganfall- und Parkinsonpatienten werden berichtet, die sowohl dysphagische als auca
dysarthrische Erscheinungen aufweisen. Die Einstufun en ihrer Sprachverstandlichkeit und ihrer
Schluckprobleme standen in keiner Relation zueinanA r . Die Implikationen dieses Ergebnisses
fiir die Dysphagiediagnose werden erortert.
214 KENNEDY, PRlNG AND FAWCUS

Key words: Parkinson’s disease, stroke, dysphagia, dysarthria, motor speech acts.

INTRODUCTION
As Edelman (1988) pointed out, speech and language therapists have been
involved in the assessment and remediation of swallowing disorders for some
time. Nevertheless, she identified a recent marked increase in referrals,
primarily of adult patients in hospital settings. The implications of this
‘dysphagia boom’ have been widely discussed. Involvement of speech and
language therapists in the area raises a number of ethical and professional
issues as well as placing a further burden on speech and language therapy
provision. In a survey (College of Speech Therapists, 1990a) of speech and
language therapy managers, 90% of those responding expressed concern about
aspects of the assessment, treatment and management of dysphagic patients,
and more than 50% volunteered that no formal policy on dysphagia manage-
ment existed within their district. Nevertheless, few respondents favoured
discontinuing this service.
Research on the assessment of dysphagia has revealed both the extent to
which it is conducted without radiological assistance and the potential hazards
of this approach.
Edelman (1988) found that, of 71 district health authorities surveyed, only
12 had routine access to radiography (usually videofluoroscopy) and held
regular clinics with an established multidisciplinary team. Respondents to the
later survey (College of Speech Therapists, 1990a) saw videofluoroscopy as
highly desirable but confirmed that in many districts it was either unavailable
or not easily accessible. Consequently, many speech and language therapists
must rely on bedside examinations and frequently face the difficult decision
between declining treatment or providing it with what they may consider to
be insufficient information. The problem is further underlined by the College
of Speech Therapists’ (1990b) suggestion that where radiography is unavailable
therapists should ‘consider whether they have sufficient information to plan
intervention’.
Experimental studies indicate that bedside examinations are unreliable at
detecting aspiration. Logemann’s (1983) study in which 40% of aspirating
patients were not detected is well known. Even this figure may be an
underestimate. Splaingard, Hutchins, Sulton and Chaudhuri (1988) found that
58% of aspirating patients were not detected. Not surprisingly these studies
and others (Linden & Siebens, 1983; Homer & Massey, 1988) advocate that
bedside evaluations should be combined with videofluoroscopy.
In evaluating these studies of the reliability of bedside examinations, it is
important to consider whether the experimental test presented to clinicians
was a fair representation of clinical practice. In several respects it appears to
be so. Thus, although advice to clinicians in Britain (College of Speech
Therapists, 1990b) repeatedly stresses the need for a multidisciplinary approach
to the management of dysphagia, it remains the case that speech and language
therapists are frequently asked to make recommendations about feeding after
MOTOR SPEECH ACTS IN ASSESSMENT OF DYSPHAGIA 215

a single bedside assessment. Moreover, the difficulties experienced by clinicians

