Brazilian Version of the Vestibular Activities and Participation Measure: Cross-Cultural Adaptation,Validity,and Reliability

Abstract

Background

Vestibular Activities and Participation Measure (VAP) subscales assess the effect of vestibular disorders on activity and participation. This study aimed to perform the cross-cultural adaptation and assess the validity, internal consistency, reliability, and measurement error of the Brazilian version of VAP subscales.

Methods

The cross-cultural adaptation followed the translation, synthesis, back-translation, review by a committee of experts, and pretesting phases. Structural validity was assessed using exploratory factor analysis (EFA) and confirmatory factor analysis (CFA), while Spearman’s correlation between VAP subscales and the Dizziness Handicap Inventory (DHI) was used to assess construct validity. Cronbach’s alpha measured internal consistency. Intraclass correlation coefficient (ICC) assessed intra- and inter-rater reliability, and measurement error was calculated by using the standard error of measurement (SEM) and minimal detectable change (MDC).

Results

Additional information was included in the Brazilian version of the Vestibular Activities and Participation measure (VAP-BR) after approval by one of the developers of the instrument to improve the understanding among individuals. One factor was found in the EFA for each subscale with 50% explained variance. Regarding CFA, the subscales 1 (S1) and 2 (S2) presented, respectively, adequate model fit indices (ie, comparative fit index of 0.99 and 0.97, and standardized root mean square residual of 0.04 for both subscales), but a very low factor load in item 6 of S1 (0.08). Chronbach’s alpha was 0.80 (S1) and 0.82 (S2). For intra-rater assessment, the S1 and S2 presented an ICC of 0.87 and 0.90, SEM of 0.01 and 1.16, and MDC of 0.39 and 0.46, respectively. When assessed by 2 different raters, SEM values were 1.03 and 1.53, and MDC values were 2.85 and 4.23 for S1 and S2, respectively; both subscales showed an ICC of 0.92. Correlations between DHI and VAP subscales presented coefficients above 0.57.

Conclusion

The Brazilian version of VAP subscales presents good measurement properties and may assist health professionals in identifying activity limitations and participation restrictions in individuals with vestibular disorders.

Graphical Abstract

Background

The dysfunction of the vestibular system often results in dizziness, vertigo, vestibular-visual, and postural impairments. ¹ In Brazil, the prevalence of dizziness ranges from 6% to 42%, and mainly affects women.^2,3 Individuals with vestibular dysfunction avoid activities or environments that may trigger vestibular symptoms, causing activity limitation and participation restriction. ⁴

The impact of vestibular symptoms may be assessed from the perspective of the individual using Patient-Reported Outcome Measurements (PROMs). ⁵ Among the PROMs developed for individuals with vestibular dysfunction, the Dizziness Handicap Inventory (DHI) and the Vestibular Disorders Activities of Daily Living Scale were translated and validated for the Brazilian population.^6-9 Nevertheless, only the Vestibular Activities and Participation measure (VAP) considers activity limitation and participation restriction caused by vestibular symptoms.

The English version of the VAP ¹⁰ was translated for the Indian population ¹¹ and validated in Turkey. ¹² Mueller et al ¹³ analyzed the VAP using Rasch analysis and found 2 subscales that were validated for German, Jordanian, and Saudi Arabian populations. The VAP subscales were also adapted and validated for the Chinese population ¹⁴ ; however, they are still unavailable for the Brazilian population.

Instruments developed in other languages must be cross-culturally adapted before use in research and clinical practice. ¹⁵ In addition, measurement properties of an instrument need to be assessed to determine its quality. ¹⁶ Therefore, this study aimed to cross-culturally adapt the VAP subscales into Brazilian Portuguese [Brazilian version of Vestibular Activities and Participation subscale 1 (VAP1-BR) and Brazilian version of Vestibular Activities and Participation subscale 2 (VAP2-BR)] and assess the construct validity (structural and hypothesis testing), internal consistency, reliability, and measurement error in individuals with vestibular disorders.

Methods

Study Design and Sample

This methodological study performed the cross-cultural adaptation and assessed the measurement properties of the VAP subscales in 2 phases (cross-cultural adaptation and assessment of measurement properties) between January 2020 and December 2021.

