Translation and validation of the Voice Handicap Index in Tamil language

Abstract

Voice Handicap Index (VHI) is the most commonly used tool to assess the quality of life in voice disordered patients. A validated Tamil language version of VHI is not developed yet. Hence, this study was undertaken to translate and validate the Voice Handicap Index in Tamil language, which can potentially benefit Tamil speaking patients with voice disorders. This was a translation and tool validation study done at a tertiary care teaching hospital using an analytical, follow up design according to WHO guidelines. Dysphonic (Group A) and Normal (Group B) respondents were purposively invited to fill a self administered VHI-Tamil questionnaire. The content validity, response process validity, internal consistency and clinical validity of the questionnaire was calculated using appropriate statistical analysis methods using SPSS version 24.0 software. Out of 117 respondents, 61 respondents were dysphonic (Group A) and 56 respondents were normal (Group B). Overall index for content validation was over 0.84, and response process validation was 1.00 in all domains. Overall internal consistency was excellent, with Cronbach’s alpha of 0.993. Excellent test–retest reliability was identified using the Spearman rank correlation coefficient (r = 0.96; p-value < 0.001). For clinical validity, a statistically significant difference between the dysphonic and the normal groups, for the overall VHI-Tamil scores and each of the three domain scores was noted. VHI-Tamil was found to be reliable and valid for assessing the quality of life in patients with voice disorders. It can be recommended for future use among Tamil speaking population.

Introduction

Voice is an important tool for communication. Voice disorders can adversely affect the quality of life of an individual. Voice disorders are commonly encountered in clinical practice, being the most common speech and language disorders, affecting approximately 6% of children under 14 years of age, and 3–9% of the adult population [1].

In the year 1997, Barbara H.Jacobson et al. developed and validated Voice Handicap Index (VHI) in English for the first time to quantify the psychosocial consequences of voice disorders. This tool consists of 30 items that are equally distributed (10 each) over the three domains: functional, physical, and emotional [2]. The VHI is considered a most relevant, valid, reliable, patient-friendly, and versatile tool available to assess the voice- related quality of life at present [3, 4].

The original VHI instrument is in English, since then, many non-English speaking populations have translated it to different languages all over the world [5]. It has also been translated into some Indian languages like Hindi, Kannada, Malayalam, and Bangla [6–9]. But translation version is not yet available in Tamil, which is an Indian language spoken by a large population (5.9%) in India. Tamils, with a population of around 76 million in the world, are one of the largest groups in the modern world [10, 11].

With tremendous growth in the field of laryngology in the recent years, a validated VHI-Tamil tool can greatly help its use in voice disordered patients of Tamil speaking population and help in assessing the changes in their quality of life in relation to various treatment modalities. Hence, his study was undertaken to translate the Voice Handicap Index (VHI) and validate it for usage in the Tamil language (VHI-Tamil).

Materials and Methods

Study Setting and Design

An analytical, follow up design was used for this tool validation study, conducted in the Department of Otorhinolaryngology of a tertiary care teaching hospital, located in the rural part of Puducherry, India from August to October 2020.

Study Participants and Sampling

There were two groups of study participant’s namely dysphonic and normal group. All participants were above 18 years of age and were able to understand and speak Tamil language well. Among them those with voice disorders formed the dysphonic group of respondents (Group A). Medical college students, staffs and hospital visitors without any voice complaints formed the normal group of respondents (Group B). Participants with stridor or shortness of breath were excluded. The purposive sampling method was used based on the presence or absence of voice disorders to select the participants into the study.

Data Collection Tool

The VHI tool had two major sections. The first section of the questionnaire included socio-demographic details of the respondents like age, sex, educational level, occupation, place of residence. The second section contained a set of translated questions in three parts, for Functional (F), Physical (P), and Emotional (E) domains of voice, each with 10 questions. Additionally, self-perceived overall severity of voice problem (0- normal, 1- mild, 2-moderate, 3- severe) was also collected from the respondents. The respondents were asked to tick the boxes that indicate how frequently he/she had the experience that was mentioned in the tool under various domains and was asked to rate it in the Likert scale of 0- 4 (0-never, 1-almost never, 2-sometimes, 3-almost always, 4-always). It was given as (Fig. 1).

