Correlation of the Voice Handicap Index-10 (VHI-10) and Voice-Related Quality of Life (V-RQOL) in Patients With Dysphonia Jonathan J. Romak, Diana M. Orbelo, Nicolas E. Maragos, and Dale C. Ekbom, Rochester, Minnesota Summary: Objectives. This study examines the correlation between two voice-specific patient-reported outcome measures: the Voice Handicap Index-10 (VHI-10) and Voice-Related Quality of Life (V-RQOL). Study Design. Retrospective chart review. Participants. Eight hundred four patients presenting to our voice clinic between May 2009 and August 2011. All patients completed the VHI-10 and V-RQOL in a single sitting. Methods. Correlation between the two scales was examined using Spearman rank analysis. Calculated VHI-10 score was derived from V-RQOL score by direct conversion equation and compared with measured VHI-10 score. Receiver Operating Characteristic (ROC) curves were derived for diagnostic groups. Results. Spearman correlation coefficient between the VHI-10 and V-RQOL was 0.91 (P < 0.0001). VHI-10 and V-RQOL scores were also significantly correlated among diagnostic categories. Calculated and measured VHI-10 scores were significantly different both for individuals and overall. Area under the curve (AUC) values from ROC curves were significantly different for the presbyphonia (V-RQOL AUC ¼ 0.586 [standard error, SE ± 0.033]; VHI-10 AUC ¼ 0.530 [SE ± 0.031]; P ¼ 0.0014) and muscle tension dysphonia (V-RQOL AUC ¼ 0.536 [SE ± 0.026]; VHI-10 AUC ¼ 0.508 [SE ± 0.26]; P ¼ 0.018) groups, with the V-RQOL showing relatively greater sensitivity. Conclusions. The VHI-10 and V-RQOL are highly correlated. However, VHI-10 score cannot be calculated from V-RQOL score using the tested equation. The V-RQOL may be more sensitive than the VHI-10 in detecting the impact of presbyphonia and muscle tension dysphonia. Key Words: Voice Handicap Index-10–Voice-Related Quality of Life–Patient-reported outcome measure–Dysphonia– Presbyphonia–Muscle tension dysphonia.
INTRODUCTION Evaluation of the dysphonic patient is complex, requiring assessment and integration of multiple variables. A detailed history, physical examination, videostroboscopy, perceptual voice evaluation, and measurement of acoustic and aerodynamic parameters all provide important information. However, the degree to which a voice disorder impacts a patient’s functional, physical, emotional, and social well-being is highly variable and based on many factors unique to the individual. Therefore, delineating the impact of a voice disorder on the patient’s overall quality of life is an essential part of the evaluation and may in fact determine the course of treatment. Several dysphoniaspecific patient-reported outcome measures (PROMs) have been developed for this purpose.1 The Voice-Related Quality of Life (V-RQOL), the Voice Handicap Index (VHI), and the Voice Handicap Index-10 (VHI-10) are commonly used PROMs at voice centers in the United States.2–4
Accepted for publication July 22, 2013. This publication was supported by grant number UL1 TR000135 from the National Center for Advancing Translational Sciences. Presented as a podium presentation at the Voice Foundation’s 41st Annual Symposium: Care of the Professional Voice; May 30 to June 3, 2012; Philadelphia, Pennsylvania. From the Department of Otorhinolaryngology - Head and Neck Surgery, Mayo Clinic, Rochester, Minnesota. Address correspondence and reprint requests to Diana M. Orbelo, Department of Otorhinolaryngology - Head and Neck Surgery, Mayo Clinic, 200 First Street SW, Rochester, MN 55905. E-mail: [email protected] Journal of Voice, Vol. 28, No. 2, pp. 237-240 0892-1997/$36.00 Ó 2014 The Voice Foundation http://dx.doi.org/10.1016/j.jvoice.2013.07.015
