Effect of Topical Anesthesia on the Phonation Process During High-Speed Digital Imaging Laryngoscopy in Patients With Laryngopharyngeal Reflux

Standard clinical practice for rigid transoral laryngoscopy, including High-Speed Digital Imaging (HSDI), includes the elective use of topical anesthesia of the oral cavity and oropharynx; however, clearly defined recommendations and indications for its use are lacking, and its potential effect on the phonation process has not been unequivocally established. Upon application to the oral cavity and oropharynx, anesthetic may drain into the larynx and desensitize the mucosa, potentially affecting sensorimotor aspects of laryngeal function. Existing studies on the effect of topical anesthesia on phonation during laryngoscopy are limited and report contradictory findings.

This prospective, randomized, double-blind, placebo-controlled crossover study assigned each participant a unique number via random number generator, determining the order of recording modalities (even numbers recorded first with topical anesthesia, odd numbers first without). Once assigned, the order could not be changed. Randomization and testing order were concealed from the examiner performing the HSDI recording.

Topical anesthesia (lidocaine spray, 100 mg/ml, 0.1 ml per actuation) or physiological saline (placebo) was applied by a junior physician, blinded to the substance used, to the base of the tongue, posterior pharyngeal wall, and soft palate (one actuation per site). HSDI recording was performed 5 minutes after application. Each participant underwent recording in both modalities on two consecutive days, a period considered sufficiently short to avoid confounding phonotraumatic, hormonal, or health-related changes.

HSDI recordings were performed by an experienced otorhinolaryngologist and phoniatrician (>5 years of experience) using a rigid 90° endoscope (Wolf 5562 HRES ENDOCAM with Wolf Auto LP 5132 Hlight light source, Richard Wolf GmbH, Knittlingen, Germany). Participants were seated upright with the tongue held protruded; the endoscope was passed along the midline of the tongue to visualize the full length of the vocal folds. Participants phonated the vowel /a/ dynamically and frequentially, and a 2-second HSDI recording was captured. Recordings were reviewed for quality (minimum 1,000 clear, focused, consecutive images of the full vocal fold length during the steady phonation phase); inadequate recordings were repeated. Fundamental frequency and phonation intensity were recorded via a microphone mounted on the endoscope with software processing. Maximum Phonation Time was determined from three stopwatch-timed measurements of sustained /a/ phonation, taken immediately before and after HSDI recording.

Each HSDI recording was independently assessed by two experienced otorhinolaryngologists/phoniatricians (>5 years of experience) using visual-perceptual analysis. Discrepancies in parameter scoring were resolved through joint re-analysis and discussion until consensus was reached before proceeding to the next recording. Assessment was standardized using the Voice-Vibratory Assessment with Laryngeal Imaging (VALI) form and the European Laryngological Society (ELS) form for glottal closure pattern classification, along with a modified classification of glottal insufficiency. The Reflux Finding Score was re-determined from recordings according to the Belafsky et al. classification.

Participants underwent a follow-up examination within three months of study participation, during which the Reflux Finding Score (via fiberendoscopy) and Reflux Symptom Index (via questionnaire) were re-assessed while participants continued regular proton pump inhibitor therapy and hygienic-dietary measures; participants with persistent pathological findings were excluded from final analysis.

Statistical analysis included the Kolmogorov-Smirnov test for normality of distribution. Normally distributed variables were described using means and standard deviations; non-normally distributed variables were described using medians and ranges. Differences between recordings with and without topical anesthesia were tested using the paired t-test (parametric) or Wilcoxon signed-rank test (non-parametric), as appropriate. Spearman correlation and regression analysis were used to explore associations between anatomical structures and measured characteristics. A p-value <0.05 was considered statistically significant.