Individuals aged ≥18 years, diagnosed with central and/or peripheral vestibular dysfunction by a physician, and without cognitive impairment [Mini-Mental State Examination (MMSE)] ¹⁷ were included. The exclusion criterion was the inability to complete any instrument due to understanding barriers.

Thirty individuals with vestibular dysfunction were included for the pretesting phase of the cross-cultural adaptation and 100 for the measurement properties assessment, following Beaton et al ¹⁵ and Mokkink et al ¹⁶ requirements.

Ethical Aspects

This study followed the Declaration of Helsinki principles (Respect for Human Dignity and Individuals Rights, Beneficence, Justice, Scientific and Ethical Review, Privacy and Confidentiality) and was approved by the research ethics committee of the Federal University of Rio Grande do Norte (No. 4,334,272). Individuals who agreed to participate signed an online informed consent form (Google Forms^®).

Instruments

The MMSE assesses cognitive function, considering cutoff points adjusted by education level (ie, 13 points for persons who cannot read or write, 18 for elementary and middle school graduates, and 26 for high school education and beyond). ¹⁷ It includes 30 items encompassing orientation to time, orientation to place, registration, attention and calculation, recall, language, and visual construction. ¹⁸

Sociodemographic (age, sex, marital status, education level, and occupation), clinical-functional [comorbidities, medications, and lifestyle (ie, exercise, alcohol consumption, and smoking)], and otoneurologic data (vestibular dysfunction diagnosis; the presence of tinnitus; aural fullness; oscillopsia; postural imbalance; and vertigo, dizziness, and their duration) were collected through an interview.

The VAP subscales are unidimensional and present 6 items with 5 response options based on the difficulty in activities and participation due to dizziness or imbalance (none = 0, mild = 1, moderate = 2, severe = 3, unable to perform = 4, and not applicable). The subscale 1 determines challenging-to-do activities that may cause vestibular symptoms, while the subscale 2 evaluates the immediate consequences of vestibular symptoms on activity limitation and participation restriction. Total score is calculated by summing the items of each subscale, and raw scores are transformed into interval-scaled scores. ¹³

The DHI is a 25-item instrument that assesses the self-perception of dizziness impact in 3 domains: physical (7 items), functional (9 items), and emotional (9 items). The total score is calculated by summing each domain score. ¹⁹

Measurement Properties

Structural validity refers to the degree to which items of a PROM reflect the dimensionality of the construct to be measured, whereas hypothesis testing indicates the degree to which a PROM score is consistent with the hypotheses. ¹⁶ Considering that VAP subscales assess the construct intended to be measured, a moderate to strong correlation is expected with the total and domain scores of the DHI.

Internal consistency is defined as the interrelatedness between PROM items, while reliability is the proportion of total variance between 2 or more measurements due to true differences. On the other hand, measurement error refers to the score that does not present true changes in the assessed construct (ie, systematic and random error). ¹⁶

Procedures

After permission from the authors of the VAP subscales, the cross-cultural adaptation (Figure 1) was conducted in 5 phases: translation into Brazilian Portuguese by 2 independent translators (a specialist in Otoneurology and a layperson) fluent in English and unaware of the VAP subscales (T1 and T2); synthesis of T1 and T2 to reach a consensus between translators (T3); back-translation of T3 into English by 2 independent native English-speakers translators blinded to the original VAP subscales (BT1 and BT2); review and development of the pre-final Brazilian version by an expert committee (3 physical therapist experts in vestibular physical therapy, 1 professor expert in the methodology, 1 otorhinolaryngologist, and 2 authors of the study who experienced dizziness or vertigo); and pretesting of the pre-final Brazilian version conducted with 30 individuals with vestibular dysfunction to assess their understanding of the instrument. ¹⁵

Figure 1.

Cross-cultural adaptation of the VAP1-BR and VAP2-BR subscales. VAP1-BR, Brazilian version of Vestibular Activities and Participation subscale 1; VAP2-BR, Brazilian version of Vestibular Activities and Participation subscale 2.

