Validation of the Spanish Version of the Voice-Related Quality of Life Questionnaire: Psychometric Properties and Clinical Applications

Since the last decades, according to the World Health Organization, health has been understood as a state of the human being where there is physical, social and emotional well-being and not simply as the absence of disease [1]. For this reason, it is essential to consider and evaluate quality of life, which is defined as an individual’s perception of their position in life, considering the cultural context and values in relation to their goals, expectations, standards, and interests [1, 2].

Voice problems have affected or will affect almost a third of the population at some point in their lives, but it is more frequent in those who use their voice in professional activities, which usually seriously affects their quality of life [3]. Voice disorders (VDs) affect social participation at a general level and also have an economic impact on those who use their voice professionally, so patients often report emotional or psychological problems as a consequence [4].

Sometimes it may happen that the perspective of the clinician does not correspond to the perspective of the patient [1‒5], therefore, incorporating the patient’s perception of their physical, mental, and social well-being is crucial in voice evaluation and treatment [6]. This conception of health and quality of life has led to the generation of different patient-reported outcomes measures (PROMs) that delivers a self-assessment report in relation to vocal quality [7], which characterizes the impact of dysphonia on people and their activities of daily living, considering social, functional, emotional, and economic aspects [8‒10].

The vast majority of these PROMs were developed and validated in English and aimed at the population that speaks this language [8‒10]; therefore, for these tools to be used in other languages, they must be translated and adapted based on international guidelines such as the proposals delivered by the Scientific Advisory Committee of Medical Outcomes trust (SAC) [4, 11], and their measurement properties must be demonstrated in a specific cultural context since even though two different countries that speak the same language could have cultural differences; hence, the importance of validating specially adapted PROMs for specific populations, thus obtaining more reliable results [12, 13].

Hogikyan and Serhuraman [10] in 1999 developed the Voice-Related Quality of Life (V-RQOL) questionnaire or PROMs, with two domains: physical functioning and social-emotional. Each domain comprises six and five questions, respectively. It is a statistically robust voice self-assessment tool with good performance of the extracted data [10], useful in evaluation and clinical practice, especially considering the low number of questions and items it has, with an estimated time to complete the protocol of 5 min [14].

The V-RQOL questionnaire is currently adapted and validated for different languages and specific cultural groups, including the versions from India, Brazil, Germany, Norway, Poland, and Turkey [15‒20]. The importance of the validation process within these protocols must be considered, since, with the application of psychometric tests, more reliable and dependable information is obtained in subsequent clinical and statistical analyses [9, 20]. Among the main elements to be evaluated through psychometric tests is reliability, which measures the degree to which an instrument does not have random errors, this test statistically determines the consistency or correlations between test items in the questionnaire or in individual domains. Another important element that must be evaluated is the validity of the instrument, which is defined as the degree to which the instrument measures what it purports to measure. Finally, responsiveness which measures the ability of an instrument to detect changes in the domains measured, such as pre- and post-therapy [7, 10, 21].

To date, the only validated PROMs that addresses voice symptoms in Chile is the Vocal Symptom Scale (VoiSS) [22]; therefore, informal translations of the rest of the protocols in English are used, typically in clinical situations without the international criteria based on official sources such as the SAC [11]. Currently, the Chilean linguistic and cultural adaptation of the original V-RQOL protocol in English has been carried out by Contreras et al. [23] following the guidelines of the SAC, like others works through different scientific journals in the literature about adaptation and validation of different PROMs, demonstrating that the SAC guidelines are reliable in such processes [24‒26]; therefore, the objective of this research was to validate the V-RQOL vocal self-assessment questionnaire for Chilean Spanish, through translations and adaptations of the tools, demonstrating psychometric measures of validity, reliability, and sensitivity.

This study aimed to achieve the following objectives: the V-RQOL questionnaire developed by Hogikyan and Sethuraman [10] and cross-cultural adapted to the Spanish language with the Chilean population by Contreras et al. [23]. Its reliability, validity and responsiveness should be evaluated through cross-cultural adaptation, objectives that are present in this study. Determine the cut-off value for the total score that effectively discriminates between individuals with VD and those who are vocally healthy (VH).

Ethics approval for this research was obtained from the Committee of Ethics in Research at Santo Tomás University – CEC UST N° 49/2017 and N° 191.16. All participants signed consent forms before taking part in the study.