in Splaingard et al. may be seen in the differing detection rates for patients
with (86%) and without (21%) feeding appliances (gastrostomy and nasogastric
tubes) or tracheostomy tubes where suction had been required. The presence
of such appliances was a confounding factor in the study because the more
severely at-risk patients may be presumed to have been more likely to be
using them. The higher detection rate for these patients suggests that clinicians
took their presence as an indication that the patient would aspirate; in contrast,
real clinical decisions may often be made prior to use of an appliance.
The seriousness of these findings and the potential threat they present to
patients cannot be overestimated. Videofluoroscopy is not widely available in
Britain; moreover, when it is there will be patients for whom it is inappropriate.
In view of its poor reliability the need for continued reliance on bedside
assessments will concern therapists. In what is a difficult professional dilemma,
one view may command majority support. If beside assessments are to continue
as a basis for clinical decision-making, every attention should be given to
improving them. In this respect, previous studies have been deficient in giving
little detail of the procedures used. For example, Splaingard et al. (1988)
required judgements to be made on a review of the case history and a ‘single
bedside examination’ but no further details were given. Without these it is
difficult to assess either the adequacy of the examination or the scope for
improving it.
Current assessments usually consist of checklists of behaviours. More
functional procedures have begun to appear (Cherney , Cantieri & Pannell,
1986; Warms, Champion & Mortensen, 1990), however. Existing assessments
also rely on visual observation alone to detect the potential for aspiration.
The inclusion of tactile monitoring represents a possible improvement in
technique. Of procedures known to the authors only one, the Paramatta
Hospital’s Assessment of Dysphagia (Warms et al., 1990), has suggested, but
not stipulated, the use of tactile monitoring. Although initial steps in assessing
reliability have been taken (Warms et al., 1990), further research and refine-
ment of the available procedures will be needed. The current context of
swallowing therapy in Britain suggests that a large-scale evaluation of bedside
examinations is required. This would seek to evaluate those techniques that
are in common usage and include suggestions (Linden & Siebens, 1983; Horner
& Massey, 1988; Splaingard et al., 1988) of other clinical signs which may to
a degree predict the likelihood of aspiration. Comparison between such a
bedside examination and radiography would be desirable; nevertheless, it is
accepted that no bedside assessment will be fully accurate in predicting
aspiration. As radiography may not be available to clinicians or appropriate
for the patient, we suggest that outcome measures that reflect the influence
of the assessment on the longer-term management of the dysphagic patient
also be included.
This paper is a preliminary for such an investigation. Current bedside
examinations often include the evaluation of speech and non-speech articulatory
movements. Thus, Logemann’s (1983) assessment progresses from isolated
articulatory movements to rapid alternating movements, to the determination
of diadochokinetic rate by rapid repetition of a syllable and, finally, to
216 KENNEDY, PRlNG AND FAWCUS

sentences. Price, Jones, Charleton and Allen (1987) included a full assessment
of motor and sensory function of the oral cavity. Miller (1984) also supported
their inclusion, stating that ‘speech is an extremely complex, overlearned
behaviour and, as such, serves as a barometer from which the examiner can
assess the status of the neuromuscular system that also serves swallowing’.
The inclusion of such activities reflect the longstanding view that speech
motor functions are phylogenetic and ontogenetic differentiations of more
primitive functions such as sucking, chewing and swallowing. This view is
similarly reflected in those assessments of dysarthria, such as the Frenchay
Assessment (Enderby , 1983), which include swallowing prior to the evaluation
of the oral, pharyngeal and laryngeal structures in speech.
Communication problems frequently occur in dysphagia. Martin and Corlew’s
(1990) retrospective study found that of 79 patients with swallowing problems
confirmed by videofluoroscopy, 90% had a communication disorder. Although
the co-occurrence of dysphagia and communication problems presents plausible
evidence for common impairments of speech and swallowing, it is derived
from associations between the two. Such evidence may always be severely
embarrassed by even a few cases where dissociations occur. This evidence is
available. In patients with cerebrovascular accidents (CVAs), swallowing or
speech may improve independently of each other (Netsell, 1986). In Parkinson’s
disease severe dysarthria may exist with minimal or no dysphagia and the
reverse (Sarno, 1968; Duvoisin, 1982).
These observations are consistent with Netsell’s (1986) suggestion that there
are specialised and differentiated neurons for speech and swallowing acts. The
neurophysiological evidence for this separation of function has been reviewed
by Larson (1985). Thus, though both behaviours share the same anatomical
structures, the actions of each are controlled by different command centres.
At the clinical level this suggests that assessing one dysfunction in patients
with neurological swallowing and communication difficulties may tell us little
of the other. In the present study the relationship between assessments of
dysarthric speech and dysphagia were compared in patients with both disabilities.
Lack of a relationship would be consistent with functional independence and
would have implications for the future assessment of swallowing difficulties.
Two groups of patients who experience both swallowing and communication
difficulties were sampled. Patients may be dysphagic following CVAs when
the lesions affect the brain stem or more anterior regions of the cortex.
Logemann (1983) states that the major features of the dysphagia following
CVA are delay in, or absence of, initiation of the pharyngeal phase of the
swallow pattern. Unilateral or bilateral weakness of the oral, pharyngeal and
laryngeal structures affects manipulation and transfer of material, and crico-
pharyngeal sphincter dysfunction may occasionally be present. The location
of lesions makes it likely that dysarthria will also occur in these patients, and
this was the case here. All CVA patients in this study had persistent dysarthria
but no language or cognitive impairment.
The second group had Parkinson’s disease. Speech disorders are common
in Parkinson’s disease (Logemann, Fisher, Boshes & Blonsky, 1978) and
studies have shown the value of speech therapy in improving intelligibility
(Scott & Caird, 1983; Robertson & Thomson, 1984; Johnson & Pring, 1990).
MOTOR SPEECH ACTS IN ASSESSMENT OF DYSPHAGIA 217

Dysarthria and dysphagia in Parkinson’s disease are often inconsistent, their

severity not necessarily being related to that of the disease itself. Dysphagia
may be the presenting feature (Croxon & Pye, 1988) and was found in 80%
of patients investigated radiologically (Logemann, Blonsky & Boshes, 1978).
Nevertheless, in the latter study only 20% complained of swallowing problems
suggesting that dysphagia was insufficiently severe to interfere with function
for many patients. Patients in the present study had both dysarthria and
dysphagia in varying degrees of severity.

