Background
Shoulder pain is a common disabling orthopedic condition experienced by many patients worldwide, just following low back pain and neck pain as the third most common disease in orthopedic visits [
36]. The lifetime prevalence of shoulder pain in the general population is 6.7-66.7 % [
22]. Shoulder pain is also a major problem in China. One previous study reported a 38.8 % one-year prevalence rate in the working population [
43]. The most common causes of persistent shoulder pain are rotator cuff disorders, adhesive capsulitis, and glenohumeral osteoarthritis [
25]. In addition, shoulder pain greatly affects arm function and leads to disability, which remains an enormous burden for the society [
24,
25].
Recently, more and more healthcare providers are concerned about the quality of life of patients with shoulder pain [
20,
37]. Interventions that better alleviate symptoms and improve shoulder functions are selected based on patients' reports [
6,
37]. Therefore, measurement tools for patient-reported outcomes are essential and critical in the management of shoulder pain. Although there have been several validated self-reported questionnaires for low back pain and neck pain in China [
39‐
42], scarce constructs are available for Chinese physicians to evaluate the status of shoulder pain and the effectiveness of interventions. To evaluate the quality of life in patients with shoulder pain, several constructs have been developed such as CMSOS [
8], disabilities of the arm, shoulder and hand (DASH), simple shoulder test (SST), and shoulder pain and disability index (SPADI) [
1]. Each has its advantages in measuring pain perception and functional disability. Currently, only CMSOS is available for Chinese patients [
16,
23]. However, CMSOS can be affected by surgeon bias [
30]. As a result, new patients-based constructs for patients with shoulder pain are needed. Therefore, it is imperative to develop a valid and easy administrative construct specific for Chinese patients with shoulder pain.
Cross-cultural adaptation of an existing measure may render the adapted measure as an objective uniform criterion for international or multi-center clinical trials [
3,
17]. Due to differences in idioms and traditions in daily life, translation of a questionnaire directly from another language is inadequate for application of the questionnaire in a new culture. Therefore, cultural adaptation of the tool is vital, and psychometric properties must be tested [
3,
17].
The Oxford Shoulder Score (OSS), first described by Dawson et al. in 1996, is a shoulder pain specific questionnaire for the evaluation of pain perception and daily function in patients suffering from shoulder pain, with excellent internal consistency, reliability and validity [
10]. It has been widely accepted for its simplicity and easy administration for doctors and patients, and has been applied in several clinical conditions such as shoulder surgery [
10], rotator cuff injury [
34], and frozen shoulder [
7]. Currently, it has been translated and validated in Dutch [
4], German [
19], Danish [
15], Korean [
28], Turkish [
35], Norwegian [
12] and Italian [
26]. However, no valid simplified Chinese version is available.
Therefore, the aim of this study was to cross-culturally adapt and psychometrically evaluate the simplified Chinese version of OSS (SC-OSS) in patients with shoulder pain in mainland China.
Discussion
In the present study, the English version of the OSS was successfully adapted and psychometrically validated into simplified Chinese. Statistical analysis revealed that all items of SC-OSS were well distributed and moderately correlated with each other, with excellent internal consistency, test-retest reliability and construct validity. Factor analysis demonstrated that SC-OSS was uniform and well-structured for shoulder pain in Chinese patients.
To evaluate the quality of life in patients with shoulder pain, several constructs have been developed such as CMSOS [
8], OSS [
10], disabilities of the arm, shoulder and hand (DASH), simple shoulder test (SST), and shoulder pain and disability index (SPADI) [
1]. Each has its advantages in measuring pain perception and functional disability. Currently, only CMSOS is available for Chinese patients [
16,
23]. However, CMSOS can be affected by surgeon bias [
30]. Specifically, two components of CMSOS are evaluated by the doctor, which leads to major variations among clinical settings [
30]. Also, the reliability of CMSOS has been challenged due to the lack of standardization in the assessment procedures [
8,
30]. As a result, new patients-based constructs for patients with shoulder pain are needed. Of the other four constructs mentioned above, the OSS is the most cross-culturally adapted and validated and has been proven to be simple, acceptable and reliable in many different cultures and areas [
4,
13,
16,
21,
27,
29,
36]. Hence, we decided to cross-culturally adapt the OSS, for Chinese patient with shoulder pain, as well as provide an international validated tool for multi-center research on quality of life.
