Validation of the Penn Acoustic Neuroma Quality-of-Life Scale (PANQOL) for Spanish-Speaking Patients

Abstract

Objective

To perform translation, cross-cultural adaptation, and validation of the Penn Acoustic Neuroma Quality-of-Life Scale (PANQOL) to the Spanish language.

Study Design

Prospective study.

Setting

Tertiary neurotologic referral center.

Subjects and Methods

PANQOL was translated and translated back, and a pretest trial was performed. The study included 27 individuals diagnosed with vestibular schwannoma. Inclusion criteria were adults with untreated vestibular schwannoma, diagnosed in the past 12 months. Feasibility, internal consistency, test-retest reliability, construct validity, and ceiling and floor effects were assessed for the present study.

Results

The mean overall score of the PANQOL was 69.21 (0-100 scale, lowest to highest quality of life). Cronbach’s α was 0.87. Intraclass correlation coefficient was performed for each item, with an overall score of 0.92. The κ coefficient scores were between moderate and almost perfect in more than 92% of patients. Anxiety and energy domains of the PANQOL were correlated with both physical and mental components of the SF-12. Hearing, balance, and pain domains were correlated with the SF-12 physical component. Facial and general domains were not significantly correlated with any component of the SF-12. Furthermore, the overall score of the PANQOL was correlated with the physical component of the SF-12.

Conclusion

Feasibility, internal consistency, reliability, and construct validity outcomes in the current study support the validity of the Spanish version of the PANQOL.

Keywords

acoustic neuroma vestibular schwannoma quality of life

Vestibular schwannoma (VS) is a benign tumor of the eighth cranial nerve. VS accounts for 8% of all intracranial neoplasms in adults and more than 90% of all cerebellopontine angle neoplasms.¹ The overall incidence of VS is about 1 case per 100,000 persons per year. However, this incidence is increasing due to the extended use of computed tomography (CT) and magnetic resonance imaging (MRI).

Hearing loss is usually the most common symptom, followed by tinnitus. In more advanced stages of this disease, symptoms such as facial palsy or intracranial hypertension may occur, sometimes becoming life-threatening.² Management options include surgical resection, radiation therapy, and serial observation.^3-5

Given that VS is a chronic disease with symptoms that can be very incapacitating and that treatments are not without after effects, it is essential to know and measure quality of life (QOL) in these patients. The Penn Acoustic Neuroma Quality-of-Life Scale (PANQOL) is the first disease-specific QOL instrument for patients with VS. It was recently developed and validated with US patients.⁶

However, this tool has been developed in the English language, and even though it is highly recommended for measuring QOL in VS, it is not available in the Spanish language.

The purpose of this project was to adapt the questionnaire to the Spanish language and to validate the Spanish version of the PANQOL for its use with Spanish-speaking patients.

Methods

Patients

A prospective study was conducted on patients diagnosed with VS and evaluated at the Otolaryngology Department of this tertiary referral center between January 2015 and December 2015. The study was approved by the local research ethics committee (Comité de la Ética de la Investigación del Hospital Universitario Ramón y Cajal).

Inclusion criteria were the following: sporadic VS recently diagnosed (less than 1 year), untreated VS, age older than 18 years, Spanish as the native language, comprehension of the purpose of the study, and availability to repeat the questionnaire within 2 weeks. Patients had to comply with all of the above-mentioned criteria to be included in the study.

Patients with a cerebellopontine angle tumor other than VS (suspected by MRI features) and those with neurofibromatosis type II or treated tumors (either with microsurgery or radiotherapy) were excluded.

Patients’ characteristics were obtained from a scheduled clinical visit. Physical examination, an updated audiogram, and MRI were performed on all the patients. The tumor size was measured according to the Kanzaki classification⁷ and the hearing level according to the American Academy of Otolaryngology—Head and Neck Surgery Foundation guidelines.⁸

Materials

Patients were administered the following tests:

PANQOL: This questionnaire has 26 items divided into 7 domains: balance (6 items), hearing (4 items), anxiety (4 items), energy (6 items), pain (1 item), face (3 items), and general health (2 items). Domain scores are calculated on a scale from 0 to 100 (lowest to highest QOL), and a composite score (cQOL), calculated as the average of individual domain scores, is also reported on a scale from 1 to 100.⁶