The expert committee aimed to ensure semantic, idiomatic, conceptual, and experiential equivalences between the original and target versions. During the pretesting, individuals were asked about their understanding of the instructions, items, and response options of the subscales. The subscales would be reformulated and retested if dubious sentences were identified in at least 20% of the sample. ²⁰

After developing the final Brazilian Portuguese version, individuals completed the assessment form and the DHI. Then, they were randomized into 2 groups (50 individuals for intra-rater and 50 for inter-rater reliability) ²¹ and responded to the VAP1-BR and VAP2-BR subscales. All data were collected in an interview format using videoconference platforms (WhatsApp^® and Google Meet^®) due to the coronavirus disease pandemic. The subscales were reapplied after an interval of 2 to 6 hours by the same or a different researcher, according to the allocation group. The adopted interval considered the wide variation over time in vestibular symptoms. In the second assessment, subscales were inverted, and items were shuffled to minimize memory bias.^10,14 Those who showed a maximum variation of 2 units on the Numerical Rating Scale (NRS) for dizziness intensity between assessments were considered clinically stable. ²² Item and total subscale scores were calculated after assessments. ²³

Data Analysis

The IBM SPSS Statistics (version 24.0. Armonk, NY: IBM Corp) and SPPS AMOS (version 21.0, Chicago: IBM SPSS) were used for the analyses. The Kolmogorov-Smirnov test verified data normality. Mean and standard deviation (SD) or median and 25th to 75th percentiles described quantitative variables, according to data normality. Absolute and relative frequencies described categorical variables.

Structural validity was assessed using exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Bartlett’s sphericity test determined the factorability of the observed matrix; P < .05 indicated factorability. The Kaiser-Meyer Olkin (KMO) test assessed the adequacy of the model to data; values ≥0.5 were considered acceptable. EFA was performed using the maximum likelihood extraction method with oblimin rotation. Furthermore, the number of factors retained was based on eigenvalues >1 and a minimum explained variance of 50%. ²⁴ Items with a factor loading >0.2 were retained in the subscales. However, items conceptually relevant to the construct were not excluded, even with a factor loading below the acceptable level. ²⁵

CFA was performed using the maximum likelihood estimation method to test the factorial structure of the VAP subscales identified in the EFA. Model fit index was standardized root mean square residual (SRMR). Comparative fit index (CFI) and goodness-of-fit index (GFI) verified which model fitted best with a baseline model whose items appeared to correlate. Threshold values for good fit were SRMR ≤ 0.08, CFI ≥ 0.97, and GFI ≥ 0.95.^26,27 The magnitude of factor loadings ≥0.3 represents the construct measured in each domain or subscale. ²⁸

Spearman’s correlation coefficient analyzed the construct validity between the VAP subscales and DHI. Values were classified as weak (0.1-0.3), moderate (0.4-0.6), or strong (≥0.70). ²⁹ The Cronbach’s alpha verified internal consistency for each subscale; values ≥0.7 were considered satisfactory. ²⁷

Intraclass correlation coefficient (ICC) assessed intra- (multiple-measures mean type and bidirectional mixed-effects model) and inter-rater reliability (multiple-rater mean-type and bidirectional random-effects model) with 95% confidence intervals (95% CIs). ³⁰ Reliability is considered sufficient by the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) when ICC ≥ 0.70. ²⁹ Measurement error was assessed using the estimated standard error of measurement (SEM) and the minimal detectable change (MDC). SEM was calculated using the formula SD√1 − ICC. The equation 1.96 × SEM × √2 was used to calculate the individual MDC at 95% CI. As the 95% CI group MDC was used in the present study, add division by √n to the individual MDC formula (1.96 × SEM × √2/√n).^31,32

Results

A total of 130 individuals were assessed: 30 in the pretesting phase and 100 in the measurement properties assessment (50 for inter- and 50 for intra-rater reliability). The sample for the pretesting phase had a mean age of 42.07 ± 6.04 years, mostly women (90%), and an education level ranging from incomplete elementary school to postgraduate. Sociodemographic and clinical-functional data of the 100 individuals are presented in Table 1.

Table 1.

