Legacy Patient-reported Outcome Measures Can Be Reliably Translated to PROMIS Domains for Use in Adolescent Spinal Deformity.
Academic Article
Overview
abstract
STUDY DESIGN: Retrospective cohort study. OBJECTIVE: The aim of this study was to determine whether Patient-reported Outcomes Measurement Information System (PROMIS) Pediatric Domain scores can be accurately and reliably predicted from legacy patient-reported outcome measures (PROMs) for use in adolescent idiopathic scoliosis. SUMMARY OF BACKGROUND DATA: Previous efforts have been made to develop translational models between PROMIS domains and legacy PROMs in adult spinal deformity, but translational models for pediatric and adolescent patients with spinal deformity have not yet been developed. METHODS: Scores were retrospectively collected on pediatric patients 8-17 years old who were diagnosed with spinal asymmetry or scoliosis for the following PROMs: Scoliosis Research Society-22r (SRS-22r), Trunk Appearance Perception Scale (TAPS), Hospital for Special Surgery Functional Activity Brief Scale (HSS Pedi-FABS), and PROMIS Pediatric Domains: Pain Interference, Mobility, Physical Stress Experiences, Psychological Stress Experiences, and Physical Activity. Pearson correlations were calculated between legacy and PROMIS scores. In the derivation cohort, PROMIS domains were modeled from SRS-22r or HSS Pedi-FABS using linear regression. These equations were used to predict PROMIS scores in the test cohort to determine the reliability of the translation model. RESULTS: A total of 1171 patients (68% female) with a mean age of 13.4 ± 2.3 years were included in the final analyses. PROMIS Pain Interference was dependent on SRS-22r domains Pain, Function, Mental Health, and Management Satisfaction, PROMIS Mobility was dependent on SRS-22r Pain, Function, Mental Health, and Self-Image, PROMIS Physical Stress Experiences and Psychological Stress Experiences were dependent on SRS-22r Pain, Mental Health, and Management Satisfaction, and PROMIS Physical Activity was dependent on HSS Pedi-FABS. (P < 0.05 for all). In the test cohort, all developed models for PROMIS domains demonstrated strong correlations between estimated and actual scores (r = 0.63-0.75). CONCLUSION: All of the evaluated PROMIS domains (Pain Interference, Mobility, Physical Stress Experiences, Psychological Stress Experiences, Physical Activity) were reliably translated directly from these legacy scores for pediatric and adolescent patients with spinal deformity.Level of Evidence: 3.