METHOD

Subjects
Eighteen patients were tested; nine presented with a CVA which resulted in
dysphagia and dysarthria but no language impairment. Medical records indicated
that each had a previous CVA but no swallowing problems were experienced
prior to the latest episode. There were four females and five males; the mean
age was 71.9 (standard deviation 9.9). One case was taken from the first author’s
caseload and eight were referred by other therapists as fulfilling the above
requirements. All were alert and able to cooperate with the tasks presented
and were receiving speech and language therapy for their communication and
swallowing problems. At the time of testing all were on oral feeding regimens
although one was about to transfer to parenteral (nasogastric) feeding.
The remaining nine patients had Parkinson’s disease. Eight had idiopathic
and one a postencephalitic Parkinson’s disease. Classified by Hoehn and
Yahr’s (1967) rating scale, two were grade 1, one grade 3, four grade 4 and
two grade 5. One was female and eight male; their mean age was 63.4 years
(standard deviation 10.1). Three patients lived independently, three in super-
vised residential accommodation and three in long-term care. Five were
currently receiving speech and language therapy; the others were volunteers
from the local Parkinson’s Society Group and had not seen a therapist prior
to this assessment though all experienced communication and swallowing
problems. All were on oral feeding regimens at the time of the assessment.

Design
The experiment required a measure of the intelligibility of patients’ speech
and a measure of their swallowing difficulties.
The former was obtained by use of Yorkston and Beukelman’s (1984)
Assessment of the Intelligibility of Dysarthric Speech. This contains large
numbers of words and sentences from which random selections are made.
Patients’ attempts to say these are recorded and their intelligibility assessed
by having judges transcribe the tapes. Pilot testing with similar patients
suggested that fatigue during the assessment might affect intelligibility. Con-
sequently, its duration was reduced; patients said only 25 single words and
11 sentences of varying length - half the normal requirement. Four such sets
of items were generated and each patient assigned to one set. In this way
judges could transcribe more than one patient’s speech without becoming
familiar with the items.
218 KENNEDY, PRlNG AND FAWCUS

The study required a consistent assessment for the evaluation of patients’

swallowing problems. The Rehabilitation Institute of Chicago’s (RIC) Clinical
Evaluation of Dysphagia (Cherney et al., 1986) was used as the basis of this
evaluation procedure: it is a functional assessment of dysphagia which includes
history taking, evaluation of pre-feeding skills and observation of swallowing
at the oral, velopharyngeal and pharyngeal phases with four consistencies -
thin and thick liquids, pureed food and solids. No assessments of speech-
related activities are included. Although it is not widely used in Britain and
(in common with other assessments for dysphagia) has not been standardised,
it has the advantages of being comprehensive and including in a structured
format many techniques that are used in bedside examinations.
The assessment does not produce a numerical score. Rather, the therapist is
directed to observe the responses of each anatomical structure at relevant stages
in the swallow process and for the different consistencies of intake. Diadochokine-
sis and speech acts are not included in the assessment suggesting that the authors
do not consider speech to be an overlaid function of vegetative acts.
The overall profile from the assessment is appropriate for clinical use;
however, the current research required both the identification of specific points
of breakdown during the eating and swallowing process, and a scoring system
by which they could be evaluated. The former was achieved by prior use of
the assessment with patients not included in the study. Profiles from the
assessments of these patients were examined to identify common features in
the breakdown of swallowing. These were the following:
1. Maintaining lip closure throughout the oral phase.
2. Lateral movements of the tongue during mastication.
3. Posterior propulsion of the bolus by the tongue.
4. Triggering of the swallow for liquid intake.
5. Triggering of swallow for semi-solidsolid intake.
6. Triggering of cough before, during or after liquid intake.
7. Triggering of cough before, during or after semi-solidsolid intake.
8. Voice quality following liquid intake.
9. Reflux of material with oral or nasal emission.
10. Complaints of food sticking in the throat or upper oesophagus.
These features largely coincided with the points of breakdown reported in
the literature (Logemann, 1983; Groher, 1984). In evaluation of triggering of
the swallow, the excursion and timing of laryngeal movement were measured
by both visual and tactile monitoring to allow for more accurate judgement
of the competence and effectiveness of the swallow pattern.
For each of these, a five-point rating scale was developed so that the
competence of patients in the study could be expressed in numerical form. In
each case a score of 5 represented a normal level of competence and 0 complete
inability to achieve the task. The rating scales for each of the above are given
in the Appendix.

