Cultural Adaptation of an Instrument to Assess Physical Fi Tness in Cardiac Patients

OBJECTIVE: To validate the content and to evaluate the reliability of the Veterans Specifi c Activity Questionnaire instrument, culturally adapted for use in the Brazilian population of cardiac patients. METHODS: The instrument was translated and back-translated and subsequently analyzed by a committee of judges to evaluate its semantic-idiomatic and cultural equivalences. Physical activities were replaced when indicated in the instrument, but uncommon in the daily life of the target population. Another committee of specialists analyzed the metabolic equivalence of replaced activities. The proportion of agreement of evaluation of the judges was quantifi ed by the Content Validity Index. The pre-test was performed in two stages (n1 and n2=15). Reliability was assessed using the test-retest (interval of 7-15 days, n = 50). RESULTS: In the evaluation of semantic-idiomatic and cultural equivalences, items with a Content Validity Index < 1 were reviewed until consensus among the judges was obtained. The second committee found 100% of agreement in the analysis of metabolic equivalence between original and replaced activities. Test-retest analysis indicated a Kappa coeffi cient of agreement (k = 0.86; (p<0.001), suggesting temporal stability of the instrument. CONCLUSIONS: The Brazilian version of the Veterans Specifi c Activity Questionnaire showed evidence of reliability, according to the temporal stability criterion and adequate cultural content.


INTRODUCTION
Cardiovascular diseases (CVD) represent the main cause of death and incapacity worldwide, 28 and physical fi tness is an important prognostic factor and independent predictor of mortality. 13,21Different methods are used to measure the level of physical fi tness, both in asymptomatic and symptomatic patients.
The cardiopulmonary exercise test, 27 a direct method and gold standard to evaluate physical fi tness, is necessary and it enables maximum oxygen consumption (VO 2máx ) to be measured.However, although the direct measure is more accurate and considered as the main clinical measure, its use is not viable in the entire population.Financial, 4 physical and time limitations, or yet the risk of occurrence of a cardiovascular event in patients with CVD, restrict the use of the direct method. 20As an alternative to the effort test, indirect methods are frequently used, such as questionnaires and scales to assess aerobic fi tness and performance of movements that result in energy expenditure. 9,12The low cost, the ease of use, and the amount of information make such instruments interesting for clinical practice and research.
The Veterans Specifi c Activity Questionnaire (VSAQ) was developed by Myers et al 20 (1994) and validated through the test of its correlation with maximum oxygen consumption. 19This instrument is frequently used in studies conducted in North America to estimate aerobic fi tness as an isolated or associated predictor of the prognosis of patients with CVD.The VSAQ has also been employed to optimize individual protocols for the effort test, to reach maximum response to exercise in a period of 8 to 12 minutes, as recommended by the American Heart Association. 24Due to its applicability, this instrument was validated in other cultures, such as the Japanese population. 14e VSAQ is a short questionnaire, designed to determine the maximum level of daily physical activity according to the occurrence of cardiovascular symptoms, being originally applied to patients referred to the exercise test for clinical reasons.The VSAQ consists in a list of activities shown in progressive order, according to their Metabolic Equivalents of Task (MET). 19,20In the study on VSAQ validation, 20 authors also observed that the inclusion of the patient's age in the VSAQ score improved the capacity to predict patient's exercise tolerance.Thus, authors proposed a nomogram to be applied according to the result obtained with the VSAQ score, which is expressed by the following equation: METs = 4.7 + 0.97 (VSAQ) -0.06 (age).The equation refl ects the relative weight of age and that of the VSAQ score to predict the capacity to perform physical exercises. 19e VSAQ is adequate for use by the entire multidisciplinary team in the regular follow-up of cardiac patients to assess their broader aspects, in addition to the clinical aspect, such as the impact of the disease on daily life and the level of exercise tolerance, on which interventions should be based.This instrument was developed in the English language, in the context of the American culture.As the physical activity pattern of Brazilians is different from that observed in Americans, its application to the Brazilian context requires its cultural adaptation.This complex process involves stages, in addition to the translation of the instrument, 3 which consider the differences in health perception, cultural context and lifestyle of the population in question. 5,11e present study aimed to validate the content and to evaluate the reliability of the Veterans Specifi c Activity Questionnaire instrument, culturally adapted for use in the Brazilian population of cardiac patients.

