Discrimination Validity of the Scoliosis Research Society-22 Patient Questionnaire

Abstract

Prospective observational study. To determine the discriminant validity of the Scoliosis Research Society-22 (SRS-22) health-related quality-of-life (HRQL) questionnaire for idiopathic scoliosis patients based on curve pattern and curve size. An important psychometric attribute of the HRQL questionnaire is the capacity to discriminate between and among patients with differing condition severity. Three study groups, with subjects ranging in age from 10 to 16 years, were set up: a control group (C) of 19 persons with scoliosis suspected but unproven with an average age of 13 years; a nonsurgical group (NS) of 68 patients with an average age of 14 years and an average scoliosis (Cobb) of 30 degrees; and a presurgical group (PS) of 32 patients with an average age of 14 years and an average scoliosis of 61 degrees. The NS group was subdivided into untreated (NSU) (n = 54) and braced (NSB) patients (n = 14). Also created were subgroups of the NS and PS groups having similar curve sizes: nonsurgical similar (NSS) and presurgery similar (PSS). Statistical significance of the SRS-22 domain scores by comparison groups was tested by ANOVA. Relationships between the SRS-22 questionnaire and scoliosis deformity measurements were calculated using the Pearson Correlation Coefficients. There were no significant differences between the control (C) and nonsurgical (NS) groups. The presurgical group (PS) scored significantly lower in the pain and self-image domains than the C or NS groups and lower than group C in the mental health domain. A subgroup of the nonsurgical group (NSS) with curves comparable to the presurgical subgroup (PSS) had significantly better self-image scores. There were no significant correlations between SRS-22 scores and curve pattern. For persons with scoliosis, curve size (Cobb) correlated with pain, self-image, and function; trunk asymmetry correlated with self-image. The SRS-22 HRQL questionnaire successfully discriminated among persons with no scoliosis or moderate scoliosis and large scoliosis. It did not discriminate among patients with single, double, or triple curves. Self-image and, to a lesser extent, pain and function domain scores correlated with radiographic and trunk asymmetry severity. The SRS-22 HRQL questionnaire may be useful in choosing nonsurgical versus surgical treatment in borderline cases.