Procedure
All patients had, first, the intelligibility of their speech, then their dysphagia,
assessed in a single session lasting approximately 45 minutes.
MOTOR SPEECH ACTS IN ASSESSMENT OF DYSPHAGIA 219

The modified intelligibility section of Yorkston and Beukelman’s (1984)

assessment was administered following the procedure outlined in the manual.
The responses were tape-recorded and scored by one of six judges. Each
judge was a health service employee who had some experience of dysarthric
speech, but none had phonetic training or knowledge of the assessment. No
judge heard any of the four sets of items on more than one occasion.
Each patient’s swallowing difficulty was evaluated in a variety of ways.
Patients were questioned and their clinicians, carers and speech and language
therapy notes were consulted. They were evaluated on limited intake of fluid
and semi-solids or solids, restrictions on the amount controlling for risks of
aspiration. The primary source of data was the rating scales of the ten features
derived from the RIC Clinical Evaluation of Dysphagia. Scales 1-9 were based
on observations, both visual and tactile, during the assessment and 10 on
patients’ answers to questions.

RESULTS
Intelligibility was taken to be the percentage of words accurately transcribed
by judges. This gave separate figures for individual words and for sentences
and an overall score for these combined (Table 1).
Examination of Table 1 suggests that CVA patients were more intelligible
than Parkinson’s disease patients and words in sentences more intelligible than
words alone. Analysis of variance showed that neither of these effects was
significant, however. In view of this and of the similar standard deviations,
intelligibility in the two groups may be judged to cover a similar range of
scores. Pearson product moment correlations between single word and word
in sentence intelligibility scores were strongly significant for the CVA patients
( r = 0.922, p < 0.001) but failed to reach significance for the Parkinson’s
disease patients ( r = 0.467).
The means and standard deviations of the rating scores from the ten scales
obtained from the dysphagia assessment are shown in Table 2 for each patient
group.
Had the objective been to seek relationships between intelligibility and the
dysphagia, multiple regression would have provided a suitable analytical
technique. This would have assessed the degree to which each dysphagia
variable predicted the intelligibility score. A relationship here would have
Table 1: Mean percentage scores on assessment of intelligibility.

CVA Parkinson’s disease

Single words
Mean 62.2 58.7
s.d. 27.0 27.8
Sentences
Mean 81.4 67.3
s.d. 22.8 22.7
Overall
Mean 78.6 66.1
s.d. 22.0 21.9
220 KENNEDY, PRlNG AND FAWCUS

Table 2: Means and standard deviations for each dysphagia scale for each group.

CVA patients Parkinson’s disease patients

Dysphagia scale* Mean s.d. Mean s.d.

1 5.0 0 4.4 0.83

2 3.9 1.10 3.4 1.26
3 4.5 0.96 3.9 1.10
4 3.3 1.25 3.9 0.87
5 3.7 1.15 4.6 0.68
6 2.9 1.20 3.4 1.42
7 3.2 1.31 4.2 0.92
8 3.4 1.34 4.3 0.47
9 4.9 0.31 4.0 1.25
10 4.0 1.49 3.8 1.23

* 1. Maintaining lip closure throughout the oral phase.

2. Lateral movements of the tongue during mastication.
3. Posterior propulsion of the bolus by the tongue.
4. Triggering of the swallow for liquid intake.
5. Triggering of swallow for semi-solidkolid intake.
6. Triggering of cough before, during or after liquid intake.
7. Triggering of cough before, during or after semi-solidsolid intake.
8. Voice quality following liquid intake.
9. Reflux of material through the cricopharyngeal sphincter leading to oral or nasal emission
10. Complaints of food sticking in the throat or upper oesophagus.

Table 3: Correlations between intelligibility and dysphagia scores.

