The Vietnamese version of the constitution in Chinese medicine questionnaire (CCMQ): validity and reliability

This cross-sectional study was carried out in 2 phases in Ho Chi Minh City from January to June 2021. In the first phase, we translated the questionnaire into Vietnamese language, in concurrent with assessing the content validity and face validity. The second phase assessed the reliability and validity of the Vietnamese version of CCMQ. During phase 1 the process of translation and cross-cultural adaptation of the CCMQ was performed as described by Guillemin et al [9, 10] and Wild et al [11] including 5 steps: forward translation, reverse translation, summation, expert review, and pilot trial. Step 1 – forward translation: The original version of the CCMQ was independently translated to Vietnamese by two local translators fluent in Chinese. One of them was familiar with the CCMQ while the other was not. In particular, a medical practitioner who has experience of evaluating BC provided a translation from a professional point of view while another person who had no experience in the field of research made a translation in general public. The translated versions of the CCMQ were designated T1 and T2. Step 2 – reverse translation: T1 and T2 were separately back-translated to Chinese and designated as BT1 and BT2. The reverse translation is conducted by two translators who were blinded to the original CCMQ version and did not participate in the previous period. Step 3 – summation: Above four translators compared T1, T2, BT1, BT2 with the original version of the CCMQ to correct all discrepancies and synthetize the “pre-final 1” version of the CCMQ. Step 4 – expert review: This “pre-final 1” version was reviewed by an expert board, who graduated from a university majoring in Traditional Medicine with a master’s degree or higher and ≥ 5 years of experience in practicing medical treatment or doing research on Traditional Medicine. Lynn (1986) [12] recommended using between 5 and 10 experts in the content validation process; therefore, we selected 7 Traditional Medicine experts from January 15, 2021 to January 31, 2021 in our study. Content validity is integrated into the translation stage and evaluated by experts as suggested by Conway et al [13]. The content validity indexes (CVI), the proportion of subjects who gave a positive rating, was assessed based on clarity and relevance of each item [12]. Subsequently, a “pre-final 2” version of the CCMQ was produced, this version was used in the step 5 – pilot trial. The purpose of pilot trial was to identify problematic questions in the questionnaire and offer solutions to make such questions easier to understand. Pilot trial assessed the face validity of CCMQ including understandability, clarity, acceptability, and purpose [14, 15]. Finally, a panel of nationwide 10 traditional medicine specialists (with a doctorate, second-degree specialty, or over; working for at least 5 years in the field of Traditional Medicine) reviewed and adjusted to create a complete final Vietnamese version of CCMQ. The second phase was to assess the reliability and the validity of Vietnamese version of CCMQ. The reliability included internal consistency and test-retest reliability while the validity assessed criterion validity. Phase 2 conducted a descriptive cross-sectional study with inclusion criteria involving permanent residents in Ho Chi Minh City who aged 18 years old or older are willing to participate in research with the ability to independently communicate and follow the instructions. Exclusion criteria were people with mental illness, behavior problems, serious illnesses, or inability to understand or complete the study questionnaire. All volunteers who took part in this study were recruited from the general population of 24 districts of Ho Chi Minh City. Study subjects participated in an interview and had filled in the Vietnamese version of the CCMQ questionnaire and SF-36 health survey to assess criterion validity. After that, they would be re-interviewed after 4 weeks for the test-retest reliability assessment [16]. The whole process of this study is summarized in Figure 1.

Figure 1. Two phases of study

Download Original Figure

In phase 1, the pilot study was conducted on 30 people in the residential community of Ho Chi Minh City. We recruited volunteers from the general population in the Ho Chi Minh City with the same inclusion and exclusion criteria as the official study. Participants were interviewed using a structured set of prepared questions, the subjects were asked to rate the ease of understanding of the Vietnamese version of CCMQ, assess the difficulty in the process of answering, express the purpose of the scale, and offer the adjustment output (if any). Research subjects could directly write comments on the answer sheet. In phase 2, to evaluate the reliability and validity of the Vietnamese version of CCMQ, a cross-sectional study was performed on 455 subjects with the contents of the Vietnamese version of CCMQ and the Short Form Health Survey SF-36. We also recruited volunteers from the general population in Ho Chi Minh City. Participants independently completed pre-prepared paper-based edited questionnaires that included the Vietnamese version of the CCMQ and the SF-36 quality of life survey. It was required that all questions must be completed prior to submission. To assess the test-retest reliability of the CCMQ questionnaire, all volunteers received an invitation to participate in the second survey carried out at 4 weeks after the initial assessment and 210 subjects agreed to take part in. Then, 4 weeks later, those participants were asked to fill out the same questionnaire again on paper. Data were collected only with the agreement of volunteers using informed consent.

In step 4 – expert review of the first phase, Lynn and Rubio recommend a minimum of 3 experts and a maximum of 10 experts [12]. Our study had 7 practioners. In step 5, Beaton et al recomended a sample size with a minnimum of 30 participants and maximum of 40 participants [17]. Our sample size was 30. The sample size estimations were based on the subject to item ratio, which is a method that is frequently used to determine the required sample size needed for scale validation [18]. A review of 114 studies on newly-developed scale validation found that the subject to item ratio was used to determine sample sizes in 92% of the articles, and the median subject to item ratio was 4 with a minimum of 1 and a maximum of 26 [19]. There were 60 items in the Traditional Medicine constitution scale, and the subject to item ratio was set at 7, therefore 420 subjects were needed. Our study had 455 subjects.

