Studies in the modified Scoliosis Research Society Outcomes Instrument in adults: validation, reliability, and discriminatory capacity.

STUDY
DESIGN: Observational study of patients with scoliosis and matched controls.
OBJECTIVES: To determine the validity and reliability of the modified Scoliosis Research Society Outcomes Instrument (SRS-22) for use in the assessment of deformity in adults. To demonstrate the discriminate validity of the SRS-22 in differentiating between affected and unaffected adults. SUMMARY OF BACKGROUND DATA: Spinal deformity has an important impact on the general health status of adults. The magnitude of this impact has been difficult to measure and reported variably in the literature. The development of disease-specific outcomes tools permits improved sensitivity and specificity in measuring the patient's self-assessment of health status. An instrument for measuring disease-specific health status in adults with scoliosis has not been validated.
MATERIALS AND METHODS: Observational study comparing the health status of adults affected by scoliosis and unaffected controls, matched for age, gender, and socioeconomic status. Pearson correlation analysis was used to determine the relationship of each domain within the SRS-22 and SF-36 with radiographic parameters including sagittal balance, coronal balance, and major curve correction. Discriminate validity of the modified SRS instrument was determined by a comparison of means between affected and unaffected cohorts. The validation of the SRS-22 was determined by criterion validity, using correlation analysis with comparable domains of the SF-36. The reliability of the SRS-22 was demonstrated using test-retest parity and Cronbach's alpha test for internal consistency.
RESULTS: One hundred eighty adults were included in the study, 146 with scoliosis and 34 without. Adults with scoliosis scored significantly lower than unaffected controls on every domain of the SRS-22 and the SF-36. The floor and ceiling effect of the SRS-22 were less than observed in the SF-36 instrument. Pearson correlation analysis demonstrated no significant correlation between any radiographic process measure and any specific domain within the SRS-22 or the SF-36 (r < 0.25). Comparison of similar domains in the SRS-22 and the SF-36 demonstrates high correlation between the instruments. Test-retest analysis similarly demonstrates a high degree of reproducibility in each domain (r = 0.83-0.94). Cronbach's alpha test of internal consistency within each domain demonstrates intercorrelation values greater than 0.75 within each domain of the SRS-22.
CONCLUSIONS: Adult scoliosis has a significant and measurable impact on affected patients compared with controls. There is a poor correlation between radiographic parameters of outcome and patient self-assessment of health status. The SRS-22 is a reliable instrument in adults as demonstrated by a high degree of internal consistency and reproducibility. The SRS-22 is a valid instrument for use in adult deformity as demonstrated by the criterion validity assessment with the SF-36. The study supports the use of the SRS-22 in the adult spinal deformity population.