Fig.1.

Final version of translated VHI-Tamil tool

Steps Involved in the Validation and Reliability Assessment

Institutional Ethics Committee approval was obtained before initiation of the study (IEC No: EC/48/2020). After obtaining written, informed consent, the respondents were invited to complete a self-administered VHI-Tamil questionnaire. Only dysphonic patients included in the study were instructed to come again to OPD after 10–12 days to assess the test–retest reliability of the questionnaire. There were three major methods involved in assessing validity and reliability of translated tool. They were (i) Translation and its content validation, (ii) Reliability measurement and (iii) Clinical validity assessment. Each of these were explained below.

Translation and Validation Process

The translation process was started after obtaining permission from the original VHI instrument developer, Dr. Barbara H. Jacobson through email. Translation of tool and its content validity remains the major concern of the language validation process and hence it was carried out rigorously following the WHO guidelines on “Process of translation and adaptation of instruments”, using translation and back-translation method [12, 13]. An Expert Committee was formed that included two translators, study investigators, and two language experts, one each for English and Tamil language.

It had four major steps namely (a) forward translation, (b) backward translation, (c) content validity assessment and (d) response process validity assessment. Of these four steps the first two steps were inter-related and were explained below. The flow of events is depicted in (Fig. 2).

Fig.2.

Steps taken in the translation and validation process of VHI-Tamil

Forward Translation

The forward translation involved the translation of English VHI into the Tamil language by two independent bilingual translators. One of them was an ENT subject expert, well versed with the medical terminologies, and the content of the VHI questionnaire with a long clinical experience in the field of ENT for more than 30 years. Another bilingual translator was familiar with the slang, jargon, informal phrases used in the Tamil language. This is especially important in cross-cultural translational processes. The translated questionnaire was presented to the expert committee, to find out the points of ambiguities in comparison with the original English VHI questionnaire. Ambiguities and discrepancies were discussed and resolved by the committee-consensus approach to generate a reconciled VHI-Tamil questionnaire with 30 questions, with 10 questions in each of the three domains, Functional, Physical and Emotional.

Backward Translation

After review and corrections, the VHI-Tamil was back-translated into source language (English) by two independent bilingual translators, with proficiency in English language, working as English Professors were involved in the back-translation process. They were blinded and unaware of the original VHI questionnaire. The two different back-translated English versions of VHI were discussed in the expert committee and reconciled. It was compared to the original VHI and any ambiguities and discrepancies were resolved after discussion by consensus. The VHI-Tamil was revised and modified.

After the translation process we calculated the content validity index and then proceeded with response process validity.

Content Validity Assessment

The translated VHI-Tamil questionnaire was subsequently pre-tested for content validity by another five bilingual subject experts, well experienced in the field of ENT, knowledgeable about the content of the questionnaire. Experts were asked to focus on various factors like Tamil language validity for wordings, style, sentence formation, and communication effect. All the experts were asked to rate each of 30 items of the translated VHI-Tamil as “Agree” or “Disagree” using a binary scoring system. Content Validity Index (CVI) was calculated depending upon the opinions and feedback obtained. The CVI was calculated as follows: I-CVI (Item-level Content Validity Index) for individual items, where the number of experts answering “Agree” was divided by the total number of experts involved, and S-CVI( Scale-level content validity index) for overall scale, where I-CVI for all the items were added. To calculate S-CVI, two calculations were made: S-CVI/Ave (Averaging calculation method) and S-CVI/UA (Universal Agreement calculation method). S-CVI/Ave was calculated from total I-CVI value divided by the total number of items; whereas S-CVI/UA was calculated by adding the number of items that had 100% agreement and dividing that by the total number of items [14, 15].

