2.1 Translation and cross-cultural adaptation
The original version of the IdFAI questionnaire was translated according to the principles reported in previous studies [23,24]. In brief, the translation included the following several steps. (1) Forward translation: Two native Chinese speakers highly familiar with the English language (one translator was an orthopedist in the hospital, while the other was a professional translator without any medical background) were asked to translate the IdFAI questionnaire into simplified Chinese version. (2) Discussion I: All the investigators discussed the two independently translated versions, and finally the two versions were merged into the preliminary version of the SC-IdFAI questionnaire. (3) Backward translation: Two native English speakers independently translated the preliminary version of the SC-IdFAI questionnaire into the English language. (4) Discussion II: The versions were discussed by the investigators again to address the discrepancies, ambiguities, and other language expression issues in the preliminary version of the SC-IdFAI questionnaire. Then the prefinal version of the SC-IdFAI questionnaire was obtained. (5) Pretesting: Twenty patients with CAI were asked to complete the prefinal version of the SC-IdFAI questionnaire. The questions reported by the 20 patients during the completion of the scale were also recorded. Afterward, a third discussion was conducted among the investigators to conduct the final, specific modifications to the questionnaire, and then the final version of the SC-IdFAI questionnaire was obtained.
2.2 Patients and data acquisition
The consecutive native Chinese-speaking patients with CAI treated in our hospital (the General Hospital of Western Theater Command) between February 2016 and March 2018 were recruited. The inclusion criteria were as follows: (1) patients aged >18 years, capable of signing for themselves; (2) Chinese native speakers who could independently read and complete the questionnaire; and (3) patients with a history of at least two severe ankle sprains who reported episodes of a feeling of ankle instability and/or ‘‘giving way’’ and/or chronic pain in the ankle during daily life or sports activity. The exclusion criteria were as follows: (1) patients with a history of previous surgery of musculoskeletal structures and fracture requiring realignment of either limb of the lower extremity; (2) patients with acute injury to the musculoskeletal structures of the other lower-extremity joints in the previous 3 months; and (3) patients with other chronic inflammatory diseases in the lower limbs, which might affect the ankle function. The patients who met all the eligibility criteria and volunteered to participate were included in the study. In addition, the number of included participants also met the sample size criteria of PROMs, as recommended by Terwee et al. [25]. In brief, the sample size for internal consistency analysis should be ≥100, and the sample sizes for floor or ceiling effects, reliability, and validity analyses should be ≥50. All the participants carefully read and signed the informed consents. The study was approved by the Ethics Committee of the General Hospital of Western Theater Command.
The patients were asked to provide demographic information, such as sex, age, and weight, on the first day of enrollment and independently complete the following four questionnaires: SC-IdFAI, FAAM, SF-36, and CAIT-C (for another research) in a quiet meeting room. One week after the first filling-in, on the day before starting the physical therapy, they completed the SC-IdFAI questionnaire for the second time to evaluate the test–retest reliability of the scale. Patients having related treatment in the previous week were excluded. Finally, the patients who volunteered to receive the 8-week physiotherapies in the hospital were asked to complete the SC-IdFAI questionnaire for the third time to help in assessing the responsiveness.
2.3 Instruments
IdFAI is a scale specific to the assessment of patients with CAI. It was developed by Matthew Donahue et al. based on the CAIT and AII [14]. It combines the advantages of both CAIT and AII scales, and thus has its own specific advantages as follows: (1) the IdFAI questionnaire contains 10 questions, which are easily understood, and the patients need very short time to complete them; and (2) the questionnaire has very high accuracy in the assessment. Using the IdFAI questionnaire alone can result in higher accuracy than CAIT or AII [15]. The IdFAI questionnaire consists of three subscales as follows: (1) factor 1, the history of ankle sprains (items 5, 6, 7, and 10); factor 2, the presence and severity of ankle instability (items 1, 2, 3, and 4); and factor 3, the functional performance in daily living and other physical activities (items 8 and 9). As the first question is a nonchoice question, the answer is not included into the total score, while the scores of the other nine questions are added to obtain the final score. The total score ranges from 0 to 37, with higher scores indicating poorer ankle functions. Generally, if the total score is >10, the patients are considered to have CAI [20].
FAAM is a region-specific scale designed specifically for the assessment of the feet and ankle functions [26]. FAAM contains two subscales, namely activities of daily living (ADL) and sports. The ADL subscale contains 21 questions, while the sports subscale contains 8 questions. The score for each question ranges from 0 to 4. Thus, the range of the total score of the ADL subscale and sport subscale is 0–84 and 0–32, respectively, with higher scores indicating better functions. Although FAAM is region-specific, and not a disease-specific scale, previous studies already demonstrated that FAAM had high validity when used in patients with CAI [27]. SF-36 is a versatile scale for assessing the quality of life, which contains 35 questions in the 8 subscales. The first four subscales can be categorized as “physical subscales,” which mainly assess the physiological functions of the patients, while the latter four subscales can be categorized as “mental subscales,” which mainly assess the mental status of the patients. Each subscale of the SF-36 scale has a specific scoring method, while the final score is converted into centesimal grade. The higher score of the SF-36 scale indicates the better mental status of functions [28]. Both these scales have been translated into Chinese versions and demonstrated to have high validity, reliability, and responsiveness [29,30].
