270 / 412
Despite the acknowledgment of variability in the
laryngoscopic appearance of VFP implicit in these terms,
no widely accepted rating system, or even a consistent,
unified vocabulary to describe such variability exists.
Consistency and reproducibility is fundamental in the
evaluation and comparison of cases, their outcomes, and
their treatment; even a brief reflection on the House-
Brackmann scale for grading facial paralysis reveals the
broad potential utility of such a standardized approach.
The goal of this investigation was to evaluate character-
istics of the laryngoscopic appearance of VFP with respect to
inter- and intra-rater consistency, and to identify features
for which clinical consensus exists, which might lend them-
selves to a useful, standardized description system for VFP.
MATERIALS AND METHODS
Selection of Examinations
Strobovideolaryngoscopies of patients with a clinical diag-
nosis of VFP based on history, physical examination, and
laryngoscopy were selected from a corpus of such examinations
recorded during the course of routine evaluation. All recording
was made under stroboscopic light using either a rigid glass rod
peroral laryngoscope (Model 9106; KayPentax, Lincoln Park,
NJ) or a distal chip flexible transnasal laryngoscope (VNL-
1170K; Pentax Medical, Montvale, NJ). Only patients with VFP
of known cause were included. Nineteen had pathology or
injury limited to the recurrent laryngeal nerve, and three had
paralysis from vagal neuropathy. For inclusion, the examination
had to feature a sustained, unobstructed view of the glottis, ary-
tenoids, aryepiglottic folds, and pyriform sinuses. A 20-second
sample of each exam, containing at least one example each of
phonatory adduction and postphonatory abduction as well as
several cycles of phonatory vibration, was selected and saved.
The pitch and intensity capabilities of patients were variable
from exam to exam, as one would expect in cases of VFP. How-
ever, as evaluators’ ratings of identical examinations were
assessed in this study, no effort was made to standardize these
parameters among examinations.
Patients who had been treated for their VFP in any way,
including injection augmentation, framework surgery, and rein-
nervation were excluded. Cases of vocal fold paresis in which
significant gross vocal fold mobility remained, even if it was
clearly less than normal, were excluded. The authors recognize
that the distinction between paralysis and paresis is not always
sharply defined and does not necessarily reflect the underlying
neurologic status.
Each exam was numbered, and randomly ordered lists of
these exams, one for each potential reviewer, were generated. In
each list, three exams were selected to be repeated by a random
number generator, which created a new set of three numbers for
each of the reviewers and brought the total number of exams to
25. Therefore, every reviewer had a randomly selected series of
repeat examinations to test their intra-rater reliability. The pur-
pose of de novo, random selection of the repeat exams for every
reviewer was to eliminate any possibility that one exam might be
more easily identified on repeat viewing than the others. Such a
scenario would bias the entire sample and yield an artificially
Fig. 1. Rating sheet for evaluators.
Laryngoscope 120: July 2010
Rosow and Sulica: Laryngoscopy of Vocal Fold Paralysis
45