Reliability and Concurrent Validity of the SRS-22 Patient Questionnaire for Idiopathic Scoliosis

· (c - Depuy Acromed) University of Kansas Medical Center, Kansas City, KS, USA

The purpose of this study is to evaluate the reproducibility and internal consistency of the SRS-22 outcome survey.

METHODS: This study was carried out using mail-in questionnaires of the SRS-22 and the SF-36 outcomes instruments. These questionnaires were mailed to 83 patients who had been previously surgically treated for idiopathic scoliosis at this medical center. Fifty-eight of the eighty-three subjects responded. Fifty-one of the first-time respondents also returned their second survey. The 48 responders had an average age of 25 years (range 19 - 43) and 84% were female. The average response time between the first and the second mail-in assessments was 28 days. The time interval between the surgery and SRS-22 survey was 10.8 years. Individual domain scores were evaluated with descriptive statistics, internal consistency with the Cronbach a, test and re-test reproducibility with the intra-class correlation coefficient (ICC), and concurrent validity with Pearson Coefficient Correlation (r).

RESULTS: The average domain scores (5 being the best and 1 the worst), based on a sample of 58 responses were 4.2 for pain, 4.2 for self-image, 4.2 for function, 4.1 for mental health, and 4.5 for satisfaction. The calculated Cronbach a of internal consistency for the five corresponding domains was pain 0.92, self image 0.75, function 0.86, mental health 0.98, and satisfaction 0.88. Test/retest intra class correlations (n=51) were pain 0.96, self image 0.90, function 0.90, mental health 0.87, and satisfaction 0.85. The calculated Cronbach a and ICC for the SF-36 domains are quite compatible to those observed for the SRS-22. Concurrent validity was r 0.70 or more for 14 relevant SF-36 domains.

CONCLUSION: The SRS-22 has score distribution similar to the SF-36, is reliable (internally consistent and reproducible) and valid in comparison to SF-36.