Developed by Hogikyan and Sethuraman in 1999, the V-RQOL consists of 10 questions pertaining to the physical functioning and social-emotional domains. A conversion equation is used to generate the final score, with higher scores indicating better quality of life.2 The VHI, developed by Jacobson et al in 1997, is a 30-question instrument dealing with the functional, physical, and emotional domains. Response options are 0–4 with higher numbers indicating greater impairment.4 Rosen et al in 2004 distilled the VHI down to the 10 most clinically valuable items as determined by clinical consensus and item analysis. The abridged version, the VHI-10, was then compared with the parent questionnaire and found to be highly correlated.3 Portone et al5 have demonstrated that the VHI and V-RQOL are highly correlated. However, to our knowledge, no studies have compared the VHI-10 and V-RQOL in a large group of patients with dysphonia. This study compares the VHI-10 and V-RQOL and investigates the possibility of direct conversion between scores in a large cohort of voice patients. PARTICIPANTS Approval for this study was granted by our Institutional Review Board, Office of Human Research Protection, Mayo Clinic, Rochester, MN (PR10-007714-02). Records of all patients presenting to our voice clinic between May of 2009 and August of 2011 who completed both the VHI-10 and V-RQOL in a single sitting were reviewed. The instruments were completed by the patient independently, before the clinical encounter. If multiple sets of surveys had been completed by a single patient due to
238
Journal of Voice, Vol. 28, No. 2, 2014
multiple visits, only the pair of surveys from the initial encounter was included. Only patients with a clear diagnosis recorded by the treating laryngologist on the day the surveys were completed were included. In patients with multiple diagnoses, the primary diagnosis was used. Patients without voice-related complaints were excluded. Eight hundred four patients met criteria and were included in the study; 59.5% (478 patients) were female and 40.5% (326 patients) were male. Average age was 58.6 years (standard deviation [SD] ± 16.3). The following diagnostic categories were identified: malignant laryngeal lesions, benign laryngeal lesions, laryngeal papillomatosis, presbyphonia, inflammatory/infectious, vocal fold paralysis/paresis, neurologic disorders, and muscle tension dysphonia. METHODS Diagnosis, VHI-10 score, and V-RQOL score were recorded for each patient. After Portone et al,5 a direct linear conversion equation was derived based on the score ranges. In our case, the anchors of 0–100 and 0–40 for the V-RQOL and VHI-10, respectively, resulted in the equation (y ¼ 0.4(x) + 40) that was used to calculate VHI-10 (y) from V- RQOL (x). Statistical analysis was performed using JMP 9.0.3 (SAS Institute, Inc, Cary, NC). Spearman correlation coefficients were derived for measured VHI-10 and V-RQOL scores both overall and by diagnosis. The difference between calculated and measured VHI-10 score was analyzed using a matchedpairs t test. Analysis of variance (ANOVA) was also used to compare calculated VHI-10 to measured VHI-10. Exploratory receiver operating characteristic (ROC) curves were generated (MedCalc 12.1.3; MedCalc Software, Ostend, Belgium) for the diagnostic groups and compared with all other diagnostic groups combined for V-RQOL and VHI-10 to assess instrument sensitivity and specificity for each diagnostic group. RESULTS Spearman correlation coefficient comparing overall VHI-10 and V-RQOL scores was 0.91 (P < 0.0001). Mean measured VHI-10 score was 19.16 (SD ± 10.37), mean calculated VHI-10 score was 15.4 (SD ± 10.31), and mean V-RQOL score was 61.5 (SD ± 25.8). Measured VHI-10 and V-RQOL scores were compared via linear regression analysis (Figure 1) yielding the equation y ¼ 0.36x + 41.5, R2 ¼ 0.82 (P < 0.0001). Spearman correlation analysis was also performed by diagnostic category (Table 1). All Spearman coefficients were statistically significant. Matched-pairs t test was performed comparing measured and predicted VHI-10 score. The mean difference was 3.76, which was significant (P < 0.001). Calculated VHI-10 was less than measured VHI-10 in 612 patients (74%). Regression analysis was also used to compare measured and calculated VHI-10 score (Figure 2). ANOVA was performed with P < 0.0001. The area under the curve (AUC) values from ROC curves were significantly different for the presbyphonia (0.586 [standard error, SE ± 0.033] for V-RQOL and 0.530 [SE ± 0.031]
