|Home | About | Journals | Submit | Contact Us | Français|
To validate the Persian version of the simple shoulder test in patients with shoulder joint problems.
Following Beaton`s guideline, translation and back translation was conducted. We reached to a consensus on the Persian version of SST. To test the face validity in a pilot study, the Persian SST was administered to 20 individuals with shoulder joint conditions. We enrolled 148 consecutive patients with shoulder problem to fill the Persian SST, shoulder specific measure including Oxford shoulder score (OSS) and two general measures including DASH and SF-36. To measure the test-retest reliability, 42 patients were randomly asked to fill the Persian-SST for the second time after one week. Cronbach’s alpha coefficient was used to demonstrate internal consistency over the 12 items of Persian-SST.
ICC for the total questionnaire was 0.61 showing good and acceptable test-retest reliability. ICC for individual items ranged from 0.32 to 0.79. The total Cronbach’s alpha was 0.84 showing good internal consistency over the 12 items of the Persian-SST. Validity testing showed strong correlation between SST and OSS and DASH. The correlation with OSS was positive while with DASH scores was negative. The correlation was also good to strong with all physical and most mental subscales of the SF-36. Correlation coefficient was higher with DASH and OSS in compare to SF-36.
Persian version of SST found to be valid and reliable instrument for shoulder joint pain and function assessment in Iranian population.
Outcome measures are useful instruments in orthopedics. They help assess the function and quality of life of patients in the clinical practice (1). Outcome measures are categorized as physician-reported versus patient-reported tools. Furthermore, they can also be classified in terms of assessing the general health status such as short form 36 (SF-36) versus region-specific measures particularly assessing an anatomical region including shoulder (2).
Shoulder conditions are the second most prevalent disorders bringing the patients to the orthopedic clinics. There are several patient reported outcome (PRO) measures evaluating shoulder function and disabilities including the American shoulder and elbow surgeons (ASES), Shoulder pain and disability index (SPADI), disabilities of the arm shoulder and hand (DASH), Oxford shoulder score (OSS), simple shoulder test (SST), subjective shoulder rating system (SSRS), shoulder severity index (SSI), and University of California Los Angeles (UCLA) shoulder score (3-5).
Simple shoulder test (SST) is a simple, short, and patient reported standardized tool for the assessment of shoulder function. Lippitt et al developed it in 1993 in order to assess the shoulder function before and after the treatment (6). It is comprised of 12 questions with dichotomous ‘Yes=1’ or ‘No=0’ response options. The sum of scores makes the total score ranging from 0 showing the worst to 12 showing the best function during the last week. SST is validated and used in different languages including English, Italian, Portuguese, Spanish, and Dutch (7-11). In the current study we aimed to translate it into Persian and test the validity and reliability of cross-culturally adapted questionnaire in Iranian population with shoulder conditions.
We performed the study in 3 phases. In phase 1, translation and cultural adaptation was done according to Beaton’s guideline (12) [Appendix 1]. In phase 2, was performed a pilot study to find the acceptability and difficulties in understanding the items. In phase 3, we administered the final version of the Persian-SST to patients with shoulder condition.
We used the forward-backward method according to the published guidelines for cross-cultural adaptation of health status measures to translate the original English SST into Persian (12). Initially, two native Persian-speaking researchers translated SST into Persian. The two drafts were compared in a session to reach to a consensus of about the translation and wording of the items. Furthermore, a native English speaker, who was not aware of the original SST, back translated the Persian draft into English. This version was then compared to the original SST and minor discrepancies were addressed [Appendix 1].
To test the face validity, we administered the Persian version to 20 patients with shoulder problems to find difficulties in understanding. There were minor issues that were addressed in the final Persian version.
Our Institutional Review Board approved this study and patients were verbally consented to fill the questionnaires. We enrolled 148 patients with different shoulder conditions from September 2013 to March 2014 [Tables 1; 2].
To test the validity of a questionnaire, it has to be tested over an already validated questionnaire to check whether any correlation exists in scores, which further shows that they both are measuring what they are supposed to measure. To test the validity of SST, one shoulder specific measure including Oxford shoulder score (OSS) and two general measures including DASH and SF-36 were also administered to the patients. OSS, DASH, and SF-36 have all been validated in Persian (13, 14). We used Spearman’s correlation coefficient to show the level and direction of changes in scores of the SST in respect to other applied measures.
To measure the test-retest reliability, 42 patients were randomly asked to fill the Persian-SST for the second time after one week. Patients did not receive major treatment in order to avoid substantial changes in symptoms over this period. Test retest reliability was tested using Intraclass correlation coefficient (ICC) with 95% confidence interval. P-values less than 0.05 were considered significant. Higher coefficients show better reproducibility of the tool.