METHODS
Before beginning the adaptation process, formal consent was obtained from the author of the original instrument.The cultural adaptation was comprised of four stages, according to the recommendations of Beaton et al 5 (2000) and Guillemin et al 11 (1993): cultural adaptation (translation and back-translation), content validity (assessment of semantic-idiomatic, cultural and metabolic equivalences), assessment of acceptability (pre-test) and assessment of instrument reliability, using the criterion of stability (test-retest).
The questionnaire was translated by three independent translators, who were native speakers of Portuguese and had a postgraduate degree in English-Portuguese translation, the Certifi cate of Profi ciency in English by the University of Cambridge and/or experience in the translation of specialized periodicals on the subject of this study.
Only one of the translators was informed about the objectives of the instrument and the concepts implied to promote cultural and idiomatic equivalence.The remaining translators did not receive such information to extract unexpected meanings from the original instrument. 3,5,11The three versions were compared by the translators and by the two fi rst authors of the study until consensus was reached.This comparison aimed to facilitate the conceptual and literary translation simultaneously, in addition to guaranteeing the detection of errors and ambiguous interpretations among translators. 3,5e fi nal version of the translation to the Portuguese language was back-translated to the original language by two other translators, who worked independently and did not participate in the fi rst step. 3,11These translators were bilingual speakers from the United States, fl uent in the original language of the instrument, who were not aware of the concepts and purposes that supported the instrument and did not have an academic qualifi cation in this area. 3,5,11The step of back-translation aimed to review data and ambiguous interpretation in Portuguese, guaranteeing the quality of cultural adaptation of the instrument of study. 5,11rifi cations of content validity, acceptability and reliability were performed in the university hospital, in cardiology outpatient clinics and in the exercise testing sector, situated in the city of Campinas, Southeastern Brazil, between August and October 2009.Participants had to have arterial hypertension, coronary artery disease and valvular heart disease, in an isolated or combined way, during outpatient follow-up, and at least one cardiovascular symptom had to be present.Patients with a record in medical charts or a report of cognitive defi cit, given by the patient himself or someone accompanying him (cognitive or memory impairment, induced by illicit substances, medications or neurological injuries/disorders), thus making it diffi cult for them to understand the instrument and responses to items, were excluded from this study.The instrument was always administered as an interview by the fi rst researcher.
The content of the instrument was validated in two steps.The translated version was submitted to the committee of specialists to assess its semantic-idiomatic and cultural equivalence, followed by the quantifi cation of content validity. 17Aiming to guarantee the greatest possible representativeness of activities to be replaced, two samples of participants were consulted: one was asked to indicate the activities uncommon in their routine in the translated instrument, while the other was asked to complete a 24-hour activity recall in order to identify the activities performed on a daily basis by the target population.Once redesigned, this instrument underwent a new assessment by another committee of experts in the fi eld of exercise physiology, who considered the metabolic equivalence of the activities included in the instrument, in addition to the conceptual equivalence.
A multidisciplinary committee assessed the semanticidiomatic and cultural equivalence of the instrument.This committee was comprised of bilingual individuals and specialists in the instrument's area of knowledge: 3 a nurse, experienced in research, teaching and healthcare in cardiology and in the use of measurement scales; a researcher-nurse with experience in the procedure of cultural adaptation of measuring instruments; a physical educator, with knowledge about exercise physiology; and one physiotherapist and physical educator, with experience in cardiology and physiology of exercise, in addition to the two fi rst authors.Judges were informed about the measures and concepts implied and received an instrument to assess the VSAQ.The committee conducted the review and comparison among the fi nal translations obtained: the original version, the unifi ed Portuguese version, and the two back-translations, aiming to obtain an intelligible fi nal version, with semantic-idiomatic and cultural equivalence. 3,10,11Items were individually assessed, and for each of them, the semantic-idiomatic and cultural equivalences were evaluated as: 1 = not equivalent, 2 = impossible to assess equivalence without the item being reviewed, 3 = equivalent, although needing minor changes, and 4 = completely equivalent.In this step, the judges pointed out the activities culturally not compatible with the target population.The instrument's content validity was quantified with the application of the Content Validity Index (CVI). 2,26The CVI indicates the proportion of experts who deemed the item as content valid. 2 The criterion proposed by Lynn 17 (1986) was adopted to interpret representativeness of agreement indices, according to which all four judges must agree with an item's content validity (CVI = 1), considering a signifi cance level of p = 0.05.The observation of a CVI ≤ 0.75 implies the automatic review of this item, because it means that at least one of the judges did not ratify its content validity.The items rated as "1" or "2" must be reviewed or eliminated. 2 The formula used to assess each item individually is represented below: 25