The study participants were recruited between 2017 and 2019 from four distinct locations: the Clinical Health Program at Valparaiso University (San Felipe city), Salvador Hospital, the Centro de Atención Profesional at Santo Tomás University and Hospital Clínico Universidad Católica (Santiago, city). A total of 193 individuals aged 18 and above participated in the research, comprising 103 patients with dysphonia and 90 VH subjects. Voice-related problems were reported by some participants who were asked to disclose their professions; interestingly, 22% of them considered their voices to be essential for their work, including actors, singers, professors, and nuns. The remaining 77% showed that their occupations did not heavily rely on their voices. For both groups, the ability to comprehend the questionnaire was sufficient.

There were 72 individuals with behavioral dysphonias among those who had voice complaints. Behavioral dysphonias include conditions such as functional dysphonia, aphonia, vestibular phonation, minor structural alterations, benign mass lesions like vocal nodules or polyps, and vocal fold edema. Moreover, 31 individuals displayed organic dysphonias that were identified by dysphonias that were not directly associated with voice use such as laryngeal cancer, vocal fold paralysis, spasmodic dysphonia, structural laryngeal changes linked with aging, among others. The dysphonic group (DG) had an average age of 48.5 years, while the non-dysphonic group (NDG) had an average age of 43.2 years. No individual from the NDG group exhibited vocal complaints or a diagnosis of dysphonia.

R statistical software was used to run the factanal function and the efficiency and cut-off value. For the validation process (validity, reliability, and responsiveness), STATA version 13.0 (StataCorp) was used.

The V-RQOL was subjected to exploratory factor analysis with the objective of determining the number of factors that could adequately represent the 10 variables of the questionnaire. This analysis was conducted in order to identify the contents of each item. The data were analyzed only with voice patients (N = 103). Prior to conducting the factor analysis, the suitability of the data was evaluated using the Kaiser-Meyer-Olkin (KMO) index and Bartlett’s test of sphericity to ascertain whether the correlations between variables justified the use of FA.

The psychometric properties’ translation, validation, and assessment were conducted according to the Scientific Advisory Committee of Medical Outcomes Trust (SAC) criteria. This evaluation aimed to ensure consistent measurements, examine the responsiveness and efficiency of the instrument, and determine its clinical utility and adaptability in various contexts [11].

For the linguistic and cultural adaptation of the V-RQOL questionnaire developed by Ruston et al. the Chilean Spanish translation was employed. This translation was carefully selected to ensure proper linguistic and cultural alignment with the target population in Chile [23].

To establish validity, additional criteria were incorporated alongside the established protocol, enabling a comparison with the total score. Self-assessment of vocal quality was determined using the following criteria: excellent, very good, good, fair, and poor. To simplify the analysis, the extremes of this self-assessment scale were grouped as follows: excellent voice (including excellent and very good ratings), good voice, and poor voice (encompassing fair and poor ratings). This categorization resulted in three distinct groups for analysis. Of note, a considerable proportion of both the DG (n = 103) and the HC group (n = 90) completed the questionnaire solely once [27].

The objective of this study was to assess the internal consistency of the Spanish version of the V-RQOL questionnaire in Chilean population using the Cronbach’s alpha coefficient. Additionally, test-retest reliability was measured to evaluate reproducibility, employing the Student’s t test with a significance level of p < 0.01.

To conduct the test-retest analysis, a total of 40 participants with vocal complaints were recruited. These participants completed the V-RQOL questionnaire on two separate occasions, with an interval of 2–14 days between administrations. It is worth noting that this interval was carefully chosen to ensure sufficient time had elapsed to minimize the potential influence of short-term changes in responses, thus enhancing the reliability of the results. The questionnaire was administered prior to the initiation of any intervention or treatment [22, 27, 28].

In order to evaluate changes in responses, a subset of 17 out of the 103 dysphonic participants (diagnosed by ENT specialists) underwent vocal treatment, and their responses to the V-RQOL-CL (Chilean Spanish version) were analyzed before and after the treatment. To facilitate analysis, the five response alternatives provided in the questionnaire were divided into two groups: “presence of dysphonia” and “absence of dysphonia.” These two groups were then compared to assess any significant differences.

