Introduction

Shoulder injuries cause pain and reduce articular mobility, ultimately adversely affecting the functional abilities, work activities and qualities of life of these patients [6]. For many years, rating scales and scoring systems have been used to assess the outcomes of shoulder injuries and treatment. In general, assessments are categorised in two groups: general health and disease and joint-specific. After systematically reviewing the literature, Bot et al. [5] discovered 16 objective questionnaires in English.

Questionnaires were established in order to investigate a specific disease or certain part of the body [25]. However, the characteristics of the majority of these scores reflect the language and social culture of the community in which they were established; therefore, it is necessary to translate and culturally adapt these scores for other communities.

Published in 1994, The American Shoulder and Elbow Surgeons Standardized Shoulder Assessment Form (ASES) has two sections: the patient self-evaluation section (also known as the patient self-report section) and the physician assessment section (also known as the medical professional assessment section) [20]. Because the ASES is available in English-speaking countries, it is not applicable over a wider geographical range. In order to broaden its usage, the questionnaire needs to be translated into other languages, culturally adapted to other countries and validated against the original questionnaire. The only questionnaires that have a valid Turkish version are the Shoulder Pain and Disability Index (SPADI), the Western Ontorio Rotator Cuff Index (WORC), Disability of the Arm, Shoulder and Hand Questionnaire (DASH), the Oxford Shoulder Questionnaire (OSQ) and [4, 8, 18, 23].

Compared to the other Turkish validated shoulder scores, ASES has some advantages. DASH has many detailed questions and takes more time to complete, and thus people in Turkey, specifically in the elderly population, generally do not fill out these forms. They also have some difficulties answering the VAS-type questions, such as in SPADI. However, ASES is shorter, so it requires little time for the patient to complete and for the health practitioner to evaluate. It is also not disease-specific, such as WORC, so it can be used for many shoulder pathologies.

Although there is no standard translation and validation of the ASES, it is widely used by physiotherapists and health practitioners. However, different interpretations of its meaning may affect clinical outcomes. In order to eliminate differences in methods and scores and replace them with suitable standards, we believe that cultural adaptation and its valid use in Turkish is important. The purpose of this study was to translate the ASES subjective form into Turkish and investigate its validity and reliability.

Materials and methods

Translation and cultural adaptation

Only the patient self-report section of the ASES was translated into Turkish and culturally adapted in accordance with stages recommended by Guillemin [9]. Two Turkish individuals, with a good command of English, were responsible for the literary and conceptual translation of the patient self-report section of the ASES assessment form. The translators were a nurse and an engineer who were familiar with the purpose of the study; translations were completed independently. Both translations were compared and reviewed by a bilingual person who highlighted any conceptual errors or inconsistencies in the translations in order to establish the first Turkish translation. Once the first Turkish translation was decided upon, two translators who had a good command of Turkish but spoke English as a first language and who were familiar with the purpose of the study, translated the finalised Turkish translation back into English separately; both translators were unaware of the purpose of the study and had no access to the original assessment form.

Recent versions of the questionnaire were compared to the initial translation. A committee consisting of four translators compared the English retranslation with the initial Turkish translation and then approved the Turkish version of the ASES assessment form. Once approved, the assessment form was administered to the patients.

No issues involving the translation or cultural adaptation were experienced during the pilot study conducted on 20 patients; however, patients had difficulty answering Question 8, and those that did not engage in regular sporting activities left Question 10 blank. Question 8 was changed from ‘throw a ball overhand’ to ‘over the head throwing activity’. Question 10 was changed to provide two options: ‘(a) do usual sport, (b) do not do usual sport’; in other words, option ‘a’ was similar to the original, and option ‘b’ was comprised of questions such as ‘putting up nails’, ‘beating carpets’ and ‘using a vacuum cleaner’. In addition to these changes, ‘10 libre’ in question 7 was converted to ‘5 kg’ because libre is not a Turkish unit of measurement.

Questionnaires

The ASES consists of 2 sections: a patient self-evaluation component and an assessment performed by a physician. The patient self-evaluation section has 11 items that can be used to generate a score. These items are divided into 2 areas: pain (1 item) and function (10 items). The pain question was scored on a scale (visual analogue scale) of 0 (no pain) to 10 (worst pain ever). The function questions ask patients whether they can perform 10 daily life activities. Additionally, they are asked whether they can do their usual work and take part in normal sporting activity. The items are rated on a four-point Likert scale. Scores obtained from the pain and function subsections are transformed into percentages, where each represents 50 % of the final score. Scores on the ASES range from 0 (absence of function) to 100 (normal function) [4]. The four-point Likert scale for the function questions ranges from 0 (unable to do) to 3 (not difficult). The final score (a possible maximum score of 100) is calculated by multiplying the pain section score (a possible maximum score of 10) by five (a possible maximum score of 50) and multiplying the cumulative activity score (a possible maximum score of 30) by 5/3 (a possible maximum score of 50).