Response Process Validity Assessment

Response process validity can be defined as “Do respondents understand the questions to mean what we intend them to mean?” [16] It was carried out on 12 target group individuals with voice disorders after obtaining informed consent, and validity of the domains was determined by technically untrained observers. Similar to content validity, calculation of Response process validation also included I-CVI (Item-level Content Validity Index), Scale-level content validity index (S-CVI), S-CVI/UA, S-CVI/Ave, the average proportion of items judged relevant across 12 respondents.

Reliability Assessment

Reliability is the reproducibility of the data collected from the tool. It was assessed by two means. They were assessment of internal consistency and test–retest reliability. The internal consistency of VHI-Tamil was assessed using Cronbach’s alpha coefficient. Pearson’s correlation coefficient was used to correlate between each item and each domain with total VHI scores that was inbuilt of the internal consistency. To evaluate the test–retest reliability of the VHI-Tamil, those dysphonic patients who came for follow up were asked to fill the VHI-Tamil questionnaires again after two weeks of follow up.

Consequence/ Clinical Validity Assessment

This was carried out under the two below mentioned methods.

• Agreement between total VHI-Tamil scores and the patients’ self-perceived severity of dysphonia were made by Weighted Kappa agreement.

• Comparison of the overall VHI-Tamil scores and scores obtained under each of the three domains (functional, Physical, and Emotional) between dysphonic and the normal individuals were done separately.

Statistical Analysis

The VHI-Tamil scores thus obtained were statistically analyzed. The data were entered digitally using EpiInfo7 (version 7.2.2.6) software package and it was further exported to SPSS version 24.0 for analysis. Internal consistency was checked using Cronbach’s alpha and Spearman correlation was used for assessing test–retest reliability. Comparison of mean scores of various domains between patients and normal individuals was carried out using independent sample t-test. All tests were two-tailed and a p-value if < 0.05 is considered statistically significant.

Results

Among a total of 117 respondents, Group A (Dysphonic group) included 61 (52.1%) respondents and Group B (Control group) included 56 (47.9%) respondents. In both groups majority of the respondents belonged to the age group of 31–40 years (34.4% in Group A, and 50% in Group B). Females were the major respondents in both groups (62.3% in Group A and 64.3% in group B). The majority of the respondents hailed from rural areas i.e. 36 (59%) in Group A and 35 (62.5%) from Group B. Majority of the respondents in both group were Homemakers by occupation (Group A—34.4% and Group B—26.8%). Among dysphonic patients 23 (37.7%) perceived their voice disorder as severe, 22 (36.1%) as moderate, and the rest as mild 16 (26.2%).

Content validation produced an overall validity index of over 0.84, with a universal agreement index (S-CVI/UA) of 0.88 for the Functional (F) domain, 0.87 for the Physical (P) domain, 0.84 for the Emotional (E) domain. An average index (S-CVI/Ave) of 0.98 for the Functional (F) domain, 0.96 for the Physical (P) domain, 0.98 for the Emotional (E) domain were obtained (Table 1). Response process validity was calculated similarly to content validity. All items showed an index of 1.00 for all the questions in all the domains.

Table 1.

Details of content validity measurements of voice handicap index–tamil version

Measurements	Domains of voice handicap index–tamil version
	F (Functional)	P (Physical)	E (Emotional)
S-CVI/Ave	0.98	0.96	0.98
S-CVI/UA	0.88	0.87	0.85
Average proportion of items	0.98	0.96	0.98

Note S−CVI (scale−level content validity index), UA (universal agreement calculation method) and AVE (averaging calculation method)

Overall internal consistency was excellent, with a Cronbach’s alpha coefficient of 0.993 (Table 2). Excellent test–retest reliability was identified for the 28 subjects using the Spearman rank correlation coefficient was obtained (r = 0.96; p < 0.001).