Correlation with intelligibility for

Dysphagia scale* CVA patients Parkinson’s disease patients All patients

1 O.OO0 0.229 0.300
2 0.312 -0.355 -0.011
3 -0.707 -0.594 -0.467
4 -0.111 0.053 -0.210
5 -0.215 -0.110 -0.179
6 -0.077 0.128 0.002
7 -0.154 -0.093 -0.246
8 0.132 O.OO0 0.069
9 0.137 -0.055 0.116
10 0.139 -0.281 -0.096
Overall score 0.201 -0.109 -0.077
Significance level for CVA and Parkinson’s disease patients: 0.521 (one tail)
0.602 (two tail)
Significance level for all patients: 0.378 (one tail)
0.444 (two tail)
* 1. Maintaining lip closure throughout the oral phase.
2. Lateral movements of the tongue during mastication.
3. Posterior propulsion of the bolus by the tongue.
4. Triggering of the swallow for liquid intake.
5 . Triggering of swallow for semi-solidsolid intake.
6. Triggering of cough before, during or after liquid intake.
7. Triggering of cough before, during or after semi-solidsolid intake
8. Voice quality following liquid intake.
9. Reflux of material through the cricopharyngeal sphincter leading to oral or nasal emission.
10. Complaints of food sticking in the throat or upper oesophagus.
MOTOR SPEECH ACTS IN ASSESSMENT OF DYSPHAGIA 221

supported the view that an assessment of dysarthria was predictive of swallow-

ing difficulties; moreover, a relationship with some but not other parts of the
dysphagia assessment would have implied that dysarthria coexisted with
particular types of swallowing difficulty. The objective was to demonstrate
that no relationship existed, however. Consequently, a simpler approach was
to correlate each individual dysphagia score with the intelligibility score. Table
3 gives the Spearman correlation coefficients between intelligibility and each
of the dysphagia assessments for each group of subjects and for the two groups
combined.
If the intelligibility scores were predictive either of swallowing problems in
general or of particular components of swallowing problems, significant posi-
tive correlations would be expected. With this prediction, a one-tail test of
significance would be used. None of the correlations obtained either for all
patients or for the separate groups gave positive correlations that approached
the required level of significance. The only correlations of any size were
negative ones for scale 3. These were statistically significant for each group
and for all patients combined. They suggest that intelligibility was better for
subjects who scored poorly on posterior propulsion of the bolus by the
tongue, a result that was not anticipated.
The remaining correlation coefficients were not significant, nor did they
exhibit a trend towards it. They appear equally likely to be positive or negative
and have an overall mean of -0.021. A one-sample t-test may be used to
compare this sample with a population mean of zero. This gave a t value of
0.43 (a value of 1.74 is required for significance). In other words, this sample
of correlations does not differ from the population of correlations obtained
from unrelated variables.

DISCUSSION
This study set out to obtain experimental evidence that no relationship exists
between intelligibility and measures of swallowing problems in patients with
both dysarthria and dysphagia. No relationship was found except for an
unexpected inverse correlation between intelligibility and posterior propulsion
of the bolus by the tongue. Of necessity the experiment predicted, rather than
rejected, the null hypothesis. As in all such cases, conclusions should be drawn
carefully and it is important to consider possible criticisms of the methodology.
The experiment might be criticised for lacking the necessary statistical power
to demonstrate a relationship; most simply, it might be argued that a relationship
would emerge if greater numbers of subjects were tested.
Although the numbers, particularly for the separate groups, were small,
there are reasons for doubting this as a source of the null result. If a
relationship existed the correlations should have been consistently in the
predicted direction even if statistical significance was not obtained. This was
not the case; indeed, the correlation coefficients gave the impression of being
randomly distributed about a mean of zero.
Furthermore, it is worth considering what a significant correlation would
have meant here. A correlation of 0.378 was the minimum required for
significance in the experiment. Measures of dysarthria are presumably included
222 KENNEDY, PRlNG AND FAWCUS