FIGURE 1. Bivariate analysis of measured VHI-10 and V-RQOL scores. Regression line equation y ¼ 0.36x + 41.5, R2 ¼ 0.82 (P < 0.0001). for VHI-10; P ¼ 0.0014) and muscle tension dysphonia (0.536 [SE ± 0.026] for V-RQOL and 0.508 [SE ± 0.26] for VHI-10; P ¼ 0.018) groups, with the V-RQOL showing greater sensitivity in both cases (Figure 3). The AUC values were not significantly different in the other diagnostic groups. DISCUSSION At our institution, both the VHI-10 and V-RQOL are used in the assessment of voice specific quality of life. The aim of this investigation was to determine the degree of correlation between these two scales, with the goal of limiting the quality of life assessment to a single survey. Portone et al previously demonstrated the correlation of VHI and V-RQOL. The same study also evaluated the possibility of using a direct conversion equation to convert V-RQOL score to VHI score, so that studies using these two scales could be directly compared. In their sample, the average of scores generated with the conversion equation did not differ significantly from that of directly measured scores, but the same values for individuals were significantly different.5 Because the VHI-10 has been shown to be essentially equivalent to the VHI and the VHI has been shown to be correlated
TABLE 1. Spearman Correlation Coefficients for Each Diagnostic Category Diagnosis Malignancy Leukoplakia Benign lesion Papilloma Bowing/presbyphonia Inflammatory/infectious Paralysis/paresis Neurologic Muscle tension
Spearman (r)
n
0.94 0.87 0.91 0.95 0.76 0.86 0.93 0.86 0.89
33 19 138 21 54 86 195 100 147
Jonathan J. Romak, et al
Correlation of VHI-10 and V-RQOL
FIGURE 2. Regression analysis of measured versus predicted VHI10. Regression line y ¼ 0.91x 5.13, R2 ¼ 0.82. with the V-RQOL, we hypothesized that the VHI-10 and the V-RQOL would be correlated.3,5 To answer this question, methods similar to that of Portone et al5 were applied to our database. VHI-10 and V-RQOL were found to be highly correlated. However, the derived conversion equation did not consistently predict VHI-10 score based on V-RQOL score. The fact that the conversion did not accurately predict VHI-10 score does not refute the correlation of the two surveys, only the ability to derive one score from the other. The VHI-10 and V-RQOL are also significantly correlated within all the examined diagnostic subgroups. Interestingly, in our study, the correlation was lower for the presbyphonia group (Spearman coefficient ¼ 0.76) compared with other diagnostic groups. This suggests that the VHI-10 and V-RQOL may be less comparable for patients with presbyphonia compared with other groups. This finding differs from that of the Portone et al group, who reported a strong correlation (-0.85) between VHI and V-RQOL in their ‘‘atrophy’’ group. Possible reasons for differences include sample size (n ¼ 54 in our study versus n ¼ 22 in the study by Portone et al). Also, presbyphonia is a complex problem that often accompanies other comorbidities affecting voice specific and overall quality of life.5 Therefore, heterogeneity within the groups could be contributing to a difference in correlation. Given the known correlation between the VHI-10 and VHI, one would expect similar correlations between the VHI-10
FIGURE 3. ROC curve for VHI-10 and V-RQOL for bowing/presbyphonia versus all patients.