Total number of responses
The instrument's translated version was applied to 20 patients of the above mentioned outpatient clinic, on the day of their regular consultation in the cardiology outpatient clinic.The objective of this step was to identify which of the activities listed in the instrument were not common in their routine, aiming to replace them by other activities with the same metabolic equivalence.
A 24-hour recall was applied to 24 patients who had not participated in the previous steps to identify and include activities in the VSAQ that were more common in the Brazilian population.This recall was based on the methodology to investigate the dietary pattern (24-hour food recall) and consisted in obtaining verbal information about the physical activity pattern of the last 24 hours, including data on the type and duration of all physical activities performed on the previous day.This is a method which is usually well accepted by participants, with a short time of application and low cost, in addition to its not promoting changes in the behavior studied. 6The recall was administered as an interview, and the researcher immediately recorded all activities reported by the patient, in a sequence and according to the times in which they were performed.Such records were written down on a lined sheet, where hours were specifi ed.Patients were interviewed on different weekdays (Monday and Thursday) to assess common activities and their routine, both during the week and on weekends, except for holidays.
The sampling process in the steps of identifi cation of culturally incompatible activities and 24-hour recall followed the criterion of consecutive enrollment of all individuals who showed all the inclusion criteria and none of exclusion, until saturation was reached, i.e. interviews were suspended when the information provided by new research participants was not relevant to the material already obtained, in the opinion of the researchers involved. 8ke all activities mentioned in the questionnaire, those selected from the 24-hour recall had their metabolic equivalences identifi ed, according to Farinatti et al Another committee of judges was formed to assess the instrument's conceptual and metabolic equivalence: two professionals experienced in physiology of exercise and one from the area of cardiology, with clinical experience.
This committee identified activities that were not common in the routine of patients and replaced them by others that were metabolically equivalent. 7,15,18e fi nal version of the instrument underwent a pre-test with 15 patients to detect errors and to confi rm that both the instrument's presentation and all questions were intelligible.Moreover, practical aspects of instrument application and the time of application were observed.

RESULTS
The proportion of agreement among specialists is shown in Table 1.The 11 items that showed CVI < 1 (totaling 17 activities, because more than one activity was emphasized in certain items) were reviewed, according to suggestions made by judges and subsequently approved by the author of the instrument.
Table 2 points to changes made in the instrument after the fi rst assessment of semantic-idiomatic and cultural equivalences.
In the pre-test, nine activities were pointed out by patients as uncommon and replaced by others obtained from the 24-hour recall.
In the assessment of conceptual and metabolic adequacy of activities, the second committee of judges chose to attribute the lowest metabolic equivalence to four activities that showed distinct MET attributions in the three researched compendiums.There was 100% agreement among judges in terms of the metabolic equivalence of replaced activities.All replacements made are shown in Table 3.
The two pre-tests that followed were applied to two distinct samples, which showed systemic arterial hypertension (83.0%), acute coronary syndrome (46.0%) and myocardial revascularization (43.0%), either in an isolated or combined way.
The sample of the first pre-test, comprised of 15 patients, was characterized by the predominance of the female sex (73.0%), mean age of 59.9 (SD = 9.9) years, mean level of education of 4.8 (SD = 3.4) years, mean monthly per capita income of US$ 396.00 (SD = 452.00),and monthly family income of US$ 990.00 (SD = 679.00)(Table 4).
The fi rst pre-test evidenced that the statement of the Brazilian version was not intelligible to some patients.In addition, it was diffi cult to interpret the level of intensity of certain activities to associate speed in km/h with the rhythm of the walk or run.Thus, the statement was re-written and the intensity of activities began to be described as: slow, fast, very fast and extremely fast or mild, moderate and vigorous effort.After these changes, the second pre-test was conducted.
In the second pre-test, the modifi ed instrument was applied to other 15 patients, of which 53.0% were males, with a mean age of 56.5 (DP = 12.4) years, mean level of education of 6.0 (DP = 4.3) years, mean monthly per capita income of US$ 735.00 (DP = 707.00),and monthly family income of US$ 1,245.00(DP = 1,132.00)(Table 4).
In this, patients responded to the instrument without diffi culties, and fi ve minutes were necessary for its application on average.
For the assessment of reliability of the fi nal instrument (Attachment), the criterion of temporal stability was considered, with the use of the test-retest.This instrument was applied to 50 patients (Table 4) and re-applied in an interval of seven to 15 days.An index of agreement between the test and re-test of k = 0.86 (p < 0.001) was observed, evidencing the temporal stability of the Brazilian version of the VSAQ.

