2.1 Study design
A single-center, cross-sectional trial was conducted to evaluate the reliability and validity of the C- BrQ. This study included AIS patients from all over the country. They were recommended by the spine surgery department of the local hospital to come to our clinic because they did not meet the surgical indications. The diagnosis was made by the attending physicians of our institution based on standardized diagnostic assessment results, including history, physical examination and X-ray.
2.2 Participants and sample size
The target population for this study was patients with moderate AIS who met the current indications: age 10 to 17 years, a cobb's angle between 25°and 40°, skeletal immaturity with a 0-3 Risser stage and wearing a brace for at least 3 months. Study exclusion criteria were unwillingness to participate, previous scoliosis fusion surgery, psychiatric disorders, mental retardation and previously received other types of brace. In addition, subjects with incomplete questionnaires were not included.
According to the guidelines of COSMIN Risk of Bias checklist, the sample size should be more than 100 cases. And considering that it should be seven times of 34 items of continuous variable, the sample size was determined to be 250.
2.3 Adaptation of BrQ
BrQ contains 34 Likert scale items and covers eight domains (including general health perception, physical functioning, emotional functioning, self-esteem and aesthetics, vitality, school activity, bodily pain and social functioning). It is specifically designed for children and adolescents between 9 and 18 years old to fulfill by themselves. The scoring of BrQ is as follows:
– For items 4, 5, 6, 12, 14, 15, 16 and 17: “Always”, 5 points; “Most of the time”, 4 points; “Sometimes”, 3 points; “Almost never”, 2 points; “Never”, 1 point.
– For items 1, 2, 3, 7, 8, 9, 10, 11,13, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33 and 34: “Always”, 1 point; “Most of the time”, 2 points; “Sometimes”, 3 points; “Almost never”, 4 points; “Never”, 5 points.
The score for each item is multiplied by 20, and the total score of the 34 items is divided by 34. Therefore, the minimum score is 20, and the maximum score is 100. The higher the score is, the better the QoL is. The total score of a domain can be divided by the number of items it comprises to obtain the subscale score of each domain13.
The BrQ was adapted according to the International Quality of Life Assessment (IQOLA).15 Two translators independently translated the original version from Greek into Chinese. One translator specializing in medical translation was in charge of the whole adaptation process. The other translator with no medical background was blind to the project. Then, the two translators and the authors of this paper compared the two translated versions and combined them into one. Next, two Greek translators who were unfamiliar with the original version translated the combined Chinese version back into Greek to ensure the two versions were equivalent and pinpoint the mistranslations. Lastly, a committee consisting of orthopedic experts, translators, statisticians and psychologists co-examined the Chinese version. After reaching a consensus, the Chinese version of BrQ (C-BrQ) was established.
2.4 Prefinal version test
A preliminary experiment was carried out before the formal study. During the first pilot study, 28 Chinese-speaking AIS patients (25 girls, 3 boys) aged 10-17 years completed the preliminary version of C- BrQ in the Department of Sports Medicine of Guangzhou Sport University. The patients filled out a standardized feedback form about unclear phrases and understanding difficulties in the C-BrQ. In case of ambiguity, the expert committee rephrased the expressions in the questionnaire based on the comments, and then conducted further investigation and review. All ambiguities were solved in the final version of the questionnaire. After the pilot study, the majority of the subjects understood and completed all questions in the questionnaire within 10 minutes. The expert committee made minor changes in three questions (28,29, 30) based on the results of the interview. The "friends" and "peers" in questions 28/29/30 of the questionnaire were changed into "classmates", because 10-year-old children tended to think that the concepts of friends and classmates in school were similar. A second pilot study among 14 participants showed no further equivocality and so this version emerged as the final one. Subjects in these pre-trials were also included in the final part of the study.
2.5 Measurement of outcomes
A total of 250 AIS patients, who underwent the Chêneau brace treatment from the outpatient clinic of Guangzhou Sport University between June 18, 2019 and August 20, 2020, were surveyed in the study. Informed consents were obtained from the participants and their parents. All participants were asked to complete C-SRS-22 and C-BrQ independently in the waiting room of the clinic before meeting the orthopedist. During further review, 6 cases were diagnosed with scoliosis of other causes (such as syringomyelia, unequal length of lower extremities, etc.), 9 cases had a Risser sign greater than grade 3, 3 cases were with cobb's angle greater than 40°, and 15 cases were incomplete. Therefore, 217 cases were finally included in the study.
C-SRS-22 is the Chinese version of SRS-22 and its reliability and validity has been proved.4 It contains 5 dimensions: function activity level, pain, mental health, self-image and management satisfaction. A 5-point scale ranging from “1” to “5” is applied in each item. The total score ranges from 22 to 110 points; lower scores are associated with poorer QoL.14
All participants underwent two tests. To avoid the memory effect, the two tests had a 7-day interval. For those who did not complete the second questionnaire, their responses in the first questionnaire were excluded from the test-retest reliability analysis but used for other analyses in the study.
2.5 Statistical analysis of data
Demographics and characteristics of the participants were summarized using frequency and percentage; mean and standard deviation(SD) were used appropriately. A test–retest design was used to measure the temporal stability of each domain with the intraclass correlation coeffificient (ICC). ICCs between 0.70 and 0.80, and >0.80 indicated good and excellent reliability, respectively. Then, as for content analysis, data were investigated (mean, SD, and range) and evaluated for outliers. The floor and ceiling effects (>10% of the possible minimum and maximum scores) were calculated. To assess the reliability, Cronbach’s alpha was used to evaluate the internal consistency of each domain in C-BrQ and C-SRS-22. Poor internal consistency was suggested if Cronbach’s alpha was <0.70; 0.70-0.80 indicated good internal consistency; and >0.80 showed excellent internal consistency. Then, the C-BrQ domains were compared with the C-SRS-22 domains using Pearson's correlation coefficient to assess the concurrent validity. The Pearson's correlation coefficients of less than 0.50, 0.50 -0.70, and more than 0.70 indicated poor, good, and excellent concurrent validity, respectively.
The level of statistical significance was set at 0.05. Statistical Package for the Social Sciences (SPSS) version 20.0 (IBM Corp, Chicago, IL) was applied for statistical analyses.