To further validate the findings, two Speech-Language Pathologists – specialists in VDs – performed a perceptual analysis of the participants’ voice quality using the GRBAS scale. Recordings of the vowel /a/ were collected before and after treatment for each of the 17 subjects. Subsequently, the V-RQOL results were contrasted with the perceptual analysis conducted by the clinicians. This comparison allowed for a comprehensive assessment of the relationship between self-reported V-RQOL and the objective perceptual evaluation of voice quality [22, 27, 28].

To determine the sensitivity and specificity cut-off values, the V-RQOL results underwent receiver operating characteristic (ROC) curve analysis. This test can determine if the disease is present or not. The capability to correctly identify individuals with the disease was obtained from the DG. Specificity, representing the ability to correctly identify individuals without the disease, was determined by the control group (NDG).

The ROC curve graphically represents the relationship between sensitivity and specificity. It plots the trade-off between correctly identifying individuals with the disease and correctly identifying individuals without the disease. By setting various cut-off points, the ROC curve assesses the performance of a diagnostic test in terms of its ability to discriminate between the two groups [29].

The variables corresponding to the ten items of the study exhibited significant factor loadings above 0.50, indicating a clearly defined factorial structure. Furthermore, the satisfactory communalities demonstrated the validity of the three-factor model. The Kaiser-Meyer-Olkin index achieved an impressive value of 0.87, indicating excellent data adequacy for factor analysis. Furthermore, the Bartlett test of sphericity provided highly significant support (p < 0.0001) for the correlation between variables, justifying the use of factorial techniques to explore latent structures within the dataset. The three factors identified collectively explained 68.5% of the total variance. Factor 1 was primarily linked to items suggesting a physical and behavioral construct related to voice problems. Factor 2 seemed to reflect social and emotional aspects. Factor 3 captured more direct impacts on daily activities (shown in Table 1). The statistically significant χ² statistic (31.87) with a p value of 0.0227 provides further support for the sufficiency of the three-factor model, which strikes an effective balance between complexity and explanatory power. This robust framework for understanding the underlying dimensions in the dataset is therefore well-founded.

The application of analysis of variance (ANOVA) tests (shown in Table 2) revealed statistically significant differences among the emotional domain, physical domain, and total scores when comparing voices with and without disturbances. In both the dysphonic and VH groups, there were significant differences in total scores between participants with VD who rated their voices as “poor” and “good.” Conversely, participants without VD who rated their voices as “excellent/very good” and “good” reported similar average scores across all domains and total scores.

The V-ROQL questionnaire exhibited excellent reliability in both total scores and individual domain scores for the DG, as evidenced by high internal consistency assessed using Cronbach’s Alpha (shown in Table 3). All Cronbach’s Alpha coefficients reached statistical significance (p < 0.01), indicating strong internal consistency.

To assess test-retest reproducibility, the Student’s t test was employed and examined for any statistical differences between the test and retest results in each domain (shown in Table 4). Interestingly, no statistically significant differences were observed between the test and retest scores in any domain. The obtained p values remained below the significance level of 1%, confirming the high reproducibility of the V-ROQL questionnaire over time.

Significant improvements (p < 0.01) were observed in the V-RQOL scores across all subdomains as well as in the overall aggregate score following the vocal treatment. These changes were found to correspond with significant differences in pre- and posttreatment voice perceptual analyses of vocal quality (shown in Table 5).

To assess the appropriateness of parametric statistical techniques, normality tests were conducted using the Kolmogorov-Smirnov and Shapiro-Wilk methods. The results (shown in Table 6) indicated that the V-RQOL variables did not conform to Gaussian models of probability at the p = 0.005 level. Given this departure from normality, we chose to utilize the nonparametric Mann-Whitney test (shown in Table 7) for the ROC curve analysis. The purpose of this analysis was to distinguish the presence or absence of dysphonia.

The ROC curve (shown in Fig. 1) exhibited a sensitivity of 79% (indicating the detection of true positives) and a specificity of 90% (indicating the detection of true negatives). The DG had cut-off scores equal to or greater than 92.5 points, while the VH participants had cut-off scores below 92.5 points (shown in Table 8).