2.4 Psychometric assessments and statistical analysis
The validity, reliability, and responsiveness of the SC-IdFAI questionnaire were assessed to evaluate whether it could be applied in native Chinese-speaking patients with CAI.
The reliability of the SC-IdFAI questionnaire was mainly assessed from three aspects, namely internal consistency, test–retest reliability, and measurement error. Internal consistency was described as the degree of interrelatedness among the items [31], which was assessed by calculating the Cronbach’s α value of the scale. The scale was considered with acceptable, good, and excellent internal consistency when the Cronbach’s α value was >0.7, 0.8, and 0.9, respectively [25]. However, not all the higher Cronbach’s α values indicated better validity. For instance, Cronbach’s α values >0.95 generally indicated question redundancy in the scale [32]. In addition, the Cronbach’s α value was also calculated after the questions in the SC-IdFAI questionnaire were omitted one by one to assess the influence of each question on the α value, which was also a method to assess the internal consistency [25,33]. For assessing the test–retest reliability of the SC-IdFAI questionnaire, the patients were asked to answer the questions twice, with an interval of 1 week, and then the answers were compared. The intraclass correlation coefficient (ICC), which was derived from a two-way analysis of variance in a random-effects model, was used to assess the test–retest reliability. The test–retest reliability of the scale was considered acceptable, good, and excellent when the ICC was >0.7, 0.8, and 0.9, respectively [34]. The measurement error is the systematic and random error of a patient’s score that is not attributed to true changes in the construct to be measured [35]. Moreover, it was analyzed using the standard error of measurement (SEM) and calculated according to the following formula: SD × √ (1 – ICC), where SD is the standard deviation of the scores from all patients at the first assessment [36]. The minimal detectable change (MDC) reflected the minimal individual change in the score that could be interpreted as a real change. It was calculated as SEM × 1.96 × √2 at an individual level and SEM × 1.96 ×√2/√n at the group level [36]. The Bland–Altman plots were further depicted to observe the systematic error between the first two investigations [37].
The validity of the SC-IdFAI questionnaire was mainly assessed from two aspects, namely content validity and construct validity. The content validity was assessed by the comprehensiveness and the relevance of the items [38]. The response rate of the questions, ceiling/floor effects, and patients’ feedback were the three indicators for assessing the comprehensiveness. If the response rates of all the questions were >95%, the ceiling/floor effects of all the subscales were <15%, and the patients did not report difficulties in understanding the scale, the scale was considered to be of good comprehensiveness [25,39]. Moreover, an expert of rehabilitation medicine and two experts of osteology were invited to help in judging whether the items of the SC-IdFAI questionnaire were relevant for the construct to be measured and for the patients with CAI [38]. The criterion validity of the SC-IdFAI questionnaire could not be assessed as no gold standard was available. So, the construct validity was assessed by the method of hypotheses testing. The construct validity is the extent to which the scores on a scale are consistent with hypotheses based on the assumption that the scale validly measures a specific construct [32]. In this study, FAAM and SF-36 were selected as the control scales for SC-IdFAI. Most of the questions in the SC-IdFAI questionnaire assessed the physical conditions, but not mental conditions, of the patients with CAI. Therefore, it was hypothesized that the results of the SC-IdFAI questionnaire highly correlated with the physical subscales of the SF-36 (physical functioning, role physical, bodily pain, and general health), as well as FAAM, but poorly correlated with the metal subscales of the SF-36 (vitality, social functioning, role emotional, and mental health). In addition, FAAM was a region-specific scale specifically designed for the patients with ankle injuries, while SF-36 was a generic scale with wide applicability. Therefore, although FAAM was not a disease-specific scale designed specifically for the patients with CAI, the contents of FAAM should be closer to IdFAI than to SF-36. Thus, it was further hypothesized that the correlation between SC-IdFAI and FAAM should be higher than that with any subscales of SF-36. The detailed contents of the hypotheses between the scales are shown in table 4. Based on these hypotheses, the Spearman’s correlation coefficient (rs) of SC-IdFAI with SF-36 and FAAM was analyzed, using the first answers from the patients in the scales. Then, the construct validity of the SC-IdFAI questionnaire was assessed according to the consistency between the data and the hypotheses. Good construct validity was based on meeting the criterion for at least 75% (30/40 or more) of stipulated a priori hypotheses [35]. The correlations were judged as poor (rs = 0–0.2), fair (rs = 0.2–0.4), moderate (rs = 0.4–0.6), good (rs = 0.6–0.8), or excellent (rs = 0.8–1.0) [40].
Responsiveness is the ability of a scale to detect a change over time in the construct to be measured [38]. In this study, the responsiveness of SC-IdFAI was evaluated by comparing the scale results before (the first-time filing) and 9 weeks after administering physiotherapeutics (the third-time filing). The effect size (ES) and standardized response mean (SRM) were the two indices to evaluate the responsiveness. SRM was defined as the mean change between the two time points divided by the SD of this change. ES was defined as the mean change between pretreatment and 9-week posttreatment results divided by the SD of the pretreatment SC-IdFAI score [41]. ES and SRM were considered large if >0.80, moderate if between 0.51 and 0.80, and small if <0.50 [42].
The Statistical Package for the Social Sciences, version 20.0 (SPSS, IL, USA), was used for statistical analysis. The mean values were reported with SD. The ICC values were reported with 95% confidence intervals. A P value of 0.05 or less was considered statistically significant.