SRS-22r legacy scores can be accurately translated to PROMIS scores in adult spinal deformity patients

BACKGROUND CONTEXT: Patient-Reported Outcomes Measurement Information System (PROMIS) facilitates comparisons of treatment effectiveness across populations and diseases. In adult spinal deformity (ASD), the disease-specific Scoliosis Research Society-22r (SRS-22r) tool assesses outcomes. Existing data must be translated to PROMIS to make comparisons. PURPOSE: To develop and validate a method to translate SRS-22r scores to PROMIS scores in surgical ASD patients. STUDY DESIGN: Retrospective cohort study. PATIENT SAMPLE: A total of 158 ASD surgery patients from an ongoing registry of patients who underwent spine surgery at a tertiary care center from 2015 to 2017 were included. OUTCOME MEASURES: PROMIS and SRS-22r questionnaires were completed at 387 visits (150 preoperative [derivation sample]; 237 postoperative [validation sample]). METHODS: Using the derivation sample, we modeled PROMIS domains as functions of age and SRS-22r domains using linear regression. The most parsimonious model was selected. In the validation cohort, we used the derived regression equations to estimate PROMIS scores from SRS-22r scores. RESULTS: The following significant associations were found (p<.001): PROMIS Pain Interference is dependent on age and SRS-22r Pain, Physical Function, and Patient Satisfaction; PROMIS Physical Function is dependent on age and SRS-22r Pain and Physical Function; PROMIS Anxiety is dependent on SRS-22r Mental Health; PROMIS Depression is dependent on age and SRS-22r Mental Health; and PROMIS Satisfaction with Social Roles is dependent on age and SRS-22r Pain, Physical Function (p=.011), Mental Health, and Patient Satisfaction. Correlations were strong to very strong between estimated and actual PROMIS scores in the validation cohort (p<.001): Pain Interference, r=0.78; Physical Function, r=0.66; Anxiety, r=0.83; Depression, r=0.80; and Satisfaction with Social Roles, r=0.71. CONCLUSIONS: PROMIS scores estimated from SRS-22r scores using our model correlate strongly with actual PROMIS scores. SRS-22r scores may be translated to PROMIS scores in all evaluated domains for ASD patients. Orthopedic surgeons can use this method to compare legacy measures with PROMIS scores.