Cronbach’s alpha coefficient demonstrates internal consistency over the 12 items of Persian-SST. Cronbach’s alpha between 0.70 and 0.90 is considered as good internal consistency while values above 0.90 shows excellent internal consistency. We also tested the internal consistency of the questionnaire after deletion of each item to check whether the consistency of the items increases without the deleted item.
Testing for validity showed strong correlation between SST and OSS and DASH. The correlation with OSS was positive while with DASH scores was negative showing that it is measuring what it is supposed to measure because lower DASH scores show less disability, which is similar in context to that of higher OSS and SST scores. The correlation was also good to strong with all physical and most mental subscales of the SF-36. Correlation coefficient was higher with DASH and OSS in compare to SF-36 showing that region specific measures are more sensitive than general measures and more prone to be affected by the specific regional condition [Table 3].
Intraclass correlation coefficient for the total questionnaire was 0.61 showing good and acceptable test-retest reliability. ICC for individual items ranged from 0.32 to 0.79. This variation may be because of a one-week interval between test and retest that may have affected the response to some items although major treatment was not received. However the reliability and reproducibility of the questionnaire is confirmed for the whole questionnaire [Table 4].
The total Cronbach’s alpha was 0.84 showing good internal consistency over the 12 items of the Persian-SST. Cronbach’s alpha after deletion of each item did not exceed the total coefficient showing that all items are consistent and measure in a same direction [Table 4].
The aim of our study was to assess the validity and reliability of the Persian-SST in Persian speaking population with shoulder conditions. We confirmed the validity and reliability of the Persian simple shoulder test questionnaire and usefulness of this instrument in quantifying pain and function of this specific joint.
One of the limitations to our study is that we administered the SST to patients who had the history of shoulder surgery or injection regardless of time limits in terms of recent or remote surgery. This may have affected the responsiveness of patients with recent versus patients with remote surgery. Another limitation to our current study is testing the validity of the Persian SST over the two more general health related questionnaires including the DASH and SF-36 (2). However, both of them are patient-based questionnaires.
Internal consistency and test-retest reliability indicated good reliability of Persian-SST. Also measurement of Spearman`s correlation coefficient confirmed the construct validity of Persian-SST by showing strong correlation with Persian OSS, DASH, and majority of subscales of SF-36 questionnaires. In our study, Cronbach`s alpha was 0.84 showing good internal consistency and high level of reliability over the items. Test-retest reliability of Persian SST was 0.61 that was high enough to confirm the test-retest reliability and proved to be a reproducible tool.
DASH questionnaire was used in the validation process of Spanish and Dutch versions of SST (7, 9). Correlations between SST and DASH in Spanish and Dutch versions were stronger than Persian version (r=-0.73 in Spanish version and r=-0.74 in Dutch version compared to r=0.596 in Persian version). OSS questionnaire was used in validation process of Dutch and Persian versions. Correlation between SST and OSS was stronger in Dutch version than Persian version(r=-0.74 in respect to r=0.586). SF-36 questionnaire was used in validation process of Portuguese and Persian versions of SST (8). SST has fair correlation with SF-36 in Persian version compared to significant correlation with Portuguese version. Godfrey et al published results of their study about validity, reliability and responsiveness of SST in patients with shoulder instability and rotator cuff injuries in 2007 (11). The ICC for SST was >0.90 demonstrating an excellent test-retest reliability. Membrilla-Mesa et al tested the validity and reliability of Spanish version of SST (9). Cronbach’s alpha was 0.79 and ICC was 0.91 showing high test-retest reliability. Validity determined from the relation between Spanish SST and DASH was strong and inversely related (r=-0.73; P<0.001) and quite inversely related to VAS (r=-0.53; P< 0.001). There were poor and direct correaltions with Physical Component Scores and Mental Component Scores of SF-12, respectively. Validation of SST in Portuguese Brazilian population showed that Cronbach’s alpha and ICC were 0.82 and 0.84, respectively (8). In another study by van Kampen et al, the Dutch version of SST was validated with a good internal consistency (alpha=0.78), and a excellent test-retest reliability (ICC=0.92) (7). There were strong correlations between the Dutch version of SST and the DASH (r=-0.74) as well as OSS (r=-0.74). For the Italian version of SST authors showed that the Italian versions of UCLA shoulder scale, SPADI, and SST are valid and reliable instruments (10).
Simple Shoulder Test is a reliable and valid instrument for the assessment of shoulder conditions, which showed to be applicable in different cultures and languages. Using an identical technique in our validation studies, we confirmed that Simple Shoulder Test is a standardized outcome measure for the assessment of shoulder pain and dysfunction (3, 15).