239
and V-RQOL as for the VHI and V-RQOL. The fact that Portone et al used different diagnostic groups makes our two studies somewhat difficult to compare. However, with the exception of different correlation coefficients for presbyphonia versus atrophy, our correlation values were quite similar to those of Portone et al for MTD (0.89 vs 0.93 by Portone et al), our inflammatory/infectious (0.86), which likely correlates with ‘‘chronic laryngitis/LPR of Portone et al (0.90), and overall correlation of the instruments (0.91 vs 0.82 by Portone et al)’’. Interestingly, Portone et al found a significantly lower correlation in patients with vocal fold nodules (r ¼ 0.43, n ¼ 11). We did not include vocal fold nodules as a discreet subcategory but rather included patients with nodules within our ‘‘benign lesion’’ category that was highly correlated in our study (r ¼ 0.91, n ¼ 138). ROC curve analysis showed significantly different AUC values only in the presbyphonia and muscle tension dysphonia groups. The differences between the curves, although significant, were quite small. In both cases, however, the V-RQOL showed greater sensitivity. This may indicate that the V-RQOL is a moderately more effective tool in quantifying dysphonia-related quality of life in these populations. Interestingly, the ROC curves for the neurologic dysphonia group yielded relatively large AUC values compared with those of other diagnostic categories. This may indicate that both the VHI-10 and V-RQOL are picking up differences in the neurologic subgroup compared with other voice disorders as a whole. The malignancy, leukoplakia, benign lesion, papilloma, inflammatory/infectious, and paralysis/paresis groups yielded statistically and visually identical ROC curves, suggesting that both instruments are similarly able to assess quality of life differences in these groups. Deciding which of the many available PROMs to use is complex. It has been suggested that none of the current instruments are ideal, given deficits in the development process.1 Until such time that a PROM is developed and agreed upon as the gold standard, knowledge of how currently available surveys compare with one another continues to be relevant and provides a foundation for future studies. Our data suggest that the VHI10 and V-RQOL are highly correlated and comparable although not identical. Limitations of this study include the possibility for selection bias given the fact that only patients who had diagnoses recorded on their surveys were included. Also, a normal control group was not included. Possible directions for future work include comparison of changes in score after treatment and analysis of the impact of individual question responses. CONCLUSION The VHI-10 and V-RQOL are highly correlated both overall and within specific diagnostic categories. Despite this strong correlation, direct conversion between the two scores using our derived equation is not supported on the individual or group level. The V-RQOL is more sensitive than the VHI-10 in the detection of quality of life changes related to presbyphonia and muscle tension dysphonia.
240 Acknowledgments Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.
REFERENCES 1. Branski R, Cukier-Blaj S, Pusic A, et al. Measuring quality of life in dysphonic patients: a systematic review of content development in patientreported outcomes measures. J Voice. 2010;24:193–198.
Journal of Voice, Vol. 28, No. 2, 2014 2. Hogikyan N, Sethuraman G. Validation of an instrument to measure voicerelated quality of life (V-RQOL). J Voice. 1999;13:557–569. 3. Rosen C, Lee A, Osborne J, Zullo T, Murray T. Development and validation of the voice handicap index-10. Laryngoscope. 2004;114:1549–1556. 4. Jacobson B, Johnson A, Grywalski C, Silbergleit A, Jacobson G, Benninger M. The voice handicap index (VHI): development and validation. Am J Speech Lang Pathol. 1997;6:66–70. 5. Portone C, Hapner E, McGregor L, Otto K, Johns M. Correlation of the voice handicap index (VHI) and the voice-related quality of life measure (V-RQOL). J Voice. 2007;21:723–727.