in dysphagia assessments to improve diagnostic prediction. In this context,

statistical significance is a poor substitute for diagnostic reliability. The former
requires a relationship that would be less than reliable for diagnostic
purposes. Hence, even if significance had been achieved its diagnostic value
would have been questionable.
A further criticism might be that the assessments were insufficiently sensitive
measures. Correlations might not have occurred simply because there was
insufficient variation in the scores to allow them to do so. Table 1 presents
the means and standard deviations of the intelligibility scores from which it
may be seen that considerable variability occurred. The dysphagia rating scales
are more problematic. Table 2 shows that variability here was greater for
some scales than others; however, no suggestion was found that correlations
increased where greater variability was present. Thus, even where variability
was small it is reasonable to conclude that patients’ intelligibility varied despite
their similarity in that aspect of swallowing, and was consequently unrelated
to it.
Had significant correlations been obtained it would have been of interest
whether they were for the overall swallowing score or for components of it.
Retention of speech-related assessments could be justified; indeed they might
be particularly valuable, if they predicted the severity of a particular aspect
of swallowing. If a correlation had occurred for a single scale its importance
would have been difficult to assess in view of the experimentwise type 1 error
rate. A similar problem arises in interpreting the negative correlation obtained
for scale 3. As negative correlations were not anticipated, it is tempting to
dismiss it as a type 1 error. However, the fact that the same correlation was
significant for both patient groups suggests a real, if unexpected, relationship.
A possible explanation for this effect is the fact that posterior propulsion of
the bolus is carried out by the intrinsic lingual musculature, whereas the
tongue’s movement for speech activity is undertaken predominantly by the
extrinsic muscles.
A final criticism might be of the assessments themselves. Both were
modifications of existing tests. The tests may be unfamiliar in clinical practice
and the modifications were intended for experimental rather than for clinical
use. There has been some discussion of whether single word measures are
reliable predictors of the intelligibility of continuous utterances. Schiaretti,
Sitter, Metz and Harde (1984) using deaf and hearing speakers in a poor
listening environment and Platt, Andrews, Young and Neilson (1978) using
cerebral-palsied speakers suggested that they were. In the present data, a
much stronger correlation was found between single-word and word-in-sentence
intelligibility for CVA patients than for Parkinson’s patients. Confirmation of
this finding, which would suggest that prediction of intelligibility from single-
word scores is more possible with some neurologically impaired groups than
others, would require separate research.
The assessment of dysphagia formalised conventional bedside examinations
while emphasising a functional approach to the procedure. The improvised
scoring system allowed numerical values to be given to particular elements of
patients’ swallowing difficulties. Comparisons across deficits or between patient
groups were handicapped by the ordinal scoring system. Nevertheless, although
MOTOR SPEECH ACTS IN ASSESSMENT OF DYSPHAGIA 223

the overall level of deficit was similar in each group, the profiles showed
different trends. Parkinson’s patients showed more diffuse handicaps whereas
the deficits of CVA patients were most pronounced in the areas of triggering
of the swallow and triggering of cough before, during or after all types of
intake and in voice quality following liquid intake. These findings are broadly
consistent with the suggestions of Logemann (1983) and Robbins and Levine
(1988) that stroke patients experience delay or absence in initiating the
pharyngeal phase of the swallow regardless of their lesion site. They are also
consistent with Linden and Sieben’s (1983) finding that liquid intake frequently
leads to aspiration in stroke patients and a ‘wet-hoarse’ voice quality after
laryngeal penetration. This suggests that ‘wet-hoarse’voice quality after liquid
intake may be a good predictor of aspiration for these patients.
As was discussed in the introduction, there are various sources of evidence
that swallowing and speech difficulties may be independent. In stroke patients,
swallowing and speech can improve independently of one another (Netsell,
1986) and, in Parkinson’s disease patients, they need not accompany one
another (Sarno, 1968; Duvoisin, 1982). The present data appear to be the
first to examine experimentally the relationship between the two, however.
Although replication with larger numbers and a greater variety of patients
may be desirable, these initial results suggest that there is no functional
relationship between motor speech acts and swallowing difficulties. This
conclusion would not apply to cases involving head and neck surgery where,
as Logemann (1985) points out, underlying structural abnormalities produce
a more direct relationship between speech and swallowing. In cases of
neurological origin, however, the results question the need to include motor
speech acts in the evaluation of dysphagia.
This paper began by discussing the concerns of therapists who are asked to
treat dysphagic patients without the benefit of radiological investigation. It
ends by showing that two alternative approaches to assessment produce
unrelated results. Previous research has shown that bedside examinations are
poor at detecting aspiration and invite the conclusion that there can be no
substitute for radiological investigation. If we fully accept this conclusion the
present result may be of little consequence.
This is surely too strong a reaction. The value of videofluoroscopy is not
lessened by the recognition that other assessments may be improved. More-
over, in the continuing uncertainty about its availability, this should be a
priority. The results here support the neurophysiological evidence that there
need be no relation between dysarthria and swallowing. In contrast both the
literature and observation of clinical practice suggest that clinicians are
influenced by the presence of dysarthria in their dysphagic patients. They are
likely to assess both and, as the assessments of speech-related activities are
familiar, whereas those of dysphagia have been heavily criticised, the former
may strongly influence their decisions. The greater this influence, the more
important are the findings of this study both for current clinical practice and
for future considerations of improving assessment techniques where radiological
methods are unavailable.
224 KENNEDY, PRlNG AND FAWCUS