Constitution in Chinese Medicine Questionnaire (CCMQ). The Constitution in Chinese Medicine Questionnaire (CCMQ) consists of 60 items to classify a person into one or more of nine BC types: Neutral (8 Items), Qi-deficiency (8 Items), Yang-deficiency (7 Items), Yin-deficiency (8 Items), Phlegm-dampness (8 Items), Dampness-heat (6 Items), Blood-stasis (7 Items), Qi-depression (7 Items), and Inherited-special (7 Items). Coexistence of multiple imbalanced BC types was possible which is consistent with the Traditional Medicine theories. The scoring algorithm proposed in the original CCMQ was adopted in this study. A higher score in the CCMQ BC scale indicates a higher likelihood of the specific BC type, and a score of 30 is set as the threshold for the case definition [4].

SF-36: in order to evaluate criterion validity, we conducted the survey of SF-36 (Vietnamese version) which includes the physical and mental component summary (abbreviated by PCS and MCS later) at the same time. Scores range from 0 to 100, with higher scores indicating better subjective health status. We used the SF-36 questionnaire with Vietnamese translation, which has been translated, culturally adapted, and assessed as a reliable tool to assess the quality of life for the Vietnamese population [8]. The Vietnamese translation of SF-36 includes 36 questions measuring 8 areas: general health, physical functioning, role physical, bodily pain, mental health, role emotional, vitality, and social functioning. In which, the first 4 areas assess the physical health, and the rest assess the mental health. The quality of life score is calculated as the average score of 8 areas. The results would be converted to a scale of 100 according to the convention table. A higher score reflects a better quality of life and vice versa [20]. It was hypothesized that subjects in the Neutral constitution would have the highest SF-36 scores because they were thought to be the healthiest. There is a positive correlation between the Neutral type and SF-36 scale, the higher scores of both of them indicate a better health. The eight unbalanced constitutional types of CCMQ are negatively correlated with SF-36 scale, so these types have a close association with lower quality of life. The SF-36 questionnaire was used in the study to determine the validity of the Vietnamese version of the CCMQ questionnaire [5].

Values were presented as Mean ± Standard deviation. Validity in this study included content validity, face validity, and criterion validity. Reliability in this study included test-retest reliability and internal consistency. In which, content validity and face validity were integrated into the first phase of the study; criterion validity and reliability (test-retest reliability and internal consistency) were carried out in phase 2 of the study. Specifically, in step 4 - professional assessment of phase 1, a content validity assessment form was sent separately to each expert along with an open letter instructing how to evaluate. Either relevancy or clarity was assessed for each question according to its role based on the Likert scale of 4 choices with the following meanings: the first is “Unusable, not meant to be (or not clear)”, the second is “Not usable yet, needs a lot of tweaking”, the third is “Can be used with a few minor tweaks”, the fourth is “Can be used without modification” [21, 22]. The content value is determined through the Content Validity Index for Items (I-CVI). The I-CVI index is calculated as the ratio between the number of experts assessed “usable” and the total number of experts. The threshold value of I-CVI was 1.00 when the number of judges was 5 or less, and 0.80 when there were 6 or more judges. Our study had 7 medical practitioners, corresponding to the I-CVI value greater than 0.80, indicating adequate levels of clarity [23]. The face validity was determined by 30 volunteers in step 5 of phase 1 (pilot trial). The face validity included 4 criteria: The ease of understanding of the Vietnamese version of the CCMQ scale; The difficulties in answering questions in the scale; Proposing to adjust according to the subject’s point of view to improve the quality of the translation and be suitable for Vietnamese people; Opinions of the research subjects about the purpose of the CCMQ scale and the issues mentioned in the content of the CCMQ scale. Criterion validity was assessed in phase 2 of the study, namely Concurrent validity, which was assessed using bivariate correlation analyses between SF-36 and the scores of CCMQ in order to verify whether the results of the scale were consistent with traditional medicine theory.

Reliability was also assessed in phase 2 of the study. Internal consistency was evaluated by Cronbach’s alpha coefficient for each of the 9 sub-scales. In which, Cronbach’s alpha coefficient was accepted when the threshold was ≥ 0.70 [24]. In contrast, Cronbach’s alpha coefficient was too high suggesting that some questions in the scale overlapped each other or had the same idea, the maximum recommended threshold was ≤ 0, 90 or ≤ 0.95 [25]. Test-retest reliability was evaluated by the weighted Kappa coefficient for each question and Spearman correlation coefficient for each sub-scale. It was generally said to be fair if a weighted kappa coefficient was 0.4 – 0.75; and good if it was 0.75 or greater [26, 27]. For Spearman correlation, it was said to be good if it was 0.6 or greater [28, 29]. Statistics analysis was performed by SPSS 20.0 and the significant level was set at p <0.05.