SF-12: This is a generic, multipurpose short-form survey with 12 questions that, when combined, scored, and weighted, results into 2 domains: mental and physical functioning and overall health-related QOL. It is scored from 0 to 100 (lowest to highest QOL).⁹

Translation and Transcultural Adaptation

This study was performed following the guidelines proposed for cross-cultural adaptation of health-related measures of QOL.^10-12 After requesting permission from the authors, the original version of the PANQOL was translated into Spanish by 3 Spanish-dominant bilingual speakers working independently. After this first stage, a questionnaire approved by consensus was obtained. Next, 3 independent bilingual translators translated the questionnaire back into English. These translators were different from those used in the first translation and were unaware of the existing English version. The second version of the questionnaire was given to 30 healthy individuals by a member of the team for the cross-cultural adaptation. The examiner asked each one if it was difficult to understand each question and if he or she could explain the meaning of the question in his or her own words. This phase resulted in a provisional version that was evaluated by the research group, and finally the definitive version of the PANQOL was obtained (version 1) ( Figure 1 ). A second version with the same questions in a different order was obtained as well (version 2).

The test was given to 27 patients who met the inclusion criteria during the study period. Patients filled out the 2 different versions of the questionnaire, separated by a period of 15 days. Version 1 was administered during a scheduled visit along with the general QOL SF-12 questionnaire. Version 2 was carried out via telephone calls.

Data Processing and Statistical Treatment

Feasibility, internal consistency, test-retest reliability, construct validity, and ceiling and floor effects were assessed for the present study. Feasibility was evaluated according to the percentage of no answer in every item (missing items) and the percentage of patients who did not answer some items. Internal consistency refers to the homogeneity of the questions comprising the questionnaire; it was contrasted with Cronbach’s α coefficient. The minimum acceptable value was 0.7.¹¹

Test-retest reliability was analyzed by the reproducibility of the response after repeating the test. It refers to the reproducibility of measurements by giving it to the same person or a group of people, on 2 or more different occasions, hours or days apart.¹³ It was determined using the intraclass correlation coefficient (ICC) and represented using a Bland-Altman plot. Another aspect of reliability is the consistency of the testers’ results in 2 different moments; it was evaluated, item by item, using the κ index. The possible scores of the κ coefficient are as follows: without agreement (if score is below 0), indifferent (if score is between 0 and 0.20), little agreement (if score is between 0.21 and 0.40), moderate agreement (if score is between 0.41 and 0.60), strong agreement (if score is between 0.61 and 0.80), and close to perfect (if score is between 0.81 and 1).

Construct validity was assessed with a Pearson correlation analysis between the domains of the PANQOL and physical and mental components of the SF-12.

Results

Demographic and clinical features of the 27 participants are detailed in Table 1 .

Table 1.

Demographic and Clinical Features of the Participants.

Characteristic	Value
No. (%) of participants	27 (100.0)
Age, mean (range), y	57.5 (30-74)
Sex, No. (%)
Male	13 (48.1)
Female	14 (51.8)
Tumor size, No. (%)
Intracanalicular	21 (77.7)
Extracanalicular 1-10 mm	6 (22.2)
Extracanalicular 11-20 mm	0
Extracanalicular 21-30 mm	0
Extracanalicular 31-40 mm	0
Presenting symptom, No. (%)
Hearing loss	10 (37.0)
Hearing loss + tinnitus	5 (18.5)
Sudden hearing loss	2 (7.4)
Tinnitus	8 (29.6)
Balance disorders	2 (7.4)
Hearing level (AAO), No. (%)
Class A	16 (59.0)
Class B	9 (33.3)
Class C	2 (7.4)
Class D	0
Cranial nerve deficits, No. (%)
V cranial nerve	0
VII cranial nerve	1 (3.7)
Lower cranial nerves	0

Abbreviation: AAO, American Academy of Otolaryngology.

The mean (SD) cQOL score of the PANQOL in patients in the first measurement was 69.21 (20.81; 0 to 100 scale, lowest to highest QOL). Table 2 shows mean and SD for the different domains. In the SF-12 questionnaire, the mean (SD) score was 45.2 (6.9) for the physical domain and 39.8 (5.8) for the mental domain.