Table 2.

Details of the internal consistency of VHI-Tamil version using Cronbach’s alpha coefficient method

Items	Corrected item—total correlation
F1. My voice makes it difficult for people to hear me	0.908
F2. People have difficulty understanding me in a noisy room	0.912
F3. My family has difficulty hearing me when I call them throughout the house	0.931
F4. I use the phone less often than I would like to	0.851
F5. I tend to avoid groups of people because of my voice	0.919
F6. I speak with friends, neighbors, or relatives less often because of my voice	0.908
F7. People ask me to repeat myself when speaking face-to-face	0.906
F8. My voice difficulties restrict my personal and social life	0.930
F9. I feel left out of conversations because of my voice	0.914
F10. My voice problem causes me to lose income	0.668
P1. I run out of air when I talk	0.852
P2. The sound of my voice varies throughout the day	0.842
P3. People ask, “What’s wrong with your voice?”	0.933
P4. My voice sounds creaky and dry	0.930
P5. I feel as though I have to strain to produce voice	0.928
P6. The clarity of my voice is unpredictable	0.894
P7. I try to change my voice to sound different	0.892
P8. I use a great deal of effort to speak	0.937
P9. My voice is worse in the evening	0.917
P10. My voice “gives out” on me in the middle of speaking	0.892
E1. I am tense when talking to others because of my voice	0.948
E2. People seem irritated with my voice	0.930
E3. I find other people don’t understand my voice problem	0.947
E4. My voice problem upsets me	0.944
E5. I am less outgoing because of my voice problem	0.946
E6. My voice makes me feels handicapped	0.956
E7. I feel annoyed when people ask me to repeat	0.954
E8. I feel embarrassed when people ask me to repeat	0.962
E9. My voice makes me feel incompetent	0.929
E10. I am ashamed of my voice problem	0.940
Cronbach’s alpha	0.993

For clinical validity, there was a significant correlation between the total VHI-Tamil scores and the patients’ self-perceived dysphonia. Weighted Kappa agreement between the two was found out to be 0.86 (86.1%), at a Confidence interval of 95%, p-value < 0.001. The VHI-Tamil showed a statistically significant difference (p-value < 0.001) between the dysphonic and the normal groups, for the overall VHI-Tamil scores and each of the functional, physical, and emotional domains scores separately (Table 3). The mean total VHI-Tamil score for Group A (Dysphonic) and Group B (Normal) were 64.2 (SD = 33.5) and 0.50 (SD = 1.51) respectively. Mean scores for all the three domains Functional, Physical, and Emotional were almost equal in the dysphonic group.

Table 3.

Details of clinical validity of VHI-Tamil version compared among Group A and group B respondents

Domains	Group A (Dysphonic)		Group B (Normal)		p value#
	Mean	SD	Mean	SD
F (Functional)	21.738	10.5260	.286	1.0395	< 0.001*
P (Physical)	21.131	11.6712	.089	.4380	< 0.001*
E (Emotional)	21.377	12.6269	.125	.4695	< 0.001*
Total score	64.213	33.4665	.500	1.5136	< 0.001*

SD Standard Deviation, # p value based on independent sample t test, * statistically significant (p<0.05)

Discussion

In this study, we observed that majority of the dysphonic patients were females (62.3%) and most of the patients belonged to the age group of 31–40 years (34.4%). Most of the dysphonic patients were homemakers (34.4%) and hailing from rural areas (59%). The content validity (scores over 0.84 in all domains) and response process validity (scores over 1 in all domains) for the translated VHI-Tamil tool was high. The study showed an excellent internal consistency with Cronbach’s alpha of 0.993.