INTRODUCTION Evaluation of the dysphonic patient is complex, requiring assessment and integration of multiple variables. A detailed history, physical examination, videostroboscopy, perceptual voice evaluation, and measurement of acoustic and aerodynamic parameters all provide important information. However, the degree to which a voice disorder impacts a patient’s functional, physical, emotional, and social well-being is highly variable and based on many factors unique to the individual. Therefore, delineating the impact of a voice disorder on the patient’s overall quality of life is an essential part of the evaluation and may in fact determine the course of treatment. Several dysphoniaspecific patient-reported outcome measures (PROMs) have been developed for this purpose.1 The Voice-Related Quality of Life (V-RQOL), the Voice Handicap Index (VHI), and the Voice Handicap Index-10 (VHI-10) are commonly used PROMs at voice centers in the United States.2–4
Accepted for publication July 22, 2013. This publication was supported by grant number UL1 TR000135 from the National Center for Advancing Translational Sciences. Presented as a podium presentation at the Voice Foundation’s 41st Annual Symposium: Care of the Professional Voice; May 30 to June 3, 2012; Philadelphia, Pennsylvania. From the Department of Otorhinolaryngology - Head and Neck Surgery, Mayo Clinic, Rochester, Minnesota. Address correspondence and reprint requests to Diana M. Orbelo, Department of Otorhinolaryngology - Head and Neck Surgery, Mayo Clinic, 200 First Street SW, Rochester, MN 55905. E-mail: [email protected] Journal of Voice, Vol. 28, No. 2, pp. 237-240 0892-1997/$36.00 Ó 2014 The Voice Foundation http://dx.doi.org/10.1016/j.jvoice.2013.07.015
Developed by Hogikyan and Sethuraman in 1999, the V-RQOL consists of 10 questions pertaining to the physical functioning and social-emotional domains. A conversion equation is used to generate the final score, with higher scores indicating better quality of life.2 The VHI, developed by Jacobson et al in 1997, is a 30-question instrument dealing with the functional, physical, and emotional domains. Response options are 0–4 with higher numbers indicating greater impairment.4 Rosen et al in 2004 distilled the VHI down to the 10 most clinically valuable items as determined by clinical consensus and item analysis. The abridged version, the VHI-10, was then compared with the parent questionnaire and found to be highly correlated.3 Portone et al5 have demonstrated that the VHI and V-RQOL are highly correlated. However, to our knowledge, no studies have compared the VHI-10 and V-RQOL in a large group of patients with dysphonia. This study compares the VHI-10 and V-RQOL and investigates the possibility of direct conversion between scores in a large cohort of voice patients. PARTICIPANTS Approval for this study was granted by our Institutional Review Board, Office of Human Research Protection, Mayo Clinic, Rochester, MN (PR10-007714-02). Records of all patients presenting to our voice clinic between May of 2009 and August of 2011 who completed both the VHI-10 and V-RQOL in a single sitting were reviewed. The instruments were completed by the patient independently, before the clinical encounter. If multiple sets of surveys had been completed by a single patient due to
238
Journal of Voice, Vol. 28, No. 2, 2014
multiple visits, only the pair of surveys from the initial encounter was included. Only patients with a clear diagnosis recorded by the treating laryngologist on the day the surveys were completed were included. In patients with multiple diagnoses, the primary diagnosis was used. Patients without voice-related complaints were excluded. Eight hundred four patients met criteria and were included in the study; 59.5% (478 patients) were female and 40.5% (326 patients) were male. Average age was 58.6 years (standard deviation [SD] ± 16.3). The following diagnostic categories were identified: malignant laryngeal lesions, benign laryngeal lesions, laryngeal papillomatosis, presbyphonia, inflammatory/infectious, vocal fold paralysis/paresis, neurologic disorders, and muscle tension dysphonia. METHODS Diagnosis, VHI-10 score, and V-RQOL score were recorded for each patient. After Portone et al,5 a direct linear conversion equation was derived based on the score ranges. In our case, the anchors of 0–100 and 0–40 for the V-RQOL and VHI-10, respectively, resulted in the equation (y ¼ 0.4(x) + 40) that was used to calculate VHI-10 (y) from V- RQOL (x). Statistical analysis was performed using JMP 9.0.3 (SAS Institute, Inc, Cary, NC). Spearman correlation coefficients were derived for measured VHI-10 and V-RQOL scores both overall and by diagnosis. The difference between calculated and measured VHI-10 score was analyzed using a matchedpairs t test. Analysis of