Cross-cultural Adaptation of the Oral Anticoagulation Knowledge Test to the Brazilian Portuguese

Patients’ knowledge about oral anticoagulant therapy may favor the achievement of therapeutic results and the prevention of adverse pharmacotherapy-related events. Brazil lacks validated instruments for assessing the patient’s knowledge about treatment with warfarin. This study aimed to perform the cross-cultural adaptation of the Oral Anticoagulation Knowledge (OAK) Test instrument from English into Portuguese. This is a methodological study developed in an anticoagulation clinic of a public university hospital. The study included initial translation, synthesis of translations, back-translation, review by the experts committee and pre-testing with 30 individuals. We obtained semantic equivalence through the analysis of the referential and general meaning of each item. The conceptual equivalence of the items sought to demonstrate the relevance and acceptability of the instrument. The process of cross-cultural adaptation produced the final version of the OAK Test in Brazilian Portuguese entitled “Teste de Conhecimento sobre Anticoagulação Oral”. There was a suitable semantic and conceptual equivalence between the adapted version and the original version, as well as an excellent acceptability of this instrument.


Introduction
Warfarin is an oral anticoagulant widely used in Brazil and around the world to prevent and treat thromboembolic diseases.The management of this treatment is quite complex due to its narrow therapeutic range and wide dose-response variability, which increases bleeding risk 1,2 .
Several studies have indicated that patients with better knowledge about warfarin therapy have better stability of laboratory parameters, such as the International Normalized Ratio (INR) 2,3 .People who use this drug must have adequate knowledge about the therapeutic goal (indication and effectiveness), the use process (dosage, therapeutic regimen, administration method and treatment duration), safety (precautions, contraindications, adverse effects and interactions) and its preservation 4 .
The implantation of anticoagulation clinics (AC) is relevant in the healthcare systems, considering the morbidity and mortality observed in individual users of this drug 1 .Access to ACs creates better conditions for individualized care and educational process of the patient.The oral anticoagulation quality is strongly associated with the individual's level of knowledge about own pharmacotherapy 5 .However, there are significant gaps in knowledge about oral anticoagulation in patients treated with warfarin 6 .
Previous studies have shown that more than half of the patients have a knowledge deficit about treatment with warfarin [7][8][9] .Actions aimed at improving knowledge about anticoagulant therapy can significantly increase adherence to treatment and control of the INR 10 .However, these studies have substantial methodological limitations regarding the lack of use of a reliable instrument specifically validated to assess the patients' knowledge about anticoagulant therapy 11 .
The Oral Anticoagulation Knowledge (OAK) Test 3 was a validated instrument for English language and translated for use in Saudi Arabia 12 , Malaysia 13 and Qatar 14 .Studies demonstrated that the OAK Test is valid and reliable to measure the knowledge of users of warfarin in different cultures, which justifies the proposal to adapt this instrument to the Brazilian culture.Cross-cultural adaptation of instruments is of fundamental importance for epidemiological practice and is essential for generating reliable and comparable data, maintaining the semantic and conceptual equivalence between the original version and the adapted version 15,16 .
In a study conducted in the Medline, Em-base, Central, Scopus, Lilacs and SciELO databases covering the 1994-2015 period, using the descriptors "questionnaires", "patient medication knowledge" and "warfarin", no instrument for evaluating the level of knowledge about oral anticoagulation with warfarin and that has been correctly adapted for use in Brazil and evaluated for its psychometric properties has been identified.Therefore, this study aimed to perform the cross-cultural adaptation of the OAK Test instrument into Brazilian Portuguese, evaluating the semantic and conceptual equivalence of the items between the original instrument in English and the Portuguese adapted version.

Study design and target population
This is a methodological study based on the organization and analysis of data, designed for the evaluation and validation of research instruments and techniques 17 .Research was developed in an AC of a university hospital located in southeastern Brazil, which plays a regional reference role in medium and high complexity care within the Unified Public Health System.The target population included subjects with cardiovascular disease and indication for indefinite warfarin use.Inclusion criteria were 18 years of age or older, Brazilian nationality and duration of treatment with warfarin of over two months.Pretest participants were recruited and interviewed consecutively on AC service days.The cross-cultural adaptation was performed from October to December 2014.

Instrument
The OAK Test was developed in the Northeastern United States 3 and used consistent methods to assess its validity and reliability.It consists of 20 questions with four answer alternatives, with only one correct choice.Each patient's correct answer equals one point, and the final score ranges from zero to 20 points.A higher score indicates a better level of knowledge about oral anticoagulant therapy.

