|Beitragstitel||Structural validity of the SRS-22 Questionnaire in Patients with Adult Spinal Deformity|
Introduction: Designed for patients with adolescent idiopathic scoliosis, the Scoliosis Research Society (SRS) questionnaire is now used as the outcome instrument of choice in patients with Adult Spinal Deformity (ASD). No studies have confirmed the four-factor structure (pain, function, self-image, mental health) in ASD or evaluated this in different contexts. We sought to evaluate the invariance of the factor structure of SRS-22 across different languages, age-groups, ASD etiologies, and times of assessment.
Methods: Confirmatory factor analyses (CFA) was performed on the 20 non-management items of the SRS-22 from 428 Spanish, 229 Turkish, 95 French, 195 German and 245 American ASD patients. Item-loading invariance was compared across different languages, age groups, etiologies and times of assessment (baseline, follow-up). A separate sample of 772 American SRS-22 data from surgical ASD patients was used for cross-validation.
Results: The SRS-22 factor structure fitted significantly better to the proposed four-factor solution than to a unifactorial solution. However, three items Q14 (self-image; “back affects personal relationships”), Q15 (function; “financial difficulties because of back”), and Q17 (pain; “days off in last 3 months”) consistently showed weak item loadings within their factors across the language versions and in both baseline and follow-up datasets. Other items e.g. Q8 ("back pain at rest), Q11 ("medication use"), and Q10 ("trunk appearance") showed weak items loadings in some languages only. A trimmed SRS (SRS-16) that used the 4 least problematic items in each of the 4 domains yielded better-fitting models across the languages, but equivalence was still not reached; there was, however, equivalence of item-loading with respect to treatment (surgery vs nonop), time of assessment (baseline vs 12 mo FU), and etiology (degenerative vs idiopathic), though not age (</≥ 50y). All findings were confirmed in the cross-validation sample.
Conclusion: We recommend removal of the 4 less-well-fitting items of the SRS-22, together with adaptation and standardization of other items across language versions, to provide an improved 16-item version of the instrument. Since factorial invariance of an instrument under different conditions is essential to allow for comparisons across studies, this should provide for a more valid instrument in future multicentre outcome studies in ASD.