Validation of Disabilities of Arm, Shoulder and Hands (DASH) Questionnaire in Dental Students

Objectives: To assess the validity, reliability and factorial invariance of Disabilities of Arm, Shoulder and Hand - DASH scale when applied to pre-doctoral students from Brazil.
Methods: An observational cross-sectional study was conducted with 335 students from Faculty of Dentistry of Araraquara UNESP/Brazil. In order to attend the content validity, 8 occupational health experts analyzed the essentiality of DASH's items, enabling the determination of Content Validity Ratio (CVR). The psychometric sensitivity was assessed by summary statistics, asymmetry and kurtosis. For confirmatory factor analysis, the χ²/gl, CFI, GFI and RMSEA indexes assessed the goodness of fit of the one-factor model, two-factor model and three-factor model. The Convergent Validity (CV) was estimated from Average Variance Extracted. Internal consistency was calculated using the Cronbach's alpha coefficient (α) and the Composite Reliability (CR). The significance level was 5%.
Results: 11 items of the DASH presented CVR values below the recommended and 7 items do not have psychometric sensitivity.
In confirmatory factor analysis, the models tested did not show adequate quality of adjustment ( for one-factor model: χ2/gl=4.195, CFI=0.520, GFI=0.667, RMSEA=0.102, r²=0.28; for two-factor model: χ²/gl=2.802, CFI=0.730, GFI=0.794, RMSEA=0.077, r²=0.42; three-factor model: χ²/gl=2.020,CFI=0.914, GFI=0.897, RMSEA=0.058, r²=0.51).
The need for refinement of the models, resulting in the removal of 9 items’ scale. Thus, the two-factor model showed the best adjustment rates in this population (χ²/gl=1.999, CFI=0.915, GFI=0.897, RMSEA=0.057, r²=0.42, AIC=461.735, BIC=640.467, BCC=469.198). Although the convergent validity has not been adequate (AVEphysical function= 0.33; AVEsymptoms= 0.38). The best adjustment was reached with the refined two-factor model, which presented good values of internal consistency (α physical function= 0.82; α symptoms= 0.85) and the composite reliability (CRphysical function= 0.85; CRsymptoms= 0.84) the recommended values.The factorial invariance has been classified weak.
Conclusions: The refined two-factor model structure was the one valid and reliable for dental students, although its invariance has been weak.