Jérôme R Lechien1, Véronique Delvaux2, Kathy Huet2, Mohamad Khalife3, Anne-Françoise Fourneau3, Myriam Piccaluga2, Bernard Harmegnies2, Sven Saussez4. 1. Laboratory of Anatomy and Cell Biology, Faculty of Medicine, UMONS Research Institute for Health Sciences and Technology, University of Mons (UMONS), Mons, Belgium; Laboratory of Phonetics, Faculty of Psychology, Research Institute for Language Sciences and Technology, University of Mons (UMONS), Mons, Belgium; Department of Otorhinolaryngology and Head and Neck Surgery, RHMS Baudour, EpiCURA Hospital, Baudour, Belgium.. Electronic address: Jerome.Lechien@umons.ac.be. 2. Laboratory of Phonetics, Faculty of Psychology, Research Institute for Language Sciences and Technology, University of Mons (UMONS), Mons, Belgium. 3. Department of Otorhinolaryngology and Head and Neck Surgery, RHMS Baudour, EpiCURA Hospital, Baudour, Belgium. 4. Laboratory of Anatomy and Cell Biology, Faculty of Medicine, UMONS Research Institute for Health Sciences and Technology, University of Mons (UMONS), Mons, Belgium; Department of Otorhinolaryngology and Head and Neck Surgery, RHMS Baudour, EpiCURA Hospital, Baudour, Belgium.
Abstract
OBJECTIVES: The study aimed to explore the impact of the selection of the analyzed time interval on the significance of acoustic measurements used to investigate laryngopharyngeal reflux (LPR) treatment efficacy, and based on these results to develop an alternative statistical approach in data analysis focusing on individual patient vocal behavior. STUDY DESIGN: This is a prospective case series. METHODS: From September 2013 to July 2015, 41 patients with a reflux finding score (RFS) > 7 and a reflux symptom index (RSI) > 13 were enrolled and treated with pantoprazole 20 mg twice daily and diet behavioral changes for 3 months. Voice recordings were performed at baseline and after 3 months of treatment. Most stable time intervals of 1, 2, 3, 4, and 5 seconds, and a 1-second time interval positioned at mid-production, were subjected to acoustic analysis. Based on the latter, we developed an "informativeness coefficient" for each acoustic parameter that aimed at assessing its sensitivity to clinical resolution in the case of LPR disease. RESULTS: Significant clinical improvement (RSI and RFS) was observed after treatment (P < 0.05). The acoustic analysis revealed that acoustic parameters significantly improving from pre- to posttreatment varied across time intervals. The duration and the position of the analyzed time interval in the production yielded considerable differences in the results. Analysis of the informativeness coefficient indicated that jitter, jitter percent, relative average perturbation (RAP), pitch perturbation quotient (PPQ), shimmer (ShdB), shimmer percent (Shim), amplitude perturbation quotient (APQ), and smoothed amplitude perturbation quotient (sAPQ) were the indices most sensitive to medical treatment efficacy, with a coefficient ranging from 75.86% to 86.21%. CONCLUSIONS: Depending on the selection of the time interval over which the acoustic parameters are measured, the potential effect of the treatment may or may not be statistically demonstrated. Future studies are needed to establish standardized methodological procedures for acoustic data analysis.
OBJECTIVES: The study aimed to explore the impact of the selection of the analyzed time interval on the significance of acoustic measurements used to investigate laryngopharyngeal reflux (LPR) treatment efficacy, and based on these results to develop an alternative statistical approach in data analysis focusing on individual patient vocal behavior. STUDY DESIGN: This is a prospective case series. METHODS: From September 2013 to July 2015, 41 patients with a reflux finding score (RFS) > 7 and a reflux symptom index (RSI) > 13 were enrolled and treated with pantoprazole 20 mg twice daily and diet behavioral changes for 3 months. Voice recordings were performed at baseline and after 3 months of treatment. Most stable time intervals of 1, 2, 3, 4, and 5 seconds, and a 1-second time interval positioned at mid-production, were subjected to acoustic analysis. Based on the latter, we developed an "informativeness coefficient" for each acoustic parameter that aimed at assessing its sensitivity to clinical resolution in the case of LPR disease. RESULTS: Significant clinical improvement (RSI and RFS) was observed after treatment (P < 0.05). The acoustic analysis revealed that acoustic parameters significantly improving from pre- to posttreatment varied across time intervals. The duration and the position of the analyzed time interval in the production yielded considerable differences in the results. Analysis of the informativeness coefficient indicated that jitter, jitter percent, relative average perturbation (RAP), pitch perturbation quotient (PPQ), shimmer (ShdB), shimmer percent (Shim), amplitude perturbation quotient (APQ), and smoothed amplitude perturbation quotient (sAPQ) were the indices most sensitive to medical treatment efficacy, with a coefficient ranging from 75.86% to 86.21%. CONCLUSIONS: Depending on the selection of the time interval over which the acoustic parameters are measured, the potential effect of the treatment may or may not be statistically demonstrated. Future studies are needed to establish standardized methodological procedures for acoustic data analysis.
Authors: Jérôme R Lechien; Camille Finck; Pedro Costa de Araujo; Kathy Huet; Véronique Delvaux; Myriam Piccaluga; Bernard Harmegnies; Sven Saussez Journal: Eur Arch Otorhinolaryngol Date: 2016-03-23 Impact factor: 2.503