APPENDIX

Rating Criteria for the Ten Variables included in the Evaluation of

Dysphagia
1. Maintaining lip closure throughout the oral phase:
(5) Action performed competently with no difficulty or drooling.
(4) Fairly consistent achievement with occasional difficulties and occasional drooling.
(3) Intermittent achievement with varying degrees of success and variable amounts of drooling.
(2) Minimal achievement and considerable difficulty in performing action; considerable drooling.
(1) No achievement with severe difficulty in attempting to perform action; perpetual drooling.
2. Lateral movements of the tongue during mastication:
( 5 ) Action performed competently with no difficulty, no residue in oral cavity.
(4) Occasional difficulties in performing action; occasional residue in oral cavity.
(3) Intermittent achievement with varying degrees of success;variable amounts of residue in oral cavity.
(2) Minimal achievement with considerable difficulty in attempts to perform action; considerable
residue in oral cavity.
(1) No achievement with severe difficulty in attempts to perform action; food remains unmasticated.
3. Posterior propulsion of the bolus by the tongue:
( 5 ) Action performed competently and confidently; transfer of bolus within 1 second.
(4) Fairly consistent achievement with occasional difficulty in performing the action and occasional
delay in transfer.
(3) Intermittent achievement with variable success and variable delay in transfer.
(2) Minimal achievement with considerable difficulty in performing the action and considerable
delay in transfer.
(1) No achievement, with severe difficulty in attempts to perform the action; bolus remains in
anterior of oral cavity.
4. Triggering of the swallow for liquid intake.
5. Triggering of swallow for semi-solidholid intake:
(5) Action performed competently with no delay; reflex triggered promptly.
(4) Fairly consistent achievement with occasional delay in performing action and triggering reflex.
(3) Intermittent achievement with variable delay in performing action and triggering the reflex.
(2) Minimal achievement with considerable delay in performing action; considerable delay in
triggering reflex.
(1) No achievement after considerable delay with or without attempts to perform action; no reflex
triggered.
6. Triggering of cough before, during or after liquid intake.
7. Triggering of cough before, during or after semi-solidsolid intake:
(5) Never.
(4) Occasionally.
(3) Intermittently.
(2) Frequently
(1) Perpetuallyhever triggered as reflex absent.
8. Voice quality following liquid intake:
(5) Good clear phonation with no evidence of ‘wet hoarse’ voice.
(4) Fairly consistent production of clear phonation with occasional evidence of ‘wet hoarse’ voice.
(3) Intermittent production of clear phonation with variable evidence of ‘wet hoarse’ voice.
(2) Occasional production of clear phonation with considerable evidence of ‘wet hoarse’ voice.
(1) No production of clear phonation with persistent evidence of ‘wet hoarse’ voice.
MOTOR SPEECH ACTS IN ASSESSMENT OF DYSPHAGIA 225

9. Reflux of material through the cricopharyngeal sphincter leading to oral or nasal emission:
(5) Never.
(4) Occasionally.
(3) Intermittently.
(2) Frequently.
(1) Persistently.
10. Complaints of food sticking in the throat or upper oesophagus:
(5) Never.
(4) Occasionally.
(3) Intermittently.
(2) Frequently.
(1) Persistently.