This study was approved by Council of Ethics in Biomedical Research at University of Medicine and Pharmacy at Ho Chi Minh City on January 28^th 2020, No. 52/HĐĐĐ-ĐHYD. All the participants signed an informed consent form in which the personal identification of research objects was not reported (name, address).

The “pre-final 1” of Vietnamese version of CCMQ was created after the completion of 03 steps – forward translation, reverse translation and summation. The content validity was considered in step 4 (Expert review) with a group of experts consisting of 7 Traditional Medicine practitioners, including 4 Doctors of Medicine and 3 Masters of Medicine with the average number of years of experience was 9 ± 1.4 years. There were 57 questions in the CCMQ which have satisfactory CVIs (I-CVI ≥80%) with clarity, consistency of response options, and relevance with health in all items. Among 3 questions with CVI <80%, question 3 “Bạn có dễ bị hụt hơi (khó thở được) không?” (Did you suffer from shortness of breath (fast breathing or difficulty in breathing)) has the lowest I-CVI of 57%, some comments that “khó thở” (shortness of breath) is easily misunderstood as pathological dyspnea, “khó thở được” (difficulty in breathing) is unclear. Question 1 “Bạn có cảm thấy sức khỏe dồi dào không?” (Were you energetic?) and question 26 “Bạn có bị ho hay suyễn do chuyển mùa, thay đổi nhiệt độ hoặc có mùi khác thường không?” (Did you have a cough or asthma caused by seasonal changes, changes in temperature, or an unusual smell?) were evaluated by some traditional medicine doctors with an I-CVI of 71% (Table 2). After being considered and reviewed by practioner committee, the pre-final 2 version was conducted and used for pilot trial. Step 5 – pilot trial: 30 participants included 13 males (43.3%) and 17 females (56.7%). Subjects aged 18 – 82 years old, the mean age was 48.5 ± 18.8. The educational level of participants ranged from nil to tertiary. Evaluating the face validity was based on the responses of 30 participants. All subjects agreed that the contents of the questions in the scale were suitable for the previously announced research with the purpose of assessing the BC types of people in the community, showing the face validity of the scale measured well. Participants had no difficulty in reading and understanding the contents of the questions and were able to complete all 60 questions within 15 minutes. In which, 76.7% of the participants rated the scale from “easy to understand” to “very easy to understand”. Participants commented to edit 5 sentences. Question 50 “Rêu lưỡi của bạn có dày nhớt hoặc hơi dày không?” (Was your tongue coating sticky or slightly thick?) was evaluated, the main feedbacks were that such a question is hard to understand the words “rêu lưỡi” (tongue coating), “nhớt” (sticky) and “hơi dày” (slightly thick). Question 4 “Bạn có dễ hồi hộp đánh trống ngực không?” (Did you get palpitations easily?), question 35 “Màu môi của bạn có đỏ hơn người bình thường không?” (Were your lips redder than others?), question 53 “Bạn có thể thích ứng với những thay đổi của môi trường tự nhiên và xã hội bên ngoài không?” (Could you adapt yourself to external natural or social environment change?) and question 56 “Bạn có cảm thấy phân của mình dính và đi tiêu có cảm giác không hết phân không?” (Did you pass sticky stools and /or feel that your bowel movement is incomplete?) were rated as unclear. These comments were acknowledged and then edited by an expert board of 10 Traditional Medicine specialists nationwide to complete the final content of the Vietnamese version of CCMQ. An expert board of 10 nationwide Traditional Medicine specialists completed the final content of the Vietnamese version of CCMQ. This panel of 10 experts included 5 experts from the North of Vietnam and 5 from the South of Vietnam; there were 2 Associate Professors, 9 Doctors of Medicine and 1 Second Degree Specialist with the average number of years of experience was 11.8 ± 3.9 years.

455 subjects included 198 males (43.5%) and 257 females (56.5%). The average age was 30.4 ± 13.1 years, the youngest was 18 and the oldest was 87 years of age. Among 455 subjects, 98.9 % of subjects could be classified into at least one BC type. The average time for response was 12.3 ± 4.7 minutes (Table 3). The score of each sub-scale of CCMQ was shown on Table 4. During the follow-up phase, there were 210 responders with 93 males (44.3%) and 117 females (55.7%). The test-retest reliability of the Vietnamese CCMQ varied 0.63 to 0.9 for each of the 9 sub-scales and 0.40 to 0.68 for each of the 60 questions (Table 5). The internal consistency of the CCMQ scale is shown in Table 6. The standardized Cronbach’s alpha coefficient varied from 0.70 to 0.83 for each of the 9 sub-scales. Regarding the criterion validity, the correlation coefficient between the score of Neutral type and SF-36 scale was 0.61 in the physical health score (PCS), 0.7 in the mental health score (MCS), and 0.67 in the quality of life score SF-36. The correlation coefficient between the scores of 8 unbalanced types and the SF-36 scale is from -0.27 to -0.52 in a physical health score (PCS), from -0.33 to -0.6 in a mental health score (MCS), and from -0.31 to -0.57 in quality of life score SF-36 (Table 7).