The 13-question SPADI is comprised of two subscales: pain and disability. The subscale ‘pain’ presents five questions about pain severity, and the patient answers each question based on a scale (visual analogue scale) from 1 (no pain) to 10 (the worst pain imaginable). The subscale ‘disability’ poses eight questions about the level of difficulty experienced when conducting daily living activities that require use of the upper extremities [21].

The Short Form Health Survey (SF-36): SF-36 was used to establish a health profile that consists of eight scaled scores, where each scale was directly transformed into a scale from 0 to 100 in order to identify the patient’s physical and mental state. These 8 sections include physical functioning (PF), physical role functioning (RP), bodily pain (BP), general health perceptions (GH), vitality (VH), social role functioning (SF), emotional role functioning (RE) and mental health (MH) [24].

Patients

Seventy-five patients suffering from shoulder complaints were recruited from the Istanbul University Faculty of Medicine, Department of Orthopedics. Eight patients did not return for the retest assessment, two declined to complete the SF-36 and two had received medical treatment before the retest assessment. In all, sixty-three patients (30 male and 33 female; mean age: 48.2 ± 13.4; range: 18–74 years) were included in the study. Table 1 presents the various diagnoses of the patients.

Table 1 Demographics
Full size table

The patients were clinically examined by two experienced shoulder surgeons. When necessary, X-ray and MRI (magnetic resonance imaging) were performed.

The inclusion criteria are listed as follows: (1) 18 years of age or older, (2) the presence of a shoulder problem and (3) no treatment between test–retest assessments. The exclusion criteria were as follows: (1) inability to complete the form due to cognitive impairment; (2) illiteracy or lack of understanding of Turkish; (3) patients whose conditions could not be stabilized following a second assessment for other ailments, such as cancer, serious infection or inflammatory disease; (4) the presence of neurological or musculoskeletal disorders other than the shoulder condition.

Prior to the start of the study, participants were asked to read and sign an informed consent form, which had been approved by the ethical committee at Istanbul University (IRB study protocol: 2010/898-268).

Testing protocol

Patients were asked to complete the Turkish version of ASES, the previously validated Turkish version of the SPADI and the SF-36 [4, 14].

Physical therapy undergraduate students administered the listed questionnaires to patients in waiting rooms prior to an appointment with an orthopaedic surgeon. In the event of any difficulties, students took notes on each patient.

Validity

Validity is represented by the extend to which a score retains its intended meaning and interpretation [7]. In this study, validity was assessed according to three factors: construct, convergent/divergent and content validity. The construct validity of the Turkish ASES was tested by correlation with the SPADI and the physical component score of the SF-36.

SF-36 PF, SF-36 RP and SF-36 PCS domains were used to asses the convergent validity. Divergent validity was evaluated by using SF-36 MH, SF-36 RE and SF-36 MCS domains. It is hypothesized that the physical domains of SF-36 would correlate with disease or joint-specific questionnaires better than the mental domains.

The content validity was assessed by the distribution and occurrence of ceiling and floor effects. The ceiling effect occurs when the maximum possible score of 100 is achieved. The floor effect occurs when the minimum possible score of 0 is reached. For the maximal score, we applied scores of 90 to 100 %, and for the minimum score, we used 0 to 10 % scores. Greater than 15 % of the patients scored a maximal or minimal score, and we considered this as a floor and ceiling effect.

Reliability

The test–retest reliability, which is a measure of stability or reproducibility, represents a scale’s capability of giving consistent results when administered on separate occasions [17]. The reliability of scale scores has been estimated using the internal consistency method and test–retest method across repeated administrations. To determine the test–retest reliability, 63 patients were asked to complete the ASES 3–7 days after the first assessment. To minimise the risk of short-term clinical change, no treatment was provided during this period.

Statistic analysis

Construct and convergent/divergent validities were tested with Pearson’s correlation coefficient, and a 95 % confidence interval was used for all correlation coefficients. The intraclass correlation coefficient (ICC) was used to measure the test–retest reliability of the ASES assessment form. Correlation values ≥0.4 were considered satisfactory (r ≥ 0.81–1.0 excellent, 0.61–0.80 very good, 0.41–0.60 good, 0.21–0.40 fair and 0.00–0.20 poor) [10, 15]. Cronbach’s alpha coefficient was used to determine internal consistency. A paired t test was used for the comparison of first and second measurements to compare and determine the statistically significant differences between the first and second tests.

A statistical package for the social sciences (SPSS) 17.5 was used to conduct statistical analysis. The agreed level of significance was p < 0.05. Floor and ceiling effects and the number of items answered were identical during the test and retest examination.

Results

Table 1 illustrates the main demographic and clinical characteristics of the patients. The patients were literate, but their educations were mostly at the high school level. They were of the middle socio-economic class. None were athletes or regularly participated in any sport. The duration of the symptoms was 4.8 ± 1.2 months. The Turkish ASES was completed in approximately four minutes. No difficulties were noted during the testing protocol. The Turkish ASES showed high-to-moderate correlation with the SPADI score and BP domains of the SF-36 (Tables 2, 3). Contrary to our hypothesis, the Turkish ASES did not show strong correlation with SF-36 physical domains, but there was better correlation with SF-36 mental domains (Table 4). Three out of 63 patients (4.7 %) scored between 90 and 100, and 2 out of 63 (3.1 %) scored between 0 and 10. A total of 8 % of patients ranged maximal and minimal scores, which were below 15 %. This implies that there are no floor and ceiling effects. A Cronbach’s alpha value of 0.88 was found. The Turkish ASES domains and total score showed strong ICC values. The paired t test did not demonstrate statistically significant differences between the test–retest means (Table 2).