Table 2.

Means and Standard Deviations of Penn Acoustic Neuroma Quality-of-Life Scale Domains of Patients with Vestibular Schwannoma in the Current Study.

Domain	Mean	Standard deviation
Anxiety	80.09	29.57
Face	75.30	24.29
General	53.24	15.73
Balance	68.36	29.87
Energy	66.97	28.79
Hearing	66.43	28.33
Pain	74.07	35
cQOL	69.21	20.81

cQOL, composite quality-of-life score (normalized 0-100).

Feasibility was studied in 27 questionnaires. The questionnaires were completely filled in by all the patients. The internal consistency, measured with the Cronbach α for the global questionnaire, was 0.87.

Test-retest agreement was evaluated in 26 patients who completed the 2 versions of the PANQOL. It is represented in a Bland-Altman plot ( Figure 2 ), where no tendency of higher or lower discrepancy is observed with higher or lower PANQOL scores. The global questionnaire ICC was 0.92 (95% confidence interval [CI], 0.84-0.97). The ICC for each domain is represented in Table 3 . The κ coefficient was calculated for each item. In items 6 and 25 of the PANQOL, the κ index took negative values, −0.39 and −0.62, respectively, indicating no agreement between the responses on the first and second visits. In all other items (92%), the degree of agreement was at least moderate, and in 18 items (69%), it was almost perfect.

Table 3.

Penn Acoustic Neuroma Quality-of-Life Scale Intraclass Correlation Coefficient for Individual Domains and Composite Score.

Domain	ICC, Mean (95% CI)
Anxiety	0.665 (0.373 to 0.837)
Face	0.579 (0.247 to 0.790)
General health	−0.138^a (–0.493 to 0.256)
Balance	0.682 (0.399 to 0.846)
Energy	0.725 (0.469, 0.869)
Hearing	0.603 (0.281 to 0.803)
Pain	0.501 (0.140 to 0.744)
cQOL	0.925 (0.837 to 0.966)

Abbreviations: CI, confidence interval; cQOL, composite quality of life; ICC, intraclass correlation coefficient.

This domain took negative values as it comprised only 2 items (see explanations in the text).

The construct validity of the questionnaire was measured with the correlation coefficients between the domains of the PANQOL and physical and mental components of the SF-12 ( Table 4 ). Anxiety and energy domains of the PANQOL were correlated with both physical and mental components of the SF-12. Hearing, balance, and pain domains were correlated with the SF-12 physical components. Facial and general domains were not significantly correlated with any component of the SF-12. Furthermore, the cQOL of the PANQOL was correlated with the physical component of the SF-12.

Table 4.

Construct Validity: Pearson Correlation Coefficients between the Domains of the PANQOL and Physical and Mental Components of the SF-12.

	SF-12
PANQOL Domains	Physical Dimension	Mental Dimension
Anxiety	0.42^a	0.43^a
Face	0.27	0.15
General	0.13	−0.12
Balance	0.74^b	0.25
Energy	0.54^b	0.53^b
Hearing	0.62^b	0.17
Pain	0.54^b	0.12
cQOL	0.67^b	0.35

Abbreviations: cQOL, composite quality of life; PANQOL, Penn Acoustic Neuroma Quality-of-Life Scale.

P < .05.

P < .001.

There were no patients with the highest possible score (ceiling effect), and only 1 patient (3.7%) with the lowest score (floor effect) was found.

Discussion

Quality-of-life questionnaires are tools that allow patients and physicians to evaluate how the disease and therapeutic decisions affect every aspect of their lives. With this aim, the PANQOL was developed as a specific instrument for measuring QOL in patients with VS.⁶

However, this scale was validated only for English-speaking patients. Spanish is the second most widely spoken language in the world, with approximately 400 million native speakers, and has official status in 21 countries.¹⁴ For this reason, it was necessary to design a similar tool for patients speaking this language. Thus, a systematic process of translation and cultural adaptation of questionnaires measuring QOL in health sciences was followed.¹⁰ For validation of the questionnaire, the following values were obtained: feasibility, internal consistency, test-retest reliability, construct validity, and ceiling and floor effects.