Sociodemographic and Clinical-Functional Data of the Individuals Involved in Measurement Properties Assessment of the VAP-BR Subscales.

Sociodemographic and clinical-functional data	Median (25th-75th percentiles)	n (%)
Age (years)	42 (29.25-55.00)
Number of comorbidities	1 (0.00-2.00)
Number of drugs	1 (0.00-3.00)
Sex
Women		80 (80)
Men		20 (20)
Marital status
Single/divorced/widowed		48 (48)
Married /common-law married		52 (52)
Education level
Elementary school		7 (7)
High school		30 (30)
University education		32 (32)
Postgraduate		31 (31)
Occupation
With work activity		56 (56)
No work activity		44 (44)
Exercise		42 (42)
Smoking		5 (5)
Alcohol consumption		49 (49)

Data are presented as median and 25th to 75th percentiles and absolute (n) and relative frequencies (%). Exercise is defined as planned, repetitive, and structured bodily movements performed to maintain or improve physical fitness. Alcohol consumption refers to any level of intake of beverages containing ethanol.

The median and 25th to 75th percentiles for NRS in the first and second assessments were 0 (0-5) and 1 (0-5), respectively. Multiple vestibular diagnoses were observed in 21 individuals. Vestibular dysfunction diagnoses and otoneurologic symptoms are presented in Tables 2 and 3, respectively.

Table 2.

Vestibular Dysfunction Diagnoses of the Individuals Involved in Measurement Properties of the VAP-BR Subscales.

Vestibular dysfunction diagnoses	n (%)
Vestibular hypofunction	42 (42)
Vestibular migraine	48 (48)
BPPV	17 (17)
Meniere’s disease	6 (6)
Other central diseases a	8 (8)

Data are presented in absolute (n) and relative frequencies (%).

Abbreviation: BPPV, benign paroxysmal positional vertigo.

a

Cerebellar stroke, Arnold-Chiari disease, Parkinson’s disease, cerebral palsy, elderly imbalance syndrome, and vasovagal syndrome.

Table 3.

Otoneurologic Symptoms of the Individuals Involved in Measurement Properties of the VAP-BR Subscales.

Otoneurologic symptoms	Median (25th-75th percentiles)	n (%)
Duration of dizziness/vertigo (years)	7.50 (2.00-19.50)
Dizziness		87 (87)
Vertigo		78 (78)
Postural imbalance		61 (61)
Oscillopsia		37 (37)
Tinnitus		67 (67)
Aural fullness		66 (66)

n (%): absolute and relative frequencies.

Cross-Cultural Adaptation

The experts suggested some modifications in both subscales. The instructions and the “not applicable (NA)” response options were maintained to improve the comprehensibility of the assessment proposal and prevent inappropriate responses attributed to items out of the scope of individuals. In the instructions, the section “effect of dizziness and/or balance problems” was adapted to “impacto da tontura e/ou desequilíbrio corporal,” and the term “intermittent” was translated to “não contínuos.” The term “imbalance” was determined as “desequilíbrio corporal.” In the first item of subscale 1 (“remembering”) was adjusted from “recordar” to “relembrar.” In subscale 2, 3 items were discussed: “walking” in item 1 was translated as “andar” and adapted to “caminhar”; “climbing (up or down)” was translated as “escalar” and adapted to “subir ou descer”; and the word “building” in item 4 was changed to “prédios.”

Moreover, the “NA” response option was clarified by including an example: “Por exemplo, se não compete em jogos de esporte (como vôlei, futebol ou natação), sua resposta será NA para esse item na escala.” Last, the word “recently” was changed to “na última semana.”

After consensus of the expert committee, the pre-final version was administered by asking 30 individuals about their understanding of the subscales. Six individuals presented difficulty understanding the instructions and the statement in the first pretesting. As a result, the instructions, the “NA” response option, and the term “recently” were revised, and new information was added. The second sentence of the instructions was modified to “Se seu desempenho varia por causa de tontura ou desequilíbrio corporal que não são contínuos, selecione a opção levando em consideração o dia com maior dificuldade para realizar suas tarefas diárias.” All changes were approved by one of the developers (Dr Susan Whitney) of the VAP subscales. The Brazilian version of the Vestibular Activities and Participation subscales are presented in Table 4.