Evaluation of the semantic and conceptual equivalence of the items
After authorization from authors of the original version, the cross-cultural adaptation of the OAK Test instrument into Brazilian Portuguese was planned according to the method recommended by national and international literature for instrument adaptation 15,[18][19][20] .The semantic evaluation was developed as follows: initial translation, synthesis of translations, back-translation, review by the experts committee and pre-testing 16 (Figure 1).
The first stage consisted of two translations of the original instrument in English into Brazilian Portuguese (T1 and T2).Two bilingual translators whose mother tongue was Brazilian Portuguese performed the translations independently.A translator had training in the area of health, clinical experience and knowledge about the terms and concepts of the instrument.The other translator had no medical or clinical training and no technical knowledge about the analyzed concepts.
T1 and T2 translations were compared and discrepancies identified.The two translators and four other researchers participated in the synthesis of translations.The group used the original instrument and the two versions translated into Brazilian Portuguese and, after consensus, produced a common translation (T12).At this stage, we tried to identify possible difficulties in understanding the instrument.The meaning of words in the different languages (English and Brazilian Portuguese) was thoroughly analyzed with a view to obtain similar effects in individuals of different cultures.
The T12 synthesized translation was back-translated into the original language of the instrument (English) by two independent translators, foreigners, born and literate in an English-speaking country, with linguistic and cultural mastering of both English and Brazilian Portuguese.The OAK Test was then adapted.Translators were not aware of research's objectives and did not have access to the original instrument.
Subsequently, a review of the back-translated versions (VI1 and VI2) and the synthesis Portuguese version (T12), using as reference the original version by a committee of experts composed of four researchers, the four participating translators, two health professionals with experience in the management of oral anticoagulation and a linguist.The establishment of this committee was fundamental to obtain a consensus regarding the semantic and conceptual equivalence of the items.
Semantic equivalence refers to the meaning equivalence of words, or to the correct translation of items.The equivalence between the original instrument and the back-translated instrument was evaluated from the perspective of the referential meaning of the terms and constituent words (similarity as to the literal meaning of the constituent terms of the assertive pairs) and the general meaning of each item (similarity as to the idea conveyed by assertions) 21 .
To analyze the meanings, experts used a specific form, designed to mask the origin of the evaluated items.For the analysis of the referential meaning, a visual analog scale 22 was chosen.The equivalence between assertive pairs was evaluated continuously, with results varying from zero to 100%, using the following categories: < 80% = non-equivalent, 80-89% = almost equivalent, 100% = maximum equivalence 22 .
The evaluation of the general meaning was developed using four levels for classification, namely: unaltered (UA), slightly altered (SA), highly altered (HA) or completely altered (CA) 22 .
The analysis of the conceptual equivalence of the items seeks to demonstrate whether they are relevant and acceptable in the original and adaptation-targeted cultures.Therefore, the opinion of experts and the preliminary test of the adapt- ed version in population samples, as described in the next step, become of great value 21 .Pre-testing was developed by one of the researchers who individually interviewed 30 patients in a suitable place, checking their understanding regarding the adapted version 16 .A questionnaire adapted for use in Brazil 23 was used, which addresses the general impression about the tool in terms of clarity, completion time and possible issues.Sociodemographic data were collected, including sex, age and schooling to characterize the participants.

Statistical analysis
The database was validated by double entry in the EpiData software program (version 3.1, EpiData Assoc, Denmark) and analyzed in the Statistical Package for Social Science program (SPSS for Windows, version 20.0, SPSS Inc., Chicago, Illinois, USA).We performed descriptive statistics of sociodemographic variables using frequency and central tendency measures.

Ethical aspects
This study was conducted in accordance with the Declaration of Helsinki 24 .The Research Ethics Committee of the Universidade Federal de Minas Gerais approved the research project.All participants signed an informed consent form prior to the onset of the research.