REFERENCES
CHERNEY, L. R., CANTIERI, c . A. & PANNELL, J. J. (1986). R.1.C. c h i c a l Evaluation of Dysphagia.
Maryland: Aspen Systems Corporation.
COLLEGE OF SPEECH THERAPISTS (1990a). Dysphagia working party: Interim report. College of Speech
Therapists’ Bulletin, May.
COLLEGE OF SPEECH THERAPISTS (1990b). Dysphagia working party: Dysphagia position paper, April
1990.
CROXON, s. c. M. & PYE, I. F. (1988). Dysphagia as the presenting feature of Parkinson’s disease.
Geriatric Medicine 18, 16.
DUVOISIN, R. c . (1982). Parkinson’s Disease: A Guide for Patient and Family. New York: Raven Press.
EDELMAN, G. M. (1988). D sphagia survey. college of Speech Therapists’ Bulletin, April.
ENDERBY, P. (1983). Frenciay Dysarthria Assessment. San Diego: College-Hill Press.
GROHER, M. E. (1984). Dys hagia: Diagnosis and Management. Boston: Buttenvorths.
HOEHN, M. H. &YAHR, M. D. (567). Parkinsonism: onset, progression and mortality. Neurolo 17,427-442.
HORNER, 1. MAssEY, E. w . (1988). Silent aspiration following stroke. Neurology 3, 87-319.
JOHNSON, J. A. & PRING, T. R. (1990). Speech therapy and Parkinson’s disease: a review and further
data. British Journal of Disorders of Communication 25, 183-194.
LARSON, c . (1985). Neurophysiology of speech and swallowing. Seminars in Speech and Language 6 ,
275-291.
LINDEN, P. & SIEBENS, A. A. (1983). Dysphagia: predicting laryngeal penetration. Archives of Physical
Medicine and Rehabilitation 64, 281-284.
LOGEMANN, J. A. (1983). Evaluation and Treatment of Swallowing Disorders. San Diego: College-Hill
Press.
LOGEMANN, J. A . (1985). The relationship of speech and swallowing in head and neck surgical patients.
Seminars in Speech and Language 6 , 251-259.
LOGEMANN, J. A , , BLONSKY, E. R. & BOSHES, B. (1975). Dysphagia in Parkinsonism. Journal of the
American Medical Association 231, 65L71.
LOGEMANN, J. A . , FISHER, H. B . , BOSHES, B. & BLONSKY, E. R. (1978). Frequency and co-occurrence of
vocal tract dysfunction in the speech of a large sample of Parkinson patients. Journal of Speech
and Hearing Research 43, 47-57.
MARTIN, B. J. w. & CORLEW, M. M. (1990). The incidence of communication disorders in dysphagic
patients. Journal of Speech and Hearing Disorders 55, 28-32.
MILLER, R. M. (1984). Evaluation of swallowing disorders. In: M. E. Groher (Ed.) Dysphagia, Diagnosis
and Management. Boston: Buttenvorths.
NETSELL, R. (1986) A Neurological View of Speech Production and the Dysarthrias. San Diego: College
Hill Press.
, J . , ANDREWS, G . , YOUNG, M. & NEILSON, P. D. (1978). The measurement of speech impairment
P L A ~ L.
of adults with cerebral palsy. Folia Phoniatrica 30, 50-58.
PRICE, G. J . , JONES, c. J . , CHARLETON, R. A. &ALLEN, c. M. c . (1987). A combined approach to the
assessment of neurological dysphagia. Clinical Otolaryngology 12, 197-201.
ROBBINS, J. A. & LEVINE, R. (1988). Swallowing after unilateral stroke of the cerebral cortex: preliminary
experience. Dysphagia 3, 11-17.
ROBERTSON, s. I. & THOMSON, F. (1984). Speech therapy in Parkinson’s disease: a study of the efficacy
and long term effects of intensive treatment. British Journal of Disorders of Communication 19,
213-224.
SARNO, M. T. (1968). Speech impairment in Parkinson’s disease. Journal of Neurology, Neurosurgery
and Psychiatry 46, 140-144.
SCHIARE~N , . , SIITER, R. METZ, D . & HARDE, R. (1984). Prediction of contextual speech intelligibility
from isolated word intelligibility measures. Journal of Speech and Hearing Research 27, 623-626.
226 KENNEDY, PRlNG AND FAWCUS

scow, s. & CAIRD, F. I . (1983). Speech therapy for Parkinson's disease. Journal of Neurology,
Neurosurgery and Psychiatry 46, 140-144.
SPLAINGARD, M. L., HUTCHINS, B . , SULTON, A. D. & CHAUDHURI, G . (1988). Aspiration in rehabilitation
patients: Videofluoroscopy versus bedside clinical assessment. Archives of Physical Medicine and
Rehabilitation 69, 637-640.
WARMS, T. s., CHAMPION, R. & MORTENSEN, L. (1990). PHAD validation and correlation with the
modified barium swallow. Adult Speech Pathology Department, Westmead Hospital, NSW,
Australia.
YORKSTON, K. M. & BEUKELMAN, D. R. (1984). Assessment of Intelligibility of Dysarthric Speech. Texas:
Pro-Ed.
Address correspondence to Dr T. R. Pring, Department of Clinical Communication Studies, The City
University, Northampton Square, London EClV OHB, UK.

Received July 1991; revised version accepted October 1992.