DISCUSSION
The adequacy of the Brazilian version of the VSAQ was assessed both in the perspective of specialists in the area of health and in that of the target population, seeking the greatest possible representativeness of activities for the population of interest.This step indicated 26 activities, which were not common in the routine of the target population, although compatible with the original instrument.This would compromise the instrument's capacity to more accurately assess the intensity of the activity during which the cardiovascular symptom occurs.
In the study on VSAQ adaptation to the Japanese culture, the stage of cultural equivalence was not performed. 14After the application of the Japanese version, it was observed that the activities included between 5 and 8 METs were not even once indicated by the patients; in addition, although certain patients had responded that they were not able to perform some group activities between 09-12 METs, they identifi ed activities of 13 METs as possible to be performed.This range of intensity included activities such as dancing and car washing, common in the Western-American lifestyle, although not frequent in the Japanese As the instrument aims to estimate the physical fi tness of participants, the accuracy of equivalence in METs of activities is essential.Thus, another essential step of the study was the assessment of metabolic equivalence of all activities to be replaced.
Metabolic equivalence, however, is not simple.Certain activities emphasized in the 24-hour recall showed different metabolic equivalences.The majority of activities show METs measured in laboratory or fi eld studies; other activities have MTEs estimated according to similar activities. 1 Moreover, there are variations in sex, age and physical fi tness level which account for the variability of energy expenditure. 23Thus, the committee of judges considered the target population of the Brazilian version of the VSAQ to establish the fi nal value of metabolic equivalence of activities in which there was a disagreement of METs.In this population, the aerobic physical fi tness is limited by the cardiovascular symptoms during the efforts, 23 resulting in less energy expenditure when compared to other populations, which are aimed at activities shown in other studies.Thus, the consensus reached was to use the lowest metabolic equivalence for each activity.
The adoption of simple and objective language is key to enable the wide use of the instrument in different educational strata.Consecutive adjustments were necessary until the language used in the statement was intelligible to participants, without the need to give additional explanations for the instrument to be completed.
With regard to reliability, the Brazilian version of the VSAQ showed evidence of temporal stability, given that kappa values between 0.6 and 0.8 are considered to be very good agreement. 16is adaptation of the VSAQ for its administration as an interview expands the scope of patients to which the instrument can be applied, surpassing the limits of the defi cit in reading and writing abilities.
In the Brazilian context, different validated instruments assess the routine performance of physical activities and consider their frequency, duration and intensity.These instruments enable individuals to be classifi ed as either active or sedentary, on different levels.In its turn, the VSAQ assesses the physical fitness of individuals when performing different routine activities, evidencing the level of limitation through the symptom.This is an original tool which can be used with the instruments of routine physical activity pattern, with great applicability in the context of cardiac rehabilitation, for example. 22us, it can be concluded that the Brazilian version of the VSAQ passed all stages of cultural adaptation process, obtaining relevant support for its content validity, according to the criterion of semantic-idiomatic, cultural and metabolic equivalences.In addition, there was strong evidence of reliability, according to the stability criterion, which enables the confi rmation of the importance of continuing the validation process of the Brazilian version of the VSAQ.
After adjustments, this instrument was applied to a new sample (n = 15).
The data collected were entered into an electronic spreadsheet (2003 Excel software) and analyzed in the SAS program, version 9.1.3,for test-retest agreement analysis, using Kappa coeffi cient.A signifi cance level of p ≤ 0.05 was adopted.This Project was approved by the Research Ethics Committee of the Faculdade de Ciências Médicas da Universidade Estadual de Campinas.All patients signed an informed consent form.

Table 1 .
Proportion of agreement among specialists, in terms of semantic-idiomatic and cultural equivalences for each item and respective Content Validity Indices.Campinas, Southeastern Brazil, 2009.

Table 2 .
Changes made to the instrument, according to metabolic equivalences and activities.Campinas, Southeastern Brazil, 2009.

Table 3 .
Activities that were culturally incompatible with the study population and respective replacements.Campinas, Southeastern Brazil, 2009.
METs: Metabolic Equivalents of Task.Aiming at a better distribution of the number of activities on each MET level, certain activities (from the original instrument) were only excluded, while others were replaced by different activities, compatible with the population.