Internal consistency was measured with Cronbach’s α index. This index estimates whether a set of items measures the same underlying concept (construct) or theoretical dimension; the closer to 1 the index, the greater the internal consistency of the questionnaire items. A ratio above 0.7 is considered acceptable, above 0.8 is considered good, and more than 0.9 is excellent. A low Cronbach’s α indicates a lack of correlation between the items in a scale, which makes summarizing the items unjustified.¹¹ In the present study, a Cronbach’s α index of 0.87 was obtained. The result was similar to other cultural adaptations of the PANQOL¹⁵ and the original study of Shaffer et al⁶ ( Table 5 ). When internal consistency is relevant and a factor structure has been determined previously, a confirmatory factor analysis should be performed. This analysis is visible in Table 5 , with a Cronbach’s α index between 0.67 and 0.97 in the different subscales. Therefore, except for the general health domain, the internal consistency of the PANQOL Spanish version can be considered “good.”

Table 5.

Comparison of the Internal Consistency (Cronbach’s α) of PANQOL in the Current Study and Revision of the Literature.

	Internal Consistency (Cronbach’s α Index)
PANQOL Domains	Current Study	Shaffer et al⁶	Van Leeuwen et al¹⁵
Balance	0.91	0.87	0.94
Hearing	0.85	0.77	0.75
Anxiety	0.97	0.81	0.88
Energy	0.89	0.88	0.91
Pain	NA	NA	NA
Face	0.71	0.71	0.65
General health	0.67	0.73	0.31

Abbreviations: PANQOL, Penn Acoustic Neuroma Quality-of-Life Scale; NA, not available.

The questionnaire test-retest reliability was assessed by means of the ICC and the κ coefficient. The ICC is the recommended index for evaluating the reliability of measurements associated with quantitative variables.^11,16 It allows assessing the agreement between 2 measurements on the same individual and the general agreement between 2 different observations. The ICC measures the proportion of total variability attributable to patients. Its values range between 0 and 1, with 0 being no match and 1 being the absolute consistency or reliability of data. In general, it is considered acceptable above 0.4 and excellent above 0.75.¹⁶ In our case, an ICC of 0.92 was obtained for the cQOL score, reflecting an excellent reliability. The ICC for each domain is represented in Table 3 . Although the overall ICC is excellent, the ICC for individual domains is lower, as they comprised fewer items, which increases the variability. In fact, the ICC took a negative value for the general health domain, as it comprised only 2 items, one of which is item 25, which took a negative value on the κ index (–0.62), indicating no agreement between the responses on the first and second visits. Interestingly, this item is the only one that is written in a positive sense. This fact is discussed further on.

Another aspect of reliability is to assess the agreement between 2 different time points over the same observer.¹⁷ For this aim, the κ coefficient is employed in case of categorical variables. In this Spanish version of the PANQOL, an almost perfect degree of agreement was obtained in 69% of the items, indicating a very good reliability between the answers in the first and second visits.

There are 2 necessary assumptions in test-retest reliability.¹³ The first one is that the true score does not change between examinations. For this purpose, the 6-month period used by the original author⁶ was considered too long, as significant clinical changes could take place in patients with VS over shorter periods of time, for example, sudden hearing loss. The second one is that the time period between examinations is long enough to prevent learning, carryover effects, or recall. Because of this, 2 weeks was considered a judicious interval of time.

To assess the construct validity of the Spanish PANQOL, we have used the correlation coefficients between the domains of the PANQOL and the physical and mental components of the SF-12. For this study, the SF-12 was used instead of the larger SF-36 scale used in other studies.^6,15 Several factors led us to choose the SF-12 instead of the SF-36. First, it reproduces more than 90% of the variance in the SF-36 mental and health domains both in the United States and in Spain. Second, it has an accurate reproduction of average scores for both SF-36 domains, having a correlation of 0.95 for the physical domain and 0.96 for the mental domain. Finally, its reduced length permits simplifying the administration of the test, reducing the time and number of pages needed.⁹