variance (ANOVA) was also used to compare calculated VHI-10 to measured VHI-10. Exploratory receiver operating characteristic (ROC) curves were generated (MedCalc 12.1.3; MedCalc Software, Ostend, Belgium) for the diagnostic groups and compared with all other diagnostic groups combined for V-RQOL and VHI-10 to assess instrument sensitivity and specificity for each diagnostic group. RESULTS Spearman correlation coefficient comparing overall VHI-10 and V-RQOL scores was 0.91 (P < 0.0001). Mean measured VHI-10 score was 19.16 (SD ± 10.37), mean calculated VHI-10 score was 15.4 (SD ± 10.31), and mean V-RQOL score was 61.5 (SD ± 25.8). Measured VHI-10 and V-RQOL scores were compared via linear regression analysis (Figure 1) yielding the equation y ¼ 0.36x + 41.5, R2 ¼ 0.82 (P < 0.0001). Spearman correlation analysis was also performed by diagnostic category (Table 1). All Spearman coefficients were statistically significant. Matched-pairs t test was performed comparing measured and predicted VHI-10 score. The mean difference was 3.76, which was significant (P < 0.001). Calculated VHI-10 was less than measured VHI-10 in 612 patients (74%). Regression analysis was also used to compare measured and calculated VHI-10 score (Figure 2). ANOVA was performed with P < 0.0001. The area under the curve (AUC) values from ROC curves were significantly different for the presbyphonia (0.586 [standard error, SE ± 0.033] for V-RQOL and 0.530 [SE ± 0.031]
FIGURE 1. Bivariate analysis of measured VHI-10 and V-RQOL scores. Regression line equation y ¼ 0.36x + 41.5, R2 ¼ 0.82 (P < 0.0001). for VHI-10; P ¼ 0.0014) and muscle tension dysphonia (0.536 [SE ± 0.026] for V-RQOL and 0.508 [SE ± 0.26] for VHI-10; P ¼ 0.018) groups, with the V-RQOL showing greater sensitivity in both cases (Figure 3). The AUC values were not significantly different in the other diagnostic groups. DISCUSSION At our institution, both the VHI-10 and V-RQOL are used in the assessment of voice specific quality of life. The aim of this investigation was to determine the degree of correlation between these two scales, with the goal of limiting the quality of life assessment to a single survey. Portone et al previously demonstrated the correlation of VHI and V-RQOL. The same study also evaluated the possibility of using a direct conversion equation to convert V-RQOL score to VHI score, so that studies using these two scales could be directly compared. In their sample, the average of scores generated with the conversion equation did not differ significantly from that of directly measured scores, but the same values for individuals were significantly different.5 Because the VHI-10 has been shown to be essentially equivalent to the VHI and the VHI has been shown to be correlated
TABLE 1. Spearman Correlation Coefficients for Each Diagnostic Category Diagnosis Malignancy Leukoplakia Benign lesion Papilloma Bowing/presbyphonia Inflammatory/infectious Paralysis/paresis Neurologic Muscle tension
Spearman (r)
n
0.94 0.87 0.91 0.95 0.76 0.86 0.93 0.86 0.89
33 19 138 21 54 86 195 100 147
Jonathan J. Romak, et al
Correlation of VHI-10 and V-RQOL
FIGURE 2. Regression analysis of measured versus predicted VHI10. Regression line y ¼ 0.91x 5.13, R2 ¼ 0.82. with the V-RQOL, we hypothesized that the VHI-10 and the V-RQOL would be correlated.3,5 To answer this question, methods similar to that of Portone et al5 were applied to our database. VHI-10 and V-RQOL were found to be highly correlated. However, the derived conversion equation did not consistently predict VHI-10 score based on V-RQOL score. The fact that the conversion did not accurately predict VHI-10 score does not refute the correlation of the two surveys, only the ability to derive one score from the other. The VHI-10 and V-RQOL are also significantly correlated within all the examined diagnostic subgroups. Interestingly, in our study, the correlation was lower for the presbyphonia group (Spearman coefficient ¼ 0.76) compared with other diagnostic groups. This suggests that the VHI-10 and V-RQOL may be less comparable for patients with presbyphonia compared with other groups. This finding differs from that of the Portone et al group, who reported a strong correlation (-0.85) between VHI and V-RQOL in their ‘‘atrophy’’ group. Possible reasons for differences include sample size (n ¼ 54 in our study versus n ¼ 22 in the study by Portone et al). Also, presbyphonia is a complex problem that often accompanies other comorbidities affecting voice specific and overall quality of life.5 Therefore, heterogeneity within the groups could be contributing to a difference in correlation. Given the known correlation between the VHI-10 and VHI, one would expect similar correlations between the VHI-10
FIGURE 3. ROC curve for VHI-10 and V-RQOL for bowing/presbyphonia versus all patients.