Results
The process of cross-cultural adaptation was systematically performed.The stages of translation, synthesis and back-translation were performed without major difficulties and, therefore, there were no significant modifications.
The semantic equivalence was evaluated by the review of the committee of experts, in which the equivalence between items from the back-translated versions and those of the original instrument was evaluated, which evidenced that instruments remained equivalent.As described in Chart 1, 17 (85%) of the 20 items showed maximum equivalence and general significance remained unaltered (UA).
The conceptual equivalence of the items was obtained by the analysis of experts and by the pre-testing.Chart 2 shows the main changes, highlighted in bold, made by the experts committee in the synthesis version translated into Portuguese (T12) and the pre-final version obtained by consensus.
In question 1, in order to keep the original meaning, we chose to use the expression in the sentence "forget about" and in the fourth alternative of response, instead of the verb "to observe", we opted for "be careful with".The word "Coumadin ® ", which corresponds to one of the trade names of warfarin, was also removed from the translated instrument, and the Brazilian Common Denomination was used 25 .
In question 3, the discussion took place around the English term "healthcare provider".In the synthesis version translated into Portuguese, the term "health service" was first used, but because it was a broad term, it was replaced with the expression "which monitors your treatment".We made this change in the other questions containing this term.
Concerning question 7, the term "PT/INR (prothrombin time)" was discussed.It was taken into account that most of the target population only knows the abbreviation INR to designate the laboratory examination for monitoring oral anticoagulation.The term PT (prothrombin time) was then suppressed in this question and wherever it appeared.
Regarding question 9, two expressions identified that evidenced comprehension difficulties by the target population were altered."Expected range" was replaced by "desired range" and "rash" by "skin reactions".
In question 10, to facilitate the respondent's understanding, the acronym "AAS" was added in the statement because of its common use to designate the acetylsalicylic acid drug.In the last response option, the phrase "you increase your dose" was modified to "increase of your dose of warfarin", avoiding the interpretation of dose self-management by patients, which is not provided for in the clinical protocol of the AC at hand.
According to the experts' analysis, in question 13, the expression "you take your dose of warfarin and alcohol separately" in the synthesis version translated into Portuguese is ambiguous, in which the respondent could interpret it as "simultaneous intake of alcoholic beverages and warfarin tablet".Thus, we opted for the modification of the expression to "you take your dose of warfarin and alcohol at different times".
The terms "effectiveness", "interaction" and "adverse effects", which appear throughout the questionnaire were identified by the experts committee as difficult to understand for the tar- With the pre-testing of the pre-final version of the OAK Test in Brazilian Portuguese via faceto-face interview, it was possible to verify the adequate level of language used and comprehend the translation.The instrument's mean time of response was 10 minutes.
Regarding the characteristics of the pre-testing participants (Table 1), most were female (66%), median age 55 years and incomplete elementary school education (67%).
There were no difficulties in understanding the items during interviews.All 30 (100%) par-ticipants answered the general impression questionnaire about the instrument and rated it as good, evaluating all issues as important for anyone using warfarin.Most participants (23; 77%), classified the subjects as easy-to-understand, and only seven (23%) participants classified the questions as fairly understood.
No participant made suggestions to change or add questions.Thus, at the end of pre-testing, there was good acceptance of the instrument among the participants, and there was no need to modify items.

Conclusions
The cross-cultural translation and adaptation process of the OAK Test for Brazilian culture followed internationally recommended steps and was successfully carried out.The results obtained showed that the Brazilian and American versions are conceptually equivalent.
The application of instruments using recognized scientific methods will allow the analysis of the relationship between patients' knowledge and quality of oral anticoagulation control.The results obtained may help in the identification of deficits and in the structuring of health education activities to improve knowledge about pharmacotherapy and, consequently, favor a successful treatment.

Collaborations
MFS Praxedes, SM Paiva, DD Ribeiro, MS Marcolino, MHNG Abreu and MAP Martins contributed during the review process of the manuscript and the approval of its final version.

Figure 1 .
Figure 1.Flowchart of the cross-cultural adaptation process of the instrument Oral Anticoagulation Knowledge (OAK) Test.

Chart 1 .
Semantic equivalence between the OAK Test version in Brazilian Portuguese and the original in English.Oral Anticoagulation Knowledge; TP, Tempo de Protrombina; RNI, Relação Normalizada Internacional; SA, slightly altered; UA, unaltered.A1 -Evaluation in terms of the percentage of concordance of the referential meaning.A2 -Evaluation of the general meaning.Comparison between the synthesis version of OAK Test translated into Brazilian Portuguese and the pre-final version after evaluation by the experts committee.
get population.However, keeping them was considered an appropriate procedure and, if there were any issues by the respondent, the interview-er would clarify them until they are fully understood and the question is answered in a convincing manner.