Communicative Suitability of Voice Following Radiotherapy for T1 Glottic Carcinoma

Abstract

The objective of the study was to determine whether a communicative suitability rating instrument could be used in a meaningful way to assess functionality of voice following radiotherapy for T1 glottic cancer. Seventeen naive listeners judged the suitability of voice of a patient group with T1 glottic carcinoma (n = 20) just before treatment, a group of patients (n = 40) after radiotherapy, and a matched control group (n = 20) of normal speakers. Listeners rated suitability on a 10-point scale for 10 speaking situations, which supposedly make different demands. In order to validate scores on communicative suitability, ratings were related to perceptual voice quality evaluations and videolaryngostroboscopic evaluations. Results indicate that the concept of measuring listener judgments of communicative suitability of voice is basically sound. Raters are reliable and can discriminate between groups of normal and pathological voices. Patients with T1 glottic carcinoma (assessed before the start of treatment) have on average the least suitable voices. Following radiotherapy suitability is, on average, improved, but does not approach the suitability of normal voices. Ratings on communicative suitability were clearly related to perceptual voice quality aspects and videolaryngostroboscopic evaluations. A subset of three communicative suitability rating scales is recommended as part of the protocol for evaluating voice outcome after radiotherapy for early glottic cancer, besides perceptual evaluation of voice quality by trained and naive raters, videolaryngostroboscopy, acoustical analyses, and self-ratings of vocal performance.

Introduction

The prognosis of T1 glottic carcinoma is relatively good, with reported cure rates over 90%. Treatment options for early glottic cancer are radiotherapy (RT), endoscopic surgery (with or without CO₂ laser), and external surgery (cordectomy or vertical hemilaryngectomy). Since cure rates are similarly good for these approaches, the preservation or restoration of voice has become an important factor in the selection of treatment strategy.

As Hirano and Bless¹ stated, voice is a complex phenomenon that requires multiple measures to describe its characteristics. In earlier research, Verdonck et al2, 3 concluded that assessing voice quality, vocal function, and vocal performance are essential dimensions for evaluating a patient's voice outcome after radiotherapy for early glottic cancer. Voice characteristics of patients diagnosed with early glottic cancer improved after radiotherapy; however, they were not fully normalized for half of the patients. The method of obtaining a biopsy of the lesion for initial diagnosis and smoking habits after treatment clearly affected the risk of voice impairment and associated consequences in daily life. These conclusions were based on perceptual evaluations from expert trained listeners and acoustical analyses of voice quality, stroboscopic analysis of vocal function, and self-ratings of vocal performance.2, 3

A patient's voice quality, vocal function, and even vocal performance are sensitive to many different kinds of deviation from an ideal standard, although the voice may be perfectly suitable for communicating in most normal daily life situations.

The concept of communicative suitability has been proposed by Franken, van Bezooijen, and Boves⁴ for speech of people who stutter. They defined communicative suitability as the speaking situation-dependent adequacy of speech as judged by listeners. They found that the rating instrument consisting of 10 speaking situations can be scored reliably and that more demanding situations (such as giving a lecture) received lower suitability scores than the less demanding ones (such as talking about everyday events with a friend), as could be expected. Also, the speech of the people who stutter was judged significantly less suitable than the speech of reference speakers. Furthermore, naive raters were considerably less tolerant in their judgments than clinicians or stuttering listeners.

The main purpose of our study is to evaluate and adjust this rating instrument to assess the communicative suitability of postglottic cancer treatment voices rated by naive listeners. For a new rating instrument to be readily used, it must first of all be proved that ratings are reliable. Furthermore, it is required that the instrument differentiates normal voices from pathological voices and low demanding situations from more demanding situations. Finally, in order to validate scores on communicative suitability, ratings have to be related to perceptual voice quality evaluations by naive raters and videolaryngostroboscopic evaluations of the vocal folds.