Test-retest reliability of the ASES-p shoulder scale.

INTRODUCTION: Shoulder disorders are common musculoskeletal problems. The self-assessed ASES questionnaire (ASES-p) is one of the most widely used tools for evaluating shoulder function. Its 11 items are divided in a function (10 items) and pain (1 item) dimension, assigned between 0 and 50 points each. Their sum is the scale's total score, with higher values indicating better health status. The current work explores the test-retest reliability of the Spanish version of the ASES-p score values. MATERIALS AND METHODS: The scale was administered twice to a sample of subjects with various shoulder pathologies, via telephone interviews performed at 3-7 days apart. Exact agreement was calculated on an item and score basis. Score variability was assessed with the 95% limits of agreement method (LoA). RESULTS: N = 161 subjects were initially contacted, and a total of 82 stable health status subjects provided valid test-retest replies. "Do usual sport" was the only item with missing data. Exact agreement oscillated between 67 and 89% per item. The 95% LoA ranged between -5.9 and 6.9 points for function; -13.2 to 11.9 for pain and -10.3 to 10.1 for the total ASES-p score. CONCLUSIONS: Test-retest reliability in stable patients was considered acceptable for the function and total scores, but not for pain. This may reflect usual pain behaviour, but it also implies that the pain evaluation should be further studied. The ASES-p pain subscore should not be used as the single measure for monitoring shoulder pain. Revisiting the "do usual sports" item may increase the scale's applicability.

Patient self-report section of the ASES questionnaire: a Spanish validation study using classical test theory and the Rasch model.

BACKGROUND: The aim of the current study was to validate the self-report section of the American Shoulder and Elbow Surgeons questionnaire (ASES-p) into Spanish. METHODS: Shoulder pathology patients were recruited and followed up to 6 months post treatment. The ASES-p, Constant, SF-36 and Barthel scales were filled-in pre and post treatment. Reliability was tested with Cronbach's alpha, convergent validity with Spearman's correlations coefficients. Confirmatory factor analysis (CFA) and the Rasch model were implemented for assessing structural validity and unidimensionality of the scale. Models with and without the pain item were considered. Responsiveness to change was explored via standardised effect sizes. RESULTS: Results were acceptable for both tested models. Cronbach's alpha was 0.91, total scale correlations with Constant and physical SF-36 dimensions were >0.50. Factor loadings for CFA were >0.40. The Rasch model confirmed unidimensionality of the scale, even though item 10 "do usual sport" was suggested as non-informative. Finally, patients with improved post treatment shoulder function and those receiving surgery had higher standardised effect sizes. CONCLUSIONS: The adapted Spanish ASES-p version is a valid and reliable tool for shoulder evaluation and its unidimensionality is supported by the data.