239
and V-RQOL as for the VHI and V-RQOL. The fact that Portone et al used different diagnostic groups makes our two studies somewhat difficult to compare. However, with the exception of different correlation coefficients for presbyphonia versus atrophy, our correlation values were quite similar to those of Portone et al for MTD (0.89 vs 0.93 by Portone et al), our inflammatory/infectious (0.86), which likely correlates with ‘‘chronic laryngitis/LPR of Portone et al (0.90), and overall correlation of the instruments (0.91 vs 0.82 by Portone et al)’’. Interestingly, Portone et al found a significantly lower correlation in patients with vocal fold nodules (r ¼ 0.43, n ¼ 11). We did not include vocal fold nodules as a discreet subcategory but rather included patients with nodules within our ‘‘benign lesion’’ category that was highly correlated in our study (r ¼ 0.91, n ¼ 138). ROC curve analysis showed significantly different AUC values only in the presbyphonia and muscle tension dysphonia groups. The differences between the curves, although significant, were quite small. In both cases, however, the V-RQOL showed greater sensitivity. This may indicate that the V-RQOL is a moderately more effective tool in quantifying dysphonia-related quality of life in these populations. Interestingly, the ROC curves for the neurologic dysphonia group yielded relatively large AUC values compared with those of other diagnostic categories. This may indicate that both the VHI-10 and V-RQOL are picking up differences in the neurologic subgroup compared with other voice disorders as a whole. The malignancy, leukoplakia, benign lesion, papilloma, inflammatory/infectious, and paralysis/paresis groups yielded statistically and visually identical ROC curves, suggesting that both instruments are similarly able to assess quality of life differences in these groups. Deciding which of the many available PROMs to use is complex. It has been suggested that none of the current instruments are ideal, given deficits in the development process.1 Until such time that a PROM is developed and agreed upon as the gold standard, knowledge of how currently available surveys compare with one another continues to be relevant and provides a foundation for future studies. Our data suggest that the VHI10 and V-RQOL are highly correlated and comparable although not identical. Limitations of this study include the possibility for selection bias given the fact that only patients who had diagnoses recorded on their surveys were included. Also, a normal control group was not included. Possible directions for future work include comparison of changes in score after treatment and analysis of the impact of individual question responses. CONCLUSION The VHI-10 and V-RQOL are highly correlated both overall and within specific diagnostic categories. Despite this strong correlation, direct conversion between the two scores using our derived equation is not supported on the individual or group level. The V-RQOL is more sensitive than the VHI-10 in the detection of quality of life changes related to presbyphonia and muscle tension dysphonia.
240 Acknowledgments Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.
REFERENCES 1. Branski R, Cukier-Blaj S, Pusic A, et al. Measuring quality of life in dysphonic patients: a systematic review of content development in patientreported outcomes measures. J Voice. 2010;24:193–198.
Journal of Voice, Vol. 28, No. 2, 2014 2. Hogikyan N, Sethuraman G. Validation of an instrument to measure voicerelated quality of life (V-RQOL). J Voice. 1999;13:557–569. 3. Rosen C, Lee A, Osborne J, Zullo T, Murray T. Development and validation of the voice handicap index-10. Laryngoscope. 2004;114:1549–1556. 4. Jacobson B, Johnson A, Grywalski C, Silbergleit A, Jacobson G, Benninger M. The voice handicap index (VHI): development and validation. Am J Speech Lang Pathol. 1997;6:66–70. 5. Portone C, Hapner E, McGregor L, Otto K, Johns M. Correlation of the voice handicap index (VHI) and the voice-related quality of life measure (V-RQOL). J Voice. 2007;21:723–727.
