Responsiveness of the Spinal Appearance Questionnaire (SAQ) Compared to the SRS Instrument in Adolescent Idiopathic Scoliosis

Background: Because appearance is a major issue in adolescent idiopathic scoliosis, the spinal appearance questionnaire (SAQ) was designed to assess various aspects of spinal deformity appearance. The SAQ has not previously been compared with the standard outcome tool for idiopathic scoliosis, the SRS instrument.

Methods: The SAQ and SRS instruments were both completed by 83 adolescent idiopathic patients preoperatively and 1 year postoperatively. Preoperative and postoperative scores were compared using paired t-tests for statistical significance. The effect size and standardized response mean were determined with values >0.8 considered large. The relative efficiency of the SAQ compared to the most responsive SRS domain was determined using the two sample Z-statistic.

Results: Both instruments had statistically significant improvement in all of their domains except of the SAQ Comparison and the SRS Activity scales. The effect sizes, standardized response means, Z-statistics, and relative efficiencies are shown in the table. The responsiveness of the SRS instrument was largely determined by the appearance domain (effect size 1.4). The relative efficiency of the SAQ domains to the SRS appearance domain was greater in 5, equivalent in 1, and less in 3 domains.

Conclusions: The appearance of the spine in adolescent idiopathic scoliosis shows the most responsiveness of any measured domain. The SAQ demonstrates markedly improved efficiency over the SRS instrument in this domain. We recommend use of the SAQ when spinal appearance is an important outcome for scoliosis.

Precis: The SRS and spinal appearance questionnaires (SAQ) were compared preoperatively and one year postoperatively in patients with idiopathic scoliosis. The appearance of the spine in adolescent idiopathic scoliosis shows the most responsiveness of any measured domain. The SAQ demonstrates markedly improved efficiency over the SRS instrument in this domain.

*Relative Efficienty is the square of the ratio of the specified SAQ subscale Z-Statistic and the SRS Appearance (the most response SRS scale) Z-Statistic.