Table 4.

The Brazilian Version of the Vestibular Activities and Participation Subscales.

Esta escala avalia o impacto da tontura e/ou desequilíbrio corporal na sua capacidade de realizar tarefas de atividade e participação social. Por favor, quantifique sua dificuldade sem a ajuda de outras pessoas em cada item. Se seu desempenho varia por causa de tontura ou desequilíbrio corporal que não são contínuos, selecione a opção levando em consideração o dia com maior dificuldade para realizar suas tarefas diárias na última semana. Se essa tarefa não faz parte do seu dia a dia, por favor, marque a opção NA (não aplicável). Por exemplo, se não compete em jogos de esporte (como vôlei, futebol, natação), sua resposta será NA para esse item na escala.
Devido à sua tontura/desequilíbrio corporal, quanta dificuldade você vem tendo na última semana para:
Subescala 1	Nenhuma	Leve	Moderada	Grave	Incapaz de fazer	NA
1. Prestar atenção (concentração, relembrar)
2. Deitar-se (ir para a cama ou sair da cama) ou virar-se na cama
3. Levantar-se a partir da posição sentada
4. Curvar-se ou pegar objetos no chão
5. Levantar e carregar objetos
6. Esportes (envolver-se em jogos organizados em nível competitivo, formais ou informais, sozinho ou em grupo)
Subescala 2	Nenhuma	Leve	Moderada	Grave	Incapaz de fazer	NA
1. Caminhar distâncias longas
2. Subir ou descer (degraus, elevador e escada rolante)
3. Correr
4. Mover-se dentro de prédios que não sejam sua casa
5. Utilizar meios de transporte (transporte público ou privado—sendo um passageiro)
6. Guiar um veículo: dirigir um carro ou andar de bicicleta

Abbreviation: NA, Não aplicável.

Structural Validity

In the EFA, both subscales showed significance for Bartlett’s sphericity test. KMO values for VAP1-BR and VAP2-BR were 0.80 and 0.75, respectively. One factor was identified for each VAP-BR subscale with a percentage of explained variance for subscales 1 and 2 of 50.07% and 50.17%, respectively. The factor loadings of each item and the eigenvalues of each factor are shown in Table 5.

Table 5.

Exploratory Factor Analysis of the VAP-BR Subscales (n = 100).

VAP1-BR	Factor loadings	VAP2-BR	Factor loadings
Item 1	0.56	Item 1	0.70
Item 2	0.61	Item 2	0.76
Item 3	0.68	Item 3	0.56
Item 4	0.82	Item 4	0.74
Item 5	0.85	Item 5	0.47
Item 6	0.08	Item 6	0.56
Descriptive statistics of the models
KMO measure	0.80	KMO measure	0.75
Bartlett’s sphericity test	P < .001	Bartlett’s sphericity test	P < .001
Eigenvalue	3.00	Eigenvalue	3.01
Explained variance (%)	50.07	Explained variance (%)	50.17

Abbreviations: KMO, Kaiser-Meyer-Olkin; VAP1-BR, Brazilian version of Vestibular Activities and Participation subscale 1; VAP2-BR, Brazilian version of Vestibular Activities and Participation subscale 2.

The CFA determined unidimensional models for each subscale in the Brazilian Portuguese version. The model fit indices for the VAP1-BR and VAP2-BR subscales are shown in Table 6.

Table 6.

Confirmatory Factor Analysis of the VAP-BR Subscales (n = 100).

Subscales	CFI	GFI	SRMR
VAP1-BR	0.99	0.97	0.04
VAP2-BR	0.97	0.96	0.04

Abbreviations: CFI, comparative fit index; GFI, goodness-of-fit index; SRMR, standardized root mean square residual; VAP1-BR, Brazilian version of Vestibular Activities and Participation subscale 1; VAP2-BR, Brazilian version of Vestibular Activities and Participation subscale 2.

Factor loadings are shown in Figures 2 and 3. Covariances were estimated between error terms referring to items that presented similar circumstances to obtain indices with the best fit for subscale 2.