Main aspects of voice assessment used in both clinic and research include: perceptual, acoustic, aerodynamic, visuo-perceptual, and self-assessment procedures by PROMs [30‒32]. In clinic, the PROMs of voice by using a variety of validated questionnaires are considered a primary way for identifying the impact of voice problems in different aspects of the individual’s life [10]. The creation and validation of PROMs that aim at measuring the impact of a specific health problem on quality of life have become a relevant aspect of different areas in the medicine field, and over the two last decades, several tools of treatment outcome and quality of life measure have been developed [12, 33].

The present study aimed at validating the V-RQOL PROMs questionnaire for Chilean Spanish, through translations and adaptations of the tools, demonstrating psychometric measures of validity, reliability and sensitivity. The V-RQOL questionnaire was selected to be culturally/linguistic adapted and validated into Chilean Spanish because it is an instrument specifically created to assess the impact of a voice problem on the individual’s life, something that is still lacking for our language and culture.

This study demonstrated that the Chilean Spanish version of the V-RQOL is a stable and applicable measurement tool. The results of our study are consistent with similar studies [10, 15‒20, 34‒36]. The present version of the V-RQOL questionnaire may function as an important part of the standard assessment process of people with voice complaints and may be used as an outcome of treatment efficacy in clinical trials.

An important application of quality of life instruments is the evaluation of treatment efficacy. Knowledge about the V-RQOL measures responsiveness to change after voice treatment (voice therapy, medical treatment, or vocal folds surgery) is crucial in this context. In the present study, we included pre-post voice therapy measures for the Chilean version. Future research should include the present validation as an important outcome of voice treatment procedures. Specifically, studies on different therapy programs for specific types of disorders should be carried out for measuring V-RQOL in VD patients. The V-RQOL is not an instrument specific to any category of VD [10, 20]. In present study, comparisons were not made by categorizing participants with VDs based on their diagnosis or by selecting a specific disease group while examining the reliability and validity of the V-RQOL.

The psychometric properties of the Chilean Spanish version of the V-RQOL scale were evaluated in a sample of 103 individuals with voice problems and 90 healthy controls. The factor analysis conducted on the Chilean version of the V-RQOL revealed that the PROMs are structured in three factors, as indicated by the correlations between items. Factor 1 includes items (1, 2, 3, 6, 7, and 9) of the physical domain, with items 6 and 9 highlighted for their association with perception of being understood. Factor 2 groups items (8 and 10), which relate to socioemotional aspects and the use of voice in social contexts. Finally, factor 3 is composed of items (4, 5, 7), where items 4 and 5 belong to the socioemotional domain and address psychological issues such as anxiety and depression, while item 7, from the physical domain, is related to the use of voice in the work environment. This suggests a possible connection between work and emotional and psychological factors, similar to what was observed in items 4 and 5. The other validations did not include a factor analysis in the process of validating the psychometric properties [16, 18, 19]. However, one study evaluated the subdomains of the Brazilian version of the V-RQOL by factor analysis, identifying a single factor, which differs from our study [37].

The following findings showed that all items included in the questionnaire significantly discriminate between normal and disordered voices. Analysis of variance tests showed significant differences between the emotional domain, the physical domain and the total score, considering voices with and without disturbances. Moreover, results from the ROC curve showed 79% sensitivity (i.e., detection of a true positive) and 90% specificity (i.e., detection of a true negative). The Chilean Spanish version of V-RQOL demonstrated to be a sensitive tool in identifying VD. Total scores of V-RQOL questionnaire may vary from 0 to 100, with 0 reflecting a very poor V-RQOL and 100 an excellent one [10]. Therefore, lower scores in this PROMs questionnaire mean poorer voice-related quality of life. Our results revealed that individuals with VD have poorer V-RQOL compared with people with normal voice.

The Chilean Spanish version of the V-RQOL performed well in the selected population. Measures of reliability and reproducibility are strong for the different V-RQOL scores. Specifically, no significant differences were found when comparing test and retest conditions for emotional, physical, and total scores. Comparable findings have been found by studies in other languages [10, 16, 18].

High internal consistency as assessed using Cronbach’s alpha was also observed in the present validation study. All coefficients were statistically significant (p < 0.01). The Cronbach alpha coefficient values were 0.90 and 0.84 for the physical and emotional domains, respectively. The Cronbach alpha coefficient was 0.89 for the Chilean Spanish version of V-RQOL total score. These data are in line with the results of other validation studies carried out in other languages [10, 16, 18, 38, 39].

