A validation study comparing two self-reported upper extremity symptom surveys with clinical examinations for upper extremity musculoskeletal disorders

• Published in: Work (Reading, Mass.) Mass.), 2012-01, Vol.43 (3), p.293-302
• Main Authors: Menéndez, Cammie Chaumont, Amick III, Benjamin C., Jenkins, Mark, Caroom, Cyrus, Robertson, Michelle, Gerr, Fred, Moore, J. Steven, Harrist, Ronald B., Katz, Jeffrey N.
• Format: Article
• Language: English
• Subjects: Classification; Comparative analysis; Computers; Cross-Sectional Studies; Female; Health Surveys; Humans; Likert scale; Male; Musculoskeletal diseases; Musculoskeletal Diseases - classification; Musculoskeletal Diseases - diagnosis; Musculoskeletal Diseases - epidemiology; Pain - epidemiology; Pain - etiology; Pain Measurement - methods; Prevalence; Protocol; Psychometrics - instrumentation; Psychometrics - statistics & numerical data; Reproducibility of Results; Self Report; Sensitivity and Specificity; Severity of Illness Index; Southwestern United States; Students; Surveys and Questionnaires; Upper Extremity - physiopathology; Validation studies
• ISSN: 1051-9815; 1875-9270
• DOI: 10.3233/WOR-2012-1401

Objective: Evaluate the validity of two self-report symptoms surveys with two disorder classification protocols. Participants: 100 graduate students at a private school in the Southwest United States. Methods: Study participants completed two self-report upper extremity musculoskeletal symptoms surveys: a nine item 10 cm Visual Analogue Scale (VAS) and a nine item Likert categorical scale anchored from "None" to "Very severe". Clinical examinations were administered using two musculoskeletal disorder classification protocols. Results: For the nine body regions, concordance between the two self-report symptoms scales ranged from 0.49–0.75. Overall there was greater than 80% agreement for the two disorder classification protocols. Using either symptom survey with either disorder classification protocol provided high sensitivities and specificities (Youden's J ⩾ 0.70). Three of possible six symptom survey/classification protocol pairings provided high sensitivities and specificities across all disorder groups. Conclusion: In this graduate student sample, none of the self-report symptom survey-classification protocol pairings was demonstratively more useful than any other pairing for studies of musculoskeletal disorders among computer users.