Figure 2.

Confirmatory factor structure model for VAP1-BR. VAP1-BR, Brazilian version of Vestibular Activities and Participation subscale 1.

Figure 3.

Confirmatory factor structure model for VAP2-BR. VAP2-BR, Brazilian version of Vestibular Activities and Participation subscale 2.

Construct Validity—Hypotheses Testing

The total scores of the VAP1-BR and VAP2-BR subscales were correlated with the total and domain scores of the DHI (Table 7).

Table 7.

Spearman Correlation Coefficients Between the VAP-BR and DHI Subscales (Total and Domain Scores).

DHI	VAP1-BR	VAP2-BR
Total score	0.70**	0.72**
Emotional domain score	0.58**	0.64**
Physical domain score	0.66**	0.63**
Functional domain score	0.70**	0.70**

Abbreviations: DHI, Dizziness Handicap Inventory; VAP1-BR, Brazilian version of Vestibular Activities and Participation subscale 1; VAP2-BR, Brazilian version of Vestibular Activities and Participation subscale 2.

**

P < .001.

Internal Consistency, Reliability, and Measurement Error

The Cronbach’s alpha values for VAP1-BR and VAP2-BR were 0.80 and 0.79, respectively. All individuals showed symptom stability (ie, variation of up to 2 units in the NRS between assessments). The VAP1-BR had an ICC of 0.87 (0.79-0.93), SEM of 1.01, MDC (group) of 0.39 for intra-rater assessment, and an ICC of 0.92 (0.86-0.95), SEM of 1.03, and MDC (group) of 0.40 for inter-rater assessment. VAP2-BR presented an ICC of 0.90 (0.83-0.94), SEM of 1.16, and MDC (group) of 0.46 for intra-rater assessment, and an ICC of 0.92 (0.86-0.95), SEM of 1.53, and MDC (group) of 0.60 for inter-rater assessment.

Discussion

The main results of this study showed that the VAP1-BR and VAP2-BR presented (1) comprehensible items for Brazilian individuals with vestibular dysfunction; (2) one-dimensional structure (except item 6 of the VAP1-BR); (3) adequate hypothesis testing, with results according to the hypotheses formulated for construct validity; and (4) adequate reliability and internal consistency.

Although the same guideline ¹⁵ was used to cross-culturally adapt the Brazilian, Chinese, ¹⁴ German, and Arabic ¹³ versions of the VAP, some differences can be mentioned regarding the expert committee. While Mueller et al ¹³ involved the initial VAP developers, the Chinese version included translators, study authors, and a methodologist with clinical experience. ¹⁴ In contrast, we involved specialist health professionals and individuals who experienced vestibular symptoms to ensure that the Brazilian version considered the perception of the target population.

The VAP subscales allow the interaction between the individual, health condition, and environmental factors; thus, addressing the activity and participation domain of the International Classification of Functioning, Disability and Health (ICF). These subscales are also aligned with the “4 Ps” concept (prevention, prediction, plasticity, and participation) proposed in the Special Therapeutic Exercise Project conference, which aimed to explore new theories and evidence on neurorehabilitation. These 4 concepts seek to translate scientific evidence into physical therapy practice. ³³ Specifically, prevention and participation encompass the effect of movement disorders on the activity and participation of individuals. ³⁴

A single domain was also identified for each subscale, corroborating the study by Mueller et al. ¹³ Although item 6 of VAP-BR1 showed inadequate factor loading in EFA and CFA, the item was maintained due to the Brazilian context. Nevertheless, one aspect needs to be discussed: the Southeast and South regions have the highest percentage (91.14%) of sports incentive projects funded by the Federal Government. ³⁵ In recent decades, the popularity of contact sports (eg, Brazilian Jiu-Jitsu and Muay-Thai) has increased due to the success of national athletes in Mixed Martial Arts events. ³⁶ Pinheiro and Nascimento Sousa ³⁷ reported at least one episode of concussion among Brazilian Jiu-Jitsu (61%) and Muay-Thai (86%) athletes; dizziness was one of the most reported symptoms. Thus, despite the selection bias probably arising from the convenience sample, item 6 of the VAP1-BR is included in the Brazilian context.