Furthermore, validity of V-RQOL was demonstrated by the robust relationship between self-assessment of voice and mean V-RQOL scores and by the great difference between the results of voice and nonvoice patients. The total scores for the participants with VD who rated their voices as “poor” and “good” were significantly different in both the dysphonic and VH groups. While participants without VD rated their voices as “excellent/very good” and “good” reported average in all domains and total scores similar between both rated.

Since one of the uses of the V-RQOL is to measure possible effects of voice treatment in quality of life, the responsiveness is also a relevant aspect to be considered in the validation process. The Chilean Spanish V-RQOL version demonstrated to be able to assess voice treatment outcomes. This was observed because it is sensitive to the positive changes caused by voice therapy. In each subdomain (emotional and physical) and in the total, the V-RQOL scores responded to treatment. These changes corresponded to significant differences (p < 0.01) in pre- and posttreatment voice perceptual analyses of vocal quality. The V-RQOL tool has proven responsive to change by its original version in English [10], and in the Brazilian [16], Persian [38], and Danish [40] adaptations. Our data showed a difference of 30.44 between pre- and posttreatment of total score. This difference is larger compared to previous V-RQOL validation studies. Moradi [38] reported 16.6, while Gasparini and Behlau [16] showed a difference of 12.4 between pre- and posttreatment of total score.

In general terms, the assessment of quality of life is carried out to evaluate treatment outcomes in clinical settings and demonstrate the effectiveness in research. This should promote an improved professional practice in the health field. The Chilean version of V-RQOL will allow the voice clinicians and researchers to provide a deeper evaluation of the patients’ status, make more accurate diagnosis of Chilean Spanish-speaking patients with VDs, explore the impact of dysphonia on different aspects of quality of life, and detect the strategies used to improve styles of life of people with voice complaints.

By following the recommended methodology established by the SAC committee, the validation process of the V-RQOL questionnaire for assessing vocal self-assessment in Chilean Spanish was successfully accomplished. Furthermore, the determination of a specific cut-off value for the total score was achieved, enabling the differentiation between individuals with VD and those who are VH.

The authors extend their heartfelt gratitude to the late Dr. Pedro Badía MD, a cherished colleague, mentor, and companion, for his invaluable contributions.

The research received approval from the Ethics Committee of Santo Tomás University (CEC UST) under the reference numbers 49/2017 and 191.16, and adhered to the ethical guidelines outlined in the Declaration of Helsinki. Prior to their involvement, all participants duly furnished written informed consent, in compliance with the ethical directives set forth by the CEC UST.

The authors have no conflicts of interest to declare.

The first author was supported by Agencia Chilena de Investigación y Desarrollo (ANID) (Beca Chile para Estudios en el Extranjero 2020, folio 72210019).

Francisco Contreras-Ruston, SLP of Speech-Language Pathology and Audiology Department – Universidad de Valparaíso, MSc o Behavioral and Cognitive Research at the University of Barcelona, PhD candidate in Brain, Behavior and Cognition de la Universidad de Barcelona, and Translational Cognitive Neuroscience at the Faculty of Psychology and Neuroscience, Maastricht University, Research Project contribution: conducted research, collected data, data analysis, and drafted text. Andrés Rosenbaum Fuentes, MSc of Health Sciences Research, MD of Departamento Otorrinolaringología, Pontificia Universidad Católica de Chile, Research Project contribution: data analysis and reviewed text. Karol Acevedo Encalada, SLP of Department of Rehabilitation Science, Universidad San Sebastian, Research Project contribution; Nury Gonzalez, SLP of Speech-Language Pathology and Audiology Department – Universidad de Valparaíso, Research Project contribution; and Lukas Salfate Velásquez, SLP of Departamento Otorrinolaringología, Hospital Salvador. Research Project contribution: collected data and drafted text. Norma León Meneses, SLP of Departamento de Ciencias de la Salud, Pontificia Universidad Católica de Chile, Research Project contribution: collected data and reviewed text. Carla Napolitano MD of Departamento Otorrinolaringología, Pontificia Universidad Católica de Chile, Research Project contribution: reviewed text and manuscript language editing. Marco Guzman SLP, PhD of Department of Communication Sciences, Universidad de los Andes, Research Project contribution: collected data, drafted text, and reviewed text.

All study data are included. Contact corresponding author for inquiries.