The present study was officially authorized by the American Shoulder and Elbow Surgeons Society, the original developer of ASESp. Informed consents were obtained from all participants prior to the study.
ASES questionnaires
ASES composed of 3 sections: demographic information, patient self-evaluation (ASESp), and physician assessment [8]. A clinician is responsible to provide his or her expertise to evaluate the range of motion, strength, instability and other shoulder pathology signs; however, a score index can only be derived from the ASESp section for a thorough assessment [14].
The ASESp consists of 18 questions from 3 sections: pain, instability, and activities of daily living (ADL). Among the 18 questions, 11 self-report items representing functional (ADL) dimension (10 items) and pain dimension (1 item) are derived into a 0–50 sub-score for each dimension [14]. The ADL section was scored in a 4-points- graded ordinal scale, ranging from 0 (unable to do) to 3 (not difficult) and cumulative scores were collected. The pain section was derived from the 10-points-graded visual analog scale (VAS) ranging from 0 (no pain) to 10 (maximum pain) [7]. The overall shoulder score index was calculated with the formula below, ranging from 0 (most disability) to 100 (least disability) [8].
$$ \mathrm{Shoulder}\ \mathrm{score}\ \mathrm{index}=\left[\left(10-\mathrm{VAS}\ \mathrm{pain}\ \mathrm{score}\right)\ \mathrm{x}\ 5\right]+\kern0.5em \left[\left(5/3\right)\ \mathrm{x}\ \mathrm{cumulative}\ \mathrm{ADL}\ \mathrm{score}\right] $$
Translation and linguistic validation
A 5 steps protocol of forward-backward translation was derived based on the Beaton’s method published in the Guidelines for the Process of Cross-Cultural Adaptation of Self-Report Measures [16]. The protocol included translation, synthesis, back-translation, expert committee review and pretesting [16,17,18,19,20].
Step 1: forward translation from English to Chinese
A forward translation was done separately by two native Chinese native speakers, one acting as an informed translator (senior orthopedic resident) and one acting as an un-informed translator (medical student).
Step 2: cross-culture adaptation
Several cross-cultural dissimilarities were found in the first translated version. First, in question 4, the example of pain medications given in original ASES, namely aspirin, Tylenol, Advil, Codeine, were replaced with aspirin and acetaminophen together with the respective Chinese brand name. Similarly, the narcotic medication given in question 5 was also replaced with tramadol, Ultracet and morphine together with the respective Chinese brand name. These modifications were done based on the most common, most familiar and typical prescription in Taiwan. Secondly, we clearly define the description of “Manage Toileting” in question 4 of the ADL section as a “butt-wiping” situation. Third, the “10 lbs” in question 7 of the ADL section was translated and remarked as “5 kg” as kilogram, which was a typical measurement unit in Taiwan.
Step 3: backward translation from Chinese to English
A backward translation was done by an English native speaker who was not familiar with the orthopedics field after a Chinese consensus version was completed.
Step 4: revision by expert committees
Both forward and backward versions were then revised and reviewed by an expert committee, which composed of five senior orthopaedic surgeons, including the chief of department of orthopaedic surgery. Both versions revealed no marked disparity or language difficulties. Thus, the primary Chinese ASES questionnaire (ASESp-CH) was formed.
Step 5: pre-test of ASESp-CH questionnaires
The ASESp-CH questionnaire was given to 20 patients to disclose any problem in understanding and approaching the questionnaire. There were no obstacles reported. Hence, a final ASESp-CH questionnaire was established (Fig. 1).
Study population
The study was conducted by the Department of Orthopedic Surgery in National Cheng Kung University Hospital (NCKUH) in Tainan, Taiwan, and was approved by the Institutional Review Board of National Cheng Kung University Hospital. Patients with shoulder disorder were recruited from the out-patient department of NCKUH and the public population. All patients were required to complete the test-retest ASESp-CH questionnaires twice at the interval of 7 days to 30 days before getting any intervention. A questionnaire of SF-36 was also required to complete during the retest session of ASESp-CH. A thorough explanation and informed consent were given.
Inclusion criteria were as follow: [1] patients’ age ≥ 18 years, [2] patients with clear insights, [3] patients with any shoulder disorders, [4] patients who are able to speak and write in Chinese, and [5] patients who completed the questionnaires twice at an interval of 7 days to 30 days. Patients with one of the following conditions were excluded: [1] the patient could not complete all of ASESp-CH and SF-36 questionnaires, [2] the test-retest interval was less than 7 days or more than 30 days, and [3] the patient received interventional procedures, such as shoulder injections or surgery, during the test-retest interval.
Reliability
Reliability was considered as the degree of replicable, is the extent to which the results can be reproduced when the research is repeated under the same conditions. In this case, refers to the degree of the results of ASESp-CH can be replicated in a test-retest manner and among related items on the ASESp-CH (internal consistency) [8]. Reliability can be expressed ranging from 0 to 1, indicates no reliability and absolute reliability respectively.
Internal consistency is calculated by using the Cronbach alpha, a widely recognized analysis tool to evaluate one’s reliability [21]. This method has been used in previous ASESp validation studies [4, 8, 10,11,12,13,14,15]. A cut-off value of 0.7 represents a sufficient correlation between the items of a questionnaire. Values between 0.7 to 0.79, 0.80 to 0.89 and above 0.90 imply fair, good and excellent respectively. Yet, a Cronbach alpha greater than 0.90 may indicate a highly homogenous situation and thus redundant [8].
While considering the test-retest reliability, it was assumed that 2 separate measurements in a suitable interval should be similar if no change occurs and the time-bias can be reduced to the minimum level. A short interval may cause memory bias, while a longer interval may encounter actual changes in patient health status [8]. Thus, a time interval between 7 to 30 days was considered relevant according to the guidelines of cross-cultural adaptation. To calculate the test-retest reliability, the intraclass correlation coefficient (ICC) was used [8, 22]. An ICC of 0 indicated no agreement, whereas an ICC of 1 indicated absolute agreement. Generally, an ICC greater than 0.60 and 0.74 were considered good and excellent, respectively [22].
Validity
To achieve the validation of ASESp-CH, the results collected from 3 domains of pain, instability, and ADL are compare with the corresponding 8 domains of 36-Item Short Form Survey (SF-36) [23]. As a widely accepted and validated health status measure, SF-36 also has been used as a parameter for the various translated versions of ASESp in the previous studies [4, 8, 10,11,12,13,14]. A translated and validated Taiwan version of SF-36 was eligible for the validation process [24, 25].
Pearson correlation coefficient was used to evaluate the construct validity between the 3 domains of ASESp-CH and the corresponding domains in SF-36. Statistical analysis was performed using SPSS 20.0 (IBM, Armonk, NY, USA) and Excel 2010 (Microsoft, Redmond, WA, USA). A p < 0.05 was considered statistically significant.