In our study, item 6 of subscale 1 presented a high frequency of NA responses. This result was also observed in the Chinese subscales and 34-item original versions. In addition, the present study included older individuals, and Alghwiri et al ¹⁰ and Wu et al ¹⁴ found that younger individuals were more likely to practice sports. Similarly, participation in sports decreases with age for both sexes in Brazil. Lack of time, not enjoying sports, and health problems are the main reasons for Brazilians not practicing sports.^38,39

The VAP-BR subscales showed moderate to strong correlations with the total and domain scores of the DHI, supporting the construct validity for the Brazilian VAP subscales. The DHI measures the impact of dizziness on daily life of individuals with vestibular dysfunction ¹⁹ based on items similar to those of the VAP subscales. Thus, higher scores on both instruments suggest a worse perceived disability in individuals with vestibular symptoms.

Cronbach’s alpha values were satisfactory and consistent with the study by Wu et al. ¹⁴ These values were also equivalent to the Person Separation Index evaluated in the original version of the subscales, ¹³ evidencing that interrelated items of each subscale assessed the activity and participation of the ICF. Moreover, the total subscale scores showed excellent intra- and inter-rater reliability, corroborating the ICC values found in the Chinese version of the subscales (>0.70). ¹⁴

Despite using the same equation, ³² we observed a higher measurement error than the study of Alghwiri et al ¹⁰ possibly due to low ICC values. The number of items in the VAP subscales is smaller than in the original scale, and some items and total scores were on ordinal scales, whereas the original version and the subscales used an interval scale.^10,13 Consequently, the MDC must be over 0.39 and 0.40 points for subscale 1 and above 0.46 and 0.60 points for subscale 2 (intra- and inter-rater administration, respectively) to be clinically relevant. Nevertheless, further studies are needed to confirm this hypothesis.

Although the original version was self-reported, ¹³ trained interviewers applied the Brazilian VAP subscales via videoconference to ensure measurement quality. According to cross-cultural adaptation studies, this administration mode is preferred since most Brazilians are unfamiliar with self-reported instruments.^40,41 Interview-based administration includes individuals who cannot respond independently, ensuring a high response rate.^42,43 In addition, previous studies using different PROMs for other health outcomes demonstrated similar results for self-reported and interviewing evaluations, even when assessing measurement properties.^44-48

Interviewers can administer PROMs using paper, telephone, computer, or another electronic device. ⁴⁹ According to Saarijärvi and Bratt, ⁵⁰ videoconference was the most similar format to in-person interviews. Although the measurement variance was affected by different administration methods, social desirability effects were similar for in-person and remote interviews. ⁵¹ Moreover, the remote interview can save time, reach individuals from rural areas, and may be useful during challenging circumstances (eg, COVID-19).^52-54

This study had some limitations. Although we included individuals with vestibular dysfunction aged over 18 years, younger individuals represented only a small part of the sample, suggesting limited participation in vigorous activities (eg, sports at a competitive level). In addition, illiterate individuals were not included, and few reported low education levels. At the same time, the remote format reached individuals from distant cities.

The VAP subscales were the first instruments to assess the impact of vestibular dysfunction on activity limitation and participation restriction among the Brazilian population. According to COSMIN criteria, the measurement properties presented in this study would be classified as very good quality across most criteria. ¹⁸ The use of both subscales is supported by clinical research, enabling health professionals to better intervene in activity limitation and participation restrictions and monitor the progress of individuals during and after treatment. Moreover, as VAP subscales are summarized versions of the original instrument, measures may be promptly obtained.

Conclusion

The Brazilian VAP subscales are comprehensible, reflect the content proposed by the original instrument, and present good measurement properties confirming the construct validity (structural and hypothesis test). Moreover, the subscales showed excellent intra- and inter-rater reliability. As a result, the VAP-BR subscales are recommended for identifying activity limitations and participation restrictions in the Brazilian population with vestibular dysfunction. These subscales may also assist health professionals in providing adequate interventions for this population.