In the present study, the acceptance of SC-OSS was high as all the items were well responded by Chinese patients with non-specific shoulder pain, with a responding rate of 95.9 %. High responding rates were also demonstrated in Italian [
26], Turkish [
35], German [
19] and Korean [
28] patients for their corresponding version. In the present study, the mean completion time was 2.1 min, similar to that reported for the Turkish and Korean versions. A longer completion time (3 min and 25 s) was reported for the German study, and a completion time was not reported in the Dutch and Italian studies. The short completion time and high responding rate indicates that SC-OSS is highly acceptable, and thus could be easily administered with little effort.
During the cross-cultural adaptation process some minor modifications had to be made due to cultural differences between Chinese and Western patients. Homogeneity analysis revealed that none of the items should be omitted. Item-total correlation demonstrated that each item made a contribution to the sum of the construct. Similar results were also found in English [
10], Turkish [
35], Korean [
28] and German
190 studies. The English study applied the ceiling or floor effect, which has a similar statistical meaning to the response trend, and also showed good homogeneity [
8]. Homogeneity analyses were not reported in the Italian [
26] and Dutch [
4] studies. Therefore, SC-OSS is homogeneous in measuring shoulder pain in a Chinese cultural background.
OSS was designed to assess the function of shoulder movement. The total items could be divided into two subgroups: pain related items and interferences of shoulder function related items. Yet, one factor structure was obtained in SC-OSS. No description of facture analysis were reported in other versions of OSS, including Dutch [
4], German,19 Danish [
15], Korean [
28], Turkish [
35], Norwegian [
12] and Italian versions [
26]. Therefore, the factor structure of OSS needs more investigation in different cultural settings.
SC-OSS had good internal consistency and reproducibility, indicating excellent reliability. The Cronbach’s α was 0.92, almost the same as that reported in the English and other language versions, indicating that OSS remains stable in different cultures. In addition, the ICC was 0.97, which is also consistent with that reported in most studies, suggesting that the OSS remains stable over time.
The construct validity of SC-OSS was tested against CMSOS, VAS and SF-36, as these constructs are commonly used in China. As expected, SC-OSS highly correlated with CMSOS, similar to the English, Dutch, German, Italian and Korean studies. However, regarding the correlation with VAS, the Korean study [
28] verified a low to moderate correlation (
r = 0.34), whereas our study observed a high correlation (
r = 0.70). This discrepancy is probably due to the difference in patient demographic characteristics (i.e. a female dominant sample with younger age in our study versus a male dominant sample with older age in the Korean study). However, since no other studies reported a correlation between OSS and VAS, this assumption should be further tested and other factors contributing to the discrepancy should also be explored. SF-36 has been applied worldwide for evaluating patient quality of life, especially in chronic pain patients [
18]. It has been used in most studies evaluating the construct validity of the OSS. In the present study, SC-OSS highly correlated with PF and BP, moderately correlated with RP, SF, GH and VT, and had a low correlation with RE and MH. Since the OSS focused on shoulder pain perception and functional disability, it is reasonable that SC-OSS has correlated highly with PF and BP, but had a low correlation with RE and MH. This phenomenon has also been demonstrated in the original English OSS version [
10], indicating that the OSS has excellent construct validity across different cultures. However, it should be noted that the English version [
10], Italian [
26], Dutch [
4] and German [
19] versions adapted a 1 to 5 rating scale, whereas the Turkish [
35] and present study adapted a revised scoring 0-4 rating scale. The different scoring may have led to slight discrepancies in evaluating the correlations between the OSS and other instruments.
There are some limitations in this study. First, the patients in this study are from a single center, which may not fully represent the whole country. Multi-center large sample study is favored. Second, responsiveness (i.e. sensitivity to change, reflects the ability to detect clinically significant changes) of SC-OSS was not determined, so as the German, Korean, Turkish and Italian studies.
Competing interests
The authors, their immediate families, and any research foundation with which they are affiliated did not receive any financial payments or other benefits from any commercial entity related to the subject of this article.
Authors‘ contributions
Conception and design: XZW, ZMW. Data analysis and interpretation: XMX, XZW and FW. Collection and assembly of data: XMX, FW, XLW. Manuscript writing: XMX, XZW, FW. Manuscript modifying: XZW and ZMW. Manuscript Final approval of manuscript: XMX, XZW, FW, XLW, and ZMW.