All domains of the PANQOL, except the general health and facial domains, were correlated with physical or mental SF-12 areas. One possible explanation for the lack of correlation of the general domain is, as mentioned by Van Leeuwen et al,¹⁵ that this domain is represented by only 2 questions in the PANQOL. Another aspect to take into consideration is the lack of correlation of the facial domain with any domain of the SF-12 questionnaire, which was also a finding in Van Leeuwen et al.¹⁵ This may be a consequence of the fact that in both Van Leeuwen et al¹⁵ and the present study, there was only 1 patient with facial nerve dysfunction, making correlation less likely. In addition, Shaffer et al⁶ in their original validation study had 18 patients with different degrees of facial nerve dysfunction, but they did not obtain any correlation between the facial domain of the PANQOL and any of the SF-36 domains. This suggests that facial function is not specifically well represented in general health questionnaires, although it can be a very limiting sequela.

As we can see, general QOL questionnaires do not always relate with all dimensions of the disease, emphasizing the importance of specific measuring tools for specific disease conditions.

No patients had the highest possible score (ceiling effect) in this study, and only 1 patient (representing 3.7%) had the lowest score (floor effect). To consider floor or ceiling effects, at least 15% of the respondents must achieve the lowest or highest possible score, respectively. The fact that floor or ceiling effects appear would indicate that there are items missing in both the lower and upper ends of the scale, implying limited content validity. Patients with these scores cannot be differentiated from one another, and therefore reliability is reduced.¹¹

The mean cQOL score of PANQOL in the present study was 69.21. This sample comprised patients with untreated and recently diagnosed VS. This reflects that, although patients did not have any after effects derived from treatment, the sole presence of a VS diminished their QOL significantly. The worst scores were obtained in the general perception of their health (53.24 points), followed by the hearing and balance domains (66.43 and 68.36 points, respectively). The best scores were obtained for the anxiety domain (80.09 points), followed by the facial and pain domains (75.3 and 74.07 points, respectively). It is interesting to observe that although patients had a negative perception of their health, their levels of anxiety were controlled.

These results are consistent with the most important series published to date.^6,15,18,19 In Figure 3 , we have represented the mean values of the PANQOL domains and the cQOL. Only comparable series have been used, that is, series that reported VS under observation management. In those studies with different treatment groups,^18,19 only the data from the observation group have been taken.

Although the PANQOL has demonstrated to be an excellent tool for measuring QOL in VS patients, we found some limitations for the interpretation of the results in a clinical setting. First is the need for a normalization to a 0 to 100 scale and inversion of the score to be able to compare the results with the data published in the literature.^6,15,18,19 Therefore, it is necessary to apply a formula to the original score (OS) to obtain a normalized score (NS) (see Appendix 1 available online). The second issue is related to item 25 (“my health is excellent”). As this item is written in an opposite sense to the other questions, it is necessary to invert the score of this item before obtaining a global score; otherwise, it would affect the score in an opposite way to the rest of the items. Interestingly, this fact is not discussed in any of the publications revised for this study.

The clinician must be aware of these 2 facts (score conversion and inversion of the score of item 25) to interpret the results correctly.

Conclusions

Quality of life should be evaluated in all patients diagnosed with VS with descriptive intention and results measurement. Feasibility, internal consistency, reliability, and construct validity outcomes in the current study support the validity of the Spanish version of the PANQOL and could be used in clinical practice as well as for assessing the quality of life in Spanish-speaking patients with VS.

Author Contributions

Maria del Mar Medina, project design, analysis and interpretation, major manuscript drafting, manuscript revision, final approval; Alvaro Carrillo, data acquisition, manuscript drafting, approval; Ruben Polo, project conception and design, manuscript drafting, final approval; Borja Fernandez, project design, analysis and interpretation, approval; Daniel Alonso, data acquisition, critical manuscript revision, approval; Miguel Vaca, project conception and design, critical manuscript revision, final approval; Adela Cordero, review of data, manuscript drafting, approval; Cecilia Perez, project conception and design, critical manuscript revision, final approval; Alfonso Muriel, project design, analysis and interpretation, manuscript revision, final approval; Ignacio Cobeta, project conception and design, critical manuscript revision, final approval.

Disclosures

Competing interests: None.

Sponsorships: None.

Funding source: None.