Table 2 Test–retest reliability and internal consistency of ASES
Full size table
Table 3 The correlation with ASES and SPADI
Full size table
Table 4 The correlation with SF-36 and ASES
Full size table

Discussion

In this study, the most important finding is that the Turkish version of the ASES is reliable, valid and sensitive to changes over time. The ASES demonstrated acceptable psychometric performance for patients with various shoulder pathologies within the Turkish population. It can be used to assess both individual patients and groups of Turkish patients with confidence.

The test–retest indicated excellent reliability with values of 0.95 for the pain subscale of the ASES, 0.86 for the function subscale of the ASES subjective form and 0.94 for the total ASES. The internal consistency was high (Cronbach’s alpha coefficient: 0.88). Michener et al. analysed the psychometric properties of the ASES in sixty-three patients with varied shoulder diagnoses [16]. Compared to our results, they reported a lower test–retest reliability (ICC = 0.84) but a similar internal consistency (Cronbach’s alpha = 0.86).

There are a limited number of valid ASESs discussed in the literature; for example, there are German, Brazilian, Italian and Arabic versions [11, 12, 19, 26]. John et al. conducted a validity test for the German ASES and they reported a high coefficient alpha value (Cronbach’s alpha = 0.90) and 0.93 for the test–retest reliability analysis of the total ASES [11]. The valid Arabic version had similar results, with a Cronbach’s alpha coefficient of 0.81 and an ICC of 0.96, which is similar to our results.

In recent studies, translated versions of the ASES have been correlated using DASH, a visual analogue scale (VAS) and the OSQ. Correlation coefficients with the ASES were reported between 0.48 and 0.92. Yahia et al. [26] correlated the Arabic ASES with the SPADI and they reported good correlation (ICC = 0.79) between the total SPADI score and the ASES. In our study, convergent validity was tested by comparing the ASES with the SPADI score and the SF-36 questionnaire. The correlation coefficients between the ASES and the subscale scores of SPADI were good to excellent (Table 3).

The correlation between the ASES and SF-36 is poor to moderate, even though authors report good-to-high correlations between the ASES, SPADI, DASH and OSQ. These results are not surprising and are probably due to the differences in context between condition-specific questionnaires, such as the DASH, SPADI and ASES, and generic questionnaires such as the SF-36 [2]. The correlation between the SF-36 score and scores of specific instruments is poor, which proves that the SF-36 measures additional aspects of physical health and provides more comprehensive information than condition-specific questionnaires [1]. Researchers investigated the correlation between the ASES and the SF-36 subscales in different settings, and results yielded poor-to-high correlations [16, 17, 26]. The correlation between the Italian ASES and the SF-36 and the German ASES and the SF-36 was poor to moderate (r = −0.25 to 0.60, r = 0.18 to 0.66). In our study, the highest correlation was observed with BP and the lowest correlation was observed with PCS (Table 4).

The answers to questions in the MCS of the SF-36 can be different from the results reported in the literature owing to the characteristics of Turkish culture. In general, health problems in the Turkish community may not be a reason for people not to visit relatives and friends or alienate them socially. Therefore, contrary to the literature, a higher correlation of the ASES and SF-36 MCS was observed (Table 4).

In the literature, the psychometric properties of ASES were studied for different shoulder pathologies, pre- or post-operative surgeries and in- or outpatient settings. Beaton et al. [3] have shown that the ASES questionnaire is reproducible, valid and responsive to change in patients with active disease. Skutek et al. [22] have assessed post-operative shoulder patients with ASES. Kocher et al. [13] used selected shoulder pathologies, such as shoulder instability, rotator cuff disease and glenohumeral arthritis, in their study. In our study, even though we have a small sample size, our patients varied with respect to their shoulder pathologies and pre- and post-operative conditions. It is short and easy to administer and interpret, with a minimal investment of time required for either the clinician or researcher. Therefore, we believe that the Turkish ASES is sufficient to evaluate the different states of disease.

A limitation of our study was the untested statistical power and the small sample size. However, similar studies have used equivalent samples. Responsiveness and minimal clinically important (MCID) differences are the other important psychometric considerations for clinical outcome measures. While responsiveness has the benefit of a scale to measure clinical change, MCID represents the smallest improvement considered worthwhile by a patient [10]. In the present study, responsiveness and MCID of Turkish ASES score should be studied.

Conclusion

The self-report section of the Turkish ASES contains enough questions to reveal the functional status and pain of the patients. It is found that the Turkish ASES is reliable and valid and can therefore be used as an instrument to assess the functional limitations of patients with shoulder pathologies.