Prevalence of voice disorders is higher in adult females than in adult males, with a reported ratio of 1.5:1.0 [17]. Similar results were observed in our study also, in Group A (Dysphonic), the majority of the respondents were females 38(62.3%), rest were males 23(37.7%) with a female to male ratio of 1.65:1. Prevalence has been reported to be higher in elderly adults, with estimates ranging from 4.8% to 29.1% in population-based studies. [18] In our study, among the dysphonic group, Group A, majority of respondents belonged to 31–40 years of age group, i.e. 21 (34.4%), followed by 18–30 years of age group, i.e. 19 (31.1%).

According to previous studies, the risk of voice disorders depending on occupation appears to be more in teachers, manufacturing/factory workers, salespersons, and singers. [19] In our study, the majority of the dysphonic patients were homemakers 21(34.4%), followed by teachers 10(16.4%), and nurses 6(9.8%). In our study, the majority of the respondents hailed from rural areas i.e. 36(59%) in Group A and 35(62.5%) from Group B, rest from urban areas i.e. 25(41%) in Group A and 21(37.5%) in Group B. But there was no statistically significant difference in scores between rural and urban area patients.

The validity of a tool is the ability to measure what it seeks to quantify. The content validation of the final translated questionnaire VHI-Tamil was good with a score of more than 0.84 in all the domains. Response process validity gave a score of 1 in all the domains. Overall internal consistency was excellent, with a Cronbach’s alpha of 0.993. Excellent test–retest reliability was identified for the 28 subjects who completed the VHI-Tamil twice after a period of 12 to 14 days using the Spearman rank correlation coefficient (r = 0.96; p-value < 0.001).

For clinical validity testing, two parameters were used. The first parameter, there was a significant correlation between the total VHI-Tamil scores and the patients’ self-perceived dysphonia. Weighted Kappa agreement between the two was found out to be 0.86 (86.1%), at a Confidence interval of 95%, (p-value < 0.001), which indicates that the VHI-Tamil captures the degree of severity of handicap felt by the patient.

Another parameter, the VHI-Tamil showed a statistically significant difference between the dysphonic and the control groups, for the overall VHI-Tamil scores and each of the functional, physical, and emotional domains scores separately. (p-value < 0.001). This shows that VHI-Tamil can differentiate the persons with dyphonia from the normal persons, thus making it a valid tool. These results were similar to previous studies of translation and validation of VHI in various languages [5].

In our study, the dysphonic group (Group A) had comparable mean score in all the three domains, in contrast to the results reported in some of the studies in the past.[20–23] These differences could be attributed to the differences in the age, gender composition of the study populations and also the various patients with different causes and severities of voice disorders included in the studies.

The results of this study indicate that the translated questionnaire VHI-Tamil is a valid and reliable tool that can be used in the assessment of self-perceived severity of voice problems in the Tamil speaking population. It can help the clinicians to manage voice disorders in these populations and also help in modifying the treatment strategies based on the response to treatment. However, the Tamil language is widely distributed in the state of Tamil Nadu and some neighbouring states, with variations in the dialects, grammar, and word meanings. Further studies including a larger number of populations from different regions are recommended to help in improvising this tool. The major strengths of this study are, (i) It being one of the first translated and validated VHI tool in Tamil language, (ii) Rigorous validation and reliability assessment methods were followed.

Conclusion and Recommendation

We have developed this VHI-Tamil version tool based on theoretical framework and rigorous methodology. It is a reliable and valid tool for assessing the quality of life in patients with voice disorders. It can be recommended for future use among Tamil speaking populations with voice disorders.

Funding

No funds, grants, or other support was received.

Compliance with Ethical Standards

Conflict of interest

The authors have no conflicts of interest to declare that are relevant to the content of this article

Ethical Approval

This study involved human participants and was approved by the institutional ethics committee [Sri Manakula Vinayagar Medical College and Hospital-Ethics committee (Human studies), No.EC/48/2020] and the study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments.

Informed Consent

Informed consent was obtained from all individual participants included in the study.

Availability

Availability of data and material: Not available.

Code Availability

Not applicable