Footnotes

Acknowledgements

We thank Michael J. Ruckenstein, MD, and his team for their permission to translate and validate the original version of the PANQOL questionnaire.

No sponsorships or competing interests have been disclosed for this article.

This article was presented at the 67th SEORL-CCC (Spanish Association of Otolaryngology–Head and Neck Surgery) Annual Meeting; October 18-21, 2016; Sevilla, Spain.

Supplemental Material

Additional supporting information is available in the online version of the article.

References

Brackmann

Arriaga

. Neoplasms of the posterior fossa. In: Flint

Haughey

Niparko

, eds. Cummings Otolaryngology: Head and Neck Surgery. 6th ed. Philadelphia, PA: Saunders; 2015:177.

Matthies

Samii

. Management of 1000 vestibular schwannomas (acoustic neuromas): clinical presentation. Neurosurgery. 1997;40:1-9.

Sughrue

Yang

Aranda

. Beyond audiofacial morbidity after vestibular schwannoma surgery. J Neurosurg. 2011;114:367-374.

Muzevic

Legcevic

Splavski

Cayé-Thomasen

. Stereotactic radiotherapy for vestibular schwannoma. Cochrane Database Syst Rev. 2014;12:CD009897.

Smouha

Yoo

Mohr

Davis

. Conservative management of acoustic neuroma: a meta-analysis and proposed treatment algorithm. Laryngoscope. 2005;115:450-454.

Shaffer

Cohen

Bigelow

Ruckenstein

. Validation of a disease-specific quality-of-life instrument for acoustic neuroma: the Penn Acoustic Neuroma Quality-of-Life Scale. Laryngoscope. 2010;120:1646-1654.

Kanzaki

Tos

Sanna

Moffat

Monsell

Berliner

. New and modified reporting systems from the consensus meeting on systems for reporting results in vestibular schwannoma. Otol Neurotol. 2003;24:642-649.

American Academy of Otolaryngology—Head and Neck Surgery Foundation. Committee on Hearing and Equilibrium guidelines for the evaluation of hearing preservation in acoustic neuroma (vestibular schwannoma). Otolaryngol Head Neck Surg. 1995;113:179-180.

Ware

Jr Kosinski

Keller

. A 12-item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med Care. 1996;34:220-233.

10.

Guillemin

Bombardier

Beaton

. Cross-cultural adaptation of health-related quality of life measures: literature review and proposed guidelines. J Clin Epidemiol. 1993;46:1417-1432.

11.

Terwee

Bot

De Boer

. Quality criteria were proposed for measurement properties of health status questionnaires. J Clin Epidemiol. 2007;60:34-42.

12.

Acquadro

Conway

Hareendran

Aaronson

; European Regulatory Issues and Quality of Life Assessment (ERIQA) Group. Literature review of methods to translate health-related quality of life questionnaires for use in multinational clinical trials. Value Health. 2008;11:509-521.

13.

Vaz

Falkmer

Passmore

Parsons

Andreou

. The case for using the repeatability coefficient when calculating test-retest reliability. PLoS ONE. 2013;8:e73990.

14.

Tinsley

Board

. Languages for the future. 2014. http://www.britishcouncil.org/sites/default/files/languagues-for-the-future-repor.pdf. Accessed August 20, 2016.

15.

Van Leeuwen

Herruer

Putter

Jansen

Van Der Mey

Kaptein

. Validating the Penn Acoustic Neuroma Quality of Life Scale in a sample of Dutch patients recently diagnosed with vestibular schwannoma. Otol Neurotol. 2013;34:952-957.

16.

Prieto

Lamarca

Casado

. Assessment of the reliability of clinical findings: the intraclass correlation coefficient. Med Clin (Barc). 1998;110:142-145.

17.

Landis

Koch

. The measurement of observer agreement for categorical data. Biometrics. 1977;33:159-174.

18.

Robinett

Walz

Miles-Markley

Moberly

Welling

. Comparison of long-term quality of life outcomes in vestibular schwannoma patients. Otolaryngol Head Neck Surg. 2014;150:1024-1032.

19.

McLaughlin

Bigelow

Lee

JYK

Ruckenstein

. Quality of life in acoustic neuroma patients. Otol Neurotol. 2014;36:653-656.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.06 MB