339 / 412
etiology patients. Sensitivity of the screen was 63% with
100% specificity. Sensitivity increased to 71% with the inclu-
sion of a barium capsule. In comparison to the esophagram,
the esophageal screen subjects patients to approximately 10
times less radiation dose and is relatively quickly completed at
the end of a standard VFSS procedure.
11
The aim of this pro-
spective, observational study was to investigate the prevalence
of esophageal abnormalities in an SLP-led VFSS clinic.
Methods
This study received appropriate regional ethics approval
(University of Auckland Human Participants Ethics
Committee 9263).
Participants
Data from 111 inpatients and outpatients consecutively
referred to an SLP-led VFSS clinic (between May and
December 2013) were collected. Patients were referred by
the ORL service (59) or by another speech-language pathol-
ogist (52), with a mean (SD) cohort age of 71 (14.95) years
(range, 20-95 years). Fifty-one patients were male (46%).
Referrals were all for complaints of dysphagia attributed to
mixed etiologies: 36 neurological (32%), 37 dysphagia of
unknown cause (33%), 28 otorhinolaryngology (eg, head
and neck cancer [25%]), and 10 other (eg, unwell elderly
[9%]). Participants were excluded if their VFSS recording
had no esophageal visualization or had no timer or no cali-
bration ring, since these were required to complete the
timing and displacement measures.
Esophageal Videofluoroscopic Assessment
Studies were performed in a radiology suite using a
Videofluoroscope (Toshiba, Tokyo, Japan) and recorded at
30 frames per second onto a USB drive. Timing information
was superimposed on the fluoroscopic recording in 100ths
of a second using a Horita VS-50 Video Stopwatch (Horita,
Capistrano Beach, California). A 19-mm diameter radio-
opaque ring was taped to the patient’s chin (in the lateral
plane) and shoulder (in the anterior-posterior plane) to allow
calibration for displacement measures. A medical radiation
technician (MRT) and a speech-language pathologist were
present at all procedures.
For the standard VFSS protocol, the patient was
screened in the lateral plane. The patient was presented
with 1 mL, 3 mL, and then 20 mL of thin barium (E-Z
Paque, E-Z-EM Anjou, Canada; 100% w/v) followed by
half a cup of thin barium through a straw. The patient was
then given 3 mL of barium paste (E-Z paste, E-Z-EM;
60% w/w). The procedure was truncated if required for
patient safety.
The esophageal phase involved the introduction of 2
boluses after completion of the standard VFSS protocol: a
20-mL fluid bolus and a 13-mm barium capsule. The patient
was positioned in the anterior-posterior plane, standing
whenever possible. The patient was asked to ‘‘swallow all
in one go’’ to avoid deglutitive inhibition. The MRT fol-
lowed the bolus from the oral cavity through the lower
esophageal sphincter (LES) until clearance into the stomach.
Screening was continued for up to 15 seconds. If there was
still residue in the esophagus, screening was ceased for 15
seconds, then recommenced. If residue was still present, the
patient was asked to take a dry swallow to see if clearance
occurred. If residue was still present, the patient was then
offered a water swallow as well as being asked if he or she
could feel any remaining bolus. A screen shot was used to
identify complete clearance.
Data Collected
Age, sex, and comorbidities were recorded for each patient.
Each VFSS was analyzed using real-time and frame-by-frame
viewing (Quicktime Media Player; Apple, Cupertino,
California). Videos were scored for the presence of oral, phar-
yngeal, and esophageal abnormalities (yes/ no). Prolonged
bolus manipulation, anterior spillage from lips, premature spil-
lage into the pharynx, and oral residue were all considered
oral abnormalities. Nasal regurgitation, reduced epiglottic
deflection, delay in swallowing initiation, pharyngeal residue,
penetration, aspiration, and prolonged pharyngeal transit time
were considered pharyngeal abnormalities. Esophageal bolus
stasis, bolus redirection/intraesophageal reflux, gastroesopha-
geal reflux, esophagopharyngeal reflux, hiatal hernia, pro-
longed esophageal transit, and pill stasis were all considered
esophageal abnormalities. Pharyngeal transit time (PTT) was
recorded and translated into a binary measure of (1) within
normal limits vs (2) more than 2 standard deviations (SD) out-
side of normal limits.
12
Esophageal transit time (ETT) was
also recorded. A conservative cutoff of over 15 seconds was
selected for abnormality. Previous published work has defined
normal liquid transit through the esophagus as less than 13
seconds.
11,13,14
Maximum penetration-aspiration scale (PAS)
scores were recorded, and scores 6 and above were considered
an aspiration event.
15
To explore whether esophageal abnorm-
alities can be predicted by objective pharyngeal measures, we
calculated the pharyngeal constriction ratio (PCR)
12
and phar-
yngoesophageal segment maximum opening (PESmax)
12
using
the Universal Desktop Ruler (AVPSoft). These measures were
also translated into binary measures of (1) within normal limits
vs (2) more than 2 SD outside of normal limits.
12
All measures
were taken from the largest fluid bolus ingested.
Data Analysis
Swallow studies were reported by an experienced otolaryn-
gologist, specializing in dysphagia management, and by a
speech-language pathologist, trained in quantitative analysis
of VFSS using the method developed by Leonard and
Kendall.
12
Interrater reliability for all measures was calcu-
lated on 30% of videos by a third researcher. Videos were
randomly selected by a fourth researcher. The third rater
was blinded to the first researcher’s scoring and patient
etiology. Total agreement across measures was 98%, with a
k
coefficient of 0.92. Lack of agreement was found for 1
PAS score (1 vs 2), and although there was slight variance
in PESmax (maximum variance .08) and PCR (maximum
variance .07), this did not change binary measures of within
Miles et al
114